ABSTRACT
A bilaterally symmetrical pair of serotonergic neurones in the cerebral ganglia of the snail Helisoma trivolvis has major features in common with the serotonergic cerebral cells of other gastropods, including those of Helix and Aplysia. The cells were found to receive a high level of inhibitory synaptic activity which is important in determining their firing level. In the cells of isolated cerebral ganglia, tonic inhibitory synaptic potentials, uniform in frequency and amplitude, occur continually at a rate of about 2/s. In the cells of cerebral ganglia left intact with the periphery and buccal ganglia, phasic inhibitory activity occurs in addition to the tonic activity. Phasic inhibitory input could be observed at regular intervals, correlated with feeding motor activity in the buccal ganglia. Other properties of these cells include: a linear relationship between membrane potential and current injected into the cell; communication between bilateral cells in about 50% of the preparations studied; and sodium-based somatic action potentials. Some of these properties of Helisoma SCCs are different from those of Helix and Aplysia. We propose that the Helisoma SCCs are homologous to those in other gastropods and that the differences observed are the result of evolutionary adaptation of the cells to different functional roles.
INTRODUCTION
By relating properties of homologous neurones to behavioural characteristics, we may be able to determine the cellular bases of behavioural adaptation. The similarities between homologous structures are traceable to a common ancestry, whereas the differences are due to subsequent adaptation to different functions. Comparisons of apparently homologous neurones have been undertaken in a number of invertebrates including annelids (Keyser & Lent, 1977), arthropods (Paul, 1979) and gastropod molluscs (Blankenship & Coggeshall, 1976; Dickinson, 1979; Dorsett, 1974; Rózsa et al. 1980; Willows & Dorsett, 1975). Identifiable neurones of invertebrates could well afford new opportunities for studying homologies of the nervous system at the cellular level.
In the gastropods, a bilaterally symmetrical pair of large serotonin-containing neurones has been identified in the cerebral ganglia of Aplysia, Ariolimax, Helix, Lymnaea, Limax, Planorbis, Pleurobranchaea and Tritonia (Berry & Pentreath 1976; Pentreath Osborne & Cottrell, 1973; Osborne & Cottrell, 1971; Sakharov & Zs-Nagy, 1968; Sedden, Walker & Kerkut, 1968; Weinreich et al. 1973). These neurones have four features in common which suggest that they are homologous: (a) large size, (b) serotonin content, (c) a complex projection with axons extending down the cerebrobuccal connectives and into buccal nerve trunks (Berry & Pentreath, 1976; Dorsett, 1967; Gillette & Davis, 1977; Kandel & Tauc, 1966 a; McCrohan & Benjamin, 1980 a; Pentreath & Cottrell, 1974; Senseman & Gelperin, 1974; Weiss & Kupfermann, 1976), and (d) the ability to influence neurones in the buccal ganglia (Berry & Pentreath, 1976; Bulloch & Dorsett, 1979; Gillette & Davis, 1977; Senseman & Gelperin, 1974). These cells were originally described in Helix and because of their large size and position in the metacerebral lobe of each ganglion they were designated the Metacerebral Giant Cells, or MCGs (Kandel & Tauc, 1966a,b). We have chosen to use a more general teminology, serotonergic cerebral cell (SCC), when referring to cells other than those of Helix. Whereas serotonin content is a feature of these cells across all gastropods so far examined, ‘giant’ cell size is not; furthermore, in most genera, pro-, meso-, and metacerebral lobes cannot be distinguished so that the term ‘metacerebral’ is also not appropriately descriptive. Weiss & Kupfermann (1976) have already put forward a strong case for homology between the SCCs of Aplysia and the MCGs of Helix.
A pair of cerebral cells in the aquatic snail Helisoma trivolvis (Granzow & Kater, 1977) also possess the four major features of the SCCs of other gastropods. In this paper we reveal some major physiological differences between these neurones and the SCCs of the most thoroughly characterized genera, Aplysia and Helix. We conclude that the anatomical and chemical similarities support homology, and that the numerous physiological differences might be related to behavioural and ecological diversification among genera.
