m

Toxieon, Vol. 31, No . 9, pp . 1099-111l, 1993 . 0041-0101/93 $6.00 + .00

Printed in Great Britain. 1993 Pergemon Press Ltd

PHARMACOLOGICAL STUDIES OF `SAPO' FROM THE FROG

PHYLLOMED USA BICOLOR SKIN: A DRUG USED BY THE
PERUVIAN MATSES INDIANS IN SHAMANIC HUNTING
PRACTICES
VrrroRio ERSP~' GIULIANA FALCONmu ERSPA~,' CINZIA SEvEum,' RosA LUISA
POTENzA,' DONATELLA BARRA 2 GIUSEPPINA MIGNOGNA 2 and ANTONIo BIANCHI4
'Institute of Medical Pharmacology III, 'Department of Biochemical Sciences, University of Rome `La
Sapienza', OWL Universitaria, 00185 Rome, Italy; 'Institute of Neurobiology, CNR, v.le C .Marx, Rome,
Italy ; and 'Department of Anaesthesiology II, General Hospital of Bergamo, Bergamo, Italy

(Received 17 February 1993 ; accepted 15 April 1993)

V. ERspAmR, G. FALcoNmu ERspAmER, C. SEvERIm, R. L. POTENZA, D.

BARRA, G. MIGNoGNA and A. BIANCHI. Pharmacological studies of `sapo' from
the frog Phyllomedusa bicolor skin : a drug used by the Peruvian Matses
Indians in shamanic hunting practices. Toxicon 31, 1099-1111, 1993 .-The
dried skin secretion from Phyllomedusa bicolor, `sapo', is used by the Matses
Indians of the Northern Peru, in shamanic rites mainly designed to improve
luck in hunting. When rubbed into burned, exposed areas of the skin, the drug
causes the prompt appearance of violent peripheral gastrointestinal and
cardiovascular effects soon followed by remarkable central effects (increase in
physical strength, heightening of senses, resistance to hunger and thirst, exalted
capacity to face stress situations) . All the peripheral and most of the central
effects of `sapo' can be ascribed to the exceptionally high content of the drug
(up to 7% of its weight) in potently active peptides, easily absorbed through
the burned, inflamed areas of the skin . The concentration in `sago' of the single
peptides (phyllocaerulein, phyllomedusin, phyllokinin, dermorphins and
deltorphins) has been determined by bioassay, and peptide contents were
correlated with the different symptoms of the `sapo' intoxication .

INTRODUCTION

OvER the past few years we have had the opportunity to study the biogenic amines and
particularly the active peptides in the skin of several hylid frogs from Central and South
America belonging to the Phyllomedusinae subfamily (ROSEGmm et al., 1986 ; ERSPAhlER
et al., 1986).
Particular attention was paid to Phyllomedusa sauvagei and P. hypochondrialis from
Northern Argentina, P. rohdei and P. burmeisteri from Central Brazil and P. bicolor from
the Peruvian and Brazilian Amazonas (ERSPAbiIIt et al., 1985).
We examined methanol extracts of both fresh and dried skins, with comparable results.
This material did not contain appreciable amounts of amines (except 10-20 hg bufotenine
per g dried skin in P. rohdei), nor bioactive alkaloids, but enormous amounts of a variety
of active peptides belonging to at least seven families : caeruleins, tachykinins, bradykinins,
1099
1100 V . ERSPAMER et al.

A Phyllomedusa bicolor SPECIMEN, WITH DIVARICATED LEGS, HOUND TO FOUR SMALL STAKES .
FIG . 1 .
The frog is ready for collection of sapo (bamboo stick). Photograph by Peter Gorman, with
permission (December 1990, Galvez River, near Angamos, Peru) .

bombesins, sauvagine (corticotropin-releasing hormone-like peptide), tryptophyllins and

opioid peptides, selective for either y- or 6-receptors (dermorphins and deltorphins) . The
pharmacological actions of these peptides have been described in a number of papers
(ERSPAMER and MELCHIORRI, 1983 ; ERSPAMER et al., 1985) .
In the course of our research on skin extracts of P. bicolor, our attention was attracted
by a report by GORMAN (1990), dealing with some shamanic practices by the Matses
Indians of the north-east corner of Peru, a subdivision of the Mayeruna tribe, in which the
use of the drugs of animal and vegetable origin played a prominent role .
One of these drugs, `sapo', was of particular interest because it consists of a dry, resinish
substance obtained from a large tree frog which the Matses call `dow kiet'. It is now well
established that the frog is actually P. bicolor (cf. DALY et al., 1992) .
Sapo is employed in magic hunting rituals designed to improve the hunter's luck. To
collect sapo the Matses catch the frog and keep it for 3 days during which time its legs and
back are periodically and gently scraped with a stick of bamboo (Figs 1 and 2).
To dry the secretion collected from the frog, the stick is placed in a leaf bag and hung
above the cooking fire between scrapings. The same stick is used over and over ; by the end
of 3 days the stick is covered with a yellow substance and the dow kiet is set free. The sapo
retains its potency for at least a year (GORMAN, personal communication) .
To use sapo, a Matses hunter first scrapes a bit of the dried resin from the stick, then
mixes it with saliva until it has a consistence somewhat like mustard. The recipient's arm
or chest is then burned with a smouldering twig, the burned skin is peeled away and the
liquified sapo introduced into the raw area, approximately the size of a matchhead .
 Frog Skin Peptides 1101

FIG. Z . BAMBOO STICK WITH THE SAPO MATERIAL USED IN THIS STUDY.

