Pyrrolizidine Alkai,Oids

Home , Hepatotoxin, Riddelliine

This report contains the collective views of an international group of experts and does not necessarily represent the decisions or the stated policy of the United Nations Environment Programme, the Interna- tional Labour Organisation, or the World Health Organization

Environmental Health Criteria 80

PYRROLIZIDINE ALKAI,OIDS

Published under the joint sponsorshipof the United Nations Environment Programme, the International Labour Organisation, and the World Health Organization

World Health Organization Geneva, 1988 The lnternational Programme on Chemical Safety (IPCS) is a joint venture of the United Nations Environment Programme, the International Labour Organisa- tion, and the World Health Organization. The main objective of the IPCS is to carry out and disseminateevaluations of the effects of chemicals on human health and the quality ofthe environment. Supporting activities include the development of epidemiblogical,experimental laboratory, and risk-assessmentmethods that could produce irtternationally comparable results, and the development of manpower in the field of toxicology. Other activities carried out by IPCS include the development of know-how for coping with chemical accidents,coordination of laboratory testing and epidemiological studies, and promotion of researchon the mechanisms of the biological action of chemicals.

rsBN 92 4 t54280 2

@World Health Organization 1988

Publications of the World Health Organization enjoy copyright protection in accordancewith the provisions of Protocol 2 of the Universal Copyright Conven- tion. For rights of reproduction or translation of WHO publications, in part or in toto, application should be made to the Office of Publications, World Health Organization, Geneva,Switzerland, The World Health Organization welcomessuch applications. The designationsemployed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or of certain manufacturers' products does not imply that they are endorsed or recommendedby the World Health Organiza- tion in preferenceto others of a similar nature that are not mentioned. Errors and omissions excepted,the names of proprietary products are distinguished by initial capital letters.

rssN0250-863X PRINTED IN FINLAND DASS_VAMMALA-5500 RIA FOR PYRROLIZIDINE ALKALOIDS

PREFACE L2

INTRODUCTION- PYRROLI NE ALKALOIDS AND ITI'MANHEALTI{ . 13

1. ST'MMARY.A,ND 19

1.1 Sumary L9 t.2 Sources'and c alstructure... 19 1.3 Mechanisms and tures of toxicity 20 t.4 Effects on man 22 1.4.1 Nature extent of health risks . 22 1.5 Methods for tion . 24 1.6 Reconurendat ions 24 1.6.1 General tions . 24 t.6.2 tions for research 25

2. PROPERTIESAND ANAL METHODS 27

2.L ChemicaL struc re and properties 27 2.2 AnaLvtical me 32 2.2.L ExtractidnExtract.t-on 33 2.2.L.L Pl-ant tissue . . . 33 2.2.1.2 Biological- ffluids l-uids and tissues 33 2.2.2 Analysis for pyrrolizidine aLkaloids 33 2.2.2.1 jthin-Layer chromatography (TLC). 33 2.2.2.2 High-performance 1-iquid chromatography (HPLC) 34 2.2.2.3 Gas chromatography (GC) and mass spectrometry (MS) . 34 2 .2.2.4 Nucl-ear magnetic resonance (NMR) spectrometry . . . 35 2.2.2.5 The Ehrlich reaction 35 2.2.2.6 Indicator dyes . 36 2.2.2.7 DirecE weighing 36 2.3 Determinationo metabolites in animal tissues . . 37

J. SOURCESAND PATHWAYSOF EXPOSUPG

3.1 Hepatotoxic pyrfolizidine alkaloids and their sources . . 38 3.2 Pneumotoxic and other toxic pyrrolizidine alkaloids I -4-

Page

3.3 Pathvays of exposure 42 3.3.1 Contamination of staple food crops 43 3.3.2 Herbal infusions 43 3.3.3 Use of PA-containing plants as food 47 3.3.4 Contaminated honey 50 3.3.5 Mitk 51 3.3.6 Meat 54 3.3.7 Use of PAs as chemottrerapeutic agents for cancer 54

4. METABOLISU ))

4.L Absorption, excretion, and tissue distribution . . 55 4.l.L Absorption 55 4.I.2 Excretion and distribution 55 4.2 Metabolic routes 58 4.2.L Hydrolysis .. .58 4.2.2 N-oxidation 59 4.2.3 donversion to pyrrolic metabolites 59 4.3 Effects of treatments affecting metaboLism . . . . 61 4.4 other factors affecting rnetabolism ...... 63 4.5 Other metabolic routes 63 4.6 Metabolism of pyrrolizidine N-oxides ...... 64 4.7 Metabolism in man . . . 64

5. MECHANISI.{SOF TOXICIIY AND OTHER BIOLOGICAT ACTIONS . . 65

5.1 Metabolites responsibLe for toxicity . . . 65 5.l.L MetaboLic basis of toxicity 65 5.I.2 Isolation of pyrrolic rnetabolites . . . . 66 5.1.3 Chemical aspects of pyrrolic metabolites . 66 5.1.3.1 Preparation 66 5.1.3.2 Chemistrv associated with toxic actions 67 5.L.4 Possible further metaboLites 68 5.2 Toxic actions of pyrrolic netabolites 69 5.2.L Aninals 69 5.2.t.L Pyrrolic esters (dehydro- alkal-oids ) 69 5.2.L.2 Pyrrolic alcohoLs (dehydronecines) .....70 5.2.2 CeLl culturea . . 7| 5.2.3 Possible participation of membrane lipid peroxidation...... 72 5.3 ChenicaL and metabolic factors affecting toxicity .....72 5.3.1 Structural featureg of a toxic alkaloid . . 72 -5-

Page

5.3.2 Activat and toxication 73 5.3.3 Factors ffect the toxicity of active rnetaboLi ea. 74 5.3.3.1 React ity of the metabolite . . . 74 5.3.3.2 The r of reactive groups 74 5.4 Metabolites 1a with the biological actions of pyrrolizid alka ids.. 75 5 .4.L Acute h city 75 5.4.2 Chronic totpxicity 75 5.4.3 Pneumoto:iicity 76 r tissues 76 77 itv.. 77 5.) t of pyrrol-izidine 78 5.5.1 Modified diets 78 5.5.2 Pre-treadment t enhance the detoxication of activd metat lites . 79 5.5.3 Other trdatment 80

6. EFFECTSoN ANIMALS J . . 81

6.1 Patterns of disdase ca by different plant genera and of or{gan in lvement in different 81 tbreaks in farm animals . . 81 84 l.es 87 1-ver B7 hepatotoxicity of t PAs and their N-oxides . 87 affecting hepatoEoxicity . 87 ffects 92 sm of toxic acEion 97 effects 98 6.4.2 ungs 103 ffects . . 105 effects . . 110 sms of toxic action . rL4

6.4.3 entral nervous svstemoJ o LLut L2L 6.4.4 organs . . L2I 6.4.5 ..L24 6,4.6 ..L24 6.4.7 ..L26 damage ....129 6.4.8 .rt"i.ia"::::::::il; -5-

Page

6.4.8.2 PlantmateriaLs ....L67 6.4.8.3 Pvrrolizidine alkal-oid metabolites aid analogous synthetic compounds ... 171 6.4.8.4 MoLecular structure and carcinogenicactivity ..173 6.4.9 Antimitoticactivity ...L74 6.4.10 Imunosuppression .. ..L75 6.4.11 Effects on mineral metabolism . . . L75 6.4.L2 Methods for the assessment of chronic hepatotoxicity and pnetmotoxicity . . . . 176 6.5 EffectsonwiLd-life...... L77 6.5.1 Deer ...L77 6.5.2 Fish.. ..L77 6.5.3 Insects ..178

7. EFFECTSON MAN .. .L79

7.I Clinical features of veno-occlusive disease (voD) L79 7.2 Salient pathoLogical features of veno- occlusivedisease . . 181 7.3 Human cage reports of veno-occlusive disease . . 183 7.4 VoD end cirrhosis of the liver ...... 201 7.5 Differences between VOD and Indianchildhoodcirrhosis (Icc) . . . . 2O3 7.6 Chronic lung disease ...... 204 7.7 Trichodesma poisoning . . 2O5 7.8 Eaiffiil?frIp betlreen dose leveL and toxic effects . ..206 7.9 Pyrrolizidine alkaloids as a chemotherapeutic agentforcancer .;...... 2L3 7.10 Prevention of poisoning in man . . 2L4

8. BIOLOGICAL CONTROL

9. EVALUATION OF ITI'MANIIEALTH RISKS AND EFFECTS ON THE ENVIRONMENT . . 2L7

9.1 lluman exposure conditioirs . . . .2L7 9.1.1 Reported sources of hr:man exposure . . . . 2L7 9.1.2 Plant species invoLved . . 2L7 9.1.3 Modes and pathways of exposure . . 2I8 9.1.3.,1 Contamination of grain crops . . . 2L8 9 .L.3 .2 lterbaL medicines . . . 218 9.L.3.3 PA-containing pLants used as food andbeverages ..2L9 9.1.3.4 Other food contaminated by PAs . . 2L9 -1-

Page

g.L.4 Levels of intake 220 9.2 Acute effects of exposure 222 9.2.I Acute livdr disease 222 9 .3 Chronic dffects <]rfexposure 223 9.3.1 Cd.rrhosis of the liver 223 9.3.2 Mutagenic{ty and teratogenicity 223 9.3.3 Cancer of the liver . 223 9.3.4 Effeits o{t other organs 224 9.4 Effects on the edvironment 225 9.4.L Agricultufe 225 226 9.4.3 Iosects 226 9.4.4 Soil and ter .,226

227

APPENDIXI. PYRROLIZID ALKALOIDS AND TIIEIR PLANT

APPENDIXII. TABIE 1. PI,ANTS HEPATOTOXICPYRROLIZIDINE ALKALOIDS 303

TABLE 2. PI,ANTS INING KNOWNALKALOIDS THAT ARE NON-IIEPA C (AI,IINOALCOHOLSAND ESTERS) . 337 -8-

MIO TASK GROUPON PYRROLIZIDINE ALKALOIDS

Members

Professor M.S. AbduLLahodjaeva, Uzbek Republican Centre for Pathological Anatomy, Tashkent State Medical Institute, Tashkent, USSR

Dr C.C.J. Culvenor, Comronwealth Scientific and Industrial Research Organization, Division of Anirnal llealth, Parkville, Victoria, AustraLia (Cttrir*")

Professor P.P. Dykun, Department of Biophysica, Petrov Research Institute of Oncology, Leningrad, USSR

Dr H.N.B. Gojalan, University of Nairobi, Department of Botany, Nairobi, Kenya

Dr R.J. ttuxtable; Department of Pharmacology, University of Arizona, Tucson, Arizona, USA

Dr A.R. Mattocks, MRCToxicology Unit, Medical Research Council Laboratories, Carshalton, Surrey, United Kingdom

Dr V. Murray, NationaL Poisons Information Service, New Cross Hospital, London, United Kiogdom

Dr B. Snith, Division of Food Regulatory Affairs, Food Directorate, HeaLth Protection Branch, Tunneyrs Pasture, Ottawa, Canada

Professor It.D. Tandon, NationaL Academy of Medical Sciences (India), Ansari Nagar, New DeLhi, India (Chairnan)

Academician F.YU. Yunusov, Department of Chemical Sciences, Uzbek SSR Academy of Sciences, Tashkent, USSR

Secretariat

Dr R. Montesano, Unit of Mechanisms of Carcinogenesis, Inter- national Agency for Research on Cancer, Lyons, France

Dr J. Parizek, International Programne on Chemical Safety, World llealth Organization, Geneva, Switzerland

Dt Z. Gtegorievskaya, Centre of International Projects, Moscov, USSR NOTE TO READERSOF TTTE IA DOCI'UENTS

Every effort has nade to present information in the criteria documents as atel.y as possible ltithout unduly delaying their publicat In the interest of all users of the environmental heallth criteria documents, readers are kindly requested to c0rmrunicate any errors that may have occurred to the Managelr of the Interaational Prograrune on Chernica!- Safety, Wolrld ltealth Organization, Geneva, Switzerland,Swltzerland, in1n order ttiatErlaf they may beDe incLuded in corrigenda, which will appear in subsequent voLumes. -10-

ENVIRONMENTALHEALTH CRITERIA FOR PYRROLIZIDINE ALKATOIDS

A WIIO Task Group on EnvironmentaL l{ealth Criteria for Pyrrolizidine Alkaloids met in Tashkent, USSR, on L-5 Decenber 1986. Dr M. Gounar opened the meeting on behalf of the three co-sponsoring organizations of the IPCs (UNEP/ILO/WHO). The Task Group reviewed and revised the draft criteria document and made an evaLuation of the health risks of exposure to pyrroliz idine alkaloids . Access to the original papers on the subject published in the USSR was made possibLe by PROFESS0RM. ABDULLAIIODJAEVA. DR A.R. MATTOCKSlrrote the first drafts of the sections on Properties and Anal-ytical Methods, Metabolism, and Mechanisms of Toxicity and Other Biological Actions. DR C.C.J. CULVENOR assisted PROFESSORH.D. TANDON in the finalization of the document after the Task Group meeting. Dr J. Parizek, who was originalLy the TPCS staff member responsible for the preparation of the document, and was to be Secretary of the Task Group, could not attend the meeEing because of sudden iLlness, and the Task Group was assisted in his place by Dr M. Gounar, former IPCS staff member. Dr A. Prost l47as responsibLe for the final version of the document. The Secretariat acknowledge the help of both Professor It.D. Tandon and Dr C.C.J. Culvenor. The Task Group meeting in Tashkent was organized by the Centre of International Projects, USSRState Committee for Science and Technology. The efforts of all who helped in the preparation and finaLization of the document are gratefulLy acknowledged.

Partial financial support for the publication of this criteria document was kindly provided by the United States Department of llealth and Human Services, through a contract from the National Institute of Environmental lleal-th Sciences, Research Triangle Park, North Carolina, USA - a WHO Collaborating Centre for Environmental lteaLth Effects.

***- A comorehenslve data base on pyrroliztdine alkaloids has been made available CSIRO Dlvision of Aninal llealth' Private Bag No. 1, Parkv l1e, Vic. 3052, Australia. The data base consists of alkal d occurrence tables and keYworded bibltography readable bv SCI-MATE software systerD (Btbllographic Manager, nstltute for Scientific Information), but adaptable to other s tems. It is available fron CSIRO on IBM - PC diskettes; prl on application to L.W. Snith. -12-

PRXFACE

A disease caused by the consunption of plants containing pyrrolizidine al,kaloids (pAs) has been recognized independently as an endemic disease in certain parts Jf the I'lest Indies and in Uzbekistan in the USSR. Outbreaks of the disease have affected significant segments of populations or large nunbers of peopLe in geographically confined areas in Afghanistan, India, and Uzbekistan. The outbreaks have been caused through contamination of the staple food crops with the s_eeds of pLarrts containing PAs, growing among the crops; such plants a_re likely to thrive foLlowing periods tt arougir-t.- It is notable that the same fanily of plants that caused endemic disease and large-scale outbreaks in uzbekistan aLso caused another outbreak of the disease in adjacent Afghanistan, long after the chemical etiology of the disease (through consumption of toxic seeds in the food) had been identified in the USSR. This happened because there lras a lack of generaL alrareness of the causal relationship between the chemical present in the plant and the disease. Sporadic cases continue to occur in different parts of the vorld through the consumption of seeds or pLant parts containing toxic PAs, as home remedies, beverages, or food. _ The IPCS recognized that this was a health problem that nigh! be lethal, and that it was entirely preventable, provided that it nas recognized in time. Ii lraa also recognized that the dissemination of knowledge, about both the disease and the sources of the chemicals invoLved, wouLd be a critical step in its prevention. Accordingly, the IpCS invited professor H.D. Tandon, who nas respoasible for estabLishing such a causal reLationship in the outbreaks in Afghanistan and India, to prepare a draft criteria document and to assist in its further deveLopm.ent and finalization after the Task Group neeting, which llas h.ld io Tashkent, USSR, on L-5 December, L986. In most episodes of toxic human disease caused by pAs, the liver has been the principat target organ, except for an outbreak in the ussR caused by Trichodesna alkatoids, in which the sJmptoms lrere mostly extra:hepiETEl- The Environmental Health Criteria document provides comprehensive coverage of the hepatotoxic PAS, but lack of relevant documentation prevented the Task Group from anaLysing the role of lrichodesma alkaloids in detaiL. -13-

INTRODUCTION. PYRROTIZIDINE ALKAI,OIDS AND HUI.'ANTTEALTH

Pyrrolizidine alkaloips (PAs) are found in plants growing in most environments an{ all parts of the worLd. The main sources are the fanilies iBoraginaceae (a11 genera), Conpositae (tribes Senecignae and Eupatoriae).r aod Leguminosae (genus c r ot a 1ar iiFna-Eh e p o t6ET-n-urnb e r o f a lk a 1o i d- c on t a i n in g ;pd;iils as high ai 6000, or 3i( of the worldrs f lowering plants (culvenor, 1980). They have long been known to be a health hazard for livestpck, at least since 1902 (Schoental, 1963), and loss of livestlck in various parts of the world has been traced to their g{azing on certain plants growing in pastures, especialLy foLllowing periods of drought or in arid climates. They have beenJ found to be toxic for all species of animals tested (Schoental, 1963), though some species, notably the guinea-pig, are resiiltant (chesney & A1lea, 1973a,' White et al., L973). lluman dipease caused by PA toxicity has been knortn to be endemic in.the central Asian republics of the USSR, at least since thf earLy thirties (Isnailov, 1948arb; Mnushkin, 1949) when sevlral outbrqaks occurred, and the cause was discovered to be tllre seeds of plants of Heliotropiurn species (Dubrovinskii, L947, Ig52; Khanin, 198)-hGE contaminated the staple food crops. A spate of reports folLowed, mostly from the West Indies, of acute and chronic liver disease (Bras et hl., L954, L961; Bras & Hill, 1956i Stirling et a1., L962), associated \,rith the ingestion by people of herbal infuslons for the treatment of certain ailments. Schoental" (1961) and Davidson (1963) suggested that, in view of the evfdence of the hepatotoxicity of PAs, consumption of plants corltaining them could be of etiologica! significanee in human 1i*er disease, especially in developing countries where they are consumed as food or herbal medicines. In spite. of this, and thQ fact that such an ubiquitous source of toxic material is calable of producing aninal .and human disease and that there h{ve been more.recent reports, the PAs have not attracted much I attention in the world as a health hazard. In fact, a recent handbook on naturaLly occurring toxic agents in food (nelnicigl, Jr, 1983) refers to them only in passing and makes no mention of human diseaee caused by them. Veno-occlusive dislease (VOO) (sras & ItiL1, 1956), which is iharacterized by the dominant occlusive Lesion of the centriLobular veins of the f.iver Lobu1e and is caused by these alkaLoids, has since bepn reported from all parts of the world, in both man and {ninals (Hi11, 1960; Bras, f973). It has been attributed to the accidental contamination of food by toxic plant products or the ingestion of herbaL infusions. There have been reports df stray cases and of smalL outbreaks -L4-

from both developing and devel.oped countries. However, in the most recent studies from Afghanistan (Tandon & Tandon, 1975; Mohabbat et al., 1976; Tandon, B.N. et al., L978; Tandon, Il.D. et al..,1978) and India (Tandon, B.N. et al.,1976; Tandon, R.K. et al., 1976; Krishnamachari et al., 1977; Tandon, Il.D. et a1.., L977; Tandon, B.N. et a1., 1978), the disease has been reported to affect large nasses of the population, resu!.ting in high mortality, and has been attributed to the accidental contamination of their staple food crops by PA-containing seeds of plants, foLLoning periods of drought. There is conclusive evidence from studies on experimental animals that the effects of a singLe exposure t; pAs may progress relentlessly to advanced chronic liver disease and cirrhosis (Schoental & Magee, t957, L959; Nolan et al., 1966), following a long interval of apparent well-being, and without any other latent or provocative factor (SchoentaL & Magee, 1959). The lowest levels of such alkaloids administered thus far to experimental animals, e.g.r I - 4 mg/kg diet, have produced chronic liver disease and tumours (Eooper & Scan!.an, L977; Culvenor & Jago, 1979). Pyrrolizidine alkaloids have also been shown to act synergisticaLLy with aflatoxin, another environmental toxin present in agricultural products, in causing cirrhosis and hepatoma in primates (f,in et 41., L974). Though there is no conclusive evidence yet of a carcinogenic roLe of PAs in man, such a possibility has been suspected on the basis of experiBental data (11i11,1960; Williams et al.,1967; IARC, L976,1983; Huxrab1e,1980; Culvenor, L983), and experimental studies have demonstrated carcinogenicity in rats given dosages equivalent to those reported to have been ingested in hunan cases (Cook et aL., 1950; Culvenor, 1983). ALkaloids/toxic metaboLites have been shown to be secreted in the miLk of Lactating dairy cattle (Dickinson et al. , 1976) and rats, and the young of both sexes have been shown to suffer toxic darnage, even when suckled by rnothers treated with retrosine, who apparently are not affected themsel"ves (Schoental, 1959). Such suckling animals lray also be in apparent good heal-th while the livers show toxic effects. Protein-deficient and young suckling animaLs are particularly vulnerable (Schoental, 1959). Chromosomal aberrations have been demonstrated in rats and humans r'rith veno-occlusive disease (Martin et al,, I97Z). Alkaloids have been found in the honey secreted by bees feeding on the toxic plants (Deinzer et a1., Lg77). According to Culvenor and his co-workers, populations in some countr.ies are exposed to low levele of alkaloids in comronly used foodstuffs, e.g., honey in AustraLia (Culvenor et a1.1 1981; Culvenor, 1983, 1985) and comfrey io many countries (Culvenor et al., 1980ar'Culvenor, 1985). Iluman cases of acute disease following the brief ingestion of the alkaloids have known to progress to cirrhosis (Stuart & Bras, L957; Braginskii & Bobokhadzaevr' 1965; Stillman et al., L977; Tindon, B.N. et al., L977; Tandoa, !1.D. et aL. , 1977) in as shorf a period as 3 months from the acute phase (Stuart & Bras, 1,957). The initial disease may be cryptic (Braginskii & adzaev, 1965) and may not be ascribed to herbal- c tion, and yet may progress to cirrhosis (Huxtable, L9 ). Veno-occlusive disease was stated to be the most cormron c of cirrhosis in infants in Jamaica been believed to be a significant etiologicaL factor adult cirrhosis, especialLy in developing countries a et al., 1963). ted to contain toxic aLkaloids are purposes as home remedies all over the world, vrithout systtrnatic testing for safety (Schoental, 1963; snith & Culvenor, 1981) and some are even used as food (Schoental & Coady, 1968i Culvenor, 1980). There are several reports of the cont use of such herbs for medicinal purposes in technically advanced countries (Culvenor, 1980). senecio jacobaea contin to be sold at herbalists shops in the United Kingdom ( tal, 1963; Burns, 1972), and Symphytum spp. (cornfreyf are sti11 used as a vegetables, beverages, or remedies ({attocks, 1980). Both these herbs are known to be carcinogeni{ (tenc, 1976,. Hirono et al., 1978). Young flower stalks qf Petasites japonicus Maxim, the pre-bloom flower of coltCfoot, Tuss.ilago farfara, the leaf and root of comfrey, of fETiiil6l-7iidT-e young leaves and stalks of Far Senecio cannabif,olius, which are a1-1 u foffi, are known to be carcino$enic for rats (Hirono et a1., 1983). Symphytum x uplandicum Nyman (Russian comfrey), which contains several toxic PAs (Culvlenor et al., 1980b) echinidine and 7 acetylycopsamine being the main constituents, is used as a saLad p1ant, green dr , and medicinaL herb. It has been estinated that the rate of ingestion of alkaLoids from this herb may, over a period.qf tine, exceed the levels reported to have been taken during ttie Afghan outbreak. There is a report of at least one patienit who developed toxic effects as a result of consuming a cbmfrey preparation (Culvenor et a1-., 1980a; Ridker et a1., I.985). Arseculeiatne et a1. (1981) found that 3 of the 50 Sedicinal herbs cormnonly used in Sri Lanka contained PAs that had been proved to be hepatotoxic for animals. They suggested that consumption of such herbs rnight contribute to the high incidence of chronic liver disease, incLuding primary liveb cancer, in Asian and African countries, especially a4 they may act synergistically with aflatoxin and hepatitis B virus. The risk of toxic effects due to these alkal-oids y be particularly high in children -16-

(Schoental, 1959r' Jago, 1970) and protein malnutrition, which exists in some countries, may potentiate them (Schoental & Magee, L957). Recent studies from Hong Kong (Kunana et a1., 1985; Culvenor et a1., 1986), the United Kingdon (UcGee et. aL,1976; Ridker et a1., 1985), and the USA (Stillnan et a1., 1977; Fox et aL., 1978; Ridker et al., 1985) report instances of human disease that have been caused by the use of such herbs, resulting in fatality or the deveLopment of cirrhosis, even in c6untries with lrelL-developed health services and among the higher economic and educated strata of society. Indeed, Stillnan et a1. (L977), from the USA, called PA toxicosis the rriceberg diseasett, inplying that cases of this disease night be more frequent than reported in the USA, especially among populations of Mexican-American origin. In general, the use of. herbal remedies is not elicited in the clinicaL history and patients do not volunteer this infornation themselves. Furthermore, the alkaloids are eliminated within 24 h (Huxtable,. 1980) and, even though nethods are available . for their detection in biological tissues and fluids, the suspicion.cannot be confirmed, as the s]rmptoms may take several days or months to apPear. Contamination of food crops is particularly likely to occur in parts of the world nith arid climates, poor or uncertain rainfall, poor irrigation facilities, and following periods of drought, all of which promote the growth of the PA-containing plants that grorr as weeds . among cultivated crops, as has been found in studies on the outbreaks in Afghanistan, India, and the USSR (Terekhov, L939; Dubrovinskii, 1947; Isrnailov, 1948arbr' Tandon & Tandon, 19751 Mohabbat et al., L976; Tandon, B.N. et aL., L976; Tandon, R.K. et a1., L976; Tandon, I1.D. et al., 1978) and in gtazirr.g pastures. The use of traditional medicines is cosrnon in these countries and there is insufficient awareness of this hazard, the disease condition, and its diagnostic pathologicat picture. Furthermore, health se:rvices are poor1y developed. Thus, many of the cases or even outbreaks may go unnoticed or unrecorded and may even be ascribed to malnutrition (Lancet, 1984). Also, many of. the reported cases of so-called ItBudd-Chiari syndromerr, a condition associated with obstruction of najor hepatic veins and/or inferior vena cava, may actually be cases of veno-occlusive disease (Sherlock, 1968), in which only the central. veins of the Liver lobule or sublobular veins are occLuded. Another type of PAs, Elglglg alkaLoids, has been known to cause a human outbreak of disease in the USSR, through contamination of the stapl"e cereaL with the seeds containing thes.e PAs; in this outbreak, the s)rmptoms lrere principaLl,y extra-hepatic (Isnailov et a1., 1970). _L7_

This document is aimep at focusing on a health menace that is insufficiently recogniied, in order to evaluate the health risks on the basis of pullished data, and to draft a set of recorunendations that dould heLp in its recognition, prevention, and controL. -1.9-

AND RECOMMENDATIONS

.1 Sumrary

The ingestion of py Iizidine alkaloids (PAs) in foods and rnedicinal herbs resu ts in acute and chronic effects in man, affecting mainly he 1iver. Data from experimental- anirnal studies indicate t PAs represent a potential cause of cancer in man. The alkaLoids are pr by numerous pLant species and occur throughout the td. In the present document, the alkaloids and their prope ties are described together with the sources of human exposur and the diseaseS that they produce in man and animals. The risks for human health are evatuated and recomrendations are for reducing such risks.

L.2 s and Chemical structure hepatotoxic, are producpd by plant species within the f ollowing f arnilies: Borapinaceae (Ilelig3ggpiu*, Trichodesna, Sympnycum, ano mosE oqher g"rre.a-l-, uo.posrcaeG, Eupat.orium,Eupatorium, and other gfneragenera of the tribes SenecionEeSenecioneae and (genus Eupatoriae)., Leguminosa-e] Crotalarla), and Scrophul- ariaceae (genus castillgja).- Tffi---$enera ar9 rnainly herbaceous and very wiaql-distributed, some species being found in most regions df the wor1d. The majority of the species within these genfra have not yet been investigated, but are expected to contaifn pyrroLizidine atr"kaloids. The hepatotoxic alkaloids have a lr2-double bond in the pyrrolizidine ring and $ranched chain acids, esterifying a 9-hydroxyl and preferably also the 7-hydroxyL substituent. Modified seco-pyrrol-izidipe a1,kaloids, in which the central bond between the tl anld CB atoms is broken, are also hepatotoxic. Some Setrecio species contain non-basic derivatives that are 5-{xopyrroles. The toxicity of these derivatives may be similaf to that of the aLkaLoids, but this aspect has not been investiigated. The alkaLoids occur as free bases and N-oxides. The latter are reduced to the free bases in the gaiErointestinal {ract of animals and have a similar toxicity when ingested ordlly. Suitable analytical pfocedures are avallable for screening plant species, includin5i a sinple field test for toxic alkaloids. Thin-Layer ctlroraatography (TtC), high-perforrnance Liquid (HPLC) r gas chro{ratography (cc), and gas chromatography-mass spectrometry (GC-US) have been applied for separating, characterizirfg, and quantifJfing the aLkaloids -20-

present. Effective use of these procedures requires authentic alkaLoids for standards, few of which are availabLe. Imrproved anaLytical nethods are required for the determination of very 1ow levels of alkaloids in some foodstuffs.

1.3 Mechanisns and Features of Toxicitv

The toxic effects of pyrrolizidine alkaloids are due to activation in the 1iver. Metaboliso of the alkaloids by mixed-function oxidases leads to pyrrolic dehydro-aLkaloids, which are reactive alkylating agents. Reaction of initial metabolites with constituents of the liver ceLl in which they are formed are probably the nain cause of liver celL necrosis. Metabolites are released into the circulation and are believed to pass beyond the Liver to the lung causing vascular lesions characteristic of primary pulmonary hypertension, especialLy when alkaloids, such as monocrotaLine, are administered to animals. In experimentaL animals, PAs are quickly metaboLized and are almost completely excreted in 24 h, so that no residual products are detectabLe in the biologicaL fluids or body tissues after this period. The rate of forrration of pyrroLic metabo!.ites is influenced by the induction or inhibition of the rnixed- function oxidases in the liver, but the relationship between the rate of metabolism and expression of toxicity is uncertain. Several pyrrolizidine alkaloid-derivatives and rel-ated compounds are known to cause chromosome aberrations in plants, leukocyte cell cuLtures of the marsupial (?otorus tTidaclylus), and in hamster cell lines. soni pyrroliTidG? alkaloids induce micronuclei formation in erythrocytes in the bone marrow and fetaL !.iver in mice, sister chromatid exchanges in a Chinese hamster cell- line and human lynphocytes in vitro, and repair DNA synthesis in rodent hepatocyte ce1l cultures. Chromosone aberrations have been reporEed in the blood cel1s of chil"dren suffering from veno-occLusive disease VOD, presumably caused by fulvine. A number of pyrrolizidine alkaloids have been shown to be mutagenic in the SaLnonella typhisruriun assay, after metabolic activation. The carcinogenic activity of pyrrolizidine alkaloids appears to parallel- their mutagenic behaviour, but not their hepatotoxicity. HeLiotrine at doses of 50 mg/kg body !,reight or more, administered .to rats during the second week of gestation, has been shown to induce severaL abnormaLities in the fetus. Doses of 200 rng/kg body weight resuLted in intrauterine deaths or resorption of fetuses. Dehydroheliotridine, the metabolic pyrroLe derivative of heliotrine, was 2.5 tines more effective -2L-

on a moLar basis than i[s parent pA in inducing teratogenic effects. The ability of PAs to cross the placental barrier in the rat and to induce prenatlre delivery or death of litters has been demonstrated. The embryo in utero appears to be rnore resistant to the toxic elfects oiffiTizidine a!.kaloids than the neonate. PAs are known to have passed through the motherrs nilk to the suck]l.ings. Megalocytosis, the presence of enlarged hepatocytes containing latge, hyper-{trronatic nuclei, I is a characteristie feature of pyrrolibidine alkaloiil-induced chronic hepatotoxicity in expprimental animals. The enlarged hepatocytes arise through the powerful ailtinitotic action of the pyrrole metaboLites of pyrrolizidine alkaLoids. This change has not been obsefved in the human Liver, though human fetal liver ceL16 in lritro culture become enlarged when exposed to PAs, indiE'iifi-f?usceptibiliry ro the anlimitotic effect of the alkaloids. In experimental anihaLs, protein-riih and sucrose-onLy diets have given some jmeasure of protection against the effects of the alkaloids, as has pre-treatinent of animals with thioLs, anti-oxidants, or zinc chloride. PAs are noted mainl-y for the poisoning of !.ivestock due to the animals grazing oh PA-containing toxic weeds, and large-scaLe outbreaks halie been recorded. Such episodes have been reported from most parts of the norld, including those with temperate or cold gLimates. Studiep carried out on a wide variety of farn add laboratory aninals have revealed generalLy cormon featurps of toxicity with sone species variations. The Liver fs the principal target organ. In smaLl laboratory aninal.d, doses approactiing a Let6al dpse produce a confluent, strictly zonal haemorrhagic necrosis in the liver lobule, within L2 - 48 h of adninislration of pAs. SinultaneousLy in non-hum{n primates, or after a short tine in the rat, chicken, and swifre, changes begin to occur, and later become organized, in the subintina of the centraL or subLobuLar veins in the liver resulting in their egqlusi6n. The reticulin framework in the central aone of the lobule collapses following necrFsis leading to scarring. Repeated adminietration of suitabld doses leads to chronic liver lesion characterized by negalocyiosis, and increa$ing fibrosis, which may resuLt in cirrhosisf Chronic liver disease including cirrhosis has been show{r to deveLop in the rat following administration of a sinlle dose of a pA. In a number of animaL species, the lungs develop vagcuLar lesions characteristic of prirdary pulmonary hypertension vith secondary hypertrophy of fhe right ventricle of the heart. In rats, appropriately low fepeated doses of several alkaloids have been shown to inducp tumoure, mainly in the liver. In some studies, a single dode has been carcinbgenic. -22-

The central nervous system is the target organ of the toxic PAs contaioed in Trichodesma, which produce spongy degeneration of the brain.

L.4 Effects on Man

In manr PA poisoning is usualLy manifested as acute veno-occlusive disease characterized by a duLL dragging ache in the right upper abdomen, rapidly filling ascites resulting in rnarked distension of the abdomen, and sometimes associated with oliguria, and massive pleural effusion. It can also manifest as subacute disease with vague symptoms and persistent hepatomegaly. Children are particularly vulnerable. Many cases progress to cirrhosis. and, in some cases, a single episode of acute disease has been demonstrated to progress to cirrhosis, in spite of the fact that the patient has been removed from the source of toxic exposure and has been given synptomatic treatmect. Mortality can be high with death due to hepatic failure in the acute phase or due to hematemesis resulting fron ruptured oesophageal varices caused by cirrhosis. Less severely affected cases may show clinical, or even apparently compLete, recovery. The Task GrouP ritas not arilare of any substantiated rePort of prirnary pulmonary hypertension resulting from PA toxicity. Ilowever, in view of the evidence in experimental animals and circumstantial evidence in one case report, the possibility of the development of toxic puLmonary disease in man cannot be ruled out. There is a report of an outbreak of Trichodesma poisoning irr the USSR in which the s)tmPtoms nere mainl"y neurologicaL.

1.4.1 Nature arld extent of hea].th risks

The two main sources of pyrroLizidine alkaloid poisoning reported in hualan beings are the consunption of cereal grain contaminated by weeds containing the alkaloids and the use of aLkaLoid-containing herbs for medicinal and dietary purposes. A third form of exposure, with the potential to affect large populations is the possible Low-level contanination of some foodstuffs, such as honey and niLk, but the Task Group ltas not aware of any casea of human toxicity having been caused through the contamination of these foods. Liver disease caused by the contamination of cereal grains has been reported in rural populations in Afghanistan, India, South Africa, and the USSR. A contributing factor appears to be abnornaLly dry neather, resuLting in the growth of an exceptionally high proportion of the alkaloid-containing weeds in the crops, the seeds of which contaminate the cereal grain on harvesting. The weeds responsible for known outbreaks have -23-

been He!.iotropiurn, Tridhodesna, @!3[, and Crotalaria species. Mortality in subh outbreaks has been reported to be high. In the largest reported outbreak in northwestern Afghanistan, an estimated 8000 people were affected in a total population of 35 000 with 1600 - 2000 deathe. lluman poisoning through the nedicinal use of herbs containing pyrrolizidine {lkaloids has been reported fron all parts of the world. PAs were responsible for a corunon liver disease in children in Jamaica, and individual cases in Ecuador, llong Kong, Indi{, the United Kingdom, and the USA. The plants invoLved were species of Crotalariq, HeliotroqiYn, Senecio. Symphytum. and Gynura. Symphytunrcontainlng preparatlons present a partr.cular nazatd because ot tnelr widespread use and the lenerally high 1-evel-s of individual exposures. The use oI herbs is altnost universal in traditional folk rnedici{e and is increasing in developed countries. Some of the herbs used cotrtain pyrrolizidine alkaloids and have a longfterrn toxicity thdt is unsusPected by the people taking them. Knowledge of the species used in herbal rnedicine and the {requency of such use is very limited in the scientific literfture. About 40 such species are listed in this report, about one-third of which are in use in developed countries. TheY are often presctibed by herbalists,

auq vLusr naturopaths,lt4Lurvy4L.tot and other noh-orthodox practitioners.l,lqlLrLrvt^vte. The extent of the contribution to adute and chronic liver disease cannot be accurately assessed. tt tnay also constitute an etioLogicaL factor in cirrhosis of the Liver and, once this stage r-s reached, it may not be pogsible to identify the cause as a PA. PAs are knor^7nto be transmitted frorn the feed of dairy animals into milk and td cause toxic darnpge in the suckling young. One instance of l]arge-scaLe contamination of honey is knohrn to have been caused by a conunon weed rich in PAs, which was the source of nectar and pollen for the honey-secreting bees. No reports of {ases of acute toxicity caused by consumption of contamin{ted dairy products or honey \ilere availabLe to the Task Grlup. Furthermore, no information is available on the possible presence of PAs or their rnetabolites in the meat of animals fed toxic weed$ before slaughter; however, the possibility pf toxic disease being caused through this medium is considered to be Low. There are no substafrtiaL, long-term follow-up data to assess whether exposure !o PAs results in increased incidence of chronic liver disease br cancer in man. AvaiLable clinical and experimental- data s t that a single episode of PA toxicity and possibl,y also a long-term low level exposure may lead to cirrhosis of thE liver. PAs could aLso be possible carcinogens in man, s e a number of them have been demonstrated to induce fancer in experiurental anirnals, the main target organ being ihe liver. These include sone which -24- have caused episodes of human toxicity, and some others which are found in herbs traditionally used as it.ems of food. Also, in several instances of human toxicity, the reported daiLy rates of intake of such PAs were in close range of those known to induce tumoura in rats. Ilovever, these risks cannot be adeguately assessed on a quantitative basis. There are indications that PA intoxications leading to liver disease are more prevatrent than the reported frequency of cases irould seem to indicate. Because of their known involvement in human poisoning and their -possible carcinogenicity, exposure to pyrrolizidine a_Lkaloids- shotrld be kept as iow as practicallt-achievabLe.- The setting of regulatory tolerance lLvels for certain food products may be required in some situations.

1.5 Methods for Prevention

The only known method of prevention is to avoid consumption of the aLkaloids. In the USSR, a set of agricultural (or agrotechnicaL) legislative, phyto-sanitary and educational measures has prevented new outbreaks of poisoning due to lteliotropium and IIigIg@, since t947.-

1.6 Recomendations

1.6.1 General recodmendations

1. CereaL crops should be assessed throughout the world for possible contamination by weeds likely to contain pyrrolizidine alkaloids. Appropriate grain inspection systems are desirable in order. to achieve near-zero Levels of contamination by such weeds

2 There is a need to create alrarenessr €mong the general. population and those .responsible . for the delivery of health services, with regard to the hazards of consuming such pl.ants ae contaminants in food or as food, or for nedicinaL purposes. Advice on hazards should include mention of possible increased risks, if the alltaloid intake is associated nith drug treatment, (e.g. phenobarbitone) or foods which increase the level of liver metaboliziag enzytnes.

3. Ethnobotanical and taxonomic studies are required in many countries to provide specific information on the use of plant species containing pyrrolizidine aLkaloids for medicinai and dietary purposes. ?here may be a need to controL the sale of some species, and their prescription by herbalists and other practitioners of traditional syatems of medicine. -25-

4. Iloney and dairy proqucts, both local ancl bulk supplies, should be assayed for pylrrolizidine alkaloids in all regions where a risk of contarnihation of these foodstuffs has been identified.

L.6.2 Recormendations research

I. Long-term follow-up studies of the survivors of both alkaloid poisoning in beings and animal outbreaks are required, in order to termine the possible development of chronic liver disease cancer. Sinilar studies are also desirable on individua who regularLy consume comfrey or other PA-containing herbs over a substant,ial period of time.

2. Epiderniological studips should be carrled out in countries with a high incidence of prirnary liver dancer, in order to determine whether there Ls an association with the intake of herbs containing pyrrolizfdine alkaLoids.

3. A network of referedce laboratories is needed to assist nember states in identify]ing plants and their seeds suspected of producing toxic effecfs and for the a$say and identification of PAs. Provisions pay be made for the easy availability of pure alkaLoids for use as reference standards for assays.

