Neuroanatomy of the rhinophore
The anatomy of the rhinophore of Aplysia
was previously investigated only with respect to the location of sensory cells [13
] and as part of a phylogenetic study of seahares [34
]. The presence of neuromodulators such as catecholamines has been demonstrated in the rhinophore of Aplysia californica
by Croll [35
]. This is the first study to focus on the functional neuroanatomy of the rhinophore of Aplysia punctata
with respect to number and location of glomeruli, serotonergic innervation and neuronal pathways. The rhinophore of Aplysia punctata
contains a groove, which extends to about two third of the total length of the rhinophore. Figures and show an example of a longitudinally sectioned rhinophore labelled with phalloidin and serotonin-immunoreactivity (IR). The rhinophore usually contracted during the process of dissection, and prominent longitudinal muscles became strongly labelled with phalloidin binding to muscular f-actin (Figs ). Serotonin-IR was detectable in various regions of the rhinophore: the rhinophore nerve, the rhinophore ganglion at the basis of the groove, and the glomeruli (Figs ). Serotonergic fibres proceeded from the rhinophore nerve via the rhinophore ganglion to the glomeruli. Since no serotonin-IR was detectable in cell bodies within the rhinophore, all serotonergic neurons innervating the rhinophore should be of extrinsic origin. Croll [35
] described tyrosine hydroxylase immunoreactivity in both large and small somata beneath the epithelium in the rhinophore, homogeneously distributed over walls of the entire structure, whereas serotonergic immnureactivity was found in cellular processes but not the somata inside the rhinophore. Croll et al. [26
] found no serotonergic cell bodies in the periphery in the nudibranch Phestilla
, similar to our findings. This indicates a physiological role of catecholamines and serotonin in olfactory processing mediated by centrifugal neurons in the case of serotonin and more local modulation in the case of catecholamines. Serotonergic innervation of olfactory glomeruli is commonly found in insects and was shown to enhance the response of olfactory projection neurons (e.g.: [36
Figure 1 Anatomical overview of the rhinophore. Neuroanatomy of the rhinophore: confocal images of sections labelled with an antibody to serotonin, fluophore-conjugated phalloidin and propidium iodide. A, B: Sagittal sections of the rhinophore at different planes. (more ...)
A series of cross sections demonstrates phalloidin-labelled muscle-fibres bundles oriented in the longitudinal and horizontal axis of the rhinophore (Figs ). Glomeruli were situated beneath the sensory epithelium close to the inner wall of the entire groove (Figs , ). Staining of cell nuclei with propidium iodide indicates that the glomeruli are not surrounded by a regular border of periglomerular cell somata (Figs , ). In histological sections, however, glomeruli appeared to be surrounded by a layer of glia-like processes (Fig. ). In contrast to insects and vertebrates, where olfactory glomeruli became brightly labelled with phalloidin due to aggregation of neuronal f-actin [38
], glomeruli within the rhinophore of Aplysia punctata
were not labelled with phalloidin, and the rhinophore ganglion was only lightly stained with phalloidin (Fig. ). Histological sections revealed that the rhinophore ganglion has a folded structure (Fig. ). Retrograde labelling of the glomeruli by insertion of 4-(4-(dihexadecylamino)styryl)-N
-methylpyridinium iodide (DiA) in the rhinophore ganglion demonstrated the connection between the rhinophore ganglion and the glomeruli (Fig. ). DiA stainings show the neuronal connection between the rhinophore ganglion and glomeruli. Retrograde labelling with DiA revealed mostly neuropil staining indicating that neuronal processes extending from the rhinophore ganglion to the glomeruli were preferentially labelled. We do not know if the cell bodies of these neurons lie in the rhinophore ganglion (because of massive staining) or further out in the periphery, or on adjacent sections. Future studies will focus in more detail on individual projections of sensory neurons using patch clamp/single neuron recordings and dye filling techniques.
Figure 2 Anatomical details of the rhinophore. Neuroanatomy and histology of the rhinophore. A: Cross sections labelled with serotonin-immunoreactivity (green) and propidium iodide (red). Nuclear labelling shows the layers of cell nuclei in the sensory epithelium (more ...)
The number of glomeruli was estimated in a complete series of sections of the rhinophore labelled with 5-HT antibody. The total number of glomeruli counted in one rhinophore was 36 in one rhinophore, and the mean diameter of individual glomeruli averaged 49 μm +/- 27 μm with a range from 25 μm to 135 μm.