The effects elicited by the drug are described by GoRMA.N (1990), in a rather journalistic
style, as follows:
"The effects are astounding : the moment the drug is placed upon your skin your body
begins to heat up . In moments you feel as though you're burning from the inside ; you
begin to sweat. Your blood begins to race; your heart pounds . You became aware of every
vein and artery in your body and feel them opening to allow for the fantastic pulse ofyour
rushing blood. Your stomach cramps and you vomit violently. You lose the control of
your bodily functions; you may urinate and defecate and drool uncontrollably . You fall to
the ground and begin to lose consciousness; then suddenly, you may feel urges to do
things you've never done before . You might find yourself growling, barking or moving
about all fours. You feel as though animals are passing through you, trying to express
themselves through your body . But even this extraordinary feeling is secondary to the
speeding of your blood, a motion so fast that you think your heart would burst.
"For about fifteen minutes the rushing gets faster and faster, you are in agony. The pain
becomes so great that you wish you could die and get it over with, but you don't die. The
pounding slowly becomes steady and rhythmic and you gasp for air like a man saved from
drowning. Finally the pounding subsides and you're overcome with exhaustion . You
sleep.
"There are no dreams or visions with `sapo'; you may even wonder what it all was for
until you wake, then you are a god. Everything about you is larger than life, and your
physical strength is explosive. You can do without food for several days and run in the
jungle for hours without tiring . You can see in the dark effortlessly. You see animals
before they see you, and you sense which plants are benevolent and which are not. Every
TOX 31 :9-B
1102 V. ERSPAMER et al .

sense you possess is heightened and somehow in tune with the jungle, as though the `sapo'
put the rhythm of the jungle into your blood."
The violent visceral effects of sapo have been confirmed by GORMAN (personal com-
munication) in about ten self-applications of the drug, and correspond exactly to the
description by DOLE and CARNEIRO (quoted by FURST, 1974) .
It is suggested that the drastic cleaning out of the body (vomiting, diarrhoea, urination,
sweating) observed in the first phase of the sapo application, with alleged elimination of
`toxin', may have some `magic' effect in itself and may heighten the effects of other drugs
possibly taken prior to, or together with, sapo. Among other things, by cleansing the
body, Matses hunters would lose their `human' odour in the short time, making it easier to
approach and capture the prey .
According to the estimates by GORMAN (personal communication) the amount of sapo
rubbed into the normal two or three burned skin areas may be 20-30 mg (10 mg per area)
with maximum doses of 100 mg per day. Sapo is used by the Matses not only to help on
long hiking trips or in tapir trap setting (in this last case sapo is applied, morning and
night, for at least 15 days, on four or five burns in the forearm and chest), but also to
produce abortion (application on two burns to the interior of the vaginal labia) .
With regard to the central `magic' effects of sapo, things are perhaps more complicated.
It is possible, and in some cases certain, that before or after sapo, the Matses (especially
the Amahuaca) take other drugs such as `Banisteropsis vine' or `nu-nu' snuff, which may
have more or less potent hallucinogenic effects. Thus it is difficult to decide how much of
the ecstatic trance experience may be ascribed to the frog poison and how much to other
drugs. It should be stressed, however, that Gorman has experienced central effects, "to
feel like god", after application of sapo alone .
The purpose of the present study, using an original sample of sapo from the Matses,
was to establish: (a) the actual amount of bioactive compounds in sapo, especially amines
and peptides ; (b) the effects of the single active constituents or of a cocktail of them on the
cardiovascular system and viscera; and (c) the effects of the same constituents on the
central nervous system .
The results of this study would help us attempt to explain the two successive phases,
peripheral and central, of sapo intoxication .

MATERIALS AND METHODS

3apo material
A sample of sapo was placed at our disposal, in February 1991, by PemR GORMAN . The skin secretion was
obtained from a P. bicolor specimen weighing 72 g, captured by the Matses near the Galvez River, Peru, on 22
December 1990.
The material, resinous in aspect, weighed 126 mg . It was ground in a small mortar and then extracted with
30 ml of 80% methanol. After 2 days the supernatant was removed and extraction with a similar quantity of
80% methanol repeated twice, again for 2 days . The liquids were combined, filtered and kept at 2°C.
The exhausted sapo material, weighing 52 mg (41% of the total weight), was extracted first with 5 ml
petroleum ether and then, after removal of the solvent, boiled in 3 ml of 0 .1 M HCI, for 20 min. The methanol,
petroleum ether and acid extracts (the last two colourless) were submitted separately to bioassay .

Chromatography on alumina
An amount of the methanol extract corresponding to 50 mg sapo was evaporated and the residue, weighing
36 mg, dissolved in 50 ml of 95% ethanol. The liquid was loaded onto a chromatographic column packed with
45 g of alkaline alumina (Merck, Darmstadt, F.R .G.) and eluted using a stepwise gradient of aqueous ethanol:
19 steps from 95 to 10% ethanol. At each step two or three fractions of 50 ml were collected. The eluates were
kept at 2°C until used.
 Frog Skin Peptides 1103

Thin layer chromatography and colour reactions

Samples of the sapo methanol extract corresponding to 0 .5-2 mg drug were applied to silica gel plates and
developed in 80% ethanol : acetic acid (10 : 1 v/v) . The developed chromatograms were then sprayed with
Dragendorff reagent (for alkaloids) and the Ehrlich reagent, p-dimethylaminobenzaldehyde (for
indolealkylamines).