MATERIALS AND METHODS
The experimental animals were laboratory-raised Helisoma trivolvis of the Oregon Red strain (albinos). Electrophysiological recordings were made from either the completely isolated cerebral ganglia or the semi-intact brain preparation described by Granzow & Kater (1977). Conventional recording and display methods were employed. Intracellular recordings were made from buccal ganglia neurones with 4m potassiumacetate electrodes (DC resistance = 30–60 MΩ). 4m potassium-chloride electrodes (DC resistance = 10–20 MΩ) were used for recording from SCCs since these electrodes consistently maintained lower resistances following repeated passage of stimulating currents. Penetration of SCCs was usually facilitated by softening the cerebral ganglion sheath with a 3–4 min treatment in 0·5 % protease (type IV, Sigma Chemical Co.), a procedure which did not noticeably alter neural activity when compared ‘non-proteased preparations. The physiological saline used consisted of 51·3 mm-Na mm-KCl, 4·1 mm-CaCl2, 1·5 mm-MgCl2, and 2·4 mm-NaHCO3. In some experiments the normal saline was replaced by salines of different ionic concentrations: o Ca2+—high Mg2+ saline (CaCl2 omitted and Mg2+ concentration raised to 10 × normal by adding MgSO4), 0 Ca2+ saline (CaCl2 replaced with NaCl), and o Na+ saline (NaCl replaced with sucrose).
RESULTS
Cell size and appearance
The SCCs of Helisoma are the largest cells in the cerebral ganglia as in other gastropods. Their size does not, however, much exceed that of the next largest cells and in this they differ from the SCCs of, for example, Aplysia, Tritonia, Helix and Limax (Kandel & Tauc, 1966a,b; Dorsett, 1967; Pentreath et al. 1973 ;Weinreich, et al. 1973; Bulloch, 1977) which have been described as giant cells. In Helisoma the diameter of SCC somata range from 80–90 μm. The cells are located at the anterior edge of the dorsal surface of each cerebral ganglion, and can often be recognized by a yellowish colour which contrasts with the much paler colour of surrounding cells. Thus, while the SCCs are not as easily distinguished on cellular size alone, as are the SCCs of some gastropods, they are still readily recognizable; in approximately 95% of all preparations in which an SCC was sought it was the first cell impaled with an intracellular electrode, as verified by its characteristic activity.
Firing patterns
SCC activity was recorded mostly in semi-intact preparations. Immediately following penetration, a SCC in such preparations usually displayed spontaneous action potentials with frequencies up to 1·5 impulses/s. The pattern of firing was usually modified by apparently inhibitory synaptic input. This frequently occurred in bursts of one to several seconds in duration, producing a low-frequency periodicity in the firing of SCCs (Fig. 1 A). Generally, the firing frequency declined over the first 20-30 minutes after penetration with the recording electrode, and then the cells often became silent. This was apparently due to an increase in firing threshold of the cell since the membrane potential remained stable and the cell could be fired by the intracellular injection of depolarizing current.
In contrast, when SCC activity was recorded from completely isolated cerebral ganglia (i.e. all peripheral nerves and all connectives joining the cerebral ganglia to the rest of the central nervous system cut), the activity was very regular and at a higher frequency, reaching 2·5 impulses/s (Fig. 1B and C). Furthermore, this firing frequency was often maintained over relatively long periods of time (i.e. greater than 40 min).
Synaptic activity in isolated ganglia
The low-frequency i.p.s.p. bursts of semi-intact preparations occurred only rarely in isolated ganglia preparations and were completely absent from most. The SCCs of isolated ganglia did, however, display tonically occurring unitary potential fluctuations. These were generally very regular in both amplitude and frequency, as revealed by slightly hyperpolarizing the SCC to reduce the number of action potentials generated Fig. 1 D). These unitary potential fluctuations occurred synchronously in both cells (Fig. 2 A) indicating that they are inhibitory post-synaptic potentials (i.p.s.p.s) derived from a tonically-firing intracerebral source supplying both SCCs. The evidence that these potential fluctuations are chemical i.p.s.p.s is as follows: (1) the fast downward phase and slower rising phase of individual fluctuations were typical of the form of the classical chemical i.p.s.p. (2) The potentials were completely abolished in o Ca2+ high Mg2+ saline, suggesting a chemical synaptic origin. (3) Depolarization from resting potential augmented their amplitude and hyperpolarization diminished it, without affecting their frequency. (4) The potentials could be reversed by hyperpolarization (Fig. 3). In this experiment, the left cell was hyperpolarized with increasing steps of current while the right cell was either maintained at resting potential, or hyperpolarized just enough to reduce firing and thus more clearly reveal the i.p.s.p.s which were synchronous in the two cells. This procedure allowed the positive-going potentials in the hyperpolarized cell to be identified as reversed i.p.s.p.s by matching them with the negative-going potentials in the control cell. Gradually increasing the amount of hyperpolarization applied to the left cell first diminished the amplitude of the p.s.p.s (Fig. 3 B-C), then abolished them (Fig. 3 D) and sub sequently reversed them (Fig. 3E-F). Although in these records extra positive-going deflexions are apparent in the hyperpolarized left cell, each downward deflexion in the right cell is matched by an upward deflexion in the left cell.
