THE NEUROANATOMICAL AND ULTRASTRUCTURAL …

Acta Biologica Hungarica 63 (Suppl. 1), pp. 99–113 (2012)DOI: 10.1556/ABiol.63.2012.Suppl.1.10

0236-5383/$ 20.00 © 2012 Akadémiai Kiadó, Budapest

THE NEUROANATOMICAL AND ULTRASTRUCTURAL ORGANIZATION OF STATOCYST HAIR CELLS IN THE POND SNAIL, LYMNAEA STAGNALIS *

NATALIA L. KONONENKO, ** T. KISS and K. ELEKES ***

Department of Experimental Zoology, Balaton Limnological Center for Ecological Research Institute, Hungarian Academy of Sciences, P.O. Box 35, H-8237 Tihany, Hungary

(Received: November 15, 2011; accepted: December 12, 2011)

The ultrastructure, neuroanatomy and central projection patterns, including the intercellular connections of the statocyst hair cells of the pond snail, Lymnaea stagnalis, were studied, applying different intra- and extracellular cellular staining techniques combined with correlative light- and electron microscopy. Based on the ultrastructure different hair cells could be distinguished according to their vesicle and granule content, meanwhile the general organization of the sensory neurons was rather uniform, showing clearly separated perinuclear and “vesicular” cytoplasmic regions. Following intra- and extracellular labeling with fluorescence dyes or HRP a typical, local arborization of the hair cells was demonstrated in the cerebral ganglion neuropil, indicating a limited input-output system connected to the process of gravire-ception. Correlative light- and electron microscopy of HRP-labeled hair cells revealed both axo-somatic and axo-axonic output contacts of hair cell varicosities, and input on sensory axons located far from the terminal arborizations. Our findings suggest (i) a versatile ultrastructural background of hair cells corre-sponding possibly to processing different gravireceptive information, and (ii) the synaptic (or non-synap-tic) influence of gravireception at different anatomical (terminal, axonal and cell body) levels when processed centrally. The results may also serve as a functional morphological background for previously obtained physiological and behavioral observations.

Keywords: Statocyst – hair cells – neuroanatomy – ultrastructure – Lymnaea stagnalis – Mollusks, Gastropods

INTRODUCTION

Locomotion of mollusks, including gastropods, is influenced by several receptive fields and receptors, among others by a pair of a specific gravireceptive organ, called statocysts. Graviceptive input control locomotory reflexes and complex geotactic behaviors; changes in spatial orientation causes corrective motor responses resulting in the restoration of the proper orientation [16, 21]. Molluscan statocysts belong to the best studied models used for understanding the organization, function and evolu-

* Dedicated to Professor József Hámori on the occasion of his 80th birthday. ** Present address: Labor für Membranbiochemie und Molekulare Zellbiologie, Institut für Chemie

und Biochemie, Freie Universität Berlin, 14195 Berlin, Germany.*** Corresponding author; e-mail: [email protected]

100 NATALIA L. KONONENKO et al.

Acta Biologica Hungarica 63, 2012

tionary aspects of gravity-sensing organs. The statocysts of gastropods are relatively simple structures of small spherical chambers, attached to the pedal ganglia, and contain a number of sensory receptor cells supplied with ciliae facing the lumen. The lumen is filled with a fluid, in which one or more calcareous stones called statolith or statoconia are contained. Moving, changing position, spatial orientation make the statoliths move which then stimulate the ciliated receptor (hair) cells. These hair cells are primary sensory neurons which send their axons into the statocyst nerve, called the static nerve projecting to the cerebral ganglion. The geotactic behavior as well as the role of the hair cells in the control of locomotion [7, 14, 17], spatial orientation [8], and defense reactions [2] have been described in detail in different gastropod species such as Clione, Planorbis, Lymnaea and Hermissenda. Recent electrophysi-ological experiments indicated the connection of hair cells with central interneurons of Hermissenda [7], or with light sensitive neurons in Lymnaea [18].