4. It is necessary to deveLop improved assay procedures, suitable for the purposds of recormrendaEion (4) in section L.6.1, particularly usiirg fluorescence I and irsnunochemicaL methods.

5. There is a need for further toxicological studies, such as studies on the carcinogen]icity of echimidine and the toxicity of the 5-oxopyrrole constituents of Senecio species, and for studies that would provi[e more quaiffitive information on the various adverse biolo[ical effects of PAs. A study of the carcinogenicity of the a1!

6. Study is required of the possibLe atkaloid content of the meat, organs, and fat of animals that have recently consumed plants containing pyrroli{idine alkaloids.

7. Experimental, studi.ee j are needed on the influence of nutritional status on thl metaboLism, an{ acute and chronic effects of PAs.

8. Further metabolic studies are requi4ed to define more specifically the enzymes involved in the mlcrosomaL activation -26- and detoxification of PAs, to determine whether organelles othef, than microsomes are involvetl, and to explore further, quantitative relationships betrileen different routes of metabolism.

9. The maximm no-observed-adverse-effect dose levels fot repeated long-term administration in the rat and the pig need to be determined.

10. ExperinentaL studies should be conducted to determine:

(a) whether pyrrolizid'ine alkaLoid N-oxides may be metabolized directly into the f-yrroLic dehydro- alkaloid in rnitochondria, especialLy in tumour ceL1s1 and

(b) which P450 enzJrmes are invoLved in the activation and N-oxidation of PAs and thence in the selective Tnduction of N-oxidation enzymes.

L1. A study night be conducted of human variability and its genetic aspects in relation to factors that influence susceptibility to PAs; for example, the study of mixed- function oxidase levels in the liver by metabolism of appropriate test subst.ances recognized as harmless. -27-

t ANDANALYTICAL }IET}IODS

2.L Chemical Structure and P rties

The chemical struc of PAs in reLaEion to their toxic effects has been rev recently by Marttocks (1986). The pyrrolizidine aLkaLoids th which this document is concerned rrhepatotoxic" are those that have iously been calLed or ttnucleotoxict'. llere it is proposed to refer to them as tttoxictt PAs, because of he weight of evidence now avaiLable that they produce damage other organs as well as the liver, and the need to avoid restrict.ive term. There are other types of pyrrolizidine a 1oids, such as those that occur in the plant fanily Orchi , which are oot toxic and are not discussed here. The toxic PAs are ters of the amino-alcoho1s derived from the heterocyclic !.eus. The pyrrolizidine molecule is nade up of two rings inclined to each other as shown in Fig. 1 so that tric isomerisn is possible, and which share a cornmon nr. at position 4.

ecHroH

Fig. 1. Molecular structure o pyrrolizidine molecuLe.

Most hepatotoxic aL oids are estera of noLecules similar to that shown in Fie I (b ) ( 1-hydro:cyroethyl-1 :2-dehydro- pyrroLizidine). Ilowever a few hepatotoxic alkaloids are est.ers of the amino-a 1 otonecine, e.g., petasitenine (fig.2, tto.7). The uns turated pyrroLizidine nucleus itseLf is not toxic, but esters of branched-chain acids are. Ester linkages may be at posit ons 9, 7, or (rarely) 6. Some esters have an ttopentt nolecule, e.g., heliotrine; whereas others are macrocyclic diesters, .g. r monocrotaline and retrosine. Examples of some pyrroli idine alkaloid structures are sholtn in Fig. 2. The ring nucleus ns a doubld bond at the L;2 position, which is ess ial for the toxic effects of the alkaLoid, but not for lated effects. -28-

L. Echinidine Chemical structure: CHr I cH3-?-oH I cH,-o-co-c -cH -cH3 tl HO OH

Chenical formula: C2gH31NO7 Relative molecular mass: 397 CAS registry number: 520-68-3

2. Ileliotrine cheurfEffiture: H.C CH" HC OCH" II cH:-o-co-c -cH -cH3 I OH

Chemical formula: C16H27NO5 Relative molecular mass3 313 CAS registry number: 303-33-3

3. Indicine-N-oxide cheniETllF[cffi H.C'\/ CH. HC or ti cd:-o-co-c-ch _cer "? f-i-\ dr i \-|J 5 Chernical formula: C15H25NO6 Relative moLecular rnass: 315 CAS registry number: 4L708-76-3

4. Jacobine H"C \.-'H ,.OH ts ChernT,GTEucture: /o\ /.cH'-c-c-cHl ^..r' I co col L anI "*'tor- \-N-l Chemical formula: C1gIf25NO6 Relative molecular mass! 351 CAS registry number: 6870-67-3 -29-

5. Lasiocarpine Chemicalurrsrulgal structure:uLtusLutE. ,H, *-"=" c''-i- -- a"" -.o-1 c",-o-co-i-ct-c't, "f, rc 6cx' , \_i,J !

Chemical formula: ]C21H33NO7 Relat.ive molecular mass: 411 CAS registry number: 303-34-4

6. Monocrotaline lLc G chemlic:i.l1ll-6'EiiEffie: cHr_ca_ | _l _"* lll co oH co tl i i*-" I \-iJ i Chemical formula: c 16H23NO6 Relative molecular mass 3 325 CAS registry number: 3L5-22-O H' cx' o 7. Petasitenine .c",-L-- -c-on-."cti' fi.. -\ chenE6l-FEiGdEire: /c GHr i=" i=" I ^ cH,-O >-u--( LiJ i cHr ChemicaL formula: C19H2 7NO7 381 60132-19-6

N-oxide = isatidlne) Chemical structure: .. -,rH .r* F "'c -c--cHPr \^-^/tH?-c t"o *,".t=l' co/ | ..o o cH, tl fn) !I Chemical formuLa: c1gH25No5 \-iJ Relative molecular rnass3 351 CAS registry numberi 480-54-6 -30-

9. Senecionine *F H -,'" ,.* chemTffiEure: \ -CHr-C-C"-Cq j=c- \ Hrc l" co L o ll ",,, i Gf I Chernical formula: ClgH25NO5 Relative molecuLar mass: 335 CAS registry number: 130-01-8

10. Syrnphytine cheriGfsEiiGture: *">ll'

cHr- -clt o-*-l-[-"* )c=c{- | I I H- c4 o" * , ")r1 /H.i \-'lJ Chemical formula: C29I{31NO6 Relative moLecular nass: 38L CAS registry number: 22571-95-5

LL. Trichodesmine chenoGGli,iffie: *'"-"r-""-L-l-"".arc gtl H,c/lll co Ho co ? i"'-" ! ar.\\/NV Chemical formula: c 1gH27No6 Relative molecular mass: 353 CAS registry number: 548-90-3

12. Incanine chetoEl-1 structure: H,c tl rcH-S-S-C-CH, i" i" co;i i*-" Chemical formulat C1gH27NO5 Relative molecular nass: 337 CAS registry nuurberi 480-77-3

Fig. 2. Molecular structure and chemical data for sone pyrrolizidine alkaloids. 31 -

it is of historical interest to recal1 that the structures of heliotrine and lasiocarpine, the rnain alft.aloids of Heliolropitto lasioc , wereitere worked out byDy Drur c.PfG.Pf Merrtshikovffiscowl4en-snlKov anq assocraLes ln nuusuv in the L930s. This work lncluded deterninlng the structure of heLiotridine, the par compound of the amino-alcohoI, heliotridane. Dr Mentshlkovts studies rtere carried out at essentially the same t but independentlY of studies bY English and American authors on retronecine-based al-kaLoids. The alkaLoids in pldnts are often found together with their N-oxides, which ar$ also toxic, \"hen ingested oraLLy. The pfrrolizidine alkaloN.ds acquire their toxic properties (the general onLy through the toxic lyrrolic intermediates structure of which is shown in Fig. 3) formed by the mixed-function oxidases df the hepatocytes. To form these pyrrolic derivatives, the lalkaloid molecule should have:

(a) a double bond at [tre 1:Z position of the ring nucleus;

(b) esterified hydro:{yl grouPs in the nucleus at the C 9 and/or C 7 positiPns; and

Fig. 3. Molecular structure of foxic pyrrolic intemediates'

Substitution at the a position of the acid and esterification of the C-7 hydr[xy group both eohance the toxicity of the alkaloid (Robins , 1982). A group of related I alkaloids, isolated fron Senecio species by Bohlmann et al. (L979), have non-basic pyrrolic structures similar to thdse of toxic pyrrolizidine alkaLoid metabolites, but they abe chernically dbactivated by the presence of a carbonyl gfoup at position 3 of the pyrrolizidine nucleusr e.B., senlaetnine (fig. 4). Senaetnine does not possess the acute h{patotoxic characteristics of basic pyrrolizidine alkaloids. llovrever, it had a direct irritant action on tissues neaf the site of intra-peritoneaL administration and caused damage to pulmonary vascular tissue when given intra-veinous tb rats (Mattocks & Driver, L987). -32-

t'ig. 4. Molecular strucrure of senaetnine, a non-alka1oida1 pyrrolic constituent of sme Senecio soecies.

The aLkaloids are fairLy stable chemicaLLy, but the ester groups may undergo hydrolysis under alkaline conditions. Some alkaloids in plant material may decompose during drying (null et al., L968), but others appear to be stable under similar conditions (Pedersen, L975t Birecka et a1., 1980). The N-oxides of unsaturated pyrrolizidines are more readily d-ecomposedby heat than the basic alkaloids, especialLy \^rhe; dry. Ilowever, the stabiLity of the alkaloids and N-oxides in hot water as, for example, in cooking, is not known.- Some pyrrolizidine aLkaloids have a linited r,rater solubility, unless neutralized with acid; but others (..g., indicine), and all the N-oxides, are readily soluble.

2.2 Analytical Methods

tr'Ihenanalysing for PAs, it is important to recognize that this group consists of many different compounds (section 2.L) and that these often occur as mixtures in plants or in materials of plant origin. They may vary in structure, relative molecular mass, response to anaJ.ytical procedures, and toxicity. Both ,basic alkaloids and corresponding N-oxides may be present at the same time. Thus, where such frixtures are present, anaLyses will inevitably be approximate, unless the individual components are separated and identified. Nevertheless, such estirnates can be useful. In particular, all hepatotoxic PAs are unsaturated in the sense that they possess a L:,2-doubLe bond in the pyrrolizidine nucleus, and analytical methods that are specific for this st.ructure can be of value in screening for potentiaL toxicity. A sirnple qualitative fiel-d test for screening plant materials for the presence of such aLkaloids and their N-oxides, without the need of high technol-ogy equipment, is described in section 2.2.2.5. -33-

2.2.L Extraction

2.2.L.1 Plant tissue

Pyrrolizidine alkaloids are usually extracted from dried, nilLed plant material witlr hot or cold alcohol. The alcohol is evaporated, the bases taken up in dilute acid, and fats extracted with ether or ipetroleun. It is usual, at this stage, to reduce any N-{xides present to the corresponding basic alkaLoids with zifrc, before naking the solution alkaLine and extracting the alkalfoidg with chlorbform (Koekemoer & tJarren, 1951). Alternatively, alcohol can be continuously circulated through the pLdnt material and then cation exchange resin, and the alkaloids subsequent.ly eluted from the resin (Mattocks, 1961; Deagen $ Deinzer, L977). PAs can also be extracted by soaking platrt materiaL in dilute aqueous acid (Briggs et al., 1965,'Crai[ et al., 1984).

2.2.L.2 Bio ical fluid and tissues

Pyrrolizidine a1kalo have been extracted for analytical purposes from honey (Dei r et aL. , 1977), milk (Dickinson et a1-., L976), blood-p!"asma & Powis, 1978,' McConish et al.,

-'vv (Mattocks, L980),t , urine \rseLve^o, 7a; Jago et al.r 19697 Evans et al., L979), and bile (Ja et a1., L969; Lafranconi et al., 1985). When attempting to i ate PAs from animal tissues, it must be appreciated the toxic all

2.2.2 Anal sis for Lizidine aLkaloids

2.2.2.1 Thin-l (TLC

For TLC, silica p1a s are usually used, eLuted with chLoroform: methanol :aq arnmonia mixtuf,es (Sharrna et al., 1965; Chalmers et al-., 1965); solvents suitable for the N-oxides, which are more ter-solubLe, have been described by Mattocks ( 1967b) and tr'Ia r et al. (1981). The most sensitive methods for detecting P on TLC are those using Ehrlich reagent (4-dinethylaminob nzaldehyde) (Mattocks, 1967b). The unsaturated alkaloids ar best visualized by spraying the -34- plates first with a solution of orthochloranil, then with EhrLich reagent, heating after each spray (Molyneux & Roitman, 1980). The N-oxides of unsaturated pyrrolizidines are detected by spr-aying a solution of acetic anhydride, heating the plate, and then spraying EhrLich reagent (Mattocks, 1967b). Pyrrolizidine alkaloids with a saturated base moiety must be detected in other rdays (which are not specific for pyrrolizidines), e.g., by exposing the dried pl.ates to iodine vapour, or by spraying with an iodobisnuth (Dragendorff) reagent (Munier, L953).

2.2.2.2 High-perforsrance liquid chrouratography (IIPLC)

AnaLytical or preparative scale IIPLC separation of pyrrolizidine alkaloids has been described by Segall (1979arb) and Dimenna et al. (1980), and an inproved method has been reported by Ramsdell & Buhler (1981). Alkal.oids from Symphvtum officinale (confrey) have been separated on an analytical scal-e by TitteL et al. (L979), and partially separated on a preparative scale by Huizing et al. (1981). UV detectors are usual-ly used for the IIPLC of pyrrolizidine compounds (Mattocks, L986).

2.2.2.3 Gas chromatography (GC) and mass spectrometry (MS)

The GC characterization of PAs using packed colururs has been described by ChaLmers et aL. (1965) and Wiedenfeld et al. ( 1981). Mixtures of alkaLoids f rom comfrey GI*$yt* ), normall,y hard to separate, were resoLved by Culvenor et "p.al. (1980a) and Frahn et al. (1980) by GC of the methylboronate derivatives. Gas chromatography combined with mass spectrometry (GC-MS) has become a valuabLe and highly sensitive means for both the identification and the quantitative determination of pyrrolizidine alkaLoids. Thus, a1ka1-oids extracted from honey vere separated and identified by Deinzer et al. (L977) and (as butylboronate derivat.ives) by Culvenor et al. (1981). Deinzer et al-. (1978) described a method for the recognition (but not the individuaL identification) of retronecine-based pyrrolizidine aLkaLoids, by hydrolysing them to retronecine (the ami.no alcohol noiety) foLlowed by GC-MS of its bis-trifluoro- acetate. The use of capillary GC has greatly improved the sensitivity of pyrrolizidine alkaloid analysis, especialLy when used rnrith MS (l,uthy et al. , 1981). The MS of pyrrolizidine compounds has been reviewed (SutL et al., 1968; Mattocks, 1986). Pyrrolizidine N-oxides generally undergo thermal decomposition, when subjected to GC, but they can first be reduced to the corresponding basic alkaloids (Koekemoer & I'larren, 1951). Al-ternatively they may be derivatised. Thus, -35- trioethylsilyl.ation of in4icine N-oxide or heliotrine N-oxide can lead eiitrer to the trinethyGilyl (TMS) derivative-of the parent alkaloid or to the TMS derivative of the dehydro- alkaLoid (pyrrolic derirlative), depending on the reagents used, and these products dill run successfully on GC-MS (Evans et al. , L979, 1980).

2.2.2.4 NucLear (NMR) s

A convenient, but relatively in$ensitive, method, specifically for the d ination of unsaturated PAs, has been described by Mol et al. (L979). The basic alkaloids are extracted, then subjepted to NltR spectrometry along with -to quantitative measurement! be made of the signal(s) representing the II2 prot$n(s) in unsaturated pyrroLizidines, and thus the aLkaloid(s) tan be determiried. Quantitative NUR analysis of pyrrolizid alkaloid mixtures from Senecio yglgar_-1" has been descr by Pieters & VLietinck (198-5ffi? compared with an HPLC Qualitative aspects of alkaLoids have been Mattocks (1986).

2.2.2.5 The Ehrlich react

This method (t"tat , 1967a, 1968b) is specific for unsaturat,ed pyrroliziiline I alkaloids and io not suitable for other alkaloids. Thus, lt is the most lrsefuL colorimetric method for potentially rpatotoxic pyrrolizidine compounds. The procedure converts t e alkal-oid into its N-oxide, using hydrogen peroxide. The t reacts nith acetic anhydride to form a pyrrolic derivaiive (dehydro-a1kaloid) that gives a magenta coLour with a s l,Ly rnodified Ehrlich reagent. The latter contains boron trifluoride to give maximum sensitivity. As tittLe as 5 pg of nrosf, unsaturated pyrrolizidines can be measured by this method. If the oxidation stage is omitted, only the unsaturated pyrrolizidine N-oxides can be determined. The determination of pyrfrolizidine N-oxides has also been discussed by Mat,tocks (1971b) A sinplification of tfire above colorimetric procedure was described by Mattocks (1971d) to provide a quaLitative test that could be used to scfeen large numbers of plant sampLes for the presence of r]rnsaturated pyrrolizidine alkaloid N-oxides. An improved vbrsion of this field test is now a:vaiLab1e (t{attocks & Jukps, 1987). It is suitable for any plant parts, such as leavfs, stems, flowers, seeds, or roots, or materials of pl-ant ori$inn such as cereals or herbal teas, but has not yet been applipd to cooked food. -36-

The plant material (O.Z - I g) is extracted by grinding it with aqueous ascorbic acid (5%) and a small amount of sand. The solution is filtered and divided into two equal portions (tttesttt and rrbLankrr). An aqueous solution (0.2 rol) of sodium nitroprusside (5%) containing sodium hydroxide (10-3 nol) is added to the trtesttt samole. Both portions are heated for approximately I nin at 7b - 80'C; then Ehrlich reagent is added and heatlng is continued for 1 min. The Ehrlich reagent contains 4-dinethylaninobenzaLdehyde (5 g) dissolved in a mixture of acetic acid (60 nl)' water (30 n1)' and 607. perchloric acid (10 nl). A magenta colour in the rrtestrr compared with the |tblankrr indicates the presence of an unsaturated PA N-oxide. The flblanktt may show a colour if the plant contains conpounds, such as indoles or pyrroLes, which can themselves give a colour rrith Ehrlich reagent. The intensity of coLour in the rrsamplett compared nith the [blankrr can give a rough idea of the amount of alkaloids present, and indicate whether further chemical or toxicologicaL testing of the plant material is adviseable. In practice, the majority of PA-containing plants contain enough alkaloid in the N-oxide form (often a large proportion) to react positiveLy iri ttris test. The main exceptions are some seeds (Crotalaria), which may contain much alkaloid base, but Little or no N-oxide. These (and any other sample not containing chlorop-hyll) can be tested for basic PAs by grinding then arith chloroform, heating the fiLtered extract with a solution ( 0 . 1" nL ) of orthochloranil (0.5"/") in acetonitrile, and then heating it with Ehrlich reagent. A inagenta colour indicates the presence of an unsaturated PA. Non-toxic pyrrolizidine alkaloids having a saturated pyrrolizidine nucleus, and pyrrolizidine aLkaLoids that are otonecine esters, such as petasitenine, will not respond to this test.

2.2.2.6 Indicator dyes

A method generally appl-icabLe to tertiary bases has been adapted for pyrrolizidine alkaloids by Birecka et al. (1"981-). It is sensitive, but is not specific for this group of al.kaloids, and it does not distinguish betneen the saturated and unsaturated alkaloids. A chloroform solution of the alkaloid is shaken with acidified aqueous nethyl orange. The yellow alkal-oid;dye complex is subsequently released from the chloroform phase, using ethanoLic sulfuric acid, and measured s pectrophotometrical ly.

2.2.2.7 Direct weighing

An insensitive way to determine the alkaloids in, for example, a pLant sample, providing enough is available, is to -37-

extract the alkaloids. (se]ction 2.2.I) and weigh then. This will provide a rough measulre of the total bases present in the sample; however, these rn{y not necessarily be PAs. Never- theless, the sample can {hen be subjected to further tests, e.g., GC-MC, nuclear nagnetic resonance (NMR), or colorinetric analysis. Furthermore, pfrroLizidine N-oxides are generall.y too rilater sotuble to be 4ppreciably ex-tractabLe from aqueous solution by chloroform. fhus, if two portions of the sampLe are extracted, and one of them is reduced to convert N-oxides to bases, the weight diff$rence betrileen the tno prodults will represent the alkaloid existing in the form of N-oxide in the original sampLe.

2.3 Determination o MetaboLites in Aninal Tissues -38-

3. SOURCESAND PATIIIilAYSOF EXPOST'RE

3.1 Hepatotoxic Pyrrotizidine Alkaloide and Their Sourceg

Plants constitute the only natural source of pyrrolizidine alkaloids (PAs) that cause toxic reactions in man and animals. PAs occur in a number of species in the families Boraginaceae, Compositae, Leguminosae (genus Crotalaria), Ranunculaceae (genus Caltha), and Scrophulariaceae (genus castille.ia) (Table 1).---TG most important genera of PA- Effiffi; toxic olanta are crotalaria (Leguminosae), Senecio (compositie), Heliotropium, -fiidffil6ngg., Amsinckia, E?h-ffi; andsymphytum-i(-Boraginaceae)-@1973)-ErZ6?aea cases of human toxicity have mainly been caused by at Least 12 different pyrrolizidine alkaloids, mostly derived from Ileliotropiurn, Senecio, and Crotalaria genera. The Senqcio spp. grow throughout the world; the Crotalaria sPP. are mainly found in the tropics and subtropics (CuLvenor, 1980).

Table 1. List of plant genera containing toxic pyrrolizidine alkaloids (with nunber of species investigated)

! anl ry Gen6a

Apocynacea€ Fernaldia (1), Parsonsia (4),

Boraginaceae Alkanna (1), Amsinckia (4), Anchusa (2), Aspe{ugo ITf-. Boraso eD. caainia di. cvnoelossum (9), E4!gs---Tt, uact

Compositae *r**=,:-1., s'j"h"gr'ttl: {r).,..lgggl4 !+r' Conoclintum (I), Crassocephalum (f), Doronlcum (2). Echinacea (2). Emilia (2) ' Erechtites (r, ' eupato;i[-n (8). Farfugium (1), cynura (2), ffiiT"ria (5). Pffi (4). senElo (L42), Syneilesis (1), Tussilago (1)

Leguminosae Crotalaria (60)

Ranunculaceae caltha (2)

Scrophulariaceae Castilleja ( 1)

An alphabetical list of pyrrolizidine alkaLoids with their plant sources has been published by Smith & Culvenor (1981) -39- and ltattocks (1986). {n updated version is attached as Appendix I. The ptant gelrera containing toxic PAs are listed in Table I indicating th$ nunber of species investigated for PAs. A comprehensive list of species of irlants belonging to each of theae genera, thp alkaloids isolated from each, and the part of the pl.ant confaining the alkalOid are presented in Appendix II. Table 1 in Appendix II incLudes species kno!,rn to contain alkaloids of provpd hepatotoxicity, or of a moLecular structure that would naNe them very probabLy hepatotoxic. Table 2 in Appendix II lncludes species containing pyrrolizidine amino-alcohols or esters, which, wtrile not having all the features of hepatdtoxicity, vould need only minor structural- nodifications lto render them hepatotoxic. Plants of the same taxononic grQups as the pLants of proven hepatotoxicity are l-isted in flart (a) of the table. There is a possibility that, on further examinition, hepatotoxic alkaloids may be found, {s rninor constituents, in strains or parts of these plants no$ yet investigated or under specific conditions of growth. Ip should be noted that the species that have been investigfted and are Listed are only few cornpared with the total niumber of species in each genera. It has been recomended by Slnith & Culvenor (1981) that it would be prudent to regard al.l" species in the fanily Boraginaceae and the genera Crotal4ria, Senecio, and Eupatorium as potentially hepatot6lffi It is pertinent to that the alkaloid content in different parts of the lant (".g., roots, leaves, stalks, fLowers, and buds) var s and is subject to fluctuations according to the cli e, soiL conditlons, and tirne of harvesting (Danninger et 1., 1983; Hartmahn & Zirrner, 1986). Mattocks (1980) demonstra that the alkaLoid content of the leaves of lynphytun spp. Russian confrey) p which are used as an item of food, varies r{ith their maturity. The toxic PA content is highest at beginning of the vegetative period and declines as the s mature. The PA content of the roots is much higher than that of the leaves, and dried l-eaves contain a higher concent ation than fresh leaves (Mattocks, 1986). According to Dann r et al. (L983), in some species (Symphytum asperun), rel tively long storage may lead to a reduction in the aLkaloi content, presumabl"y because enz)rmes are released during dryin . Candrian et aL. (1984b) studied the stability of PAs in hay and silage containing various amounts of Senecio a1 . The PA conternt of hay remained constant for several mont s, but the PAs in silage were mainly degraded. However, the degradation of PAs rdas much less complete in the lower tration rangei, A quantitatively signif icant PA-degradat product in silage was identified as retronecine. Silage with an S. aLpinus pefcentage of 3.5 - 23 stiLl contained macrocvc lc PAs at a concentration of about -40-

2O nglkg wet weight. Such siLage lras not considered safe for cattle bearing in nind that a 600-kg calf eats about 30 kg silage/day, amounting approxirnately to a daily intake of about L ng PAs/kg body neight. In feeding trials r4rith Senecio jacobaea, Johnson (Jg79) found that the minimum letha-i-d6i6 for cattle was between I and 2 ng PAs/kg body weight per day. PAs knolrn to have been associated nith instances of human toxic liver disease in different parts of the world are Listed in Table 2. Two groups of alkaloids that, according to Culvenor (1983), are consumed in significant amounts by people in different parts of the world include:

(a) Echinidine acetvll. am1-ne and related alkaloids countrt.es

Leaves (j:gglyt* .of pLants of rhe Symphytun Bp. of f icinale (confrey) and Synphyturn x uplandicun) E?Else

(b) Echimidine and related al-kaloids (Australia)

PAs derived from Echium plantagineum, with echimidine as the major component, have been found in honey secreted by bees feeding on the plant (Culvenor et a1., 1981). The plant is a major source of honey (section 3.3.4).

3.2 Pneumotoxic and Other Toxic Pyrrolizidine Alkaloids

Not all hepatotoxic alkaloids are pneumotoxic. The cofimonest ones used to produce experimentaL Lung injury are fulvine (Barnes et al-., L964; Kay et al., 1971-a; $lagenvoort et aL. r 1974a,b) aad monocrota!.ine (f.alich & Ehrhart , L962; Chesney & Allen, L973b; Iluxrable er al., 1977). These are also the most active (Mattocks, 1986). The seeds of Crotalaria spectabilis, which contain monocrotaline, have also been used to study pneumotoxic effects on experimental animaLs (Turner & Lalich, 1965; Kay & Ileath, 1966r. Kay et a1. , t967a) and C. spectabilis has been called the pulrnonary hypertension- pl"anE@-Eeath, L969), because of Lhe- pul"monary hypertensionogenic properties of the pAs it contains. Culvenor et a1. (L976a) screened 62 PAs for hepatotoxicity and pneumotoxicity. Chronic luog lesions rrere produced by most compoundd that induced chronic liver lesions, though high -4L-

Table 2. Instances of humn toxicity caused by Pyrrolizidine alkaloids9

PrincipaL Country/ Cause of Reference alkaloid Region int4ke

HeLiotrine and tleliolTopium Afghanistan contanin- Tandon & Tandon other alkal-oids popovrl ation ( 1975) ; Tandon, similar to B.N. et al. lasiocarpine (1978); Tandon, It.D. et al. ( 1978) ; Mohabbat er aL. (1975)

Senecionine senecio South contamin- tlilnot & i11ic ifornis; Africa ation Robert6on senecie ( 1920 ) b;;EJTi

.@ig spp. south contaEin- selzer & Africa ation Parker ( 1951)

Alkaloids of Crotalaria Ecuador redlcine Lyford et al. trichodeemine lUnCee ( 1976 ) and senecionine ryPe

Heliotrine and Hong Kong nedlciae Kumana et al. lasiocarpine (1985): Culvenor et aI. ( 1e86)

Crotananine and Crotalaria India contamin- Tandon' R.K. cronabumine ngna ation er al. (1975); Krishnanachari et al. ( 1977) ; Siddiqui et al. ( 1978a,b)

tleLiotrine lleliotrooiun India nedlcine Datta et al. N-oxide eichwaldii (1978a,b)

Monocrotal ine Crotalaria west Indies redicine Bras et al. (L954, fulvine @; L957) Crotalaria fulva Stuart E Bras ( 1957)

E United nedticine Mccee et a1. "p. Kingdon ( 1976)

Riddelline senecio USA nedi-cine StilLnen et al. retrorsine Iffitouos (1977); rox et N-oxide al. (1978); (with others) l{uxtable (1980) -42-

Table 2 (contd).

Indicine N-oiide purified USA nedicine Letendre et a1. chemical ( 1984)

Symphytine, syngl- Symphytum sp. USA medicine Ridker et al. andine, and other (1985); synphytum alka- Huxtable loids er al (1986)

I-asiocarpine and USSR contamination Dubrovinskii heliotrine (1952); iinushkin ( 19s2)

Trichodesmine and Tiichodesna USSR conEamination qhranhAro I. incanine lncanum nrlarra IrOii). iunosov & Plekhanova ( r959)

a Adapted from: Culvenor (1983) and Matfocks (f986). Refer also ro Table 15 for details and section 7. d_oses,lrere required in some instances. It is possible that chronic lung lesions may not occur in experimentar. animals because of early death due to acute toxicity. I{ovever, the authors identified a number of pAs that were particularl"y prone to produce chronic lung damage in rats incLuding crispatine, senecionine, seneciphylline, and usaramine (l2-nenbered rnacrocyclic, retronecine diesters), anacrotine and madurensine (crotonecine esters), and the heliotridine esters, heLiosupine, lasiocarpine, and rinderine. The molecuLar structureactivity requirements for pneumotoxicity are the same as those for hefatotoxicity. it L" is consistent with their both being caus;d by the same toxic metabolites and by the metabolic activation of the alkaloids in the liver celLs to form a reactive pyrrolic dehydro-alkaloid (Culvenor et al., 1976a) Trichodesmine and incanine, found in the seeds of Triclrodesna incanum (yunusov & piekhanova, 1959), are believed to have been the causative factors of the rrozhalangar encephaLitisrr that riras endemic in Uzbekistan, USSR (Lg42-51), in which the synptons and signs were related prinarily to the central nervous system (Shtenberg & OrLova, 1955) (section 7.7).

3.3 pathways of Exposure

Naturally-occurring aninal disease is caused bv the alkaloid-containing plants gronring in fields and pastures or -43- being fed accidentally as fodder. They af,e mostly herbaceous or smal1 shrubs and many thrive in dry and arid climates. One such plant containing toxic PA alkaloids has been rePorted to gror 1tt the lrestern desert of Bgypt (Hamouda et a1.' 1984). Ttre growth of thie group of plants is particularly prolific during, and following, periods of drought, as has been reported in association with the outbreaks of human disease in Afghanistan (Tandon & Tandon, 1975,' Mohabbat et a1., 1976) and India (Tandon, B.N. et al.r 1976). Alkalold-containing plants are widespread in the tropics, especially Crotalaria, of which there are over 300 species in Africa. Ordinarily' the alkaloid-containing pLants have a bitter taste and grazitg animals wiLL reject them, unless their normal fodder is scarce. llortever, PAs often occur l-argely as N-oxides, which are said not to be bitter, and Plants containing PAs are readily eaten by some animal species. lluman intoxication may result from ttre ingestion of the toxic substance in either contamined food or herbal infusion.

3.3.1 Contamination of staple food crops

The products of pyrrolizidine alkaloid-containing plants' generally seeds, may contaminate the staPle food and may be eaten over Long periods of time. The fagt that these plants nury cause disease is generally not recognized by the people and such contamination is known to have resulted in Large- scale outbreaks of poisoning (Dubrovinskii, L952i Mnushkin, L952; Shtenberg & Orlova, 1955; Tandon & Tandon, L975; Mohabbat et a].., I976; Tandon, B.N. et al., L976, L977; Tandon, R.K. et al., L976; Krishnamachari et al., 1977; Tandon, H.D. et al., 1977) (table 2, section 3.L).

3 .3.2 l{erbal infusions

Plants have been used traditionally for nedicinal purposes all over the world. Herbs have been the nainstay of the indigenous systems of medicine, especially in China, Greece, and India, since ancient times. Table 3 includes a list of some planis that are suspected, or knownr to contain PAs and have been used as herbal medicines in different countries (Mattocks,1986). Several PA-containing plants are included arnong the 1i8t of plants used in indigenous systems of rnedicioe in India (Chopra, 1933). As a part of a research study on the etiological factors of chronic liver di$ease in Sri Lanka, Arseculeratne et a1. (1981) chenicaLl'y screened the first 50 plants used in Sri Lankars traditionaL medicine pharmacopoaea, and found that 3 of them contained PAs. A1.1 3 were hePatotoxic in rats. Of the 3, the presence of alkaloids i" @!1 -44-

Table 3. Sme plants containing (or suspected of conEaining) pAs, which have been used by people eicher as herbal medicines (I1) or foods (F)

Plant Mode Country or Reference9 of use region

BORAGINACEAE

Anchusa officinalis M Europe Broch-Due & Aasen (1980) B

Borago officinalis M USA Delorme et al. (l-977) A

Cynoglossum M East Schoental & Coady (1968) A geometricum Af ri-ca

CynogLossum M Iran coady (1973) officinale

Ileliotropium India candhi et al. (1966a); B eichwaldii Datta et a1. (1978a,b)

Il. europaeum l nd ia, IARC (1976) Greece

H. lasiocarpun Hong Kong Kumana et al. (L985); CuLvenor et a1. (1986) A

H. indicun India, Africa, scnoenEar ( I9b6ar; Souch America, Hoque et a1. (1976) b and elsewhere

H. ramossissimum Arabia Macksad et a1. (1970); B ]?aiffi)- Coady (1973) B

H. supinum M Tanzania Schoental & Coady (1968) A

Pulmonaria spp. M usA Delome et al . ( 1977) A

Syrnphytun officinale F'M Japan and Ilirono ec al. (1978, 1979b) A (confrey ) elsewhere

lt USA Furuya & Hikichi (1971); A Delorme et al. (1977) A

S. x uplandicum F, M Genera 1 Hills (1.976) B

USA Culvenor et al. (1980a,b) A

S. agperuE M USA Pedersen (1975) A

COMPOSITAE

Cacalia deconposita M USA Sullivan (1981 ) B (natarique )

C. yatabei Japan Hikichi & Furuya (1978) B -45-

Table 3 (contd).

PLant Mode Country or ReferenceS of use region

Farfugiun japonicun M Japan Furuya et al. (f971) B

Ligularia dentata F Japan Asada & Furuya (198/r) B

Petasites japonicus F, M Japan llirono et al. (1973, 1979b)

Senecio abyssinicus M Nigeria Williams & Schoental (1970) B

S. aureus M USA I.lade ( L977) B

S. bupleuroides M Africa WaLt & Ereyer-Brandwijk ( 1962)

S. burcheLli F, M South Africa Rose (1972)

S. coronatus M South Africa Rose (1972)

S. discolor M Janaica Asprey & Thornton (1955) B

S. doronicum M Germny Roeder dt al. (1980a) B

S. inaequidens F South Africa Rose (1972) B

S. iacobaea M Europe schoent4l & Pullinger (1972)t B ( ragwort ) Wade (1977) B

S. lonsilobus USA srillmn et al. (L977); A (S. douglassi) I{uxtable (1979a) B

S. monoensis USA Huxrabl€ (1980)

S. nemorensis M GerMny Ilabs et al. (1982) A "PP:.@ S. pierotci F Japan Asada & Furuya (1982) B

-Ts.-IiEi?6TlsilS. fetrorsus M South Africa Rose (1972)

S. vulqaris Europe Watt & Breyer-Brandwijk (comon groundsel) ( 1962)

Netherlands I,Iade (1977)

M Iran coady (1973)

Svneilesis Dalnata F Japen Hikichi & Furuya (1976) B

Trichodesma africana Asia omar er al. (1983) B -45-

Table 3 (contd).

Plant Mode Country or ReferenceS of use region

M Japan Culvenor et al. (1976a)

u China llirono er al. (1976b)

M Norway Borka & Onshuus (1979) b

I1 USA Borka & Onshuus (L979); Culvenor et al. (1976b)i

'.EGII}IINOSAE

Crotalaria brevidens F East Africa Coady (1973) B

C. fulva M Jamaica Barnes et al. (1964); -q Mclean (1970, 1974) A

C. incana M East Africa schoental. & coady (1968)

C. juncea M, India Chopra (1933); Watt & Breyer-Brandwijk (L962)

C. laburnifolia M Tanzania Schoental & Coady (1958) A

F Asia coady (1973) B

C. mcronata M Tanzania Coady (1973) B

C. recta Tanzania Schoental & Coady (1968); Coady (1973) B

C. retusa M, F Africa rARc (1976)

India Watt & Breyer-Brandwijk A (L962)

C. verrucosa t', Sri Lanka Araecul.eratne et a1. (1981-) g A = Reference in the reference list of this document. B = Reference in Mattocks (1986). -47-

all, of the plants reported to be etiological agents in human cases of veno-occlusive disease can be found in an inventory of medicinaL pLants used in different couotries (WflO, 1980), vhich also indicates the countries that they are used in. The above lists may not be complete as rnany Buch pLants may be used in foLk nedicine but have not been mentioned in the scientific literature. llorrrever, the lists do indicate the wide and varied use of such toxic herbs in all parts of the wor1d. Lately, there has beea a growing interest in the deveLoped countries in organically grovm products fotr food, as weLl as home remedies (Table 3), and some of the PA-containing herbs have been freel.y avaiLable in herbal shops (Schoental-, 1968; Burns, L972). Danninger et al. (19B3) listed plants containing PAs that are comonly used in tlhe Federal Republic of Germany as medicaments (Table 4). Ile also listed 9 plants in which aLkaloids have only been identified qualitatively, the toxicity of which has not been, or has been insuf- ficientLy, irnrestigated (TabLe 5). Sinilarl.y, Roitman (1983) listed L0 plants, in which the presence of pAs is suspected or has been proved and which are used as herbal teas in the USA. The f.ists incLude 10 plants containing pAs, most of which have been proved hepatotoxic experimentally, gome having highly carcinogenic promoter activity. Some of these aLkaloids have been associated with human case reports of PA toxicity. The more recent reports (Table 2) of instances of pA poisoning through the use of herbal medicines are from deveLoped countries (Lyford et al., L976; SrilLman et aL., L977; Fox et al., 1978; Kumana et at., 1985; Ridker er al., 1985). Such use of the herbs is the reason that veno-occLusive disease is endemic in Jamaica (Bras et al., L954; Jellife et al., 1954a,b; Bras & Watler, 1955; Stuart & Bras, 1"955, Lg57). There are obvious difficulties in exercising any kind of controL to restrict this use only to pla[ts that have been tested and certified as safe for human uge. It is inpossible to identify many such herbs, as they ate I sold as piaats or their amorphous products in the herbal shops. Manufactured preparations may aLso contain pA-containing herbs,, !.g.1 comfrey-pepsin capsules sold as a digestive aid (Iluxtable et al., 1986).

3.3.3 Use of PA-containing plants as food

Several PA-containing plants are used as food as can be seen in Table 3 (Mattocks, 19S6). petasites japonicus Maxim, Tussilago iarfara L. (colisfoot), anmale L. (comfrey or Russian cornfrey) are kno@ in Japan, and have been proved to co4tain carcinogenic -48-

Table 4. Medicinal plants containing PAs of knom hepatotoxicity, reported as comonly used in the Federal Republic of Germny, and the PAs contained in thern3

Fanily Genug Species Pytrolizidine alkaloids

Cmpositae Eupatorium E. cannabinum ambil ine I G;.p ag-ffifiy) supinine!

Petas i tes P. hybirdus senecionine!'9 inteeerrinina! sentirtine!

Senec io S. nenorensis fuchs isenec ionin e (ffii-a".r) 'P@ senecionine!€ (Fuch's groundsel)

S. vulgaris senecionin&'9 ac.-.i6iiilED' senecloPnyl I lng rerrorslnd riddelline!'9

jacobine! senecionine9'9 seneciphyl 1ine9 jacoline, jaconine chlorinated PAs9

S. aureus senec ionine!r9 l16EiTili' golaen ragwort )

Tussilaso T. farfara senkirkine! (coltsfoot) (coltsfoot) senecionine! '9 tuss ilagine

Alkanna A. tincroria 7-ange ly lretronec ine tr iangularine d ihydroxytr iangularine

Anchus a A. officinalis lycopsamine

Boraginaceae Borago lycopsanine/ intermedine I acetyL lycopsanine/ acetyl intermed ine amabil.ine supinine

Svmphvtum S. officinale symphytine! ( comlrey, Goffie|)- echinidine(?) lycopsamine acetyl lycopsanine9 lasiocarpine!'! heliosupine N-oxide -49-

Table 4 (contd).

