Jacob Bell
New Member
BY S. LOEWE, Ph.D.
Today's status of the pharmacology of Cannabis, to be
reported here as it has emerged from a fourteen year
investigation,* part of which is as yet unpublished, contrasts
sharply with the textbook presentations which even now
often deal with the subject on the level of a-quarter-of-acentury-
old knowledge. In fact, the pharmacological spectrum
of the cannabis drug, formerly derived in part from
misinterpreted actions of crude preparations of either
Oriental (hashish) or American hemp (marijuana), has been
classified and re-evaluated; botanical relations between the
various "species" of Cannabis have been revised by aid of
determinations of their content in active principles; the
SARt of the cannabis-active substances has been elucidated;
a new class of chemical agents, comprising products of
laboratory synthesis as well as of plant origin, has been
opened up; and, last but not least, experimental and clinical
investigation of the pure substances has turned a subject of
merely toxicological interest into a source of therapeutic
potentialities.
* The status of the Cannabis problem immediately before the discovery
of the active principles (1938) is best surveyed in Walton's book and
Blatt's chemical review. The literature from then on to 1942 will be
found in Roger Adams' lecture and in the review which the author in
1941, contributed, upon the La Guardia Committee's invitation, to that
committee's book published in 1944.
t "SAR" (relationship between chemical structure and biological activity)
has found entrance into the American literature as a concise
symbol suggested (Fed. Meetings, 6 :352, 1947) for that phrase so
difficult to compress into a brief expression, and will be used in that
sense (Structure-Activity Relationship) hereinafter.
Originally published in German in Arch. Exper. Path. u. Phurmakol., vol. 2 11,
1950, pp. 175-193. Translation courtesy of Dr. Carl C. Pfeiffer, New Jersey
Bureau of Research in Neurology and Psychiatry, Princeton, New Jersey.
That the active substances are contained in the so-called
active principles
of hemp. "Red Oil," a high-vacuum distillate from cannabis extracts,
had been known for many decades. Only in 1937 to 1942,
however, was it ascertained by the work of an American
research team* that the Red Oil consisted essentially of two
inactive substances, cannabinol (CZ, H,, 0, ; III) and cannabidiol
(CZl HSO 0, ; I), and a varying mixture of tetrahydrocannabinols
(C,, HSO 0, ; THC; II a, b, and c), the representatives
of "cannabis activity."t The conclusion of this first
research phase was marked by a detailed description of the
first and apparently maximally potent natural agent, a THC
from Indian hemp resin, designated as charas tetrahydrocannabinol;
its potency (P; cf. below) was 14.6.
Motor ataxia in the dog was the test reaction which served
as a guide in the procedures to isolate the active compounds.
It served to determine their content in the starting materials,
to fractionate the crude oils, and to assay the potency of the
purified substances. That the research plan resulted in success
is due to two circumstances, namely, the choice of the right
test response and its adaptability to quantitative purposes. It
is now evident that motor ataxia is the one action in
experimental animals which closely parallels psychic cannabis
action in man; although this action is not accessible to
evaluation by the customary biostatistical procedures because
of the excessive variation of inter-individual sensitivity of the
dog, a biostatistically adequate assay method could be developed
by the introduction of a new principle of "intraindividual
potency comparison."
The history of the discovery of the cannabis-active substances
is characterized by the following paradox. On the one
hand, those investigators who followed analytical procedures,
by employing tons of starting material, large-scale molecular
* H. J. Wollner, J. R. Matchett and J. Levine, Narcotics Laboratory,
U.S. Treasury Department, Washington, D. C.; Roger Adams, et al.,
Department of Chemistry, University of Illinois; and this author.
t This is the term to be employed hereinafter for the characteristic
psychic action and its equivalent in the animal experiment, which in
some countries are designated as "hashish action," in others as "marijuana
action."
distillation, and chromatographic and elution methods
rapidly got hold of numerous, obviously chemically pure
active substances but were unable to identify them structurally,
particularly since even today they foil all attempts at
crystallization. On the other hand, the discoverer of the two
inactive by-products predicted the structure of the plant
principles without going into an analysis of any one of the
natural agents. As early as 1930-33, Cahn in England had
obtained cannabinol as a major component of the hashish
resin formed as a surface excretion on Oriental hemp, and
had identified it correctly except for minor inaccuracies in
the positions of the hydroxyl and the amyl group. In 1939,
Adams demonstrated it to be a 1-hydroxy-3-n-amyl-6,6,9-trimethyl-
6-dibenzopyran and at the same time prepared it
synthetically. Cannabidiol, of which hashish resin contains
only minimal quantities, was discovered in 1940 by Adams in
extracts from freshly harvested North American hemp as a
similarly considerable part (33 percent) of the crude oil,
identified as menthadienylolivetol, and synthesized. The
chemical relationship between the two substances, their
inactivity, and the age differences of the raw materials gave
rise to the ingenious hypothesis that cannabidiol signified the
starting material and cannabinol the end-product of a phytochemical
conversion process and that the intermediates on
the way of this process, namely, hydroaromatic cannabidiolisomeric
precursors of cannabinol, may be the searched-for
active principles.
This was confirmed on various avenues of approach: (1)
Adams, by intramolecular condensation of cannabidiol,
obtained-depending on the process of isomerization
employed-two such semi-synthetic tetrahydrocannabinols
(II b and c), both of which Loewe found to be very markedly
cannabis-active; the two isomers differ in optical rotation,
position of the double-bond, and potency (7.3 and 8.2). (2)
At once, the oily substances which Wollner and Loewe and
their associate? had isolated from hashish and American
hemp and found to be highly potent could now be identified
as THC. (3) Soon, Adams mastered the difficult synthesis of
another THC (IIa) from 5-methylcyclohexanon-2-carboxylate
and olivetol, the much lower potency of which was from
then on employed as standard (P = 1). (4) Part of the
presumable processes which in nature create the active
substances could be imitated in vitro; in an inactive cannabidiol
synthesized by Adams, which had been irradiated
with ultraviolet light, active substances could be demonstrated4
in an amount which indicated that about 2 percent
had been converted to THC.'
Like the synthetic cannabidiols, the completely synthetic
THCs are racemates, whereas the natural and the semisynthetic
THCs are laevorotatory. In 1942, Adams and his
associates dissolved the synthetic rat. THC into its two
optical isomers (rot. +152 and -114, respectively), and here,
too, the I-isomer (P = 1.66) turned out to be superior to the
d-isomer (P = 0.38).*
Stereo- and optical isomerism make possible a large num- Content of
ber of isomeric THCs which differ in the spatial arrangement ~~~r$&,,,e,
of the 9-methyl group and of the planes of the hydroaromatic
ring, and in the position of the double-bond in this
ring. Probably all these isomers are cannabis-active, though
they differ greatly in potency, and quite a number of them
has been prepared from hemps of different origin.4 The
potency of fluid extracts from hemps of different origins also
varies quite markedly. Botanically, however, it is noteworthy
that on the one hand fluid extracts from hemp cultivated in
Rumania, Manchuria, Italy, Tunis and climatically very different
parts of U.S.A. were cannabis-active (P: max. 0.52 in
Tunisian hemp, min. 0.003 in an American hemp).4 On the
other hand, a hemp grown from Oriental seeds on an
experimental field near Washington, D.C., and re-seeded there
for three subsequent years maintained its high potency.4 It is
therefore quite possible that the composition of the active
fraction constitutes the only difference between various
varieties of Cannabis sativa. American hemp appears to be
characterized by THCs of P 6.0 to 7.3 and 8.0 to 9.5, which
may be identical with Adams's semisynthetic THCs from
cannabidiol. In Oriental hemp, THCs of P 12.0 to 14.6
occur, in which the position of the double-bond in the
hydro-aromatic ring is as yet undefined and which have not
yet been obtained from other hemps. At closer analysis, the
cannabinol component of the hemp plant proved also to be
cannabis-active; completely synthetic as well as natural,
frequently re-crystallized cannabinols had the same potency
of 0.049 , which obviously is only of academic interest. The
hexahydrocannabinols are also active4 and so are presumably
the dihydrocannabinolst which can be assumed to be intermediate
products in the conversion of THC to cannabinol;
but none of these have as yet been demonstrated in hemp. In
the plant, the cannabinols appear to occur in part as esters of
aromatic acids.'7C
The content in active substances is not limited to the
"flowering tops of the male plant," but occurs in many parts
of the plant; for example, even seedlings of a few centimeters
in height4 and seeds6b yielded active extracts.
Synthetic cannab- According to the preceding data, cannabinol and all reducinols
and S~XUC- tion products
ture-Activity
of its toluene ring can be considered to
Relationship embody cannabis activity, and, in view of the activity of the
(SAR) parent substance, insignificant as it may be, it appears
chemically and pharmacologically justified to designate the
new class of chemicals as the class of cannabinols. After the
gates had been opened by the disclosure of the natural agents
and of synthetic procedures, the already mentioned representatives
of the class were joined by many other compounds
which served in our studies of SAR. Also to a team of British
investigators39 4 who had already for quite a time devoted
themselves to the problem of the Cannabis drugs, that gate
opened up a field of successful and, in part, independent
chemical research whose results, unfortunately, for reasons to
be discussed below did not become serviceable to the comparative
study of cannabis activity in its proper sense.