Early and recent morphological and neuroanatomical studies on the statocysts of different gastropod species revealed its general (gross) morphology and cellular organization [5, 9, 12, 15, 20]. It was also shown that the statocysts, in general, con-sists of 13–14 hair (sensory) cells which project with their sensory axon to the cor-responding, ipsilateral, cerebral ganglion. Recent studies have also revealed contra-lateral projections as well [6]. Immunocytochemical investigations on the statocyst demonstrated that in the majority of the hair cell population histamine was present whereas other sensory neurons displayed FMRFamide, glutamate, nitric oxide or small cardioactive peptide B (SCPB) immunoreactivity ([3, 10, 15], Kononenko et al., unpublished observations).

Concerning the higher resolution of the functional morphology of the hair cells, including their ultrastructural organization and intercellular connections involved in different regulatory processes our knowledge is limited. Until now two early papers are known reporting on the fine structure of the hair cells in Aplysia [5] and different terrestrial, Limax and Arion, species [20], dealing with general features such as mem-brane attachments between hair cells and supporting cells, and the description of some of the cytoplasmic organelles. Here we provide a detailed analysis, aiming at the organization of the Lymnaea statocysts with special attention to the possible clas-sification of the hair cells, their neuroanatomical (projection) characteristics, and possible intercellular contacts in the cerebral ganglion. By this way we wanted to provide a functional morphological background for earlier and future physiological-behavioral studies.

MATERIALS AND METHODS

Animals

Adult specimens of Lymnaea stagnalis L. were collected in their natural surroundings and kept in aquaria under laboratory conditions until use.

Statocyst hair cell morphology in Lymnaea 101


Intracellular staining of sensory neurons

The circumpharyngeal ganglionic ring (CNS), including the paired pedal ganglia with the statocysts attached, were quickly dissected and pinned put in a Sylgard-coated Petri dish. To ease the intracellular penetration the connective tissue sheath covering the CNS was treated with Pronase (Protease, P 5147, Sigma) for about 1 minute.

Neurobiotin and Lucifer Yellow labeling

Following intracellular recording with glass microelectrodes filled with 0.1 M potas-sium chloride (tip resistance of 15–35 МΩ), 5% Neurobiotin or or 5% Lucifer Yellow dissolved in distilled water was injected iontophoretically by applying depolarizing current pulses of 3–5 μA amplitude and 100 ms duration at a repetition rate of 0.5–1 Hz for up to 1 hour.

Following both types of intracellular staining, preparations were kept for 2–4 hours at 7 °C in saline to allow dye diffusion, and then fixed in 4% paraformaldehyde diluted in 0.1 M phosphate buffer (PB, pH 7.4). After fixation, the Neurobiotin preparations were washed three times in Tris-HCl buffered (TrisBS) saline at pH 7.4, each for 15 minutes, followed by a pre-incubation with TrisBS containing 10% nor-mal goat serum (NGS, Sigma) for 5 hours and an incubation with Cy2-conjugated streptavidin (Jackson) diluted 1 : 500 in TrisBS containing 10% NGS for 12 hours. Incubations were performed at 4 °C. Both Neurobiotin, and Lucifer Yellow filled preparations were washed in PBS, dehydrated in graded ethanol, cleared in methyl-salicylate, mounted in Permount, and viewed in a Zeiss Axioplan compound micro-scope equipped with the appropriate filter set and a Canon PowerShot G5 digital camera.

Horseradish peroxidase (HRP) labeling

Electrode tips were filled with 5% HRP (Type VI-A, Sigma) diluted in 0.1 M Tris-HCl buffer containing 0.5 M potassium chloride. The enzyme was injected iontopho-retically by applying depolarizing current pulses of 3–5 μA amplitude and 200 ms duration at a repetition rate of 0.5–1 Hz for 20–30 min. The isolated CNS (circum-pharyngeal ring) was then kept in Lymnaea physiological saline for 5–7 hours at room temperature or overnight at 10 °C to facilitate axonal transport of the enzyme, fol-lowed by a fixation in a mixture of 2.5% glutaraldehyde and 1% paraformaldehyde diluted in PB (pH 7.4) at 4 °C overnight. After fixation the preparations were washed in 0.1 M PB, then rinsed three times in 0.1 M Tris-HCl buffer (pH 7.6), each for 15 minutes. Development of the histochemical reaction was performed as follows: (i) pre-incubation in 0.05% 3,3’-diamino-benzidine (DAB) diluted in 0.1 M Tris-HCl buffer (pH 7.6) for 30 minutes, (ii) development for 15 minutes in 0.05% DAB con-



taining 0.01% H2O2. The preparations were briefly rinsed in 0.1 M Tris-HCl buffer and in 0.1 M PB, dehydrated in graded ethanol series, cleared in methylsalicylate and mounted in Permount for light microscopy, or embedded in Araldite for correlative light- and electron microscopy (see below).