Fmily Genua Species Pyrrolizidine alkaloids

S. peregrinum LvcoosaninJ S. x uolandicum interedine9 synphytine| echinidine9. 7-ace!yllycopsanine 7-ecetyl internedine symlandine upLandicine

S. asperum asperunine (prickly confrey) helioeupine N-oxide echinidine! echinatine

Gynoglogsun heliosupine N-oxide ( hound I s echinetine tongue) acetyl helio- supineS -tri

Ileliotropiun H. europaeum heliotrine!.'9r9 IiCIT6-otGI lE6iil6FGTiouope) lasiocarpine!€€ suPlnlne heLeurine europine acetyllas iocarpine9

Modified from: Danninger et al. (1983). h Tqic alkaloids. Al-kaloids known to have caused human toxicity. d Alkaloide with highly carcinogenic promoter activlty. e Used only in honeopathy. pyrrolizidine aLkaloids (Itirodo er al. , L973, L979a,b). The young flolref,-stalks of P. japonicus and the buds of coLtsfoot have been used in Japan as human food or herbaL remedies. The l-eaf and root of comfrey are also used as an edible vegetabLe or tonic (Hirono et al. , L978) in Japan and other countries (Culvenor, 1985). The carcinogenic PAs in these plants are petasitenine (!_. jgpqo&ng), senkirkine (coltsfoot), and the group includingE;FE EiiE(cornfrey). They were also mutagenic in the Ames system of SalnonelLa typhinurium and V79 hamster cell line and induced t@reserved hamster embryonic cells (Hirono et a1., 1979b). Other such PA-containing plants, used as food in Japan, include young leaves of $yneilesis paLnata, various Caialia species, and young Senecio pierotti (Mattocks, L986). According to -50-

Table 5. Medicinal plants containing PAs, reported as comonly used in the Federal Republic of Gernany, lhe toxicity of which has not been, or has been insufficiently, investigatedS

Fanily Species

Compositae Eupatorium E. perforatum

Brachyglottis B. repens

Arnica A. montana (mountain atnica)

Boraginaceae Lappula L, intermedia (stickseed)

Pulmonaria P. officinalis ( lungwort )

a Modified from: Danninger et al. (1983).

Culvenor (1985), consumers of comfrey could be ingesting up to 5 rng PAs per day. Rose (1972) listed a number of pLants of the genus Senecio that are used as spinach in South Africa. These include S. burchelli, which is known to have caused an episodeotra!@roughtheingestionofcontaminated bread (Wilnot & Robertson, 1920).

3 .3.4 Contaminated honey

In the USA, Deinzer et aL. (L977) reported the presence of aLl" PAs contained in Senecio jacobaea (ragwort) and proved to be hepatotoxic, in the honey secreted by bees feeding on the pLant. The total alkaLoid content ranged from 0.3 to 3.9 rng/kg. It has been estimated that an average annual human intake of honey (6OO g) at the highest alkatoid level quoted r,rould contain less than 3 rng of PAs (Mattocks, 1986) . CuLvenor et al. (1981) and Culvenor (1983, 1985) drev attention to the same potential hazard in honey from Echium pLantagineum, a weed that grows wideLy in Southern AusEffi and is a major source of honey, yielding an estimated 2000 - 3000 tonnes per annum for human corisumption. Echirnidine is the major component of the alkaloids of Echium, which are present in concentrations of up to L ng/kg-culvenor (1983) estimated that individuals may consufle up to 80 g honey/day with a corresponding alkaloid intake of 80 vg/da1y, if only the Echium honey rdere used. No reports of acute human toxicity through this source are available. - 51 -

3.3.5 Milk

PAs have been shown to produce toxic effects via trans- ference into the niLk of dams (Schoental, 1959). Retrorsine nas administered orally to 17, and intraperitonealLy to 6, lactating rats weighing 185 - 350 g in 5 - L0 mg doses daily, the first dose being given during the first 24 h after parturition. The rats received from 1 to L4 doses, the total intake amounting to 2L - 335 olglkg bodf weight. The Litters were separated from the mothers for Ll2 h following the administration of PA to avoid direct contamination of the forner by Licking. Apparently the nilk production lras not affected as the stomachs of many of the young, examined postoortem, lrere distended lrith railk. All animals whose mothers had received a total dose of 138 ng PA or more died within 30 days. Many of the young whose mothers had received smaller doses survived until they were killed at 6 months. Biopsy of the liver of the young at various intervals or at autopsy sholred marked changes, even in caseg nhere the mothers did not appear to be affected. Aninals dying at 18 - 30 days showed hydropic or fatty vacuolation of liver cells. In the liver of animals dying or killed later, various degrees of haemorrhagic necrosis and increase in the centrilobular reticulin of the liver, and some thickeninrg of centrilobuLar veins nere seen. In anir"als that survived 6 months, the appearance was less abnormal, but some hyperplastic nodules and bile-duct proliferation were aeen. The Lactating rats dosed with the PAs generally survived longer than the suckling animals and usually did not show any iLl effects, suggesting that the susceptibility of the suckling rats vas greater than that of the mothers. Dickinson et al. (1976) demonstrated the presence of pAs in the niLk of dairy cattLe fed or dosed with ragnort (Senecio .iacobaea). When 4 cows were administerea ttre Arieifl E naterial at levels of up to 10 g/kg body weight per day through rumen cannula, PA levels of up to 0.84 ng/kg were observed in the nilk. Holrever, only one (jacoline) of the several PAs contained in the plant nas secreted. Calves, bucket fed on the milk did not show any signs of PA toxicity. Dickinson (1980) repeated the study on goats. Four milk goats lrere freshly prepared with rumen carrnulae. The kids lrere separated from their dams and were fed nilk twice a day. Dried tansy raglrort plant material vith a PA content. of O.L6"A (dry weight) was administered through the cannulae to each goat at a dosage rate of 10 g/kg body weight per day over 125 days. During this period, each of the 4 kids received milk from their dams at approximately L25 nl/kg per day in addition to ad lib feeding on alfalfa grass hay. Six PAs lrere isolated from the plant material: jacobine, jaconine, jaconline, -52-

jacoziae, senecionine, aod seneciphylline. uilk samples collected twice daily showed PA contents of 225 530 ug/litre with a rnean of 381 pg/litre. No apparent health effects were noted in the kids, and only nild hepatic damage \ras suspected in the dams, on the basis of Liver function tests. Fifty percent of the kids lrere killed after l-0 weeks. No lesions of PA toxicity were seen. The dams were rebred and appeared nornal throughout the gestation period. Itowever, three dams aborted at aLmost fuLl tero, and the fetuses were born dead. One of the dams died shortly after parturition and showed evidence of severe liver damage characteristic of PA toxicity. Another, which delivered normall.y, also showed a lesser degree of liver danage at b iopsy. Data relating to PA secretion rrere compared with sinilar earlier data on cows. Mean secretion of PAs in cows appeared rnuch higher, e.g., 684 ug/1itre. The authors concluded that the amount of PAs secreted in the goatrs milk did not cause any serious deleterious effects in the kids. Johnson (L976) fed long-term lethal doses of Senecio jacobaea, by stomach tube, to 6 cows. Feeding started ffi E-iffifin 30 days post-partum, and continued until what was considered to be a lethal dose had been fed. The daily dose of the plant ranged from L to 4.4 g/kg body weight, the total amount fed repre'senting 5 - L5% of body weight over a period of 54 - L26 days. Five cows died within 98 days; one, in a moribund state, was killed on day 126. The calves suckled for 30 - 126 days. Suckling started iurnediately after birth i.n the case of 4 calves and 10 and 30 days later, respectively, in the 2 renaining calves. Three calves rilere killed with their dams or soon after, and 3 were retained for 1 year for observation. Milk samples from 2 cows were collected and pooled in 14- to L6-day lots during 64 days of feeding of the Senecio plant. Each pooLed sample was administered intragastrically to a group of rats in daily doses of 12 nl for L5 - 30 days. A control group of rats were fed rav nilk from cowa not fed Senecio. Blood samples of the dams and the calves were an?@d- for glutanic oxaloacetic transaninase (cOT), Lactic dehydrogenase (f.off) , and gama-glutamyl transpeptidase (GGTP). Serurenzyme leveLs in all colts iadicated statistically significant deviations suggesting liver dysfunction, and ihe livers at autoPsy had characteristic features of PA toxicosis. The LDII and GOT levels in calves were generally abnormal after 20 - 45 days of suckling. The abnormalities ranged from mild to a 15- to L70-fo1d increase. One calf was autopsied at the peak increase of serurenzymes and was found to have mild focal hepatitis. No significant pathological features nere seen in the livers of other animals, nor of the rats, some of which were retained for up to 150 days. -53-

Goeger et a1. (1982) fed dried Seneclo jacobaea (tansy ragwort) to Lactating goats in a pr6!ffii-ii-757- of the feed. The niLk contained 7.5 uE PAlkg dry weight. The nilk produced by the goats nas pooled and then bottle fed to appetite to 2 Jersey bu1l calves (1 day old) that also had access to tansy ragwort-free hay for 109 and 124 days, respectiveLy. They were then weaned and given nornal feed and observed for 6 months, after which they lrere kiLl.ed and autopsied. In anoEher study, rata nere fed a diet containing the freeze-dried nilk at 807" level for 180 days with a calculated total PA intake of 0.96 mg/rat. other groups of rats nere fed tansy ragnort at dietary LeveLs of 0.01 L0 g/kg (corresponding to PA intakes of 39.77,5.04, 0.52, and 0.05 mg/rat). The calf livers only showqd very nild non- specific changes, but the Livers of rats fed tansy ragrirort or the nil"k from tansy ragwort-fed goats shoned definite, but mild, changes including swoLlen hepatocytes, megalocytosis, biLiary hyperpLasia, and fibrosis. Histopathological changes in rnilk-fed rats lrere similar to those in the group fed tansy ragnort in the diet at 0.01 g/kg. The authors concluded that there was evidence of PA transfer into milk, which proved hepatotoxic for rat\ It was aLso noted thaE the goats had been fed high LeveLs of tansy ragnort at the upper linit of their acceptance, and that the hepatic changes observed in rats fed high levels of Inilk, for extensive periods, were s light. Luthy et al. (1983) produced direct evidence of excretion of macrocyclic esters of retronecine of the senecionine and seneciphylline-type into rat milk. sll-retronecine, an 3H-necic acid-labelLed senecionine, and seneciphylline were prepared biosynthetically with seedLings of Senecio vulgaris L. Two lactating rats (Ivanovas, Sprague Dawley), weighing 300 - 400 g, were fed the first of the second compouad by stomach tube, in doses of 2.7 mglkg and 5.5 mg/kg body weight, reepectively. Samples of blood were examified 1, 3, and 6 h after treatment, and those of nilk 1 and 3 h after. Anirnals were kilLed after 6 h. They were found to have excreted approximateLy 0.087" of the applied radioactivity in the nitk within 3 h, mainly as unidentified tetronecinederived metabolites, and approxinately 0.02% as unchanged PAs. The highest Levels of PAs and metabolites in tissues were found in the liver and t-ungs, 6 h after administration. Candrian et a1. (1984a) al-so demonstrated that Drosophila melanogaster flies fed on nilk from lactating rats that had been adrninistered an oral dose of seneciphylline showed 1.22 sex-l.inked recessive lethals, compared with 0.3% in controls, indicating the transfer of the mutagenic properties of the PA via miLk (section 6.4.7). -54-

The irnplications of the above etudies on the possibility of carrfover of PAs into foodstuffe of aninal origin are obvious. However, no reports of hrrman caaes of acute PA toxicity, ascribed to the conaumption of contaminated nilk, are available.

3 .3.6 lleat

There have not been any reports of the detection of PAs in meat products from livegtock expoeed to them.

3.3.7 Use of PAs as chemotherapeutic agents for cancer

An alkaLoid of @ L. (indicine N-oxide) has been found to hffiity and has b&n used in experimentaL clinical chemotherapy for cancer (section 7.9). -55-

4. METABOLISM

4.1 Absorption, Excretion, and Tissue Distribution

4.1.1 Absorption

There have been few studies on the absorption of PAs in man, but absorption has been inferred fron studies on tissue tlistribution and the amounts of alkaloicls and their metabolites excreted in the urine, faeces, and biLe of animals ( section 4.L.2) . Swick et al. (1982c) measured the transfer of a mixture

4.L.2 Excretion and distribution

The excretion and distribution of heliotrine in rats has been reported in Bull et a1., 1968. Young rats (150 g), given the LD5O of heliotrine by ip injectio4, were killed at intervaLs, bled quickly, and the organs and tissues analysed. Heliotrine $ras present in the liver after 2 nin (3.77" of totat dose), the level peaking at 5 min (6.3D, and dropping xo 2.27" at t h and 0.5% at 2.5 h. In adult rats, the level in the liver at 5 h was O.O7"Aof the total dose. Five min after dosing,30 - 401( of the initial dose remained in the peritoneal cavity, and the blood Level of heliotrine rras 60 nglLitre, dropping to 3 mglLitre at I h. The urinary excretion of base and metabolites other than pyrrolic metabolites, collected and measured 16 h after adminisiration of several alkaloids by ip injection, is shown in Table 6. The proportion of base excreted unchanged increased with the hydrophiticity of the alkaloid, being 62'l for heliotrine N-oxide, 30% for heliotrine, and only I L.5% for lasiocarpine. Ileliotridine, the hydrolysis product from heLiotrine -56-

Table 6. Urinary metabolites oi pyrrolizidine bases in the rat ( 16-h urine)3

urine constituent (amount in percentage of dose injected)

Base adninistered Unchanged Base ttelie Helio- Helie ltelio- ( ip injection) base N-oxide tridihe tridine tridine tridine trachel- trachel- N-oxide anthage anthate N-oxide

Heliotrine 30 rrace lu 3 15 I{el iotr i-ne 62 (62) 2.7 ca. 6 ca. 1 ca. 10 N-oxide

L?siocarpire I-I.) l. )-J b Heliotridine ca. l (35) ( ca. 1) 5ZO trachelanthate lieliotridine 40 zo (40) (20)

3 From: Bull- et al. (1968).

and lasiocarpine, was excreted in the form of the N-oxide in larger quantities after the administration of each- of these alkaLoids. The distribution and excretion of monocrotaline lraa studied in rats by Hayashi (1966) who found that 50 - 7O"Attas excreted in the urine vithin the first day. llonever, the analysis was by a non-specific chemicaL method that did not distinguish between the unchanged alkaloid and its metabolites. Uattocks ( 1968a) gave toxic pyrtolLzidine alkaloids intraperitoneally to maLe rats and measured the urinary excretion of the unchanged alkaloid, and of N-oxide and pyrrolic metabolites. The excretion of N-oxiEe and unchanged alkaloid was rapid and almost compLete ln the first 24 h. Excretion bf pyrroles nas also rapid but continued for a Little longer. For example, in rats given retrosine (60 ng/kg body weight), the urine in the first 24 h contained 10.62 unchanged alkaloid, 13.32 N-oxide, and 13.42 pyrrolic metabolites. During the second da1y, onLy 0.12 al.kalola, O.27. N-oxide, and 1.82 pyrroles were excreted. Biliary excretion also occurred. About one-quarter of an iv dose of retrosine in rats lras excreted in the bile as pyrrolic metaboLites, and 4Z as unchanged aLkaloid; most of this excretion occurred during the first hour afrer rhe injection (I,thire, Lg77). Jago et al. (1969) gave heLiotrine iv to sheep; urinary excretion of the unchanged al.kaloid together with metabolites (!-oxide, and denethylation and hydrolysis products) occurred -57-

rapidly and continued for up to 8 h. Excretion in the bile was onLy 2% ot thax in the urine. Ths tissue distribution of radioactivity from a tritiated toxic pyrrolizidine alkaloid analogue vas Etudied by Mattocks & White (L976) using synthanecine A bis-N-ethyLcarbamate (40 rng/kg body weight). The highest concentrations of radio- ."ti"-ity-t.". in the LivEr (where uetabolism occurs), lungs, kidneys, "."iand spleen (respectiveLy, 3.97", O.Lgi(, 0.182' and 0.277 of the dose given), and about 69i( of the dose was eliminated in the urine during the first day. Radioactivity in the expired air waa negligible. The binding of radioactivity in the liver, and especially the lungs, was more persistent than in other organs. Sinilar results were given by the semisynthetic pyrrolizidine atkaloid analogue, retronecine bis-N-ethylcarbamate (ltattocks, L977). The distribdtion of the uniformly tlC-labelLed natural pyrrolizidine alkaloid senecionine in lactating mice rrtas itrrdied by Eastman et aL. (1982). After 16 h, 757" of the radioactivity had been recovered in the urine, L47" in the faeces, but only 0.04% was in the nilk; the Liver -contained L.gzit.' The nice were milked using teat cups. Candrian et a1. (19S5) studied the distribution of radioactivity in rats given sna11 doses of senecionine or seneciphylline (0.3 (retronecine) 3.3 ng/kg),- tritiated in the pyrrolizidine moiety. Most radioactivity ltas eliminated in the urine and faeces within 4 days. Using mass specttooetry, Dickinson et (1976) found a concentration of up to 0.84 ng ?As/litre in the"1. milk of cows fed Senecio jacobaea. Blood levels of senecionine in rats gir6i-6@S wet. determined by Culvenor (197S). The levels nere 0.38' O.32, and 0.14 ng/l-itre at 0.5, 1, and 2 h after injection, resPectively. To sumarize, the avaiLable evidence suggests that ingested toxic pyrrolizidine alkaloids are rapidly metabolized and that the excretion of unchanged alkaloid and of most metabolites is al-so rapid. Thus, within a fer'r hours, only a relatively smalL proportion of the dose r€mains in the body, much of ttris in the form of netabolites bound to tissue constituents. It appears inprobable that a significant amount of unchanged alkaloid will remain in the body after the first day' Pyrrolizidine N-oxides are much more atater soluble thar their parent alkaloids. Indicine N-oxide (which is exceP- tionaLly water soluble) is very iapidLy excreted, either unchanged or conjugated. Thus, indicine N-oxide given iv to mice, monkeys, or rabbits disappeared ftom the serum with initial ha!'f-lives ranging fron 3 to 20 nin (Powis et al" 1979; El Dareer et a1., 1982). over 8O7"of tritiurlabelLed indicine N-oxide given iv ltas excreted in the urine of mice or monkeys w-ithin 24 h (El Dareer et a1., L982)i at' 2 h, the highest concentrations of radioactivity nere in the kidneyst -58-

liver, and intestines. Urinary excretion of indicine N-oxide was also rapid in rabbits, but somelrhat slower I.n numan Der.ng6 (Porie et al., L979).

4.2 Metabolic Routes

The major metabolic routea of unsaturated pyrrolizidine alkatoids in animals are: (a) hydrol.ysis (of the ester groups),'(b) I-oxiaationi and (c) dehydrogenation (of the pyrrolizidine nucteus) to dehydro-aLkaloids (pyrrolic derivatives). Other minor routes of metaboliam are known, but the three pathways account for the major known toxic effects of these alkaLoids (Fie. 5). Routes (l) and (b) are be!.ieved to be detoxification mechanisms. Route (c) Leads to toxic rnetaboLites and appears to be the major activation mechanism. Route (a) rnay occur in various tissues, including the tiver and blood. Routes (b) and (c) are brought about in the Liver by the microsomal mixed-function oxidase system.

DEHYDBO. ^.._^..fu-;.H*.ALKALOID primarytoxic metabolite \ \ ?\ DEHYDRO. NECINE

!-oxrDE possiblefurther metabolites

Fig. 5. Major metabolic routes of unsaturated pyrrolizidine alkaloids .

4.2.L HydroLysis

The hydrolysis of a PA Leads to the formation of the amino-alcohol moiety (necine base) and the acid moiety. Neither of these is hepatotoxic (Schoental & Mattockst 1960; Culvenor et al., L976a). The highly nater-soluble necine base is readily excreted, is not accessibLe to the microsomal -59- aystem, and iB not activated to a toxic metabolite. Thus, pyrrolizidine alkaloids that are very susceptible to (enzynic) hydrolysis have 1ow toxicity (Mattocks, 1.982). A major factor contributing to resistance to esterase is the steric hindrance in the acid moiety. Thus, the chain branching near the carbonyl groups slows hydrolysie allowing the fornation of relatively high levels of pyrrolic metabolLtee; a conformation of the basic moiety, which brings the tlro ester groups close together, thus leading to mutuat steric hindrance, cao also prevent hydrolysis (Mattocks, 1981a). The influence of hydrolvsis in vivo on alternative metabolic pathways is demonst"aied bfrhe fact that treatment of rats nith an esterase inhibitor, before giving pyrrolizidine aLkaloids (or synthetic anaLogues), can lead to greatly incteased production of pyrolic metabolites fron alkaloids that are nornally susceptible to hydrolysis, but little increase in those from alkaloids norlial.ly resistant to hydrolysis (Mattocks, L98La).

4.2.2 N-oxidatioo

The N-oxidation of pyrrolizidine alkaloids is induced by the hepatic microsomal enzJrmes. The N'oxlde metabolites are highly water solubLe and are rapidly excreted in the urine (Mattocks, 1968a). Pyrrolizidine N-oxides are not converted to any significant extent to pyrr;lic metabolites by nicro- somal enzymes (Jago et al., 1970i ltattocks & llhite, 197la), and there is no evidence that they ar€ toxic, unless first reduced to the corresponding basic alkaloids, which can then be activated by the microsomaL system (Mattocksr l971c). Thus, it appears that the formation of N-oxi.des represents a detoxif ication pathway.

4.2.3 Conversion to pyrrolic metabolites

In Laboratory animals, toxic pyrrolizi.dine alkaloids are netabolized to pyrrolic derivatives, so-called because the unsaturated ring of the pyrrolizidine system loses 2 hydrogen atoma to form what is in effect a pyrrol.e ring (though the structure is more correctLy a dihydropyrol.izidine). Pyrrolic netabolites are easily detectable in the tissues shortly after giving a toxic pyrrolizidine alkaloid to an animal, by treating the tissue with an Ehrlich reageot containing boron trifluoride, when a red colour is produced; this reaction also occurs lrith the urine (Mattocks, L968a; Mattocks & !'lhite, 1970). In rats given retrosine, pyrrolic metabolites were found prineipally in the Liver, with highest levels associated with the microsomaL and soLid debris fractions and less in the mitochondrial fraction; low levels were found in the Lungs, -60-

heart, spleen, and kidneys, wiBhin 4 h of giving retrosine. Rats given 60 ug retrosine/kg body weight excreted L4"l of rllre dose in the urine, within 48 tr. Pyrrolic metabolites are formed by the hepatic mixed- function oxidase ayatem, with a requirement for cytochrome P450, oxygen, and NADPH, as has been demonstrated in vi.tro (Jago et a1., 1970; uattocks & white, 1971a). convffiiTF pyrrolizidine aLkaloids to pyrro!.ic metabolites by the lung tissue of the human embryo (Armstrong & Zuckerrnan, 1970), rat (llattocks & White, L97La; HitLiker er al., 1983), or rabbir (Guengerich, 1977) was negligible. The fornation of pyrrolic metabolites does not proceed via N-oxide intermediates, but appears to resuLt from an initial hydroxyLation of the unsaturated pyrrolizidine ring adjacent to the nitrogen atom (Mattocks & tltrite, L97La:, Mattocks & Bird, 1983). This wouLd lead to a chemically unstable intermediate that would be expected to decompose spontaneousLy to the pyrroLic product. The prinary pyrolic metabolites (or dehydro-alkaloid) formed by dehydrogenation of pyrrolizidine alkaloids are chemicaLly dehydropyrrolizidine esters (fig. 5). These are highly reactive compounds that can rapidly react. with tissue constituents or hydrol"yse to the corresponding pyrrolic alcohols, or dehydro-necines, which can thus be regarded as secondary metabolites. The Latter can also react with tissue constituents, but nore slowly. Because of their high chernical reactivity, the prinary metabolites would be expected to have a short life in the liver cel1 (minutes or seconds) before they are hydrolysed or react with nucleophilic tissue constituents. Sone rnight escape into the blood stream and reach other organs, especialLy the lungs. Dehydro-necines are more stable and also more nater soluble, and can become more widely distributed throughout the body. However, they are also capabLe of reacting with tissue constituent's. Thus, measurements of pyrroles formed from pyrrol"izidine alkaloids in tissue samples, using a colour reaction (Mattocks & White, 1970), wil-L not represent a single metabolite, but mixtures of the netabolites together with various reaction products of these nith tissue constituents. It wiLL be seen (section 5) thbt pyrrolic rnetabolites are believed to be responsible for major toxic actions of pyrroLizidine aLkaLoids (Mattocks, L972a). A pyrrolic metabolite with reactivity nidway between that of dehydromonocrotaline and dehydroretronecine has been reported to be formed from monocrotaline in isoLated, perfused rat liver (Lafranconi et al., 1985). Studies on this metabolite (isolated from bile) indicated that it is a monoester, and that it is toxic in perfused rat lung. This suggests that, monoester pyrroLic metabolites may play a part in the toxic actions of PAs in extra-hepatic tissues. -6L-

Vlhen rats or other laboratory animals are given a toxic pyrroLizidine alkaloid, pyrrolic netabolites accumtrlate rapidly in the liver (Mattocks, L973i White et a1., L973), reaching a peak within I - 2 h, then falling slowly during the next 24 }J.', the metabolites may still be detectable after 2 days. AccumuLation is especially rapid after intraperitoneal injection, a very high level of pyrrolic metabolites being attained within 20 minl this indicates how rapid the netabolisn of pyrrolizidine aLkaloids can be. The level of pyrroLic metabolites in rat liver is generalLy directly related to the amount of alkaLoid given 2 h previously, at least up to an acute LD5g dose. The pyrrole level depends on the alkaloid used, and is related to the acuEe hepatotoxicity of the alkaloid (Mattocks, I972a). To be converted to the type of chenically reactive, toxic pyrrolic metabolites described above, an aLkaloid must possess a l-hydroxynethyl pyrrolizidioe systeo, urrsaturated in the L r2-position (this makes the ring susceptible to dehydrogenation), and at least one hydroxyl group Eust be esterified, usually by a branched-chain acid. Otonecine esters are converted to similar pyrrol netabolites by a different media involving N-demethylation. Pyrrolizidine amino-aLcohols (e.g., retronecine) are not metabol.ized to more than srDall amounts of pyrroles (Jago et al., 1970; Mattocks' 1981a), possibly because they are too lrater soluble to reach the rnicrosomal enzJrmes. The metabolic fornation of pyrroles is catalysed by cytochrome P450 and specificity exists in the various isozymes (Guengerich,1977; Juneja et al., 1984). A few non-toxic pyrrolizidine aLkaloids (e.g., rosmarinine and hygrophylline) are converted to pyrrolic metabolites in yilrg (Mattocks, 1973). Such metab;lites are cheroicali! EiT?erent from the pyrroles of toxic alkaloids, and they are neither reactive nor toxic (Mattocks & I'lhite, 197lb). The balance of structural features necessary f,or a pyrrolizidine alkaloid to be converted to give high concentraEions of toxic pyrrolic metabolite has been discussed by Mattocks (1981a); the optimum conditions appear to be met in some alkaloids that are macrocycLic diesters, such as retrosine.

4.3 Effects of Treatments Affecting Metabolism

The fornation of pyrrolic metabo!.ites (and of N-oxides) is aLtered by treatments that affect the hepatic microsomaL enzymes. Such effects have been studied by measuring rates of metabolism of pyrrolizitline alkaloids in vitro using microsomal preparations from animals pre-tr-6EGi-ii- varioui ways (rable 7). For exanple, microsomes from rats given the microsomaL erl.zyme inducers phenobarbitone or DDT (but not -62-

Table 7. Effect of pre-treatment of male rats on the conversion of pAs to pyrrolic derivatives and to N-oxides by !-iver microsomes in vitro

A lkaloid Pre-treetEent, and EnzyEe activity as Reference time before enz)me Z of controL vaLues measurementa for fomarion of: Erroi6-s E-oxfdes

retrorsine phenobarbitone, ip, 311 232 llattocks & 3 x 100 ng/kg, white (197la) l-3days

retrorsine DDI, ip, 75 ng/kg, 4O7 203 Mattocks & 3 days $rhite (1971a)

retrorsine 3nethylcholanthrene, 95 (na) 116 (ns) Mattocks & ip,3x20ng/kg, white (197la) I - 3 ilays

retrorsine retrorsine, ip, 63 - Mattocks & 35 ng/kg, 20 h white (197la)

retrorsine protein-free diet, 39 - Mattocks & 3 days r.rhire (1971a)

nonocrotaline phenobarbitone, sc, 448 165 Ch€sney et a1. 4 x 75 mglkg, (L974) 1 -t 4 daye

nonocrotaline chloramphenicol, se, 10 109 (ns) cheoney et al. 200 ng/kg, t h (r974)

monocrotaline SKF 5254, ip, 75 ng/kg, 10 87 (ns) Chesney et aL. lh (L974)

nE = not significantly different frm controls.

those from rats given 3-nethylchotanthrene) induce greatly increased pyrrole fornation and smalLer increases in N-oxide-White, formation, from the alkaloid retrorsine (Mattocks & 1971a). Bnz]me preparations from rats treated with inhibitors of microsomal enzymes, including SKF 525A and chloramphenicol, are nuch less active in converting monocrotaline to pyrroles (Chesney et al., L974). The ability to metabol-ize retrorsine is dininished in microsooes from rats fed a protein-free diet, or from rats acuteLy poisoned lrith retrorsine (Mattockg & White, 197la). -63-

The effects of in vivo treatrnent with several tyPes of enz)rme inducers oi-ffrf toxicity of lasiocarpine and senecionine for prinary rat hepatocyte cultures waa investigated by llayes et al. (1985). Pre-treatment with phenobarbitone potentiated the cytotoxicity of senecioniae tonards the cultured celLs, whereas pre-treatment with 3-nethylcholanthrene dininished the toxic action of senecionine, but had little effect on lasiocarpine cytotoxicity. The cytocidal effects of both alkaloids ltef,e substantialLy inhibited in the presence of $KF 525A.

4.4 Other Factors Affecting Metbbolism

Variations between anirnal species have been investigated by Vttrite et al. (1973) and Shull et al. (1976). For instance, metabolism to form pyrroles is high in rats and very Low in guinea-pigs, which, however, have trrigher rates of N-oxidation. For exampLe, 2 h after an ip dose of retrosine (100 rng/tg body weight), the Liver-pyrrole level in maLe rats was 13 tines higher than that in nale guinea-pigs (!i?rite et aL., L973). Liver microsome preparations from nale rats were 28 tinres more active than microsomes fron nale guinea-pigs in the dehydrogenation of monocrotaline (Chesney & A11en, t973a). The development with age of the ability of Wistar rats to metabolize retrosine was studied by Mattocks & White (1973). the ability to form pyrroles is very Lowr in nerrr-born rats, but, by 5 days of age, it is neatly as high as in adult males. This activity continues at a sinilar level in maLe ratsr but, in females, it fal1s after the age of about 20 days until, by 60 days, it is about one-eighth that in males. Such a sex difference lras not observed in mice (Wtrite et al. , L973).

4.5 other Metabolic Routes

The actions of hepatic microsomaL enzynes on pyrrolizidine alkaLoids can produce other metabolites as weLl as pyrroles and N-oxides, but there are felr reports of these. Eastman & segalL (19S2) demonstrated hydroxylation of the acid rnoiety of senecionine by liver microsomes from femaLe mice. Such metabolism shouLd not prevent the subsequent conversion of the product to pyrrolic or N-oxide metabolites. The formation of other microsomal metabolites of senecionine has been rePorted by Segall et al. (1984). The O-demethyLation of the acid moiety of heliotrine has been dernonstrated by Jago et al. (1969) and rePresents a partial detoxification mechanism, since the Product is about half as toxic as heliotrine. Other detoxification nechanisns exist in the rumen of sheep (Dick et 41.' 1963; tanigan & -64-

Srnith, 1970arb), which are, thus, particularly resistant to the effects of pyrrolizidine alkaloids.

4.6 Metabolisn of Pyrrolizidioe N-Oxides

As mentioned in section 4.2, the N-oxides of pyrrolizidine alkaloids are not converted to pyrrolic netabolites by liver microsomes. It appears that their main route of metabolism in animals is reduction to the corresponding basic alkaloids, which nay then be further metabolized as already described. This reduction has been shonn to occur in the rat or rabbit gut (Mattocks, 1971c; Powis et al., L979), and may be brought about by intestinal bacteria or possibLy by gut enzymes. Such reduction can also be brought about by hepatic microsonal fractions (Powis et a1., 1979) in the presence of NADII or of NADPIi, and by sheep rumen fluid (tanigan et aL., L970a, b). The reduction of pyrrolizidine N-oxides in vivo is of great importance as a step in the bioactivation of these compounds (Mattocks, l97Lc), as shown in section 4.2.2.

4.7 Metabolisn in llan

Powis et al. (1979) found that indicine N-oxide given iv to 3 human patients as an antitumour drug wao partial"Ly reduced to indicine base, detectable in the urine and plasma. Armstrong & Zuckerman (l-970) showed that human embryo Liver slices, but not lung slices, converted the pyrroLizidine alkaloids lasiocarpine, retrorsine, and fulvine to pyrrolic metaboLites in vitro. -65-

5. MECHANISI.{SOF TOXICITYAND OTHER BIOI,OGICAT ACTIONS

5.1 Metabolites Responsible for Toxicity

5.1.1 Metabolic basis of toxicity

. The toxic effects of pyrrolizidine alkaloids are rnediated through their toxic metaboLites and not by the alkaloids themselves. The folLowing observations are evidence for the above statement (Mattocks, 1972a) :

(a) The alkaLoids are chenically rather unreactive and it is hard to envisage reactions with cell constituents that t.hey could undergo readily under physioLogical conditions. On the other hand, chenically prepared derivatives, similar or identical to known metabolites of these alkaLoids, are highly reactive and are capable of causing toxic effects simiLar to those of PAs, often at dose leveLs much lower than those required by the alkaloids themselves.

(b) The l-iver is usuaLly the nain organ affected, $rhat- . ever the route of adninistration of the alkal-oid. The aLkaloids are known to be metabolized in the liver.

(c) Direct appl.ication of these alkaloids to the skin does not cause locaL toxic effects (Schoental- et aL., L954), nor do cytotoxic effects occur at sites of injection.

(d) The susceptibility of animals to the toxic actions of PAs is related to the ability of the animaL to metabolize the alkaloids. For example, the hepatic microsomal. enzJrmes of rats less than t h old have very 1ow activit,y towards retrorsine and these rats are relatively resistant to it, whereas rats aged severaL days have a high enzyme activity and are highLy susceptible to the aLkaloid (Mattocks & White, L973). Guinea-pigs are very re$istant to retrorsine, unl.ess they have been given phenobarbitone, which potentiates the enzymes that metabolize i.t (Wtrite et al-., 1973)" Rats pre-treated with microsomal enz]rme inhibitors, such as SKF 525A or chloramphenicol, have increased resistance to retrorsine or monocrotaLine (Allen et al.; t972;' Mattocks, 1973) . In general, there is a good relationship between the rate of hepatic metabolism of pAs to pyrrole in vitro (shul1 et Lg76) and chronic toxicity. nTg-EIfEsistant species,"1., €.g., guinea-pigs, Japanese quail, and sheep, have a Low rate of pyrrole formation, while susceptible species, such as the horse, -66-

cattle, and rat, have a high rate. Notable exceptions are the rabbit and hamster, which have high rates of pyrroLe formation, but are resistant. It is possibLe that this may be due to changes in the balance betlreen activation and the involvenent of other factors, sueh as activity of detoxification. For example, sheep have a high epoxide hydrolase activity in the liver (swick et al., 1983), which rnay affect PA detoxification (Cheeke & Pierson- coeger, 1983).

5.L.2 IsoLation of pyrrolic metgbolites

There is plenty of evidence that many unsaturated PAs are converted into pyrrolic esters (dehydro-alkaloids) in the maromalian liver (section 4.2.3). These prinary pyrrolic metabolites cannot be isoLated, because of their high reactivity and rapid rate of hydrolysis. Ilolrever, their more st.able hydrolysis products (pyrrolic alcohols; dehydronecines) have been isolated and identified. Thus dehydroheliotridine has been obtained from the in vitro incubation of both the heliotridine-based alkaloids@arpine and heliotrine, with rat f.iver microsomes (Jago et aL., l97O) and dehydroretronecine nas found to be the main detectabLe pyrrolic metaboLite in the 1iver, bLood, and urine.of rats injected with the retronecine-based alkaloid, monocrotaline (Ilsu et aL., 1973). There is evidence that these material.s are identical, i.e., the (t)-forn resulting from racemization during hydrolysis of the parent pyrroLic esters (Kedzierski & Buhler, 1985). The results of these studies confirm that rat liver enzJrmes convert PAs into metabolites with known cytotoxic activity (section 5.2), and imply that these metabolites are formed via the yet more toxic and short-Lived dehydro- alkaLoids (Jago et al., 1970).

5.1.3 ChemicaL aspects of pyrroLic metabolites

5.1.3.1 Preparation

Chemical methods are availabLe for converting unsaturated PAs into pyrrolic esters (dehydro-alkaloids), the putative prinary toxic metabolites, enabling the physical, chemical, and toxicological properties of the latter to be studied. Snal1 amounts of dehydro-pyrrolizidine alkaloids are usually prepared by the reaction of the corresponding alkaloid N-oxides with either acetic anhydride (Mattocks, L969; duLvenor et al. , L97Oil or methanolic ferrous sut fat.e (Mattocks, 1969). The products mugt be protected from -67-

moisture and from acids, which can cau$e their irmediate decomposit ion. A variety of reagents can dehydrogenate the alkaloid bases to pyrrolic derivatives, these include nnnganese dioxide (Culvenor et aL., L970arb; Mattocks, L96g), potassium permanganate (CuLvenor et al., 1970a), chloranil (Culvenor et a1., 1-97Oa), 2r3-dichloro-5r6-dicyanobenzoquinone (Mattocks, L969), iodine (Culvenor er al., 1970b), and aryl thiols (Juneja et aL., 1984). Socre PAs are s1ow1y oxidized to pyrroles by moLecular oxygen (Bick et aL., 1975). The more stable pyrroLic- - alcohol, dehydroretronecine (rig. 6) , is prepared from retronecine .t"iog chloranil (Culvenor et al., 1970a) or aqueoua potas;iun nitro- sodisulfonate (Mattocks, 1981c) or from retronecine N-oxide ( isatinecine) using ferrous sulfate (Mattocks, 1969): The enantiomeric dehydroheliotridine can be prepared from heliotridine in similar lrays. Racemic dehydro-heliotridine has been synthesized (Viscontini & Gilhof-Schaufelberger, 1971; BohLnannet al., L979).

Fig. 6. structures .f (nl = , 9")Igt"h_:liotridine ontnz = H) and dehydroretronecine(Rr = HlRz= OH).

5.L.3.2 Chemistry associated lrith toxic actions

Dehydro-pyrrolizidine alkaloids and dehydronecines (pyrrolic esters and alcohols) act chemically as alkylating (electrophil"ic) agents, i.e., they can react with compounds possessing electron-rich (nucleophilic) groups, as amines, thiols, and some hyilroxyl compounds. The products"rrch of ttpyrrole[ al"kylation consist of the moiety covalently bonded to the substrate molecule. The mechanism of alkyLation is illustrated in Fig. 7 (Mattocks, L972a). An esrer (R = COR) or hydroxyL group (R = H) attached to the pyrroLe ring via one carbon atom (i.e. , at C7 or C9) is highly reactive, being easily cleaved, leaving a positively charged pyrrole moiety with a high affinity for electron-rich substrales. pyrrolic esters are the most reactive, RCOo being a better trleaving group" than t{O. When 2 oxygen functionst are present (as illustrated), either (in turn) can act as an alkylating centre. Such bifunctional aLkylation could lead to cross l inking of macro'molecul.es (l,fattocks, 1969r. tr{hite & Mattocks, -68-

5d-:*;dd": ." + ROH

Fig. 7. Mechanism of reaction of a pvrrolic alcohol or ester with a nucleophil (Nu).

1972; Petry et aL., 1984, 1986). tlhen the groups (R) are the same or similar, C7 is the more reactive site. Examples of such alkylations using pure chemicaLs (amines or alcohol) have been given by Mattocks (1969) and Culvenor et al. (1970a). Mattocks & Bird (1983) showed that a variety of nucleophiles of bioLogical interest could be alky1-ated by dehydroretronecine. Black & Jago (1970) demonstrated the in vitro alkylation of DNA by dJhydroheliotridine, and Roberts6i-@ and Wickramanayake et al.. (1985) the alkylation of deoxyguanosine by dehydroretronecine. The alkylation of mouse or rat liver DNA by pyrrolizidine alkaloids has been shown in vivo by Eastman er ;1; (1982) and candrian et al,. (1985)

5.L.4 PossibLe further metabolites

The possibility that pyrrolic metabolites of PAs night themselves be metabolized by microsomal enz)rmes to further cytotoxic derivatives $ras suggested by Guengerich & Mitchell (1"980). These authors showed that the tritium-Labe11ed model compounds 1r2,3-trimethylpyrrole and l-methyl-3-4 bishydroxy- methylpyrrol-e could be metabolized in rats or by rat liver rnicrosomes to unidentified derivatives abl-e to bind covatentLy to proteins and nucleic acids. It is possible that liver damage, seen in some rats given iv injections of pneumotoxic -69- pyrrolie eaters, night have been due to metabolites of the latter forned in the liver (Uattocks & Driver, 1983). Segall et al. (1985) have identified trans-4-hydroxy-2-hexena!- in an in vitro mouse Liver microsomal system metabolizing the pA ffiine and suggested that it might have been formed fron the alkaloid via a pyrrolic intermediate. The compound is capable of causing liver damage and rnight contribute to the acute hepatotoxicity of senecionine and other aLkaloids. Honever, this has not been proved, and the highly reactive and toxic prinary pyrrolic rnetaboLites from PAs are themselves capable of causing the knol,rn hepatotoxic effects of these aLkaLoids.