Bioassay
The occurrence and concentrations of the various active components of the sapo extracts were determined
using the following test preparations: caenrlein-like peptides: guinea-pig gall bladder; bmdykinin- and bombesin-
like peptides : rat uterus; tachykinins: guinea-pig ileum and rabbit colon; dermorphins: electrically stimulated
guinea-pig ileum; deltorphins: electrically stimulated mouse vas deferens ; and sauvagine: rat diuresis .
All the above methods have been described in detail in previous papers (ERspAmR and FALCONIM ERSPAMI?R,
1962; ERspAmm et al., 1972, 1980; BRoccARDo et al., 1981).
The guinea-pig gall bladder, rat uterus, mouse vas deferens and rat antidiuresis tests gave reliable qualitative
and quantitative results even with the crude sapo extract . However, the quantitative evaluation of the
tachykinins and dermorphins was possible only after separation by chromatography.
In a further set of experiments the crude methanol sapo extract as well as the petroleum ether extract were
injected intracerebroventricularly in rats, in order to assess the direct central effects elicited by the drug .

RESULTS

Bioassay
(1) Crude sapo methanol extract. The content of active peptides was as follows:
phyllocaerulein, expressed as caerulein (equiactive to the former) 30-35 lug per mg sapo ;
phyllokinin 181íg/mg; sauvagine 31íg/mg, finally, Ala-deltorphins, expressed as
Ala-dermorphnn I 5.31íg/mg . Phyllomedusin and the dermorphins, as already stated, could
not be determined in the crude extracts .
Intracerebroventricular injections in rats of the extract, in amounts corresponding to
20, 50, and 2001íg of drug per rat, produced only symptoms attributable to the opioid
peptides : increased locomotor activity with rearing, grooming and sniffing, characteristic
for the deltorphins (LONGo1VI et al., 1991), and analgesia, characteristic for the dermor-
phins (BROCCARno et al., 1981). Threshold doses ranged from 20 to 50 Pg ; with 2001íg,
effects were intense. No other obvious neurological or behavioural manifestations could
be observed .
(2) Petroleum ether extract. The residue of the extract, taken up in water (slightly
opalescent) did not display appreciable effects in any of our test preparations, up to
amounts of 0.3 mg sapo/ml bath fluid . Similarly, the extract was completely inactive when
given intracerebroventricularly in doses up to 3 mg sapo .
(3) Acid extract. This extract displayed < 1 % of the activity of the methanol extract in
our test preparations . Thus, only trace amounts of active compounds remained in sapo
after methanol extraction.
(4) Eluates from alumina column. Figure 3 presents the elution profile of the various
peptides from the alumina column. Sauvagine is retained by alumina and therefore
appears only in trace amounts in the eluates. (a) Phyllocaerulein was eluted by 40 to 10%
ethanol (261íg/mg sapo), with a recovery of 80%; (b) phyllokinin appeared in 60, to 50,
eluates (141íg/mg), with a recovery of 78%; (c) phyllomedusin emerged in eluates 70, and
703 (181íg/mg) ; assuming an 80% recovery (as inferred from other experiments) the
amount occurring in the crude methanol extract would be 221íg/mg ; (d) the ala-
deltorphins coeluted with phyllomedusin (4 .61íg/mg) with a 86% recovery (Fig. 4) ; (f)
finally, [Lysl-dermorphin-OH was found in eluates 40 and 30 (0 .2-0.25 ug/mg) ; assuming
1104 V . ERSPAMER et al.

Phyllomedusin
[D.AIáldeltorphins
0 Phyllokinin
0 0 Phyllocaerulein
a
Q
ao
E

701 702 703 60, 602 501 502 40 30 20 10

ETHANOL ELUATES
FIG. 3. ELUTION PROFILE FROM AN ALUMINA COLUMN OF THE FOUR MAIN PEPTIDES OCCURRING IN SAPO
(IN 'Ig PER mg SAPO) .
Chromatography was obtained with a stepwise gradient of aqueous ethanol, from 95 to 10% .
Numbers 1, 2, 3 (e .g . 70  70 2, 703) refer to the 50 ml eluate fractions obtained successively at the
same ethanol concentration. [Lysl-dermorphin-OH was eluted by 40 and 30% ethanol (not
shown). Sauvagine was retained by alumina. Peptides were quantitatively estimated by bioassay,
against their synthetic standards, using the smooth muscle preparations shown in Fig . 5 .

a 70-80% recovery from alumina, the actual content of the peptide in sapo would be
0.25-0.33,ug/mg.
Methanol extracts of fresh P. bicolor skin also contain similar amounts of [Trp a,
Asnl-dermorphin-OH . The peptide could not be detected in the sapo extract probably
because it was largely inactivated during drying of the cutaneous secretion. We have
demonstrated (unpublished observations) that [Trp°, Asnl-dermorphin-OH is attacked
with exceptional rapidity by tissue endoproteases.
From data obtained in the bioassay of the crude sapo methanol extract and of the
eluates from the alumina column, it can be calculated that active peptides account
altogether for 7.2% of the sapo weight and for 12% of the weight of the methanol-
extractable constituents of the drug.