Information on SAR of the class was largely obtained by
measurements of the ataxia potencies. It was demonstrated
that the I-hydroxyl group is an important component of the
phenol ring of the THCs; its effective blockade greatly
diminishes activity. 3y 4 Also significant is the position of the
substituents of this ring. For example, the 1-n-hexyl-3-
hydroxy isomer of parahexyl, * the 3n-hexyl homolog of the
completely synthetic THC, has a P of only 0.05-t; i.e., only
one-fortieth that of the substance of comparison. The role of
the hydro-aromatic ring was studied in many analogs of the
"synthetic THC" (P = 1). Lack of the 9-methyl group brings
P down to 0.13 ; its substitution by ethyl, to 0.22 ; transfer to
position 10, to 0.25; to position 8, to 0.14; an additional
methyl group in position 7, to 0.75, but in position 8, to
0.11; replacement of the 6, 6-methyls by ethyls reduces P to
0.12, by propyls to 0.04. All such alterations at the cyclohexene
ring, as well as opening of the ring so as to leave in its
place a 6-n-butyl and an 11-methyl group (P = 0.04),
reducing it to an 11-methyl group 0.033) or replacing it by
cycloheptan (0.21) reduce but do not completely abolish
activity.
Not until the significance of the 3-alkyl side chain was
studied were increases in potency observed. The changes of P
with changes in the length and arrangement of this side chain
are presented in condensed form in Figure 1. Essential facts
on SAR are as follows: (1) Shortening of the 3-n-amyl chain
decreases potency.4 (2) Peak-effective, however, in the series
of homologs of the "synth. THC" standard (Series A of
Figure) is not this n-amyl compound, but its 3-n-hexyl
homolog (parahexyl).4 which is about twice as potent (1.8),
yet far inferior to the natural THCs. The same relation
repeats itself in the other homologous series (C, D; cf. below)
having an unbranched side chain." (3) Branching of the side
chain can result in substances of far greater P, but in these
series (E, F, and G" -15 ) peak potency is associated with a
six- but not with a nine-C-atom side chain. (4) Optimum
effectiveness was always found in side chains branched near
to their "root"; for example, among the methyl-amyl isomers,
P decreases consistently from the value 3.65 of the
1'-methylamyl-R (Series E) with transfer of the methyl group
into position 2' (H; 1.58), into 3' (K; 1.25), and 4' (L; 1.14).
(5) Methyl branching under otherwise equal conditions,
appears to grant the greatest activity [compare in Series E
l'-methylamyl-R (3.65) with 1'-ethylbutyl-R (M; 1.68), and
l'-isopropyl-propyl-R (N; 3.18), and in Series G l', 2'-dimethylbutyl-
R (3.84) with l'-ethyl-2'-methylpropyl-R (P;
3.40)]. (6) Double branching at the same C-atom appears not
to offer any advantage over single branching (compare the
two Series E and F). In contrast, two methyl branches at two
adjoining positions 1' and 2' (Series G) result in representatives
of outstanding potency which is only diminished by
prolongation of the branched chain (Series P) or by addition
of a third branch (Q).*
The role of the third (pyran) ring is illustrated by the
observation that analogs of the hypothetical type IV, prepared
from pulegon and appropriate resorcinol derivatives,
are in tetra- as well as in hexa-hydrated from (Series C and D,
respectively) approximately as potent as the corresponding
cannabinols. These compounds may differ from the cannabinols
in the mutual position of the two carbon-rings and
in the structure of the heterocyclic ring. 1'-methylation
appears not to increase potency in this series (cf. R).
Naturally, the problems of SAR are by no means consummated
by the comparative bioassays of this scanty hundred
of compounds hitherto studied. Many questions, among them
that of the applicability of experiences in one series to other
series, e.g., to homologs of the natural THCs, are still open.
The role of the configuration of the 3-side-chain could
certainly be further elucidated by the study of another
hundred congeners; for example, substances have been
synthesized in which the 3-alkyl chain is linked to the phenol
ring by an ether linkage I6 - "' ; if it were permissible to draw
conclusions from the results of the Gayer test, a weak
Cannabis activity would have to be ascribed to some of these
ethers.'7a
Only recently, in a study of 2-cyclohexyl-3-hydroxyl-6, 6,
9-trimethyl-7, 8, 9, lo-tetrahydro-dibenzopyran," we found
that cyclic arrangement of the side chain is superior to a
straight and unbranched side chain (compare S in Figure 1;
potencies of the corresponding n-alkyls are below the frame
of the graph)-a finding which points to new possibilities.
The substances of the cannabinol class are viscous oils, Conditionsof
some solid at room temperature, and extremely poorly ~t~~,?~'~
water-soluble. Even in the best solvents for cannabinols
(acetone, ethanol, glycols), certain representatives with long
side chain dissolve only slowly and yield only colloidal
solutions. Accordingly, all cannabinols are difficultly absorbed;
even after intravenous injection, thirty to sixty
minutes and often more elapse until peak effect is attained,
and the peak effect persists for hours-even for five days, as
was observed after a single mediumsized intravenous dose of
the l'-methylnonyl representative of Series E. Oral administration
requires twenty to thirty times the intravenous dose,
and subcutaneous administration hardly ever produces a
demonstrable effect. Even after high doses, we were as yet
unable to demonstrate urinary excretion of active substances."
In the blood, however, 1.6 to 10.5 percent of
parahexyl given by vein were found circulating after two to
three hours and 2.4 percent after eleven hours; the lungs of a
dog contained 0.9 percent of a dose injected twenty hours
before.18 Probably because of these physicochemical characteristics
of the compounds, the margin of safety is enormous;
per kg body weight, the oral L.D.,, of Charas THC in the
mouse is more than two hundred thousand times the intravenous
threshold dose for ataxia in the dog and more than
forty thousand times the oral threshold dose for psychic
action in man."
Not all the actions of crude hemp preparations or oil
redistillates reported in the literature could be reproduced
with pure active substances. At that, many of them are of
minor importance at the present state of the Cannabis
problem. This holds true, for instance, for the effects upon
heart rate, respiration and pupil. Therefore, the following
discussion will be limited to those, in part newly revealed,
sections of the large action spectrum, which have been
studied with the aid of pure substances.
1. Psychic Actions. Analysis of the psychic actions in
man, still more so in animals, is a task as yet unsolved. Even
with pure substances, neither the host of experimental
psychologists of the LaGuardia Committee" nor the effects
with a "battery of psychological tests" both elsewhere** and
in test persons in our department*' were capable of disclosing
more than a large range of variation of all phenomena
tested. Thus, the availability of numerous pure substances of
the class has not as yet given impetus to the analysis of the
unique "pleasure action," but a few experiences have been
collected which deserve to be mentioned partly as modest
beginnings, partly with a view to therapeutic aspects. For
instance, both parahexyl** and compounds "RA 122" and
"RA 125A"*l were found to have no effect upon the
normal human EEG or upon musical appreciation.22b Therapeutic
effects of parahexyl have been reported in depressive
states23 a but in consideration of the negative experiences of
Pond23b and of unpublished ,personal reports to the writer
they are in urgent need of further verification. Charas THC
has been extensively employed for "psychic relaxation" and
i s reported24 to be equivalent to the customary barbiturates
for purposes of narcoanalysis, narcocatharsis and
narcosuggestion. The idea, however, that Cannabis and morphine
may be comparable in euphorizing effects has not
proved fruitful in its application to morphine withdrawal;
withdrawal symptoms were neither shortened nor abated by
parahexyl treatment.2sa Noteworthy, though difficult to
evaluate, are reports on psychiatric episodes, mostly in
psychopaths, occurring in some instances under prolonged
treatment with, and in other instances upon withdrawal of
cannabis-active drugs .25 a -g
2. Cataleptic Effects. They are never missing in the picture
of the hemp "jag," always accompany the ataxia effects in
the dog, and are produced by all ataxia-active pure substances4*
19 They can be seen in many species of animals, and
a catalepsy test has been describedi which occasionally has
been found quite serviceable.
3. Ataxia Action. This has gained ever-increasing prominence
among all cannabis actions, as an experimental tool. It
is not only highly serviceable,* but is closely correlated with
psychic action. This has been confirmed for all those crude
and pure products which have been both bioassayed by the
writer and evaluated in man20p24 ; particularly good agreement
was found between ataxia and psychic potencies of
four cannabinols at the occasion of their especially careful
and expert evaluation by Dr. L. Kubie. There are only vague
notions about the locus and mechanism of the ataxia action.4
In view of its outspoken parallelity to the psychic action, a
close correlation in the mechanisms of the two actions has to
be considered seriously. As a matter of course, final proof of
both identity in mechanism and the quantitative parallelity can
only be expected from an evaluation in man of all the
compounds available.
4. Central Stimulant Actions were observed with all substances
of the class. From medium degress of ataxia upward,
dogs at times exhibit convulsive motor symptoms.4F l9 They
are possibly related to the compensatory counteractions
against the disturbance in coordination; their frequency
varies with the dog's individuality and strain. Indisputable
and unique convulsions appear consistently after doses
greater than twenty times the threshold ataxia dose after
intravenous administration of the high-potent "RA 122" and
dominate the picture of action of lethal doses. Mice often
become pugnacious after medium doses of cannabinols."
