The Effects of A9-Tetrahydrocannabinol on Action Potentials in the Mollusc Aplysia

Depcrrbm~rat c,fBic~lngy, 2CfcGill University, Morztreal, Qurhec H3C3GI Received February 20, 1975

~ c o s ~ n - U ~ ~ v r 1 ~ 1 , J . , and CHASL. K. 8972. The effects of A9-tetrahydrocannabinol on action potentials in the mollusc Aplysda. Can. J. Physiol. Pharmacol. 53,793-798.

The effects of A9-tetrahydr~~cannabinol (A9-THC) on action potent~al~ were examined during intrason~atic recordings from the isolated buccal and parieto-visceral ganglia of A p l y ~ i a califor- nica. When added to the saline solution bathing the preparation, the compound (in concentrations 1 0-4-10-5 M) caused areduction in spike overshoot (15-20% oftotal amplitude) and increased the lability of responses to electrical stimulation. The somatic membrane appeared to be more affected than the axonal membrane. Diffusion barriers in the ganglion probably account for the high degree of variability in drug response. such that both of the characteristic changes were observed in only about 30% of the te\ts. This i$ the first report to descl-ibe effects of 2,'-THC on invertebrate neurones. The results indicate that A'--THC causes a depression in nerve ce81 excitability. and thebe data are comistent with reported effectsof THC compounds in mammals.

ACOS.KA-URQ~J~D~, J. et CHASE, R. 19'75. The effects of b9-tetrahydrocannabinol on action potentials in the mollusc Apfys ia . Can. J. Physiol. Pharmacol. 53,793-798.

Les efTets du A'-tetrahydrocannabind (A9-THC) sur les potentiels d'action ont ete examines par enregistremenat intrasomatique effectues sur les ganglions isoles buccaux et parieto- visceraux dlAplysiu cmljfbrnica. 1,orsqu'on ajoerte la solution saline dans laquelle bnigne lapreparation le composC ades concentrations de 10--4-10-' M cpn obtient Line riductionde I'over- shoot de potentiels d'action (de 15-20% de l'amplitude totale) et une augmentation du nombre d'essais w'enti-ainant pas de reponse lors de stimulations ilectriques. La membrane somatique semble plus affectee que Ia membrane axonale. Des barritres a diffusion b I'intkrieur du ganglion sont probablement responsables pour le haut degre de variabilite de reponse i la stimulation chimique de svrte que ies deux modifications caracteristiques n'ont ete observke quc dans 30% environ de ces tests. Ceci est le premier rapport decrivant I'effet du A9-THC sur les neurones d'invertebres. P,es rksu8tats indiquent que le A9-THC entrnine une depm-essionde I'excitabilite des ceilules nerveuses. Ils ssnt coherents avec ceux observes chez les rnan~miferes.

Introduction Although there is now a large body of litera-

ture describing the pharmacology of marijuana, it is not yet known how the active derivatives of Clrnnubi,~ affect the nervous system to pro- duce the characteristic behavioral and psychs- logical effects of marijuana intoxication. Since technical complexities make difficult the ncces- sary analysis in mammalian brains, particularly at the intracellular level, we investigated whether a familiar molluscan nervous system might serve as a model for the study of this

effects of the principal active component of marijuana, a"-tetrahydrocannabinol ( A ~ T H C ) , on nervous function in the marine mollusc Aplysiu. The present report is based on a larger study (Acosta-Urquidi 1974) in which the effects of thc drug were evaluated with respect to the electrical properties of nerve cell mem- branes, synaptic transmission and the genesis of action potentials. In that study, drug actions werc most consistently recorded in relation to action potentials, and it is these results which are reported here.

The usefulness of invertebiate ner- vous preparations for pharmacological investi- Methods

Specimens of Aplysiu calijornicu were obtained gations has become evident in s'v'"' recent from pacific ~ i ~ ~ ~ ~ i ~ ~ h p p l y Co., Venice, Calif, articles (r.8. ~halazonitis 1967; Fau@er and Either the parieta-visceral ganglion or the paired WilIows 1973; Chase 1975). We studied Ihe buccal ganglia were removed from the animal and

pinned to a layer of Sylgard epoxy resin in a lucite 'Present address: Bepartnlent of Zoology, Univer- chamber. The preparation was bathed in a saline

sity of Toronto, Toronto, Ontario M5S 1A8. solution which contained 394 rnM NaC1. 1 1 m M gTo whom reprint requests sho~ild be addressed. KCI, 19 mM MgCls, 11 mM CCL, 30 mA4 MgSO,,

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.