Extracellular labeling with a mixture of HRP and Streptolysin O

Streptolysin O (SLO) is an immunogenic, oxygen-labile toxin, which is reversibly activated by dithiothreitol and can be used for cell permeabilization or hemolysis [1, 13].

One mg of Steptolysin O (lyophilized powder, Sigma) was diluted in 2 ml of dis-tilled water containing 100 mM dithiothreitol (Sigma) to a final concentration of 0.5 mg/ml. 10% HRP (Type VI-A, Sigma) was prepared in 0.1 ml 0.1 M Tris-HCl buffer containing 0.5 M KCl. Borosilicate glass capillaries (type 1B150F, World Precision Instruments) were filled with a mixture of 0.1 ml 10% HRP and 0.1 ml 0.5 mg/ml SLO, resulting in a final concentration of 5% HRP and 0.25 mg/ml SLO. Before use, the solution was filtered through a 0.2 Millipore filter or centrifuged.

For the labeling of a hair cell population, a mixture of HRP and SLO was injected into the statocyst cavity through an electrode, the tip of which was broken or beveled (resistance 40–50 M) to improve the penetration. The pressure injection system con-sisted of a gas supply connected via a timing circuit and a pressure regulator, to a pipette holder. The pressure (40–50 psi) and timing of the pulse (100–500 ms) was roughly established by measuring the diameter of a drop expelled from the pipette tip into liquid paraffin. After the statocysts were filled with the mixture of HRP and SLO, the CNS preparations were kept in Lymnaea physiological saline for 5–7 hours at room temperature to facilitate the penetration of the enzyme and the axonal transport. The HRP histochemical reaction was developed as above either in whole-mounts or in 50 μm thick Vibratome slices. The latter were processed for correlated light- and electron microscopy as described below.

Electron microscopy

Circumpharyngeal ganglion rings (CNS) were dissected, pinned out in a Sylgard-coated Petri dish, and fixed in a mixture of 2.5% glutaraldehyde and 1% paraformal-dehyde diluted in 0.1 M PB (pH 7.4) for 4 hours or overnight at 4 °C. Following washing in 0.1 M PB, post-fixation was performed in 1% OsO4 diluted in 0.1 Na-cacodylate for 1 hour at room temperature. The preparations were then dehy-drated in graded ethanol and propylene oxide, infiltrated by 1 : 3, 1 : 1 and 3 : 1 mix-tures of propylene oxide and Araldite, and finally embedded in Araldite (Durcupan ACM, Fluka). In the course of dehydration, block staining was performed in 70% ethanol saturated with uranyl acetate. Vibratome slices containing HRP-filled hair cells were first flat embedded on slides, viewed in a Zeiss Axioplan compound micro-



scope light microscope attached to a Canon PowerShot G5 digital camera or traced in an Olympus B2H light microscope combined with a camera lucida drawing tube, then re-embedded for ultrathin sectioning. Following orientation in 1 μm semi-thin sections stained with 1% toluidine-blue, serial 50–60 nm ultrathin sections were taken, placed onto mono-hole copper grids, stained with lead citrate, and viewed in a JEOL 1200EX electron microscope.

RESULTS

Cellular organization of the Lymnaea statocysts

Morphology of the hair cells

The statocysts of Lymnaea stagnalis are paired spherical organs, both with a diameter of about 120 μm, attached to the dorso-lateral side of the pedal ganglion of the CNS and surrounded by a thin (approx. 5 μm) connective tissue sheath (Fig. 1A–D). Each statocyst contained 13–14 sensory hair cells, forming actually the outer “wall” of the sensory organ and facing with their ciliated inner surface towards the cavity (Fig. 1C, D). In the statocyst cavity statoconia occurred, which sometimes occupied about one-third of the total volume of the cavity (Fig. 1D).