5.2 Toxic Actions of Pyrrolic Metbbol"ites

Pyrrolic derivatives prepared chenicalLy from PAs, as well as some analogous compounds, have been tested in experimental animals and in vitro systems, and shown to have a variety of toxic actiottsT

5.2.1 Animals

5.2.t.L PyrroLic esters (dehydro-alkaloids)

Dehydro-pyrrolizidine alkaLoids are very reactive and their effects in vivo are largely confined to the first tissues they enGGE When given oralLy to rata, they are destroyed almost inurediately in the aqueous acid of the stomach and show no toxic action. I,lhen given ip, they cause severe loca1 irritation and peritonitis (!lattocks, 1968a; Butler et al., L970),' subcutaneous injection leads to skin Lesions (Hooson & Grasso, 1976). After iv injection of pyrroles, such as dehydromonocrotal-ine (monoct'otaline pyrrole), into the tail veins of rats, the toxic injuries appear principally in the lungs. Depending on the dose, these include vascular lesions and pulmonary oedema (Plestina & Stoner, 1972); a progressive alveolar proliferation similar to that produced by very rmrch larger doses of the pareot alkaloid (Butler et al., L970) and pulmonary hypertension (nilLiker et al., 1983). Dehydromonocrotaline does not require further metabolism to expresa its pneumotoxicity, bnd it is rapidly rendered inactive after exposure to aqueous media (Bruner et al., 1986). Sirnilar pneumotoxicity is produced by totalLy synthetic pyrrolic esters having a simpler structure but the same type of chemical reactivity as the aLkaloid derivatives (Mattocks & Driver, 1983), thus confirming the chemicaL mechanism of this action. Injections of dehydro-pyrrolizidine alkal.oids or synthetic analogues into mesenteric veins of rats lead to l-iver damage -70- after smaLler doses than the alkaloids themseLves (Butler et al., 1970; Shumaker et al., L976). The Liver danage differs somewhat from the alkaloid damage, consistent with the toxin being introduced via the hepatic vascular system rather thao being produced within the hepatocytes, as is the caee nith the alkaLoids. Nevertheless, the progressive liver lesions are very similar to those produced by PAs (Butler et a1., 1970). The Lung damage after tail vein injections bears a cloeer resemblance to pyrrol.izidine danage, since the latter is also believed to be caused by netabolites entering the lungs via the bloodstream (Barnes et al. , 1964).

5.2.L.2 Pyrrolic alcohols (dehydro-necines)

Dehydroheliotridine (Fig. 6), a secondary pyrrolic metabo- Lite from heliotridine-based PAs, such as heliotrine aod lasiocarpine, is less acutely toxic than its parent alkaloids; it has an LD56 (Z aays) of about 250 urg/kg body weight in mice (Percy & Pierce, L97L). Its effects on 14-day-o1d rats were studied by Peterson et aL. G972). A11 rats given ip doses of 0.4 mrnol/kg body veight survived, but a dose of 0.6 mo1/kg kiL1ed most animals !'rithin 10 days. Toxic effects r,rere mainLy found in rapidly deveLoping tissues. In young rats, it caused fur 1oss, tooth defects, and atrophy of hair folLicles, gut mucosa, spleen, thymus, testis, and bone narrorr. The Lungs vere not affected. Pathological effects in the Liver were confined to necrosis of isolated cell"s and antimitotic action, wtrich was manifested as a rniLd megalocytosis (development of giant hepatocytes) in rats surviving 4 weeks or more. The persistent antimitotic action of dehydro- heLiotridine and of its parent aLkaloid lasiocarpine in the tiver of rats was investigated by Samuel & Jago (1975), who Located the mitotic block as being either late in the DNA synthetic (g) phase or early in the post synthet.ic (G2) phase of the cell cycle. Dehydroheliotridine is aLso carcinogenic. Peterson et al. (1983) showed that rats given 9 ip injectioos of this compound (50 - 76.5 ng/kg body weight) over 23 weeks had a shorter life span and suffered a significantly higher incidence of tumours than control rats. The authors concluded that dehydroheliotridine is responsibLe for some, or possibly a1.1, of the carcinogenicity of its parent alkaloids. Dehydroheliotridine was found to be teratogenic when given ip to female hooded rats on the 14th day of pregnancy. A dose of 4O mg/kg body weight produced effects siniLar to those produced by the alkaloid heLiotrine at a dose of 200 ng/kg (Peterson & Jago, 1980). For the immunosuppressant activity of this compound, see section 6.4.10. -7L-

The toxic actions of dehydroretronecine (DHR) (Fig. 7) when given sc to rats are similar to those of dehydroheliotridine (Itsu et al., t973; Shumaker et aL., L976). RePeated large doses aLso caused ulceration of the glandular stom:rch. Daily sc doses (4 ng/kg body weight), adminlstered to rats for 1 week, caused lung damage leading to right ventricular hypertrophy (IluxtabLe et al., 1978). DIIR was carcinogenic when applied repeatedly to mouse skin (Johnson et al., L978; Mattocks & Cabral, L982).

5.2.2 CeLl cuLtures

Dehydroheliotridine and dehydrosupinidine both have an i.nhibitory action in cultures of KB cells (human epidernoid carcinoma of the nasopharynx) wittr ED50 concentrations of -4 -5 10 mol and 10 mol, respectively (Culvernor et a1. , 1969). Bick & Culvenor (1971) found dehydroheliotridine (lun) to be considerably more effective than the alkaloid heliotrine in suppressing ce11 division and causing chromosome breaks, in cultures of leukocytes from the marsupial Potorus tridactylus; at a concentration of 6 x Lo-s mol, the mitotic index was zero, and more than half the cel1s had disintegrated. In a study by Mattocks & Legg ( 1980) , dehydroretronecine and several synthetic analogues conpletely inhibited celL division in a cultured rat liver cel1 1-ine at a concentration of 10-a mol. Ord et al., (L985) found that DIIR induced sister chromatid exchange in human Lynphocytes without the need for metaboLic activation. Analogous pyrrol.es with only one functional (reactive) group $rere much less dffective. DllR was also weakly active in induciog mutations in the SalmonelLa typhirnuriurn base substitution strain, TAg2, and gav-lGfffi results in an in vitro cel.l transformation test using a culture derived F-mster kidney celLs (styles et al., 1980) . The toxicity of the pyrrolic ester, dehydrononocrotaline, for cuLtures of mouse fibrobl-asts was studied in vitro bv Johnson (1981). The level of exposure was appro*ffiGTi-i ng per ce1L. Cel1 death was preceded, first by the swelling and disruption of organelles, incl.uding mitochondria, and then by the rupture of plasma membranes with the release of celL components. Bick et al., (1975) investigated whether the effects of PAs on leukocyte cuLtures of Potorus tridactylus ri/ere due to pyrrolicmetaiolites.Levels-i-F-d-irryd:ffi .zines,which could be demonstrated in the culture media, were insufficient to account for the observed effects of heliotrine, lasiocarpine, and monocrotaline on the cells, but the actual- -72-

amounts formed within the cells may have been higher than those observed.

5.2.3 Possible participation of membrane lipid peroxidation

Distinct increases in NADPU- and ascorbate-dependent peroxidation of microsomal membrane f.ipids were found in rats given heliotrine subcuraneously (300 ng/kg body weight) (Savin 1983). The primary biochenicaL interactions and cellular mbcromoLecular targets for the pathogenesis of PA-induced toxicity remain unidentif ied.

5.3 Chenical and Metabolic Factors Affecting Toxicity

The toxicity of an alkaloid depends on the extent to which it is converted into active metabolites and on the disposition and reactivity of these metabol"ites, once formed.

5.3.1 StructuraL features -of a toxic aLkaLoid

The essential structural features of a hepatotoxic pA (Fig. 8) are:

(a) a l-hydroxynethylpyrrolizidine ring system unsaturated in the 112-position, with preferably a second hydroxyl group in the 7-position;

(b) esterification of at least one of the hydroxyls, though toxicity is much greater when both hydroxyls are esterified; and

(c) ester groups that are resistant to enzymic hydrolysis, wtrich usually means that there is a high degree of chain branching in the acid moiety.

The above requirements apply to natural pAs but, strictLy speaking, onLy the right hand (pyrroline) ring is essential, being the ring that is metaboLized to a pyrroLe derivative. Thus, esters of 2,3-bis-hydroxynethyl-L-rnethyl-3-pyrroline (synthanecine A) (fig. 9) have pyrrolizidine-like heparo- toxicity (Mattocks, L97la; Driver & Mattocks, 1984). StructuraL requirements for N-oxides are the same as those for the hepatotoxic alkaloids. lloriever, it is. important to note that a PA N-oxide is not hepatotoxic itself; toxicity depends on it Teing reduced to the corresponding basic alkaloid, chiefly in rhe gur (Mattocks 1971c), bur possibly in other organs, such as the liver (powis et aL., 1979). -73-

o tl whereX = RCO or HO orH

Fig.8. Essential structural featurea of a hepatotoxic pyrrolizidine alkaloid.

HO CH,OH lr*fr. t) --N--,/ CHg

Fig. 9. Structure of synthanecine A.

5.3.2 Activation and detoxication

Factors affecting the proportion of aq ingested alkaLoid that is converted into toxic netabolites irl an animal include the following:

(a) Lipid solubility

Highly vater-soluble alkaloids (such as indicine) are easiLy excreted and have low toxicity. A,Lkaloids that ,ere more lipophilic are mof,e open to acdivation by liver microsomes (Mattocks, 1981a). -74-

(b) Subceptibility to hydrol.ysis

This is detersrined by the molecular structure and conformation of the alkaloid (Mattocks, 198l.arb). If the alkaloid is open to esterase attack, it may be Largely detoxified by hydrolysis.

The relative amounts of an alkaloid converted by hepatic microsomal enzymes to N-oxide and to pyrrolic metabolites depends on its molecular- structure and conformation (Mattocks & Bird, 1983). N-oxidation represents a detoxication pathway (Mattocks, L972b)-

5.3.3 Factors affecting the toxicity of active metabolites

5.3.3.1 Reactivity of the metabolite

Toxic metabolites are formed in Liver cells. Primary pyrrolic metabolites (dehydro-alkaloids) are very reactive and, thus, are quickly hydrolysed or deactivated by reaction with celL constituents. To damage tissues other than the ceLls in which they are formed, active metabolites must cross the ceLL membrane and survive whiLe being transported in the bloodstrearn. The more stable pyrrolic metabolites, such as dehydromonocrotaLine from the alkaLoids monocrotaline, are able to reach, and become bound to, Lung tissue (Mattocks, L973) . Thus, monocrotaline frequently danages the lungs, whereas retrorsine, which yields a more reactive pyrroLic metabolite, normalLy does not. Secondary metabolites (pyrrolic alcohols, e.g., dehydroretronecine), formed by the hydrolysis of primary ?yrrolic metabolites, are water soluble, relatively stable compounds that can become nore widely distributed throughout the body or excreted; these are not acutely toxic.

5.3.3.2 The number of reactive groups

The toxicity of a pyrroLic alkylating agent is affected by the number of reactive ester or hydroxyl groups (f or 2) present as the following examples show:

(a) Many pyrrolic esters can cause acute lung damage when given iv to rats, but only bifunctional ones aLso cause delayed effects on the lungs (Mattocks & Driver, 1983). -75'

(b) Bifunctional pyrrolic alcohols are more effective irririUitort of nitosis in cultured cells than monofunctional pyrrotes (Mattocks & Legg, 1980).

(c) Bifunctional pyrrolic alcohols are much better inducers of sister chronatid exchange ($cE) in lynphocytes than monoalcohols (Ord et al., 1985).

Reasons for these differences night be that the bifunctional pyrroles are able to crosslink macromolecules or sinply that they can bind more strongly to target molecules'

5.4 Metabolites Associated with the Biological Actions

5.4.1 Acute hePatotoxicity

The fo!.lowing is good evidence that acute liver necrosis is caused by prinary pyrrolic ester metaboLites ( dehydro-alkal-oids ) :

(a) The liver, in which these metabolltes are formed, is the only organ exposed to them in relatively high concentrations.

(b) There are good correlations between amounts of pyrroles bound to liver tissue and acute hePatotoxicity (Mattocks, L973).

(d) Pyrrolic esters injected iv into the liver are much more atutely hepatotoxic than the parent alkaloids (Butler et al. , 1-970).

It is possible that other metabolites; such as 4-hydroxy 2,3-unsaturated aldehydes, might aLso cootf,ibute to the acute hlpatotoxicity of some PAs (SegaLL et a1', L985)' Hovever' this has stil1 to be confirmed.

5.4,2 Chronic hepatotoxicity

The persistent antimitotic action on the liver that leads to the fornation of giant hePatocytes can be produced both by (Hsu pyrro!.ic ester metabolit"", such as dehydromono-crotaline it al.. 1973). and bv pvrroLic alcoholsr such as dehydroheliotridine (Peterson eC al., 1972). Both kinds of metabolites can lead to similar alkylation Products and both are -75-

likely to be present in the liver when the aLkaloids are netabolized. Thus, either could be responsible for chronic hepatotoxic effects. llo$ever, the antimitotic action alone is not sufficient. It mrrst be accompanied or followed by a stimulus of ceLl division. This nay be provided by-or the acute necrotic effect of prinary pyrrolic melabolites by any other cauae of acute liver injury that leads to ii""rr. regeneration. In very young animal-s, the stimulus can be the erfianced rate of replication that already exists in them.

5.4.3 Pneumotoxicity

Characteristic pyrrolizidine Lung damage is produced by iv injections of pyrroLic ester metaboLites, which are effective at much lower doses than the parent alkal.oids. The Latter are not metaboLized in lung tissue; thus, lung darnage from pAs is believed to be due to pyrrolic esters reaching ihe lungs from the liver (ButLer et aL., L}TO). Chronic lung damage Jpp..." to be caused by bifunctional rather than by monotunctional qyrrolic alkylating agents (Mattocks & Driver, L9g3) (sectioa 5 .3.3.2) . There is some evidence that pyrrolic alcohoL metabolites rnight al-so be able to contribute to chronic (but not acute) pneumotoxicity (Huxtable et al-., L978).

5.4.4 Toxicity in other tissues

Chronic heart damage including right ventricular hypertrophy is a consequence of pyrrolizidine lung damage (pulmonary hypertension) (Hayashi et al-. , 1967). Brain damage is attributed to armonia intoxication secondary to severe pyrrolizidine liver injury (tlooper, L972). This view has been contested and some PAs are knovrn to have direct effects on the central nervous system (section 6.4.3). There is no evidence that PAs are appreciably metabolized in tissues other than the 1iver. Thus, damage to other organs is probably due to metabolites transported from the 1iver. For example, in the relativel-y unconmon cases of chronic kidney ainrage after pyrrolizidine intoxication (Hooper, 1974: Itooper & Scanian, L977,) megalocytosis in this organ suggests that pyrroLic metabolites (either ester or alcohol.) are involved. Overall, patterns of disease, as observed in extra-hepatic sites, are rrspilloverrr consistent with a effect of the pyrroles produced in the Liver (Hooper, 1978). Toxicity of an al-kaloid ieflects its rate of netabolism to a pyrroLe (Tuchweber et al. , Lg74) and so the spiLLover effect is likely to be more evident at higher doses. Studies of Culvenor et al. 1976il suggest that the PAs that are hepatotoxic for rats should also le Dneumo- toxic when administered at higher doses. In acute poisoning, -77-

the hepatotoxic effects could outrileigh the pneumotoxic effects or those on other organs, to such a degree that the latter are not manifested. Variation in expression of disease (prinarily hepatic or extra-hepatic) also depends oo the reactions of host tissues in different species of animals, in addition to the quantities of the pyrroLes (Hooper, 1978). The sensitivity of the blood vessels might explain severe interstitial pdeumonias in some animals, or severe nephroses in pigs (McGrath et aL., L975)

5.4.5 Carcinogenicity

The pyrrolic aLcohoLs dehydroretronecine and dehydroheliotridine are known carcinogens (John$on et al., L978; Peterson et aL., 1983), whereas the pyrrolic eaters dehydromonocrotaline and dehydroretrorsine are only carcinogenic in conjunction with a tumour promotor (Mattocks & Cabral, L979, L982). This suggests that the more persistent secondary metabol-ites (pyrrolic alcohols) night accolrnt for the rather weak carcinogenicity of some PAs.

5.4.6 Antitumour activity

Some PAs and their N-oxides are active as tumour inhibitors in test system-s (Culvenor, 1968; Suffness & Cordel1, L985). Indicine N-oxide, in particular, showed high activity against 816 meTanoma, su$mary r xenograft, M5076 sarcona, P388 leukaemia, and Walker 256 carcinoma. In clinical studies, indicine N-oxide has shom significant activity against some forms of leukaemia, with dosage linited roainly by nyelosuppression and sometimes by hepatotoxicity. It is tempting to suppose that this action is reLated to the polverfuL antinitotic action of their pyrrolic metabolites, even though some of these alkaloids and derived pyrroles are themselves carcinogenic. On the other hand, there is evidence suggestitrg that indicine N-oxide owes it activity to a property of the compound itEelf rather than to the pyrrolic metabolites, which could be formed through reduction to indicine (Powis et al. , L979). The evidence, that indicine is less effective than indicine N-oxide. is not conclusive and other structure-activity data-(Uiltowsky, L985) point to a need for a structural capability to form a pyrrolie metaboLite. It is also possible that irrdicine N-oxide is converted directly to dehydroindicine by nitochondrGl enzynes in Liver or tumour cells, since the type of reaction required has been observed in the nitochondrial metaboLism of the N-oxides of tryptamine alkaLoids and certairn nethyLated amino acids (Fish et al., 1956; Srnith et al., 1962). -78-

5.5 Prevention and Treatment of Pyrrolizidine Poisoning

There is no known vay to prevent pyrroLizidine liver damage, once a hepatotoxic dose of the alkal.oid has been ingested. A number of dietary regimes have been found to partially protect animaLs (chiefly rodents) from the acute effects of subsequent alkaLoids ingestion. None of these are of any practical use for preventing pyrroLizidine intoxication in Livestock. Furtheroore, chronic toxic effects in the liver or in other organs are sometimes more severe in animals receiving higher doses of alkaloids after being protected against acute hepatotoxicty.

5 .5 . 1 Modif ied diet.s

The mechanism of action of modified diets is not clear, but they may be associated with the decreased metabolic activation of the alkaloids. Sone examples foLlow:

(a) A protein-rich diet can give some protection to rats against Senecio jacobaea alkaloids (Cheeke & Gorman, 1974). 6ts--?a?:;TEE-asein dier survived longer than rats given a'norrnal diet, when poisoned with retrorside or riddelline, but the survivors lrere more liable to develop liver tumours (Schoental & Head, L957). However, vhether this ltas sinply due to a prolongation of Life of the animaLs by the diet is open to question.

(b) Male rats previousLy fed a sucrose-only diet fot 4 days were considerably protected against the acute hepatotoxicity of retrorsine (lnt6 l2O ng/ke body weight compared with 34 ng/kg in normaL rats). Ilovever, lung damage, rare in control rats, rdas frequently seen in rrprotectedrr rats given high doses of retrorsine (Mattocks, t973) .

(c) Restriction of feed intake to 4O"l of norm€rl attenuated the increase in Lung weight and lavage protein concentration in ceLl-free bronchopulmonary lavage fluid and aboLished the right ventricular hypertrophy in mono- crotaLine-treated rats. Furthermorer the percentage of diet-restricted animals that survived was significantly higher than that in aninals that had eaten ad libitun up to day 28, but, from this time onwards, there was no difference. Alterations of dietary sodium intake alone did not affect the results of monocrotalineinduced toxicity (Ganey et al., 1985). -79-

5.5.2 Pretreatment to enhance the detoxication of active

Treatments that have afforded some protection against pyrrolizidine hepatotoxicity (probably by increasing the liver Level of suLfydryl compounds, which are known to react rtith pyrrolic netabolites) (white, 1976) include the following:

(a) Pre-treatment of rats with mercaptoethylamine (L50 rng/kg body weight ip) partially protectdd rats against the acute hepatotoxicity of monocrotaline Siven 15 min later (ltayashi & Lalich, 1968); it gave no protection when administered 2 h after the alkaloid. Mercaptoethylamine, when given orally (300 mg/kg body weight) at the same time as the lasiocarpine, also increased ther resistance of rats to the alkaloid (Rogers & Newberne, 1971).

(b) Cysteine (12 in the diet) partially protected rats against Senecio aLkaloids (Buckmaster et al-., L977) and mice against monocrotaline (Miranda et a1., 198lc ) .

(c) The antioxidant ethoxyquin fed at a 1evel of 2.5 glkg diet to female mice for 38 days, increased the liver thiol concentration and raised the acute LD

(d) Rats or mice also had increased resistance to acute pyrrolizidine hepatotoxicity when fed the antioxidant butylated hydroxyanisoLe (BHA) (,rp to 7.5 elke diet) (Miranda et al., 1981c, 1982a,b; Kim & Jones, L982).

(e) Hel-iotrine-induced toxicity can be modified by the co-administration of cupir (a copper-containing complex) at a level, of 1 rng/kg per day for 20 days. It prevented the exit of hepatic cytosolic enz5rmes into the bLood and improved all the energy reactions studied in the mitochondria of heliotrine-intoxicated rats (Yuldasheva & Sultanova, 1983). Inhibition of lipid peroxidation by cytoplasrnic copper was shown later (Wittig & Stephen, 1964). Savin (1983) found that lethality to rats of heliotrine (300 nglke sc) r"as completely prevented by co-administration of o-tocopherol (6 mL/kg ip).

(f) Rats pre-treated with ip doses of zinc chloride (72 wol/kg body weight) had increased resistance to the hepatotoxicity of Senecio jacobaea alkdloids, as assessed -80-

by histology and enz)rme measurements (Miranda et al., 1982c). The zinc treatment increased the liver level of metallothionein, a sulfhydryl-rich protein that might react lrith pyrrol"ic metabolites.

Metabolic inhibitors of the microsomal p450 mixed-function oxidase systemr SKF 525A, metyrapone, and alLylisopropyl acetamide, which inhibit the formation of toxic pyrroles in the liver, have been tried successfully in the prevention of the toxic effects of monocrotaline in rats (Eisenstein & Huxtable, L979). The use of P450 inhibitors was stated to rrpotentiaL show therapeutic promisett. llovever, this would seem impracticable considering that, at Least in the rat, PAs undergo a high rate of metabolism coumencing a fen minutes after ingestion (Mattocks, L972b). In some instances, they have been known to lead to an increase in toxicity, e.g., with Lasiocarpine as reported by Tuchweber et al. (1974).

5.5.3 Other treatments

Lanigan & Whittem (1970) attempted, unsuccessfulLy, to protect sheep against lleliotropium europaeum poisoning by treating them rrith cobalt, in the hope that this would enhance the vitamin 812-mediated detoxication of the alkaLoids in the rumen (oick et aL., 1963). Lanigan et aL. (197e) found that the resistance of sheep to dietary lleliotropium europaeum was increased by giving thern Large daily doses of the antimethanogenic drug, iodoform. Ilolrever, Swick et al., (1983) found that Senecio jacobaea alkaloids were not detoxified by incubation for ZB h -IEh sheep rumeo fLuid in vitro. -81 -

6. EFFBCTSON ANIMALS

6.1 Patterns of Disease Caused bv Different Plant Geaera

The most important genera of PA-containing plants list'ed insection3.1area11hepatotoxic.Amongthese'@ spp. cause damage in the broadest range of tissues in most domestic species. In pigs, tbey are known to be severely nephrotoxic (Peckham et al., L974; McGrath et al., 1975; Hooper & Scanlan, L977). Some species are known to be pneumotoxic for horses (Watt & Breyer-Brandwijk, 1962; eardiner et. al., 1965), cattLe (Sanders et al-., L936; Berry & Bras, 1957), sheep (Laws, 1968), and pigs (Peckharn et a1.., L974; Ilooper & Scanlan, L97'1), as welL as hepatoEoxic. Al-though several Crotalaria spp. are known to be pneumotoxic for horses (caEi;,e-?-?fAT;, 1965) , c. rerusa is an exception. It is an important cause of diseiffi-*-rses in northern AustraLia (Hooper, 1978) and has been shown to be pneumotoxic for pigs in the same area (llooper & Scanlan, L977)i yet it produces only hepatic disease lin horses (Rose et al., 1957a,b). Sirnilarly, Seoecio spp. are primaril-y tlepatotoxic, but S. jacobaea has b6-?em6iSirated to be pneumotoxic for pigs (ttarding et al., L964), though it could probably be an inconCistent change (8u11, et al., 1968). This plant is also known to cause pulmonary disease in rats and mice (Hooper, 1974). Ilowever, there are no reports of its affecting the lungs in cattle, sheep, horses, or chicken. Renat megalocytosis and miLd nephrosis are reported in most species poisoned with S. jacobaea (Harding et al-., L964; Bu1l et al., 1963). Heliotraifn-spo- Amsinckia spo.. and Echium sDD. are all mainly hepatot,oxt c. Roitnan (1983) surmrarized the pattern of organ involvement observed in man and different species of farm and experimentaL anirnals affected by pyrrolizidine alkaloids (Tab1e 8). Even within a singl"e species, the nature of a toxic effect, as well as the organ affected, can be altered .by changing the dose rate and duration.

6.2 Fiel-d Observations - Outbreaks in Farm Anirnals

The veterinary probLem of PA toxicity has been reviewed by Bu1l et al.. (1968) and Mclean (1970). Mattocks (1986) listed the cases of Livestock poisoning and feeding trials sinee 1968, and cited relevant l.iterature. Peterson & Culvenor (1983) produced a useful and comprehensive table of the plant species known or suspected of causing natural outbreaks of -82-

Table 8. Anioal species and organs affected by pyrrolizidine alkaloids3

Spec ies Liver Lung Kidney Heart Pancreas Gastric Muscle mucosa

Man + ltonkey + + + Ilor s e + + + Pig + Sheep + + Goat + Cat t 1e + Dog + Itouse Rat Chicken + Turkey + +

3 From: Roitmn (l-983). poisoning in each aninal species. The influence of factors such as speciesr 4g€r sex, and diet, on toxicity is also reviewed in the same paper. The first cases of pyrrolizidine poisoning rirere described in cattLe as earl-y as L903 (Gilruth, 1903). Since then, there have been numerous reports from most parts of the world, of poisoning among farm animals caused by grazing or feeding on PA-containing weeds (8u11 et al., 1968; Mattocks, 1986). One of the first cLues to the etiology of the human disease in Jamaica, came from a study in which calves fed with CrotaL- aria fulva (Bras et aL., 1957) deveLoped characteristiJ:i6-o- occlusive disease in the liver. Lar,rs (1968) described a field outbreak in sheep in a herd of 100 adult merino enes, which developed lrithin 2 weeks of moving into a coastal- farm in Australia, where they grazed CrotaLaria mucrocata. The etiology was confirned by feeding @, 4 of which died wirhin 24 h of feeding, r^rith severe puknonary oedema. Ilolrever, the rapidity of poisoning and the atypical lung lesions suggest that possibly a toxin other than pyrroLizidine alklaoids was also present. An outbreak of Crotalaria retusa poisoning was observed in a piggery near DarwG]lGf?fr--Effiitory, Australia (Ilooper & Scanlan, 1977) containing approximately 350 sows. It rras caused by feeding sorghum contaminated by CrotaLaria seeds at the rate of about 0.12 by weight for about 3 weeks, and at a rate of about 0.057" for a further !ileek. The disease \das indicated by reduced body weight gain and inappetence. The dominant pathol-ogical features at autopsy nere severe nephrosis with chronic uraemia, and to a lesser degree, severe -83- diffuse interstitial pneumonia. Both wete accompanied by nicroscopic disease in the liver, and both the liver and kidney shor^Tedrnegalocytosis. Walker & Kirkland (1981) reported outbrdaks in the Hunter river valley of New South Wales ia Australia, in cattle that had been grazing a pasture in which Senecio lautus \ras growing. There were sforadic deaths amongffi? welL as tlro protracted outbreaks affecting calves 3.5 months of age and older animaLs, in which groups of 3 - 16 head of cattle died in addition to sporadic deaths of animals over periods of 1 and 6 months. Clinical signs were weakness, emaciation with recumbency, airnLess wandering and ataxia, which suggested neurological involvement. At autopsy, the Liver showed characteristic rnegal-ocytosis, periportal fibrosis, and focaL necroses. An aged animal had cirrhosis. The etiology of S. lautus was proved in feeding studies on 3 calves Knight et a1. (1984) reported the dedths of 10 horses during a 3-year period after being fed tray from the same pasture. The animals became sick in the spring after being fed only the suspect hay throughout the wirrter. The hay was found to be contaminated with CynogLossum officinaLe (houndts tongue), which contained two e^0,@hinatine) in much higher quantities than is generaLLy reported in Senecio species. The animals developed weight loss, icterus, ataxia, and sfmptoms of hepatic failure. At autopsy, there $ras diffuse severe megalocytosis, biliary hyperplasia, and fibrosis. The C. qfficinal-e was proved as the etiological factor in a fEetTl.....ffaf on a l4-year-old pony, that devetoped cLinical features and patho!.ogica1 changes in the f.iver suggesting PA poisooing. The signs of PA toxicity in horses are mostLy neurological, though non-specific gastric and oesophageal Lesi'ons have aLso been reported (Mclean, 1970) (section 6.4.3). Rose et aL. (1957a,b) described a disease in which the dominant symptom rras ttcompulsivett walking in a straight Liae. It occurred in areas where CrotaLaria retusa was growing, and was ascribed to a steep riseffi??iEffiiE levels, which accompanied chronic liver failure. Farm aninaLs differ widely in their sensitivity to pAs, sheep and goats being fairLy resistant, cattle and horses, less sor and poultry and pigs, rather $ensitive (section 6.4.L.2). Sheep are not imediately affe-cted and generally survive one season, after feeding on heliotrope and Senecio (Bull & Dick, 1959). During rhe second season of f66arilE; they die of neurological symptoms caused presumabLy by rising blood-armonia Level-s associated with chronic liver disease, or with haenoglobinuria and very high copper l"evels in the blood (BuLl" et al., 1956, 1958) (seetion 6.4.1.2). with crotalaria, the1ungseemstobethetargetorgan(Hooper,]-978)ffi -84_

acute responses to a single feed of the plant Crotalaria spectabilie were described in cattle by Enrnel (1946)-A-ifG tJre-cTfrEE; by Piercy & Rusoff (1946). Poisoning of cattle in northwestern USA has reached such proportions that it has become a considerable economic problem (Johnson, 1982). Culvenor (1985) has reviewed the problem of livestock losses due to PA toxicosis in Australia, where it has been estimated that about 10 million sheep are exposed to heliottope and Echium pLantagineum (Patersonrs curse) to a greater or lesser extent and may suffer a shortening of their productive life by as much as two years. Most of the PA-containing plants are reported to grow in faLlow fields and pastures and thrive particularly in a dry clinate or following periods of drought. I{owever, instances of cattle poisoning have been reported from most parts of the world, incLuding countries with temperate or cold cl-imates, which do not ordinariLy suffer drought. Three herds of cattle have been reported to have been affected in the alpine/subalpine region of Switzerland, after they grazed pastures that had Senecio alpinus growing on thern. Nine different hepatoto*ic PAffi found in the weed, the main o:re being seneciphylline. Analyses of urine samples from one con confirmed the presence of PA metabolites. Several corrs had to be slaughtered, because of cirrhosis of the liver (Luthy et al., 1981).

6.3 Studies on Farm Animals

There are several reports of the production of disease characteristic of PA toxicity in farm animals, by feeding them PA-containing plants. Senecio jacobaea (tansy ragnort) is a weed that corrnonly grows in pastures and has been the cause of extensive livestock losses in the United Kingdorn (Forsyth, 1968) and the USA (Johnson, 1,982). Extensive studies on this pLant have been carried out using a variety of farm animals. Dickinson et a1. (1976) fed tansy ragr^rort to cows through a rumen canula at the rate of L0 rng/kg body weight per day, for 2 weeks. Liver biopsies showed characteristic megalocytosis and fibroplasia, and autopsy al-so showed centrilobuLar necrosis. A PA, jacoline, was found in the mi1k, but when the calves were bucket fed the rnilk, there lrere no detectable effects on them. Thorpe & Ford (1968) made similar observations in 5 calves fed ragwort in their diet. Animals eventually dying of toxicity showed characteristic necrosis, negalocytosis, and veno-occlusive lesions. In a study by Goeger et al. (1982), goats were fed dried ragr^rort nixed in the diet at 250 g/kg. Four of the 11 goat kids end lactating dairy goats died. Characteristic megalocytosis was seen in the liver. Goats are more resistant to tansy ragwort toxicosis than cattle and _85_ horses, the chronic lethal dose for cattLe or horses being 0.05 - 0.2 kg ragwort/kg body weight and, for goats, t.25 - 4.04 kglkg body weight. The aLkaloid levels in the pLant, and thus its toxicity, varies with season and locality. Hooper & Scanlan (L977) studied the loog-tern effects of feeding very lov levels of ground C. retusa seeds, nixed with the feed, to pigs and chickens. Seven groups of 4 pigs each (sex not nentioned) bred fron SaddLeback-large vhite cross sows and Large White or Landrace boars, were maintained on diets containing 0 (control), 0.004%, 0.01%, O.O2"1",0.052, O.l%, ot 0.52 body weight ground seeds. Another 8 pigs were fed a diet containing 0.12 C. retusa fot 2L days followed by O.O5% fot 7 days and then kept on a C. retusa-free diet. Pigs either died or were killed when .oribliffiF:at the end of 1.36 days of feeding. In a second study, groups each of 4 2*week-old chickens were fed diets containing 0 (control), 0.0052, 0.012, O.O5%, 0.12, and 0.5% ground seeds of C. retusa. Chickens fed 0.57" started dying 12 days after tnea6ffiEEemeot of feeding and were all dead by day 45. Five out of 8 birds fed 0.1% or more died between days 22 arl'd 56. No deaths occurred in animaLs fed 0.012 - 0.0052. Ia the study on pigs, all animaLs died between days 63 and 107 except for 2 that survived 136 days. In these animals, pulmonary disease lras the main cause of death. llepatic and renal megalocytosis lras seen in alnost all atrimals in both the field outbreak and study group. The Lungs showed extensive consolidation and oede-a. Besides megalocytosis in the gLomeruLi and tubuLes, the kidneys showed glomerul"ar atrophy and tubular necrosis. In the study on poultry, the major disease was hepatic necrosis of irregular distribution. The kidneys showed miLd negalocytosis. In the above study, the low levels of contamination that produced serious disease are worthy of note. Johnson & Molyneux (1984) fed 55 cattle, by gastric lavage, with hay nixed with threadleaf groundseL (Senecio douglasii var. longilobus), which grolrs comonly it the pastures of south!f,estern USA. The PA dosage in different groups ranged from 5 to 40 rng/kg, daiLy, and the total intake ranged frorn 80 to 284 rng/kg body weight. The groups were fed for periods ranging from 2 to 20 days. One hundred percent mortaLity occurred in 3 out of 9 groups, each consisting of 2 - 8 caLves, receiving doses of 13 ng/kg or more. Mean survivaL time was generalLy inverseLy proportional to the dosage received. AlL sick calves had typical clinicaL signs of seneciosis. At autopsy, the principal ldsion was seen in the liver and consisted of swell-ing of the hepatocytes, necrosis, bi1-iary hyperplasia, and marked fibrosis. The estinated minimrm lethat dose of the PA was 13 rng/kg body _86_ weight for 15 days, or a total intake of approximately 200 mg PAlkg, over a 15-day period. Cattle that consumed up to 600 ng PA/kB, in hay, in a 20- to 100-day period, were unaffected or only slightly affected. The authors concluded that the time-dose reLationship for PA toxicosis in cattLe is important and that there is a threshoLd level that must be exceeded for the toxicosis to devel-op. A similar study was conducted by Johnson et aL. (19Q5) in which the dry whole or ground leaves of @!13!1!1!1i!!!, nixed with the feed, were fed to calves in geLatin capsules or by gavage. Forty-trro female Ilereford calves were divided into 3 groups. One group of 12 was fed the leaves rnixed with alfalfa hay feed estinated to have 20 - 40 ng PA/kg body weight per day over a 20-day period in different regimes, receiving a total of 400 - 800 mg PAs per animal. The second group of 12 animals received the plant packed into gelatin capsules, receiving an estimated PA content of 10 - 2O mglkg body weight, daiLy, over 20 days, with a total of 200 - 400 arg per animal. The third group of 18 animals was administered (by gavage) finely ground leaves in a water slurry at various PA dosages ranging frorn 10 to 60 ng/kg body weight per day and a total of 200 - 500 rng over 20 days. Calves that received 10 rng PA/kg body treight per day for 20 days did not develop clinical signs of disease or show any changes in serureozyme. However, feed containing the plant that provided 15 - 20 rng PA/kg per day or morer administered by gavage or fed in capsul-es, resulted in high mortality. Ma1-aise, depression, erratic or unprecedented behaviour, aimless waLking, and ataxia, nere observed in the affected calves; diarrhoea with tenesmus aod rectal collapse were frequentl-y observed. The feed intake decreased progressively and was negligible terminally. The animals that died and those that were moribund or in a state of irreversible wasting, were autopsied. Hepatobiliary lesions were present in aLl such animals. The most consist.ent change was portal biliary hyperplasia and periportal fibrosis. Centril,obular or zonal haemorrhage and necrosis $rere observed in some lobules. Fibrosis of some central veins was cormon, often encroaching on the lurnen, resulting in compLete occlusion. Hepatocytes also showed nonspecific changes. CentraL nervous system changes were present in all animals with clinical signs of seneciosis, consisting mainly of spoogy degeneration of the brain. The plant mixed in the hay ration was eaten s1owLy and reLuctantLy and was tolerated at dosages > 20 mg/kg per day, ernphasizing that the toxicity depended on the rate at which the dosage vras consumed and that mortality rrras not necessarily dependent on the cumul-ative dosage. -87-

Burguera et al. (1983) produced the disease in turkey pouLts by feeding them seeds of C. spectabilis. Simultaneous addition of sodium selenite at doses of 0.L, 5, or 10 mg selenium/kg diet did not provide any protection.

6.4 Experinental- AnimaL studies

6.4.1 Effects on liver

6.4.L.1 Relative hepatotoxicity of differe[t PAs and their

The LD56 vaLues for rats, f.isted in Table 9, are for some of the most coamonLy used hepatotoxic alkaloids, caLcuLated from data on animals dying from acute haemorrhagic necrosis of t.he 1iver,3 - 7 days after intra-peritoneal administration of a singLe dose. It is evident that the toxicity varies widely between the alkaloids. The most toxic are certain nacrocyclic diesters of retrorsine and the Least toxic are the monoesters of heliotridine, retrorsine, and supinine (l{attocks, 1986). . The relative toxicity of N-oxides compared vrith that of their basic al-kaloids depends on the route of administration. The N-oxides of Lasiocarpine, monocrotaline, and fuLvine were repoited to be as toxic as their basic aLkaloids (Schoental & Magee, 1959) when adninistered oraL1y; however, when given by the ip or intravenous (iv) routes to rats, they were much less toxic (Mattocks, 197lc). Sirnilarly, the LD50 of retrorsine N-oxide when administered ip to mal-e rats r4ras 250 rng/kg fraut. p) but when given oraL1y, ir was 48 mg/kg. This has been explained by the observations on the metabolic pathvays of the basic alkal-oids and their N-oxides. The PAs or their N-oxides exert toxic effects only-after being rnetabolized to fyrrol-es by the hepatic microsomal enzymes (section 5.1.1). Itepatic microsomes act directly on the N-oxides (Mattocks & White, 1971b) only after they have been cdnverted to the basic alkal-oids (Mattocks, 1986); this mainly occurs in the gut (Mattocks, L97Lc; Powis et aL., 1979). This matter is of practical importance as the aLkaloids are often present as their N-oxides in veeds grazed by farrn anirlal-s.

6.4.I.2 Factors affecting hepatotoxicity

These factors have been reviewed by Mattocks (L986).

