Toxicity of peptides of Amanita virosa mushrooms in mice

FUNDAMENTAL

AND

APPLIED

TOXICOLOGY

5, 1144-

( 1985)

1152

Toxicity of Peptides of Amanita virosa Mushrooms

in Mice

A. LORANGER, B. TUCHWEBER,C. GICQUAUD, S. ST-PIERRE,AND M. G. COTI? Departments of Pharmacology and Nutrition, University of Montreal, Departments of Chemistry and Biology, Universite’du Quebec d Trois-RiviPres, and Department of Pharmacology, University of Sherbrooke, Quebec, Canada Toxicity of Peptides ofAmanita virosa Mushrooms in Mice. LORANGER, A., TUCHWEBER, B., C., ST-PIERRE, S., AND C&6, M. G. ( 1985). Fundam. Appl. Toxicol. 5, 1144-l 152. A study was performed on the hepatic reaction of mice to acute intoxication with virotoxins (alaviroidin, viroisin, deoxoviroisin, viroidin, deoxoviroidin) and phalloidin, cyclic peptides isolated from Amanita virosa mushrooms. Purified fractions were administered intraperitoneally at various dosages to determine the LDso which ranged from 1.0 to 5.3 mg/kg, with viroidin, phalloidin, and viroisin being the most potent. The virotoxins and phalloidin induced hemonhagk necrosis of the liver. The development of hepatic lesions was followed by enhanced serum alanine aminotransferase (ALT) activity as well as by light and electron microscopic changes. In additional groups, bile ducts were cannulated and bile was collected for 2 hr after injection of the peptides (1 mg/kg) to determine their cholestatic potential. The earliest changes in hepatocytes were plasma membrane invagination and cytoplasmic vacuole formation.At later time periods, erythrocyte accumulation was evident in vacuoles and in the cytoplasm. The severity of hepatic damage, as judged by morphologic analysis, was correlated with serum ALT activity. Two of the peptides tested (viroisin and phalloidin) decreased bile flow by more than 50% over control values. Mild ultrastructural alterations in the bile canalicular pole of hepatocytes were observed during the GICQUAUD,

development of cholestasis. Sincevirotoxins,likephalloidin, areboundto actin,it ispossible that their affinity to cellular actin may be responsible for their hepatotoxicity.

The toxicity of peptides isolated from the mushroom Amanita phalloides has attracted considerable attention (Wieland and Faulstich, 1978). Two main groups of peptides, the amatoxins and phallotoxins, have been studied extensively and differ in their mode of action. While amatoxins inhibit RNA polymerase and, hence, protein synthesis in various mammalian cells (Stirpe and Fiume, 1967), the phallotoxins exert their effect on hepatocytes, possibly in interacting with cellular actin (Tuchweber et al., 1973; Agostini et al., 1975; Petzinger and Frimmer, 1982). Faulstich et al. (1980) recently described a group of toxic peptides found singularly in Amanitu virosa mushrooms. Purification and resolution by high-pressure liquid chromatography (HPLC) demonstrated that their structure was similar to that of the phallotoxins. However, while phallotoxins are bicyclic pep tides containing cysteine, the virotoxins are 0272-0590/85 $3.00 Copyri&t Q 1985 by the Society of Toxicology. All rights of mpmduction in any form reserved.

Q 1985 SoeietyofToxkology.

monocyclic peptides containing Dserine (Figs. 1 and 2). The principal component (viroisin) of the five fractions separated from the virotoxins exerts a biologic activity which is somewhat comparable to that of the phallotoxins. A preliminary in vivo and in vitro study (Gicquaud et al., 1982) of the effects of two peptide fractions obtained from A. virosa mushrooms collected in Quebec determined that, like the phallotoxins, they caused hemorrhagic necrosis of the liver and bind F-actin. In the present investigation, we report the biologic activity of the five fractions derived from this initial mushroom extract after purification and resolution by HPLC. METHODS Isolation of virotoxins and phalloidin. A. virosa mushroomsweregathered in theautumnof 1981 at Part Na-