Vomiting occurs in dogs rather regularly and late, usually one
to five hours after injection, following doses upward from a
medium-effective intravenous dose. The emetic effect as well
as the hyperexcitability of the scratch reflexes can probably
be ascribed to a central stimulant action.4' l9
5. Hypnotic Activity. Elements of central nervous stimulation
among the actions of hemp are of interest in view of
the question of its soporific action. Probably cataleptic
symptoms in man have often been interpreted as signs of a
sedative action. In the dog, concomitance of catalepsy with
higher degrees of ataxia which prevent upright posture results
in a syndrome which might impress a superficial observer as
sleep. In experiments with combinations of aqueous hemp
extracts and hypnotics, Burgi believed he had demonstrated a
hypnotic component of Cannabis action.27 We were able to
duplicate his observation with fluid extracts of hemp; they
prolonged sleep duration in mice after a pernocton dose
which is just enough to suppress the righting reflex. Sleep
prolongation, in our experiments, was significant indeed, but
only for this one hypnotic; however, even the consideration
prolongation by a large dose of Cannabis equals no more than
the effect producible by an additional 10 to 20 percent of
the pernocton dose employed. Above all, however, a study of
pure substances gave proof that this action is due not to any
one of the cannabis-active components, but to the otherwise
inert cannabidiol.28 * Moreover, the natural as well as the
synthetic cannabis-active substances lack any other indications
of central nervous depressant action. Even in states of
severest hypomotility and impairment of attention, up into
the terminal phases of lethal effect, the animals, quite
contrary to those under the influence of hypnotics, still
respond, with frustraneous movements to moderate acoustic,
tactile, or pain stimuli. This is in accordance with the
characteristic criterion of Cannabis action both in dog and in
man, namely, that the drugged individual can readily be
diverted by environmental stimuli.4
6. Cornea1 Areflexia. Abolition of the wink reflex in
rabbits after administration of hemp extracts or crude oils
was discovered in 1928 by Gayer in Straub's laboratory and
considered to be a faithful expression of Cannabis activity.29
First of all, at closer analysis this systemic action has proved
inappropriate for quantitative purposes because of the great
variation in inter- and intraindividual susceptibility. Only
after a tedious search in a large stock of animals by way of
numerous recalibrations can one find a number of individuals
suitable to yield conclusive data in strictly intraindividual
potency comparisons4> 3o that there are two types of areflexia
producing agents, namely, (a) ataxia-active THCs which
lose neither in ataxia nor in areflexia activity by such
oxidation, and (b) other extractives possessing little or no
ataxia activity, but a marked areflexia activity which can be
destroyed by oxidation. In the search for the as yet unidentified
substances of this type (b), the Gayer test in its
present30 more complicated but more reliable form is thus
the test reaction of its choice.
7. Anticonvulsive Action. It was only by the aid of
the highly active synthetic cannabinol congeners and due to
the methods elaborated in this laboratory by Goodman and
Toman and associates3' that the antiepileptic activity of
cannabinols could be disclosed and specified. All cannabinols
tested were effective. Their activity is of the type of anticonv&
ant activity of diphenylhydantoin. Like this drug, they
abolish the tonic hind leg extensor component in the pattern
of supermaximal electroshock in rats and cats33 (and presumably
also in man34), whereas they prolong the duration of
metrazol convulsions, modify their appearance and display a
strong lethal synergism with metrazol.33 The cannabinols
differ from diphenylhydantoin and its congeners by exhibiting,
in their maximum-potent representatives, much greater
potency and, in reference to lethal effectiveness an incomparably
greater margin of safety. When referred to psychic
side effects, the therapeutic indices of the various agents vary
considerably. In a preliminary experiment,35 "RA 122"
lacked the ability of diphenylhydantoin, discovered by
Toman et a1.,36 to raise the threshold of the isolated nerve
preparation for electrical stimulation and to prevent the
repetitive discharge elicitable by immersion of the nerve in
neutral isotonic phosphate solution; it is true that, in view of
the minimal water-solubility of the cannabinols, the result is
inconclusive.
Alteration of the electroshock pattern points to anti-grand
ma1 activity, suppression of metrazol convulsions to anti-petit
ma1 activity. 32 Accordingly, clinical experiments were undertaken
with some of the agents. A first series3' in diphenyl
hydantoin-refractory grand-mal epileptics proved noteworthy
effectiveness and absence of psychic side effects when "RA
122" was given orally for several months in daily doses of
one mg or less, or the weaker isomer, "RA 125A," in
somewhat higher dosage. A second series3' consisted of five
institutionalized children with severe grand-n-al epilepsy and
mental underdevelopment, in whom daily doses of 0.13 gm
phenobarbital combined with 0.3 gm diphenylhydantoin or
0.2 gm Mesantoin had proved inadequate. "RA 122," in daily
doses of 1.2 to 1.8 mg, was in three children "at least as
effective" as prior therapy; the fourth became almost free
from attacks and the fifth completely free. Following transfer
to 4 mg of "RA 125A," the attacks remained infrequent
in one patient, the other one suffered exacerbations. The first
patient had a brief paranoid episode, similar to others he had
repeatedly experienced prior to the cannabinol therapy. In
contrast to their ineffectiveness upon the normal EEG the
cannabinols normalized the EEG of grand-ma1 patients.
The protective activity of cannabinols against electroshock
and grand-mal attacks is of theoretical interest in view of the
apparent absence of structural relationship between the
cannabinols and the antiepileptics from the classes of hydantoins,
barbiturates and oxazolidinediones. The practical evaluation
of cannabinols as antiepileptics, notwithstanding their
superior potency and persistence of action, will largely
depend upon the problems of their side effects which will be
discussed below.
8. Analgesic Action. It was probably the conceivable tendency
to compare every euphorizing drug with. morphine
regarding anodynic properties, which brought British investigators
to take up anesthesia tests and to ascertain the
intravenously injected hashish extracts were somewhat effective,
and that some cannabinols of the 1-hydroxy-3-alkyl-R
and one of the 2-alkyl-3-hydroxy-R type were considerably
effective (see Table 1). Using a modification, introduced in
this department by Nickerson, of the customary methods of
rat-tail heat stimulation which had also been employed in the
afore-mentioned studies, we observed an even greater effectiveness
in two other cannabinols.
9. Lethal Action. As already mentioned, large doses are
required for a lethal effect, which are not always available
and are not easy to administer because of the poor solubility
of the substances. Data for quite a number of older preparations
(previous to 1947) have been presented previously.", l9
Only some particularly illustrative observations need be
reported here. The values of intravenous L.D.,, in the
mouse4 indicate that the lethal toxicity of pure substances is
markedly lower than that of the crude preparations. The
necessity of employing solvents such as propylene glycol,
which are by no means indifferent, and the minimal watersolubility
of the drugs are probably responsible for the poor
reproducibility of the lethal-dose values. Such shortcomings
are less important than the experience that the lethal
mechanism of the same substance can be different with
different routes of administration. The late death of dogs
after oral administration of parahexyl is associated with
profuse intestinal hemorrhage, after intravenous injection
with severe pulmonary edema. In contrast, incomparably
smaller intravenous doses of "RA 122," which is ataxiaeffective
even in doses of a few micrograms, cause death
associated with convulsions within a few hours, obviously
owing to some primary central nervous mechanism. This
suggests the interpretation that the lethal effect of the
lower-potent cannabis-active substances is due to non-specific
mechanisms and that only that of the high-potent agents
originates from a mechanism more closely related to the main
activity, which in the low-potent substances is masked or
outdone by non-specific toxicity.
10. Habituation. The clinical authors of the La Guardia
Committee's report, on the basis of observations in chronically
marihuana-smoking prisoners, deemed the danger of
habituation negligible. 2o This has evoked a vehement controversy
.40 The most lucid answer to the problem of hemp
habituation will be found with Goodman and Gilman.4'
Conclusive animal experiments are not available. For its
curiosity value, the "beginner's habituation" may be mentioned:
The more than one hundred fifty dogs from the
writer's twenty-six hundred bioassay experiments, some of
which had been tested twice per week for many years, almost
invariably exhibited a certain decrease in sensitivity during
the period of the three to four initial experiments, and from
then on maintained a rather constant susceptibility. This
habituation was the more pronounced, the less the individual's
postural behavior was dominated by tenseness. Accord
ingly, the apparent decrease in susceptibility seems to be due
to the "learning" of postural responses compensating for the
ataxic incoordination, at the mercy of which the individual
finds itself quite helpless during its first experiences in ataxia.
All other intraindividual variations in susceptibility, as they
developed over long periods, were completely irregular,
varied in intensity, and consisted in increases as well as
decreases; in bitches, they were not without relation to sex
cycles and pregnancy, for which reason all assays were
conducted in males.
Interrelation It has been repeatedly emphasized above that up to now
between different
cannabinol actions.
all observations have pointed to an even quantitatively close
relation between the psychic and the ataxia activity. The
ratio between psychic potency in man and ataxia potency in
dogs of an individual substance is a constant which varies
very little in the entire class. Contrariwise, it has been
demonstrated above that neither of these two activities has a
constant relationship to lethal activity.
That the cornea1 areflexia action is not correlated with the
ataxia action, is evidenced by the fact that in crude hemp
products the two activities can be dissociated by oxidation. It
is true that not only the unknown cannabis-inactive substances
disclosed by that procedure, but also ataxia-active
cannabinol possess areflexia activity. However, the potency
ratio between the two activities varies from substance in a
wide range, as has been shown earlier4 for a series of
cannabinols and is confirmed in Table 1 in nine, partly new
substances. A similar inconstancy of the potency ratios is also
evident when analgesia and ataxia activity are compared
(Table 1, column 5).
The relation between antiepileptic (anti-electroshock) and
ataxia activity is also marked by a certain though lesser
inconstancy of the potency ratios (Table 1, column 4). Here,
however, the results should be evaluated with greater reserve.
For the electroshock experiments, for better comparability
of the cannabinols with other antiepileptics, were performed
after oral administration of aqueous, lecithin-homogenized
emulsions of oily drug solutions, whereas the potency values
for ataxia and all other actions were determined after
intravenous administration. Since in these comparisons numerator
and denominator of the potency ratio originate from
experiments with different routes of administration, it is
possible that inequality of the ratios is due to differences in
absorption, distribution and detoxification. A comparison of
the E.D.,, values for cornea1 areflexia of the British investigators
with the writer's ataxia values indicates that in some
cannabinols the areflexia dose is smaller than the ataxia dose.