Action potentials in the SCC of the isolated ganglia appeared to be generated off the rising phase of the recurrent i.p.s.p.s (Fig. 1C and D). In each case the slope of the rising phase was the same, but the cell did not always reach threshold before the next i.p.s.p. The firing frequency of the SCC is therefore apparently determined by the frequency at which the tonic i.p.s.p.s occur. In some cases two or more action potentials may follow an i.p.s.p. if the next i.p.s.p. is sufficiently delayed. Thus, it appears that the normal ‘resting’ potential of a SCC is actually above firing threshold, but that the cell is under tonic inhibition, with the frequency of the i.p.s.p.s determining the actual firing level. Alternatively, it might be that each individual i.p.s.p. momentarily lowers the threshold of the cell and that action potentials are generated as a result of post-inhibitory rebound. We have observed post-inhibitory rebound following hyperpolarization of the cell with the intracellular injection of DC current.
After relatively long time periods of recording from the SCCs of isolated ganglia (usually >40 min) the cells became less excitable and the i.p.s.p.s no longer gave rise to action potentials. Sometimes when this occurred, two different sizes of i.p.s.p.s could be observed (Fig. 2B). The larger i.p.s.p.s seem to be the summed result of two of the smaller i.p.s.p.s which become apparent when they occur slightly asynchronously. This suggests that each of the regularly occurring i.p.s.p.s is composed of input from two, probably bilaterally symmetrical, intracerebral sources which are usually strongly coupled.
Although the phasic bursts of i.p.s.p.s seen in the semi-intact preparations (e.g. Fig. 1 A) were absent from most SCCs of isolated cerebral ganglia, they did occur rarely in a few preparations (Fig. 4). This suggests that the presynaptic source(s) for at least some of the phasic i.p.s.p.s is located within the cerebral ganglia, but that activation of this source is usually dependent on inputs from either the periphery or other ganglia. The phasic i.p.s.p.s observed in isolated ganglia appeared as if they might be generated by the same source as the tonic i.p.s.p.s; the phasic nature resulting from a short duration increase in the firing frequency of that source. The small amplitudes of the unitary potentials in such barrages would be accounted for by their increased frequency which results in non-linear summation as the cell hyperpolarizes toward the reversal potential of the i.p.s.p.s.
Synaptic activity in semi-intact preparations
Synaptic activity in semi-intact preparations consisted of both the tonic i.p.s.p.s seen in isolated ganglia, and additional phasic bursts of i.p.s.p.s. No spontaneous excitatory synaptic activity was apparent. The phasic inhibitory input in these preparations generally resulted in an irregular bursting of the cell (Fig. 1A). In about 15% of all semi-intact brain preparations the SCCs received prominent cyclical bursts of i.p.s.p.s (Fig. 5). Simultaneous recordings made from feeding motoneurones of the buccal ganglia revealed that these cyclical hyperpolarizations in SCCs were correlated with the rhythmic activity of these motoneurones which comprises the feeding motor program. In particular, occurrence of the cyclical hyperpolarizations in SCCs was correlated with the occurrence of the cyclical hyperpolarizations recorded from identified protractor motoneurones (Fig. 5). The onset of each SCC hyperpolarization generally preceded the onset of the shorter duration hyperpolarization in the protractor motoneurone by approximately 0·5 s, but the two terminated almost simultaneously and bursting in the two cells also occurred in the same phase of the cycle. Thus, one type of inhibitory input to the SCCs functions in correlating activity of SCCs with that of buccal ganglia feeding motoneurones. This may be due to direct or indirect feedback from the premotor pattern-generating neurones of the buccal ganglia feeding motor program.
Compound synaptic potentials could be evoked in SCCs by electrically stimulating the cerebral nerves in semi-intact preparations. These potentials were predominately inhibitory with the exception of those derived from the large frontal lip nerve. Stimulation of this nerve produced an antidromic action potential followed by a compound p.s.p. which had an early excitatory component and a later inhibitory component.