I144

TOXICITY

1145

OF Amanita virosu IN MICE

were under urethane anesthesia ( 1.08 g/kg ip). Body temperature was monitored with a rectal probe and maintained ~~-00--WI-~--CO--NH-~-CCh-~-CnzRZ at 37°C by means of a heat lamp. Bile was collected for NH I20 min at 15-min intervals before and after the virotoxin I “t 2 OH and phalloidin injections were given. Bile secretion was co ws Ii measured gravimetrically. cx-RJ Light and electron microscopy. For light microscopy, ._G -co--dn I tissues were fixed in alcohol-form01 and embedded in RS

Nii-CO

-CM-NH-

CO

Ho-cai I w

FIG. 1. Chemical formula of phallotoxins or phalloidin: R,=OH, Rr=H, R,=CH,, &=CH,, Rs=aOH. MW 800. (From Wieland T. and Faulstich H., 1978, CRC Crit. Rev. Biochem. p. 197.) tional de la Mauricie (Quebec, Canada); details of authentication of the mushroom have been described elsewhere (Brosseau, 1983). The peptides were extracted with organic solvents according to the method of Yocum and Simons (1977). The mixture of virotoxins was then subjected to HPLC following the technique described by Faulstich et al. (1980) as modified by Gendreau et al. (198 1). This method facilitated the resolution of the five peptides ala-viroidin, voiroisin, deoxovirosin, viroidin, and deoxoviroidin; HPLC retention times, ammo acid analysis, and reference compounds (kindly supplied by Faulstich) were used to identify the toxins (Turcotte et aL. 1985). Aladeoxoviroidin, which was identified by Faulstich et al. (I 980) in mushrooms collected at different locations, was absent in those gathered for the present study. Animals and treatments. Female Swiss white mice with an average body weight of 32 g were injected intraperitoneally single dosesof the phalloidin, ala-viroidin, viroisin, deoxoviroisin, viroidin, deoxoviroidin in a 0.9% Nail solution as indicated in the table and tigures. Detailed necropsies were performed at the end of the experiment when body, liver, and gallbladder weights were registered. To obtain the gallbladder weight, including bile care was taken to dissect and remove the organ after placing a ligature in the common bile duct to avoid collapse. To determine the LDx,, each peptide fraction, including phalloidin, was given at eight different dose levels to four to eight mice per group. Mortality was recorded for up to 5 days postinjection, and the number of dead mice was expressed as number of dead/total. The LDso values and confidence limits were ascertained according to the moving average method of Weil(l952, 1983). Enzyme activity. Blood was drawn from the abdominal aorta while the animals were under ether anesthesia. Serum aianine aminotransferase (ALT) activity was amessedwith a Dade kit via a modification of the technique described by Reitman and Frankel(l957). Bile secretion. To measure bile volume, a catheter (PE 10) was placed in the common bile duct while the animals

para5n. Sections (4-6 gm thick) were cut and stained with hematoxylin-eosin. Since light microscopic studies of various tissues showed that the liver was the only organ a&ted by the toxins, electron microscopy was only performed on hepatic tissue. The livers were fixed in the universal fixative ( 1% ghrtaraldehyde, 4% formaldehyde, sodium phosphate buffer, pH 7.2) of McDowell and Trump ( 1976) and embedded in Epon resin or Araldite. Semithin sections were cut and stained with tohridine blue for general orientation. Ultrathin sections were cut and stained with uranyl nitrate and lead citrate. Statistical analysis. Statistical comparisons were made by analysis of variance and Student’s t test. p values less than 0.05 are considered statistically significant.

RESULTS L&o In the 5-day observation period chosen for these studies, death always occurred between 3 and 24 hr following administration of the

FIG.

2. Chemical formula of virotoxins or fractions:

Ala-viroidin Viroisin Deoxoviroisin Viroidin Deoxoviroidin

X

RI

R2

MW

SOr

CH, CH,OH CH,OH CH, CH,

:&H,), CH(CH,), CH(CH3)2 CHUM2

868 912 898 898 880

so2

SO SO2 SO

(From Faulstich et al., 1980, Biochemistry 19, 3338.)