Previous investigations on Aplysia californica
have shown that the groove houses various types of sensory cells [13
]. Emery and Audesirk [13
] suggested that intraepithelial cells with 30 μM long cilia produce a steady water flow around the epithelial cells of the rhinophore to facilitate olfaction. Anterograde labelling experiments in a terrestrial snail (Achatina fulica
) by Chase and Tollozcko [21
] have shown that sensory neurons project to the glomeruli and directly to the rhinophore ganglion or to further centres in the cerebral ganglion. Future experiments are needed to find out if similar projection patterns of sensory neurons are present in Aplysia
Labelling of cell nuclei with propidium iodide and histological sections showed the high amount of cell nuclei in the sensory epithelium (Figs , ). The layer of sensory cell nuclei is clearly separated from the more basal nuclei (Fig ). Glomeruli were located underneath this epithelium (Figs ).
Propidium Iodide and Mallory's stain revealed cell nuclei of variable sizes surrounding the glomeruli. Both methods revealed very few nuclei inside the glomeruli similar to the conditions in vertebrates and in insects (Figs. ). Propidium Iodide has very high affinity to nucleic acids, and since we used agarose sections with excellent penetration properties we are convinced that all nuclei were labelled. In contrast to the glomeruli, cell nuclei were present inside the rhinophore ganglion (Fig. ). Cell nuclei can be from both neuronal and glial cell bodies.
The anatomy of the rhinophores with distributed glomeruli around the groove leaves different possibilities open for central projections of olfactory receptor neurons (ORNs). The general view of a uniglomerular projection pattern, which seems to occur in most insects [39
] and vertebrates [41
], would be complicated with the present arrangement of glomeruli. Possibly, ORNs of the same type are situated beneath one glomerulus and are not uniformly distributed over the sensory epithelium, or, in the other case, axonal projections of individual ORNs to their target glomeruli would be very long and across the rhinophore. Recent studies in Xenopus laevis
described an innervation of more than one glomerulus by individual ORNs [43
]. Multiglomerular projection patterns also appear in crustaceans [44
]. In Aplysia
, future studies on the projection pattern of ORNs combined with functional imaging studies of the glomeruli will provide further insight in the functional role of glomeruli in molluscs.
Calcium imaging of odour evoked responses within the rhinophore ganglion
To investigate whether the rhinophore ganglion responds to olfactory stimulation of the sensory epithelium of the rhinophore, we applied different amino acids as odorants and recorded the responses optically. First of all we tested our system by the application of artificial sea water (ASW) with high K+ (Fig. ). The Ca2+-response induced by a high K+-solution could be reproduced and showed nearly a similar increase after repetitive application.
Figure 3 Fluorimetric measurements. Measurement of changes of Ca2+-levels in separated ROIs in response to high K+ and amino acids in the rhinophore ganglion (stimulus duration was one minute in each case). A: Application of artificial sea water with high K+. (more ...)
We chose different amino acids, as they were used in previous experiments in gastropod molluscs [32
]. To avoid direct excitation of neurons within the rhinophore by the stimuli, we did not use glutamic and aspartic acid, which induced highest responses in the nudibranch Phestilla sibogae
]. In more than 10 experiments we found no Ca2+
– response within the rhinophore ganglion induced by the application of methionin at different concentration (200 μm – 20 mM).
Clear Ca2+-responses within the rhinophore ganglion were found during stimulation with alanine (ALA). Figure shows three responses to ALA from three different animals. In all experiments 2 mM ALA was applied. We found a decrease of Ca2+-levels as well as an increase (Fig upper trace). Application of different amino acids (2 mM each) during one experiment revealed differential responses of intracellular Ca2+-levels in distinct regions (Fig ). The traces in Fig show Ca2+-measurements in two different regions of interest. Arginine (ARG) and glutamine (GLN) induced a response in one region, whereas the other region showed no response, but both regions responded with a change in Ca2+-level to the stimulation with glycine (GLY) and valine (VAL). Both regions showed a decrease in Ca2+-levels to all stimuli (Fig. ) indicating an inhibition or reduction of cellular activity by these stimuli. Stimulation with 2 mM phenylalanine (PHE) induced no detectable responses.
The responses to different concentrations of ALA (2 μM – 20 mM) were highly dependent on the recording position within the ganglion. Ten responses from ten regions within the rhinophore ganglion are demonstrated in figure . Regions of interest (ROIs) were selected more or less randomly across the rhinophore ganglion at the beginning of an experiment, and only responsive regions are shown in the figures. The size of individual ROIs can be estimated from the scale bar (Fig. ). We assume that in many cases ROIs may have not included a single neuron since the diameter ROIs were between 15 and 60 μM and only the largest cell bodies reach 50 μM.
