Morphology and Odor Sensitivity of Regenerated Snail Tentacles

RONALD CHASE and RIFA’AT KAMIL

Department o f Biology, McCill University, 1205 Avenue Docteur Penfield, Montreal, Quebec H3A 1B1, Canada

Received February 22,1982; revised August 17,1982

SUMMARY

This study examined certain structural and functional aspects of the olfactory system in regenerated posterior tentacles of the terrestrial snail Achatina fulica. Regeneration of the epithelial sensory pad occurs with accurate size regulation. All five neuronal cell types which are normally revealed by horseradish peroxidase backfilling are also regenerated. The sensory cells attain normal numbers a t about 20 weeks postlesion. The organization of neuronal elements within the tentacle is chaotic, however, at early stages of regeneration. Even later, the di- gitlike extensions of the ganglion, which are characteristic of intact tentacles, fail to appear. The recovery of odor sensitivity was evaluated using a tentacular olfactometer and a behavioral assay that involved locomotor orientation towards the odor stimulus. Thresholds and concentration-dependent response rates were equivalent for regenerated and intact tentacles, tested in the same animals, a t 10 weeks post-lesion.

INTRODUCTION

The regenerative capability of the tentacles in pulmonate molluscs has been recognized in the scientific literature for over 200 years (see ChQtail, 1963). Not only does the tentacular wall regenerate, but also the eye and the specialized epithelium which is sensitive to wind and odors (the “sensory pad”). Previous studies have documented the normal appearance of regenerated tentacles, de- scribed the successive stages of regeneration, and concluded that the process of regeneration resembles embryological development (ChQtail, 1963; Eakin and Ferlatte, 1973; Flores and Pelligrino de Iraldi, 1980).

The present study is concerned with the sensory pad and its neural innervation in the terrestrial snail Achtina fulica. This system is anatomically well devel- oped. It consists of an epithelial surface area approximately 1.5 mm2, a large peripheral ganglion and a nerve, connected with the cerebral ganglion, which is approximately six times thicker than the optic nerve. The neuronal elements in the ganglion and its digitlike extensions have been described (Chase and Kamil, 1983). There are at least five distinct neuron types, each found in a characteristic location and each apparently sending an axon into the nerve towards the CNS. The quantitative sensitivity of the tentacle to wind and odor stimulation has also been reported. Wind velocities as slow as 25 mm/s evoke an electrophysiological response (Chase, 1981). Pure organic volatiles in con- centrations as low as 10-8 M evoke a behavioral response (Chase, 1982). Fur-

Journal of Neurobiology, Vol. 14, No. l , pp. 43-50 (1983) 0 1983 John Wiley & Sons, Inc. CCC 0022-3034/83/010043-08$01.80

44 CHASE AND KAMIL

thermore, it is clear that the tentacles have a critical importance in the snail’s olfactory orientation behavior (Chase and Croll, 1981).

The purpose of the present investigation was to determine whether the structural and functional complexity of the intact olfactory apparatus is retained in regenerated tentacles. Neuronal components and fiber pathways were ex- amined by backfilling the tentacle (olfactory) nerve with horseradish peroxidase (HRP). Odor sensitivity was measured by means of a two-chambered tentacular olfactmeter. The same methods were applied to both intact and regenerated tentacles.

METHODS

The specimens were mature terrestrial snails of the species Achatina fulica, obtained from a laboratory colony. Regeneration was induced by cutting the posterior (optic) tentacles just beneath the terminal knob. Since the cuts were done with scissors in freely moving animals, the exact location of the lesion was variable. A given animal suffered only a single lesion, so that the remaining tentacle could be used as an intact control.

Measurements of sensory pad surface area were obtained from tentacles that were removed from the animal, cut open, and pinned out in a Ringer solution. An image of the pad was projected by means of a camera lucida onto a piece of graph paper. The area of the pad was determined by counting the number of lined squares contained within the outline of the pad.