. (a) Route of administration

Most studies on LD5g val-ues have been carried out usiag the ip route, and very few experimental dat4 are available on -88-

Table 9. lBO" in nale rats after a single intraperitoneal dose of some hepatotoxic a lkalo ids

Alkaloid LD59 (ng/kg) Tine range Reference (days )

heliotrine 296 J Bull et af. (1958) '17 lasiocarpine Bu1l et al. ( l-958)

lasiocarpine N-oxide 547 Bull et a1. (1958)

monocrotal ine tI) 3 Bull et al. (1968)

retrorsine 34 4or7 llattocks (L972a)

retrorsine N-oxide 250 7 ilattocks (1972a)

senecionine )U 7 Mattocks (L972a)

senecipiylline 77 Bul-l et al. (1968)

senkirkine 220 Ilirono et al. (1979a)

synphytine 130 Ilirono et al. (1979a)

.3 Not stated.

toxicity using the oral route. There is a close sinilarity betveen the iv data and the ip data. Furthermore! toxicity data on rats administered PAs by the oral" route (Schoental & I*Iagee, L959), including retroraine, lasiocarpine, heliotrine, and monocrotaline, closely resembLe those relating to the LDqn vaLues for the same strain administered PAs intra- peiltoneally (Mattocks, Lg72b). Thus, ihe hepatotoxicity of PAs in rats does not differ very much, irrespective of the route of administration. However, rabbits appear to be less susceptible to PAs in the plant Senecio jacobaea nhen administere4 orally than when adnr nousl-y 'J,977). (Pierson et al. ,

(b) species

Wide differences have been observed in the hepatotoxic effecls of PAs and alkaLoid-containing pLants between different species of both farm animals and Laboratory animals, and in the same animal exposed to PAs derived from different plants. Sheep are resistant to PA-containing plants (section 6.2) and nhen fed Echium pLantagineum pellets containing 1".3 g alkaloid/kg as the-so!.e diet for 12 mooths, over a period of _89_

2 years, showed almost no liver damage (Culvenor et al., 1984). llowever, aduLt rats fed the same pellets as only 502 of the diet for 14 days died 4 - 13 weeke later (Peterson & Jago, L984). Pigs r"ere found to be 5 - 10 times as susceptible to PAs in Crotalaria retusa as chickens (Hooper & Scanlan, 1977). ov"tafi-Gtllilfriie ratios of quantities of plant material required to prove toxi.c in the various species f.isted are about 200 for the sheep (approximately the same for the goat), 150 for the mouse, 50 for the rat, 14 for cattle (approxirnately the same for the horse), 5 for the chicken, and 1 for the pig (Hooper, 1978). Cheeke & Pierson-Goeger (1983) studied the chronic toxicity of Senecio jacobaea for severaL laboratory anirnals by feeding tfre EiiEE-lTiffiFa component of a mixed diet. The degree of susceptibility to PA poisoning was compared in terms of the chronic Lethal dose of the dried pLant as a percentage of the initiaL body weight among the animals themselves, and with sinilar data on livestock in other studies. Gerbils, hamsters, and guinea-pigs were resistant to chronic toxicity, gerbils being the most resistant, consumi.ng over 35 times their body weight of the dried pLant. Comparison with similar data in other studies indicated that the rabbit (Pierson et al., L977), Japanese quail (Bucknaster et at.., 1977), and goat (Goeger et al., 1982) rdere resistant, requiring a long-term lethal dose of the plant of LL3Z or more df the initial body weight, whereas the rat was highly sensitive requiring only 2L7" (Goeger et a1., 1983). Chicks and turkey poults were also susceptible with severe inhibition of grotrth occurring when there $ras 57" and 10% contarnination of the feed with the pLant; survival time was short (Cheeke & Pierson-Goeger, L983). In a study by Fushimi et a1. (1978), on the carcinogenicity of the flower stalks of Petasites japonicus Maxim in mice and syrian golden harst@ strain differences were observed, not only with regard to hepatotoxicity, but also with regard to the carcinogenicity of PAs. Mice of ddN, Swiss, and C57BL/6 strains and Syrian golden hamsters were fed on a diet containing young flower stalks of the plant for 480 days. lligh incidences of lung adenoma and adenocarcinoma were observed in ddN mice, but no sigaificant differences in tumour incidence were obgerved between the experimental groups of Swiss mice and hamgters and the corresponding control group. No tumours were induced in an experimental group of C57BL/6 strain mice. These differences have been expLained rby the differences in the rate of metabolic conversion of PAs to toxic metabolites (pyrroles) by the hepatocyte microsones in the different animal species (Utrite et a1., 1973; Shull et aL., l-976; Peterson & Jago, 1984). -90-

The resistance of sheep has been ascribed to destruction of the alkaloids in the rumen by a reductive conversion into non-toxic - l-urethylenepyrrol.izidine derivatives (nul1 et aL., 1968; Lanigan, L97L, L972). Ir has also been suggested thai the resistance of sheep is due to a low Level of aciivation in Liver cell,s (Strul1 et aL., Lg76), but this factor rras not prominent in some Australian sheep, which were as sensitive as rats to PAs injected intraperitonealLy (Itooper, L974). Thus, it is possibLe for ruminants to Eraze plants containing PAs for a period of months nithout evident harm, e.g.r cattle eating Crotalaria juncea in Africa (Srungboonmee & Maskasame, 198L), but Long-tern effects may arise in animals exposed over several years. Considerable differences in LDcn values have been reported for the same aLkaLoids in different species. For example, the LD56 for retrorsine varies from 34 me/ke for male rats to 279 n'glkg for quail and over g00 ng/kg- body weight for guinea-pigs (White et al. , lg73). Guinea-pigs are aLso resistant to monocrotaLine (Chesney & Allen, L973a), but not to jacobine or to mixed aLkaloids of Senecio jacobag.l, which are highLy toxic (S\,rick et aL. , L982a)

Significant differences in the hepatotoxicity of the same alkaloid have been observed between sexes in some species. Male rats are much more susceptibLe to the acute toxicity of retrorsine or monocrotaline than females (Mattocks , Ig72b). Mattocks & White (1973) reported a higher level of metaboLic transfornation in young mal-e rats to form pyrrol-es from retrorsine, compared with females (section 4.4):- Jago (1971) reported a higher susceptibility in maLe rats to the chronic hepatotoxic effects of hel-iotrine, while femal-e rats were more susceptible to lasiocarpine. It is possible that this rnay be due to the potentiating effect of male sex hormones. Campbell (1957a,b) reported that diethylstilboesterol protects against the effects of seneciphylline and promotes repair of danaged Liver in poultry. Protein-deficient rats of both sexes, or female animals pre-treated with testosterone, $rere more susceptible to monocrotaline (Ratnoff & Mirick, 1949). (d) -&

Available data on the effects of age are highly confLicting. It has been stated that young rats, particularly suckling animaLs (Schoental, 1959), are more susceptible than adults to the hepatotoxic effects of some alkaloids (Jago, 1970), such as monocrotaline (Schoental & Head, 1955), and - 91 - retrorsine and Lasiocarpine (Schoental, 1959). Rata' I - 4 days o1d, rilere far more susceptibLe to retrorsirre and senkirkine than rats aged 25 - 30 days (Schoental, 1970); yet new-born rats (within t h of birth) wete relatively more resistant to the hepatotoxic effect.s of relrorsine than 1- to 4-day-oLd rats (Mattocks & White, 1973). Mclean (1970) has critically reviewed the data. In conparing the data on snraLl animals from several studies, newborn and 4-veek-old animals appear to have about the same susceptibility as adults. Data for the intervening period obtained by l{arris et al. (1957), schoental (1959), and l{ayashi & Lalich (1968) are conflicting, suggesting decreased susceptibility in some studies (Ilarris et a1., 1957 and one series of Schoentalrs studies, L959), and increased susceptibility in others (Ilayashi & Lalich, 1968 and the second series of Schoentalrs studies, L959). Furthermore, Jago (1971) demonstrated that, while rats aged L - 2 weeks \rere more susceptible to the acute effects of heliotrine and lasiocarpine than older rats, sensitivity to the effects of long-term administration of these alkaloids increased with agerafter2-3months.

(e) Diet

Effects of both qualitative and quantitative changes in diet on the hepatotoxicity of PAs have bqen investigated in severaL studies. Restriction of protein l-evels in the diet enhanced the acute hepatotoxic effects of retrorsine in rats (selzer & Parker,1951) and the chronic effects of a single dose of oraLLy administered lasiocarpine (Schoental & Magee, L957) (section 6.4.5.1) as we!.l as the toxicity of PAs in jacobaea, protein a protective +'Senecio whereas a high diet had eftect (Cheeke-& Gorman, L974). Likewise, l-ow lipotrope diet enhanced the toxic effects of orally administered Lasiocarpine in pregnant rats and also in the fetal Livers (Newberne, 1968). On the other hand, it protected young male rats against the acute toxicity of monocrotaline, becairse of the reduced metaboLic conversion of the alkaloid into pvrrol-ic metabol-ites (Newberne et al. , 197t, L974). CaLoric restrictioo reduced the cardippulmonary toxicity of a single dose of monocrotaline in rats (Hayashi et al., L967). This was ascribed to the reduced gror,rth rate in animals on a restricted diet rather than to a reduction in the rate of rnetabolic conversion of the alkaloid, since dietary restriction started only after administration of the alkaloid. When the animals srere put back on the ad Libitun- feeding regimea, they developed signs of increased fo" A high copper content in the diet has been shown to enhance the toxic effects of PAs (Miranda et a1., 1.98lb). Incorporation of copper sulfate at 5O ng/kg, in the diet -92-

containing the plant Senecio jacobaea increased the hepatotoxicity in rats, as judged by enz]rme measurements. The inplications of this observation are obvious Lt sone PA-containing pLants being grazed by farn animals also have a high copper content. Mattocks (1972b) demonstrated the lrrotective effects of sucrose against the acute hepatotoxic effects of retrorsine in rats, if adroinistered for 3 days prior to alkaLoid administration (section 5.5.1).

6.4.1.3 Acute effects

Experimental aninal studies on the pathological effects of PAs on the liver have been reviened by BuL1 et al. (1968) and Mclean (1970). Most studies have been carried out on the rar (Schoental & Magee, L957, L959; Bu1L & Dick, 1959; Schoental, 1963; Barnes et al., L964; Mcl,eao et aL., \964; Nolan et a1., 1966; Jago, 1969; Butler et al., 1970; Peterson & Jago, 1980), but severaL other species of animals have been studied incl-uding the nonkey (Wakirn et al., L946; Rose et a1., 1959; Allen & Carstens, L968, t97L; AlLen et al., 1969), turkey (Allen et aL., 1963), chicken (Allen er a1., 1960), hamgter (I{arris et a1., 1957), mouse, guinea-pig (chen et al-., 1y40), quai1, cat, rabbit, and pig (Eqmel et a1., L935i Ilooper & Scanlan, 1977). A11 animaLs tested, except the guinea-pig (Chen, t945), have been found to be susceptible in studies using purified alkaloids and their N-oxides and crude extracts of PA-containing plants. Typically, the most cormon Lesion produced in snall laboratory animals by doses cLose to the LD56 is a confluent haemorrhagic necrosis in the liver, which appears nithin about 12 h of exposure and peaks at 24 - 48 h. lt is strictly zonal in distribution in different species of animals but nay vary within the same animal, depending on the alkaloid used, species, nutritional status, and pretreatment with other chemicals. .Retrorsine produces centrilobular necrosis in the rat, mouse, and guinea-pig, periportal necrosis in the hamster, ard focal necrosis in the fowl and in the monkey (Wtrite et a1., 1973). In the monkgy, monocrotaline produces centriLobular necrosis (Allen & Carstens, 1968), but senecionine produces necrosis in the periportal and midzonaL areas of the liver 1obule (Wakim et al., 1946). Almost simuLtaneousLy, or shortly after the development of acute haemorrhagic necrosis of the liver cel1s, various levels of change appear in the central and sublobular veins of lhe Liver lobuLes, coneisting of subintimal oedema or even necrosis, deposits of fibrin, thrombosis, and occlusion of the lumen, lrhich later becomes organized. tr'Ihile haemorrhagic necrosis is a constaot feature, -93- attempts to produce occlusive Lesions in the veins of experimental animals have produced variabl-e results (A1len et aL. L967). In man and non-human primates, hepatocel-1ular necrosis and venous occlusion occur simultaoeousLy but, in the rat (Mclean et al., 1964), chicken (Allen er al., 1960), and swine (nrmeL et al., 1935), the vascular changes fol1ow hepatic necrosis. Selzer & Parker (1951) produced a lesion comparable to human veno-occlusive disease in al-bino rats by adninistrating retrorsine hydrochloride, the active alkaloid of Senecio iLicifolius, as weLl as the crude pLant extract. Four FEEhE of rats were administered aLkaloids orally in a single dose of 1 - 1.5 ng/10 g body weight or repeated doses of 5 - 50 mg/kg body weight for 31 days or as single subcutaneous injection oi L00 ng/kg body weight. One batch was fed on a diet of Senecio iLicifolius constituting LOT"of the ration as crude pia;-A G;ct; the animats livea for 2L - 84 days. sone groups were kept on a nornaL diet, and others on a diet that was protein-deficient. Aninals, administered a eingLe dose oral-Ly, developed the earliest degenerative changes in the centrilobular hepatocytes and sinusoidal dilatation, and the vascular lesioir appeared after 36 h. protein deficiency enhanced the toxic effect. Onlv 5 out of 9 animals administered repeated doses orally showed early centrilobular fibrosis and noae sholred the vascular lesion, possibly due to scarring. Bu1L et aL. (1958) studied the effects of a single ip LD76. dose of heliotrine (3ZO rrr1lke body weighr), lasiocarpine (80 nglkg), or lasiocarpine N-oxide (629 ng/kg) in rats of a hOoded strain of both s.*e": Eighty-one rats \rere used and 3 rats from each treatmeot were ki11ed at interval-s of 4 - 36 h. Iteliotrine produced narked centrilobular necrosis at 24 h, but venous changes were not evideat, except for some aggregation of mononucLear macrophages on the endotheLium. tr'fith lasiocarpine, the hepatic changes were similar, but the necrosis was not clearly centrilobular and, with its N-oxide, it was midzonal at 34 - 49 h. The earLiest toxic eFfect of the PAs was manifested as a temporary loss of mitochondria at 8 h. The pAs authors concluded that have an early lglic effect on the hepatocytes and that this does not foLlow vascular injury, as Buggested by Davidsonrs earlier studies (1e3s). Mclean et aL. (1964) administered an aqueous extract of Crotalaria fulva to Wistar rats in a single intragastric dose 6Fdffi5!/l.g body lreight. Lesions idenrical-to rhose of human veno-occlusive disease vere produced in the animals by adjusting the dose to permit survival for 8 - L2 days. Loss of cytopLasnic gLycogen in the centrilobular cel,ls occurred 3 h following adninistration. CentriLobular necrosis, which -94-

occurred after 24 h, increased with time. The central- veins gradually filled up with thickened endorhelial cells at abour 7 days, later progressing to collageni,zation. Evidence was presented that the histological occlusion of the central veins was preceded by several days by a functional blocking of the blood f1ow. Barnes et a1. (L964) observed similar results in rats administered a single oral dose of fulvine N-oxide at 50 mg/kg body weight and studied at intervals of I = 4 days after the administration. One hundred and thirty-five rats of both sexes were used. Acute lesions resembling human disease were observed during days 1 - 8. During days l-9 - 21,3 out of 25 animals showed liver damage consisting of some centrilobular haemorrhage and fibrous thickening of the central veins. Of the 78 animals studied at 22 - 44 days,50% stiLL had centrilobular congestion and some had fibrous thickening of the central veins. The effects of pyrrolic derivatives of pAs on rats were studied by Butler et al. (1970). Male albino rats of porton strain were administered solutions of pyrrole derivatives of monocrotaline and retrorsine in dimethyl formamide as a single injection of 0.05 - 0.1 ml solution. When injecred in the tail vein at a concentration of 5 mg/kg body weight, it produced progressive proliferation of alveolar epitheLiurn of the lungs and the animals developed signs of respiratory distress in 2 - 3 weeks. When injected in the mesenteric vein at a concenLration of 15 mg/kg body weight, as a rule, the animals remained well in the postoperative period and only 1/26 animals died of mesenteric vein thrombosis; the f.ivers developed rnultiple infarcts in the left l"obes that deveLoped into mulciple coarse nodules at 6 - 12 weeks. The above studies substantiated the view that PAs act only when converted in hepatocytes to pyrroles. When pyrroles vrere injected, they affected the smaller vessels at the portal of entry; in anirnals injected through the mesenteric vein, the rnain target r^7as the portal vei.n with only secondary damage to the parenchymal ce1ls, thus sparing the animals from the effects of hepatocellular injury. On the other hand, pyrroles injected through the tail vein \rent directly to the pulmonary arteries through the hearc and damaged the alveolar capillaries. Acute veno-occlusive disease r4ras produced in monkeys administered monocrotaline (A11en et a1., 1967, 1969) and ground !:r.Jglerla_ spgclabilis seed (Allen & Carstens, 1968). In a study published in 1967, these authors used 14 monkeys (Macaca speciosa) of both sexes, each weighing approximately 4 kg. Seven of the animals were administered I mg monocrotaline in distilled water by gastric intubation on days 1 and 14. The remaining 7 were used as controls and received distill_ed -95- water only. !iledge biopsies of liver were examined weekly. The survival tiare ranged frorn 14 to 38 days, the mean being 21 days. The livers of treated animals were small and firm and showed changes characteristic of human veno-occlusive disease incLuding centrilobular necrosis, and vascular changes in the centraL veins of f-iver lobules ranging frogr subintimal oedema to progressive colLagenization and extension of collagen fibres into the sinusoids. Sirnilar observations were rnade in studies on Macaca mulatta monkeys (A11en & Carstens, 1968), administeredlffiffilaria ipectabilis seed. Sixty-four animals, averiging 6me divided into 3 groups. Group I, comprising L0 experimental animaLs (4 control animals), received seeds in the diet containing the equivalent of 0.074 mg monocrotaLine/kg body weight, daiLy, up to death. Group II, consisting of L4 treated animaLs (4 controL anirnals), received a single dose of seeds containing 1.3 g monocrotaLine/kg body weight, and Group III, consisting of 26 experimeotal animals (6 controls), received 3 weekly doses containing the equivaLent of 0.116 g monocrotaline/kg body weight, by gastric intubation. Liver biopsies were carried out each month in Group I and each week in Groups II and III. Animals of the Last 2 groups were killed when in extremis. The mean survivaL tirnes for the groups r"ri cr."p I, 269 d,ays (L76 - 425 d,ays); croup II, 28 days (6 - 43 days); and Group III,4L days (23 - 91 days). In Group I animals, occlusive lesions of the central and sublobular veins of the liver r,rere seen in 7/10 animats at autopsy. These consisted of oedema, haemorrhages, and fragmentation of the vessel wal1s, the lumina being fi11ed lrith fibrin, and degenerating f-i.ver ce11s. The lobuLar pattef,n was distorted because of connective tissue bands encircl-ing smaLl- groups of liver cells, especially in the central zones of the Lobules. In Groups II and III, various 1evels of focal or centriLobular necrosis nere observed and the liver cells vtere replaced by stromal tissue. VascuLar lesions, as described above, \rere seen in 25 monkeys, but no co1-Lagenwas demonstrated. In the studies of Wakirn et aL. (1946) on the rhesus monkey, senecionine administered iv as a 2% solution at doses of 10 - 30 ng/ke body weight to 4 animals, daily, until they appeared to be sick, produced periportal, or midzonal necrosis in 3 animals accompanied by haemorrhage. No rnention was made of any vascular changes.

ELectron microscopic studies

Svoboda & Soga (1966) studied the effeets of Lasiocarpine and Crotalaria fulva on the livers of male Sprague Dawley rats weigfr@:-86-l each. one group of 22 rats was given an ip injection of lasiocarpine at 80 mg/kg body weight and pairs -96- of animals llere killed at various intervals ranging from 15 rnin to 6 days. A second group of 22 animals r'tas administered a singl-e dose of an aqueous extract of Crotal"gria fulva at 0.5 mg/g body weight, by gastric tube, andETIEl-d-E ffi-3ame intervals. A third group of 8 rats was administered a total of 3.2 tines the LD56 dose of Lasiocarpine in smal1- doses, 3 times a week, and killed at 9 - 20 weeks. The changes prinarily invoLved the nucleus and interchromatin granules. The first change, seen after 30 min in the nueleoli of the hepatocytes and Kupffer cel1s of animals receiving lasiocarpine or crotalaria exCract, consisted of a separation of the fibriLl,ar and granular components. The hepatocyte nuclei had returned to normal after 72 h and remained so throughout the rest of the study. In animaLs receiving a singl-e dose of lasiocarpine or crotalaria, round periodic acid schiff (PAs) positive eosinophilic bodies appeared in the cytoplasm after 12 h, consisting of dense masses of cytoplasmic naterial. Five days after treatment with crotalaria, large cells lined the luminal surface of the centraL veinsl the centrilobuLar cells had undergone necrosis by this stage. Animals receiving 3.2 times the LD50 of lasio.carpine developed megalocytosis after 9 weeks (section 6.4.L.5). The cytoplasm shoned vesicles of smooth endoplasmic reticuLum with nitochondria of various shapes and sizes. The appearance resembLed an exaggerated regenerative response. AlLen et al. (L967, 1969) studied uLtrastructuraL changes in the liver tissue, in general, including the hepatic veins in Macaca speciosa monkeys treated with PAs. In the study on hepatic veins (Allen et al., L969), 18 treated and 6 control aduLt animals were used with an average weight of 5.8 kg. AnimaLs wete divided into 3 groups of 6 treated and 2 control animaLs each. The experimental animals received 0.125 g monocrotatine/kg body weight by ip injection. Livet wedge biopsies were examined at various intervals in Group I during hours 1 - 48, in Group II at 4 - 12 days, and in Group III at L6 - 32 days. The earliest changes, observed by light microscopy, nere seen at 24 h and consisted of progressive Loss of endothelial cell-s and other associated changes in the lumen and wall leading to occLusion by coLLagetization by the third week. Under the electron microscope, within 24 h of administration, marked changes vrere observed in the endothelial celLs resulting in their rupture and release of organelles in the l"umen. This vas followed by penetration of fLuid though the vessel walls in the first week and changes in the fibroblasts. By the third to fourth week, hepatic veins showed various levels of occlusion and the vessel r^tas scattered with cell debris, free organelles, etc. The authors concluded that, in this species, hepatocelluLar necrosis was not the prirnary factor causing veno-occLusive disease, as the -97- association of cel-lular necrosis and venous occlusion occurred only in the central area of liver Lobules, aod the hepatocytes surrounding the sublobuLar and nediun sized hepatic veins were unaffected. In their study of 1967, Allen et al. al.so investigated the ultrastructural changes in the liver of 1,1. speeteee monkeys afteradministeringIdo'""of1gmonocro-5ffiondays l- and I-4. At autopsy, after a mean survival tioe of 21 days, a wide spectrum of changes was observed in the hepatocytic organel"les, many of wtrich rrere 1ying, discharged into sinusoids, and also phagocytosed by the Kupffer cells. By the third week, proliferation of connective tissues had started in the sinusoids near the central veins,and al-go in the wal1s of central veins. The authors concluded that the vascular and parenchymal cell changes were sinultaneous and appeared to be equally instrumental in the development of the occlusive lesion.

6.4.L.4 Mechanism of toxic action

The mechanism of toxic action in acute pyrrolizidine hepatotoxicity and the sequence of events, judged from the collective experimental studies, appears to be as follows. The PA, which is inactive as a cell poison by itself, becomes cytotoxic through its metabolisrn in the hepatic parenchyrnal cells to pyrroles, which act preferential-ly on the hepatocytes and the endotheliurn of blood vesseLs in the liver or 1ungs. In the hepatocytes, the imediate action is a rapid fall in cytoplasmic protein synthesis reaching 30% of. control levels at 15 rnin artd 6% at t h (Harris et alr, 19.69). This is manifested as disaggregation of polyribosomes and is followed by failure of pyruvate oxidation, loss of glycogen, structural damage to the mitochondria, 1-ysosomaL actlvity, failure of mitochondrial nicotine-adenine-dinucleotide (NAD) systemb and nuclear NAD synthesis, and necrosis (Mclean, 1970) . The necrosis is zonal in the liver lobule, the particular zone affected depending on the metabolic enzymic, geography of the lobule in the particular animal species, and aLso in man and monkey on the vascular endothelium of the central and sublobular veins . The sequence of events of the vasculaf lesion has been studied by Mclean et al (1964). After a singl-e dose of Crotalaria fulva extract in the rat. centrilobular necrosis is present after the first day, but collagenous veno-occlusion of the central veins of the l-iver lobule only appears between 7 and 10 days later. Evidently, the necrosis of the liver ceLls is not secondary to venous occlusion. Centrilobular haemorrhage is seen from day 2 onwards and signs of hepatic venous outflow tract obstruction appear after 2 - 5 days (Mclean & Hill, L969).

7 -98-

Rappaport et al. (1967) and Mclean (1969) demonsrrated, through transillumination studies on rats, that the outlet end of the sinusoids is blocked by stationary coluurs of red cells, L6 - 24 h foLLowing adninistration of PAs. The reaction is typically patchy and results in stasis and extravasation of red ce1ls spreading backwards from the centre of the lobuLe. For at least 3 days, no circulatory detail can be seen with transillurnination. Portal pressure is significantLy raised 3 days after administration of fuLvine (Rappaport et al. , 1967), notabl"y before the first appearance of collagenous venous occLusion at 7 days. Mclean (1969) observed that 6 - 10 days after PA adninistration, a new irregular pattern of vascular flow, contrasting nith the uniform radial pattern of flow in the normal liver Lobule, develops, which corresponds to the bypass channels represented by dilated paraseptal sinusoids, as observed in human liver biopsies (section 7.3). Segments of central vein into which the blocked sinusoids open, are gradually abandoned in favour of such by-pass routes and undergo occlusion first by oedematous connective tissue and then by fibrosis. The mechanism of closure of the sinusoids is not clear. A toxic action on the sinusoidal or venous endothelium, which swe!.1.s and occludes the lumen, seems possible, as suggested by A1len & Carstenrs (1968) electron microscopic studies on the monkey, and studies on chil"dren (Brooks et a1. r 1970). The endothelial lining of the vessels is denuded and replaced by a fibrinous and proteinaceous precipitate, which, together !,rith the oedenatous wall of the vessel, becomes organized and slowly replaced by fibrous connective tissues. The occlusion of sinusoids is further contributed by the discharge of cellular debris into the space of Disse. The lumen of the sinusoids becomes occluded simuLtaneousLy lrith the fibroeis occurring in the centraL vein. Collagen fibres extend into the space of Disse and sinusoids leading to a ereeping fibrosis. The proximate toxin that escapes from the liver is returned to the heart, after wtrich it damages the first portal of entry into the alveolar capillaries of the lung and pulmonary arteries.

6.4.1.5 Chronic effects

The chronic hepatotoxic effects of PAs have been described in a nunber of studies on a variety of animaLs and have been reviewed by Bull et al. (1968) and Mclean (1970). A notable feature is that an appropriate single dose of PA has been demonstrated by Schoental and Magee (L957, 1959) to lead to a relentlessly progressive course and eventualLy kilL rhe -99- animal, more than 18 rnonths after administration. Schoental & Bensted (1963) demonstrated that rats receiving a single dose of PA may develop chronic liver disease aild finally hepatocellular carcinoma more than 13 rnonths after administration. The morphological changee in the liver are similar in a given species of animal., whether a singLe sublethal dose is administered or multipLe srnall doses. Schoental & Magee (1957) studied the lolrg-term effects of a single dose of l-asiocarpine on rats receiving normal and protein-deficient diets. Albino rats of Porton strain were used. In the first study, 66 rats fed a normal diet were administered a single oral dose of lasiocarpine at one of 3 dose levels (50 - 74 rng/kgr 75 - 100 rng/kg, or 101 - 150 rng/kg body weight); 24 animals served as controls. In another study, 46 young female rats nere administered a protein- deficient diet. Of these, 1.3 and 10 anirnals received a singLe oral dose of lasiocarpine at 50 - 100 mg/kg and 50 - 75 urg/kg body weight, respectively. Each of these grouPs was pair-fed with an identical number of animaLs that did not receive any PA. 0f the 66 rats fed a normal diet, vet'y few died in the first 10 days. Thirty-one animals survived longer than 3 months. They continued to be in good health until shortly before death. The numbers of animal-s that survived for 13 months after administrat.ioa of lasiocarpine vrere 8/10 rnales and,717 females (lowest dose) and 5/25 rnalee and 11/18 femal-es (intermediate dose). In the group that received the highest dose, neither of the 2 male animals survived longer than 35 days, and only Ll4 female animals survived longer thaa 3 months. In the animals that died or,nere killed when moribund, parenchymal damage was invariably present vith prominent megalocytes, ductular proliferation, and invasion of Lobules by oval or eLongated cells, thought to be derived from the bil,e-duct epithelium or the reticuloendothelial ce11s. Anirnals that survived showed various degrees of fibrosis and noduLar hyperplasia and, in sorne, a mild thickening of the central- veins. No obliterative lesions of the veins were seen. The 31 animals that survived 13 months showed similar changes, but to a much lesser extent. In the livers of animaLs that had repeat liver biopsies, there nas no tendency to regression of the lesions. The above data indicate that very fev animal-s died of acute disease. In most anirnals, there was a tatent period of 3 - 4 weeks, during which they remained well and showed little evidence of liver cell- injury, followed by a progressive course often leading to fibrosis and nodular hyperplasia. The low-protein diet adversely affected the growth of all the rats in the control as well as the treated group. Only 3 out of 23 treated animal-s remained alive and in apparent good health, B - 11 months after the treatment. Liver biopsies -100-

taken at various intervaLs between 4 and l0 months showed very severe fatty changes in the liver ce11s. There was 1ittle fibrosis and no bile-duct proliferation. In areas where the fatty changes r^rere less ievere, characteristic megalocytes rirere seen. Control animals had either normaL Livers or showed only slight fatty changes that rirere not comparabLe in severity with those in the livers of PA-treated animals. Thus protein deficiency in the diet was shown to enhance the toxic effects of the PA. - Schoental & Magee (L959) extended these studies on young Wistar rats using several other pAs incLuding freliotiine, retrorsine, riddelliine, seneciphyl-line, monocrotaline, and its N-oxide in various dosages ranging frorn 25 to 300 urg/kg body-weight; the animals wJre stud]ied- at death from 1 - l0 - days to 18 30 months. Pathological changes were similar to those observed with lasiocarpine in the previous study. Notable observations r^7ere that necrosis did not necessaril.y precede subacute or chronic changes. The livers of some aaimals became severely damaged and showed noduLar hyperplasia. Liver biops!,2 - 3 days after pA treatment, did not show pathologicaL changes in some animals, but a repeat biopsy at 41 days showed characteristic changes. Fibrous thickening of the central veins was observed in some animaLs, more often with monocrotaline N-oxide, but no occlusion of the hepatic veins was seen. The studies of Nolan et al. (L966) confirmed the observations of SchoentaL & Magee (L957, 1959). They gave a single dose of lasiocarpine at L2O mglkg body weight, by stomach (.equaI tube, _ to 108 numbers of both sexes) Sprague Dawley weanling rats (60 - L35 g body weight). thirty aninals of both sexes served as controls. Groups of lO animals, each consisting of 8 treated and 2 control animals, were kiLLed at various intervaLs from 1 to 123 days. Of the g0 treated animals, 28 died within 26 days. No delayed hepatic lesions were found in 59 rats between days 1 and 18. Betveen 19 and L23 days, delayed lesions were found in 34/49 rats. These 34 rats showed megalocytosis, but no ductul-ar proliferation or f ibros is . In a second study, 127 twenty-one-day-oLd male l,listar rats were given a single oral dose of lasiocarpine at 80 ng/kg body weight; 65 animals served as controls. Liver biopsy was carried out on day 3 in 108 and on day 9 in j.5 of the treated animal.s. In 47 animal-s, additionaL biopsies rrere carried out at intervals. 0f the 127 PA-Eceated rats, 98 died during the first 9 days, and 29 after L0 - 50 days. In contrast to the first.study, 32158 survivors exhibited deLayed subacute and chronic lesions, as described by Schoental- & i{agee (1957). Of these, 8 animals developed cirrhosis. The observations indicated that the lesions of acute zor.a]- necrosis. which -101 -

appeared on, or before, the third day, healed without residual Lesions. However, 55"1 of the 58 sur'vivors developed subacute/chronic lesions that tended to be progressive after a latent period of 2 - 3 weeks. There agpeared to be an intimate relationship between chronic lesions and megalocytosis, which rf,as seen ln 52158 surviving animal-s. No obliterative vascular changes were observed and so the lesions couLd not be ascribed to impaired circulation. Schoental (1959) demonstrated the toxic effects on the newborn of PAs administered to lactating mother rats . l'listar rats (200 - 300 g) were administered lasiocarpine, orally or by ip injection, at 25 - 40 mg, in 5 - 10 doses of 5 - 10 mg, twice weekly or more (24 rats), or retrorsine aE 4 - 10 mg per dose, in 1 - 14 approxirnateLy daily doses (23 rats). The litters were left with the mothers fot 24 days or more (except for L/2 hour separation during the PA treatment). The litters were examined by biopsy at frequent intervals or at autopsy when they died or were killed in a moribund state. Litters of the lasiocarpine-treated rats showed onLy insignificant fibrosis or some megalocytosis. In the retrorsine-treated group, the majority of the young rats survived for about 18 days, but all rats died before reaching the age of 30 days. The miLk secretion of the mothers was appartintly not affected by the PA treatment. Of the 98 animals in 9 litters, 45 died by the 20th day and 45 survived 30 days. Animals dying in the first fortnight did not sholv gross liver lesions, but those that died at weaning time or 1-ater, all showed signs of liver disease. Animals dying at 18 - 30 days showed hydropic or fatty vacuol-ation of hepatocytes and haemorrhage into distended sinusoids. The change $ras severe in anirnals dying at 1 - 2 months, and some central veins showed a narrowed lumen. The l-actating rats Ehat receivdd the alkaloids survived longer than their young, and most showed no i11 effects from the treatment. This evidently indicates either a high susceptibility of suckling rats or a high concentration of PAs in rniLk. In studies by Atlen et al. (1963), 2 grbups of A-week-old turkeys, each consisting of L2 birds, $rere fed diets containing ground Crotalaria spectabilis seed at 2.5 g/kg and 5 glkg, respectively, for 120 days. Twelve animals served as controLs. At the end of the study, LU12 birds receiving 5 elke sqed and 6lL2 receiving 2.5 glkg seed in the diet developed cirrhosis. The minimum period of feeding required to produce cirrhosis was 75 days, provi{ed the diet \,ras reduced to a level that was not l-ethal. AlLen & Carstens (1971) induced the Budd-Chiari syndrome in monkeys by monocrotal-ine. Six adult female and 9 adult male Macaca speciosa monkeys, weighing 5.2 -,6.5 kg each, were -toz- divided into 2 groups, each comprising 5 control (3 males and 2 females) and 10 treated animals (6 males and 4 fenales). The treated group lras given a subcutaneous (sc) injection of monocrotaline at 60 ng/tg body weight at monthLy intervals for 3 months. Needle biopsy of the Liver was carried out every month for 5 months and laparotooy, 6 months after PA treatment. The treated animals showed marked vascular changes and various degrees of occlusion in the centrilobular and sublobuLar veins as weLl as the larger vessel"s. There was also characteristic haemorrhagic necrosis in the centrilobular zones and megalocytes \rere seen. The portal, venous pressures were raised. The aninals lrere autopsied at 6 months. The livers were markedly shrunken weighing an average of 68 g in contrast to those of control animals, which weighed 130 g. There were severe occlusive vascular changes and irregu!.ar fibrosis in the lobules. The adjacent sinusoids were dilated as a comPensatory mechanism. Swiek et al. (1982a) studied the effects on guinea-pigs of Long-term dietary adninistration of Senecio jacobaea and compared then nith the toxic effects--6?-ffiffi-6Fes of injected Senecio alkaloids and monocrotaline. The possible protective-Fect of cysteine lras aLso examined. Fifteen guinea-pigs of 250 - 300 g initial body weight were divided into 2 treated groups, being fed 102 Senecio jacobaea, or LOi( Senecio jacobaea plus 1% cysteine in EE-Ee:TiiitT control group. The whole plant of Senecio jacobaea was air dried and powdered for incorporation into the diets. The animals were fed for 365 days. They were autopsied at death of, at the termination of the study. In a second study, 7 guinea-pigs of 500 g body weight were injected intraperitoneally with either monocrotaline, jacobine, or nixed Senecio jacobaea PAs. The chronic lethal dose LD166 of -EEffiETET-uoaySenecio jacobaea ttas L264 glkg initial body weight B"zlozoe weighr witir an average survival time"t of 279 days. No mortality was observed in controL animals. This contrasts lrith the chronic LDrno of Senecio jacobaea for rats of 581l of initiaL body wei!f,i (SwiiI-Et a1-.,-T929-) and that of cattle equivalent to 5 - 2O% body weight (4u11 et a1., 1968). Addition of cysteine to rhe diet was only s1ight1y, but not significantLy, protective. Pathological examination of the livers of the guinea-pigs fed Senecio jacobaea revealed extensive megaLocytosis and severe cytoplasmic vacuoLation with biliary hyperpLasia and fibrosis, prinarily in periportaL areas. The centrilobuLar and midzonaL areas were spared. Monocrotaline was noa-toxic at doses up to 1000 ng/kg body weight, whereas jacobine and mixed alkaloids from Senecio jacobaea were Lethal at much lower Levels. Sinilar ;€ffi' showing resistance to monocrotaline in guinea-pigs rrere also -103- reported by Chesney & AlLen (1973a). In in vitro studies, they relatld this resistance to lack of coiffi of PA to pyrroles by guinea-pig microsomes. A morphological pecuLiarity of chronic hepatotoxicity in a large variety of laboratory and farm animals is megaLocytosis (Bull, l-955,' Mclean, 1970), i.e., the appearance of exception- aIly large hepatocytes, 10 - 30 tirnes the voLume of normat cells with proportionately Large nuclei. Relevant literature has been reviewed by Jago (1959), Mclean (1970), and Mattocks (1986). Advanced megalbcytosis was produced by Jago (1969) within 4 weeks in 2-week-old rats by administering a single dose of lasiocarpine at 76 urnol/kg body weight. MegaLocytes tend to appear in the periportal and midzones of the Liver lobules with normal sized ceLls around the central veins. The nuel-ear chromatin is proportionately increaeed, but the cells appear incapable of entering into nitosie, as only abnormal mitoses are seen. Jago (1969) demonstrated a fall in the mitotic index (from L.61 to 0.04) in liver cel1s of 2-week-o1d rats, one day after injection of 50 umol lasiocarpine/kg. The eLectron microscopic appearance also Eupports the above observations (Afzetius & SchoentaL, 1967). A striking proliferation of rough endoplasmic reticqLum and multipLe centrioles is seen in the cytoplasm, and the cytoplasmic organelle6 are disorganized, suggesting increased metabolic activity but inability of the ce1ls to divide. Such ce1Ls may persist for the life-time of the animal (up to 2 years in the rat) and the liver never returns to normal (Mattocks, L986). Megalocytes have also been described in the kidney (fu1l et al., 1968), the l-ung (Baroes et al., L964i Butler et al., l-970; llooper, 1974), and the duodenum (Iloopet, L975c). Data on the total chronic lethal dose of heliotrine in rats were discussed by Bul1 & Dick (1959) and Bull et al. (1968). For a variety of dosing rates, and with withhoLding periods of 10 - 20 veeks interposed, the total doses ranged from 2.2 to 7.8 LD5O. In Table L0, these data are extended with results for olher alkaloids. The overall range of the total lethal dose is 1.2 - 10.3 LD56.

6.4.2 Effects on lungs

Current Literature has been extensiveLy reviewed by Kay & Heath (1969), and Mattocks (L986). PAs have been shown to produce pulmonary hypertension with associated vascular changes in the puLnonary circulation in a number of experimental animal species incLuding the rat, morirse, frog, turkey, pig, sheep, rabbit, and horse (Mclean, 1970) as welL as in non-human primates (Allen & Chesney, L972; Chesney & A1,Len, 1973b) and the dog (Miller et a1., L978). The alkaloids have been administered by feeding the animaLs with: PA-containing _104_

Table 10. Total chronic lethal doses in rats (ip administration, 2 or 3 tines per week)

Dose (x L\q) Time to Total lethal Reference death (dayg) dose (x LD59)

Helrofrane GffiEG, unless otherwise stated)

o.2 58 5 Bull & Dick (1959) 0.1 L23 5.1 0.04 303 4.L 0.02 508 4.L 0.01 5.2 - 5.3 Bull & Dick (1960) (with inrerval of 10 - 20 weeks afcer 21 days) 0.11 2.2 - 4.3 Bull e Dick (1959) 0.1 7.8 Jago (1971) (35-day-old male rats) 0.1 4.7 Jago (1971) (337-day-old mle rats) 0.1 5.8 Jago (1971) (35-day-o1d fenale rats) 0.1 4.) Jago (1971) (337-day-old femle rats)

Las iocarpine G;T-;CT

0.t 210 9 Culvenor & Jago 0.05 482 10.3 (1979) 0 .02 o/o 0.01 59s 2.6 ( 0.005) ( 638) ( 1.4) 0. t- 81 - 175 2.4-5 Bull- & Dick (1959) 0.1 6.3 - 10.9 Jago (1971)

0.1 1.08 4.6 Culvenor & Jago 0.05 274 5.8 (L979) 0 .02 47L 4 0.01 487 a1 (0.005) (692) ( 1.5) 0.1 2.4-7 Jago (1971)

Monocrotaline 1.2 - 2.4 Bu1l et al. (1968) 0.05

SenecioninC 0.2 2 - 7.4 Bul1 et al. (1968) 0.1 0.04 (3 survived)

9 Assuming LD5g mg/kg (c.f., Mattocks, 1986). -105- seeds of pLants (notably Crotalaria spectabiliq) (Turner & talich, 1965; Kay & lteath,-T966; Kay et ai., L967a) or the dried plant itself (..g., Senecio jacobaeb) (Burns, 1972)' .q,r"o,r" soLutions or rut-vine@ 1964; wagenvoort et al., 1974arb) or monocrotaLine (r,atictr'& Ehrhart, 1962; Iluxtable et al., 1977), subcutaneous injections of monocrotaline (A1len & Chesney, L972; Chesney & A1len, 1973b) and seneciphylline (Ohtsubo et a1., I977), or intravenous injections of some pyrrolic esters and analogues of pyrrolizidine alkaloids aod their metabolites (Mattocks & Driver, 1983). Lafranconi &. Iluxtable (1984) studied the hepatic metabolism and puLmonary toxicity of monocrotaline in in vitro perfusion studies. Some of the representative studies on the morphological effects of toxic lung injury are listed in chronological order in Table 1L. Chronic lung lesions have been produced by most compounds that produce chronic liver lesions, though higher doses were required in some instances (culvenor et al-., I976a). Ilowever, not all PAs that are hepatotoxic are also pneumotoxic. Among the pneumotoxic alkaLoids, fulvine (Barnes et al.., 1964) and monocrotaline are particularly active,(Mattocks' 1986). Molecular structure activity requirements are the same as for hepatotoxicity, since both are caused by the same toxic metabolites produced in the hepatocytes.