The form and general morphology of the hair cells of the statocyst were studied in 1 μm Araldite semi-thin sections. Sensory hair cells were found to be strongly flat-tened, measuring about 8–15 μm in diameter along their shorter and 20–40 μm along their longer axis (Fig. 1D). Much smaller (1–2.5 μm wide) supporting cells were located interposed between the receptor cells. The static nerve left the statocyst at the medial-lateral side (Fig. 1C). According to the staining intensity of the cytoplasm, the hair cells differed; even within the individual sensory neurons different (light and dark) cytoplasmic regions could be distinguished (Fig. 1C, D), which actually cor-responded to the electron dense perinuclear and the electron lucent granule-containing cytoplasmic regions resolved at ultrastructural level, respectively (Fig. 2A–D).

Ultrastructural organization of the hair cells

Statocyst hair cells displayed, in general, a similar ultrastructural appearance. Two different cytoplasmic regions could be observed in them, clearly segregated from each other. A moderately electron transparent cytoplasm located peripherally, adja-cent with an electron dense perinuclear cytoplasmic region. The hair cells were covered by a multilayer sheath of glial cells and basal membrane facing the hemo-coel, meanwhile they had a free membrane surface bearing sensory ciliae towards the statocyst cavity (Fig. 2A, B). Higher magnification revealed that the electron transparent cytoplasmic parts were filled almost exclusively with masses of vesicles/granules of different types (Fig. 2D–F), meanwhile the perinuclear region contained



Fig. 1. Light microscopy of the statocyst in 1 μm semi-thin Araldite sections. Toluidine-blue staining, horizontal section plane. A: Low power view of a statocyst (arrow) attached to the pedal ganglion (PG). B: Higher magnification of a right statocyst (st). Note its identical size (approx. 120 μm along the longer axis) with the nearby located giant dopaminergic RPeD1 neuron (asterisk). C: Hair cells (asterisks) dis-playing different toluidine-blue staining intensity, which indicates the different cytoplasmic organization. Arrow – exit of the statocyst nerve. PG – pedal ganglion, NC – nerve cell. Insert: The statocyst nerve seen in a consecutive section. D: High magnification view of a statocyst, showing “dark” (arrows) and “light” (double arrow) hair cells, and numerous statoliths (arrowheads) located in the cavity. PG – pedal ganglion. A–D – left is rostral, right is caudal. cv – statocyst cavity, ct – connective tissue sheath. Scale

bars: A 120 μm, B – 50 μm, C – 10 μm; insert: 8 μm, D 5 μm



Fig. 2. Ultrastructure of the sensory hair cells. A, B: Low magnification views of hair cells showing the electron lucent vesicular cytoplasmic region (asterisks, A), or the clear segregation of the electron dense perinuclear (pn) and the vesicular (asterisks) cytoplasmic regions (B). Note cross-sectioned ciliae (arrow-heads) and a large statolith (arrow). C: Higher magnification detail from the perinuclear region. Note large number of mitochondria (m) and the granules of low electron density (arrowheads) nearby in the vesicular cytoplasm (asterisks). D–F: Details of the vesicular cytoplasm, containing either large number of eccentric dense-core vesicles (arrowheads, D), or 120–150 nm granules of different electron density (arrowheads in E), or a population of 80–100 nm granular vesicles (arrowheads in F). cv – statocyst cav-

ity. Scale bars: A–C 1 μm D–F: 0.5 μm



mitochondria, elements of the ribosomic endoplasmic reticulum system and Golgi-units in a dense arrangement (Fig. 2C), among which vesicular/granular elements also occurred. The vesicle/granule types observed could be classified in three groups as follows: (i) 80–120 nm dense-core or eccentric dense-core vesicles, (ii) 80–120 nm granules with finely granulated content of low or medium electron density, (iii) large, 140–160 nm vesicles of low electron density (Fig. 2C–F). According to the presence of the different vesicle and/or granule types, the following hair cell types could be distinguished: type 1 containing only large dense-core vesicles (i) (Fig. 2D, F); type 2 showing a mixed population eccentric dense-core vesicles and granules with finely granulated internal content; type 3 with dense-core vesicles and granules with finely granulated internal content; type 4 containing mostly large granules of low electron density and dense-core vesicles (Fig. 2E). Most frequently type 2 hair cells could be observed.

