ABSTRACT
In studies of the activity of isolated organs it is desirable to bathe them in a fluid similar to that found in the intact animal. Previous work on nervous systems indicates that their electrical activity is greatly affected by the concentration of ions in the medium, except in cases where they are enclosed in a thick impermeable sheath. In gastropods the suitability of a 0·6-0·75 Locke solution for in vitro experiments on isolated organs was demonstrated by Cardot (1921), who studied the effect of different dilutions of mammalian Locke solution on the frequency and amplitude of the heart beat in Helix. A 0·7 dilution (sometimes called Cardot’s solution) has generally been used by previous workers on the nervous system of gastropods, and their preparations retained constant electrical activity for many hours, but there is an absence of studies on the effect of variations in the composition of the bathing fluid on the gastropod nervous system. This is of interest because the blood of slugs and snails is known to show considerable fluctuations in concentration. The problem may also be of significance in relation to the general behaviour of the animal, since it is known that these fluctuations are related to the activity of the animals.
Thus Duval (1930) and Kamada (1932) found that the body fluid is more concentrated in hibernating and aestivating snails than in active animals. Similarly, Brand (1931) showed that the water content of active snails was substantially greater than that of inactive ones.
Changes in body fluid concentration may occur quite rapidly. Arvanitaki & Cardot (1932) observed a decrease in the depression of freezing-point of the blood from 0·47 to 0·2° C. following rain, and Pusswald (1948) has shown that the major part of the water loss from slugs comes from the blood. She also found that it occurred more rapidly in Limax than in Arion, and related this to the greater sensitivity of the former to desiccation and its increased activity after rain. In contrast to the findings of these authors, who have assumed a direct relationship between water content and activity, are the conclusions of Wells (1944) and Dainton (1954). Their experiments suggest that high water content is not the immediate cause of activity but that it determines whether a response to a particular sensory stimulus will take place.
In this paper observations are given on the effect of different dilutions of Locke solution on the frequency of discharge from individual units within the pedal ganglia. The work was made possible by the development of a technique for recording the electrical activity, while the ganglia remained in the bathing fluid for long periods and so were not subjected to drying or oxygen lack.
MATERIAL AND METHODS
Specimens of Agriolimax reticulatus were collected from gardens and were usually kept for a few days on damp blotting-paper prior to an experiment. They were fed on pieces of carrot.
The dissection was performed on an unanaesthetized animal pinned to a wax block. The viscera were removed to expose the brain and two pedal nerves running along the foot. The brain consists of two cerebral ganglia (supraoesophageal) and a series of suboesophageal ganglia. The largest of these are the paired pedal ganglia, each giving off a large pedal nerve. In addition, there is a small ring of ganglia situated between the pedal ganglia and the buccal mass. The brain was removed from the animal, together with the long pedal nerves which were ligatured with nylon threads. The preparation was completed by removal of the cerebral and other small ganglia. The pedal ganglia now remained with the two pedal nerves attached to the nylon threads which were used to manipulate the preparation into any required position.
In the first series of experiments the ganglia were hooked beneath chlorided silver wires mounted in a Perspex holder which could be dipped in and out of different solutions.
In the second series the preparation remained in the bathing solution throughout the experiment and did not suffer from periodic drying (Fig. 1). The nylon threads were fastened to a microscope slide by means of elastic bands. A coating of black paraffin wax on the slide made the ganglia more easily visible, and provided a surface which was less likely to damage the fine electrode. The slide was fixed in a Perspex holder contained in a trough filled with the bathing solution. A series of baffle plates surrounding the slide served to protect the preparation from disturbing currents when the solution was changed. This was done by using a filter pump to drain one solution before the next was run in.
The electrodes consisted of sharpened 0·002 in. tungsten wire fused into a Pyrex glass tube filled with mercury. By means of a Peterfi micromanipulator the electrode was easily thrust into a ganglion. The bathing solution was earthed by a tungsten wire. Impulses were amplified by a Grass P 4 amplifier and led to a Cossor double-beam oscilloscope. They were also passed through a level-selecting device before being fed to a Dynatron scaling unit. The input of this was connected to the second beam of the oscilloscope so that it was possible to see exactly which part of the activity was being counted. In some cases negative pulses were counted, but in other preparations the positive part of the wave-form proved more suitable. Readings were taken every 30 sec.
The bathing solutions were made by diluting standard Locke solution ; the osmotic pressure of these and other solutions were determined by the depression of freezing-point method of Ramsay (1949).