6.4.2.L Acute effects

Pulmonary lesions produced by PAs have been extensively investigated, mostly in rats, but al-so in non-human primat.es. Monocrotaline has been the alkaloid most frequently used, but lung l-esions have also been seen in rats following fulvine and seneciphylline administration. Besides pure alkaloids, PA-containing seeds of some plants, most notabl-y Crotal-aria spectabilis. have also been used. Mi11er et a1. (1978) gave a single iv injection of monocrot.aline at 60 ng/kg body weight to 10 mongrel dogs. Toxic effects, recorded within 2 h, incLuded ultrastructural changes in the endothelial ce1ls of .the alveolar capillaries, prominent accumulation of plateLets, and Ehe appearance of interstitial oedema (table 11). Valdivia et al. (1967a,b) used 25 Sprague DawLey rats in their study and made similar observations oo the rat Lung within 4 h of a single subcutdneous injection of monocrotaline at a dose of 60 rng/kg body weight (Table 11). Ioterstitial oedema and elastolysis of the alveolar wal1, increase in number of mast cells, and other associated changes r^7ere observed within 4 h of the injection, foLLowed by alterations in endothelial and interstitial ce11s. A11 of the changes progressed steadily for up to 3 weeks. It was concLuded that the initial changes - t_06-

N o. @ o !dd a a o o !! q.o o o q o > 6 F^ 6 d H^ g 5! .dF oo !o 06 Qn N o !o iN .il@ do 16 so FO o 6A .dr aa (3 pa o )@ Fv !

dr!-N@ o o eeL.i o O . O d Fg ! ldo !l oso .io o oru o o dr! k E o I O O O€ 3q O q g 6 I o s !o il - d od 6.d F6 r! 6i o 6 i d k tr q' F .rd.r ./CO )9t l^ 9.d @ o.d o o E a h h o .ri>. .t 6OO kor ! od I o H 6 > ts 6N tr @! o o o 6 > p E o o !u! F o.r 6!tso o! o ooo odd FEoia Euo o !Es udo o@a0 r.ro c 6 cud QEE .i.rh9 tlo o .r o O .il u ! E ts.c 3 AvE t a tl .3, x Od oo 6 0 o !6 JE h E H> oo d !o.d € N BU O O oo @6J>'N +o k N d ak o€l po I t! 'E I .r90 ddo od qd v oq Ndd N O h N o O.t @ OA ooo do ddJ SP qo usu .dC .d2 rd .d .a @o @e oo6 ots oo oo poa otog ooo, 3l Ov EEC! to oo6.i .il* !>so lo otl F! 5Sr.J g.d .i dO oHo- l9 d|! !o g tso o!o lo lo €!o to Elo EE < oE O.dOd olq OJi ls o 66 @ !{!o ts^ hE xt hI ! ! d oi\ !@ €€ 900 | o oo to o@ O.d O.d o.d O'i 4tl o€ geg Nv 6 3eg seg 3

o 3 .31

h ooo o .d .d .l X did s s*lilr o o 6d6 o !!! ooo ! f :'[l 3 o FFh t i"'l I o 3t3 k X d ggs e *t!lE 6 o do !q6L E oF! ho-Sl lo 2 .aC z 90 sl ; qq * * - o .El Hlol o Hl F ;l 3.4 ! $r 'il :3"r .r guO!x ;l *[.Er B'9. *EE; H[Er < oo @4ts ;lOl koAh -107-

q I 6 o 4 !6 do ON € F '1 o x6 oa dq lZv >/) e

I q xF I ll !Q EI oi I o ! 61 8r r FE ho o o k E o >r ts E o >. E ts E ii h I h.d Fk.rEO diO F S-lOF.r 6dAts!!! dc A e ts d.d n t x q n t xq k d t >d o o o E k o o o 5 oE 9O 9 ou O o 3^=Ef ;*'E*Ei E AH -F .d .i F6 .d 'dFH .d ^rFiEC d xQ xt k >.6 d F h6d ! hsd d .i tl 5 )E €.i !3./ J u €.d 5 ^ e"C.i 5 ts 5 -! co a hoq cq! '.: 0o dFo oOo 5&b*.i 53b',94.:E >! I d s kd > ts ! >Et;.:5 >!!'o '!u E.r E ! d ! ! ;.: O ! O !.t ! ts d Ek !tsdErE -rHF'ih'if, -op .r i|, F od ! o .-&*BgEg" a !uil qH sgh tg €"&-tg fl O ! 5 Xd .d h! H! .ihtstskE .dx!F!!!6 rJ ! 6 a! o kE 6 ud h56!6! kE6!6!dr

q >| o o o x F. N q d F ts € s F I o I I I I o6 o 6 N F

o o o o o0

o o o

o t. k o ! u d o d @ 6o 6 o !o I !E

!! t! u x5 >5 h.c oq EU a o.. 3.: o€.P 3.S O'd aa! o0 Np o o! o op o dv iv9 6v9 ovB

o @l 6l.dl .JIJI d tsFd I dl!l oo i|;l rlOl EITI, dlulo O.d !lol! o> EIH,|E elHs cl ol olo otolo FH

$ do ld

s NO N N- s o lo .d5h Ots o'/ k dH o d @o i d 6 trs o:dd 6s D F o !u eoo! n9k3! gd! s 6 t QA d o.d d F kSrF h 33 ! >o@Ah k;3F -108-

>E o-. c: oj ()€

o {x^ F o bN d6 3l o oq

a ll :.1 i I :':l!. " o H{!FEr:-ifI hi.i xt rEiEi." o {s Ei'i!'.: i: !Ft?6?6'[sIi. E g d i i; i! :i :i t; F; .i EH.l:.: o o ieI ilefie*l; Bfr !aiF:i.ir o sil5f iE f sf EY.I F$ir eflI: o : 89119 !e Fk 9h! g-. uk6ts 6H A.; o 'iFr!F.!iE &E&E.= 3.s f"3.'i3j&;: 3 3.FeFee3J'd .l-asteFH3 F !6 R.t*.ih'i"iE:;YoS8o th v.a r.i ' I O O^ q d. , I I ", 6>,>d ->!i 'd !o -lgG N o9 1g-9* e.sJi : €J$ r. >, _- o qo too o , is-:qir-L^ 3d "SSoEgf.g88.9- Eet:i.5'i B

o o l. o o 6

E !o VDO < o9

@ x5 Xj !@ o.i op o 33 s8's o"3

o o .i d @ o

F o

! )s -.== 4 -!^ c :w

i5 !LON ol ?r 0l dl!l .tdl h ol.dl ^ X uFil ^ o 610l! o o qlolE o 'E!6Ed I olul! a F dlolo x a o q jltE =tFI6 v E E dlv o E-oF-o I -109-

0.o d oa gd !d @ !s o 6qr iI o ,F. id6 o !+ Ur !6 x6 3d ov EV ttts I F.r F o I trok o dr I d I x d9A!.E i5k d da uE 6 d E hE >td O O O D 6 d & tr d tr./dtO3O .i.d! dd .r oOat okt >a> !tsqejd !o,r>a .aorF O o!dE.rE iDEu jJ !!6!!HugvO6.i!6.il!L ots ^dEuO !O .a U, ! !Eo .dohosF o d>xL!6 dho@d k E.i.i o @o E d 'f !.i x u>q!tsotrH xpo-dN xo!tsi gHEg -F'Ee Hfo9"Zc H3'.-* OE9Hh qlO E.dujl .r'd C+h^d 9OX6€F5> O!hots otstsdn 3l&xe.E.:o l- 3.9:6 lE.Et : !!tl,FsrgE ,Fx o.il sFtoN.d > OE s! u! 6 A6!o >! i\E E- F

@06 rth oo! 3! €on llr

Jd6

oo @s

oo a tsto EI 9.do al @rd dod .ll'o uadu G sd 5 el.5'atJ ts Fo o -16 ! uc ! c tro 3 ol0 ,iU dlq O^ u!6u hEXEXt! E.i _ E.q' , E.3 3.9 3e9b8eg 3eg 93

s O I N

o OF FO o O O^ Rh alli Fg o Ei E 69 0.J d IHEH H KBts EEEst SSOF -110-

of destruction of pulmonary capillaries and the other components of the alveolar wal1 preceded the arteriolar lYpertrophy and arteritis observed by other investigators folLo-wing monocrotaline administration, and alone lras sufficient to cause right ventricular hypertrophy. Sugita et al. (L983arb) adninistered a single dose of monocrotaline at 40 ng/kg body weight to 5 Sprague Dawley rats and adduced further evidence, by biochemical and radioisotopic studies, of microvascular leak in the alveolar wall within the first 3 days of- injury, which preceded right ventricular hypertrophy observed 2 weeks following administration (Table 11). A histological and eLectron microscopic study rdas nade by llurley & Jago (1975) of the Lungs of rats administered dehydromonocrotaline. FemaLe bl-ack and white hooded rars - weighing 80 100 g were used. Dehydromonocrotaline dissoLved in dinethylformamide (DlF) nas administered iv as a single dose at 30 mg/kg body weight ro 12 rats and ar 15 mg/kg body weight to 7 rats. Four control rats were administlret outr. alone. A coLloidaL suspension of carbon black was injected - iv, 6 18 h after injection of dehydromonocrotaline, .rid th" animaLs killed 19 - 44 h after treatment. After an interval- of 6 - 8 h, there was a direct toxic effect on the endotheLial cells of puLmonary capillaries and sma1l venuLes. Many endothelial cel1s had prominent nuclei and thickened cytoplasm, which contained more RNA granules than usual. There was also an increase in the number of mitochondria. The endotheliaL damage did not seem to have caused permanent disruption of the srnall blood vessels , and, 2 days after injury, all vessels were patent. Large numbers of mononuclear cells, which appeared in the interstitiaL tissues -o-f the lung 44 h after injury, seemed to be altered emigrated bl.ood monocytes.

6.4.2.2 Chronic effects

Lalich & Erhart (L962), fed 35 Sprague Dawley rats a diet containing monocrotaline at L0 - 30 ng/kg (Table 1l). Anirnals receiving a daily dose of monocrotaline of 2O mg/kg diet. or more, showed progressive changes in the lungs after 24 days of feeding. Of the 23 anirnals receiving 20 - 30 mg/k1, 12 showed pulnonary arteritis, 4 of these even at the dose levet of 20 mg/kg diet. PuLmonary haemorrhages were observed in 16 animals. No changes were observed in aninaLs fed alcohol- extracted seeds or other derivatives of monocrotaLine. Identical. _ -changes were observed in similar studies by Turner & Lalich (1965) on tlro strains of rats, Sprague Dawley (19) and Wisrar Furth (2D. The first groop of 19 femaLe Sprague-Dawley rats was fed a diet containing monocrotal"ine at an initial level of L0 ng/kg diet. nependirig on the response -111- of each individuaL animal, the monocrotaline Level was raised to a naximum of 75 ng/kg diet (table 11). Fourteen rats survived for more than 100 days and 8 reached the naximrm dietary 1eve1 of monocrotaline, the last aninal dying after 232 days. The second group of 24 femal-e Wistar Furth rats lqas fed a diet contaminated ttith Crotalaria epectabilis seeds, initialLy at 0.2 mg/kg diet, grffielke by increasing the Levels by 0.2 ng/kg every week. A11 test animaLs survived 100 days and 15 reached the maximum levels of Crotalaria fed. The Last animal died after L72 days of E dTn& /{nimaLs developed signs of toxicity and right ventricular strain, e.g., cyanosis, etc. Progressive thickening of rnedia in the muscular pulmonary arteries, progressive muscularization of art-erioles, and changes characteristic of pulnonaiy hypertension, nere seen. Some pulmonary arteries showed mediaL necrosis. No changes were observed in the pulmonary veins. Significant hypertrophy of the heart, as judged by the heart weight in relation tor body weight, was seen in aLmost all animals that survived 100 days or more (32/38), and the right ventricles lrere dilated. The hypertrophy affected the right side of the heart onLy, and geoeraLly corresponded with the vascular chalrges. There wae a marked hyperplasia of the me.st cel1s in the, mediastinal lymph nodes and around bronchi and pulmonary arteries. Sinilar observations were nade by Barnes et al. (1964) and Valdivia et a1. (1967a,b). Kay and his group studied cardiac and puLmonary vascular changes in rats fed CrotaLaria spectabilis seeds (Kay & Heath, 1966; Kay et a1., lffiea furvine (Kay et al. , 197la). A group of 10 fenale weanling Wistar albino rats lrere fed a diet containing L g powdered seeds of Crotalaria spectabilis/kg until th-ey d-ied of cardiorespirato@r 36 - 60 daye of feeding. Thirty-four control rats were fed a normal diet. At autopsy, the atria of the heart, the right ventricle, and the left ventricle with the interventricular septum were weighed separately. The rnediatr thickness of the muscul-ar puLmonary arteries was measured, and expressed as percentage of external diameter. The medial thickness of the muscul-ar pulmonary arteries increased in al.l test rats; acute or healing pulmonary arteritis lras seet in 3 aninals. Statistically significant cardiomegaly was present in aL1 rats fed the seeds, contributed chiefly by the right ventricle. The readings from al.1 the test rats vere welL outside the upper 951l confidence Linit. The increase in the nedial thickness of the pulmonary arteries correlated well with the weight of the right ventricle (fig. 10). Th:ee rats showed pulmonary arteritis (indicated by a solid triangle). It lras presumed that the organic basis for increased pulmonary -Lt2-

resistance r{ras the abnormal muscularization of the radicles of the pulnonary arteria! system. Essentially sinilar resuLts were obtained in an identical study repeat.ed on 8 test rats and 5 controls (Kay et aL., L967a). The test rats developed pulmonary hypertension in 37 days, levels of which ltere correlated lrith the mediaL thickness of the muscular pulnonary arteries and that of the pulnonary trunk, as well as with the weight of the right ventricle.

15 s q 6 (1) :c .e E 10

(d € o

0.05 0.15

RVweights x 100 body weight

Fig. 1"0. Ttre relation between the average percent nedial thickness of pulnonary arteries, and the weight of the right ventriele, expfessed as a Percentage of the total body weight, in control and test rats (Kay & Heath, 1966). Control aninals - oi test animals - o; animals showing pulnonary arceritis - A.

Ghodsi & Will (1981) made similar observations in Sprague Dawley rats given a single subcutaneous injection of monocrotaline at 60 ng/kg body weight. Forty rats rileighing 180 - 200 g nere used; of these, 20 constituted the control group. The control animals received the same volume of saline. Each week, 3 rats fron each group \rere catheterized and pulmonary artery pressures were measured. In the treated - t-13 -

group, 2 out of 5 animals showed a rnild increase in pulmonary artery pressure at the end of 8 days. A further 4 out of 5 animals showed a nild to moderate rise ic pulmonary artery preasure after 2 weeks. The highest value recorded in test rats rilas 56 mrnllg cornpared with a normal upPer lirnit of 22 mrnlfg. The mediaL thickness of pulmonary arteries was correlated with pulnonary artery pressures (P < 0.02) as rrras the thickness of the right ventricle. The correLation between the pulmonary artery pressures and right ventricular hypertrophy was statisticalLy significant (P < 0.05). Kay et al. (197La) administered a single dose of fulvine to rats, intraperitoneally at 50 mg/kg body weight, or through a stomach tube aE 80 ng/kg. Of the 30 treated rats, L7 survived 23 days. A11 of the animaLs showed changes characteristic of hypertensive pulmonary vascular disease with right ventricuLar hypertrophy and muscular hypertrophy of the pulmonary trunk and the muscular pulmonary arteries. Pulmonary arterioLes !'rere also muscul"arized and contained fibrin thrombi. Four animals showed pulmonary arteritis. EssentiaLly sirnilar changes in the pulrnonary arterial system were produced within 20 - 28 days by llayashi & Lalich (1967) in 22 male suckling rats administered a single injection of monocrotaline at 120 rng/kg body weight, and within 28 days by Ohtsubo et al. (L977) in L6 maLe, 4-week-old rats given a single injection of seneciphyl-Line at 80 ng/kg body weight. Ilooper (1974) did not find any such effect on feeding powdered Senecio jacobaea at 100 - 200 ng/kg diet to 9 rnale white rnice for up to 193 days. In studies by Allen & Chesney (1972), don-human prirnates (Macaca arctoides) were administered 4 dosee of monocrotaline a?-To--=--,66-Te/te body weight by subcutaneous injection (tabte tt). Twelve infant monkeys (30 days o1d) and 12 adults (L5 rnonths oLd) were studied r^rith different results. Vascular changes, characteristic of pulmonary hypertension and resultant cor pulmonale, were observed in the infant monkeys, as described by earlier workers in the rat. Only isol-ated smalL hepatic veins were occluded. On the other hand, the adult animals showed a more severe involvement of the liver with changes characteristic of veno-occlusive disease and onl-y an occasiona!, pulmonary blood vessel rtas invol-ved. The authors postulated that the different responses in infant and adult animals were due to the different stages of rnaturation of the enzyme systems of the hepatocyte in the two age groups. It is possible that the different reactions in the Liver and lung in the 2 groups may be due to the fact that the enzymatic pathways responsible for producing metabolites that cause hepatic darnage are poorly developed in the infant, but those responsible for causing pulmonary lesions are better developed. -LL4-

Chesney & Allen (1973b) made observations sinilar to those of AlLen & Chesney (L972) in twenty, 30-day-o1d monkeys in a similar study using monocrotaline injections and recorded, in addition, endocardial fibrosis of the right heart. The treated animals developed classical clinicaL features of cardiopulmonary distress, which was also evidenced by changes in the blood-gas parameters. The raised right heart pressures were confirmed by actual measurements of the blood pressure in the right ventricle, pulmonary artery, and descending aorta. The authors considered this study to be a good experimentaL model to investigate hypertensive pulmonary vascular disease, or pulmonary and endocardial fibrosis. The type of vascular changes seen in the animals were comparable rn?ith those associated with puLmonary hypertension in man in cardio- pul-monary disease (Barnes et al., L9643 Kay & Ileath, 1966; fay et al., L967a; Chesney & Allen, 1973b). Wagenvoort et al. (L974arb) made light rnicroscopic and ultra-structural studies on 50 female Wistar rats, 1 - 6 weeks folLowing a single oraL dose of fulvine at 80 mg/kg body weight or an ip dose at 60 rng/kg body weight. Twelve animal.s served as controls. Vasoconstriction of muscular pulmonary arteries and arterioles was seen initiaLLy, one week fol.Lowing administration. This was evident by the coiLed appearance of the muscular nuclei and excessive crenation of the internal eLastic lamina. The nucLei of smooth rnrscle cel1s as wel"1 as those of endothel-ial cel-1s were partl-y squeezed betroeen the folds of the lamina. After 3 - 4 weeks, these blood vessels began to thicken, with muscuLar hypertrophy and fibrinoid necrosis of the arteriaL muscle. Animals surviving administration of fulvine developed right ventricular hypertrophy, proliferation of endothelial ceLls in the arteries and even thickening of the veins. In a study by Meyrick and Reid (1979), 2L Sprague Danley rats v/ere fed a diet containing 1 g powdered seeds of Crotalaria spectabilis lkg for various periods ranging from 3 to 35 days. The earliest demonstrable change in the puLmonary arterial system of the animals rilas seeo on day 3 and consisted of the appearance of muscle in normaL1y non-muscular arteries of the lung. The muscular puLmonary arteries began to show hypertrophy of the media from day 7, which reached statistically significant levels on day 10 in srnaller arteries and on day Lh in the larger arteries. Significant right ventricular hypertrophy was seen on day 21. These ehanges rdere confirmed by fi-thymidine uptake studies (Meyrick & Reid, 1982).

6.4.2,3 Mechanismsof toxic action

Considerable progress has been made recently in the understanding of biochemicaL and pharmacological changes that occur in ?A-induced lung disease. -115-

Turner & LaLich (1965) and Takeoka et al. (1962) postulated that pulmonary hypertension nas rnediated by the release of 5-hydroxytryptamine from the tnast celts, which became hyperplastic in the rnediastinat Lymph nodes and around bronchi and pulmonary arteries (Turner & Lalich, L965) (section 6.4.2.2) following administration of monocrotaline, causing vasoconstriction. 0n the other hand, Kay et aL. (1967b) found that the number of mast ce!"lsi corresponded with the severity of exudative changes in the lung and $rere not related to the genesis of pulmonary hyperplasia. Besides the medial muscuLar hypertrophy of pulmonary arteries reported in several studies cited above, swelLing and tysis of the endotheliaL cell,s, contributing to LuminaL narrowing and thickening of the walL with fibrosis, have been described (A11en & Carstens, 1970). Weanling rats are more susceptible to these changes than older animals, and the changes follow a strict temporal sequence. Oral administration of mooocrotal"ine to rats at 20 mg/litre in drinking-nater produced a sequence of changes over 3 weeks, that included an increase in lung mass, which was significant by d,ay 9, stirmrlation of puLmonary RNA and protein synthesis (rnaxinal on day 10), increased pulmonary arterial blood pressure (significant by day 10), and right ventricular hypertrophy by day L4 (Huxtable et al., 1"978i Lafranconi et aL., 1984). The increase irl the lung mass was not accompanied by change in the totaL collagen content and was contributed possibly by hypertrophy of endothelial cells, but the increased mass of the right ventri,cle was associated with a 4-fo1d increase in collagen content (Lafranconi et al., 1985) . An early event is inhibition of seiotonin removal by pulmonary eodotheliurn (Ituxtabl-e et a1., 1978). This phenomemon, combined with che increased reilease of serotonin by mast ceLls that has been observed, may be involved in the development of pulmonary hypertension (Caril1o & Aviado, 1969). Right ventricuLar hypertrophy r is blocked by propanolol, whereas the developmeot of pulmonary hypertension is unaffected (Iluxtable et al., 1977). Novpl metabolites have been found to be reLeased by livers perfused with monocroEa- line in vitro, and these metabolites bLock serotonin transport in vitlg, r,rhen perfused through isolated lungs (Lafranconi & Iluxtable, L984). These data suggest that the sLow release of netabolites from the liver into the circulation foLlowing low- level exposure to monocrotaLine resuLts in specific inhibition of endothel.ial cel.l function (HuxtabLe et a1,, L978). The effect of monocrotaline treatment on pulmonary angiotensin converting enz)rme (ACE) activity in the rat i; disputed. Itayashi et al". (1984) observed a reduction in the ACE activity of pulmonary tissue in pyrrolizidine-exposed rats -116- in parallel- with the deveLopment of pul-monary alterations, while the ACE activity of the plasma remained uachanged. However, other authors have reported that, though the specific activity of ACE fall-s in the isol-ated perfused lungs of monocrotaline-treated rats, or in lung homogenates from such animals, when activity is expressed as totaL activity per lung, there is no significant alteration in the lungs of lreated animals compared with those of untf,eated animals (Iluxtable et a1., I978; Lafranconi & Iluxtable, 1983). Therefore, the significance of changes in ACE activity is open to question. MoLteni et al. (1984) also found evidence of endothelial ce11 damage by monocrotaLine in their ul-trastructural and biochemical- studies on rats. Eighty roale Sprague Da\rley rats were used; half were administered monocrotaline at 20 mg/1itre in the drinking-water and half were given plain r^rater. The average daily water consumption was 35 ml/rat. Thus, the treated rats r^rere estimated to have received 2 mg/kg per day. Five animal-s each from the treated and controL groups rirere killed at intervals of L - 12 weeks after the start of the study. The endothelial darnage lras measured by ACE activity, plasminogen-activator (PtA) activity, and prostacycLin (PGI2) production. These vrere correlated with pulmonary arterial perfusion and ultrastructural changes in the Lung. In the treated groups, after an initial rise at 1 week, the ACE activity showed a steady decline from 1 to 6 veeks, after wtrich it plateaued at 55i( of normal. pLA activity did not change for 2 weeks, but decreased by 59 and 79% of the controL value after 6 and L2 weeks, respectiveLy. On the other hand, the PGI2 production increased progressively reaching 140 and 2707" of the control level after 6 and 12 weeks, respectively. These endotheLial functional- changes were not accompanied by significant changes in pulmonary arterial perfusion as visual-ized by s mTc-labelled macroaggregated albtunin perfusion studies. The activities of ACE and PLA and the production of PQIt are considered sensitive indices of endothelial functio-n in rats. The above results indicated endothelial celL dysfunction. The uLtrastructuraL studies also reveaLed oedema of capiLlary subendotheLial, perivenous and periarterial tissues at I Lreek, and interstitial inflarsnatory infiltrates at 2 weeks. At 6 - 12 weeks, there rtas thiokening of the pulmonary arteries and enlargement of right side of the heart. Stenmark et al-. (1985) studied the role of alveolar infLanrnation and arachidonate metaboLism in monocrotaline- induced pulmonary hypertension in rats. Five groups of male Sprague Dawley rats were treated as follows: (a) 20 rats received 40 rng monocrotaLine/kg body weight, scl (b) 20 rats received mooocrotaLine, 4O ng/kg sc plus diethylcarbamazine _ 117 _

(DEC) 100 mg/kg sc, every 12 h; in additioo, 250 mg DEC was added to 100 srL of drinking-\rater. This treatrnent started 2 days prior to the start of the study and wab continued dai!-y for 3 weeks,' (c) LZ controL rats received normal saline pLus monocrotaline at 40 mg/kg sc; (d) 12 rats ieceived indomethacin at 2 mglkg sc for 2 days, prior to receiving monocrotaline at 40 mg/kg and then daily for 3 weelts; (e) 6 animals each received a single sc injection of normal saline and served as additional controls. One, 2, and 3 weeks after monocrotal-ine or saline injection, lung Lavage was carried out for cell counts and assay for enz)rme activity and cyclooxygenase metabolites, the degradation products of prostacyclin (fCf7) and thronboxane A2(TXA2), 8s 6-keto-prostaglandin (fClro) and TXB2, respectively. At 3 weeks, the animals wetre anaesthetizea, right ventricular pressures measured by catheterization and the heart removed. The 2 ventricles srere separated and weighed for the determination of heart-weight ratio (right ventricl-e/left ventricle + sspluns RV/LV + S) an indicator of right ventricular hypertrophy. The right ventricle shoved hypertrophy at 2 weeks and the right ventricular pressure was increased at 3 weeks following monocrotaline administration (Fig. 11). The leukocyte count in the lavage fl,uid increased at 3 weeks, with a rise in the percentage of poLynorphonucLear Leukocytes artd Large, abnormal alveolar nacrophages in the test animals. B-N-acetyl--D-g1uco- seaminidose activity was aLso elevated at 3 w1eks, indicating activation of leukocytes. There was also arrise in the concentration of 6-keto-PGF1o at 1 and 3 weekE, as well as in TX% at 3 weeks, compared with those in contrbl aninals. The administration of DEC inhibited both the increase in heart-weight ratio (nv/r,v + S) and the increase in ouLmonarv artery pressure (fig. 11) that occurred 3 weeks after monocrotaline administration, and reduced the percentage of polymorphonuclear ce11s, abnormal- alveol_ar macrophagesr aod hexoseaminidase activity in the lavage fl_uid, compared with that from animals that had received monocroEaline onl-v. The riseinthe1eve1sof6-keto-PGF1n\'asinhibited(p< - 0.05) by 73% and thar of TXB2 by 74"1 i; the 1-ung lavage. The administration of indomethacin did not have any effects on either the heart-weight ratio or the puLmonary arterial pressure 3 weeks after monocrotaline administration (rig. L1), bur ir inhibited (p < 0.05) rhe rises in 6-keto-PcFlo (by 9O%) and TBr CUy gl-"/") rhat occurred in the lung lavage of monocrotalineltreated animals at 3 weeks. The above studies indicate that both the cyclo-oxygenase and the lipo-oxygenase pathways of arachidonate metabolism are activated by monocrotaline as early events in its toxic effect. -118-

O control O monocrotaline A DEC+ monocrotaline O Indo+ monocrotaline tr NS+ monocrotaline

60 Fpr

t (mmHg) 5(! 40

0 1 2 3 Sweeks WEEKS

Fig. 11. Effects of monocrotaLine administration on pulmonary artery pressures and right ventricular hypertrophy in rats and of the pharmcological blockade of pathways of arachidonate metabol-ism by diethy).carbanizine and indonethacin.

Note:

Left axis: measurements (SE) of heart/weight ratios Iright re"ttfrT7 left ventricle + septum (nv/r,v + s)l, 1, 2; and 3 weeks after administrarion at time zeto of monocrotaline. Right axis: mean pulmonary arterial pressure at 3 weeks [40 i-[7['[f-6]-tean of 15 ratsl or of saline (o, nean of 6 rats), subcutaneously. a = mean values for 15 rats given monocrotaline at time zero pLus daily diethylcarbanazine (osc) for 2 days before and 3 weeks after monocrotaline. v = mean vaLues for 12 rats given monocrotaline at tioe zero plus daily i.ndomethacin. fi = mean values for 12 rats given monocrotaline plus daily nomal saline injections. :r' = values different (p < 0.05) from those in untreated controls. + = values different (P 0.05) from those in nlonocrotal ine- treated rats. + = values different (P < 0.05) from those treated with nonocrotaline and nornal "ilir". Source: stenmark et al. ( 1985 ) -119-

Activatioa of the cyclo-oxyenase pathway, demonstrated by increased concentrations of the prostaglalrdin netabol.ites 6-keto-PGF1o and TXB2 in lavage fluid, was inhibited by indomethacin, but this inhibitioo did not prevent the monocrotaline-induced injury. DEC attentiated both the inflarmatory response and pulmonary hypertension and inhibited the formation of slov reacting substances iincLuding leuko- triene D4. Since DEC produces a pharmacolosicaL bLockade of the f.ipo-oxygenase pathway, it seems that the latt.er, rather than the cyclo-oxygenase pathway, is lresponsible for perpetuating the pathophysiological mechanism Leading to monocrotal ine- induced pulmonary hyper tens ion. HiLliker et a1. (1984) demonstrated that antibody-induced thrombocytopaenia att.enuates right ventricuLar hypertrophy induced by monocrotaline in rats. In anothei study, Hilliker & Roth (1984) also produced evidence that hydrallazine, a vasodiLator and inhibitor of platelet prostaglandin synthesis, dexamethason, an antiinflarnnatory agent and inhibitor of phospholipase, and sulfinopyrazone, an inhibitor of pLatelet prostaglandin synthesis inhibited monocrotaline-induced right ventricuLar hypertrophy in rats, supporting the hypothesis that platelets and vasoconstrictor agents play a role in monocrotaliae-induced pulmonary hypertension. Likewise, prior chemicaL sympathectomy with 6 hydroxydopamine (100 mg/kg) or inhibition of serotonin synthesis with p-chlorophenyl-alanine (500 ng/kg) reduced the degree of monocro-taline-induced right ventricular hypertrophy in rats, but did not prevent or reduce puLmonary vascular muscularization (Tucker et al., 1983). Thus, the synpathetic nervous system and serotogenic mechanisms seemed to be invoLved in the development of right ventricular hypertrophy, but not in the developmenc of the pulmonary vascular lesion induced by nonocrotaline. Kay et al. (1985) also demoostrated that pretreatment vith g-chlorophenyLalanine, which inhibits 5-hydroxytryptamine (5t{P) synthesis, also significantly (P < 0.05) reduced right ventricular systolic pressure, right ventricular hypertrophy, and nedial thickness of anuscular pulmoilary arteries in monocrotaLine-treated rats. SimiLar observations were made in rats exposed to hypoxia. It was therefore suggested that 5HP might play a role in monocrotaline-induced or chronic hypoxic pulmoaary hypertension. The biosynthesis of rat lung polyamines, putrescine, spermidine, and spermine, generally considered to be important regulators of cel1 gro$rth and differentiation, is increased prior to the evolution of monocrotal-inqinduced puLmonary hypertension in rats. Continuous adminigtration of a-di- fLuoromethyLornithine (OfUO), which is a highl.y specific irreversibLe inhibitor of ornithine decarboxylase (OOC), a rate-tiuriting enzyme in polyamine biosynthesis, atienuated the -L20- development of monocrotal.ine-induced pulmonary hypertension in rats (Olson et a1., 1984). This effect rrras mediated by the DFMO, by inhibiting the synthesis of putrescine aad s.permidine, and not by blocking the hepatic rnetabolism of monocrotaline to pyrroles (Olson et al., 1985). Thus, it was suggested that lung poLyamine biosynthesis rnight be essentiaL for the expression of monocrotaline-induced perivascular oedema as well as rnedial thickening in the development of monocrotaline-induced pulmonary hypertension vascular disease. On the basis of the preceding studies, mostly on the rat, the mechanism of chronic long-term injury to the lung by monocrotaline seems to be as folLows. Inlithin hours of pA administration, there is damage to the pulmonary endotheLial cells accompanied by vascular leak leading to pulmonary oedema. Platelet aggregation also occurs. The endothelial damage indicated by ul-trastructural and biochemical studies activates the production of prostacyclin and lipogenic products, which mediate increases in vascular permeability and inflanrnatory reaction. There is simultaneous production of 5 hydroxytryptamine and several polyamines. The injected monocrotal-ine is completely metabolized within hours, and no significant quantity is found in the body at 24 h (Hayashi, f966) and, though some active metaboLites may still be detectable by isotope studies, even at L4 days (Hsu et al., 1974), the rats do not have any lung lesions. The slow evolution of vascular changes suggests that it is not caused by monocrotaline but through biological pathways activated by the initial injury. Methylprednisolone (l'{P) , which reduces acute lung oedema caused by monocroEal-ine (MCT), has been shown to reduce MCT-induced .pulmonary hypertensive vascular changes in rats and the resultant right ventricular hypertrophy (Langleben & Reid, 1985). Daily ip administration of MP ar 5 mg/kg body weight, was found to be more effective than 2 large doses of MP at 30 mg/k1, 2 h before and 2 h after a single sc injection of MCT at 60'nrg/kg. It was suggested that secondary changes, though triggered by the acute MCi injury, become self sustaining and are more significant for vascuLar structural remodelling. Structural arterial remodelling with vasoconstriction, nedial- hypertrophy of the muscular pulmonary arteries, and muscularization of the pulmonary arterioles follow as late effects, resulting. in pulmonary hypertension and right ventricular hypertrophy of the heart. The results of the above studies suggest a direct toxic effect af the alkaloid on the endothelial cells of the aLveolar capillaries and on the pulmonary arteries, as well as a pulmonary hypertenoive effect on the heart. -121-

6.4.3 Effects on the central nervous system

The dominant signs of pyrrolizidine poisoning in horses are neurological- (Rose et al., 1957arb1' Mclean, 1970). Sirnilar signi caa also occur in cattle and sheep. It has been clained thlt such signs are probabLy non-speci-fic secoodary effects following ptit"ty Liver disease resul'ting in hyPer- amnonaemia (nose et aL., L957a). Ilowever, neurological abnormalities in which animals walk in a straight f ine uotil they come to ao object, and then stand with their heads prelsed against the object, indicate specific Lesions in the ceatral nervous system. Spongy degeneration of the central nervous system occurs in cattle, sheep, and pigs (Hooper et aL, , 1974; llooper, 1975a,b). Trichodesma al-kaLoids, in particular, apPear to be neur6ffiEere is a considerabLe body of literature in the USSR on Trichodesma intoxication of mice, rabbits, and dogs, which hail6iln reviewed (rsnailov et a1-., 1970). Mice given Trichodesma alkaloids subcutaneously at 0.5 ng/kB deveLop ;€rrJffiTthe hind f.irnbs vithin L2 - 17 days. opisthotonus - and clonic convulsions are also seen. Doses of 10 15 ng/kg - of alkaloids produces death in aLl animaLs within 2 6 h, as the result of respiratory depression. Higher doses produce imnediate death.

6.4.4 Effects on other organs

Right ventricular hypertrophy, secondary to the prinary effects on the pulmonary arteries, and the resultant pulmonary hypertension in animaLs treated with PAs or PA-containing pLants have been

120 ng/kg body weight. RenaL changes were seen, ro sorne extent, in aLL animals surviving for 4L - 47 days. Carstens & AlLen (1970) studied rhe effects of feeding 9rotaLaria spectabiLis seed on the rat kidney. Fifty nate Sprague Dawley rats lrere fed- a diet containing ground - @__"p""8!:1is seed at 0.2 0.8 g/kg fot 8 months. Tffied to contain approximaiely 3.5 g monocrotaline/kg; 10 aninaLs served as controls. Renal Jhanges nere seen in 33/50 PA-treated rats. In 22 rate, over 75"A of the glomeruli were hyalinized and capsules thickened. In the less severely affected kidneys, the glomerular baeement membrane was thickened and homogeneous deposits were seen in mesangial areas. Afferent arterioLes and interlobular arteries rtere narkedly thickened. In the most severely affected vessels, the internal elastic lamina rilas necrosed and the larger arteries showed fibrinoid necrosie. Renal tubular megal.ocytosis nas the dominant lesion described by Eooper (L974) in mice. Nine male white nice, 10 weeks of age, were fed Senecio jacobaea, which contained a concentration of altaT6Ta-iftconin-nine, jacobine, and seneciphylline) of 2.7 glkg and a concentration of N-oxide of 0.9 gl1rg, nixed with the diet. The S. jacobaea wal eiven ar 190 e/te diet for 9 weeks, before b=ffiEEffid to Ioo g/tg diet. Five aninals served as controls. The animals rrere killed from 63 to 193 days afrer the srarr of the srudy. A11 treated aninals, except 2 killed on day 63, shoved changes. The large cells occurred in both the proxinal tubules and the loop of llenLe. Sinilar ce1ls were seen in the al_veolar and bronchiolar epitheliun. No glomerular lesions were described. Th€ author mentioned having seen the above changes in rats given repeated subletha!. injections of fu!.vine and specta- biLine. On the other hand, Kurozumi et al. (1983) observed glomerular l-esions in rats given a singLe injection of monocrotaLine. A variety of renaL lesions has been observed in pigs, a conmon pathological feature being renaL negalocytosis, wfrictr was observed, in pigs poisoned by at Least 4 different plant genera containing a variety of toxic alkaloids (Harding et a1., 1964; Peckham et aL., L974) and has aLso been observed in wilh pigs grazir.g in areas rich in pA-containing plants in northern Australia (Hooper, 1978). McGrath et aL. (1975) described gLomeru!.ar lesions in pigs given Crotalaria spectabilis seed daily for 43 days. Severe renalf--LEET6ns comprising tubular dilatation, megalocytosis, and necrosis of tubuLar epitheliaL cells with casts in the lumen, interstitial and periglornerular fibrosis, and glomeruLar hyalinization lrere reported by Hooper & Scanlan (L977) in pigs fed Crotalaria retusa seeds containing monocrotaLine. Rena!. reg;i;Fffi EiEA-lso been reported in C. retusa poisoning in h'orses, -L23-

sheep, and mice Poisoned by S. jacobaea but not by H' europaeum, and in vervet rnonkeys tiFt, ch.orri" retrorsiiZ poisonifi (van der watt et aL., L972). tesions have been reported in the stomach and intestines in fieLd and experimental animals after poisoning with pyrrolizidine aLkaloids, but are difficult to identify as siecific PA injury. Hooper (1975c) conducted studies on rats, and mice. In the study on s-heepr t2 male "he.p,crosi-bred lambs, 7 - 8 weeks of age, were newly, neaned on to a standard cotunercial calf grorter diet. Lasiocarpine rtas administered at the rate of 15 - 20 rng/kg body weight every 2 - 4 days. Each animal was killed when in terminat coma' Survival time ranged from 4 to 17 days. In the rat study, - young tJistar-Furth rats (sex not stated) weighing 150 200 g were used. In one group of 11 rats, each animal received an ip injection 6f lasiocarpine at the rate of 40 rng/kg body ,".igti. Three animals received isotonic saline. Animals were kiLled or died 2 - 6 days after the injection. A second group of 13 rats received dose of 35 mg Lasiocarpine/tg uoay " weight; d control animals received saLine. A11 rats received - a second injection 48 h 1ater. They were killed 3 60 days after the second administration of l-asiocarpine. In the mouse study, 3 rnature male white mice received 6 injections each of lasiocarpine at the rate of 45 mg/kg body weight followed by 4 injections of 90 mg/kg body weight at 48-h intervals. There rdas one control animaL. A11 animals showed characteristic hepatic lesions. Sheep also showed severe oedema, haemorrhage, and epithelial necrosis in the gall bladder; Lesions were al"so found in the central nervous system and occasionaLly in the kidney. ALl animals showed severe intestinal atrophy. There was inhibition of -crypt cel.1 mitosis leading to rnitotic irregularities, abnormal 1,atge celLs and syncitial ce11s, especially in the duodenum of sheep, and severe vil-l-ous atrophy with ulceration' Lesions in the intestines were similar to those caused by radiation and radiomimetic agents. It nas suggested that the local intestinal radiomimetic effect r'tas due to I'ocal exPosure to the pyrroLe metabolite of lasiocarpine after excretion through the bile duct. It ltas proposed that a more and rapid development of duodenal megalocytosis ras"oo"pi"rrous dne to very rapid turnover of cel1s in the duodenum. Other probabLy secondary effects incLuded haemolysis in sheep in. associa-tion with advanced liver disease and high livei-copper levels (sul1 et al., L956), anaemias and disturbance in iron metabolism and haematopoiesis (Schoental & Magee, 1959; Schoental,, 1963; Peckham et aL., t974i Hooper & Scanlan, lg77) (section 6.4.11), pancreatic oedema and fibrosis (Bras & Ilil1, 1956; Schoental & Magee, 1959), cerebral oedema, haemorrhage, and congestion in the rat brain (Davidson, 1935; RosenfieLd & Beath' 1945). -t24-

Tumours in the different organs have been dealt lrith separately under carcinogenesis (section 6.4.8).