That makes it tempting to base all evaluation upon the Gayer
test which, superficially judged, is more convenient than the
ataxia test. For SAR studies and for the assay of cannabis
activity of crude preparations and unknown mixtures, this
must be urgently discouraged because the potency ratios are
so inconstant. Quite inversely, the analgesic dose of all
cannabinols and the anti-electroshock dose of some are high
as compared with the ataxia dose. In such substances, each of
the two therapeutic effects may be obtained only at the price
of considerable psychic side effects. Were such an unfavorable
potency ratio common to all cannabinols, the outlook
would be small indeed that further search could reveal other
substances of the class more devoid of these side effects. The
inconstancy of the ratios leaves the possibility open that
there are cannabinols of a greater therapeutic index both for
analgesia and for epilepsy therapy. Actually, available experience
already teaches that in clinically equieffective doses the
only two cannabinols which have as yet undergone clinical
examination are not equally liable to produce psychic side
effects : "RA 122" not only exerted greater therapeutic
action than "RA 125A" but also had a smaller incidence of
psychic Cannabis effects. That the two substances differ
clinically with regard to the risk of side effects agrees with
the fact that the experimental potency ratios of different
cannabinols differ; that "RA 122" has fewer side effects than
"RA 125A" is contrary to what one would have expected
according to the size of the two potency ratios and may thus
indicate that ratios of experimental potency values which are
obtained after administration by different routes are not the
last word on this subject.
Theoretically, the dissociation between psychic activity on
the one hand and analgesia, cornea1 areflexia and antiepileptic
activity on the other hand, as demonstrated by the
inconstancy of the respective potency ratios, is probably an
indication that the actions compared must be ascribed to
different reactive groups or, more generally expressed, to
different parts of the total configuration of the same drug
molecule.
Summary. 1. This review attempts to present some evidence for the
proemial statement that our knowledge of the cannabis drugs
has undergone a striking change.
2. After a report on the isolation of the natural Cannabis
agents and their identification as tetrahydrocannabinols, their
subsequently studied synthetic isomers and analogs are surveyed
and some data are reported on the structure-activity
relationship in this new class of chemicals, the cannabinols.
3. The pharmacology of the cannabis-active substances,
as revised with the aid of a study of the pure substances, is
briefly presented, and examples are given for the decisive
elucidation obtained by the study of the most recent,
highest-potent synthetic agents, the potency of which was
found to be up to seventy times the average and up to
thirty-five times the maximum potency of natural tetrahydrocannabinols.
4. Some details and therapeutic trials of the psychic
cannabis action are reported.
5. The significance of the ataxia action in the dog, which
appears to parallel closely the psychic action in man, for
identification and bioassay of cannabis-active compounds is
illustrated and, contrariwise, the cornea1 areflexia in the
rabbit is discussed as a quite detached property of some of
the hemp products and congeners, which is only of very
limited usefulness.
6. The old question of the hypnotic activity of the
Cannabis drugs is answered in the negative from a study of
the pure substances; closest to hypnotic activity is the
practically insignificant, very limited capability of crude
preparations to prolong the hypnotic action of certain barbiturates,
an activity which is absent in the pure cannabisactive
substances and embodied only in an otherwise inert
by-product, cannabidiol.
7. The analgesic activity of cannabinols, only recently
disclosed and not yet tested for its practical applicability, is
briefly discussed.
8. In their anticonvulsant activity, also an only recently
discovered property, the cannabinols are demonstrated to be
closely related to diphenyl-hydantoin, according to both
experimental criteria and clinical experiences; the most
potent synthetic congener of the cannabinols appears to
possess more than one hundred fifty times the anti-grand ma1
potency of diphenyl-hydantoin.
9. The problem of the therapeutic indices of the five
major actions and that of their mutual relations in the roles
of main and side actions are dealt with on the basis of
experimental data; the potency ratios hitherto established do
not preclude the possibility that continued search of the new
class of compounds will lead to congeners having a still
greater margin of antiepileptic effectiveness than the as yet
best examined "RA 122," and perhaps also to congeners
having an adequate therapeutic index of analgesic activity.
10. Contributions are presented to the problems of habituation
and of the mechanism of the lethal action; as a
menace of habituation, addiction and tolerance the cannabisactive
drugs appear to rank lowest among the narcotics, and
according to the margin of safety some of them rank
uppermost among all drugs.
REFERENCES
Marijuana: Medical Papers
1 . Walton: Marihuana. Philadelphia-London 1938.
2 . Blatt: J. Washington Acad. Sci. 28, 465 (1938).
3 . Adams: Harvey Lectures, Ser. 37, 168 (1941-42).
4. Loewe: Studies on the Pharmacology of Marihuana. In: The
Marihuana Problem in the City of New York. Lancaster 1944.
5. Wollner, Matchett, Levine & Loewe: Jour. Amer. Chem Sot.
64, 26 (1942).
6a. Loewe: Jour. Amer. Pharmaceut. Ass. 28,427 (1939).
6b. Matchett & Loewe: ibid. 30, 130 (1941).
6c. Loewe: Science 106,89 (1947).
7 . Levine: Jour. Amer. Chem. Sot. 66, 1868 (1944).
8. Adams, Smith & Loewe: ibid. 64, 2087 (1942).
9 . Loewe: Science 102, 615 (1945).
1 0 . Madinaveit,ia, Russell & Todd: Jour. Chem. Sot. 1942, 628.
11. Avison, Morrison & Parkes: ibid. 1949, 952.
1 2 . Adams, Chen. & Loewe: Jour Amer. Chem. Sot. 67, 1534
(1945).
1 3 . Adams, Aycock & Loewe: ibid. 70, 662 (1948).
1 4 . Adams, Mackenzie & Loewe: ibid. 664.
15. Adams, Harfenist & Loewe: ibid. 71, 1624 (1949).
1 6 . Alles, lcke & Feigen: ibid. 64, 2031 (1942).
17a. Bergel, Morrison, Rinderknecht, Todd, MacDonald & Woolfe:
Jour Chem. Sot. 1943,286.
17b. Bergel &Wagner: Ann. d. Chem. 482, 55 (1930).
18. Loewe: Jour. Pharmacol. and Exper. Therap. 86,294 (1946).
1 9 . Loewe: ibid. 88, 154 (1946).
20. The Mayor's Committee on Marihuana, Clin. Study in: The
values referred to Parahexyl. - c: = Anti-electroshock potency, according to
unpublished experiments of Dr. E. A. Swinyard. - d: This value results from the
author's own experiments (analgesia comparisons after oral administration),
whereas all other values of this column are calculated from the analgesia doses
which Avison C.S. l1 determined after intravenous injection. Hence this value is
not directly comparable with the other values of this column, but this is of no
significance for the overall outcome. - e: Loewe, unpublished. - f: s.e. as
percentage of mean.
Marihuana Problem in the City of New York. Lancaster, 1944.
21. Davis: Unpublished.
22a. Williams, Himmelsbach, Wikler, Ruble & Lloyd: Publ. Health
Reports 61, 1059 (1946).
22b. Aldrich: ibid. 59,43 1 (1944).
23a. Stockings: Brit. Med. J. 1947, 918.
23b. Pond: J. Neurol. 11, 271 (1948).
24. Kubie & Margolin: Psychosomat. Med. 7, 147 (1945).
25a. Himmelsbach: South. Med. J. 37, 26 (1944).
25b. Fraser: Lancet 1949, 747.
25~. Chopra, Chopra & Chopra: Indian J. Med. Res. 30, 1 (1942).
25d. Marcovitz & Myers: War Medicine 6, 382 (1944).
25e. Bromberg: J. Am. M. A. 113,4 (1949).
25f. Reichard: Fed. Probationer, Oct.-Dec. 1946, 15.
25g. Allentuck & Bowman: Amer. J. Psychiat. 99, 248 (1942).
26. Loewe: Federation Proc. 6, 352 (1947).
27. Gisel: Z. f. exper. Path. u. Ther. 18, 1 (1916). Burgi: (German).
med. Wschr. 1924, Nr. 45.
28. Loewe: J. Amer. Pharm. A. 29, 162 (1940).
29. Gayer: Arch. exper. Path. u. Pharmakol. 129, 312 (1929).
30. Loewe: J. Pharmacol. and Exper. Therap. 84, 78 (1945).
3 1. Alles, Haagen-Smit, Feigen & Dendliker: ibid. 76, 21 (1942).
32. Goodman & Toman: Proc. First Nat. Med Chemistry Symp. of
the Amer. Chem. Sot. 1948, June 17-19, 93; Toman & Goodman:
Physiol. Rev. 28, 141 (1948); Goodman, Toman & Swinyard: Arch.
Internat. de Pharmacodynamie 78, 144 (1949).
33. Loewe & Goodman: Federation Proc. 6, 352 (1947).
34. Toman, Loewe & Goodman: Arch. Neurol. a. Psychiat. 58, 312
(1947).
35. J. E. P. Toman: Personliche Mitteilung.
36. Toman, Greenhalgh, Carlson & Bjorkman: Federation Proc. 8,
284 (1949).
37. Belknap, Ramsey & Loewe: Unpublished.
38. Davis & Ramsey: Federation Proc. 8, 284 (1949).
39. Davies, Raventos & Walpole: Brit. J. Pharmacol. and Chemotherapy
1, 113 (1946).
40. Bouquet: Jour. Amer. Med. Ass. 124, 1010 (1944). Bowman:
125, 376 (1944); (editorial) 127, 1129 (194.5); Walton: 128, 383
(1945); Bowman: 899. Anslinger: 1187; Marcovitz: 129, 387 (1946);
41. Goodman & Gilman: The Pharmacological Basis of Therapeutics,
New York, 2nd edition.