The absence of anomalous rectification in SCCs
Anomalous rectification has been reported as a property of the MCGs of other gastropods in which the current-voltage (I-V) relationship of the cell’s membrane has been studied, including Helix (Kandel & Tauc, 1966b), Aplysia (Weiss & Kupfer-mann, 1976) and Pleurobranchaea (Gillette & Davis, 1977). We analysed the I-V relationship of the SCCs of Helisoma in several different preparations, by penetrating cells with two intracellular electrodes, one for passing current and one for measuring voltage. As shown in Fig. 6 A, membrane potential (E”m) was a linear function of this amount of current injected, i.e. the membrane resistance remained constant at and levels of membrane potential (Fig. 6B). Thus, anomalous rectification is not a property of the SCCs. We considered the possibility that the large amount of synaptic activity which occurs in these cells may have masked their anomalously rectifying properties. However, the same result -a linear I-V relationship - was also obtained in preparations bathed in o Ca2+ high Mg2+ saline which blocked all signs of synaptic input to the cells.
Communication between SCCs
Interactions occur between the two SCCs in Helisoma (Granzow & Kater, 1977). In 30% of the preparations, each action potential evoked in one cell could drive an action potential in the other cell at up to 10 impulse/s (Fig. 7A). In 20% of the preparations this communication was weaker, with post-synaptic spikes failing to follow when the pre-synaptic cell was driven at frequencies greater than 2 or 3 impulse/s. Usually this weak communication was bidirectional (Fig. 7B), but two cases of unidirectional weak communication were found (Fig. 7C). In 50% of the preparations, the SCCs did not communicate at all (Fig. 7D). We have also noted changes in the efficacy of communication within individual preparations, both from strong to weak and from weak to strong communication. Sometimes this occurred ‘Spontaneously’ and at other times the change occurred during various experimental Manipulations, such as stimulating cerebral nerves. However, attempts to alter predictably the communication with such manipulations were unsuccessful.
The absence of communication in preparations did not appear to be attributable to damage to the cells since activity was normal in all other respects.
The nature of the communication between SCCs was difficult to ascertain. Injection of D.C. current of either polarity into one SCC never resulted in a D.C. membrane potential shift in the other SCC, even when the two cells were strongly communicating with one another. However, communication between the two SCCs survived long periods of exposure to o Ca2+ high Mg2+ saline which abolished all signs of synaptic input to these cells. This suggests that the bilateral SCCs which do communicate are electrically coupled (but at a site far removed from the soma) as has been reported in some other genera including Planorbis (Berry & Pentreath, 1976), Ariolimax (Senseman & Gelperin, 1974) and Lymnaea (McCrohan & Benjamin, 1980a).
Each SCC apparently has more than one spike initiation zone. This conclusion follows from the observation that spontaneously occurring action potentials, recorded from the soma, i.e. those associated with the i.p.s.p.s described above, could be abolished with very small amounts of hyperpolarizing current (less than 0·1 nA). However, action potentials in one SCC which were driven by action potentials in the contralateral SCC were suppressed only with much larger amounts of hyperpolarizing current (greater than 2 nA). This indicates that spontaneous action potentials are initiated at a site which is located near the soma, but that action potentials driven by the contra-lateral SCC are initiated at a site which is relatively distant from the soma.
Ionic basis of SCC action potential
Bathing the cerebral ganglia in o Na+ saline reversibly abolished all ongoing SCC action potentials, and action potentials could not be evoked by the injection of de-polarizing current (Fig. 8 A). However, SCC action potentials survived repeated washings in o Ca2+ saline with little or no reduction in amplitude (Fig. 8D). Thus, the currents underlying somatic action potentials of SCC are apparently due to the influx of sodium ions with little or no contribution by calcium ions.
DISCUSSION
I. Comparison of SCCs of Helisoma and other gastropods
In Table 1, features of Helisoma SCCs are compared with those of SCCs (MCGs) of several other gastropods. This table extends the comparisons made between the MCGs of Helix and the SCCs of Aplysia by Weiss & Kupfermann (1976). We have not included data on certain features such as responses to pharmacological agents and heterosynaptic facilitation as did Weiss & Kupfermann (1976) since these areas have so far not been investigated in Helisoma and no new data is available in other.gastropods.
(A) Synaptic activity
Spontaneous levels of firing which occur in Helisoma SCCs in semi-intact brain preparations are similar to those recorded from the SCCs of semi-intact brain preparations of Planorbis (Berry & Pentreath, 1976) and Lymnaea (McCrohan & Benjamin, 1980 a). On the other hand, the high level of firing in Helisoma SCCs of isolated cerebral ganglia is in contrast to the low level of activity recorded from the MCGs of isolated ganglia in Helix (Kandel & Tauc, 1966a,b) and the SCCs of isolated ganglia in Aplysia (Weiss & Kupfermann, 1976). Both the Helix and Aplysia cells are characteristically silent unless excitatory synaptic input is evoked by stimulation of cerebral nerve trunks.