Thin layer chromatography

No Dragendorf or Ehrlich-positive spots appeared on chromatograms following
application of 2 mg crude methanol extract, confirming the negative results obtained on
paper chromatograms of P. bicolor skin extracts, using amounts of up to 0.2 g fresh skin
(unpublished observations) .

DISCUSSION AND CONCLUSIONS

Sapo, the dried secretion obtained from P. bicolor skin, contains enormous amounts of
bioactive peptides, potently affecting the cardiovascular system and the visceral functions:
up to 32 hg caerulein per mg sapo, 18 Ag phyllokinin, 22 jug phyllomedusin and 3 jig
sauvagine. In addition it contains 5.2 hg Ala-deltorphin (expressed as Ala-deltorphin I)
and 0.25-0.33 leg [Lysl-dermorphin-OH. Sapo does not contain, like P. bicolor skin
extracts, detectable amounts of biogenic amines (especially indolealkylamines) and
 Frog Skin Peptides 1105

FIG. 4. SMOOTH MUscLE PREPARATIONS usED IN THE QUALITATIVE AND QUAN71TATIVE BIOASSAY OF
THE DIFFERENT PEPTIDES OCCURRING IN THE ETHANOL ELUATES FROM AN ALUMINA COLUMN LOADED
WITH SAPO .
Organ bath, 10 ml . GPGB, guinea-pig gall bladder. Responses elicited by amounts of eluates
corresponding to 2 hg sapo . For comparison 10 ng caerulein (CAER). Peak of caerulein-like
activity was found in 30% eluate . GPI, guinea-pig ileum. Responses elicited by amounts of eluates
corresponding to 3 pg sapo. Peak of phyllomedusin activity was found in 702% eluate . MVD,
mouse vas deferens. Response elicited by amounts of eluates corresponding to 10 ug sapo . For
comparison 2 ng Ala-deltorphin I (DELT) . Clear-cut peak of deltorphin activity was found in
702% eluate . RU, rat uterus . Responses elicited by amounts of eluates corresponding to 10 kg
sapo . For comparison 100 ng phyllokinin (PHK). Peak of phyllokinin activity was found in 602
and 50,% eluates. Time marks, 5 min.

Dragendorfl-positive alkaloids. Other peptide peaks, in addition to those produced by the

above peptides, are present in sapo HPLC chromatograms. Their study is in progress, but
none of them displayed any activity in our systems.
At this point, before passing on to a detailed discussion of our results, a preliminary,
fundamental question must be answered, that of the rate of absorption of the sapo
constituents through the burned areas of human skin.
It is well known that exposed inflamed dermis (like inflamed mucosae) easily absorb
drugs, as a consequence of vasodilatation and increased capillary permeability . Moreover,
bradykinins and tachykinins, occurring in large amounts in sapo, are potent pro-
inflammatory agents, further increasing vascular dilatation and permeability produced by
the thermal injury .
Both bradykinin and eledoisin (a tachykinin related to phyllomedusin) are active,
following intradermal injection in man, at doses as low as 1-5 ng (DE CARO, 1963).
As previously stated, it has been calculated that the sapo material applied to a burn has
an approximate weight of 10 mg. If so, the peptide content of a single application would
be as follows: caerulein 320 ug, phyllomedusin 220,ug, phyllokinin 180 ug, sauvagine
30 ug, Ala-deltorphins 53 ug and [Lysl-dermorphin-OH 3 ug .
110 6 V. ERSPAMER et al.

On the basis of these qualitative and quantitative data we will try to explain both the
peripheral and central symptoms observed in the two successive phases of sapo
intoxication .