Neuronal structures were viewed after backfilling the tentacle nerve with a 20% solution of HRP, as described in the preceding article (Chase and Iiamil, 1983).

Odor sensitivity was evaluated by means of a two-chambered tentacular olfactometer. The design and use of this device is described in full elsewhere (Chase, 1982). Briefly, it consists of a pair of glass chambers which fit over the snail’s tentacles. In the present experiments, the vapor of amyl acetate was circulated through one chamber to maintain a known concentration of odorant within the local vicinity of one of the tentacles. The other tentacle was exposed to a flow of clean air at the same rate of circulation. Responses to the odor stimulus were recorded as locomotor turns towards the side of odor presentation, subject to criteria of latency and extent. The odor stimulus was pre- sented to a regenerated tentacle and an intact tentacle, in the same animal, on alternate trials.

RESULTS

Morphology

The general appearance, in uiuo, of the fully regenerated tentacle is very similar to the intact one, except that the regenerated tentacle is slightly shorter, the terminal knob is lighter in color, and perhaps the knob is more bulbous. At about 3 weeks postlesion the inner and outer tentacular cylinders join together (see Chetail, 1963; Flores and Pellegrino de Iraldi, 1980), permitting extension and retraction of the appendage. The sensory pad, however, is not evident until about 4 weeks postlesion [it is barely discernible in the 4-week specimen illus- trated in Fig. 3(A)].

The sensory pad of regenerating tentacles increases in surface area until it is approximately as large as the pad belonging to the intact tentacle in the same animal (Fig. 1). In fact, size regulation is so precise that after 90 days postlesion, the mean ratio of regenerated pad area to intact pad area does not differ signif- icantly from 1.0 (t-test; p >0.05). The apparent equivalence does not mask a hidden left-right bias since the mean ratio of pad areas in the left and right tentacles of unoperated animals is likewise not significantly different from 1.0 (t-test; p >0.5). The absolute size of the sensory pad shows considerable vari- ation in individual animals, and is closely related to the weight of the animal (Fig. 2). Size regulation is effective throughout the range of weights represented in

REGENERATED SNAIL TENTACLES

- % 1.4- E - 1.2-

p 1.0- ?

0.8-

0 m

c $ 0.6-

45

O2

I , I

50 100 150 200 250 Regeneration Period (Days)

Fig. 1. Size regulation in the regeneration of sensory pad epithelium. The ordinate indicates the ratio of sensory pad surface area on the regenerated tentacle to surface area of the pad on the intact tentacle, measured in the same animal. For comparison, the mean ratio of surface area on the pads of left and right tentacles in unoperated snails is 1.025 f 0.07 SD ( N = 17; see Fig. 2).

Figure 2, but the means by which information pertaining to tentacle size is in- corporated into the mechanism of regulation is unknown.

Backfills of the tentacle nerve with HRP reveal that regenerated sensory pads are innervated in a manner similar to intact tentacles (Fig. 3). For example, all five neuron types which are apparent in intact tentacles after backfilling (see Chase and Kamil, 1983) are also found in regenerated tentacles. In some re- spects, however, the regenerated neural structures are clearly different. Most obvious is the absence of the digitlike extensions of the ganglion which are characteristic of intact tentacles. In Figure 3(C), for example, there are nu- merous fibers emanating from the region of the ganglion, and to some extent they are bundled together, but the pattern of branching is much more ramified than in the intact structure [Fig. 3(D)]. Moreover, while normally the somata of neurons which innervate the sensory pad (excluding the receptors) are located only in the ganglion and in the digits [Fig. 3(D), and Chase and Kamil, 19831, in regenerated tentacles the cells are distributed in an apparently random fashion across the entire internal face of the sensory pad [Fig. 3(C)]. Regenerated tentacles also frequently lack a clearly articulated peripheral ganglion. For

1.67 .. .*

8: ..

L , , , , , , , , , ,

Weight (Grams) 10 20 30 40 50

Fig. 2. The variation in sensory pad surface area with weight of the snail. Two points are plotted at each weight, representing the intact left and right tentacles of single animals. The number “2” beside a point indicates that the two values are identical, or nearly so.