In Helisoma SCCs, spontaneous synaptic activity is predominately, if not exclusively, inhibitory, whereas in Helix and Aplysia spontaneous activity consists largely of e.p.s.p.s. In all three genera the spontaneous p.s.p.s occur in completely isolated cerebral ganglia at relatively high frequencies. Two intracerebral sources of p.s.p.s have been postulated for both Helix (Kandel & Tauc, 1966 a) and Aplysia (Weiss & Kupfermann, 1976), one source producing tonic p.s.p.s and the other phasic p.s.p.s. In Helix, individual e.p.s.p.s of both the phasic and tonic type are common to both right and left MCGs (Kandel & Tauc, 1966a). In Aplysia, individual e.p.s.p.s of the phasic type are also common in both right and left SCCs. However, the tonic type of p.s.p. (actually an e.-i.p.s.p.) occurs continuously in bursts which are synchronized in the two cells but individual p.s.p.s within these bursts are not matched one for one in the two cells (Weiss & Kupfermann, 1976). In Helisoma our data suggests a dominant intracerebral source for the spontaneous synaptic input to the SCCs. In the isolated cerebral ganglia this source generates tonic i.p.s.p.s which are common to both right and left cells. This source sometimes appears to consist of two components (presumably bilaterally symmetrical cells) which are usually tightly coupled to one another. The phasically occurring barrages of i.p.s.p.s seen in the semi-intact brain preparation could either derive from sources outside the cerebral ganglia such as primary sensory neurones or result from inputs which modify the activity of the intracerebral source which otherwise gives rise to the tonic i.p.s.p.s in the isolated cerebral ganglia. In the MCGs of Helix, both the tonic and phasic e.p.s.p.s, which have their pre-synaptic sources in the cerebral ganglia, can be activated by electrical stimulation of cerebral nerve trunks (Kandel & Tauc, 1966 a).
In some semi-intact brain preparations of Helisoma cyclically recurring barrages of i.p.s.p.s are found to correlate with cyclical hyperpolarizations occurring in identified protractor motoneurones of the buccal ganglia during generation of the feeding motor output. These cyclical hyperpolarizations could be the result of feedback from components of the buccal ganglia feeding motor program. Primary candidates for the source of this feedback would be the premotor pattern-generating neurones, the cyberchrons, since bursts in these neurones generate the cyclical hyperpolarizations in protractor motoneurones (Kater, 1974; Kaneko, Merickel & Kater, 1978). Because the onset of the hyperpolarizations in SCC actually precede the onset of the hyperpolarizations in the protractor motoneurones, another likely explanation might be that a separate source for the SCC hyperpolarizations is located in the cerebral ganglia and becomes phase-locked with cyberchron bursting during feeding. The cyclical hyperpolarizations in Helisoma SCCs serve to phase-lock the SCC bursts to those occurring in protractor motoneurones. Such phasic activity in the SCC is not required to generate rhythmical buccal ganglia motor output (Granzow & Kater, 1977). However, evidence has been obtained that the SCCs also excite protractor motoneurones directly (B. Granzow, in preparation) and in this way intensify the protractor motoneurone bursts. Phasic synaptic input to the SCCs also has been found to be correlated with rhythmical motor activity in the buccal ganglia of Limax (Gelperin & Forsythe, 1976; Prior & Gelperin, 1976), Lymnaea (McCrohan & Benjamin, 1980b) and Pleurobranchaea (Gillette & Davis, 1977). In both Limax (Gelperin & Forsythe, 1976) and Pleurobranchaea (Gillette & Davis, 1977) the cyclical synaptic input to SCCs consists of two phases—an excitatory component during the retraction phase and an inhibitory component during the protraction phase of the feeding motor program. Thus, firing of the SCCs tends to occur during the retraction phase of the cycle. In Lymnaea the SCCs also receive a two phase cyclical input-excitation followed by inhibition-but in this case the SCCs tend to burst during the protraction phase of the feeding motor program (McCrohan & Benjamin, 1980b). It therefore appears that phasic firing in SCCs may act to intensify one phase of the feeding cycle relative to another but that the particular phase affected is different in different genera. In Helisoma it has now been found that activity in protractor motoneurones can excite the premotor pattern generating neurones of the buccal ganglia (B. Granzow, in preparation). Thus, intensification of the protractor phase of the cycle could thereby excite the rhythm generating neurones and in this way account for the SCCs’ ability to initiate or accelerate the whole motor program.