RESULTS
(1) The nervous system in situ
Observations were first made on the electrical activity of the nervous system when the animal was opened and the viscera displaced to one side. The activity recorded from electrodes placed on one of the pedal nerves or adjacent to the pedal ganglia is composed of an apparently random sequence of potentials of varying height and duration (Fig. 2A). Larger spikes occasionally stand out from this background activity. Stimulation of the foot produced a burst of larger waves as did stimulation of the head (Fig. 2B). With the external electrodes it was difficult to detect any definite pattern in the recordings, but a tungsten electrode inserted into the pedal ganglia gave much simpler recordings (Fig. 2E). Rhythmically active units were then readily obtained from almost all the ganglia. The type and frequency of such ganglionic activity were variable, and in no case was there any obvious correlation between the potential changes and movements of the foot, heart, alimentary canal or other parts of the animal. It was possible to modify this electrical activity by stimulating the animals as mentioned above.
(2) Isolated preparations
A. External electrodes’
At the commencement of this work many experiments were carried out using a pair of hooked silver-silver chloride electrodes. As with the in situ recordings there was considerable background activity and rhythmically active single units were difficult to distinguish (Fig. 3A). Bullock (1945) obtained similar recordings from the cerebral ganglia of Agriolimax. In many preparations well-defined unit activity seemed to be absent, but they frequently showed rhythmical changes in amplitude of the background (Fig. 3B). Such records suggest variations in the degree of synchronization of several units, such as have been found in other invertebrate ganglia (cf. Arvanitaki, 1942).
Quantitative studies were only carried out on preparations having well-defined unit activity. Although most preparations in air showed several active units, the half-minute count remained relatively constant for periods of up to 8 min. (Fig. 4). It was apparent that after being exposed to air for this time the preparation dried up and the electrical activity was inhibited. This effect was reversible since activity returned when the preparation was immersed once more in Locke solution. In addition to the drying effect, it was found that the removal of the nerves and ganglia from the solution stretched and stimulated the nerves. Drying was over-come to some extent by raising the preparation into liquid paraffin, but the tension at the interface still disturbed the position of the preparation on the electrodes. This was a serious disadvantage for a quantitative study because the size of the potentials recorded from a given unit varied. As this criterion was being used to discriminate between different units, the accuracy of the counts was not constant. Furthermore, the periodic movement of irritable tissue in and out of the solution must have some stimulating effect and thus decrease the ‘spontaneous’ nature of the recorded activity. It is apparent that this technique, which has been largely used by previous workers, has many disadvantages if a study is to be made of the effect of variations in the composition of the bathing fluid on the activity of isolated nerve centres.
B. Internal electrodes
The use of fine tungsten electrodes piercing the sheath surrounding the ganglia largely overcame these difficulties and enabled a study to be made of single units which were easily discriminated from the background (Fig. 3C). By this means a continuous record of ganglionic activity could be obtained over long periods. In many preparations the electrical activity persisted for about 24 hr. The pattern of activity during the first 3½ hr. after isolation is remarkably constant for all preparations (Fig. 5). The half-minute count is high (30) to start with, and usually falls to a lower level (5) after half an hour. Recovery to a count of 10-15 then takes place and this level is maintained with remarkable constancy for about 2 hr. This provides a convenient time during which the effects of changes in the environment can be studied. A steady decline in the frequency of discharge ensues until the preparation dies.
The size, shape and polarity of the potentials varied in different preparations pre-sumably because of differences in electrode position with respect to their source. They were usually triphasic in form and about 200μV. The largest potentials were obtained with the electrode in the centre of the ganglion.
It was rare for there to be more than one active unit recorded, but in cases where this occurred it was noticed that the rhythms were independent of one another (Fig. 3D).
(3) The effect of variations in the concentration of the bathing fluid
The same dilution of Locke solution (0·7) used by previous workers on the gastro-pod nervous system was adopted as the standard solution in the present investigation. The depression of freezing-point of this solution falls in the middle of the range found in the blood of slugs brought in from gardens (Fig. 10) and therefore the tissue was not subjected to harmful osmotic forces during isolation. As has been described above, the pedal ganglion preparation remained in good condition for many hours and shows electrical activity very similar to that recorded in situ.
The technique described for changing the solution proved very satisfactory as the change was completed within 30 sec. and the activity of the preparation was not significantly influenced (Fig. 6). Initially, preparations were allowed to equilibrate for 30 min. in 0·7 Locke solution before making any change in the surrounding fluid. As a result of a change in concentration, it was always found that a more dilute solution accelerated the rhythm, whilst a more concentrated solution reduced the activity. The rate of development and magnitude of these effects are roughly proportional to the difference in concentration of the two solutions. Fig. 6 shows the reduction in the half-minute count from 12-14 to 2-3 when the solution surrounding a fresh preparation was changed from 0·7 to o·8 Locke. This lower level was maintained for 10 min. and immediately rose to about 11 when the original concentration was restored. Similar experiments involving changes from 0·7 Locke to dilutions of 1·o, o·6, 0·5 and 0·25, were carried out and some of the results shown in Fig. 7. Most preparations were sensitive to a change of the order of 0·05 Locke, i.e. from 0·7 Locke to 0·65 Locke.