794 CAN. J . PHYSIOL. PHARMACOL. VOL. 53, 1975

and 10 mM Tris base, buffered to pH 7.7. The experi- ments were carried out at rooin temperature (22- 24 "6). All recordings were intrasomatic, using glass microelectrodes filled with 3.0 M KC1 and having resistances between 5-15 Ma, measured in the saline solution. The a~nplification and display of transmem- brane potentials was conventional. Often two elec- trodes were inserted into a cell, one for recording and one for stin~~llation (called 'direct stimulation'). Poly- ethylene suction electrodes were attached to appro- priate nerve trunks for purposes of antidromic stimu- lation. Stimulation parameters (intensity, duration, and frequency) were set at values which consistently elicited a spike upon each shock before application of A"-THC. These settings were maintained during testing of drug effects, except when measurements of threshold were obtained during direct stimulation, in which case stimulation current was adjusted to thresh- old levels.

With respect to spike initiation by either antidromic or direct stimulation, a 'test' of drug action (Table 1 ) is defined as an identical series of stimuli, typically consisting of 10 trials, presented to one cell before and after delivery of a single collcentration of AS-THC. A test of action potentials during spontane- ous activity consisted of equivalent periods of obser- vation (1-15 min) on one cell before and after delivery of a single concentration of A"-THC. On occasion, when a cell showed no response to a par- ticular drug concentration during 45 n ~ i n of observa- tion, or if complete recovery of drug effects was ob- served, the preparation was washed thoroughly (see below) and then another (usually higher) drug con- centration was tested in the same manner. While it is possible that the procedure of repeated testing on a single cell created cumulative drug actions, even despite washing. we have no evidence that this in fact occurred. The study is based on a sample of 72 re- corded cells, derived from about 30 animals.

Two forms of A"-THC were employed, both ob- tained through Health and Welfare Canada. AS-THC dissolved in ethanol was prepared by the United States National Institute of Mental Health and de- scribed as 95% pure. A water soluble ester derivative of AKTMC, r-morpholinobutyric acid hydrobromide, compound SP-lI lA, was prepared by the Sheehan Institute, Cambridge, Mass. (Zitko et a/. 1972); Brady and Carbone (1973) reported at least 95% purity for this compound, but we have no direct assay of our sample. The AtTHC in ethanol was made soluble in saline by the addition of a small quantity of the surfactant agent Pluronic F-68 (Wyandotte Chemicals Corp., Wyandotte, Mich.) according to the method described by Byck and Ritchie (1973). The SP-11 1A was first dissolved in a few drops of distilled water by gentle warming. followed by shaking. Once it had obtained am opales- cent color, the solution was diluted in saline to the desired concentration. The drugs were delivered to the preparation by replacing the saline in the record- ing chamber (volume of 3.0 ml) by a saline solution containing the dissolved drug. Similarly, the drug was removed by perfusion with normal saline. In general, once an adequate replacement had been completed

(approximately 30 times the volume of the chamber), perfusion was not continued during periods of re- cording.

Control experiments were performed with the car- rier vehicles for both forms of the drug and no physiological effects were noted. The final alcohol concentration in the one type of 3"-THC solution amounted to approximately 0.05% for a AtTHC concentration of 3 X lo-" M, but this is well below the critical level for detectable physiological effects in both the present preparation and a closely related one (Chase 1975). While the possibility of an ethanol-THC synergism cannot be excluded, this would have had little consequence in terms of our total results, since over 85% of the experiments employed the SP-111A solution. Because no differ- ences were detected in the effects of the two forms of AS-THC, they will not be distinguished in the de- scription of the results.

Results In concentrations as low as 4.5 x 18-W,

AS-THC reduced the overshoot amplitude of action potentials and increased the lability of evoked spikes. Since all stimuli were supra- threshold before application of ARTPIC, an increase in lability was defined as response failure on more than 10% of the test trials dur- ing the presence of the drug. These effects were apparent for spikes initiated either 'spontane- ously', or by direct or antidromic stimulation. The results are summarized in Table 1.

The increase in lability is interpreted as a change in excitability because the response to direct somatic stimulation could be re-estab- lished by increasing the stimulus strength. The levels of depolarization required to produce a spike by direct stimulation were seen to in- crease after exposure to AKTHC, in one case by as inuch as 25 mV coinpared to the control level. In three cases a decrease in threshold was observed.

Application of the drug was often followed by changes in the resting membrane potential, but these were small and highly variable (Acosta-Urquidi 19'74). From a total of 72 cells, a depolarization was observed in 20 cells (range 2-14 mV) and a hyperpolarization in 27 cells (range 2-8 mV). These alterations in membrane potential were not correlated with the observsd increases in lability.

