Pharmacological activity of three fractions obtained by smoking cannabis through a water pipe


Material and methods


Pages: 49 to 56
Creation Date: 1976/01/01

Pharmacological activity of three fractions obtained by smoking cannabis through a water pipe

Jomar CUNHA, 2
E. A. CARLINI, Head, Department of Psychobiology, Escola Paulista de Medicina 04023 Sao Paulo, Brazil, and
T.A. KEPHALAS Department of Biological Chemistry, University of Athens Medical School Goudi, Athens, 609, Greece


In a collaborative work, six chemical fractions obtained from smoke of cannabis, prepared by the Department of Biological Chemistry, University of Athens, were assayed pharmacologically at the Department of Psychobiology, Escola Paulista de Medicina, Brazil. The chemical fractions, obtained through burning cannabis mixed with a special tobacco through a water pipe similar to that used by Greek smokers, were: fraction Is composed of water-soluble compounds, fraction IIs made up of water-insoluble substances that condensed at the top of the water phase due to their low solubility; and fraction IIIs, of water-insoluble highly volatile substances, corresponding to the part that reaches the human lungs. Fractions IB, IIB and IIIB were placebo fractions obtained from burning only the special tobacco, called "tubeki", which is mixed with cannabis by the Greek smokers.

The six fractions were assayed pharmacologically, in a blind fashion, using the following methods: barbiturate sleeping time in mice, catatonia in mice, corneal reflex in rabbits (Gayer test), and spontaneous motor activity in mice. Fraction IS, and its placebo fraction IB, were inactive in all methods employed, revealing no water-soluble active substances in cannabis. Fraction IIS revealed to be the most active as it induced significant effects in all 4 methods used; its IIB placebo fraction was inactive. Fraction IIIS, composed of the only substances that reach the human lungs, was moderately active; it produced significant effects on barbiturate sleeping time of mice and on the corneal reflex of rabbits, but was inactive in inducing catatonia and in decreasing spontaneous motor activity of mice.

These results indicate that devices such as the water pipe used by Greek smokers retain active principles from cannabis.

* * *

1 Permanent address: Department of Biological Chemistry, University of Athens Medical School, Goudi, Athens, 609, Greece.

2 Through the Central de Medicamentos (CEME), Brazil, working full time at Escola Paulista de Medicina.

Recent reports state that the activity of the cannabis plant cannot be explained solely on the basis of its (-) delta-9- trans-tetrahydrocannabinol (delta-9-THC) content (Isbell, 1971; Karniol and Carlini, 1972; Kubena and Barry, 1972; Galanter et al., 1973; Carlini et al., 1974). These comparative studies have shown that in laboratory animals and in human beings the extract or the smoke obtained from the plant induce stronger effects than those obtained by an amount of pure delta-9-THC which is equal to that present in the crude material; furthermore, qualitative differences in the effects have also been reported (Karniol et al., 1974; Volavka et al., 1974). The situation is obviously further complicated by the way in which cannabis is smoked in different places. For example, during the meeting on the chemistry of cannabis smoke, organized by the United Nations Narcotics Laboratory in Greece, in April 1975, two participants (T. A. Kephalas and E. A. Carlini) had the opportunity to observe and discuss the way in which marihuana is usually smoked in their respective countries. In contrast to the simple method used in Brazil-packing the dried plant in a loose cigarette and smoking it directly-a much more elaborate procedure is employed in Greece. The material, usually hashish, wrapped in water-proof paper, is further rolled in a piece of newspaper which is immersed in water and then ironed (90°) several times. The thin slab of cannabis thus obtained together with a special variety of tobacco, is put into a water pipe, the head of which is made of a fresh potato. The mixture is burnt and the smoke bubbles through a water container before reaching the lungs.

These data indicate that more research is needed on the chemistry and pharmacology of cannabis smoke obtained by the use of different procedures. The present paper represents the collaborative work of two institutions; several fractions obtained from cannabis burnt according to the way the plant is smoked in Greece were prepared at the University of Athens, Greece; pharmacological assays of these fractions were undertaken at the Escola Paulista de Medicina, Brazil.

Material and methods


The experimental model refers to the actual way of smoking cannabis by experienced Greek users, i.e. through a water pipe. The water pipe device employed (fig. 1) is almost identical to that used in Greece. The pipe head is made of a fresh potato and the volume of the water, 100 ml, is changed after each use.