Source: The Active Principles Of Cannabis And The Pharmacology Of The Cannabiiols
Today's status of the pharmacology of Cannabis, to be
reported here as it has emerged from a fourteen year
investigation,* part of which is as yet unpublished, contrasts
sharply with the textbook presentations which even now
often deal with the subject on the level of a-quarter-of-acentury-
old knowledge. In fact, the pharmacological spectrum
of the cannabis drug, formerly derived in part from
misinterpreted actions of crude preparations of either
Oriental (hashish) or American hemp (marijuana), has been
classified and re-evaluated; botanical relations between the
various "species" of Cannabis have been revised by aid of
determinations of their content in active principles; the
SARt of the cannabis-active substances has been elucidated;
a new class of chemical agents, comprising products of
laboratory synthesis as well as of plant origin, has been
opened up; and, last but not least, experimental and clinical
investigation of the pure substances has turned a subject of
merely toxicological interest into a source of therapeutic
potentialities.
* The status of the Cannabis problem immediately before the discovery
of the active principles (1938) is best surveyed in Walton's book and
Blatt's chemical review. The literature from then on to 1942 will be
found in Roger Adams' lecture and in the review which the author in
1941, contributed, upon the La Guardia Committee's invitation, to that
committee's book published in 1944.
t "SAR" (relationship between chemical structure and biological activity)
has found entrance into the American literature as a concise
symbol suggested (Fed. Meetings, 6 :352, 1947) for that phrase so
difficult to compress into a brief expression, and will be used in that
sense (Structure-Activity Relationship) hereinafter.
Originally published in German in Arch. Exper. Path. u. Phurmakol., vol. 2 11,
1950, pp. 175-193. Translation courtesy of Dr. Carl C. Pfeiffer, New Jersey
Bureau of Research in Neurology and Psychiatry, Princeton, New Jersey.
That the active substances are contained in the so-called
active principles
of hemp. "Red Oil," a high-vacuum distillate from cannabis extracts,
had been known for many decades. Only in 1937 to 1942,
however, was it ascertained by the work of an American
research team* that the Red Oil consisted essentially of two
inactive substances, cannabinol (CZ, H,, 0, ; III) and cannabidiol
(CZl HSO 0, ; I), and a varying mixture of tetrahydrocannabinols
(C,, HSO 0, ; THC; II a, b, and c), the representatives
of "cannabis activity."t The conclusion of this first
research phase was marked by a detailed description of the
first and apparently maximally potent natural agent, a THC
from Indian hemp resin, designated as charas tetrahydrocannabinol;
its potency (P; cf. below) was 14.6.
Motor ataxia in the dog was the test reaction which served
as a guide in the procedures to isolate the active compounds.
It served to determine their content in the starting materials,
to fractionate the crude oils, and to assay the potency of the
purified substances. That the research plan resulted in success
is due to two circumstances, namely, the choice of the right
test response and its adaptability to quantitative purposes. It
is now evident that motor ataxia is the one action in
experimental animals which closely parallels psychic cannabis
action in man; although this action is not accessible to
evaluation by the customary biostatistical procedures because
of the excessive variation of inter-individual sensitivity of the
dog, a biostatistically adequate assay method could be developed
by the introduction of a new principle of "intraindividual
potency comparison."
The history of the discovery of the cannabis-active substances
is characterized by the following paradox. On the one
hand, those investigators who followed analytical procedures,
by employing tons of starting material, large-scale molecular
* H. J. Wollner, J. R. Matchett and J. Levine, Narcotics Laboratory,
U.S. Treasury Department, Washington, D. C.; Roger Adams, et al.,
Department of Chemistry, University of Illinois; and this author.
t This is the term to be employed hereinafter for the characteristic
psychic action and its equivalent in the animal experiment, which in
some countries are designated as "hashish action," in others as "marijuana
action."
distillation, and chromatographic and elution methods
rapidly got hold of numerous, obviously chemically pure
active substances but were unable to identify them structurally,
particularly since even today they foil all attempts at
crystallization. On the other hand, the discoverer of the two
inactive by-products predicted the structure of the plant
principles without going into an analysis of any one of the
natural agents. As early as 1930-33, Cahn in England had
obtained cannabinol as a major component of the hashish
resin formed as a surface excretion on Oriental hemp, and
had identified it correctly except for minor inaccuracies in
the positions of the hydroxyl and the amyl group. In 1939,
Adams demonstrated it to be a 1-hydroxy-3-n-amyl-6,6,9-trimethyl-
6-dibenzopyran and at the same time prepared it
synthetically. Cannabidiol, of which hashish resin contains
only minimal quantities, was discovered in 1940 by Adams in
extracts from freshly harvested North American hemp as a
similarly considerable part (33 percent) of the crude oil,
identified as menthadienylolivetol, and synthesized. The
chemical relationship between the two substances, their
inactivity, and the age differences of the raw materials gave
rise to the ingenious hypothesis that cannabidiol signified the
starting material and cannabinol the end-product of a phytochemical
conversion process and that the intermediates on
the way of this process, namely, hydroaromatic cannabidiolisomeric
precursors of cannabinol, may be the searched-for
active principles.
This was confirmed on various avenues of approach: (1)
Adams, by intramolecular condensation of cannabidiol,
obtained-depending on the process of isomerization
employed-two such semi-synthetic tetrahydrocannabinols
(II b and c), both of which Loewe found to be very markedly
cannabis-active; the two isomers differ in optical rotation,
position of the double-bond, and potency (7.3 and 8.2). (2)
At once, the oily substances which Wollner and Loewe and
their associate? had isolated from hashish and American
hemp and found to be highly potent could now be identified
as THC. (3) Soon, Adams mastered the difficult synthesis of
another THC (IIa) from 5-methylcyclohexanon-2-carboxylate
and olivetol, the much lower potency of which was from
then on employed as standard (P = 1). (4) Part of the
presumable processes which in nature create the active
substances could be imitated in vitro; in an inactive cannabidiol
synthesized by Adams, which had been irradiated
with ultraviolet light, active substances could be demonstrated4
in an amount which indicated that about 2 percent
had been converted to THC.'
Like the synthetic cannabidiols, the completely synthetic
THCs are racemates, whereas the natural and the semisynthetic
THCs are laevorotatory. In 1942, Adams and his
associates dissolved the synthetic rat. THC into its two
optical isomers (rot. +152 and -114, respectively), and here,
too, the I-isomer (P = 1.66) turned out to be superior to the
d-isomer (P = 0.38).*
Stereo- and optical isomerism make possible a large num- Content of
ber of isomeric THCs which differ in the spatial arrangement ~~~r$&,,,e,
of the 9-methyl group and of the planes of the hydroaromatic
ring, and in the position of the double-bond in this
ring. Probably all these isomers are cannabis-active, though
they differ greatly in potency, and quite a number of them
has been prepared from hemps of different origin.4 The
potency of fluid extracts from hemps of different origins also
varies quite markedly. Botanically, however, it is noteworthy
that on the one hand fluid extracts from hemp cultivated in
Rumania, Manchuria, Italy, Tunis and climatically very different
parts of U.S.A. were cannabis-active (P: max. 0.52 in
Tunisian hemp, min. 0.003 in an American hemp).4 On the
other hand, a hemp grown from Oriental seeds on an
experimental field near Washington, D.C., and re-seeded there
for three subsequent years maintained its high potency.4 It is
therefore quite possible that the composition of the active
fraction constitutes the only difference between various
varieties of Cannabis sativa. American hemp appears to be
characterized by THCs of P 6.0 to 7.3 and 8.0 to 9.5, which
may be identical with Adams's semisynthetic THCs from
cannabidiol. In Oriental hemp, THCs of P 12.0 to 14.6
occur, in which the position of the double-bond in the
hydro-aromatic ring is as yet undefined and which have not
yet been obtained from other hemps. At closer analysis, the
cannabinol component of the hemp plant proved also to be
cannabis-active; completely synthetic as well as natural,
frequently re-crystallized cannabinols had the same potency
of 0.049 , which obviously is only of academic interest. The
hexahydrocannabinols are also active4 and so are presumably
the dihydrocannabinolst which can be assumed to be intermediate
products in the conversion of THC to cannabinol;
but none of these have as yet been demonstrated in hemp. In
the plant, the cannabinols appear to occur in part as esters of
aromatic acids.'7C
The content in active substances is not limited to the
"flowering tops of the male plant," but occurs in many parts
of the plant; for example, even seedlings of a few centimeters
in height4 and seeds6b yielded active extracts.
Synthetic cannab- According to the preceding data, cannabinol and all reducinols
and S~XUC- tion products
ture-Activity
of its toluene ring can be considered to
Relationship embody cannabis activity, and, in view of the activity of the
(SAR) parent substance, insignificant as it may be, it appears
chemically and pharmacologically justified to designate the
new class of chemicals as the class of cannabinols. After the
gates had been opened by the disclosure of the natural agents
and of synthetic procedures, the already mentioned representatives
of the class were joined by many other compounds
which served in our studies of SAR. Also to a team of British
investigators39 4 who had already for quite a time devoted
themselves to the problem of the Cannabis drugs, that gate
opened up a field of successful and, in part, independent
chemical research whose results, unfortunately, for reasons to
be discussed below did not become serviceable to the comparative
study of cannabis activity in its proper sense.