The effects were typically observed aftor a latency of 5-1 5 min, however, the total range of latencies was 1-30 min, with both sf tBac extreme values coming f ron~ experiments in

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.

ACOST.4-URQCTIDI AND CHASE: A9-TETRAHYDROCANNABINOL EFFECTS

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.

796 CAN. J . PHYSIOL. PHARMACOL. VOL. 53, 1975

FIG. 1. Reversible changes in lability and spike amplitude following application of 7.3 X 10-' M 3"-THC. Responses were recorded in a 'follower cell' of the buccal ganglion during antidromic stimulation of the BN2 nerve (terminology from Gardner 1971). A, before application of A"-THC. B, 10 min after application. C, 35 min after application. D, 5.5 h after Mb-THC was removed from the preparation by saline wash. Each photograph shows 10 superimposed responses obtained from a series of I0 stimuli. Stim- ulation consisted of 50 V, 0.3 ms pulses delivered through a suction electrode at 0.5 Hz. Calibration is 20 mV, 5 ms.

FIG. 2. Effects of 1''-THC on the axon spike and the sonla spike during antidromic stimulation of the giant cell RI. The right connective nerve was stim- ulated through a suction electrode. Each photographic group consists of two successive responses super- impored. A, before application of A"-THC. The first spike in each response is an antidromic spike and the two are exactly superimposed; an arrow marks the point of inflection between the axon spike and the soma spike. The delayed spikes are efferent spikes triggered by the antidron~ie discharge (for a descrip- tion of this phenomenon see Tauc 1962). B, spikes from left to right were evoked, respectively, 5 and I S rnin after application of 5.4 X 10 % Mu-TKC. Note the separation between the axon spikes and the soma spikes; also. the second sonla spike is depressed. C, 2 h after saline wash. The soma spike is absent from both responses despite an increase in stimulus voltage. Calibration is 20 mV, 20 ms.

which high drug concentrations (5 x M ) were employed. When affected by the drug, action potential overshoot usually declined by 15-20% of the total spike amplitude after a single treatment. Repeated exposure to A"-THC often lccH to progressive seduction of the action potential (up to 50%) but the amplitude stopped declining following a wash with drug- free saline.

A range of A"THC concentrations from 1.6 x 18- W to 7.3 x M was employed, and while some dose-response dependency was noted, both with respect to latencies and substantive effects, the relationship was not consistent.

When the h9-THC was washed out and the cell's behavior was subsequently monitored for at least 1 h, complete recovery of spike ampli- tude and/or responsiveness was observed in approximately one third of the cases (see, for example, Fig. 1) . Recovery was judged to be partial in another third of the cases, and in the remaining third of the experiments recovery was never evident. The longest observation period without apparent recovery was 6 h after the wash. If a cell's resting membrane potential deviated more than 10 mV from pre-drug levels, its ultimate recovery from A"-THC effects, or lack thereof, was considered irrele- vant to the account of recovery given above. This occurred in less than 10% of the record- ings. There was no clear relationship between drug concentration and degree of recovery.

During antidromic stimulation we saw evi- dence that the drug caused conduction block in the initial regions of the axon or at the axon-soma junction. The cell body is excitablc in Aplysia but it has the highest threshold of all parts of the neuron. The transition zone be- tween the soma and axon is also an area in which spike conduction commonly blocks (Tauc 1962). During intrasomatic recording the axon spike (A spike) is distinguishable from the soma spike (S spike) either by an iisflection on the rising phase of the action potential or by a definite temporal separation of the two events, as was originally shown for the giant cell of the parieto-visceral ganglion (R" by Tauc ( 1962). In the prcsermt experi- ments, we found that application of A!'-THC often selectively reduced the amplitude of the S spike and delayed its appearance, indicating reduced somatic cxcitability or partial coi~duc- tion block. This was most frequently seen dur- ing antidrornic stimulation of R2, as illustrated in Fig. 2. Occasionally the S spike was com- pletely absent from the antidromic response (Fig. 2C), indicating conduction block in the axon or at the axon-soma junction.

Table 1 shows that apglicatio~~ of A!~-THC was not always effective in prodtecing the re- sults described above. Often thcre was a reduc- tion in spike amplitude with no increase in lability, or a lability change but no amplitude change, or in approximately one third of the tests, no observable drug effect on either of these measures. In an attempt to accoiant for

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.