Compound inhibitory synaptic potentials are evoked in SCCs by electrically stimulating most of the peripheral cerebral nerves in Helisoma. In contrast, synaptic potentials evoked by cerebral nerve stimulation are predominantly excitatory in Helix and Aplysia (Kandel & Tauc, 1966 a; Weiss & Kupfermann, 1976), although a later inhibitory component can be detected following stimulation of the tentacular nerve in Helix (Szczepaniak & Cottrell, 1973) and following stimulation of the lip nerves in Aplysia (Weiss & Kupfermann, 1976). This inhibitory component has a reversal potential near resting potential and can therefore be detected if the MCG is held slightly depolarized during nerve stimulation (Weiss & Kupfermann, 1976).
The difference in signs of both the spontaneous and evoked synaptic potentials between SCCs of Helisoma and those of Helix and Aplysia constitutes the most puzzling finding derived from our comparisons. It is possible that the balance of inhibitory and excitatory inputs which can be evoked by cerebral nerve stimulation are the same in all three genera but that the inhibitory inputs have a lower threshold to electrical stimulation in Helisoma and the excitatory inputs have the lower threshold in Helix and Aplysia. In Limax, inhibitory potentials are the predominant responses of SCC to cerebral nerve stimulation; however, chemosensory stimulation of the lip epithelium has an excitatory effect on the cell (A. Gelperin, personal communication). It is also possible that recordings made from the somata of the SCCs reflect different synaptic inputs although all are present in the different genera. For example, the location of synapses onto the cells may vary.
It is most interesting, however, to consider that the SCCs of Helisoma, Helix and Aplysia might integrate inputs from different sources or that the same inputs have different post-synaptic effects on the cells. Perhaps in some gastropods such as Aplysia and Helix, SCC activity requires excitation by stimuli relevant to feeding whereas in Helisoma the SCCs normally tend to fire and thus excite the buccal ganglia feeding motor program unless they are actively inhibited, for instance when aversive stimuli are encountered. It was found that light prods to the head epithelium of Helisoma resulted in strong inhibitory synaptic input to the SCCs (B. Granzow and C. H. F. Rowell, unpublished observations). Thus, we may speculate that a basic difference between the SCCs of Aplysia and Helix and those of Helisoma is the active initiation of SCC firing at appropriate times in the former two animals and the active termination at appropriate times in Helisoma. Such a difference could reflect an ecological necessity for gastropods such as Aplysia to be alerted to the presence of a discontinuous food source, whereas for Helisoma, a microherbivore, the food source is more widely distributed and thus feeding behaviour initiated at almost any time could possibly result in food ingestion. However, on the basis of results on intact snails, one would predict that food stimuli might initiate activity in the SCC, since quiescent snails do commence feeding behaviour and already feeding snails increase the rate and vigour of feeding movements in response to novel food stimuli, such as ‘Trout Chow’, applied nearby in the aquarium (B. Granzow and C. H. F. Rowell, unpublished observations). The excitatory component which can be evoked in the SCC of Helisoma by stimulating the frontal lip nerve, therefore, may be the result of activity in chemosensory receptors which impinge on the SCC.
Unfortunately, little data is available on the specific inputs to the serotonergic cerebral cells in any of these genera although excitatory effects of chemosensory stimuli to the lip epithelium have been reported in both Limax (Gelperin & Forsythe, 1976) and Pleurobranchaea (Gillette & Davis, 1977). Recent work in Aplysia has demonstrated the existence of a chemical synaptic connection from another identifiable cerebral ganglion neurone, C-2 (Weiss et al. 1978); however, the function of C-2 as yet remains to be determined. The cellular analysis of sensory systems in gastropods has in the past been hindered by problems associated with locating and recording from the cell bodies of sensory neurones, many of which are located in the peripheral
Rody tissues (e.g. Kater & Rowell, 1973; Janse, 1974; Emery & Audesirk, 1977), However, an increasingly large number of primary sensory neurones are now being identified in gastropod central ganglia (e.g. Byrne, Castellucci & Kandel, 1974; Getting, 1974; Alkon, Akaide & Harrigan, 1978; Audesirk, 1979; Audesirk & Audesirk, 1979; Rosen, Weiss & Kupfermann, 1979). Thus, a thorough examination of the cerebral ganglia in Helisoma as well as in other gastropods may well reveal primary sensory neurones which could be tested for their effects on the serotonergic cerebral cells. Such an approach seems essential for furthering our understanding of the factors which determine the level of activity in these cells.