Peripheral aspects of the sapo intoxication

The single active peptides will be discussed in turn.
(a) Caerulein and the equiactive phyllocaerulein display a moderate action on blood
pressure (threshold dose by s.c. injection in dog 5-10 hg/kg; in man possibly 1-2 ug/kg),
but a potent action on the gastrointestinal smooth muscle, and gastric and pancreatic
secretions (BERTACCINI et al., 1968a, b; ANAsTAsi et al., 1969 ; BERTACCINI, 1980).
Intramuscular doses as low as 0.06 leg/kg produced in man relaxation of the sphincter of
Oddi, with facilitation of bile flow, and doses of 0.5-1 ug/kg (threshold 0.1 leg/kg)
provoked contraction of the gall bladder, potent stimulation of gastric and pancreatic
secretions and potent stimulation of intestinal motility both in healthy subjects (AGosTI
et al., 1970; BERTAccnm and Atom, 1971) and in patients with post-operative intestinal
atony (HAAS and FuuF, 1978). Side-effects, observed with 1 ug/kg, were nausea, vomiting
(3-10% cases), facial flush, mild tachycardia, changes in blood pressure, sweating,
abdominal discomfort and urge for defecation.
Contraction of the gall bladder and increase in acid gastric secretion were observed also
following nasal insufflation of 1 leg/kg caerulein (threshold 0.1 pg/kg) (AcosTi and
BERTACCIM, 1969). All the above effects and side-effects were obtained, in persons
weighing 60 kg, at a total caerulein dose of 30-60 Mg, an amount 15 to 20 times lower than
the 640-690 Rg of peptide present in 20-30 mg of sapo.
The obvious conclusion is that the gastrointestinal symptoms (vomiting, diarrhoea)
observed in the early phase of sapo intoxication may be predominantly, if not entirely, an
expression of caerulein intoxication . The peptide could also play a role in producing the
cardiovascular effects elicited by the frog secretion.
(b) Phyllokinin is present in 20-30 mg sapo in amounts of 360-540 pg . Since the activity
of this peptide on blood pressure in the dog is three times greater than that of bradykinin,
the cardiovascular effects of the amount of phyllokinin occurring in sapo would be the
same as those elicited by 1000-1500 leg bradykinin . In addition, the effect of phyllokinin
on blood pressure lasts considerably longer than that of bradykinin, indicating a lower
inactivation rate (ANAsTAsi et al., 1966). No data are available on the amount of s.c. or
i.m. phyllokinin or bradykinin necessary to cause a fall in blood pressure in dog or man.
By i.v. injection 0.1 Feg/kg phyllokinin is hypotensive in dogs and rabbits. Thus, it is
possible that 5-7.5 ug/kg of s.c. phyllokinin may affect, at least transiently, blood pressure
in man, contributing to the initial violent cardiovascular effects of sapo.
(c) Phyllomedusin. It is well known that this peptide (ANAsTAsc et al., 1970), like all
other tachykinins, is potently active on blood pressure, gut motility and lachrymal and
salivary secretions (ERSPAMER and FALcoNIERi ERspAmER, 1962 ; BERTACCINI et al., 1965;
DE CARo et al., 1966). The threshold hypotensive dose of the peptide in the dog, by s.c.
injection, is 1-2 Eeg/kg ; the stimulant action on gastrointestinal motility even lower:
0.2-1 Rg/kg in the dog and probably also in man (BERTACCIM, 1980).
Thus, it is conceivable that the 440-660 ug phyllomedusin rubbed into the skin with
20-30 mg sapo ( = 15-20 ug/kg) may substantially contribute to the cardiovascular effects
of the drug and, even more, to its gastrointestinal effects.
 Frog Skin Peptides 110 7

(d) Sauvagine. The amount of this peptide present in 20-30 mg sapo is of the order of
60-90 ug. Peripherally, sauvagine causes in dogs, rabbits and rats a long-lasting fall in
blood pressure, accompanied by intense tachycardia (ERSPAMER et al., 1980). In the dog,
pressure decrease is due mainly to an intense vasodilatation in the mesenteric vascular
area. In rats, the threshold hypotensive and heart rate-stimulating dose of sauvagine, by
s.c. injection, is 0.5 pg/kg (MELcmomu and NEGRI, 1983).
From the above data it seems likely that the amount of sauvagine rubbed into the skin
with 20-30 mg sapo (corresponding to 1-1 .5 hg/kg) may contribute to the cardiovascular
effects ("blood racing, heart pounding") seen in sapo intoxication, as well as, through its
potent vasodilator action on gut mucosal vessels, to diarrhoea. The effects of the
activation of the pituitary-adrenal axis produced by sauvagine will be discussed later.
(e) Opioid peptides . No peripheral effects whatsoever can be expected from the amounts
of Ala-deltorphins (105-160 hg) and [Lysl-dermorphin-OH (6-9 pg) occurring in
20-30 mg sapo.
In summary, it may be reasonably concluded that the intense peripheral cardiovascular
and gastrointestinal symptoms observed in the early phase of sapo intoxication may be
entirely ascribed to the known bioactive peptides occurring in large amounts in the frog
material . A predominant role seems to be played by caerulein, with potentiation of the
caerulein intoxication by phyllomedusin, phyllokinin and sauvagine. Abortion ascribed to
sapo may be due either to direct effect of the peptide cocktail on the uterine smooth
muscle or, more likely, to the intense pelvic vasodilatation and the general violent physical
reaction to the drug.

Central aspects of the sapo intoxication

(a) Caerulein, given s.c. in rats, at a dose of 2 ,ug/kg, displayed clear-cut satiety and
analgesic effects (JuRNA and ZETLER, 1981). In man, an analgesic effect was observed,
following s.c. administration of caerulein, at doses of 0.06-0.2 pg/kg (4-12,ug/60 kg) in
patients suffering from renal colic, rest pain due to peripheral vascular insufficiency and
even cancer pain (ERSPAMER, 1981, 1983; GRAIFF, 1990). It has been suggested that the
analgesic effect, independent of any action on smooth muscles, may be related to the
release of ß-endorphin or to a direct effect of the peptide on the aminergic system in the
CNS. On the other hand, caerulein given by i.v. infusion at a rate of 2 ng/kg/min (total
7 pg) caused a significant reduction in hunger and food intake in human volunteers
(STACHER et al., 1982).
The amount of caerulein rubbed into the skin with 20-30 mg sapo (11-16,ug/kg) would
be largely sufficient to produce satiety and analgesic effects in man.
(b) Sauvagine, given to rats by the s.c. route, from threshold doses of 0.3,ug/kg, caused
the release of corticotropin from the pituitary, with consequent activation of the pituitary-
adrenal axis and an increase in plasma corticosterone levels . With 0.5 and 2,ug/kg, levels
rose by 100% and 400%, respectively . The effects lasted for 2 to 4 hr. Simultaneously,
there was a release of fl-endorphin and an elevation of plasma catecholamine and glucose
concentrations (VALE et al., 1981 ; ERSPAMER and MELCHIORRI, 1983).
It should be remembered that according to GORMAN (personal communication), the
maximum amount of sapo rubbed into the skin over the course of a day could be as high
as 100 mg (corresponding to 3000,ug caerulein and 300 lAg sauvagine) and that Matses
hunters setting traps could apply 20-30 mg sapo to their skin twice daily for 10-15
1108 V . ERSPAMER et al.