Neuroanatomy of the sensory hair cells

The neuroanatomical characteristics and projection patterns of the statocyst hair cells were studied both by intracellular labeling or extracellular staining through the stato-cyst cavity, applying different fluorescence dyes or HRP (Figs 3, 4).

Intracellular staining with Neurobiotin or Lucifer Yellow

Intracellular staining with Neurobiotin or Lucifer Yellow provided a detailed view of the neuroanatomy and axon projections of the statocyst sensory hair cell axons, including the static nerve which originated from the antero-dorsal part of the statocyst (Fig. 3). The static nerve passed over the pedal-pleural connective and ran parallel with the pedal-cerebral connective before entering the ipsilateral cerebral ganglion. In case of successful injection(s), both fluorescence dyes (and also HRP) clearly delineated the path of the sensory axons within the corresponding cerebral ganglion, including their terminal arborization in the neuropil (Fig. 3A–E). Along the statocyst surface all sensory cell perikarya showed an irregular, multi-angled form. The com-parison of whole-mount Neurobiotin and Lucifer Yellow preparations (and also the HRP-stained hair cells) made possible to determine the three dimensional size of the sensory cell, that was the following: 40–60 μm length (= diameter along the longer axis), 30–40 μm width (= diameter along the shorter axis) and 10–15 μm height (= ”thickness” of the cells).

Hair cells injected intracellularly projected usually with a thin sensory axon along the outer surface of the statocyst (Fig. 3), and they displayed a terminal arborization varying as to the branching pattern and the neuropil area occupied (Fig. 3C–E). The terminal arborizations had either knob-like (Fig. 3C) or fine varicose (Fig. 3D, E) endings.



Fig. 3. Anatomy of the statocyst hair cells visualized following intracellular fluorescence labeling. Whole-mount preparations. A, B: Intracellular labeling by Neurobiotin/Streptavidine. A: Low power view of the right statocyst (arrow), after injecting several hair cells. The labeled bundle of sensory axons (arrow-heads) project in the static nerve to the cerebral ganglion (rCG). rPG – right pedal ganglion, rPlG – right pleural ganglion, cpc – cerebro-pedal connective. B: Higher magnification view of a single hair cell injected with Neurobiotin/Streptavidine. Its sensory axon delineates the outer surface of the statocyst. C–E: Different patterns of arborization (arrowheads) of sensory axon processes (arrows) in the neuropil of the cerebral ganglion, following Lucifer Yellow staining of individual hair cells. A 200 μm, B 40 μm,

C–E 10 μm





Intra- and extracellular staining with HRP

Hair cells were also successfully labeled by both the new method applying extracel-lularly a mixture of HRP and Streptolysin O (Fig. 4), or intracellular HRP injection (Fig. 5). The first method allowing a better penetration of HRP proved to be a fast and efficient way for the simultaneous labeling of up to 4–5 sensory neurons (Fig. 4A inset, C). Light microscopic examination and subsequent camera lucida tracing of HRP-streptolysin O labeled statocyst preparations revealed the pathway of the axon projections in the ipsilateral static nerve and their arborization in the neuropil of the cerebral ganglia (Fig. 4A–D). The sensory axon arborization pattern occupied a small, well-delineated area in the cerebral ganglion neuropil, showing local arrange-ment of the branching fibers (Fig. 4B, D). Following intracellular HRP injection, labeled hair cells could also be followed from the very origin of the thin sensory axon up to processes in the cerebral ganglion (Fig. 5A–C).