1146

LORANGER TABLE 1 LD50 VALUES AND CONFIDENCELIMITS 95% Confidence limits

Toxic peptides

LD50 (mg/kg)

Lower

Upper

Phalloidin Ala-viroidin Viroisin Deoxoviroisin Viroidin Deoxoviroidin

1.38 3.7 1.68 3.35 1.0 5.1

1.2 1.19 2.7 0.67 4.7

1.58 2.4 4.1 1.3 5.4

Note. The toxic peptides were dissolved in 0.9% NaCl and administered intraperitoneally to female mice. The number of dead animals was recorded after 24 hr. The LD50 values and confidence limits were computed according to the method of Weil(l952, 1983).

toxic peptides. After this time period, all the surviving animals recovered. Table 1 shows an increasing sequence of toxicity with deoxo-

ET AL.

viroidin, ala-viroidin, deoxoviroisin, viroisin, phalloidin, and viroidin. All six toxins were active, but viroidin, phalloidin, and viroisin were the most potent. Necropsy Findings Detailed examination revealed that the liver was the sole target for those toxins. In the more severely affected mice, the liver was enlarged, dark, and friable. Liver

and G&lad&r

Weights

Figure 3 reveals that all six toxins significantly augmented hepatic weight generally in a dose-dependent manner. The maximal increases in liver weight ranged from 40 to 60% over control values. Gallbladder weight, in-

7c

6a 2 p s B > i z I 5 a 4 f

50

40

30

20

10

0 DOSE

OF TOXIN

hglkg)

FIG. 3. Effect of various toxins on liver weight. The values represent percentage increases when compared to controls given saline solution. The results are the means +SE of 4-8 animals per group. *Significantly different (p < 0.05) when compared to the controls.

TOXICITY

OF Amanitn virosu IN MICE

1147

nificant only at the highest dose level and between 50 to 80% over control values. The response probably could be related to the cholestatic action of the toxins. The weight increase was attributed to bile retention since no hemorrhages or modification of empty gallbladder weights were observed.

ALT Activity

CIXE OF TOXN ( mQlkQ )

FIG.4. Comparative effectson serum ALT activity 90 min after administration of various doses of viroisin and phalIoidin. The controls were given an injection of saline solution. The results are the means &SE of 4-8 animals per group. *Significantly different (p < 0.05) when compared to the controls.

eluding bile (results not shown here) increased significantly (p c 0.05) in all experimental groups given phalloidin. For the other toxins, however, the differences were statistically sig-

TIME

Figure 4 summarizes the data on serum ALT activity in animals receiving phalloidin and viroisin. Both toxins exerted a similar effect, the response being dose dependent and maximum enzyme activity corresponding to an approximately fivefold increase.

Bile Secretion Bile secretion as an index of hepatic function (Fig. 5) was measured in mice given phalloidin or viroisin chosen as representative of the six peptides. The results demonstrate a gradual dose-dependent decrease of up to 50% of the initial flow rate at 120 min after ad-

( min 1

FIG.5. Comparative effectsof phalloidin and viroisin on bile flow. The results are the means +SE of 5 animals per toxin. *Significantly different (p < 0.05) when compared to the initial bile flow rate.

1148

LORANGER

ministration of the toxins. Doses equivalent to the LDso reduced bile flow by 100% (results not shown here). Histopathologic and Ultrastructural Observations Light microscopy. Histologic analysis revealed a similar hepatic response induced by the various toxins at doses greater than the LDsO. The lesions were characterized by massive hemorrhagic necrosis resembling that already described in detail in phalloidin intoxication (Fig. 6). In surviving animals, moderate postnecrotic fibrous accumulation and bile duct proliferation were evident. At doses lower than the LDso, the hepatic lesions involved hepatocytic vacuolization, which decreased in proportion to the dose administered. No vacuolization occurred at LD5. Examination of

ET AL.