46 CHASE AND KAMIL

Fig. 3. Wholemounts of the tentacular olfactory structures at several stages of regeneration. In each case, the nerve, shown at the bottom of each micrograph, was backfilled with HRP. (A) 4-week regenerate. The sensory pad is poorly developed and only a few nerve fibers are labelled. The la- belled cell bodies are not neurons. The dark material at the junction of the nerve and the epithelium is eye pigment. (B) 10-week regenerate. The dark band a t the base of the sensory pad is caused by tissue folding. (C) 24-week regenerate. (D) Intact tentacle. The scale bar is 250fim and applies to all micrographs.

example, a ganglion is evident in the 24-week specimen of Figure 3(C), but there is no obvious ganglion in the illustrated specimens with shorter postlesion survival times [4 weeks, Fig. 3(A); 10 weeks, Fig. 3(B)]. Even in some specimens with regeneration periods as long as 22 weeks, ganglia were not evident.

At early stages of regeneration, the nerve fiber structures in the tentacle are highly abnormal [Figs. 3(A) and 3(B)]. Fiber diameters are unusually thick, with some fibers appearing two to three times as large as any seen in intact tentacles. Some fibers have a beaded shape. Probably the most striking characteristic of early regenerative stages is the chaotic arrangement of the fiber pathways, in which there is not even a suggestion of the normal digit structure. Since the fiber systems do eventually become more orderly, there is evidently a selective loss of misdirected fibers, coupled, perhaps, with the bundling of existing fibers.

Nerve cell bodies do not stain with HRP before about 6 weeks postlesion.

REGENERATED SNAIL TENTACLES 47

Prior to this time, there are cells in the tentacular periphery which evidence endogenous peroxidase activity, but they are not nerve cells [Fig. 3(A)]. The small neurons, the ganglion cells and the sensory cells, appear first, with the lateral cells, gamma cells, and delta cells only becoming evident in fills after about 17 weeks postlesion. We do not know whether the lack of the larger cells at earlier stages represents an absence of the cells or simply an inability to fill them.

The absolute number of cells of all types increases with the period of regen- eration. This was established quantitatively for the sensory neurons, whose endings can be clearly seen and individually counted in wholemount preparations viewed from the external face of the sensory pad (see Fig. 11 in Chase and Kamil, 1983). Figure 4 shows three representations of epithelial innervation by receptor neurons: a t 11 weeks postlesion, a t 16 weeks, and in an intact tentacle. Ob- viously, the density of endings is less in the regenerated tentacles than in the intact one. It can also be noted that the regenerated tentacles preserve, indeed accentuate, the gradient of receptor density in which the greatest number of endings is found near the dorsal side of the pad and a lesser number towards the ventral side. When the number of labelled sensory endings is plotted as a function of regeneration time (Fig. 5) it is seen that the number of endings in- creases rapidly after about 15 weeks postlesion, and the full complement of endings is reached at approximately 20 weeks postlesion. Again, we cannot say whether unlabelled endings are absent or just not filled.

In comparison with the olfactory sensory structures, we observed that regen- eration of the visual system is neither as reliable nor as accurate. Some regen- erate tentacles have no eye, others have multiple eye spots. The location of the regenerated eye is also variable and often abnormal. Occasionally, regenerated eye spots are even embedded in the tentacle ganglion. It is especially curious that the regenerated eyes never possess a discernable optic nerve. Either the afferent visual fibers fail to regenerate, or they travel in the tentacular (olfactory) nerve.

A. 1 1 Weeks 6. 16 Weeks

C. Intact Dorsal

0.5 mm

Ventral

Fig. 4. Camera lucida drawings of the distribution of sensory endings on the surface of the sensory epithelium, derived from tentacles that were backfilled with HRP and viewed in wholemounts. The dorsal/ventral designations and the scale apply to all drawings.