Information on SAR of the class was largely obtained by
measurements of the ataxia potencies. It was demonstrated
that the I-hydroxyl group is an important component of the
phenol ring of the THCs; its effective blockade greatly
diminishes activity. 3y 4 Also significant is the position of the
substituents of this ring. For example, the 1-n-hexyl-3-
hydroxy isomer of parahexyl, * the 3n-hexyl homolog of the
completely synthetic THC, has a P of only 0.05-t; i.e., only
one-fortieth that of the substance of comparison. The role of
the hydro-aromatic ring was studied in many analogs of the
"synthetic THC" (P = 1). Lack of the 9-methyl group brings
P down to 0.13 ; its substitution by ethyl, to 0.22 ; transfer to
position 10, to 0.25; to position 8, to 0.14; an additional
methyl group in position 7, to 0.75, but in position 8, to
0.11; replacement of the 6, 6-methyls by ethyls reduces P to
0.12, by propyls to 0.04. All such alterations at the cyclohexene
ring, as well as opening of the ring so as to leave in its
place a 6-n-butyl and an 11-methyl group (P = 0.04),
reducing it to an 11-methyl group 0.033) or replacing it by
cycloheptan (0.21) reduce but do not completely abolish
activity.
Not until the significance of the 3-alkyl side chain was
studied were increases in potency observed. The changes of P
with changes in the length and arrangement of this side chain
are presented in condensed form in Figure 1. Essential facts
on SAR are as follows: (1) Shortening of the 3-n-amyl chain
decreases potency.4 (2) Peak-effective, however, in the series
of homologs of the "synth. THC" standard (Series A of
Figure) is not this n-amyl compound, but its 3-n-hexyl
homolog (parahexyl).4 which is about twice as potent (1.8),
yet far inferior to the natural THCs. The same relation
repeats itself in the other homologous series (C, D; cf. below)
having an unbranched side chain." (3) Branching of the side
chain can result in substances of far greater P, but in these
series (E, F, and G" -15 ) peak potency is associated with a
six- but not with a nine-C-atom side chain. (4) Optimum
effectiveness was always found in side chains branched near
to their "root"; for example, among the methyl-amyl isomers,
P decreases consistently from the value 3.65 of the
1'-methylamyl-R (Series E) with transfer of the methyl group
into position 2' (H; 1.58), into 3' (K; 1.25), and 4' (L; 1.14).
(5) Methyl branching under otherwise equal conditions,
appears to grant the greatest activity [compare in Series E
l'-methylamyl-R (3.65) with 1'-ethylbutyl-R (M; 1.68), and
l'-isopropyl-propyl-R (N; 3.18), and in Series G l', 2'-dimethylbutyl-
R (3.84) with l'-ethyl-2'-methylpropyl-R (P;
3.40)]. (6) Double branching at the same C-atom appears not
to offer any advantage over single branching (compare the
two Series E and F). In contrast, two methyl branches at two
adjoining positions 1' and 2' (Series G) result in representatives
of outstanding potency which is only diminished by
prolongation of the branched chain (Series P) or by addition
of a third branch (Q).*
The role of the third (pyran) ring is illustrated by the
observation that analogs of the hypothetical type IV, prepared
from pulegon and appropriate resorcinol derivatives,
are in tetra- as well as in hexa-hydrated from (Series C and D,
respectively) approximately as potent as the corresponding
cannabinols. These compounds may differ from the cannabinols
in the mutual position of the two carbon-rings and
in the structure of the heterocyclic ring. 1'-methylation
appears not to increase potency in this series (cf. R).
Naturally, the problems of SAR are by no means consummated
by the comparative bioassays of this scanty hundred
of compounds hitherto studied. Many questions, among them
that of the applicability of experiences in one series to other
series, e.g., to homologs of the natural THCs, are still open.
The role of the configuration of the 3-side-chain could
certainly be further elucidated by the study of another
hundred congeners; for example, substances have been
synthesized in which the 3-alkyl chain is linked to the phenol
ring by an ether linkage I6 - "' ; if it were permissible to draw
conclusions from the results of the Gayer test, a weak
Cannabis activity would have to be ascribed to some of these
ethers.'7a
Only recently, in a study of 2-cyclohexyl-3-hydroxyl-6, 6,
9-trimethyl-7, 8, 9, lo-tetrahydro-dibenzopyran," we found
that cyclic arrangement of the side chain is superior to a
straight and unbranched side chain (compare S in Figure 1;
potencies of the corresponding n-alkyls are below the frame
of the graph)-a finding which points to new possibilities.
The substances of the cannabinol class are viscous oils, Conditionsof
some solid at room temperature, and extremely poorly ~t~~,?~'~
water-soluble. Even in the best solvents for cannabinols
(acetone, ethanol, glycols), certain representatives with long
side chain dissolve only slowly and yield only colloidal
solutions. Accordingly, all cannabinols are difficultly absorbed;
even after intravenous injection, thirty to sixty
minutes and often more elapse until peak effect is attained,
and the peak effect persists for hours-even for five days, as
was observed after a single mediumsized intravenous dose of
the l'-methylnonyl representative of Series E. Oral administration
requires twenty to thirty times the intravenous dose,
and subcutaneous administration hardly ever produces a
demonstrable effect. Even after high doses, we were as yet
unable to demonstrate urinary excretion of active substances."
In the blood, however, 1.6 to 10.5 percent of
parahexyl given by vein were found circulating after two to
three hours and 2.4 percent after eleven hours; the lungs of a
dog contained 0.9 percent of a dose injected twenty hours
before.18 Probably because of these physicochemical characteristics
of the compounds, the margin of safety is enormous;
per kg body weight, the oral L.D.,, of Charas THC in the
mouse is more than two hundred thousand times the intravenous
threshold dose for ataxia in the dog and more than
forty thousand times the oral threshold dose for psychic
action in man."
Not all the actions of crude hemp preparations or oil
redistillates reported in the literature could be reproduced
with pure active substances. At that, many of them are of
minor importance at the present state of the Cannabis
problem. This holds true, for instance, for the effects upon
heart rate, respiration and pupil. Therefore, the following
discussion will be limited to those, in part newly revealed,
sections of the large action spectrum, which have been
studied with the aid of pure substances.
1. Psychic Actions. Analysis of the psychic actions in
man, still more so in animals, is a task as yet unsolved. Even
with pure substances, neither the host of experimental
psychologists of the LaGuardia Committee" nor the effects
with a "battery of psychological tests" both elsewhere** and
in test persons in our department*' were capable of disclosing
more than a large range of variation of all phenomena
tested. Thus, the availability of numerous pure substances of
the class has not as yet given impetus to the analysis of the
unique "pleasure action," but a few experiences have been
collected which deserve to be mentioned partly as modest
beginnings, partly with a view to therapeutic aspects. For
instance, both parahexyl** and compounds "RA 122" and
"RA 125A"*l were found to have no effect upon the
normal human EEG or upon musical appreciation.22b Therapeutic
effects of parahexyl have been reported in depressive
states23 a but in consideration of the negative experiences of
Pond23b and of unpublished ,personal reports to the writer
they are in urgent need of further verification. Charas THC
has been extensively employed for "psychic relaxation" and
i s reported24 to be equivalent to the customary barbiturates
for purposes of narcoanalysis, narcocatharsis and
narcosuggestion. The idea, however, that Cannabis and morphine
may be comparable in euphorizing effects has not
proved fruitful in its application to morphine withdrawal;
withdrawal symptoms were neither shortened nor abated by
parahexyl treatment.2sa Noteworthy, though difficult to
evaluate, are reports on psychiatric episodes, mostly in
psychopaths, occurring in some instances under prolonged
treatment with, and in other instances upon withdrawal of
cannabis-active drugs .25 a -g
2. Cataleptic Effects. They are never missing in the picture
of the hemp "jag," always accompany the ataxia effects in
the dog, and are produced by all ataxia-active pure substances4*
19 They can be seen in many species of animals, and
a catalepsy test has been describedi which occasionally has
been found quite serviceable.
3. Ataxia Action. This has gained ever-increasing prominence
among all cannabis actions, as an experimental tool. It
is not only highly serviceable,* but is closely correlated with
psychic action. This has been confirmed for all those crude
and pure products which have been both bioassayed by the
writer and evaluated in man20p24 ; particularly good agreement
was found between ataxia and psychic potencies of
four cannabinols at the occasion of their especially careful
and expert evaluation by Dr. L. Kubie. There are only vague
notions about the locus and mechanism of the ataxia action.4
In view of its outspoken parallelity to the psychic action, a
close correlation in the mechanisms of the two actions has to
be considered seriously. As a matter of course, final proof of
both identity in mechanism and the quantitative parallelity can
only be expected from an evaluation in man of all the
compounds available.
4. Central Stimulant Actions were observed with all substances
of the class. From medium degress of ataxia upward,
dogs at times exhibit convulsive motor symptoms.4F l9 They
are possibly related to the compensatory counteractions
against the disturbance in coordination; their frequency
varies with the dog's individuality and strain. Indisputable
and unique convulsions appear consistently after doses
greater than twenty times the threshold ataxia dose after
intravenous administration of the high-potent "RA 122" and
dominate the picture of action of lethal doses. Mice often
become pugnacious after medium doses of cannabinols."