ACOSTA-URQUIDI AND CHASE: A9-TETRAHYDROCANNABINOL EFFECTS 797

this variability, we examined the possibility that different cells might vary in their sensitivity to the drug. We therefore selected about ten particular cells for repeated observation, using for these purposes the criteria of Frazier et al. ( 1967) for the parieto-visceral ganglion and the map of Gardner (1971) for the buccal ganglion. We found, however, that there were no consistent differences between cells in their response to the drug. In fact, the giant cell Rz was tested over 30 times and showed a vari- ability of response equivalent to that of the entire population. Furthermore, cells in the buccal ganglia had the same average sensitivity as cells in the parieto-visceral ganglion.

Discussion These results represent the first description

of pharmacological activity of As-tetrahydro- cannabinol on nervous function in invertebrates. The principal effects were a reduction in the amplitude of action potentials and an increase in the lability of spike conduction. Two earlier publications reported an absence of effects when AKTHC was tested on isolated squid giant axons or lobster nerves (Brady and Car- bone 1973; Byck and Ritchie 1973). Possible reasons for the discrepancy are discussed be- low. We also point out that the effects we ob- served were in most instances minor and they were highly variable; and, of course, we dealt with a different species.

We have attributed the increase in lability to hypoexcitability because, when stimulating the soma directly through a microelectrode, response reliability could be re-established by increasing stimulus strength. We do not con- sider hyperpolarization as a probable cause of increased lability because hyperpolarizations were slight and were not correlated with in- creases in lability. Lability in our experiments would therefore appear to be due to elevated thresholds. Ths explanation for reduced spike amplitudes is not so clear. When spikes orig- inate in the axon, either spontaneously or dur- ing antidromic stimulation, amplitude reduc- tion could be due to incomplete invasion of the soma following increased thresholds, as in Fig. 2B. Even with direct stin~ulation of the soma, action potentials will usually originate in the axon, and are therefore vulnerable to conduc- tion block near the soma, except when exccp-

tionally strong stimuli are employed (Tauc 1962). If spikes are initiated in the soma during direct injection of current, it is unlikely that portions of the somatic membrane would remain unexcited. However, since we are not certain about the site of initiation during direct stimulation, we can only conclude that spike amplitude reduction is due to the depression of electrically excitable ionic channels, with threshold elevation perhaps a contributing factor.

It was disconcerting to find that the drug effects were not consistently reproducible, even when studying the same identifiable cell. In this regard, it is worth noting that other studies of drug actions in invertebrate preparations have also reported a high degree of variability (Chalazonitis 1967; Faugier and Nrillows 1973). In her study on the pharmacology of volatile anesthetics, Chalazonitis found, as we did, that the somata of Aplysia neurones were more sensitive than were the axons. Consistent with this, we failed to observe any depressive effect of axTHC on synaptic transmission (Acosta- Urquidi 1974). Since molluscan synapses are made on the axon, active responses of the somatic membrane are not involved in synaptic phenomena.

There are several possible explanations for the relatively high sensitivity of the soma. First, as we have already mentioned, the somatic membrane is the least excitable portion of the neurone (Tauc 1962) and may therefore be more affected than other regions by a given reduction in excitability. Secondly, somatic action potentials in Aplysia apparently depend more on early inward calcium currents than do axon spikes, which are sodium dependent (Kado 1973), and it is possible that A~-THC has selective effects on these currents. Indeed, we have some cvidence that calcium spikes are more sensitive to AX-TMC than sodium spikes ( Acosta-Urquidi 1974). Finally, invertebrate ganglia are so organized that thc somata lie in closest physical proximity to the bathing solution. Thc neuropile is located in the centre of the ganglion beneath overlying glia, neuronal somata, and connective tissue. Thus, in con- sideration of the problems of diffusion, drugs delivered in the bathing solution are likely to reach the somatic membrane more rapidly and in greater concentration than the axonal mem-

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.

798 CAN. J . PHYSIOL. PHARMACOL. VQL. 53, 1975

brane. The entire ganglion is also enveloped in a fibrous connective sheath which represents an additional barrier to diffusion. When this sheath was removed by microdissection we usually observed more pronounced effects with a shorter latency. A further impediment to drug diffusion are the external laminae (basement membranes) that coat the cells embedded within the sheath (Coggeshall 1967). These membranes could not be removed by dissec- tion. Because of the presence of the various tissues described, the AS-THC may have had difficulty penetrating to potentially vulnerable plasma membranes. It is likely that the irreg- ular nature of these diffusion barriers contri- buted to the variability in drug responses by limiting in a random fashion the attainment of a critical drug concentration at the plasma membrane.