6.4.5 Teratogenicity

The teratogenic potential of pAs was demonstrated by Green & Christie (1961) who produced a variety of dose-related fetal abnormaLities in the rat, with a singl_e intraperitoneal injection of heliotrine administered during the second week of gestation. The dosages ranged fron 15 to 300 mg/kg naternal body weight. Litters exposed to a dose of Less than-5O ng did not show any abnornalities, but abnormalities were observed in lit-ters exposed to higher doses, and increased in frequency and severity with increasing dose. The abnornalities includei retardation of development, musculoskeletal defects, especialLy hypoplasia of the Lower jaw, cleft palate, and other abnormalities. Doses above 200 mg resulied in the intrauterine death or resorption of many fetuses. Sinilar studies were perforned by peterson & Jago (1930) who compared the effects of heliotrine with its metabolic pyrrole derivative dehydroheliotridine (DfH), when administered in a sing!.e ip injection to rats on the 14th day of gestation. IteLiotrine was administered at 200 ng/kg body weight and DIIH at 30 - 90 ng/kg, 14 days after gqagsp;ien. Effects on errbryos, evaluated on the 20th day, showed that both heliotrine and DIIH retarded grovth and weie teratogenic, but that the effects of a 40 ng/kg dose of DIIII were equivalent to those of 200 ng/kg heliotrine, i.e., the rnetabolite was 2.5 times as effective on a molar basis. DHII produced a number of skeletal abnormalities including retarded ossification, distorted ribs, long bones, cleft palate, and feet defects. At higher doses, growth almost ceased in many tissues and the fetuses lrere very irmature. Horilever, the embryonic liver parenehyma did not show the antimitotic effects of DIIH. The teratogenic properties of heLiotrine rrere also demonstrated in Drgggphilg Larvae fed low levels of rhe al.kaLoid (Briak, 19EZ)1--

6.4.6 Fetotoxicity

_ The subject of fetotoxicity has been reviewed by Mattocks (1986). Sundareson (1942) demonstrated the of pyrroLizidiae alkaLoids to cross the rat placenta."Ulflty Twice weekly injections of the pA, starting at, or after, the 12th 9ly of gestation, resuLted in premature delivery of some litters and many were born dead. The same author jhowed ttrat the alkaloids themselves and not just the pyrroles formed in ! th: dams livers, could pass the pLacentaL barrier by injecting senecionine into 19-day-o1d rat f etuses in utepr -L25- which produced the characteristic toxic lesions in the dams. The fetuses were also found to be more resistant to the Lethal effects of the PA than the mother rats. When 4 fetuses were each administered 1..25 ng of PA, representing about 200 400 rng/kg body weight, which is much higher than the LD56 for an adult rat, 3 of then were sti11 alive after 2 days. creen & Christie (1961) did not find any Liver damage in fetuses from pregnant rats given teratogenic doses of heLiotrine. Only nild liver damage wag found in the embryo rats whose mothers had been injected nith PAs (heliotrine, Lasiocarpine, retrorsine, or monocrotaLine) (Bhattacharya, 1965). In contrast, Schoental (1959) demonstrated that lasiocarpine and retrorsine, when administered to lactating rats produced littLe effect on the mothersr but produced acute liver lesions in the suckling infants. The lesions were most severe in 3- to 7-week-old animals. It lras suggested that the infants were affected by the milk from the lactating mothers, which possibly contained the metaboLic products of the PAs. It woul-d seem from the above studies that the embryo is reLatively more resistant to the toxic effects of PAs in utero rhan it is after birth. Mattocks & White (1973) postulated that this could be due to the low capacity for the metabolic activation of PAs of the embryo liver, as they had shown that the ability of liver enzymes to convert retrorsine to toxic metabol.ites was Low in rats, imnediateLy after birthr but picked up rapidly afterwards. The susceptibiLities of rats of various ages to the hepatotoxic effects of the PAs r,tas proportional to their capacity to form and retain the pyrrolic metabolites. Twenty-day-old rats rilere found to be more sensitive than older animals. The effects of fuLvine administration on pregnant rats between 9 and L2 days of gestation r'rere studied by Persaud & Iloyte (1974). Dose-related fetal resorPtions were observed, but no hepatic lesions r^tere seen in the fetuses. On the other hand, Newberne (1968) observed damage, in both the maternal and fetat livers, when lasiocarpine roas administered to pregnant rats. Acute liver necrosis was observed in the livers of mothers as well as fetuses in animals that had received L00 arg Lasiocarpine/kg body weight on day 13 of gestation. Ilowever, in animals that received 2 doses of 35 mg/kg body weight on days 13 and 17 of pregnancy, liver necrosis iras seen in the fetaL Liver but not in that of the mother. It is not known why lasiocarpine acts differently from other alkaloids and has a greater effect on the fetal liver. uattocks (1986) has postulated the possibiLity that fetotoxicity was caused chiefly by toxic metabolites formed in the maternal liver, and that a greater proportion of such metabolites reached the fetus from Lasiocarpine than from other PAs. -L26-

6.4.7 Mutagenicity

A number of PAs that have been shown to be oowerful dose dependent mutagens in DrosophiLa melanogaster have been listed by ltattocks (1986). A@eparotoxic though the degree of nutagenicity is not necessarily proportipnal. TabLe L2 provides a surnnary of the mutagenicity tests on different PAs, related conpounds and plant extracts. Clark (1959) demonstrated the mutagenic effect of heliotrine in Drosophila, in which a considerable increase in sex-linked recessive lethals was produced, apparently by interfering with the maturation of germ cells, so that as soon as the available spermatozoa were used, the males were no longer capable of breeding. The ce11 damage was irreversible. The mutagenic effect of feeding Drosophila males for 24 h with a medium - containing 10 3 moffifficro-ialine rdas comparabl-e to about 1000 R of X rays (Clark, L976). The Basc test wirh Drosophila melanggaster _is considered a highly sensitive .uFage-@ test for PAs (Candrian et aL., 1984a). SeneciphylLine and senkirkine, known to occur in animal feeds and medicinal herbs, respectively! were tested for their ability to produce sex-linked recessive lethaLs in males of DrosophiLa melanogaster using the Basc (3-day feeding method) by Candrian et aL. (1984a). SenecipbyLline was found to be mutagenic at concentrations of L0-5r 10-r, and 10-3 moL, otrich produced 3.82 (983 chronosomes tested), 9% (708 chrosro- somes tested), and L5.3i4 (327 chromosomes tested) sex-linked recessive lethals, respectively. Senkirkine (10-s. mol) $ras found to produce 4.4"/" sex-linked recessive lethals (254L chromosomes tested) against 0.172 maximum sensitivity in the late spermatid stage of spermatogenesis indicating that pAs act as indirect mutagens. Flies fed with niLk frorn lactating rats given an oral dose of !J ng seneciphylLine/kg showed 1.22 sex-linked recessive lethaLs (L477 chromosomes tested) compared lrith 0.3"A (1533 chromosomes tested) in controLs. Mutagenic properties of 7 PAs extracted from plants to Salmonella typhinuriun TA100 have been demonstrated by a @ yamanaka et al. (Lglg). The pAs were clivorine, fukinotoxin, heliotrine, l-asiocarpine, ligularidine, LXC201, and senkirkine. preincubation of these alkaloids with liver 59 mix and bacteria in liquid medium was essential for demonstration of the property. pAs in the hetiotridine ' and otonecine fanily $rere mutagens, while retronecine bases were inactive. MonocrotaLine and heliotrine were not active mutagens to Escherichia coli tilp2, even though they were quite cytoroxic (cG-ffiiTilLg7r. They were active in repair deficient strains. Retrorsine was active in inducing mutations on the Ames Salrnonella/rnicrosome assay (wehner et al., 1979). Exrracrs Tffi'GdTcinal pLanrs and noxious weeds were mutagenic towards Salmonella in the Arnes assay (Pool , L982; White et a1., 1,983; K;Tdffi? a1., 1978). -t27-

TabLe 12. Mutagenicity tests on pyrrolizidine aLkaloids, related compounds, and source plants

Compound or material Type of tes€ Respons&

C livorine A HPC

Echinidine D

E chinat ine

Fulvine D

Heliotrine D P F cc A B Til c!1/cc

Integerrimine D

Jacobine n

Lasiocarpine D

l{Pc clr/cc Tlt

Ligularidine

L indelofine

Lycopsamine

Monocrotaline n P

HPC CT

Petasitenine ( fukinotoxin) A HPC clt -L28-

Table 12 (coqtd).

Compound or naterial Type of test3 Response!

P latyphyll ine D 0

Retrors ine A + D CT +

Rosmarinine CT 0

Senecionine D + A 0

Seneciphyll ine A o P + +

Senkirkine + HPC c}l/ cc + D

Supi.nine D ! +

Mixed alkaloids from Senecio jacobaea 0

senecio numorensis ffiI@T&':tract) CM 0 +

Senecio jacobaea (extract) +

Serecio longilobus (extract) 0

Symphytun officinale (confrey extract) 0

Retronec ine b is-q-chloro- P benzoate

Synthanecine A bis-N-ethy1- CT carbamale

Retronecine 0

Heliotrid ine D 0

vlrlcollorlc acld 0

Heliotric (heliotrinic) acid D I HPC + -L29-

Table 12 (contd).

ComFound or macerial Type of test9 Respons e!

Dehydroretronec ine CT A scE t

Dehydrohel iotrid ine CM

Pyrrole HPC 0

2 ,3-Bishydroxynethyl-1-me thy 1- CT + pyrrole 1 scE

2-Hydroxy lnethyl- 1-nethy 1- pyrrole scE I

3-Hydroxy lnethy 1- 1-me thy 1- + pyrrole scE +

A = Salnonella ("Anesrr) test. B = Other bacterial lests. CC = Clastogenic activity in cultured ce.lls. CM = llutagenicity tn cultered mmlian cells. CT = CelI lransformation test. D = Mutasenicitv in Drosophila. F = Tests in fungus (Aspergillus nidulans). HPc = He pa t ocy t e pr inl:y-6ilfEFbT[-iEfiii r te s t . P .= Chromosoml aberrations in plant cells. SCE = Sister chromtid exchalge. TM = Transplacenlal micronucleus test.

b + = active. I = narginally active. 0 = inactive.

From the lirnited data avaiLable, it aeems that the carcinogenic activity of individual alkaloids par411els their mutagenic behaviour, but not their relative hepatotoxicities (Culvenor & Jago, 1979).

6.4.7.1 €hromosome damage

Pyrrolizidine alkaloids have been shown to be capabLe of danaging chromosomes in pLants, fungi, bacteiia, tissue cel1 cultures, and the fruit fly (Drosophila nelanogaster). Literature on this .topic has been@l . (1968), Mclean (l-970), and Mattoeks (1986). -130-

Several PAs are known for their ability to damage the chromosomes of growing plant cells (Mattocks, 1986). Sinilar properties have been demonstrated in leukocyte cultures from the marsupiaL (Potorus tridactylus) (Sick & Jackson, 1968; Bick, 1970). 8m971) found dehydroheliotridine, a metabolite of heliotrine, to be L0 times more active than the alkaloid. Infusions of Symphytum officinale L., described in PoLish pharmacopoei. .s Bg!!1gp!y!i., are recormnended as expect- orants, especially foi Childien. Furmanowa et al. (1983) demonstrated the mutagenic effects of an alkaloidal fraction and infusion in this plant in the meristematic cells of the lateraL roots of Vicia faba L. var minor. Lasiocarpine, a proven carcinogen, =ffi-Fa positiveGErol. Chromosomedamage by PAs in the hamster lung cell line was demonstrated by Takanashi et al. (1980). Stoyel & Clark (1980) used the transplacental. micronucleus test in pregnant femaLe mice and showed the chromosome damaging properties of heLiotrine (225 rr'glkg body weight) and lasiocarpine (86 rng/kg) within 20 h of the injection. The genotoxicity of heliotrine, monocrotaline, seneciphylline, and senkirkine was studied by Bruggeman & Van der Iloeven (1985) using the sister-chromatid exchange (sCE) assay in V79 Chioese hamster cel1s co-cul-tured with primary chick embryo hepatocytes. Exposure to these PAs resulted in the high induction of SCEs, a more than 5-fo1d increase in the SCE rate with 2.5 mg heliotrine/Litre, 4-foLd with monocrotaline at 5 mg/Litre, 8-fold with seneciphyline at L.2 mg/litre, and more than 5-fo1d response with senkirkine at 2.5 mg/Litre. For alL compounds, a dose-response relafionship hras observed at concentrations that did not seriously affect survival. PAs are also knoriTn to induce DNA repair in rodent hepatocytes (Green et al., 1981; Mori et al., L985). DNA repair synthesis was elicited by 15 alkaloids, including 11 of unknown carcinogenic potentiaL (Mori et a1., 1-985). There are also a few reports of chromosome damage by PAs in man. Martin et al. (1972) found chromosome damage in the bl-ood, ceLLs of chiLdren with veno-occlusive disease, probably caused by fulvine. It has also been shown by Ord et a1. (l-985) that dehydroretronecine, is abLe to induce SCE in human Lymphocytes. Kraus et aI-. (1985) studied the Pa? senkirkine and tussiLagine, which occur in a medicinal plant lgif.3g. farfara, for their ability to induce chromosome damage in human lyrnphocytes in vit.ro. They were irot found to enhance the number of chroEffiberrations up to concentrations of '.^How€ver, 1000 'uno1. heliotrine, used for comparison, 'chromosomal ' induced aberfations at concentrations of 1-00 unol . In addition, heliotrine r^ras also f ound to be capable of damaging unstimulated eg Go-phase lymphocytes. - 131 -

6.4.8 Carcinogenesis

Carcinogenesis has been reviewed by Mclean (1970), IARC (1976, L983), and Matrocks (1986). A number of purified pAs, purified or crude extracts of plants containing them in a mixture or the actual plant, dried and mill-ed, and several PA metabolites or synthetic analogue compounds have been tested for carcinogenecity. llowever, these include only relativel-y few of the known cytotoxic PAs. Data reLating to some of the iepresentative studies on rats are sumarized in TabLe L3. Studies on liver tumours found in rats given ?As and pLant materials are sumarized in Table 14. AL1 experimentaL animal studies, lrith the exception of one on chickens (Canpbell, 1956) and one on Syrian golden hamsters (Fushirni et a1., 1978), have been carried out on rats. The liver is the most connnon organ involved in experimental studies. Trmours produced are mostly of epitheLiaL origin, but a significant number are also vascuLar. Lack of precision and diversity of terns used to describe similar or identical tumoura makes it difficult to compare the types of carcinogenic effect in different studies. Some terms have been used interchangeably, e.g., hepatomas, hepatocellular carcinomas, haemangiogenic and cholangiogenic tumoursl nodular hyperplasia, pre-neoplasma, neoplastic nodul-es, and hepatocel"lular tumours. In most studies, there are no supporting photomicrographs to draw any inference as to nhether the tumours were malignant. Difficulties in the interpretation of data have been comrented on by Schoental et al. (1954) and Mclean (1970). Lasiocarpine has produced the largest yield of tumours. In the studies of Svoboda & Reddy (1972), t6lL8 aninals surviving for more than 56 weeks after receiving ip multidoses of lasiocarpine developed nalignant tumours of the liver. Of these, 10 aninals had more than one tumour. Continuous feeding of rats on a regimen containing lasiocarpine resulted in all animaLs (24124) developing tumours (NcI, 1978). In one study, a single oral administration of retrorsine (Schoental & Bensted, 1963) to weanling rats resulted in 7 of the 29 animals that survived for more than one year developing 11 tumours of a wide variety, at least of 5 which were maLignant. It is of note that this PA is known to have caused two caaes of human toxicity together with riddelline in two cases, though the total intake was proportionately lower (Stillman et aI., L977; Fox et al., 1978,. Ituxtable, 1980) (Table 15). Tumours produced covered a very wide range in unrelated tissues and organs, for example, the pancreas, urinary bladder, pituitary, bone, retro-peritoneal tissues, and skin, among others. IlepatoceJ.Lular carcinoma and haemangiosarcoma vere the most cormon. -L32-

o I o o I !o ! q ci !o ov 6 o o @q @ o rc c.rb0 !dC O'i O \'N o k o >i dNF I \\ t 6'3g -+qo E O h.r t.r t G FAOO

I @ 60 q to ooo I €a! dda oo od o eaF k9 c o> ooo !!! >rod 666 EO aoo ! 5P Ghq ooo 6{G sEs d o s !> s oh r,c rt .d OE .i9 .d! ts oo !6 96 oq3.d co Eo s aoo! t€ 9! o o oJa o qt0 t. io33& o d .-l! ! ! ! p Ead606 CF \t oF o o 5 o oo OO E € € E q .rE gJdq E 6 !9 6r! ESgo i i E i d !o BBog s s P 3k qd E E O Cl e. A9 d dv U a A N t Jd o 0 to ca I o o o.d E -\ q 'd^ 6 ! k! 'dNil t x a crq d6 E O a! d '!F3O O ..d 9€ EOHd o! ovs B.{' ts^ o o oo d 9F O O O O d! >. kJ Lg ! >rkg do0 E.d ! I o I O !p! O HC td.ig,.dord.io I F o@oF. P; E <{. <.€ <.3{t g d .r oo .i E so uo oB @o o o o l€ E J6 0 h h O x ! ca d 6 o o o hdo a! d o 5 E€ o0 .dd or oE c o i ?o ?59,o o ! co(dts 0v q oo oo oqoo 6 c o € .c u -o u o o 6 6 0 0 E -o o o h 9d -o.d o O -d - O a dd Od> Od Od Ot t c Hqd .l 0od HEEq .9E'Etg.E5 .eg ! Q.O F F O Eq E 0 e€ -.5 oo Eq 5 q E F q !5'oE .ddN O Q \t 6gO6 Bvi odv d<- €H o\t qts

! o o I o o EO .iE

o Fg k O'd o h! !o ts a oo E ifl EH -133-

o q ns 6 !o !t Qd tr o^ Ov o!o o o 4€ 5. 3 C6 I Oi @6 o lov

hhn6nn \6NNdNNN dndNNd \ \ \ \

€ o tsC oto o Qd 6 d 6 { ! 6 hEu e E E E E E I d O o >\o 6 6 q 6 q q O Od E E a ts s A A .d E ll E ! .l 5.d O E 5 !r d.r 9.a I 6 Edg.Co EiO Oq Ca Oo * 9 e. qd ts ! 6 F E t F E k d k q q o o o 6.d ! { - t o o a! o d E t ,t ,Cogto E, ra Hs gg

6 d |C, E !6 60 qo dE

! c o E

ttk !OFO.. G!tqo€ € k o hd o E NAE € H OO E! d .i6.r oE o o!\>od6.i o6d.ro\o! S.dO >il$5 o .i*?,.!x- ig"o. q >!'H. uO{! '-cj'dtE*93aoNd .1o -Ja '-"P'6E€\00 eor3ao.rlo.d3 d.r tr 6 d !< d ko @u'doi9e5 o 0!6X Eo.dJ'aOix6@€ .rOqgtsOX.dOd.iQqO o.-F!qo!ot oOOo

-o o ooo O O d or E I od O; d d !.r d ts 6 E o oi.! E ao E Ee i t f -o Ots O.r O 66 9vF ABBhv

o o o

k o 9o !c F oo d OO a F Hv -L34-

€ o o d !€ a q O^ ! I o o9 g 0 tr6 oFv

lc .rg l6666o00o 5 i'r 6 \\\\ 9o?E i-d6dddF f .r t c Erood

EE5 'E. E r.E q OO " " ko .c'$! H .Ei E * E o6 .5'EHf,"'d.t 99 FE! 3 S E=} 9 H5 o! gE !q, FT ?i; EE! =-isE EE aEftq t !> OH ri OE ! oo sd roo!

Oq e.E Po

0 o

O6 o

Ou d

o ! G@ !C o !qo o.:3 o s.d 5 o o c€ E O O\ 3 6 ? E9 e @o E >\ ./ &d6! s OO O opo!

E Fqd C !o o9.d9 .iooc o o s! E gq c! F X.d h 6 c p50CE oO oo @co6 AB- 66 ts O

o o o

d o ! l! o q o :6 ts &e -135-

€ s !t or tro^ O oo Ov

@Fv @d

on 64 NNooeooJ dd

NidNilil dil

5oE= .oEoo .i o o tr I or ho qg Q o S .JJ !x :: - 'd ! t { @e uool s dt ! ol X.E ?o E 6 6 E ro 6Jrd d= hF S rd P]iE 8.5. $ 3 - 3 C'=^ g;5 i.:99 oo3 J o q ! 693 ooc i:Eg : El.l t! 6 ) t i r 5! 6 Cj o ;E oq ed.- o 6 H:Y !o 9, al ao.; 6 oo 3"5'X .: .: .s oov d o , i 6ov O.i o = s f ! qi

AE AS o 'lu.:! o co !a ! s tro d o

h c6 o! .J dO s oo tso .ds 'd ud o€ ! o olu do s F a5

Itl doo od ts oo > ts o u o 6 ! E ! o sE E!@vO O C! t@ .i J ! ! .r | \.i g\^qdo o'@Q o'iq)96ao o,doE3 A Toi'-."!H tof,Ro - c9o'ixFo ca o! o .r !!,6 I 6 S O .i d ! 6 Q [email protected]

o t^ d@ F oC xd s;; a.r o .i i ! trn oi EqEts.rd o G! dtr t dE o o o/ E !C E3 !!!!c cs F o F o o6 qB- F 6> d k O.i or av dv fi!dB Fv

d ts OO o qr o dl! 96 q o slF o X€ tso tsd ! @ clo qo o €.rl! qo x o olq q>' o o6 .i dv o El.i F E @

o o r 0 ots €o !o o ov ! Av o o hcl F. !o o 6E od !(()

tq d.i @ .rtr o.d o dN I , E'd d 9.@?E a.Aad Hoo6 d

a od qd € .r tdil .id O !o! o!!o!o! @o ooE Ed6 d 0d o06 6 to oo .40F O E O E u k! o Hho klo kso o oo 906 9OE 9Oq 9OE o qco trdo Eno ! fF d o€ q.d ld€ do€ oo€ o HA O o oo o.i 6 q'd 6 I I o d €^tsH E> o€od EO ox d5.d o iE sE o d!.iE .d 0g .dr.l!!! .ducqx .t!E F @O q q,o.t I O O 6 ! A OE !o!d o c o aq u Eio d !d E O.rO aoo! ,!v o 5!v 5 g td

o ! EE:

.99 ! d6 HO Au

! o F! oo !o uk O! o v9 \9o o I SO !r P 69 e ! o ! kQ u I I 'i o@ € hE E E€8tjg o O .i O'ddq q o ood q .ioo x | 9c o o ri O'iOF d o c! E e dFa od N d E{ iluNh€€ n 5 E! @ Nr{o t9 95.r F oEtO o O'ir.i .d @ s Q i{ O EOAk d i B€ o@o0 9r o

d d t) I;.E ? oti o E tsqd " .r od o t o'9E oo6 tl d !1 d90 6 6a d [ € x.F Eg a E! u r9l @ o 600 @ Eo( Nv € Eer ! o o tl ! u.il k oA{ o o * .gl .il . s! ql€l s3t" :l a il r *lglg .41El orjl o k 36lEl o E€.5tgt8 > Errt u t! o 9lElo 6 d .d o Ftsdlv 3*'91 ts E E o

F do € o ov I ^ O >rN HI t€F !0 oto o.c P Od EO 6dv

ooFF6COOO \\\\\\\\NN-t$dddd sH:-3d6€

6Q: d6 -8 do E o o | 3. tE tE fl g kg .98 o oE e I EEr o e at d qt | 6 O o6 d a 6 .4 q'c A Q r onc AA 6 6i H'! 'ie!u F'd o a !5.r E.i O E.d F gi gi'6 ! 3.=c ii P6 .:5c E 53 &g =3 6 t .s33 €3.?€ *F3 a o EO ts d t'i€ | | o !d !g o o tx ! dlo @F 'r No>l o F 0.!o N 6 6 r N Q oo t€3N "+€€ Ns€ €t

o trO t O do 9c, E€€ ! 60 dc 5 ot dEao h idE ,{, 'r!.q5 d6 ! o c!

€h o Eo !! C!!O9.O I a OE Sd O 00d o s!.d cJ.r .d!CEN.iltr6d ! I I O 05 !E.iF to.dE .iEOirOEO€N * E f, 5 5 * g; b t b.9<.*els6 o o c Eq o F cq O! N! h -d Cd d ! o dg O o E Cn o 6 d o N ne oE !J * ot !J ts o orgo dg 6!!{.|OO 6kD9OO .9 d.4.d o . o-.!.i.r o E'?1'dE3S6.98 oe o'€B oqolol .d.dFlBBoq!t

! o o .a ql^ -o -o oo k o o i @ od o od oE o G dd6 Hd rE o6e od E6 do qo tsEO Se ev E9 ! q O E6 06 oo c)o OCr O:o .iN ds 6 6 gvd tav

o o o

c c o o 0 o o 6 o ts 6 -138-

q d € F 0 €o q 6E (dv e00 !s o E tda C 6^ k Oh o Ed o Ed o 4€ o oF o aN OE 3 Po o -c >o q&>o o G- o Gi @9v O 9v

tg C.r S .dc o.d o o @€6c'OO ! A >i ddNddrh 5 d'J 6 N \\\\\\\ddN;ntd E € H.i a.a A A Hood e I 6r t d I E H'iA bt r I ! H O i.d E o a Q C d X ! oo 6 o g -F,5 .l=, F!o a 3;E H=. b qq i3'ib- d3 E?t i8 Et .Eg oo alo q!n qal .i96 'rrC GE 6s 'FE€ F!s gA iE g.:3 .?E S8i 3i !E _Hs I o t> .o h .d oE I OO o q -o.r aoo!

o trc oo

66 FO 9k

od g@ !ts o!

o o o GU oX kc o tro! o o oo o oooc .$t8t .dOo o os g og u q E! E d9t03 notoE d @o J @ @o jl @ , e€ s tr'-o\./ q'aO\.r otEo .iEouo gd6! 'd g'g 6 .:3 o.d d E B u q O .O { o o o ri oo E t d ! oil od ol .d oOd 3E 5H,3 rg s! E 3E ! u! | Eo do ts E X.il ol 6 ch !otr..l q Eg I tOC IN oN o.-das€ @ cod Ov ABd!vo o

o o o o I o ots o ts o d O^ OE 9H X o H. ots o o k ! oo o OO olE ts E Jv -139-

OF o od Ev o

ild

l6 60 lHd olH \ o.l -i €d 6l ktts FF (,lN Nd l*; ON o d! OH Ei t 5 do oq o oi O.d i O o9 ! oo .dd o E d 06 uo 0 0 E,* !E 6 50 o ko H d ! F.d I oa: oc I >O s6€ g3 g f g'i o .r€ .'.!68 .46 9 @!o E d6 -E U F 9 o 3 -r .d.d € srgg.8o !FC

P 6 o

o € ! c.- oth'* F o!o €>rou o!€o o- q Ei E d.i F q E 6 Ot!.dl OO@ d s5p ! c@q .d cH^ E/.d ! !! g9 tl, sq O I o.-c o.i.i!r o.i .no.ro'93oOvFFEUE O >rji F dq6.d riloo.ipB€ € H o Co ! o O O TE 5E O ts }! Pq a5 xdd k o @ aq E chi qs H ! d c 6rd c x qEoq!dd Od t OE O.i 6.i o99 o 0 ! oo UEF o o e3 gt E h NOF6kFqCOAd'/pS N 5Xi O O NO n!HiaqotHoo'd6o -d.o o € O F E $E

o o -o .. -o t'" o@ Hoo qoo !H O E H Od Ftd dO d dE dd odd EH ! ad o Ad E H E E ou 5eo too H6 *tE* o g Fqs okN6 Hr o qo 49di g|i@ qF@

! g o ori 3l3l*? o d $IEIlI H 3l-'t" -140-

o € o 0 o ts o &o E^ o ! oF 3so OF o o oi E6 zdv

tc g.r N .dg o'i o N F o >r $ ? c.i 6 9 0> E \' a.d 3 C tsood

d6 OE OE oo ao tso ao -a a 9€ qro FO !ra o ft HA !id o & E> o! o oc x 4E 2 o u*,3 .g NO !d c AoO! r3 E Oh o ' Ev e

65 ':1 E q6 FO EX

o k.J Od to

c o . .$ o d@ .: -f i.fi !F 'rejeffgs.rs.l - iE<.:6 6 * ;^- :.33 6 'i:l;E;6xn gE?,^ ! e .;p :l ce E -€.'f&r $ ;;cE'i 3E€ S E60ts g:EfislesifuE?i i gE.gi* s 6 ts !o q @ ..dood ) xo @o E 6 F r.E 3 FE x.: ! 90 ;egE oov aB 3 $

o ! o o q o 6 o ! o o k o F o u o 6 E H o o -141 -

d Ea ON ,CO @v

\\\JilOO \€ao \ \ iddo idt

6 dal e oF od 6 9i !l k.r, .ra h a .drtl .ol d 6!o e o F ts ! o o 6 E d'i E sad 6d d 6oE E >rFo auA o E u e o t o E o F od o I ! FUO ./6 06 FrO ! q€r .dH o a o c a a 9.4 I o c 6 Es.i ! 9lr CdO !O ! qiO O .dOA o:d S oE .d€ .dq O @t O Fn 6 O q O A.r d ad q O O.J 6 E AS O !'d O

@ o €@ AE O I O .C ! .i! i C,.d€ !6u !6 rOlC .A c 66 6 FO fEVE NE 5€bi€e39rE

I ol tg .o I I c h o o.d I o'a fl 3 3.9'Es &3€ -E b:€ 5 66O -Otqi€q trhD^qPO6.d co4^q € o6.i € oqt ! rrv u oN s P !, o9!! d o 6 ! oE ! o o a o oJ o t c I ooJs @OliCt o€\.d-€o'r! o9\.d -a6Q.AP .d.rq90aE6!o .il.il@OOdO6!o tr E E B.r 6 o s.d cEErrdoEo!'r >rt xx EO o! t P" E 6 .o O.d ^ o .9 > O I H s @ .:€ s o JE I !> ! E! \.-oirh \'.oo O!uJ E!S o so o Peia P3.t; O.rrB b ooo 3'd A{fi 3'6 te o 3Fo N B Fd 6 N F h!

o o a @ d I a € d N

o E. FE o 9c "1 0 o :tq F -L42-

o !9 o a OF o o la od ts Oh Cv OF 50 o @v lCo

tg c.r s .d6 O.r O NNNN H O >d N€iddiO t Ei.r 6 \\\\\\ qo>E do@dddd ).A A d HOOaS g9 lonololtE >r h O.r d - d d O.d F Ho .id OOI tsE E .dd oo qE 6E ooq od ! d6 !6 so HO ! F ! E 65 E Orc o; ai .d O .r O d u F O ! J.d 5d ! ql s g€ vE ut EQ !rE ldq r.i ogo {.lo q0 qdk da 5F 'iE '6.8 erq 2o >o€ HA .E5 5E .333 i3.'i-86 & oo E> eqhe OH ESE€ !d o6 | 6! | m! I OO .E .E o o os!.i P{OO! 3RE 3Fg 3 3

o ! o o oo h .dA a a q6 FO tk o A! o I

to€ 60 o C !.o .r6 ! F,E !N! o6 o ox s .co o t\.d !k .E.E o! Hg

o .d F ! !h o E ! -dE .d 600 O Oo5, ! KE ! E .r'i ! o u I € dN 6 I S €a € ".:3 6 !6 0 d dg Et E k .i$ O 9l E rduA 6€? s o o Nt.gE X E dq H .d F Oq ! O ^.3 6 ^o{ o oo ooE -o o d o ! oi o E !qd o od dd d F OE .d oo6 do f,,4-e ! !E "E5-€q 'I F F H Eq Oo o€ io ;:g'H N Ov

!a OF o oo od o o

I o9€€ I \\ NNNd

tq to o0, o66\ ho @ 516 ! E .i ! o Ed | 60 te d6 oE!9 o E od g .d iD c E ! o.d o t o ! 5.d o o d o o o o ooto o9 6r!, ldja ac E C dk p d!adHJor-o o o .c 66 O .co Ood oxo t! i c k ov E !o !oo dEd a6 oo doq 998 I ! o o oo € € ae dE

o6 oos9 hh!{ 6dac €!oo oo OONN 9€-d

hh AA 66 9! tB ?3 .1oq .d!66F! Jo, .r!6Q9!g !drEo! !ddEO! rG .tuc FA 'd!F &.s6EHE g"Et$tE

66N I Clh oG do o d€ ! .d ! hod xF .d o! o !xddq €o !trlao rt 9il o >,oq6 Qd H! € otoo.!o G otoq ! .-o! ! s! .n05!C a F o!sE€oEE trO'>EO .d o .rbEJ@oJUo .&'d .iB9o9eo'6 N I .f g'b'6 I E ..1 * $ c "o!$E3dE3@ otqBi

! o -o 6 ooo l'.o HO q a d od O H d HO o d oid 5 x 4 ae lde @oAqt rd E3 :E3 ilETE O@ Aa! ts6 Qh av€ oAOr ON sl fil Hl 5 o ql k ol t d @tts ! dl^ O o rl! ts o F o .dt! ! c o ot d >.d o d 313 f 3 o E{tv a F -L44-

i! OO^ o o ! rF o oo o Td ts En Av o OF E d6 = . o dd EA 4v od tg 'rEE.d S O'd O o oooooo ts o >d o o900€€ , C'd d \ \\\\\\ oo>E F d60-tod E€ F.d J.d ' C: HoOd

o Ed hdoot E a iC hct d 6 b ! o 9d c ! E F E .d \@ o oo c Eq9E- ea I do tE Ed d6 6 ko o a) abi,i 3?.: ..dB €E gt o! 6.5 E6 bE t! 66 ba.EsE Ei 3.=l Eit €i E6 €6 HO ea EeEgE e"3 3ai f 3 tE iE I "s33 €> o ok o ,r oEi ! ts @o c Oq-O.r o6 0rOOs dg

H o ! QO cd oo 9! !E AQ 66 FO EB 5! NN o! Ad

EF ! ol o 0 r q Fg! q o o c! ts! 'd! c o6 q d6 0 0 GO UK HBctE o! a5o€+ c o .o g 60i- 6g so0 FE !*ooE 5E b o.:3 6 lE.E; .iE33a q c€ E qOAO FroU.€ 'd ! q .d 6JzJ u I d q o\.d r€Er o.d o J o d o ooqt E66F ODE.O OOBEI

E !^ .d o do o o &0 td O d E !HT o o , xd doo6 FE sog I€ x.E e sd uSoc 63 H.53 DCOd ov oav

€ o o o o o d o O^ tsEl o ts o OC p EO o q F a =v -L45-

p^ 0F !ts .NOts .io o ov Cv 60 h. ts. Ql .di E6 ;dd

€l €l o o06 € .i Ndd \\\ o N €n6

s @ t6 od tq od o.i a !o ldd o @ FB*.9 dc 6C oo 6 d g E o! qE O.t 'iqa e !o o EO tsd o!o Fo 3 9oo oc 9' O E0q oc s h9.- HO Eo E oo I > o 5 d.d tl 6d o d F o .d€ EE A.d C !oo dd

o o h o -cd I 6 cp k c !! d oq o, N ao8 r e 2 ea

iE o ot AD a E o'i E € .dE ! N 4 dt $ 50

HO o O.d tr 0 go q pp .dI !o I o I o! o ! ai oN xddhio! E ,/ !c €0 €dBc 9C @ | O o tsg 0qxo o @ o o.d uo ft o o.o .o g .-uo o! E .iJ O O o !:\9 .d [email protected]€U!do "1N .d 6J ji oog Od ToBH* O\.d 'ddtB 5; B\.d ! k .F'Fo€ 9i ctgpg.eggss O.dS! !v O @

!^ o 6@ ^ I o o I o o.d .o o .:| d d o d ! 09 I q 5 >,E (d E 3E o H E O ao O o E ! 6 o 2 E) > qr EE^ 6o 5Sgg,3 EEt 5--3 I ovv Ud @ Av

o 66 oo ts ec >A -L46-

!N o or il o ^q Ae € h o OE |1. o! 'iil o a E 6 &&

IQ E.J d ol 0l oOOoCt FNNNN o o> = d6rd ?'A a C

o E! ts^ ts tsE c c6 dt o o Fo E"T3 ET".: EE o oo !o @x5J 0.c s60 2)'i, 9: O'i>s !d9.dooo.dao Fg ES3 E€;: E5 E I ol o 9k a ! 6: 3 oq!.: a ooii

Oq o 66 o .lo E 3k

o ts! OO

dS !q !oG .d!@ 'luuqt c'Ei I o c o\ ' Bo3"P F XoCO _d€: i, o .d o E6ci q o.io C E O !6

6^ oo !o oOO 0 !d 9ag oq o6 I.1EE EF FOH 6O i,aa E Od .rN OJ OvO kv € g

o Oo Htr Fd { 9?r o ts o tri .H 6 ts 13 $ av I -L47-

o@ ts .!o

0O6O60 oooo66

S9$d-NJJd NJ o? EO ELddO I 9 E rd | 6 E E Ed 5 O.d O.d O o Ag d .ri 6 .dd rd ! !O dqt 4Od ql .r UOd O a aE AE I CtE E E >rE!E tt ir Hoo Hdpo o o o69o oo 99 qo 9.ro arEoe E Jl { EEoo oE E Fh >!! totk >, ' E Oooh riD r ? 6 .d o d .r 6 Et d .c o I .4 o tr 6 ad d i9 dho dJcoo ! J d xEoo a,4 oo h>, ci! E€ Oo 66 s{

o o q e

€ o $ s

9d t.5 gJ Xodg $ dq o

o o o g @i

.rlElBst.:th o 4toto stSiE o l.-iv H AI -148-

o oo o ts .=o o F o o aa & h6

IE q.i 4 .rc o.d o € € ! o >r N N , Ei.r d orl 0q dl 04 90>E g ts! x.d )'e a E HoAa! Eg lo o o q .:'i g oo - €6 'd6 oE 6 HO E'E E'-g ,P o€ *E j :^ E E ..8 EO Ee ' rsE>o€ E.3ok P g " 96 t! t€ q .d q c F d 5 0 0 5€ FA -o d dEo !a d i Ct d6 oooa & hx €> o! rd Essg OE k ao Qooo Oqt.i o9o-6 tuoo! s.n.c-t

o 9C o o oo oo >r 'r 'dE o q 6d t I !! 6c tso 3F @ @ €€ au s s ss

o ts! od

!! O!

o H ! o du ol !g 6o €o o6oc .dOO !q 'r!ot.9 ooo O EO E " .ngja c €€ ! E er o 5 E€ U od .dC t qdtro N od .t Nts9 Edd! { dq c s9 0

o ooo OO o .iooE 3t -13t o "io6 HqO" "i E !qi EdE o;E \ d E oH E .dood o.i q o ! d o a9 e @ s Fs ! eq ts E X.r 'if,oo !aoc 8-o 6'4ooFsEn Es3ooc, @cod @ ONN ONN OFNN OdN

0 o o o o o !