(B) Anomalous rectification
Anomalous rectification presumably involves a voltage-dependent change in membrane conductance to potassium and is related to the external potassium concentration [Ko] (Marmor, 1971; Apland & Livengood, 1979). Marmor (1971) has demonstrated that the G cell of the marine gastropod Anisorodis manifests anomalous rectification when bathed in artificial sea water ([Ko] 10 mm) but that anomalous rectification was absent when [Ko] was reduced to zero. Thus, the presence of anomalous rectification in the MCGs of the terrestrial snail Helix (Kandel & Tauc, 1966b), and the marine gastropods Aplysia (Weiss & Kupfermann, 1976) and Pleurobranchaea (Gillette & Davis, 1977) and the absence of anomalous rectification in the SCCs of the freshwater snail Helisoma may be related to the differences in the potassium concentration of the bathing salines which in turn reflect the potassium concentrations of the haemolymph of these animals. The [Ko] in the snail physiological saline used to bathe Helisoma cerebral ganglia is low (1·7 mm) compared to that in the artificial sea water used to bathe Aplysia and Pleurobranchaea cerebral ganglia (10 mm) (Weiss & Kupfermann, 1976; Gillette & Davis, 1977) or the physiological saline used for Helix cerebral ganglia (5·8 mm) (Cottrell & Macon, 1974). Thus, whether or not an SCC in a particular genus manifests anomalous rectification may be a consequence of that genus’ external ionic environment. Whether there are also differences in the ionic channel constituents of the membranes of these cells remains to be determined. In either case, the presence of anomalous rectification in the SCCs of some genera and its absence in others will cause the cells to respond differently to the same types of synaptic inputs, all other factors being equal. Such a difference may be especially important in influencing the interactions between synaptic potentials when they occur in close temporal proximity. For example, heterosynaptic facilitation is thought to be a consequence of the anomalously rectifying properties of the SCC in Aplysia (Weiss & Kupfermann, 1976).
(C) Ionic basis of the action potential
We have found that the generation of somatic action potentials in the SCC of Helisoma involves a large contribution by sodium ions and little or no contribution by calcium ions. Whereas little comparable data is available on the SCCs in other gastropods, Cottrell (1977) reported that somatic action potentials in the MCG of Helix remained in reduced external sodium and were probably mediated by calcium. In Planorbis, action potentials are propagated in zero calcium high magnesium saline Just as in Helisoma (Berry & Pentreath, 1976). Further comparisons of this type could indicate similarities and differences among SCCs at the membrane level.
(D) Communication between SCCs in Helisoma
Another notable difference between the SCCs of Helisoma and those of Helix and Aplysia is the existence of communication between these cells in Helisoma. There is also electrical coupling between the bilateral serotonergic cerebral cells of Ariolimax (Senseman & Gelperin, 1974), Planorbis (Berry & Pentreath, 1976) and Lymnaea (McCrohan & Benjamin, 1980 a). The variability in communication seen among preparations in Helisoma has also been reported for the serotonergic cerebral cells of Planorbis (Berry & Pentreath, 1976). In Ariolimax, the serotonergic cerebral cells could be directly coupled to one another in the cerebral ganglia since each cell sends a process into the contralateral ganglion. However, in Helisoma repeated intracellular injections of Lucifer Yellow CH into the SCCs have so far revealed no processes of the SCC extending towards the contralateral cerebral ganglion. Similar results have been found in Lymnaea where the site of electrical coupling has been localized to the buccal ganglia. In this case cutting the buccal commissure usually abolished the coupling, although some weak coupling sometimes remained (McCrohan & Benjamin, 1980a). The coupling contacts between the two cells are apparently made within the buccal commissure itself since axon processes do not extend across the commissure into the contralateral buccal ganglion. The weak coupling which survives cutting of the buccal ganglia commissure may be mediated via contacts made between the two cells in the subcerebral commissure since axon processes extend along this pathway. In Helisoma, putative sites of contact between SCCs also exist within the buccal ganglia since each cell does send processes into the contralateral buccal ganglion and out contralateral as well as ipsilateral buccal nerve trunks (Granzow, 1979). We have also found that in some preparations cutting a cerebrobuccal connective will abolish communication, but that in other preparations the communication remains relatively strong. Thus, while it is highly likely that the neurones are coupled within the buccal ganglia it is also necessary to postulate another pathway, e.g. an interposed interneurone which mediates coupling within the cerebral ganglia, since no direct sites of interaction occur outside the buccal ganglia.