The experiments were carried out with a United Nations reference cannabis sample (UNC 351), for comparative purposes. Before burning it in the experimental device, the sample was subjected to a special treatment which is the same as used by the smokers. Ten grammes of the sample were wrapped in a water-proof paper envelope and further rolled in a piece of newspaper. The small packet so formed was immersed in water and then ironed several times with a hot (90°C) electric iron until the newspaper wrapping was almost dry. In this way a thin slab of cannabis was obtained. This slab of cannabis is smoked mixed with a special variety of tobacco known as "tubeki". Similarly, dried leaves of "tubeki" were ground, washed repeatedly with tap water and left to dry at room temperature for a couple of days.

FIGURE 1 Experimental model of smoking a mixture of cannabis and tobacco through a water-pipe. ( a ) Nut charcoal, ( b ) pipe head, ( c ) water-pipe device, ( d ) water, ( e ) double-coiled condenser, ( f ) smoking machine.

Full size image: 38 kB, FIGURE 1 Experimental model of smoking a mixture of cannabis and tobacco through a water-pipe

Three grammes of the cannabis slab and 4 g of "tubeki" were put in the pipe head, in alternate layers (usually three layers of tobacco, one of which always covered the top of the pipe head and usually two layers of cannabis).

Nut charcoal was used for burning; according to the smokers' experience it is the best burning agent because it leaves very little ash. Inevitably, a smoking machine was used (type ethel MK VII, tobacco smoking machine (fig. 1). The frequency and duration of puffs were regulated to one puff/min and 8 sec., respectively.

Under these experimental conditions three types of material were collected:

  1. Particles of the smoke which were electrostatically precipitated in the glass tubes of the machine and which comprise what is called "the sublimate" (fraction IIIS); the sublimate is normally inhaled by the smoker. Analytical work with this material (Kephalas et al., in press), revealed many known and/or unknown substances present together with the well-known cannabinoids, delta-9-tetrahydrocannabinol (delta-9-THC), cannabidiol (CBD) and cannabinol (CBN). The percentages of the main cannabinoids in this material were as follows: delta-9-THC 7.18 per cent, CBD 3.29 per cent and CBN 25.28 per cent;

  2. Water-soluble components of cannabis and tobacco, and/or their pyrolysis products which are also soluble in water. The above-mentioned material was then extracted with ether and after acidification of the remaining water phase, new extraction with ether was performed. The water phase was then lyophylized comprising fraction IS. Both ether extracts (neutral and acidic) as well as the remaining water phase did not contain known cannabinoids when analysed by means of TLC and GC; they were not assayed pharmacologically;

  3. Cannabis and tobacco components and/or their pyrolysis products which appear mainly in the walls of the water container and on top of the water as small particles (fraction IIS). There are insoluble in water but they remain there owing to their low volatility when the smoke is passed through the water. This material can be separated from the water-soluble one (fraction Is) by filtration through a filter paper (Whatman No. 50). Preliminary analysis of the above mentioned fraction indicated the following percentages of cannabinoids: delta-9-THC 2.37 per cent, CBD 0.89 per cent and CBN 38.6 per cent.

For reasons of comparison analogous types of material (IB, IIB, IIIB) have been obtained by smoking only "tubeki" tobacco.

Pharmacological assays on fraction IS - IIIS and IB - IIIB

Preparation of suspensions for the pharmacological study. All 6 fractions were received from Greece as dry materials in glasses labelled in such a way as not to permit their identification. Therefore, all assays were carried out in a blind fashion. The suspensions were prepared in saline-Tween 80 mixture (99:1 part:v/v); the latter solution alone served as a control solution.

Four experiments were carried out.

Pentobarbital sleeping time: Three to five months' old male albino mice were used. Groups of 10 animals were first injected i.p. with the control solution or the appropriate dose of the drug under study. One hour later 50 mg/kg of pentobarbital sodium was administered i.p. The interval of time between losing and recovering the righting reflex was recorded. This reflex is considered to be intact if the animal, placed on its side, recovers from this position within 1 minute; it is considered lost when the recovery requires a longer period. The total sleeping time is expressed in minutes. The maximum time of observation was 4 hr. Three comparisons were carried out:

  1. Control solution and 10 mg/kg of the fractions IB, IS, IIB, IIS, IIIB and IIIS.

  2. Control solution and 40 mg/kg of the same fractions.

  3. Control solution, 50 mg/kg of CBD and 80 mg/kg of the previously mentioned fractions.

Each of the 3 above-mentioned experiments were statistically analysed through one-way analysis of variance followed by Duncan's new multiple range test.