the kidney, intestine, gallbladder, and pancreas did not reveal any histologic damage. Electron microscopy. The hepatic ultrastructure of control mice was similar to that described by this laboratory in an earlier work (Tuchweber et al., 1979). Administration of the virotoxins and phalloidin at doses greater than the LDSo resulted in the development of lesions characterized by endocytic vacuolization and subsequent accumulation of erythrocytes in sinusoids and in hepatocytic vacuoles, which were limited by a single smooth membrane and contained homogenous fine granular material. Microfilamentous material was noted in close proximity to the vacuolar membrane. At the biliary pole, the canaliculi were dilated and most microvilli were absent. A prominent microfilamentous web surrounded these dilated canaliculi. Cytoplasmic foci of degeneration were seen frequently with interruptions of the plasma membrane. The

FIG. 6. Portion of liver of a mouse given 3 mg/kg of viroisin and sacrificed 90 min later. Note the marked endocytic vacuolization and erythrocyte accumulation in both hepatocytic vacuoles and sinusoidal spaces. A few necrotic cells are visible (arrows) X 144.

TOXICITY

OF Amanitn virosa IN MICE

endoplasmic reticulum was disorganized and single short arrays of rough membranes often surrounded the mitochondria, which remained unaltered (Fig. 7). In contrast, at LDs , the hepatocytic changes were comprised of moderate lipid droplet accumulation and hypertrophy of the Golgi apparatus. The bile canaliculi appeared to be virtually normal, except for a slight accumulation of pericanalicular filamentous material (Fig. 8). DISCUSSION Among the peptides of A. virosa, the amatoxins and phallotoxins have received the most attention with regard to their biologic effects (Wieland and Faulstich, 1978). The virotoxins

1149

as a group have been only recently isolated, characterized (Faulstich et al., 1980) and their toxic properties examined (Gicquaud et al., 1982). The present study demonstrated that the LDSo of the six peptides tested varied from 1.O to 5.3 mg/kg, with viroidin, viroisin, and phalloidin exerting the strongest potency. The liver was the main target organ following peptide treatment, as evidenced by light and electron microscopy. The pattern of early endocytic vacuolization of the hepatocytes with the consequent accumulation of erythrocytes after administration of the virotoxins was similar to that caused by phalloidin and cytochalasins (Glinsukon and Lekutai, 1979; Jahn, 1975, 1979). The early congestion, that is, the augmented numbers of erythrocytes within the

FIG. 7. Hepatocytes of a mouse given a high dose of viroisin (3 mgjkg). A cytoplasmic vacuole, surrounded by a single membrane, contains amorphous material (Vl). Other cytoplasmic vacuoles (V2) contain several erythrocytes (ERY). Note the accumulation of amorphous and filamentous material in proximity to a membrane vacuole and in the pericanalicular area (arrows). The endoplasmic reticulum is disorganized, and many mitochondria exhibit changes in matrical density X8 160.

1150

LORANGER

ET AL.

FIG. 8. Hepatocytes of a mouse given a low dose of viroisin (1 mg/kg). The cell organelles appear normal, but the Golgi apparatus is conspicuously enlarged (arrows). Moderate accumulation of lipid droplets can be observed (L) x6640.

sinusoids and the space of Disse, and the subsequent circulatory alterations may have contributed to mortality, which occurred l-4 hr after injection of the peptides. Serum ALT activity, which increased moderately in proportion to the doses administered, was correlated to the changes in hepatic morphology. Indeed, the characteristic features of the hepatic lesions were severe vacuolization, phagocytosis of erythrocytes, and mild cell necrosis. The mechanism mediating the virotoxininduced cell injury could be similar to that hypothesized for the phallotoxins in general and phalloidin in particular. The initial step could be an interaction of the toxins with the outer surface of the hepatocyte through a process which is not yet clear. This interaction could be unique to the hepatocyte and would explain the marked target organ specificity of

the toxins. Considerable experimental evidence supports the theory of Frimmer (1975) of a receptor site on the cell surface. Following this interaction, the toxins could react with actin filaments (Gicquaud et al., 1982; Gicquaud and Turcotte, 1983) which are intimately associated with the membrane. It could result in actin polymerization, seen as the microfilamentous material accumulated beneath the membrane and close to the vacuoles. Recent findings indicate that the plasma membrane alterations may in turn modify the permeability of the barrier to Ca2’ ions. Thus, the resulting hepatocytic injury could be related to disturbed Ca2’ homeostasis in altered cells (Russo et al., 1982). In mice given viroisin or phalloidin in doses equivalent to the LDs , a statistically signilicant decrease in bile flow was observed at 15 min postinjection with a 50% diminution at 2 hr