Correlative light- and electron microscopy of HRP-labeled elements

Single intracellular injection of HRP resulted in finer, evenly distributed reaction product (Fig. 5D, E), in contrast to the rather roughly labeled axon profiles found after the application of extracellular labeling with the mixture of HRP-Streptolysin O (Fig. 4E–G). This latter often masked completely the ultrastructure of the labeled ele-ments (Fig. 4E–G), meanwhile after single intracellular labeling mitochondria, gran-ule/vesicle contents and membrane contacts could clearly be distinguished (Fig. 5E). On the other hand, following both labeling techniques applied, HRP-labeled struc-tures were found forming contacts with unlabeled pre- (Figs 5G, H) or postsynaptic (Fig. 5D, E) profiles in the cerebral ganglia. The labeled hair cell axons ended both on axons and nerve cell bodies (Figs 4F, G, 5E).

←⎯⎯⎯⎯⎯Fig. 4. Correlative light- and electron microscopy of hair cells following extracellular staining with a mixture of HRP and Streptolysin O. A, B: View of the left statocyst (double arrow) containing a number of labeled hair cells (insert, arrowheads), which project (arrows) to (A) and arborize in (B) the left cere-bral ganglion (lCG). Subsequent 50 μm Vibratome slices, embedded in Araldite. C, D: Camera lucida reconstruction from the preparations seen in A, B. C: The statocyst (Ci, arrow), the sensory axons (Ci, arrowheads) of the labeled hair cells (Cii). Small clear symbols in Cii indicate statoliths. D: Local arborization (arrows) of the sensory axons in the neuropil of the lCG, traced in two different focus planes (Di, Dii) in the region shown in B. lPeG right pedial ganglion, lPlG – left pleural ganglion cpdc – cerebro-pedal connective. E–G: Identification of HRP-labeled axon processes at ultrastructural level. E: Axon processes (A) traversing in the neuropil of the cerebral ganglion. F, G: Higher magnification of labeled varicosities (T), contacting (arrows) a granule-containing perikaryon (P) in F, and an axon profile (A) in

G. Scale bars: A, B 200 μm, C 400 μm, D 40 μm, E 1 μm, F μm, G 1 μm



Fig. 5. Correlative light- and electron microscopy of a hair cell, following intracellular HRP-labeling. A: The hair cell (arrow) and its axon process (arrowhead) seen in a 50 μm Vibratome section embedded in Araldite. B, C: Consecutive 1 μm semi-thin sections taken from the Vibratome section shown in A. B: The labeled hair cell body (arrow) and a short segment of its axon process (arrowhead) entering the cerebral ganglion. C: A short segment of the sensory axon (arrowhead) within the neuropil of the cerebral ganglion. D: Electron microscopic depiction of the HRP-labeled axon process (A) shown in C. Note the similar morphological appearance. A1, A2 – unlabeled axon profiles. E: Enlarged view of an axo-axonic contact (arrows) rectangled in D, established between an unlabeled (A2) and a labeled thick sensory (A) axon. Note the long segment of the closely apposing axon membranes (arrows). Arrowheads – granular

vesicles; m – mitochondria. Scale bars: A 100 μm, B, C 50 μm, D 1 μm, E 0.5 μm



DISCUSSION

In our present study we have provided a detailed neuroanatomical and ultrastructural characterization of statocyst hair cells of a gastropod mollusk the pond snail, Lymnaea stagnalis, including individual tracing and visualization of intercellular contacts in the cerebral ganglion. In this respect the results are to be considered as novel in, combining intracellular tracing and ultrastructural observations as well as analyzing the fine structure of the different gravireceptive sensory cells, revealing characteris-tics not known before. They provide a background for earlier and recent physiological investigations that were conducted on the gravireceptive system of different gastro-pod model species such as Planorbis, Lymnaea and Clione [2, 11, 12, 17]. Our results are especially new in high resolution neuroanatomy describing the ultrastructure and the intercellular connections of hair cells. Apart from two early reports on the stato-cyst ultrastructure [5, 20] data are missing on the intracellular organization of the hair cells concerning their possible difference in appearance, organelle and vesicle/gran-ule content.