Catatonia test. Three to five months' old male albino mice were used. Thirty min. after intraperitoneal injections, the mice (10 per group) were placed with the forepaws onto a horizontal rod at 4.5 cm from the floor. This was done 3 times at 20-min. intervals each, up to 2 hr. The total times the animals remained in this position were calculated and analysed statistically through one-way analysis of variance followed by Duncan's new multiple range test.

The doses used in this experiment were: 20 mg/kg for IB, IS, IIB, IIS, IIIB and IIIS and 10 mg/kg for delta-9-THC.

Spontaneous motor activity. Three to five months' old male albino mice were i.p. injected with three different doses from each fraction (5, 10 and 20 mg/kg). Thirty minutes later they were introduced into cages provided with 3 photocells. The number of light beam interruptions were cumulatively recorded every 15 min. during 1 hr. Nine to ten animals were used per each dose and fraction.

One-way analysis of variance followed by Duncan's new multiple range test were performed only with the results obtained after the last measurement (60 min.)

Corneal areaflexia in rabbits (Gayer test). The corneal reflex was elicited 10 times in each eye, every 3 min., with the help of a hair. The lack of at least 80 per cent of the responses was taken as a positive assay. The fractions IB, IS, IIB, IIS, IIIB and IIIS, suspended at a concentration of 2.0 mg/ml, and delta-9-THC, at a concentration of 0.5 mg/ml, were injected through the marginal vein of the ear, at a constant speed of 0.04 ml per min. Five animals were used for each fraction.


Pentobarbital sleeping time. Fig. 2 shows that only the fractions IIS, IIIB and IIIS at a dose of 80 mg/kg, significantly increased the pentobarbital sleeping time when compared to control animals. Also IIS and IIIS significantly increased the pentobarbital sleeping time when compared to their corresponding placebo fractions, IIB and IIIB.


Full size image: 18 kB, FIGURE 2

Effects of placebo fractions (IB, IIB and IIIB) and of cannabis fractions (IS, IIS, IIIS) and CBD on the pentobarbital sleeping time of mice. The columns represent the percentage of increase or decrease of the means as compared to the control values (zero per cent). The control sleeping times in the 3 comparisons were, respectively, 50.7±22.2, 31.5±15.7 and 50.7±17 min. (mean ± S.D.). The column of CBD represents the percentage of increase of the median under the control value. The doses (in mg/kg) used for the experiment are represented above or below the columns. Asterisks indicate statistically significant differences from the respective control groups (Duncan's new multiple range test; 1 per cent level). § indicate statistically significant differences (p x 0.05) from the corresponding cannabis placebo IB, IIB, IIIB).

The animals injected with 50 mg/kg CBD were sleeping longer than 4 hr. This means that CBD at this dose is much more active in potentiating pentobarbital sleeping time than the fractions IIS, IIIB and IIIS.

Larger doses of fractions IIIB and IIIS revealed to be toxic as the injected animals died within 4 min. after administration of 160 mg/kg. The deaths were preceded by tonic-clonic convulsions. The other 4 fractions caused no deaths at this dosage level.

Catatonia. Fig 3 shows that only IIS and delta-9-THC produced statistically significant catatonia when compared to control animals.

Spontaneous motor activity. The table shows the results obtained with all fractions injected 30 min. before introducing the animals into the cages. Only 20 mg/kg of IIS significantly decreased spontaneous motor activity.


Full size image: 19 kB, FIGURE 3

Effects of 20 mg/kg of fractions IB, IS, IIB, IIS, IIIB, IIIS and 10 mg/kg Δ 9-THC on the catatonia of mice. Asterisks indicate statistically significant differences from the control group (Duncan's new multiple range test; 1 per cent level). Each column averages the total catatonia in two hours' observation, on 10 animals. The vertical bars represent the S.E. of the means.

Spontaneous motor activity of mice i.p. injected with several fractions of cannabis smoke


Number of animals

Dose mg/kg

(Number of light beam interruptions ± SD)

10 5
  10 10
  9 20
9 5
  9 10
  9 20
10 5
  9 10
  10 20
10 5
  10 10
  10 20
62±32 *
10 5
  10 10
  9 20
9 5
  10 10
  10 20

Asterisk indicates statistically significant difference from the control group (p x 0.05; Duncan's new range multiple test).

Corneal areflexia. The fraction IIS, at a dose of 0.658 ± 0.064 (mean ± S.E.), IIIS, at 0.788 ± 0.101, and Δ-9-THC, at 0.096 ± 0.008, abolished the corneal reflex. Fractions IS, IB, IIB and IIIB were inactive up to 6.0 mg/kg.