48 CHASE AND KAMIL

25001

3 ml 2000

w

* In tac t

L A I I I I I I t / 10 12 14 16 18

Regenera t ion P e r i o d ( W e e k s )

Fig. 5. Number of labelled sensory endings in regenerated and intact tentacles.

Odor sensitivity

Two factors constrained the use of the tentacular olfactometer and limited its application to animals with 10-week regenerated tentacles. First, it was necessary for the tentacles to extend a minimum of 8 mm into the olfactometer; tentacles at regeneration times less than 10 weeks were too short to be tested. Second, the testing protocol required the snails to remain active, and moving, during a continuous period of several hours; animals at long regeneration times had reached a chronological age at which this demand became unreasonable.

Six animals were tested, each with one intact tentacle and one 10-week re- generated tentacle. The results, shown in Figure 6, indicate that there is no significant differences in sensitivity between the intact and regenerated tentacles. The threshold sensitivity is identical, within the error of measurement, and the relationship between concentration and response frequency is also the same.

’I:] Regenerated tentacle (10 weeks) - T 1

C O N C E N T R A T I O N OF AMYL A C E T A T E ( log f rac t ion of s a t u r a t e d v a p o r )

Fig. 6. Odor sensitivity of intact and regenerated small tentacles. Bars represent means plus SE for six tentacles. Each tentacle was presented with odor on a total of 11 trials at each concen- tration while the opposite tentacle was presented with an equivalent flow of clean air during the same time. Intact and regenerated tentacles were tested in the same animals, on alternate trials, 10 weeks after the tentacular lesion. In the trials marked “clean air,” both tentacles were presented with clean air. Ipsilateral turns indicate a tropotactic orientation response (additional methods and inter- pretation in Chase, 1982).

REGENERATED SNAIL TENTACLES 49

The performance of animals with one regenerated tentacle and one intact tentacle (Fig. 6) can be compared with the performance of unoperated animals tested under identical conditions [reported in Fig. 4(A) of Chase, 19821. Cu- riously, although the threshold sensitivities are the same, the rate of responding elicited by stimulation of either tentacle in the operated animals is consistently higher, a t all odor concentrations, than for the unoperated animals. Even in the clean air (control) trials, the difference (28.1% vs. 12.5%) is statistically sig- nificant (t-test; p <0.5). Comparison of these experiments therefore leads to the conclusion that the frequency of spontaneous turning is greater in animals which have suffered a tentacular lesion. This in turn suggests that the lesion, even after regeneration, induces an instability in the bilateral balance of exci- tation (for a discussion of locomotor control mechanisms, see Chase, 1982).

DISCUSS I0 N

The main finding in this study is that regenerated snail tentacles possess normal or nearly normal structure and function in the olfactory sensory system. With respect to morphology, we found, for example, that the epithelial sensory pad returns to its normal size and shape. Further, the HRP experiments dem- onstrated the regeneration of nerve cells which were normal in both morphology and number. Indeed, the exactitude of the observed neuronal regeneration is remarkable, especially since the tentacle ganglion can be considered as a part of the central nervous system. Previous authors have taken this latter view because the tentacle ganglion and the conventional CNS have a common em- bryological origin in the ectodermal cephalic plate (Samassa, 1894; Hanstrom, 1925; ChQtail, 1963). It has recently been demonstrated that the molluscan cerebral ganglia also share with the tentacle ganglion the capability of regener- ation (Price, 1977; Moffett, 1980).

The evaluation of functional competence involved a test of odor detection in a procedure demanding olfactory orientation behavior. It was found that the regenerated tentacles were as sensitive as the intact tentacles, and equally capable of mediating the behavioral response.