consecutive days. It is clear that this would cause a persistent activation of the pituitary-
adrenal axis by sauvagine and full manifestation of the analgesic and satiety effects due to
caerulein.
(c) No appreciable central effects are to be expected by the amounts of phyllomedusin
and phyllokinin present in 20-30 mg sapo, unless they increase the blood-brain barrier
permeability, thus facilitating access to the brain not only of themselves but also of the
other active peptides . This would perhaps allow the appearance of the anti-dipsogenic
effect of phyllomedusin, with consequent enhanced resistance to thirst (DE CARo et al.,
1988).
(d) We do not know whether the Ala-deltorphins present in 20-30 mg sapo
(120-180 ug) could have any central action in man, when administered by the s.c. route. It
seems very improbable . In rats, intracerebroventricular doses as high as 0.1-0.5 ,ug/kg are
required to elicit an increase in locomotor activity and other behavioural changes
(ERSPAMER, 1992). Volunteers, infused with a total of 5 mg of Ala-deltorphin I over a
30 min period, showed no symptoms whatsoever (DEGLI UBERTI et al., 1989). The same
can be said for [Lys]-dermorphin-OH present in 20-30 mg sapo in an amount of 5-8 Fig.
In rats an analgesic effect could be appreciated only after 1-2 mg/kg of the peptide, given
by s.c. injection.
In conclusion, some central effects of sapo (increase in physical strength, enhanced
resistance to hunger and thirst and, more generally, increase in the capacity to face stress
situations) may be explained by the presence of caerulein and sauvagine in the drug; the
role played by the other peptides is uncertain. Of course, it cannot be ruled out that the
cocktail of peptides occurring in sapo, owing to its effect on the vasculature, may increase
the permeability of the blood-brain barrier, facilitating access of all peptides to the brain.
At any rate, no hallucinations, visions or `magic' effects are produced by the known
peptide components of sapo. Similarly, biogenic amines are beyond consideration,
because P. bicolor skin, like that of other Phyllomedusinae, does not contain detectable
amounts of these compounds.
In theory, it is not possible to exclude that other unknown sapo constituents, not
detectable in our bioassay procedures, display some central effects in man. This, however,
seems highly improbable because intracerebroventricular injections ofsapo (methanol and
petroleum ether extracts) in rats did not produce neurological or behavioural effects other
than those expected from their content in known peptides, especially opioid peptides .
Recently, DALY et al. (1992) have isolated from a sample of P. bicolor skin secretion
(sapo) a 33-residue linear peptide, very rich in alanine and lysine (11 and 6 residues,
respectively) and possibly including in the sequence a D-amino acid residue:
Gly-Leu-Trp-Ser-Lys-Ile-Lys-Glu-Val-Gly-Lys-Glu-Ala-Ala-Lys-Ala-Ala-
Ala-Lys-Ala-Ala-Gly-Lys-Ala-Ala-Leu-Gly-Ala-Val-Ser-Glu-Ala-Val-OH
The peptide, named adenoregulin, enhances binding of agonists to A,-adenosine
receptors in rat brain membranes. It is accompanied by peptides, as yet not sequenced,
that inhibit binding. It is difficult to conceive that adenoregulin has any important
participation in the peripheral and central effects of sapo. The enormous s.c. dose of
6 kg/mouse (approx. 240 hg/kg) caused only modest behavioural depression. The s.c.
injection into mice of methanol or aqueous extracts of P. bicolor dried skin secretion in
doses equivalent to 0.1-0 mg/mice of the dried material ( = 4-240 mg/kg) resulted in a
dose-related reduction of spontaneous locomotor activity. At higher doses, mice were
completely inactive for several hours, although when touched they would respond briefly
and then reassume a lethargic state. According to DALY et al. (1992) the inactive state is
 Frog Skin Peptides 1109

reminiscent of the inactive but responsive state caused in mice by injection of some non-
natural adenosine analogs.
Unfortunately no data are presented on the blood pressure situation of the injected
mice . Reduction of locomotor activity and lethargic-like state, with preservation of
responsiveness to stimuli are typical of severe hypotension, that may be produced by the
enormous amounts of vasoactive peptides (up to 70 ug/mg) occurring in the dried skin
secretion. The Matses share the `magic' use of frog poisons with other Peruvian Indians,
the Amahuaca (CARNEIRO, 1970) and the Cashinaua (KENNEIT KENSINGER, quoted by
FURST, 1974) as well as with the Brazilian Indians Mayoruna (of. DALY et al., 1992) and
Marubo (MONTAGNER-MELATTI, 1985). The Cashinaua seem to employ a different frog
species, and to interpret the experience as one of purification, designed to expel a sickness-
like condition and bad luck, especially in hunting.
RoTH (1915) has reported similar `magic' practices among the Indians of the Guineas.
The poisonous exsudation and spaw of certain frogs or toads are rubbed into cuts made in
the skin, or introduced into the eyes, nose, mouth and ears of the hunters, with the same
drastic symptoms experienced during sapo intoxication .
We have studied as many as 12 Phyllomedusinae frogs collected from Northern
Argentina to Mexico, and have shown that all the examined species contained the same
cocktail of active peptides (bradykinins, tachykinins, caeruleins, sauvagine, opioid
peptides) as P. bicolor skin (ERSPAMER et al., 1985). Hence, the skin secretion of all these
frogs could provoke an intoxication similar to that elicited by sapo . Several Phyllomeditsa
species, e.g. P. tomoptera, P. palliata and P. tarsius, are sympatrid to P. bicolor. In
accordance with this assumption, the injection of a methanol extract from a skin of
P. lemur, that was equivalent to 100 mg of wet skin, produced in mice an inactive state
similar to that caused by extracts of dry P. bicolor secretion (DALY et al ., 1992).