Although effects of a"THC have not been previously reported in invertebrates, Brady and Carbone (1973) found that the oxidized metabolite of A"-THC, 1 1 -hydroxy-A"-tetra- hydrocannabinol, irreversibly blocked impulse conduction in squid giant axons. While similar to the effects reported here, their results were more substantial and more reliable. It would therefore appear that isolated invertebrate ner- vous systems are unable to metabolize a9-THC, or have only a limited capacity to do so, and that the hydroxylated cannabinol compound has a more potent pharmacological action on nervous tissue. Unfortunately, since a sample of 1 I-BH-A"-THC was not available, we were unable to compare its potency with that of AKTHC. However, if we assume that AS-THC has only a weak effect on invertebrate neurones, then it is perhaps significant that in those ex- periments which failed to reveal any effect, the preparations were isolated axons which pre- sumably did not include regions of low safety factor. In our experiments, the relatively inex- citable soma appeared to be most affected by the drug.

The present results are consistent with ear- lier descriptions of depressed neural excitability following application of AKTHC or other cannabinoid compounds. In addition to the squid nerve study cited above, Byck and Ritchie (1973) found a reduction in the size of the compound action potential in rabbit nonmyelinated peripheral nenre using A'-THC concentrations as low as 3 pM. The anti-con-

vulsant properties of cannabinoid drugs in mammalian brains have been frequently re- ported (e-g . Loewe and Goodman 1947; Wada et al. 1973; Karler et al. 1974).

The authors are grateful to Lorraine Pawson for technical assistance. This research was supported by a grant from the National Research Council of Can- ada to R. Chase. J. Acosta-Urquidi was the recipient of a predoctoral fellowship from the Conseja Na- cional de Ciencia y Technologia. Mexico.

ACOS~A-URQUIDI, J. 1974. The effects of A9-THC on nervous function in Aplysia. M.Sc. Thesis. MeGill University, Montreal. Que.

BRADY, R. O., and CARBONE, E. 1973. Comparison of the effect? of A9-tetrahydrocannabinol.ll-hydroxy-A9-tet- rahydrocannabinol, and ethanol on the electrophys- iological activity of the giant axon of the squid. Neu- ropharmacology, 12,601-605.

BYCK, R., and RITGHIE, J. M. 1973. A9-Tetl-ahydro- cannabinol: effects on mainmalian nonmyelinated nerve fibers. Science, 180,84-85.

CHALAZONITIS. N . 1967. Selective actions of volatile anesthetics on synaptic transmission and auto- rhythmicity in single identifiable neurons. Anes- thesiology, 28, 11 1-122.

CHASE, R. 1975. The suppression of excitatory synaptic responses by ethyl alcohol in the nudibranch mollusc, Tritonia diornediu. Comp. Biochem. Physiol. 50C, 37-40.

COGGESHALL, R. E. 1967. A light and electron microscope study of the abdominal ganglionofAplysia culifornicu. J. Neurophysiol. 30, 1263-1287.

FAUGIEH, S., and WILI_OWS, A. 0. D. 1973. Behavioral and nerve cell membrane effects of an epileptic agent (metrazol) in a mollusk. Brain Res. 52, 243-260.

FRAZIER, W. T.,KANDEI,, E. R.,KUPFERMANN,I., WALIRI, R.. and COGGESHALL, R. E. 1967. Morphological and functional properties of identified neurons in the abdominal ganglion of Aply sia californicca. J. Neurophysiol. 30, 1288-135 1.

GARDNER, D. 1971. Bilateral symmetry and interneuronal organization in the buccal ganglia of Aplysia . Science, 173,550-553.

KADO, R. T. 1973. Aplysica giant cell: soma-axon voltage clamp current differences. Science, 182,843-845.

KARI.ER, R.. CELY, W. and TURKANIS, S . A. 1974. Anticon- vulsant properties of A9-tetrahydrocannabinol and other cannabinoids. Life Sci. 15,931-947.

LOEWE, S., and GOODMAN, L. S. 1947. Anticonvulsant action of marihuana-active substances. Fed. Proc. 6, 352.

TAUC. L. 1962. Site of origin and propagation of spike in the giant neuron of Aplysia. J. Gen. Physiol. 45, 1077- 1097.

WADA, J. A., SATA, M., and CORCOR~N, M. E. 1973. Antiepileptic properties of A9-tetrahydrocannabinol. Exp. Neurol. 39, 157-165.

ZITKO, B. A., HOWES, J. F., RAZDAN, R. K.. DAI~ZELL. B. C.. DALZEI-L, H. C., SHEEHAN, J . C., PARS, H. G., DEWEY, W. L., and HARRIS, L. S. 1972. Water-soluble derivatives of A'-tetrahydrocannabinol. Science, 177, 442-444.

Can

. J. P

hysi

ol. P

harm

acol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV W

IND

SOR

on

11/1

5/14

For

pers

onal

use

onl

y.