TOXICITY

OF Amanita

after treatment. The cholestatic effect was not associated with hepatocytic damage, as evidenced by serum ALT activity and light microscopic examination. Electron microscopy revealed virtually normal bile canaliculi. Occasionally, the microvilli were absent, and a slight accumulation of microfdamentous material was noted around the bile canahculi. The other cell organelles were normal, but mild lipid accumulation and enlargement of the Golgi structures were evident. In an earlier study, microlilament dysfunction was thought to be a mechanism mediating phalloidin-induced cholestasis (Gabbiani et al., 1975; Tuchweber et al., 1981; Watanabe et al., 1983). It remains to be shown that this mechanism is also responsible for cholestasis in the present study, in which the toxins produced minimal changes in microfilaments, at least as demonstrated by ultrastructural analysis. The accumulation of fat droplets and multivesicular bodies could be attributed to an effect of the peptides on lipoprotein secretion. Earlier investigations by Prentki et al. (1979) on isolated rat hepatocytes have revealed that both phalloidin and viroisin (identified as a toxin extracted from A. virosa) inhibit lipoprotein secretion in vitro. These authors assumed that the effects of the toxins could result from altered microfilament function. In summary, six peptide fractions isolated from A. virosa mushrooms were shown to be markedly hepatotoxic in mice. These peptides exerted similar effects which, at high doses, presented as hepatocytic vacuolization and intracytoplasmic accumulation of erythrocytes. At low doses, they induced cholestasis in the absence of cell damage. These actions of the peptides are probably related to abnormalities of the cell membrane and of associated microfilaments. REFERENCES B., C~VINDAN, V. M., HOFMANN, W., AND T. (1975). Phalloidin induced proliferation of actin filaments within rat hepatccytes. Z. Naturforsch.

AGOGTINI, WIELAND,

3oe, 793-795. BROSSEAU, M.

(1982). Etude de la toxicitd

et du contenu

virosu

IN

en toxines Amanita.

1151

MICE

de certains

champignons

du Quebec du genre

Maitrise en sciences de l’environnement, Universite du Q&bec a Trois-Riv&es. DUBIN, M., MAURICE, M., FELDMAN, G., AND ERLINGER, S. (1978). Phalloidin-induced cholestasis in the rat. Relation to change in microfilaments. Gastroenterology 75, 450-455.

FAULSTICH, H., BUKU, A., BODENMULLER, H., DABROWN, J., AND WIELAND, T. (1980). Virotoxins: Actin binding cycle peptides of Amanita virosu mushrooms. Biochemistry 19,3334-3343. FRIMMER, M. (1975). A membrane specific toxin. In Pathogenesis

and Mechanisms

of Liver

Cell Necrosis.

(D. Keppler, ed.), pp. 163-173. Univ. Park Press, Baltimore. GABBIANI, G., MONTESANO, R., T~JCHWEBER,B., SALAS, M., ANDORCI, L. (1975). Phalloidin-induced hyperplasia of actin filaments in rat hepatocytes. Lab. Invest. 33,562-569.

GENDREAU, M., TURCOT~E, A., GICQUAUD, C., ANDSTPIERRE,S. (198 I). Separation des peptides toxiques de l’dmanita virosa par chromatographie liquide a haute performance. In Annales de I’Association Canadienne Francaise pour IAvancement de la Science, Vol. 48, p. 35, ACFAS, Montreal, Que, Canada. GICQUAUD, C., TURCOTTE, A., GRUDA, J., AND TUCHWEBER,B. (1982). Effects in vivo et in vitro de peptides extraits de Amanita virosa. Rev. Canad. Biol. Exp. 41, 23-34.