Acknowledgements-This work was supported by grants from the Consiglio Nazionale delle Ricerche, CNR,
Rome, and Fidia Research Laboratories, Abano Terme, Padua (Italy) . We are indebted to Miss P . TAJARIOL and
Miss C. CREMONA of the same laboratories for linguistic supervision .

REFERENCES

AGosn, A. and BERTAcciN4 G . (1969) Nasal absorption of caerulein . Lancet 1, 580-581 .

AGosTI, A ., BIASIOLI, S . and BERTACCINI, G. (1970) Action of caerulein on gastric secretion in man .
Gastroenterology 99, 727-730.
ANAsTAsl, A ., BERTACCINI, G . and ERspAmlt, V . (1966) Pharmacological data on phyllokinin (bradykinyl-
isoleucyl-tyrosine-O-sulfate) and bradykinyl-isoleucyl-tyrosine. Br . J. Pharmac. 27, 479-485.
ANAwAsi, A ., BERTAccim, G ., CH, J . M ., DE CARo, G ., ERspAmER, V . and IMPICCIATORE, M . (1969) Structure
and pharmacological actions of phyllocaerulein, a caerulein-like nonapeptide: its occurrence in extracts of the
skin of Phyllomedusa sauvagei and related Phyllomedusa species . Br. J. Pharmac. 37, 198-206.
ANASTASI, A . and FALmNmu ERSPAMER, G. (1970) Occurrence of phyllomedusin, a physalaemin-like
decapeptide, in the skin of Phyllomedusa bicolor. Experientia 26, 866.
BERTACCIm, G . (1980) Peptides of the amphibian skin active on the gut. 1. Tachykinins (substance P-like
peptides) and caeruleins . Isolation, structure, and basic function. In : Gastrointestinal Hormones, pp . 315-341
(JERZY GLAss, G . B ., Ed .) . New York: Raven Press .
BERTAccml, G . and AGosn, A . (1971) Action of caerulein on intestinal motility in man . Gastroenterology 60,
553.
BERTACCINI, G ., CEI, J . M. and ERspAtmat, V . (1965) The action of physalaemin on the systemic arterial blood
pressure in some experimental animals. Br . J. Pharmac. 27, 479-485.
BERTAccim, G ., DE CARo, G ., ENDEAN, R ., ERSPALmR, V . and IMPICCIATORE, M . (196ßa) The action of caerulein
on the systemic arterial blood pressure of some experimental animals. Br . J. Pharmac. 33, 59-71.
BmTACCINI, G., DE CARo, G ., ENDEAN, R ., EItsPANER, V . and ImncciAmRE, M . (196ßb) The action of caerulein
on the smooth muscle of the gastrointestinal tract and the gall bladder. Br . J. Pharmac. 34, 291-310.
1110 V. ERSPAMER et al .