Although the hair cells of the Lymnaea statocyst differ greatly in size (8–30 μm), their general ultrastructural organization appears to be uniform. The cytoplasm is segregated for two clearly different regions: mostly peripherally localized vesicular region and a perinuclear region tightly filled with cellular organelles. This unusual arrangement of the cytoplasm may indicate a kind of functional differentiation at the hair cell body level. The vesicular/granular cytoplasm containing almost exclusively a mass of granular vesicles/granules of different morphology/electron density may correspond to a storage depot of the transmitter/modulator content synthesized previ-ously in the perinuclear region, and which could be mobilized upon the functional request of the nerve cell periphery at the synaptic level.

On the basis of the vesicle/granule content different hair calls can be distinguished in the Lymnaea statocyst. Accordingly, in most of the hair cells the cytoplasm con-tained large population of eccentric dense-core vesicles, meanwhile large size gran-ules of medium dense or low electron density, or large dense-core vesicles occurred in a smaller part of the sensory neurons. This kind of ultrastructural versatility may correspond to the different transmitter content of the hair cells. Indeed following the application of immunohistochemistry, 7 histamine, 3 glutamate, 3 NO and 1 FMRFa containing hair cells were found in the statocyst of Lymnaea ([3, 10], Kononenko et al., unpublished observations). The most frequently occurring cell type that contained the eccentric dense-core vesicles is assumed correspond to the histamine-containing hair cell.

The high resolution of the arborization of the hair cells following the application both of the fluorescence and HRP tracers revealed important details of functional significance. The terminal arborization of the sensory axons in the cerebral ganglion is rather limited. This kind of local arborization suggests that the hair cells are involved in a relative low number of interactions be synaptic or modulatory. The knob-like endings visualized by the fluorescence tracers resemble those described for



the visual sensory cells and the monopolar neurons of arthropods [19], whereas the typical local arborizations resolved after HRP-labeling resemble those described for local interneurons in the locust thoracic ganglia [4]. The comparison of fluorescence tracers and HRP-techniques in view of the visualization of terminal arbors following intracellular or extracellular (multicellular) labeling shows that intracellular labeling provides a more precise resolution of the input-output area of a hair cell. The multi-cellular labeling delineates the entire synaptic area of the statocyst in the neuropil of a cerebral ganglion. Similar results were obtained by Sakakibara et al. [18] who showed that in case of combining the results obtained with several individually filled hair cells resulted in a broader arborization area in the neuropil. Whether each of the individually restricted arborization corresponds to different output-input connection systems is to be clarified. We were also able to detect input-output contacts on thick sensory axon processes and hair cell bodies. Physiological observations reported both the activation of Lymnaea statocyst hair cells [18] by light and the activation of inter- and motorneurons of Hermissenda by input from the statocyst hair cells [7]. These findings seem to correspond to our electron microscopic observations obtained after HRP-labeling, showing that hair cells possess input and form output contacts. The occurrence of hair cell input on unidentified thick axons and cell bodies, however, suggests that the synaptic area of the hair cells in the cerebral ganglion might range beyond that suggested by the camera lucida tracing. On the other hand, vesicle/gran-ule containing axon profiles contacting thick sensory axon processes suggests that the gravireceptive information might be influenced not only at the level of axon arboriza-tion and their boutons but earlier en route to the terminals. Consequently, it involves the possible role of en passant modulatory events which may shape sensory informa-tion before being transmitted to interneurons in the cerebral ganglia.

ACKNOWLEDGEMENTS

A grant from IBRO-CEERC awarded to Dr. Natalia N. Kononenko to work in Tihany is greatly acknowl-edged. The skilled technical contribution of Ms. Zsuzsanna N. Fekete, Ms. Zita László and Mr. Boldizsár Balázs is highly appreciated. This work was supported by an OTKA grant, No. 49090.

REFERENCES

1. Alouf, J. E. (1999) Streptococcal toxins (Streptolysis O, Streptolysin S, erythrogenic toxin). Pharmacol. Ther. 11, 179–186.

2. Arshavsky, Y. I., Deliagina, T. G., Okhstein, G. N., Orlovsky, G. N., Panchin, Y. V., Popova, L. B. (1995) Defense reaction in the pond snail Planorbis corneus. III. Response to input from statocyst. J. Neurophysiol. 71, 898–903.