Figure 4 Ca2+-Responses of ROIs to different concentrations of alanine. A: Schematic overview of the rhinophore of Aplysia punctata: Measurements were performed within the rhinophore ganglion (box). Glomeruli (GL) with putative projections to the ganglion. Rhinophore (more ...)
In the course of a single experiment, odour induced responses to amino acids recorded from different regions remained similar. We applied ALA in concentration from 2 μm to 20 mM. At low concentration (<2 mM) it was difficult to decide whether an observed Ca2+-response was stimulus dependent. At a concentration of 2 mM ALA, clear responses could be observed in three regions (Fig VIII-X). Regions VIII and IX showed a decrease of the Ca2+-levels, whereas in region X the Ca2+-level increased.
In response to 20 mM ALA all of the selected regions showed a change in intracellular Ca2+-levels. It was either a decrease or an increase of the Ca2+-levels, or even a combination of both as shown in trace II. Remarkably, responses from neighbouring regions (II and III; IX and X) showed opposing Ca2+-responses indicating a possible inhibitory coupling. The spatial response patterns were complex. Stimulation with 20 mM ALA induced an increase in Ca2+-levels in four regions (Fig. -I,-V,-VI,-X), whereas in five other regions we found a decrease in Ca2+-levels (Fig. -III,-IV,-VII,-VIII,-IX), and in region II a decrease was followed by an increase in Ca2+-levels.
In all of our experiments, the highest Ca2+
-responses were induced by ALA. Since the rhinophore is cut during the preparation, we cannot exclude partial injury of the rhinophore ganglia. However, the calcium imaging experiments demonstrate that parts of the rhinophore ganglion respond differentially to stimulation of the sensory epithelium with amino acids indicating that olfactory stimuli are relayed and processed in the rhinophore ganglion. In a previous investigation we tested sponge alkaloids and revealed clear responses to sceptrin at 200 μM concentration [20
]. The term olfaction is used in terrestrial animals for airborne chemicals, but also in aquatic animals for water borne chemicals detected by primary receptor neurons in the nose (e.g. fish, amphibians) or on the antennae (crustaceans) [45
]. In vertebrates it is referred to as long distance reception in contrast to close range chemoreception via taste receptors (secondary receptor cells). In Phestilla sibogae
different densities of subepithelial sensory cells and intraepithelial sensory cells were found in the oral tentacle and the rhinophores [46
]. These morphological data together with the electrophysiological evidence for greater sensitivity of the rhinophores [33
] led Boudko and colleagues [46
] to the conclusion, that rhinophores serve for long-distance chemoreception or olfaction. To our knowledge, a clear separation based on receptor cell morphology has not been done for Aplysia
; therefore we prefer to use the term "olfaction" because of the apparent similarities regarding the organization of the central pathway in glomeruli.
, processing of olfactory information was mainly investigated at the level of the procerebrum [47
], and complex odour induced oscillations were described. In Aplysia
, using calcium imaging techniques we did not find any indications for odour induced oscillations at the level of the rhinophore ganglion. However, the complexity of changes in intracellular Ca2+
-levels indicates that processing of odour information takes place within the rhinophore ganglion.
The spatial distribution of odour induced activity recorded in this study indicates that the activity in the rhinophore ganglion depends on the chemical nature and intensity of the stimulus.
Various studies in insects as well as vertebrates suggest that olfactory glomeruli represent functional units for odour processing, and olfactory information is represented in a spatial map of differentially activated glomeruli (e.g.:[50
]). It will be interesting to determine if this is also the case in Aplysia
and if glomeruli represent functional units for odour processing. Glomeruli diameters ranged between 25 μm and 135 μm indicating different numbers of neurons converging in one glomerulus. In insects, macroglomeruli were found to play an important role in pheromone communication (e.g.: [52
]). As the rhinophore of Aplysia
plays an important role in pheromone detection [17
], it would be interesting to investigate the large glomeruli and their potential function in pheromone processing. Measuring the responses of sensory neurons in the epithelium may also help to identify olfactory sensory neurons. Many marine organisms produce secondary metabolites for defence, deterrence and as pheromones. Sea slugs as well as many other marine animals depend on chemosensory information from their aquatic environment, as the optical sense plays a minor role in shallow muddy waters or the deep sea and acoustic/mechano-sensory senses give only limited information about for example food quality. The chemical sense probably is the primary sense in many marine organisms, and Aplysia
represents a promising model system for future investigations of the chemical ecology of sea slugs.