Vomiting occurs in dogs rather regularly and late, usually one
to five hours after injection, following doses upward from a
medium-effective intravenous dose. The emetic effect as well
as the hyperexcitability of the scratch reflexes can probably
be ascribed to a central stimulant action.4' l9
5. Hypnotic Activity. Elements of central nervous stimulation
among the actions of hemp are of interest in view of
the question of its soporific action. Probably cataleptic
symptoms in man have often been interpreted as signs of a
sedative action. In the dog, concomitance of catalepsy with
higher degrees of ataxia which prevent upright posture results
in a syndrome which might impress a superficial observer as
sleep. In experiments with combinations of aqueous hemp
extracts and hypnotics, Burgi believed he had demonstrated a
hypnotic component of Cannabis action.27 We were able to
duplicate his observation with fluid extracts of hemp; they
prolonged sleep duration in mice after a pernocton dose
which is just enough to suppress the righting reflex. Sleep
prolongation, in our experiments, was significant indeed, but
only for this one hypnotic; however, even the consideration
prolongation by a large dose of Cannabis equals no more than
the effect producible by an additional 10 to 20 percent of
the pernocton dose employed. Above all, however, a study of
pure substances gave proof that this action is due not to any
one of the cannabis-active components, but to the otherwise
inert cannabidiol.28 * Moreover, the natural as well as the
synthetic cannabis-active substances lack any other indications
of central nervous depressant action. Even in states of
severest hypomotility and impairment of attention, up into
the terminal phases of lethal effect, the animals, quite
contrary to those under the influence of hypnotics, still
respond, with frustraneous movements to moderate acoustic,
tactile, or pain stimuli. This is in accordance with the
characteristic criterion of Cannabis action both in dog and in
man, namely, that the drugged individual can readily be
diverted by environmental stimuli.4
6. Cornea1 Areflexia. Abolition of the wink reflex in
rabbits after administration of hemp extracts or crude oils
was discovered in 1928 by Gayer in Straub's laboratory and
considered to be a faithful expression of Cannabis activity.29
First of all, at closer analysis this systemic action has proved
inappropriate for quantitative purposes because of the great
variation in inter- and intraindividual susceptibility. Only
after a tedious search in a large stock of animals by way of
numerous recalibrations can one find a number of individuals
suitable to yield conclusive data in strictly intraindividual
potency comparisons4> 3o that there are two types of areflexia
producing agents, namely, (a) ataxia-active THCs which
lose neither in ataxia nor in areflexia activity by such
oxidation, and (b) other extractives possessing little or no
ataxia activity, but a marked areflexia activity which can be
destroyed by oxidation. In the search for the as yet unidentified
substances of this type (b), the Gayer test in its
present30 more complicated but more reliable form is thus
the test reaction of its choice.
7. Anticonvulsive Action. It was only by the aid of
the highly active synthetic cannabinol congeners and due to
the methods elaborated in this laboratory by Goodman and
Toman and associates3' that the antiepileptic activity of
cannabinols could be disclosed and specified. All cannabinols
tested were effective. Their activity is of the type of anticonv&
ant activity of diphenylhydantoin. Like this drug, they
abolish the tonic hind leg extensor component in the pattern
of supermaximal electroshock in rats and cats33 (and presumably
also in man34), whereas they prolong the duration of
metrazol convulsions, modify their appearance and display a
strong lethal synergism with metrazol.33 The cannabinols
differ from diphenylhydantoin and its congeners by exhibiting,
in their maximum-potent representatives, much greater
potency and, in reference to lethal effectiveness an incomparably
greater margin of safety. When referred to psychic
side effects, the therapeutic indices of the various agents vary
considerably. In a preliminary experiment,35 "RA 122"
lacked the ability of diphenylhydantoin, discovered by
Toman et a1.,36 to raise the threshold of the isolated nerve
preparation for electrical stimulation and to prevent the
repetitive discharge elicitable by immersion of the nerve in
neutral isotonic phosphate solution; it is true that, in view of
the minimal water-solubility of the cannabinols, the result is
inconclusive.
Alteration of the electroshock pattern points to anti-grand
ma1 activity, suppression of metrazol convulsions to anti-petit
ma1 activity. 32 Accordingly, clinical experiments were undertaken
with some of the agents. A first series3' in diphenyl
hydantoin-refractory grand-mal epileptics proved noteworthy
effectiveness and absence of psychic side effects when "RA
122" was given orally for several months in daily doses of
one mg or less, or the weaker isomer, "RA 125A," in
somewhat higher dosage. A second series3' consisted of five
institutionalized children with severe grand-n-al epilepsy and
mental underdevelopment, in whom daily doses of 0.13 gm
phenobarbital combined with 0.3 gm diphenylhydantoin or
0.2 gm Mesantoin had proved inadequate. "RA 122," in daily
doses of 1.2 to 1.8 mg, was in three children "at least as
effective" as prior therapy; the fourth became almost free
from attacks and the fifth completely free. Following transfer
to 4 mg of "RA 125A," the attacks remained infrequent
in one patient, the other one suffered exacerbations. The first
patient had a brief paranoid episode, similar to others he had
repeatedly experienced prior to the cannabinol therapy. In
contrast to their ineffectiveness upon the normal EEG the
cannabinols normalized the EEG of grand-ma1 patients.
The protective activity of cannabinols against electroshock
and grand-mal attacks is of theoretical interest in view of the
apparent absence of structural relationship between the
cannabinols and the antiepileptics from the classes of hydantoins,
barbiturates and oxazolidinediones. The practical evaluation
of cannabinols as antiepileptics, notwithstanding their
superior potency and persistence of action, will largely
depend upon the problems of their side effects which will be
discussed below.
8. Analgesic Action. It was probably the conceivable tendency
to compare every euphorizing drug with. morphine
regarding anodynic properties, which brought British investigators
to take up anesthesia tests and to ascertain the
intravenously injected hashish extracts were somewhat effective,
and that some cannabinols of the 1-hydroxy-3-alkyl-R
and one of the 2-alkyl-3-hydroxy-R type were considerably
effective (see Table 1). Using a modification, introduced in
this department by Nickerson, of the customary methods of
rat-tail heat stimulation which had also been employed in the
afore-mentioned studies, we observed an even greater effectiveness
in two other cannabinols.
9. Lethal Action. As already mentioned, large doses are
required for a lethal effect, which are not always available
and are not easy to administer because of the poor solubility
of the substances. Data for quite a number of older preparations
(previous to 1947) have been presented previously.", l9
Only some particularly illustrative observations need be
reported here. The values of intravenous L.D.,, in the
mouse4 indicate that the lethal toxicity of pure substances is
markedly lower than that of the crude preparations. The
necessity of employing solvents such as propylene glycol,
which are by no means indifferent, and the minimal watersolubility
of the drugs are probably responsible for the poor
reproducibility of the lethal-dose values. Such shortcomings
are less important than the experience that the lethal
mechanism of the same substance can be different with
different routes of administration. The late death of dogs
after oral administration of parahexyl is associated with
profuse intestinal hemorrhage, after intravenous injection
with severe pulmonary edema. In contrast, incomparably
smaller intravenous doses of "RA 122," which is ataxiaeffective
even in doses of a few micrograms, cause death
associated with convulsions within a few hours, obviously
owing to some primary central nervous mechanism. This
suggests the interpretation that the lethal effect of the
lower-potent cannabis-active substances is due to non-specific
mechanisms and that only that of the high-potent agents
originates from a mechanism more closely related to the main
activity, which in the low-potent substances is masked or
outdone by non-specific toxicity.
10. Habituation. The clinical authors of the La Guardia
Committee's report, on the basis of observations in chronically
marihuana-smoking prisoners, deemed the danger of
habituation negligible. 2o This has evoked a vehement controversy
.40 The most lucid answer to the problem of hemp
habituation will be found with Goodman and Gilman.4'
Conclusive animal experiments are not available. For its
curiosity value, the "beginner's habituation" may be mentioned:
The more than one hundred fifty dogs from the
writer's twenty-six hundred bioassay experiments, some of
which had been tested twice per week for many years, almost
invariably exhibited a certain decrease in sensitivity during
the period of the three to four initial experiments, and from
then on maintained a rather constant susceptibility. This
habituation was the more pronounced, the less the individual's
postural behavior was dominated by tenseness. Accord
ingly, the apparent decrease in susceptibility seems to be due
to the "learning" of postural responses compensating for the
ataxic incoordination, at the mercy of which the individual
finds itself quite helpless during its first experiences in ataxia.
All other intraindividual variations in susceptibility, as they
developed over long periods, were completely irregular,
varied in intensity, and consisted in increases as well as
decreases; in bitches, they were not without relation to sex
cycles and pregnancy, for which reason all assays were
conducted in males.
Interrelation It has been repeatedly emphasized above that up to now
between different
cannabinol actions.
all observations have pointed to an even quantitatively close
relation between the psychic and the ataxia activity. The
ratio between psychic potency in man and ataxia potency in
dogs of an individual substance is a constant which varies
very little in the entire class. Contrariwise, it has been
demonstrated above that neither of these two activities has a
constant relationship to lethal activity.
That the cornea1 areflexia action is not correlated with the
ataxia action, is evidenced by the fact that in crude hemp
products the two activities can be dissociated by oxidation. It
is true that not only the unknown cannabis-inactive substances
disclosed by that procedure, but also ataxia-active
cannabinol possess areflexia activity. However, the potency
ratio between the two activities varies from substance in a
wide range, as has been shown earlier4 for a series of
cannabinols and is confirmed in Table 1 in nine, partly new
substances. A similar inconstancy of the potency ratios is also
evident when analgesia and ataxia activity are compared
(Table 1, column 5).
The relation between antiepileptic (anti-electroshock) and
ataxia activity is also marked by a certain though lesser
inconstancy of the potency ratios (Table 1, column 4). Here,
however, the results should be evaluated with greater reserve.
For the electroshock experiments, for better comparability
of the cannabinols with other antiepileptics, were performed
after oral administration of aqueous, lecithin-homogenized
emulsions of oily drug solutions, whereas the potency values
for ataxia and all other actions were determined after
intravenous administration. Since in these comparisons numerator
and denominator of the potency ratio originate from
experiments with different routes of administration, it is
possible that inequality of the ratios is due to differences in
absorption, distribution and detoxification. A comparison of
the E.D.,, values for cornea1 areflexia of the British investigators
with the writer's ataxia values indicates that in some
cannabinols the areflexia dose is smaller than the ataxia dose.