I t can be noted that the restoration of function, as determined 10 weeks postlesion, occurs prior to the complete regeneration of neuronal elements, as revealed by HRP backfilling. This conclusion is drawn from a comparison of the behavioral results (Fig. 6) with the pattern of HRP staining at 10 weeks [Fig. 3(B)] and the quantification of sensory endings at about the same period [Figs. 4(A) and 5)]. In fact, the tentacle illustrated in Figure 3(B) is one of those tested in the olfactometer and included in the data of Figure 6. Despite the poorly developed and highly abnormal appearance of its neural structures, this tentacle showed only a slight behavioral deficit relative to the contralateral intact tentacle, and it was clearly responsive to odors in a concentration dependent manner.

There are several possible explanations for the apparent discrepancy between the behavioral and morphological results. For one, the HRP backfill method may result in incomplete staining when applied to tentacles in the early stages of regeneration. This would produce an underestimation of the number of neurons present. It could not, however, account for the abnormal appearance of individual nerve fibers and their chaotic organization. Moreover, the fact that the size of the sensory pad is still below normal at 10 weeks (Fig. 1) suggests that indeed neuronal regeneration is likewise still incomplete. One must therefore consider the possibility that some of the neuronal structures which are

50 CHASE AND KAMIL

normally present are redundant or required only for the control of behavioral capacities other than those tested here. An inference of a similar nature was recently drawn by Wright and Harding (1982) in respect to their experiments with mice, Animals which had experienced bilateral removal of the olfactory bulb were nevertheless able to perform tasks that required odor discrimination, orientation, and olfactory memory. The performances were based on the for- mation of direct synaptic connections between regenerated primary olfactory neurons and the forebrain. Clearly, however, a behavioral deficit could probably be demonstrated in both the mouse study and our own provided that an appro- priate test were devised. So far as the snail is concerned, deficits in 10-week regenerate animals might be revealed with tests that required either odor dis- crimination, or responses to odors other than amyl acetate, or the detection of winds.

The authors thank Colombe Evanko and Sue Cocker for their assistance and Gerald Pollack for his critical review of the manuscript. This work was financially supported by the Natural Sciences and Engineering Research Council of Canada.

REFERENCES

CHASE, R. (1981). Electrical responses of snail tentacle ganglion to stimulation of the epithelium with wind and odors. Comp. Biochem. Physiol. 70A: 149-155.

CHASE, R. (1982). The olfactory sensitivity of snails, Achatina fulica. J . Comp. Physiol. 148: 225-235.

CHASE, R., and CROLL, R. P. (1981). Tentacular function in snail olfactory orientation. J . Comp. %ysiol. 143,357-362.

CHASE, R., and KAMIL, R. (1983). Neuronal elements in snail tentacles as revealed by HRP back- filling. J . Neurobiol. 14: 29-42.

C H ~ ~ T A I L , M. (1963). Etude de la regCnCration du tentacule occulaire chez un Arionidae (Arion rufus L.) et un Limacidae (Agriolimax agrestis L.). Arch. Anat . Microsc. Morphol. 52: 139- 203.

EAKIN, R. M., and FERLATTE, M. M. (1973). Studies on eye regeneration in a snail, Hrlix nspersu. J . Exp. 2001. 184: 81-96.

FLOHES, D. V., and PELLEGRINO L)E: IKALDI, A. (1980). Studies of the regeneration of sensory tentacular structures in a pulmonate snail. Reu. Can. R i d . 39: 209-217.

HANSTROM, B. (1925). Uber die Sogennaten Intelligenzspharen des Molluskengehirns und die Innervation des Tentakels von Helix. Acta 2001. Stockholm 6: 183-215.

MOFFAT, S (1980). Ultrastructure of regenerating central nervous connections and implanted ganglia in the snail Melampus. Neurosci. Abstr. 6: 386.

PRICE, C. H. (1977). Regeneration in the central nervous system of a pulmonate mollusc, Melampus. Cell Tissue Res. 180: 529-536.

SAMASSA, P. (1894). Uber die Nerven der augentragenden Fuhler von Helix pomatia. 2001. Jahrb. Abt . Anat.. 7: 593-608.

WRIGHT, J. W., and HARDING, J. W. (1982). Recovery of olfactory function after bilateral bulb- ectomy. Science 216: 322-324.