BRoccARno, M., ERSPASn?R, V., FALCONmRI ERsPAmER, G., IMPROTA, G., LINARI, G., MELCHIORm, P. and
MoNTEcuccm, P. C. (1981) Pharmacological data on dermorphins, a new class of potent opioid peptides from
amphibian skin . Br . J. Pharmac. 73, 625-631.
CARNEmo, L. R. (1970) Hunting and hunting magic among the Amahuaca of the Peruvian montaíia . Ethnology
9,331-341.
DALY, J. W., CACERES, J., MoNI, R. W., GusovsgY, F., Moos, M., JR, SEAMON, K. B., MILTON, K. and MYERS,
CH. H. (1992) Frog secretion and hunting magic in the upper Amazon : identification of a peptide that interacts
with an adenosine receptor . Proc. nain . Acad. Sci. U.S.A . 89, 10,960-10,963.
DE CARO, G. (1963) Azione dells eledoisina sulla permeabilitii capillare nell'uomo, nella cavia e nel ratto. Arch .
Int. Pharmacodyn. Ther . 146, 27-39.
DE CARO, G., FARRuGIA, L., MINARDI, E. and NOVARIHI, A. (1966) Hypotensive effect of eledoisin, physalaemin
and related peptides in man. Naunyn-Schmiedeberg's Arch. Pharmac. 254, 194-198.
DE CARo, G., PERFUMI, M. and MAssi, M. (1988) Taehykinins and body fluid regulation . In: Progress in
Psychobiology and Physiological Psychology, pp . 31-66 (EPSIEIN, A. N. and MORRwN, A. R., Eds) . New
York: Academic Press.
DEGLI UBERTI, E. C., SALvADoRi, S., TRAsPoium, G., MARGurn, A., AamROSIo, M. R., Rossi, R., MARAsToNi,
M. and PANSIM, R. (1989) Effetto della deltorfina, oppioide ad alta afflniti per i recettori delta, sulla
secrezione di prolattina, ormone della crescita e TSH nell'uomo. J. Endocrinol. Invest . 12, (Suppl . 1), 22 .
ERsPAmER, V. (1981) Caerulein sus Amphibienhaut . Ein Analgon des Siugetiercholezystokinins. Therapiewoche
31, 2690-2705 .
ERSPAMER, V. (1983) Caerulein and dermorphin as potential analgesic and sedative natural peptides . In :
Pharmacological Basis of Anaesthesiology . Clinical Pharmacology of New Anaesthetics, pp. 127-140 (TIENGO,
M. and CousiNs, M. J., Eds) . New York: Raven Press.
ERsPAmER, V. (1992) The opioid peptides of the amphibian skin. Int. J. devl . Neurosci . 10, 3-30.
ERsPAmER, V. and FALcoNmRl ERsPAmER, G. (1962) Pharmacological actions of eledoisin on extravascular
smooth muscle . Br . J. Pharmac. 19, 337-354 .
ERSPAINER, V. and MELcmoRRI, P. (1983) Action of amphibian skin peptides on the central nervous system and
the anterior pituitary. Neuroendocrine Perspectives 2, 37-106.
ERSPAMER, V., FALCoNwRI ERsPAmER, G., INSELVIm, M. and NEGRI, L. (1972) Occurrence of bombesin and
alytesin in extracts of the skin of three European discoglossid frog and pharmacological actions of bombesin
on extravascular smooth muscle . Br. J. Pharmac. 45, 333-348.
ERsPAmER, V., FALCONIERI ERSPAMER, G., IMPROTA, G., NEGRI, L. and DE CASIIGLIONE, R. (1980) Sauvagine, a
new polypeptide from Phyllomedusa sauvagei skin . Occurrence in various Phyllomedusa species and
pharmacological actions on rat blood pressure and diuresis . Naunyn-Schmiedeberg's Arch . Pharmac. 312,
265-270.
ERSPAMER, V., MELCHIORRI, P., FALCONmRI ERSPAMER, G., MoNTEcuccrn, P. C. and DE CASTIGLIONE, R. (1985)
Phyllomedusa skin: a huge factory and store-house of a variety of active peptides. Peptides 6 (Suppl . 3), 7-12 .
ERsPAmER, V., FALCONB"RI ERSPAMER, G. and CSI, J. M. (1986) Active peptides in the skin of two hundred and
thirty American amphibian species . Comp. Biochem. Physiol. 85c, 125-137.
FURST, P. (1974) Hallucinogens in precolombian art. In: The Toad as Earth Mother: a Problem in Symbolism of
Ethnopharmacology, pp . 88-100 . Special Publications Museum, Texas Technical University .
GoRmAN, P. (1990) People of the jaguar : shamanic hunting practices of Matses . Shaman's Drum, Fall 1990, pp .
4049 .
GRAIFF, C. (1990) Caerulein versus diclofenac in the management of cancer pain : double blind study. J. Cancer
Res. 116 (Suppl.), 626.
HAAS, W. and FuEFF, F. L. (1978) Caerulein in der Therapie der postoperativen Darmatonie and des Ileus.
Therapiewoche 28, 8939-8944 .
JuRNA, I. and 7ErLER, G. (1981) Antinociceptive effect of centrally administered caerulein and cholecystokinin
octapeptide (CCK-8). Eur. J. Pharmac. 73, 323-331 .
LoNGoNI, R., SPINA, L., MuLAs, A., CARBONI, E., GARAu, L., MELcinoRRI, P. and DI CIüARA, G. (1991)
[D-Alai]-Deltorphin 11: D,-dependent stereotypes and stimulation of dopamine release in the nucleus
accumbens. J. Neurosci. 11, 1565-1576 .
MELCMORRI, P. and NEGRI, L (1983) Actions of sauvagine on the mesenteric vascular bed of the dog. Regul.
Peptides 2, 1-13 .
MONTAGNER-MELATn, D. (1985) O Mundo des espiritos: Estudo etn6grafico dos ritos de curs Marfibo. Tese de
doctorsdo, Universidad de Brasilia.
RosEGHnvi, M., ERsPAmER, V., FALCoNmRI ERSPAMER, G. and CEI, J. M. (1986) Indole- imidazole- and
phenyl-alkylamines in the skin of one hundred and forty American amphibian species other than bufonids.
Comp . Biochem. Physiol. 94c, 455460 .
ROTH, W. E. (1915) An inquiry into the animi in and folklore of the Guinea Indians. 30th Annual Report of the
Bureau of American Ethnology, 1908-1909, pp . 103-386.
 Frog Skin Peptides 111 1

STACI*R, G ., STnxsuacER, H ., Scam, G., Sct-uvmER, G. and WwxtII-IIVn, S. (1982) Ceruletide decreases
food intake in non-obese man. Peptides 3, 607-612.
VALE, W., SrnEss, J ., RivmR, C. and RtvmR, J. (1981) Characterization of a 41-residue hypothalamic peptide that
stimulates secretion of corticotropin and ß-endorphin . Science 213, 1394-1397.