GICQUAUD, C., AND TURCO~, A. (1983). Des peptides de YAmanita virosa viroidine et viroisine sont plus efficacesque la phalloidine pour proteger l’sctine in vitro contre les effets de l’acide osmique. Eur. J. Cell Biol. 32, 171-173.

GLINSUKON, T., AND LEKUTAI, S. (1979). Comparative toxicity in the rat of cytochalasins B and E. Toxicon 17, 137-144. JAHN, W. (I 975). Endocytosis in the isolated perfused rat liver, induced by cytochalasin B. Naturwissenschajier 62,445-446.

JAHN, W. (1979). Cytochalasin D is able to mimic the effects of phalloidin on the rat liver. Experientia 35, 1638-1639. MCDOWELL, E. M., AND TRUMP, B. F. (1976). Histologic fixatives suitable for diagnostic light and electron microscopy. Arch. Pathol. Lab. Med. 100,405-410. PETZINGER, K., AND FRIMMER, M. (1982). Energy linked uptake of demethylphalloin by isolated rat liver cells. Naunyn-Schmiedeberg’s

Arch. Pharmacol.

319,87-92.

PRENTKI, M., CHAPONNIER, C., JEANRENAUD, B., AND GABBIANI, G. (1979). Actin microfilaments, cell shape, and secretory processesin isolated rat hepatocytes. Effect of phalloidin and cytochalasin D. J. Cell Biol. 81,592607. REITMAN,

S., AND FRANKEL, S. (1957). A calorimetric method for the determination of serum glutamic oxalacetic and glutamic pyrnvic transaminases. Amer. J. Clin. Pathol.

58, 56-63.

1152

LORANGER

Russo, M. A., KANE , A. B., AND FARBER, J. L. (1982). Ultrastructural pathology of phalloidin-intoxicated heaptocytes in the presence and absence of extracellular calcium. Amer. J. Pathol. 109, 133-144. STIRPE, F., AND FIUME, L. (1967). Studies on the pathogenesis of liver necrosis by cy-amanitin on ribonucleic acid synthesis and on ribonucleic acid polymerase in mouse liver nuclei. Biochem. .I 105,779-782. TUCHWEBER, B., KOVACS, K., KHANDEKAR, J., AND GARG, B. ( 1973). Peliosis-like changes induced by phalloidin in the rat liver. J. Med. 4, 327-345. TUCHWEBER, B., SIECK, R., AND TROST, W. (1979). Prevention by silybin of phalloidin-induced acute hepatotoxicity. Toxicol. Appl. Pharmacol. 51, 265-275. TUCHWEBER,B., YOUSEF,I. M., AND VONK, R. J. (198 1). Changes in liver cell plasma membrane polypeptides of phalloidin-treated rats with special reference to the actomyosin complex. Canad. J. Biochem. 59, 165- 170. TURCOTTE, A., GICQUAUD, C., GENDREAU, M., ANDSTPIERRE, S. (1985). Separation of virotoxins from

ET AL. Amanita virosa and comparative studies of their interactions with actin in vitro. Canad J B&hem. Cell Biol. 62, 12, 1327-1334. WATANABE, S., MAMORU, M., OSHIO, C., SMITH, C. R., AND PHILLIPS, M. J. (1983). Phalloidin alters bile canalicular contractility in primary monolayer cultures of rat liver. Gastroenterology 85, 245-253. WEIL, C. S. (1952). Tables for convenient calculation of median-effective dose (LD,,,) and instructions in their use. Biometrics 8, 249-263. WEIL, C. S. (1983). Economical LDr,, and slope determinations. Drug Chem. Toxicol. 6, 595-603. WIELAND, T., AND FAULSTICH, H. (1978). Amatoxins, phallotoxins, phallalysin and antamanide, the biologically active components of poisonous Amanita mushrooms. CRC Critical Rev. Biochem. 184-260. YOCUM, R. R., AND SIMONS, 0. M. (1977). Amatoxins and phallotoxins in Amanita speciesof the Northeastern United States. Lloydia 40, 178-190.