Our results show that the water-soluble fractions Is (from hashish + tobacco mixture) and IB (only from tobacco) did not possess known cannabinoids and were devoid of clear cannabinoid-like effects. Thus, they did not increase pentobarbital sleeping time (fig. 2), did not induce catatonia (fig. 3), had no effects on spontaneous motor activity (see table), and up to 6 mg/kg were not able to block the corneal reflex of rabbits. These data confirm many previous reports stating that cannabinoid compounds are water-insoluble. On the other hand, the data disclosed a large variation in results from animal to animal, which can be better evidenced in the table (spontaneous motor activity) where the S.D. are very large.

These variations, which were also observed with all other fractions, could be due to active substances from tobacco, probably nicotine, which possesses both stimulant and depressant actions. This suggestion is strengthened by the results with 160 mg/kg of fractions IIIS and IIIB which caused convulsions and deaths to the mice. It could also be that these tobacco substances were responsible for some "placebo" effects, as those observed with 80 mg/kg of IIIB on barbiturate sleeping time in mice (fig. 2), and the reasonable activity of "placebo" fraction IIB on the catatonia method (fig. 3).

Fraction IIIS, the most important as it is the one which reaches the lungs of smokers, was discretely active. Thus, 80 mg/kg provoked a larger barbiturate sleeping time than both the control solution and its IIIB placebo fraction (fig. 2); it also abolished the corneal reflex of rabbits at a dose of 0.788 ± 0.101 mg/kg. However, at the doses employed, IIIS failed to induce catatonia (fig. 3) and to modify the spontaneous motor activity of mice (see table).

Finally, fraction IIS, containing the low volatile material retained by the water phase of the pipe, revealed to be the most active. Thus, it gave positive results in the 4 methods employed by us: barbiturate sleeping time (fig. 2), catatonia (fig. 3), spontaneous motor activity (see table) and abolished corneal reflex at a dose of 0.658 ± 0.064 mg/kg. This rather larger activity of fraction IIS (when compared to fraction IIIS) was somewhat surprising, as it possessed only about 1/3 of the content of delta-9-THC present in fraction IIIS. One possible explanation of this discrepancy lies in the different proportions of CBN and CBD present in these fractions. Fraction IIIS was poorer in CBN and richer in CBD than fraction IIS; consequently the ratio CBN/CBD were, respectively, 11.0 and 43.4. There have been suggestions that CBD is able to block and CBN to mimic, some of the effects of delta-9-THC (Carlini et al., 1974; Karniol et al., in press) which could then explain the discrepancy. Another possibility is that low volatile principles from the "tubeki", retained in the water phase, potentiated delta-9-THC effects. However, regardless which of these mechanisms is correct, the main conclusion to be drawn from this work is that devices such as the water pipe used in Greece retain active principles of cannabis.


Carlini, E. A., I. G. Karniol, P. F. Renault and C. R. Schuster. Effects of marihuana in laboratory animals and in man. Br. J. Pharmac. 50, 299-309 (1974).

Galanter, M., H. Weingartner, T. B. Vaughen, W. T. Roth and R. J. Wyatt. Δ 9-tetrahydrocannabinol and natural marihuana. Arch. gen. Psychiat. 28, 278-281 (1973).

Isbell, H. Clinical pharmacology of marihuana. Pharmac. rev. 23, 337-338 (1971).

Karniol, I. G. and E. A. Carlini. The content of Δ 9-tetrahydrocannabinol does not explain all biological activity of some Brazilian marihuana samples. J. Pharm. Pharmac. 24, 833-835 (1972).

Karniol, I. G., I. Shirakawa, N. Kasinski, A. Preferman and E. A. Carlini. Cannabidiol interferes with the effects of Δ 9-tetrahydrocannabinol in man. Eur. J. Pharmac. 28, 172-177 (1974).

Karniol, I. G., I. Shirakawa, R.N. Takahashi, E. Knobel and R. E. Musty. Effects of Δ 9-tetrahydrocannabinol and cannabinol in man. Pharmacology (in press).

Kephalas, T. A., J. Kibouris, C. M. Michael, C. J. Miras and D. P. Papadakis. Some aspects of cannabis smoke chemistry. In: "Marihuana", Sattelite Symposium of the 6th International Congress of Pharmacology, Helsinki, Finland, July 1975. Springer Verlag (in press).

Kubena, R. K. and H. Barry III. Stimulus characteristics of marihuana components. Nature (Lond.) 235, 397-398 (1972).

Volavka, J., P. Crown, R. Dornbush and M. Fink. Marijuana, hashish and Δ 9-tetrahydrocannabinol: effects on EEG, heart rate and mood. J. Psychiat. Res. 10, 159-165 (1974).