Although in general it is possible to relate the frequency of discharge to the concentration of the medium, this is complicated by several factors. The first of these is the condition of the animal from which the ganglion was isolated. The slugs were kept in fairly standard conditions for some time before the operation, but variations in the nutritive level and/or the state of hydration of the animal (Howes & Wells, 1934) were unavoidable. Arvanitaki & Cardot (1932) experienced similar difficulties when studying the isolated heart of the snail. They found that whereas the heart of one snail might beat perfectly well in a given dilution of Locke, yet the same solution would stop the heart of another snail. Slow acclimatization following isolation allowed both hearts to continue beating normally.
A second factor concerns alterations in the sensitivity of the preparation which occur with time. In many experiments the first change in concentration produced the most marked effects. Changes in activity brought about by a series of identical changes often became progressively smaller (Fig. 6). This suggests some sort of adaptation similar to Arvanitaki and Cardot’s findings with the isolated heart. That the preparation is not completely fatigued by such a series of changes is shown by its response to a more dilute solution after it has become adapted to the change from 0·7 Locke to o·6 Locke. Such a preparation will still be very responsive to a change to 0·5 Locke.
Further evidence for adaptation is provided by studying the effects of prolonged immersion in a given solution (Fig. 8). It has always been found that the increase in activity is gradual with a small change in the concentration, and in this case (O·7-O·6) it reaches a maximum about 20 min. after the change, which is unusually slow. There follows a decrease in activity, which is usually maintained above the original level.
It is apparent that nerve cells in the pedal ganglion are sensitive to changes in the concentration of the bathing medium. The precise cause of these changes is not yet clear, but preliminary experiments show that the effect is mainly an osmotic one and not due to the change in the concentration of individual ions. Thus 0·7 Locke solution with mannitol added to make it iso-osmotic with a 1·o Locke solution had an identical effect on the preparation as a 1·o Locke solution. Experiments on the pedal ganglion of the snail have given similar results.
(4) The range in water-content of the slug
It is desirable to know if the osmotic changes used in these experiments on isolated preparations fall within the physiological range to which the ganglia may be subjected in the life of the animal. Slugs were kept under different conditions of humidity and samples of their body fluid taken. Those kept in very dry conditions soon ‘set’ as rigid jellies and were completely immobile. Animals dehydrated in a desiccator lost about 35 % of their weight in 3 hr. When slugs treated in this way were allowed to come directly into contact with water, they regained their initial weight within an hour. No change in weight took place if they were kept in an atmosphere saturated with water, but continued desiccation resulted in further loss in weight (Fig. 9). These changes in weight were due to evaporation and not to loss of slime, as the animals were immobile.
During the life of the animal there will be a certain amount of water-loss due to evaporation, particularly in relation to the seasons, but losses consequent on the secretion of slime during locomotion (Kunkel, 1916) probably affect the animal even more markedly. Dainton (1954) found that after 49 min. locomotion in a saturated atmosphere there was a 17 % loss in weight and the slug showed reduced activity in response to further stimulation. A similar decrease in activity followed waterloss due to evaporation. In both cases the greatest fall in water content would occur in the body fluid (Pusswald, 1948) which bathes the nervous system.
Animals maintained in moister conditions showed much greater activity and were more reactive to mechanical and other forms of stimulation. Blood was obtained from these active animals and values for the depression of freezing-point were about 0·3° C. under very moist conditions. The blood of desiccated slugs had a much higher osmotic pressure (Δ = o·8° C.). Determinations of the osmotic pressure of a range of dilutions of Locke solutions shows that this range is equivalent in depression of freezing-point to dilutions of 0·4 to 1·4 Locke solution (Fig. 10). There appears to be little doubt that changes in osmotic pressure of the body fluids of the slug occur which are comparable in magnitude with those used in the experiments on the isolated ganglia. However, it must be remembered that in the present series of experiments these changes have been relatively rapid, whereas in nature they would presumably occur more slowly.
DISCUSSION
The persistence of electrical activity in parts of the central nervous system when isolated from the rest of the animal appears to be a property of the nervous system of all groups that have so far been investigated. The interpretation of such activity and its significance in the behaviour of intact organisms continues to present problems at all levels of study in the attempt to bridge the gap between behavioural studies and neurological experiments. A major difficulty is to preserve the isolated nervous tissue in conditions as normal as possible. The disruption of the blood or tracheal supply inevitably leads to conditions of oxygen lack in the nerve cells and hence to abnormal activity. Desiccation during the course of an experiment also produces abnormalities. A priori it would seem that these dangers are less severe in a study using nervous tissue which is normally bathed in a haemocoel. Such ganglia when isolated in a solution similar to the haemocoelic fluid can be expected to remain in fairly normal or even in unusually favourable conditions with regard to the oxygen concentration. There remain of course the injury effects produced by the severance of the nerves, but these are unavoidable in any isolation experiment.