The existence of communication between SCCs in Helisoma appears redundant considering that the synaptic inputs to the two cells are usually so similar. In this regard it is also relevant to consider the range of apparent redundancy across genera in the serotonergic cerebral cells’ innervation of the buccal ganglia (Fig. 9). In all genera, the serotonergic cerebral cells send an axon branch down the ipsilateral cerebrobuccal connective (CBC). In Ariolimax (Senseman & Gelperin, 1974), Helix (Kandel & Tauc, 1966a; Pentreath & Cottrell, 1974) and Tritonia diomedea (Dorsett, 1967) the cells also send processes down the contralateral CBC, whereas in Aplysia (Weiss & Kupfermann, 1976), Helisoma (Granzow, 1979), Limax (Senseman & Gelperin, 1974), Lymnaea (McCrohan & Benjamin, 1980a), Planorbis (Berry & Pentreath, 1976) and Pleurobranchaea (Gillette & Davis, 1977) processes neither extend to the contralateral cerebral ganglion nor down the contralateral CBC. In Tritonia hombergi, based on intracellular injections of cobalt chloride, Bulloch (1977) found that eight out of ten SCCs filled lacked contralateral CBC axons whereas the other two had contralateral CBC axons in addition to the ipsilateral CBC axons. In all genera but ArioUmax (for which no data is available) and Lymnaea, axons continue into both ipsilateral and contralateral buccal ganglia and out ipsilateral and contralateral buccal nerve trunks (although not every nerve trunk). In Lymnaea, axon processes extend out ipsilateral buccal nerve trunks but apparently do not cross the contralateral ganglion (McCrohan & Benjamin, 1980a). Considering the presence of communication between bilateral SCCs in Ariolimax, Helisoma, Lymnaea and Planorbis and the different number of parallel descending CBC axons of the SCCs of various genera as well as their extensions into both ipsilateral and contralateral buccal ganglia, there are varying degrees of apparent redundancy in the SCCs’ innervation of the buccal ganglia among the gastropod genera (Fig. 9). Such redundancy seems to be a common feature of the bilaterally symmetrical neural systems in gastropods (Bahls, Kater & Joyner, 1980) apparently assuring that bilaterally symmetrical elements function in effect as a single unit (McCrohan & Benjamin, 1980a).
(E) Functional role in buccal ganglion activity
The SCCs of Helisoma can excite the buccal ganglion to produce the motor program for food ingestion (Granzow & Kater, 1977). The role of the SCCs in modulating components of the feeding motor output of the buccal ganglia has been demonstrated to various extents in a number of different gastropod genera including Planorbis (Berry & Pentreath, 1976), Pleurobranchaea (Gillette & Davis, 1977), Aplysia (Weiss, Cohen & Kupfermann, 1978) and Tritonia (Bulloch & Dorsett, 1979).
II. The case for homology
Since selection pressures act on behaviour, evolutionary relationships among animal groups should be reflected in homologies in the nervous system. The strength of the case for homology between identifiable neurones in different animal groups depends not only on the number of shared properties, but also on the extent to which those cells are characterized by properties which make them unique from other neurones (Weiss & Kupfermann, 1976). On the other hand, the functional role of homologous neurones can differ among animal groups, and this implies structural or chemical differences between the cells or thier associated neurones.
We conclude that the gastropod SCCs are homologous because of their unique combination of structural and chemical features. However, there are an impressive number of differences, especially apparent between Helisoma and the best-studied genera, Helix and Aplysia. These differences all appear to be relevant to the integrative role of the cells and we believe they will ultimately be explicable in terms of adaptive function. There are already indications of such a link between life style and physiological specialization in the above comparison of Aplysia and Helisoma.
The gastropods differ widely in the habitats which they occupy and include marine, freshwater and terrestrial groups. In addition they exploit widely disparate food sources which require the use of different feeding modes. In genera already studied neurophysiologically these range from grazing on macroalgae (Aplysia) or microalgae (Helisoma) to rapid ingestion of relatively large pieces of animal matter (Pleurobranchaea and Tritonia) (Kandel, 1979). This class of molluscs therefore offers a unique opportunity to compare across many diversified genera an identifiable Kurone whose basic behavioural relevance is already known. It will be especially interesting to determine how cellular properties might be related to physiological or behavioural properties across genera. Such a comparison might reveal the means by which natural selection acting at the behavioural level influences the evolution of the nervous system at the level of individual neurones.
ACKNOWLEDGEMENTS
This study was supported by NIMH predoctoral traineeship MH16109 to B. G. The authors are grateful especially to Dr S. B. Kater, in whose laboratory this work was carried out, for assistance in some phases of the study, as well as constructive criticisms throughout all phases, and also to Dr A. G. M. Bulloch for a critical reading of the manuscript.