3. Braubach, O. R., Croll, R. P. (2004) Evidence that histamine acts as a neurotransmitter in statocyst hair cells in the snail, Lymnaea stagnalis. J. Grav. Physiol. 113, 57–66.

4. Burrows, M. (1986) The neurobiology of an insect brain. Oxford University Press, Oxford, New York, Tokyo.

5. Coggeshall, R. E. (1969) A fine structural analysis of the statocyst in Aplysia californica. J. Morphol. 127, 113–132.



6. Chase, R. (2002) Behavior and its neural control in gastropod mollusks. Oxford Univ Press, New York.

7. Crow, T., Tian, L.-M. (2004) Statocyst hair cell activation of identified interneurons and foot contrac-tion motor neurons in Hermissenda. J. Neurophysiol. 92, 2874–2883.

8. Deliagina, T. G., Arshavsky, Y. I., Orlovsky, G. N. (1998) Control of spatial orientation in a mollusk. Nature 393, 172–175.

9. Geuze, J. J. (1968) Observations on the function and the structure of the statocyst of Lymnaea stag-nalis (L.). Neth. J. Zool. 18, 155–204.

10. Hegedűs, E., Kaslin, J., Hiripi, L., Kiss, T., Panula, P., Elekes, K. (2004) Histaminergic neurons in the central and peripheral nervous system of gastropods (Helix, Lymnaea): An immunocytochemical, biochemical and electrophysiological study. J. Comp. Neurol. 475, 391–405.

11. Janse, C. (1982) Sensory system involved in the gravity orientation in pulmonate snail Lymnaea stganalis. J. Comp. Physiol. 145, 311–319.

12. Janse, C, van der Wilt, G. J., van der Roest, M., Pieneman, A. (1988) Intracellularly recorded respons-es to tilt and efferent input of statocyst sensory cells in the pulmonate snail Lymnaea stagnalis. Comp. Biochem. Physiol. 90, 269–278.

13. Johnson, L. R., Moss, S. B., Gerton, G. L. (1999) Maintenance of motility in mouse sperm permea-bilized by Streptolysin O. Biol. Reprod. 60, 683–690.

14. Levi, R., Varona, P., Arshavsky, Y. I., Rabinovich, M. I., Selverston, A. I. (2004) Dual sensory-motor function of a molluskan statocyst network. J. Neurophysiol. 91, 336–345.

15. Ohsuga, K., Kurokawa, M., Kuwasawa, K. (2000) Mosaic arrangement of SCPB-, FMRFamide-, and histamine-like immunoreactive hair cells in the statocyst of the gastropod mollusk Pleurobranchea japonica. Cell Tissue Res. 300, 165–172.

16. Orlovsky, G. N. (1991) Gravistatic postural control in simpler nervous systems. Curr. Opin. Neurobiol. 1, 612–627.

17. Panchin, Y. V., Arshavsky, Y. I., Deliagina, T. G., Popova, L. B., Orlovsky, G. N. (1995) Control of locomotion in marine mollusk Clione limacina. IX. Neuronal mechanisms of spatial orientation. J. Neurophysiol. 73, 1924–1937.

18. Sakakibara, M., Aritaka, T., IIzuka, A., Suzuki, H., Horikoshi, T., Lukowiak, K. (2004) Electro-physiological response to light of neurons in the eye and statocyst of Lymnaea stagnalis. J. Neuro-physiol. 93, 493–507.

19. Strausfeld, N. J. (1980) Neuroarchitecture of brain regions that subserve the compound eye of crus-tacea and insects. In: Autrum, H. (ed.) Handbook of sensory physiology. Vol. VII/6B. Comparative physiology and evolution of vision in invertebrates. Springer Verlag, Berlin, Heidelberg, pp. 1–132.

20. Wolff, H. G. (1969) Einige Erkenntnisse zur Ultrastruktur der Statocysten von Limax maximus, Limax flavus und Arion empiricorum (Pulmonata). Z. Zellforsch. 100, 251–270.

21. Wolff, H. G. (1975) Statocyst and geotactic behaviour in gastropod molluscs. Fortsch. Zool. 23, 63–84.

Documents

THE NEUROANATOMICAL AND ULTRASTRUCTURAL …