That makes it tempting to base all evaluation upon the Gayer
test which, superficially judged, is more convenient than the
ataxia test. For SAR studies and for the assay of cannabis
activity of crude preparations and unknown mixtures, this
must be urgently discouraged because the potency ratios are
so inconstant. Quite inversely, the analgesic dose of all
cannabinols and the anti-electroshock dose of some are high
as compared with the ataxia dose. In such substances, each of
the two therapeutic effects may be obtained only at the price
of considerable psychic side effects. Were such an unfavorable
potency ratio common to all cannabinols, the outlook
would be small indeed that further search could reveal other
substances of the class more devoid of these side effects. The
inconstancy of the ratios leaves the possibility open that
there are cannabinols of a greater therapeutic index both for
analgesia and for epilepsy therapy. Actually, available experience
already teaches that in clinically equieffective doses the
only two cannabinols which have as yet undergone clinical
examination are not equally liable to produce psychic side
effects : "RA 122" not only exerted greater therapeutic
action than "RA 125A" but also had a smaller incidence of
psychic Cannabis effects. That the two substances differ
clinically with regard to the risk of side effects agrees with
the fact that the experimental potency ratios of different
cannabinols differ; that "RA 122" has fewer side effects than
"RA 125A" is contrary to what one would have expected
according to the size of the two potency ratios and may thus
indicate that ratios of experimental potency values which are
obtained after administration by different routes are not the
last word on this subject.
Theoretically, the dissociation between psychic activity on
the one hand and analgesia, cornea1 areflexia and antiepileptic
activity on the other hand, as demonstrated by the
inconstancy of the respective potency ratios, is probably an
indication that the actions compared must be ascribed to
different reactive groups or, more generally expressed, to
different parts of the total configuration of the same drug
molecule.
Summary. 1. This review attempts to present some evidence for the
proemial statement that our knowledge of the cannabis drugs
has undergone a striking change.
2. After a report on the isolation of the natural Cannabis
agents and their identification as tetrahydrocannabinols, their
subsequently studied synthetic isomers and analogs are surveyed
and some data are reported on the structure-activity
relationship in this new class of chemicals, the cannabinols.
3. The pharmacology of the cannabis-active substances,
as revised with the aid of a study of the pure substances, is
briefly presented, and examples are given for the decisive
elucidation obtained by the study of the most recent,
highest-potent synthetic agents, the potency of which was
found to be up to seventy times the average and up to
thirty-five times the maximum potency of natural tetrahydrocannabinols.
4. Some details and therapeutic trials of the psychic
cannabis action are reported.
5. The significance of the ataxia action in the dog, which
appears to parallel closely the psychic action in man, for
identification and bioassay of cannabis-active compounds is
illustrated and, contrariwise, the cornea1 areflexia in the
rabbit is discussed as a quite detached property of some of
the hemp products and congeners, which is only of very
limited usefulness.
6. The old question of the hypnotic activity of the
Cannabis drugs is answered in the negative from a study of
the pure substances; closest to hypnotic activity is the
practically insignificant, very limited capability of crude
preparations to prolong the hypnotic action of certain barbiturates,
an activity which is absent in the pure cannabisactive
substances and embodied only in an otherwise inert
by-product, cannabidiol.
7. The analgesic activity of cannabinols, only recently
disclosed and not yet tested for its practical applicability, is
briefly discussed.
8. In their anticonvulsant activity, also an only recently
discovered property, the cannabinols are demonstrated to be
closely related to diphenyl-hydantoin, according to both
experimental criteria and clinical experiences; the most
potent synthetic congener of the cannabinols appears to
possess more than one hundred fifty times the anti-grand ma1
potency of diphenyl-hydantoin.
9. The problem of the therapeutic indices of the five
major actions and that of their mutual relations in the roles
of main and side actions are dealt with on the basis of
experimental data; the potency ratios hitherto established do
not preclude the possibility that continued search of the new
class of compounds will lead to congeners having a still
greater margin of antiepileptic effectiveness than the as yet
best examined "RA 122," and perhaps also to congeners
having an adequate therapeutic index of analgesic activity.
10. Contributions are presented to the problems of habituation
and of the mechanism of the lethal action; as a
menace of habituation, addiction and tolerance the cannabisactive
drugs appear to rank lowest among the narcotics, and
according to the margin of safety some of them rank
uppermost among all drugs.
REFERENCES
Marijuana: Medical Papers
1 . Walton: Marihuana. Philadelphia-London 1938.
2 . Blatt: J. Washington Acad. Sci. 28, 465 (1938).
3 . Adams: Harvey Lectures, Ser. 37, 168 (1941-42).
4. Loewe: Studies on the Pharmacology of Marihuana. In: The
Marihuana Problem in the City of New York. Lancaster 1944.
5. Wollner, Matchett, Levine & Loewe: Jour. Amer. Chem Sot.
64, 26 (1942).
6a. Loewe: Jour. Amer. Pharmaceut. Ass. 28,427 (1939).
6b. Matchett & Loewe: ibid. 30, 130 (1941).
6c. Loewe: Science 106,89 (1947).
7 . Levine: Jour. Amer. Chem. Sot. 66, 1868 (1944).
8. Adams, Smith & Loewe: ibid. 64, 2087 (1942).
9 . Loewe: Science 102, 615 (1945).
1 0 . Madinaveit,ia, Russell & Todd: Jour. Chem. Sot. 1942, 628.
11. Avison, Morrison & Parkes: ibid. 1949, 952.
1 2 . Adams, Chen. & Loewe: Jour Amer. Chem. Sot. 67, 1534
(1945).
1 3 . Adams, Aycock & Loewe: ibid. 70, 662 (1948).
1 4 . Adams, Mackenzie & Loewe: ibid. 664.
15. Adams, Harfenist & Loewe: ibid. 71, 1624 (1949).
1 6 . Alles, lcke & Feigen: ibid. 64, 2031 (1942).
17a. Bergel, Morrison, Rinderknecht, Todd, MacDonald & Woolfe:
Jour Chem. Sot. 1943,286.
17b. Bergel &Wagner: Ann. d. Chem. 482, 55 (1930).
18. Loewe: Jour. Pharmacol. and Exper. Therap. 86,294 (1946).
1 9 . Loewe: ibid. 88, 154 (1946).
20. The Mayor's Committee on Marihuana, Clin. Study in: The
values referred to Parahexyl. - c: = Anti-electroshock potency, according to
unpublished experiments of Dr. E. A. Swinyard. - d: This value results from the
author's own experiments (analgesia comparisons after oral administration),
whereas all other values of this column are calculated from the analgesia doses
which Avison C.S. l1 determined after intravenous injection. Hence this value is
not directly comparable with the other values of this column, but this is of no
significance for the overall outcome. - e: Loewe, unpublished. - f: s.e. as
percentage of mean.
Marihuana Problem in the City of New York. Lancaster, 1944.
21. Davis: Unpublished.
22a. Williams, Himmelsbach, Wikler, Ruble & Lloyd: Publ. Health
Reports 61, 1059 (1946).
22b. Aldrich: ibid. 59,43 1 (1944).
23a. Stockings: Brit. Med. J. 1947, 918.
23b. Pond: J. Neurol. 11, 271 (1948).
24. Kubie & Margolin: Psychosomat. Med. 7, 147 (1945).
25a. Himmelsbach: South. Med. J. 37, 26 (1944).
25b. Fraser: Lancet 1949, 747.
25~. Chopra, Chopra & Chopra: Indian J. Med. Res. 30, 1 (1942).
25d. Marcovitz & Myers: War Medicine 6, 382 (1944).
25e. Bromberg: J. Am. M. A. 113,4 (1949).
25f. Reichard: Fed. Probationer, Oct.-Dec. 1946, 15.
25g. Allentuck & Bowman: Amer. J. Psychiat. 99, 248 (1942).
26. Loewe: Federation Proc. 6, 352 (1947).
27. Gisel: Z. f. exper. Path. u. Ther. 18, 1 (1916). Burgi: (German).
med. Wschr. 1924, Nr. 45.
28. Loewe: J. Amer. Pharm. A. 29, 162 (1940).
29. Gayer: Arch. exper. Path. u. Pharmakol. 129, 312 (1929).
30. Loewe: J. Pharmacol. and Exper. Therap. 84, 78 (1945).
3 1. Alles, Haagen-Smit, Feigen & Dendliker: ibid. 76, 21 (1942).
32. Goodman & Toman: Proc. First Nat. Med Chemistry Symp. of
the Amer. Chem. Sot. 1948, June 17-19, 93; Toman & Goodman:
Physiol. Rev. 28, 141 (1948); Goodman, Toman & Swinyard: Arch.
Internat. de Pharmacodynamie 78, 144 (1949).
33. Loewe & Goodman: Federation Proc. 6, 352 (1947).
34. Toman, Loewe & Goodman: Arch. Neurol. a. Psychiat. 58, 312
(1947).
35. J. E. P. Toman: Personliche Mitteilung.
36. Toman, Greenhalgh, Carlson & Bjorkman: Federation Proc. 8,
284 (1949).
37. Belknap, Ramsey & Loewe: Unpublished.
38. Davis & Ramsey: Federation Proc. 8, 284 (1949).
39. Davies, Raventos & Walpole: Brit. J. Pharmacol. and Chemotherapy
1, 113 (1946).
40. Bouquet: Jour. Amer. Med. Ass. 124, 1010 (1944). Bowman:
125, 376 (1944); (editorial) 127, 1129 (194.5); Walton: 128, 383
(1945); Bowman: 899. Anslinger: 1187; Marcovitz: 129, 387 (1946);
41. Goodman & Gilman: The Pharmacological Basis of Therapeutics,
New York, 2nd edition.
Source: The Active Principles Of Cannabis And The Pharmacology Of The Cannabiiols