From the experiments described in this paper it appears that a 0·7 Locke solution is suitable for long-term observations, as activity and excitability persist for over 24 hr. Since similar potentials were recorded in the ganglia before isolation it is probable that the activity studied occurs normally in the slug and is not an artifact due to isolation. Furthermore, as the blood of slugs normally shows fluctuations in water-content comparable with those described in these experiments, it seems likely that the activity of the ganglia in vivo will vary in a way similar to that found in the present work.
The potentials recorded represent some unit of activity within the ganglia as they maintain a constant shape and size and are distinguishable from other units by these characteristics. On the whole it was rare to find more than one unit active with a given electrode position. The potentials are probably produced by motor neurones as each beat is succeeded by a corresponding action potential in the pedal nerve. Whether they are derived from a single cell or from a group of cells cannot be decided, although the smoothness of outline and constancy of shape argue in favour of a single cell. It is significant to note that there are large cells in the pedal ganglia of up to 150/x in diameter. These are comparable with the large cells isolated and examined by Arvanitaki (1942) from Aplysia ganglia.
As yet the precise function of this activity in the life of the animal has not been determined. In no case has there been any suggestion of a correlation between contraction of the foot musculature (or of any other activity of the slug) and the rhythmic discharge of these units within the central nervous system. The lack of a direct relationship between impulses in the pedal nerve and movement of the foot was also observed by Turner & Nevius (1951). The co-ordination mechanism of the molluscan foot is incompletely understood, particularly the relationship between activity of the pedal ganglion and the peripheral nerve net, but it is generally accepted that co-ordinated locomotion requires the presence of the pedal ganglion. Perhaps the rhythmic electrical activity studied here represents the physiological basis of the tonic function of the pedal ganglion which forms a part of the hypothesis of several authors (Jordan, 1918; Herter, 1931).
Although no direct connexion can be established between the electrical activity studied in the present work and the locomotory activity of the animal, it is not impossible that a general connexion exists and that the discharge frequency of these units is indicative of the ‘vigilance’ or central excitatory state of the ganglia.
If this supposition is correct then from the experiments described in this paper it is probable that an increase in water content will produce greater activity of these centres and so influence the behaviour of the animal. It is not suggested that the changed behaviour is a direct consequence of the altered rhythm of the units, but rather that some properties of the central nervous system are changed in such a way as to make the animal respond more readily to peripheral stimuli. Turner & Nevius do not record any observations on the influence of the ‘spontaneous’ potentials on synaptic transmission through the pedal ganglion or vice versa. This preparation might be suitable for studies on this general problem of the relationship between sensory input to the central nervous system and its intrinsic activity in determining the pattern of motor activity.
This interpretation is largely in agreement with the most recent observations on the locomotory activity of snails and slugs in relation to water-content. Both Wells (1944) and Dainton (1954) have emphasized that high water-content is not the immediate cause of activity, which they suggest is due to some form of sensory stimulation; but they have also pointed out that these animals are more responsive under conditions of high water-content.
However, before any definite conclusions can be reached about the role of changes in ganglionic activity in vivo further work is necessary on the nature of the potentials recorded and also on the relationship of the pedal ganglion and the foot nerve net to locomotory activity of the foot.
SUMMARY
A method is described by which the electrical activity of single units in the isolated pedal ganglia of the slug Agriolimax reticulatus can be studied quantitatively for many hours. When tungsten microelectrodes were used the results were more simple and less variable than those obtained using external silver-silver chloride electrodes.
The activity studied consisted of potentials of characteristic size and shape. Their frequency was usually about 30/min. No correlation was observed between these and any rhythmic activity of the animal.
The blood of the slug shows considerable variation in its water-content, ranging from a concentration equivalent to that of a 1 ·4 Locke solution in dehydrated animals to one equivalent to 0·45 Locke in hydrated animals.
Animals taken straight from a garden in the evening had a body fluid equivalent to 0·7 Locke.
The ‘spontaneous’ activity of the isolated pedal ganglia was greatly affected by a change in the concentration of the bathing solution. Concentration of the medium decreased the activity and dilution increased the activity.
ACKNOWLEDGEMENTS
This effect appears to be due to the change in osmotic pressure rather than a change in the concentration of individual ions. The possible significance of these changes in the life of the animal is discussed.