6 o h o d 6 ts t -L49-

g^ o@ I .IOF OF o9 €v o

,d h6 E:d

ooo60G6 dilHdddd

N nddnddd Hd

o 60t qcgd!,oE03 dE9 ddd o o 't? 'r? 6oEpC E Ei.d o o.r a H EO n.$i69r hbEE€.i O 5 k OE ddrnri o ! ouE60 ! oC El6.d Cl€g uo oc! a€:r D6 H H O OF ?O .dErr .aOh 6E >O .ddo de EOt|ro o 'i€ FPO XdO ts5 .d! !rd !5 a o E! o E trt| t f tPo v! i6 5!o v! oo I >h .t o6 k !E rto *.q t .d ! E€ .ii oo sr! A EE aao co!t s.+ g€ b"e tt o!

o oo o h >r >' c 6A o €9 'o o oo o @@ s .+ -s a

o I o6 ! u Q! o o o .dO .do p €o OF EO .do qai q6 9t oo oo Qd *i dN N €q O oo c 06 od o O.. oo o O d .O N.O o! or o qi h . @o oo6 H a 6 u . H Od Hod F!dE o-E qd@ d6 dE .xH sn iE,fl ",f^iEE $E,s StOF 8-- U-638-* 3oo ui o E{filj F a"- -150-

0 o !6 ots o o:i F Cv o e !ld

It tr.d d .d9 o.i @ ddiiitdd ko>a NNNNNNNN 5 C.r d \\\\\\\ qOPE rdNddddF t.r t c tsood

f d o s O.d 6 6 I 6 h ! o .i- E A .il.d F ,d^ oo d @o d xk o h H o u E o 6 d € 3 0 to a cE g FOd tsoc so -JE ua H tso 6!o 6Ed dg'i 50 Eo .do @) oc Eoo E!.i EEo ajd oF )A !O EE > o oE k .J d o .i d k e, d o ! o d E .d€ 6c6 kd6 Fi6 ha od .r€ tst tso d 6 ! o o 5! a 5! 0 di o d aq v! I I o €> v ok i50 .i oc .T ! E€ k oo !d c oq5.d aoo! E€ 8.8

q o ! oo d urf ed tso I

F! od !o O!

o U d@ o !F to oood .ioo qd qI E€ E oo

, Eit S N a€ E o o@ E6A!

o d H oF.! HoE E Fqd .d o o-d ado EO E o !-o; 4CoG

o ! o o

o o ts o s 6 F - 151 -

6NOdNNNNO NNdd-l,-++\+N H6NOgdN90 \\\\ idddd

SE o!q! o E .d^ @ A r o .r^ o q q hts o I o h! 9 n d { } 6 E E E tod K E FoE doE ts HO ts o o€d 4 ) h o dt.i c.n 6 @ a ts o ot o tr trE.d P o o tr tr€ o op d ! o o c >o >O .rda dE >O .dS! FHO dE >O € ./! raid Ht !r9 Hio Eo€ ts' .iE "1 dS 9Of v! d O 5OO dOd v! d 6 I t. h g rEo i! o .d 9 EE .J ! EE us c: !@ E FO!5 9E op 5€ O,o

o d glo o ht 6

O! @ €5

o d! o d o! p € o o! oo o o 5 @ oo o 50 0 .do .r9 0.n@ .d I .iJ o.d €> €3 € a sld E o o @€ 6 Fd q qO !Q t"9 OF ou @6 0 o! oo6 o ou o\ o NO NO N o Xsq NOd>ro €d @d { hoF s N ts€3O

..OO o H .O r. -O H oil > Od -.9 *,9 >od >o6 E E dE ?sn 3s3 G9 0 3.E$ F$-+ tsNN tss$ !*. odd (,Ji ONN oHd

o Ft 6 5l d ! I'r hlo o Htq d hto F @l -L52-

o !6 ON o o= ts Ev a o o

tct tr.d @ .dE o.d o $s\t.+.+OOOO ts o >d NNNNNOO-- I C.r d \\\\\\\\ 90>E €.+ddF€6Hd E! F.i A'd f .il, c Food

.[$5 .$EEq .d^f o .io5 b E oo B _1" FO :! . E . Hs gE eie E€ eE*E !,Ee"HB rro H.E E.IEe,.t 3E bt UE3'53'i5'5 ?o >oF ots Hko CE >o >Oi .i€jJiD iDki Fq :E 5.Es €3 €38 J,? 53 ::g EE3 t e 9> o! oc I @O Oq!.d aoou

o d 9d o oo o o I !9d ! p! 0

o tsi Od 5O

s^"1 !l r ,6 Od o o d@ !N O UE ! O h O.dN E - !1-t .d!Ed.r!6.i cjEiic !OEr.l€d. .dOO O Et E qH6E-Eo6 o8 "3E+3gfg.e Nts R_E9 X0 N O 6 o 6 i.d | 6 FdS! dk!@!sBo> oa

H6 H -O H od >o9 >od c !qd .iloOd oco o!5 qn E 5 Fq o- t Eq u5od 8^- qc66 odi (,dr

! a

o o

o EEL hl ol! o ts gl.il! o a ts E t|ElS -153-

OF - o od o.c Cv E do o o qts !o El ov

o oooo OdNN \ d d d d d \ \ \ o-odd

od 6 d aa E o E d 33 I t 6 ! .ra E C ! n o t! a .'! td !o >!tsE d6 { h€ aE ! rqE ! (!6r i.oi !odE i.'d l 66 E .rit ooF o5 dsd todO .!o O!! !9 9{ 'aE) n6 do;e Oiod gi i.rd oE ilz !oQ .;6it 'o.bOO !kts >q EEEiE Ei .d! !t! 9..1F O!5 i;5 tsdt ,;d-€ 6;r ooo >ro 9qt O6s :tt 5Da 6qo6 Elq hq! d6 EA e.d!

.dl

O/>r F C: 66HO ' €€oE!

I o o

6J4 JO o!, o3 I N \i. 6

c -; o @!d ./XtX u\ s9 o 39"'593 B Oh dN o il o .rN >.9 EJI o o.* €'E -g3 .r o! u ! B

@o o o

OEE d9 FFO FF c6 o€6 ON ov€

o o q

I o E d o ts -t54-

o !c, o o@ o 0 I kv o s .c. o !d & v6

IC c.d o .dA O'd O oooooNNNo H o >i NdNNNidid , C.d d \\\\\\\ QO>E iodddN9oi E€ K.d ,.A t I HOOd ! rt P lo d s oE dg I o u, I o F.r t O o C! O .i dgo >' .d @ d I rd .d o ! i.d F ! ao d @d E.d 6 0 | d so o o o 5s 5 6 q to e a E .nE Eo oqt lE .iqto oo! uF k O k 6 aj O it O 3 .il.d t{ d b lid .dH .d I OC oEO dl Fo s! OEO df t !Od AA ;x > O > O ! O 3 O o o.d > O H AOE old ! o t! Fro :x 5E$ E.e',i5 gts :i3 .ggE 9.53'd.€ I I o E> ot 'ilt o 'J oC O!gE€ o F OO atr d C oh oq!.il h6ts oF 3 6q aoo! 9€O€O! +E

o p oo .rE d6 tso s

a @J JO ts,8 ao o6 9F EO !k N O! {6 FH o oo g! u I *.iii d@ .ilgt! !\@oO FF 0 s.i oo @^ @^ Oood GN tr * .Joo .gEgB6* .il6 .rl 6 O E€ E FN o d .dNhOEg i8 i 3 H 3€ES .do3!!B.* 3'd.33 to o 860ts Ev !v

d o - O .@ -o oo O O o0 Od o ! o i ! od OE E !qi o6 .ioOd d [E g do ao E TE H EH HqHq !3oc IR oa o I oo @dod N

I ! o ! o o o

o F o o s d ts @ -155-

.: uo d o@-6 ! sd o^ HV N 6 0@ po E.'ad od g d Elv

-it??,: -o -O -o .o g U so .Et q3{ eq! c8i :3cO eq; s s '' ;EE ;Efr ;EE -t ;Ef ;EEh6 dil 6 F ^ vil vQ vN vo

,3! od G ? G ? ? i.:g d, lt_ 3. t & .lEE l.gB E:E ni' sb rb *8f;"H r! o!i"* srTt!€ [H€;ss res ; eates iE tes

@ o J' +o \to io i9 dB

o o o o o o I F s

I o tsh Fo h U@dq H !$EN F ooo @oo dq! -r do p @o ! 6 ds o d @! @ Foj@qt k oJ u9jj ! o\'d F O p o\.i tr O d o @ o'd O d o so.i o .i€ E B! B .rtE3!l'

o doo -o oo F Od o oi OE i >,d E oEE so E 0 do !FO Eq ! 'o ca ooNo oNo o AvN NN 9vN

! o o

t€ o ;c .io itrE 6 EIHIH lr gv oool -156-

96 Or Ev o

'ro

o.d o 9qo6Oo6 o >d 3-NNNNNd E.d ct o>E ;i; t H.r .i5l ooql

ld oct I o o E k.d 6 .dO o r d o g o sq E ,"' o a hbl k do i3Etr.i o ttqE o E otH .riln F 5 d b.t,E I o Ea bc 9' s E: qt o o cd.d ! o o o c o .r> ok ?o Hko H6! .d6E 6r.lEi oOi t c6 99! !€f' !nq oc3 rEid d.co :E soO dAs t'oE !.ral .rd.l I 5

k oc dpo6 oo dd rd o Fq oo9 !€oo

@ o a! I 6 o F o ! s

CI o k.f od EO !H

! 660 KE N o N N. .:36 d co E ! u €?r o o 6N! ? a oO od 9rO s.q odE !i 6 ood ts EO !E sE5 Hh E5.E ois (rd-t ca6 {vH

o a

q ts fl.#t -L57-

OO o3 Ed Ov !

ooo60\t-t.+ NNdddNNN 6di600

| ts.I I d o 6 dP o i tr O t Q e a.A .l O O @ / 6 s4 d dsr{ >r!6 =ii ood 6o l6d . qE_ E .qo F tsod oots ce E oeE dE3b.E 6'ig g€i'ib8 bE3 b3 6'i3 2O!->O kkd .r6a oug >o! >o tsho ..!sq.d€ tO.g tsdd oE, .doO .rp tOt i'codd dga! 5DA !.du tE o dd dod

d! J6 d6 .do !El !g

>h >h kd Fd 6 aa 5€ >, @06 6 oo! 3* Px o € 3" P" S s

N N NNN €{r o

!!! ooo .d .i o

o60 oi or' or O6 O6. Od dFACaC oE do Eg Eg EJ EH

{$ NN NN

o t^ qli la ! ul; lo F ol Ela ! stc lr t: olo tv k al -158-

o o o@ 6 o od I tv a o o !ld

tc C.d @ .dQ .+.i.f-t6 .NNNNS , c.d d I o? E FNo 2.a21 tsood

td6 E E T O 6! tr o r0 A.d".I oo .l ts ! d O kq !tc '!bt'533;f 3E -3 3Et o! F35 39'5gE EO ouA !O oO f! !{: '1? koi FO ;€.4 3.5; 3',i 3EE t €> o! .d oi b *,3.9 QOO!

o OO .dE 6 O6 I tso tk O! I

od 50 Os

s@ FF o o aoc .noa, o c! E t d E!q @ Q€ o d E ds!

a oo

.roodI !qd o9E HtrO c6 qdad.j 5 6'E ON-+ 9vN

! o o 3l o El .el€le o ts o a a jlEl5 FI u -159-

I o6 €o OE ^q OE 9v ,o o o €o4 a oo F ;86 i t' E E OO O .d.t > a 69 6E oo 6 q'c \t -t -gE8- od El 6oo E !tl a9 0 r 5! @ ooco o E o.r 6 5g@ gu0E al 9 >! qq d.l k O H 5po 6d 00ts !g .il!p .do g !oQ Ad i'da @!E !O €o o Ei, .i€ tS.9 !.tp ad sEo9 orl ! o h6 0 o€co ! !'i "bgEo H\H Q o>r a s. €do !o !5 d kPElE. ot d qli 6 EO P 'E H oq >r : O h@ od @o o 0.ts Fd !!!od NO 9Frd o H ok - l! u dtso d o aqr.r NO ro2€ NO € u o.n o o ds so o o€ o , ! o q.d .l! !E U oassF .il8 Jto d >.d d .iEE 6 0€.r o tH o .tgc 6 60 d E'g'"5 ! / o.d E {dd .n *.Efl EE E6 . i okdt6 .ilEO 6 d c.i! E F@k dcip o @ ds9 g*3gg oqH.r 60Eao Elo d > t4 t o o q d.d O .d C EU H oQO> t o6 dE !i t o c os 6 ooo o t ! cu C o'd Q o !oo t q ! oi a ! od vq q oeE4.d 6 C E .! o 6qiHo d o k.d k o o p o! ! E€ ' 3 O d.i d o ! €!ooo o c o ei e d t !v€ @O odo5d 5 .eotoE o .d! E E !oo .i Nts 9O odg 5S tr o6!6.O T E3 o Eo il EC.53P F6 ! d o! o d o oN.+ O.i 6 E Bd JEt'€C 3 -dd o.iF o itiD OvN € ooq6 od oud o o.G"c 6d > a3 k!! o .d EWo t o.r @t o s '..:€EJ o.d Bd ! E > o Ed"EcS' 3 3.3E E ! o o.d.d d tsil d o.d o o.cE o! a9.Eilr o30 c!0 .ia C'i! o o 00 5 E9

.-o F 6l : i 9 . 3n{F. Sn$F h :r3.:-.:;'E:^R9333 >-.i^ 6 I -:E -:E 3 9d o d d d ol : d . pv o 'f o !l c ; .E .E e E >\ O O€ Od o o €E o : : e nd r- t I E € o qf H * H ; El rI r o E 3: 3 I sg 8g , o o a. > 6 6' ts & x q& z E 33 83 E

o o I F o I EE E {!i 0 .. !'i i EeE o 9t Ei : ! i s -.; EE; 9!o o o p€ ior ! F ?{ -t:; !k H: iE! .? E = . F :$B' .q.i !E; t d ea H- E E$ !i6 65 g H HE iEg gE E E: sHE o F u o HE i.i.! 3. Hs e F.i Etg o 3g tHgE j{ # J€ s€f ! 2 N€ I-- X -6 \+ oj 3:l

-o>l -d6 3. N 6o ts! 4 o ts @ d f.q 2o 6 6 qc or Q g3 o o c{ s.i d 'i @ t ^r ? i i ! E€; b .?b b 5 g g 3

lq H lo ! =*E o bl! $.iI b.:I s.:I B.:I ,.rf.. q lo h-o qa i-6 Fdo ld rda rdo i:HAh o E E$f nf E$JE$f E$f E$#E$rEH o l" .: g o o eE N g .: .5 .: g € .i' .5'i .d o t E ! ! ! 3s E N F E F F F5 o n L o 9 + 5 ,i ; !.! o z r.ss8s-si3

\' A I o I q.HE o o 9o -. eo s'Fi'H.H!H! ':{aouq I 9aa o H x9k.r! ! '-9sc. Hg '1Y O'd !qo YVOoO olF .=u8gi u oq o r6Aoo Ec.i @

ES3^ E E 3 8 I aa : :: j s 3 gg ttf;{*;.j.j u!o6 :loitt!! 990^--oo .i.i!sEFsEEE.;3I9oo =lEiE'+€.:.I HEEE

o vo 5E '5t66 o EE 6 3 EEs s F * FtE'FEdi; c r- i ; 3 i 3 .3 E+F .-gtB=H F $ H e .: lre .i EE E n.!i E HH sH Hss S E'$ ft'r"Ec,

a Er E$E{EE-b EdP g Eg -€ -J{F ;H ;HE€E s.l =:I; ;: ::;:

F N N € c'l \' s n o N N

.? ?T g T i T 8 c.:! i,! rqb rob ,,b + F !$ri€ lBt cBfEB€ gfiFi ,.H.H6??E .=83.1.t'lP'E s g F s g H *r I € I o r gt13 E tgo '' "," El-: .sl $l o ! o 0 Blj Fb tl- ! s itFs itg -jr il H .s tE ;ill,fl E t E ilfI {Hd€ ;lth F F 5 5 stsr iE gE jlEl fltti iIfl 6l -162-

Crude extracts of plants or whole plants have a!.so been demonstrated to produce a variety of tumours. For exampLe, Senecio longilobus has been shown to produce tumours in a ;idiffi;nmrce proportion of experinental animals (Itarris & Chen, 1970). Hirono et al. (L973, L976, 1.978, L979b) demonstrated the carcinogenic properties of a number of plants, used as food in Japan, in the rat. A sumary of the relevant experimental data follows.

6.4.8.1. Purif ied alkaloids

Kuhara et aL. (1980) administered clivorine in the drinking-water at a concentration of 0.05 g/Litre to 12 rats of both sexes, continuously for 340 days followed by plain ltater. There were 20 control rats. A11 the treated rats survived 440 days. Eight of the L2 animaLs developed liver tumours including 2 haemangioendothelial sarcomas and 6 that were described as ilneoplastic nodulestr. No liver tumours lrere seen in the control animals (Table 13). Schoental (1975) tested the carcinogenic properties of heliotrine, With and without prior administration of nicotinamide. .Nicotinamide protects against Liver necrosis and so may enhance tumour yeild, as shown by Rakietin et al.. ( 1971), who studied pancreatic tumours in rats given s treptozootoc in. Ileliotrine rras administered intragastricalLy to 4 groups of 26 nal"e weanling rats in L or 2 doses of 230 ng/kg and 300 mg/kg body weight; 2 groups also received nicotinarnide at 350 - 500 mg/kg body weight, administered ip 10 - 15 min before, and 2.5 h after adninistration of heliotrine, as per the dosing regimen shown in TabLe 13. There were 8 controLs. A11 animals administered the higher dose of heliotrine (300 ng/kg body weight) died within 5 months of rhe pA administration. The livers showed lesions characteristic of PA toxicity. No tumours vere seen. Only 1 out of 4 animals receiving heLiotrine alone, survived 27 rnonths and it shoriled an islet cell adenoma as ve11 as adenoma of pituitary, but so did 3 of the 8 controls. In the group receiving 230 mg heliotrine/kg and also treated with nicotinamide,4 animals died within 5.5 nonths (one with a fibrosarcoma and the other in a moribund condition), chiefly. fron toxic liver disease, and two more had to be killed. Of the 6 animale surviving more than 22 months after heliotrine treatment, islet cell adenoma was seen in 3 together with other tumours as shown in the table. This tumour is stated to be extremely rare in the animaL strain used. Ilepatoma vas seen in only one aninal. The role of nicotinamide in this studv is not clear. -163-

Svoboda & Reddy and their group have carried out 2 studiea using lasiocarpine (Svoboda & Reddy, 1972, t974; Rao & Reddy, 1978). Svoboda & Reddy (L972, L974) gave repeated ip injections to 25 rats at a dose of 7.8 mglkg body weight (0.1 LD5g) for 56 weeks as per regimen shown in Table 13. Three rats died of acute liver necrosis in the initial 4 weeks. Eighteen rats survived 56 weeks, by which time each animal had received an average cumulative dose of L25 mg lasiocarpine. Of these, 16 animals developed a variety of tnnours 60 - 76 weeks after the beginning of the study (Table 13). Ten of the 16 animaLs had more than 1 tumour. IlepatoceLlul,ar carcinoma was the most cofimon. The squamous celL carcinoma of the skin was found to be transplantabLe. When the sane PA, rnixed with the diet. at the rate of 50 ng/kg (Rao & Reddy, 1978) (Table 13), was admioistered to 20 rate for 55 weeks, L7 animals developed tumours. Angiosarcoma of the liver emerged as the most cotrmon tumour (9/20 aninals), even though hepatocelLul.ar carcinomas lrere also frequentLy seen (712O aninals); squamous cel1 carcinoma of the skin was not found, but there naa a malignant adnexal tuaour of the skin. The average cumulative dose of the alkaloid was estimated to be 190 - 200 mg per rat. Monocrotaline has been studied for its carcinogenic activity in rats by SchoentaL & Ilead (1955), Newberne & Rogers (1973), Allen et al. (L975), Shumaker et al. (1976), and llayashi et al. (L977), using different routes of administration. Atlen et al. (1975) studied the long-term effects on rats of repeated sc injections of monocrotaLine or its major detectabLe metaboLite, dehydroretronecine. Male Sprague DawLey rats lrere given biweekly injections of monocrotaline, at 5 mg/kg body weight (75 animaLs) for L2 months, or dehydroretronecine at 20 mglkg body weight (75 aninals) for 4 months followed by 10 rng/kg body weight for 8 months. Fifty control animals received phosphate buffer (Table 13). PartiaL hepatectomy was performed on L5 animals in each of the treated groups and 5 in the control group. They were observed for 10 months following cessation of the injections. Of the 60 animals surviving in each of the treated groups, those receiving monocrotaline showed rhabdomyosarcoma at the injection si.te (z anirnals), hepatocelLular carcinoma (Z anirnals), acute myelocytic leukaemia (Z aninraLs), and pulmonary adenoma (2 aninaLs). In the group receiving dehydroretronecine, 36 anirnals developed rhabdongrosarcomas, and 5 of these animals developed metastases. None of the controL group developed tumours. Tissues obtained from partial hepatectomies showed that both compounds caused inhibition of mitotic division in regenerating 1iver. The -L64- results of the study ilLu8trate the dual alkylatiag and antimitotic properties of these agents, comented on by Cul.venor et al. ( 1969) . In a sinilar study by Shumaker et al. (1976), rata ltere administered monocrotaline at 5 mg/kg body weight ou alternate weeks or its metabolite dehydroretronecine at 20 mg/tg body weight, on alternate weeks for 4 months, followed by a dose of 10 ng/kg on alternate weeks in the succeeding 8 nonths (Table 13). of the 60 rats receiving monocrotaline, L7 developed one or more tumours, but not until after the treatment lraa di.scontinued. The tine interval nas not stated. The most comon tumour was carcinoma of the lung (lL aninals) folLowed by hepatocellular carcinoma (5 aninals). A wide variety of other tumours was also seen. A notable feature !ta8 that the metabolite dehydroretronecine did not produce any tumours by systemic action, but onLy at the site of injection, wtrere significant numbers of rhabdonyosarcomaa (39160 animals) rrere seen. The marked difference in tumour sitee is explained by the fact that the parent alkaloid monocrotaline has to be metaboLized before it beco'mes a carcinogen. For this reasont the tumours are distributed in several organs of the body, whereas dehydroretronecine is itseLf careinogenic and so acts at the site of injection. VJhenmonocrotaline was delivered in a higher dose, but as a single subcutaneous injection (40 ng/kg body weight) (Ilayastri et a1. ' L977) to 40 rats there nere no matignant tumours but only adenomas of the islets in the pancreas In L6123 surviving animaLs. The results of the above studies indicate that monocrotaline is tumorigenic, but the type of tumour and the malignancy both depend on the route of adninistration and the dosage used. Hirono et al. (1977) studied petasitenine, the Pure alkaloid isolated from the flower stalke of the plant, Petasites japonicus Maxim, wtrich has been found to be @s (nirono et a1., 1973). Two groups of ACI rats of both sexes, Group I of 3 animals and Group II of Ll animals, vere given the aLkaloid in in the drinking-water' at concentrations of 0.5 glkg and 0.1 g/kg' respectively (Table 13). There vere 1.9 controls. A11 3 animals in group I died within 72 days showing marked hepatocellular damage; no tumours lrere seed. In Group II, L0 out of L1 rats survived for more than 160 days. Eight of the 10 animal"s deveLoped tumoura - liver cell adenomas (5) and haemangiosarcomas (5) (tabLe L0). Two animals had both types of tumours. The authors concluded that the carcinogenicity of the plant was due to petasitenine. Retrorsine and its N-oxide (isatidine) have been used in severaL studies. Sch6entaL et aL. (1954) administered retrorsine at a concentration of 3 rng/litre in the drinking- lrater to 14 rats and isatidine at concentrations of 5 ng/litre -165- folLowed by 3 mg/litre to 22 tats untitr death. Twenty-five rats lrere administered the alkaloid mixture from Senecio jacobaea Lin at a concentration of 5OO ng/litre fo115fr6d-Tl 566-T!-/lltre in the drinking-water. Dosing regimens are shown in Table 13. One group of 7 animals receiving isatidine received supplementary 0.57( choline in the drinking-water and another group of 5 animaLs received isatidine 2 mg as a single ip adninistration in 0.2 ml of tricapriLyn, followed by skin application of alkalids as 0.5% solution for 3 days/week. The animaLs receiving mixed aLkaloids of the pLant showed extensive Liver darnage followed by rnarked nodular hyperplasia; no tumours developed. The nodules were earlier interpreted ag hepatomas, but later only as an early stage in the progression fron hyperpLasia to neoplasia. The retrorsine group showed extensive liver damage associated with cirrhosis and nodular hyperplasia. In 4 rats, they lrere interpreted as hepatomas. In the isatidi.ne group, 10 out of the 22 rats developed hepatomas. Tumours wete present in 3 out of 7 animaLs receiving isatidine plus choline indicating that the latter had no protective roLe. One out of the 5 animals receiving the alkaloid ip and then through dernal appl.ication developed a tumour, which was also interpreted as a hepatoma. In another study (schoental & Bensted, L963), 95 weanling rats lrere administered a single dose of retrorsine at 30 rng/kg body weight, by stomaeh tube (Group II). one comparable group of 50 weanling rats (Group I) received 400 r radiation in animals surviving 100 days after retrorsine administration (3U50). A third group (Group III) of 6 aninals received 400 r radiation alone. Another group of 10 weanLing rats received the PA, 9 days after partiat hepatectony (Group IV) (Table 13). The additional treatments were given to study whether they wouLd act as co-carcinogens and induce neoplasia in hepatocytes, vhich are known to show injurious effects for long periods following PA treatment. In Croup I, 19 of the 50 animal.s receiving a single dose of the PA died before radiation couLd be given. Of the 3L remaining anirnals that received radiation after the PA, 25 survived 12 months. Among these, 19 tumours of a wide variety nere seen (Table L3). Of the 6 tumours in the f.iver, only 1 was maLignant, having metastasized. ltost of the other tumours nere malignant, including those of the breaat, one of which had aLso metastas ized. Ln Group Tl, 29 out of 95 anirnals that had received one dose of PA and no radiation survived for more than a year, with a mean survivaL time of 23 months. Arrong these, 7 animals developed tumours of a wide variety (Table 13). Five tumoura in the liver were benign. Most of the others were malignant. Tlro tumoura, a cystic tumour of the breast and a -166- carcinooa of the uterua, were preaent in one animal. Tumours seen in Groups III and IV, which rtere found in aninals surviving 12 months or more after the start of the study, are shown in the Table 13. The 2 tumours of the liver in Group IV were also benign. The number of.control animals, if any, were not indicated. The authors concLuded that the above studies did not provide definite evidence of synergistic action in the carcinogenicity of retrorsine by whote body radiation or partial hepatectomy. Ilirono et aL. (1979a) studied the carcinogenic properties of senkirkine extracted from the dried nil.led buds of Tussilago farfara (coltsfoot) and synphytine extracted from ilried niLled roots of Syna*rylum officinale (confrey). Both Tussilago farfara tniffi) and _gyrphvt* officinale (ttirono et a1., 1978) had earlier been demonstrated to-Ha:frilFrcinogenic properties. Sixty inbred ACL strain male rats ltere divided into 3 groups of 20 animals each and received repeated ip injections of eenkirkine at 22 nglkg body weight or synphytine at 13 mg/kg body weight as Per schedule given in Table 13. A11 animals treated with senkirkine survived 290 days. Nine out of 2O rats developed liver adenomas mostLy after 350 - 450 days from start of the study. Cirrhosis of liver wag frequently observed. Of the synphytine group' all animals survived 330 days after the start of the study. Three animaLs developed haemangioendothelial sarcoma and ole, liver adenoma. The sarcomas rilere noted at least 518 days after the start of the study. Schoental & Cavanagh (L972) used trilo alkaloids, retronecine and hydroxysenkirkine isolated from'dose@! Crotalaria Laburnifolia, whicli was' injected ip in single ffi-Id'dE-300 ng/kg body veight in 5 weanling Porton Wistar rats (Table 13). one animal that had received the 300 rng/kg dose developed astrocytoma of the brain after 14 1/2 nonths. Retronecine hydrochl.oride was adminietered in doses ranging fron 300 to 1000 ng/kg body weight by single subcutaneous injection to 10 newboro rats. One male rat, which rtas found to be parapl.egic 6.6 months after receiving a dose of 600 ng PA/kB, was also found to have ependymoblastoma of the spinal cord. Among the litter mates of this rat, 1 maLe that had received a dose of 1000 ng/kg died. The remaining 6 females and 2 males were killed within 22 months of being dosed. Of the femal"es, 5 had pituitary tumouts aod L had a nafinary tumour (type not stated). Two males did not show.any s ignif icant abnornalities. One group of 5, 6-month-o1d female rats, born to dams that had been fed on a diet containing 50 g dried and powdered Iteliotropium ramosissimun/kg diet, a tribal remedy, used -L67-

during pregnancy and parturition, were then. themselves fed on the same diet at 6 months of age, as per the regimen indicated in Table 13. One female rat was found to be paraplegic at 7 months of age. A tumour, possibLy of Schwann or satellite celL origin, lvas found.

6.4.8.2 Plant naterials

A number of plant materials have been tested for their carcinogenicity by adninistering either a mixture of extracted alkaLoids (Cook et a1., 1950; Schoental et al.,1954) ot, more often, the dried and nilled plant rnixed with the diet. The only study in which the plant alone was fed lras that of Canpbell (1956), who produced tumours in chickens by feeding dried and nilled Senecio jacobaea plant. It is notable that the plants tested are almost aLl those that have been reported to be used as herbal medicines atdlor food, some of which,have been reported to cause human toxicity. Schoental et al. (L970) used mixed PAs (intermedine and lycopsamine) extracted from seeds of Ansinckia interngdia, known to cause livestock losses in. the@ sten_s from HPLiotroeiup known to be.used by women 1upinun.L., in East Afr-ica as a herbat medicine, after childbirth. The dosing regimen aod mode of adrninistration are indicated in 'Of Table 13. the 15 nale weanLing rats that had received a single treatment with the Ansinckia PAs and survived for nore thai f year, 3 qats showeda-3d-a-denorna, one adenocarcinoma of the isLet celLs, and one adenoma of the exocrine pancreas, respectively (fable 13). In addition, one of the animal.s also deveLoped a pituitary adenoma.and papillary tumour of .the urinary bLadder. Eight aninals received the treatment with HeLiotropiun supinim L. (Table 13). One out of the 2 weanling rats fed on the pl-ant with the diet and one out of the 6 aninals, that received a single intragastric dose of the crude alkaLoid fraction devel.oped islet cell adenoma of the pancreas. The number of control animals, if any, used in the study is not stated. (1970) frarris & Chen tested the carcinogenicity of !9ggg!9 longilobus, which has been associated with cases of-Eman ffi(stillman et al., L977; Iluxtable, 1980; Fox et a1., 1978), HarLan rats were divided into 4 groups (equal numbers of both sexes) and fed diets containing dried and povdered stems and Leaves of the plant in the proportion of 5 - 7.5 glkg. ' The number of aninals in each group and the feeding regimen are shown in ,Table 13. Continuous feeding did not pro.duce any tuaoura, presumably because of the comparatively Low survival .rate of animals. , Significant results were obtained when the animals nere fed contatsinated diet alternating with normal diet (Group IV). Of the 100 aaimals -168- fed on this reginen for 54 weeke, 47 survived for more than 200 days. Seventeen of these aninals developed malignant trrmours in the livers - hepatoceltular carcinomas in 16 and an angiosarcoma in I aninat after a mininnrn feeding period of 2L7 days. In Group III, fed a contaminated diet fot I yeat, 23 rats lived for more than 200 days. Fout rats (3 nales and I fenale) developed hepatocellular carcinomas and one, a peritoneal mesothelioma, after a minirmrm feeding period of more than 428 days. The tumour-bearing animals lrere predoninantly male. The authors emphasized the relative rarity of liver tumours in the strain of animals used. Results demonstrated that S. longilobus is carcinogenic for rats. For a study of Schoental & Cavanagh (L972), using dried and powdered lleliotropium ramosissimum, see section 6.4.8.1. Hirono et at. tested the carcinogenic effects of 3 widely-used herbs containing PAs. Ilirono et al. (1973) studied the possible carcinogenic effects of the young flower stalks of -iilEffiE5?lPetasites japonicus Maxim, which has long been used in Japan remedy as welL as the PA, petasitenine, isolated from it. The flover stalks of the plant were dried, milled, and fed to young ACI rats of both sexes mixed with the basal diet in the proportion of 40 80 g/kg, as indicated in the feeding regimen in Table 13, until the animaLs were moribund or dead. One group of 27 rats was fed the pLant rnixed in the proportion of 4O glkg diet for 6 months foLLowed by 80 g/kg diet on alternate weeks. The second group of L9 rats was fed a contaminated diet GA glkg) continualLy. Three animals in Group I died of pneumonia. Eleven out of the remainitg 24 rats in this group developed liver tumours after 15 - 16 months of feeding. There were 2 hepatocellular carcinomas, 6 liver cel1 adenomas, and 3 haemangiosarcomas, of which 2 had netastasized. In Group I1., 2 animals died from non-tumorous causes. of the remaining 1.7, 8 developed haemangiosarconas, 4 liver cell adenonas and t had a hepatocelLular carcinona. The incidence of haemangiosarcoma was statistically higher in Group II. In a siurilar study on mice and hamsters (Fushini et al. (1978), groups comprising 20 - 24 male and 20 - 2L female 6-week-oLd ddN, Swiss, ar.d C57BL|6 mice, and 13 male and 1.7 female hamsters, were fed a diet combining 4% young, dried, and milled flower stalks of qglggiqgg'-iaponicus ltaxin for 480 days. A11 surviving animals were killed at the end of the study. Lung adenomas and adenocarcinomas were found in 30/39 surviving male and female ddN rnice combined (conpared with 1/50 in the respective controls) in addition to other tumours (Table 13). No significant differences in tumour incidence were observed between treated Swiss and C57BL|6 mice and hamsters, and the corresponding controls. No data were given -169- on the tumour incidence according to sexr in treated and control aninals or on survivat in the controls. In studies on rats (ltirono et 41. ' L976) ' the dried and (coltsfoot) nere powdered flower buds of E-1134--gg.fa!q nixed with the diet. rorfrfrffi?i strain rata ltere - divideal ioto 4 grouPs and fed diets containing 40 32O e T. farfaralkg for up to 600 days. The number of animals in 6lE-n group and the dosage and feeding reginen are given in Table 13. Two groups reeeiving a diet containing 80 320 glkg oo a conEinual basis deveLoped tumours in the liver of t6e same types as aninals in the studies using Petasites japonicus, e.g., haemangioendotheliomas, Liver cel'l adenomas, iinliT$liocellular carcinomas (Table 13). A11 12 animals in eroup i survived for more than 380 days after the start of the study. Of these, 8 (5 nales, 3 fenales) deve-loped haemangio- endolhelial sarcoma of the liver. In addition, 3 of the 8 rats developed sinultaneously hepatocellular adenomar hepatocellular carcinoma, or urinary bladder papil"Lo'na. In Group II receiving coltsfoot at 80 gl1rg in the diet, 9 out of L0 animals that survived fot more than 420 days developed a haemangioendothel ial s arcon€I. Hirono et al. (1973) fed Svnphvtum officinaLe, siniLarly dried and nilled, to the ACI JtiEin of inbred rats of both sexes at different levels ranging fron 80 to 330 glke as leaves or 5 - 40 glkg aa root, for 280 days or nore. Eight groups of animals of 19 ' 48 aninals each ltere used on different regimens of feeding. The feeding regimen, the nunber of animals in each groupr and the duration of treatment are given in Table 13. Sixty-five malee and 64 femaLes served as controls. Tumours ltere induced in all groups receiving leaves or roota. The most connon trrmour wae liver adenoma. Haenangiosarconaa vere observed but infrequently (3 animels in the whole study). No carcinonas of the liver ltere seen, but there lras a wide variety of other tumours. It i8 noteltorthy that, in Group VII, 14 of 15 animals fed the lowest doee of 10 g/kg for 275 days followed by 5 elke or basal diet alternating every 3 weeks, developed tumoursr 2 of which were malignant (haemangiosarcomas). In the control' groupr single aninals each had papilloma of urinary bladder, caecal adenoma, subcutaneous fibrosarcoma, mamnnry fibroadenoma, or retro- peritoneal teratoma. The livers of aninels that did not have itre tumours showed other features comonLy encountered in aninals administered PAs, e.g., megatocytosis, liver cirrhosis, hyperplastic nodules, etc.r suggesting that they were induced by the PAs contained in the plant. The authors concluded that the carcinogenic activity of this plant was weaker than that observed in animals fed Petasites (coltsfoot). llirono et al. (1979b) have gumarized the above studies on PAs found in edible plants in Japan. -170-

Ilabs (1982) and Habs et al. (1982) tested a crude aLkaloid extract from the plant Senecio numorensis sp. fuchsii containing fuchsisenecionin@ seneEiiiffiE !10 e/tg). The extract was administerldlntragastrically in 2 doses of 8 rng and 40 ng/kg body weight, reslectively, to Z groups of rats of both sexes, cornprising 40 animaLs each, 5 times a week for L04 weeks (Table 13). A Large, dose-related nurnber of liver tumours was produced, originating in the Liver ceLL and the sinusoidaL system, and predominantLy affecting the female animals. The tumour incidence nas higher in Group II (dose, 40 ng/kg) with 34 tumours compared nith 13 in Group I. In Group I (dose, 8 nrg/kg), 2 tumours were fouqd in 20 males compared with 11 tumours amoqg ZO female ratsr Sinilarly, in Group II, only 6 tumours were found in 20 males compared wi"th 29 arnong 20 female rats. The tumours trhepatocelluLar included 19 of origintt, 16 .of rrcholangigenic rrhaemangiogenic originr, and L2 of origin!, (table 14t. Besidds the liver tumours, L2 maLes and 9 females in the treated,groups and 4 males and 6 females in the control group deveLo-ped a variety of extra-hepatic tumours. senecioniie ii known to be hepatotoxic and capabLe of being converted in the _rat liver into c)rtotoxic metabolites. Fuchsisenecionine is a saturated PA, not previously known to be cytotoxic. It is therefore likely that senecionine was the hepatotoxic component and that perhaps the.two PAs acted synergisticaLly with each other, though there is no actual evidence for synergism. Ilowever, this needs confirmation. Moreover, there appeared to have been other unknown components in the mixture tested (Mattocks, 1986). Ilirono et aL. (1993) studied the carcinogenicity of 2 more pLants (Farf.ugium j?ponicum and Senecio cannabifoLius) of the trlbe senecloneae in the family Courpositae, the l-eaves and stalks of which are used in Japan as human food. Fresh Leaves and stalks of the pLants were dried, rnilled, and mixed with the basal, diet. Inbred strain ACI rats of Loth sexes, with preponderance of females, 1.5 months old, iTsrs divided into 6 groups. .They were fed diets containing various proportions of 'as the dried plant materials, indicated in Table 13. The study was terninated at 480 days, except for one group, which was studied for 560 days. Besides the groups shown in the tabLes, 2 mofe groups of 30 and 28 animals of equal, numbers of both sexes r47ere fed 8% and 47" of lenecio cannabifolius, respectiveLy. None of these animals sffi days and all died of hepatotoxicity. A wide range of tumoura was observed, mostly ln the l-iver, as shown in Table 13, the most cofltrnon being haemangiosarcomas, which r^rere not encountered in the controL group. The carcinogenicity of Farfugium iaponicum is considered to be due to senkirkine an@necio cannabi- -l7L-

folius containe senecicannabine, a nert macrocyclic PA' aene- carcino- ;T;TlTtio., and jacozine. It is probable that the (Ilirono, g""i"ity oi Senecio cannabifolius is due to these PAs et al., 1983). mechanisms by i.y""-pes eL al. (1gs5) srudied the biological wtrich initiate carcinogenesis in rnaLe Fischer 344 rats. i*io."tpi"e (single or- Jou"bl9 of up to 80 vnol/kg -injection up il;-;;6;tj J"tt""ionine (singLe or double injection of ".ta tr iOO riof/tgl were inactive as initiators of Y-GT-positive partial noduLes in rats to 2-acetyLaminofluorene and trep"r.cto*y. larninistration""po".a of lasiocarpine or senecionine deveLopnent of iy'f, .fa"t partial hepatectouy resulted in the double very few nodules. Lasiocarpine given in a single or by at aosl (up to 80 urnol/kg) delayed hepatic regeneration least 8 weeks after partial hepatectoflyt and pre-treatment with this pA reduced the initiating capacity of diethylnitro- and N-nittosomethylurea in rats subsequently selected wittr"".itt. Z-acetyTaminofluorene and Partial hePatecto,my'- Resistant typica1 nodules seiected with lasiocarpine aLso had the resistant nodul-e phenotype (positive for Y-GT and epoxide Lasio- iyar"ir".l and also raciea pi,-induced megalocytosis' citpitte treatment also resulted in small regenerative, nodules were ih.t ,""" distinct from resistant nodules, because they negative for Y:GT and epoxide hydrolase.

5.4.8.3 P-vrroli?idine alFaloid netabolites and analogous synth?tic comPounds

The subject has been reviewed by Mattocks (1986)' The pyrrolizidinl alkaloids are converted into pyrroLic esters in into tie hepatocytes. These may then be hydrolysed .pyrrol'ic alcohois, thi"h more nater soLuble and Less active than "t. the esters. The esters may be widely distributed throughout the body. Some of these compounds and their analogues have been tested for carcinogenicitY.

(a) PyrroLic esters

(i) DehydromonocrotaLine (monocrotaline pvrrole)

Mattocks & Cabral (lg7g) tested dehydromonocrotaline on mice' lhe skin of mice. In their first etudy on male BAtB/c 2-week they applied the compound on the back at 1- to not intervals. Thirty-three applications of 1 1nrol did produce. any skin tumours in 16 nice, but 2' developed 'Lung adenomas; no tumours occurred in L4 control mice treated with (Mattocks & the solvent (acetone). In the second study Cabral, Lg82), 11 fenale LACA mice each received 47 applications of 2.5 wrol dehydromonocrotaline; one animal -L72-

developed a maLignant skin tumour. A second batch of 10 mice nas given sirnilar treatment and subsequentLy received 61 twice-weekly applications of croton oi1 at the same site. Ita!.f of these animals developed tumours. In the control group of 10 nice, only L tumour rras seen. The results of these studies indicate that dehydrononocrotaline requires the action of a promoter to manifest carcinogenic potentiaL.

(ii) Dehydroretrorsine

This supposed pyr.rolic metabotite of retrorsine lras applied to the skin of maLe BALB/c nice at 1- to 2-week intervals; 3.3 treatments of 0.5 or L ymol fail_ed to produce tumours in 15 nice that survived for up to 60 weeks (Mattocks & Cabral, L979>.

( i i i ) l-Methyl-2, 3-b i gtr ine thy Lacetoxyne thy lpyrro Le

Mattocks & Cabral (Ig7g, IgB2) nade 2 srudies on this compound. In the second study, which vas nore significant, 22 female LACA nrice received 47 dernal applications of 0.5 pnol each. ?his caused marked skin damage ,with ulceration and scarring. llalignant skin tumours developed in 19 out of 21 surviving animals. Two hydrolysis products of this esrer, pivalic acid and _L-nethyl-2r3-bi-shydroxyrnethylpyrroLe, - lrere sinilarly tested (Mattocks, 1986). Tumours wlre produced, though fewer. The results of the above studies suggest that the intact pyrrolic ester is a carcinogen.

(b) Pyrrolic alcohols

Studies have been conducted using dehydroretronecine (retronecine pyrrole) and dehydroheliotridine, which are secondary metaboLites of monocrotaline and heliotridine-based PAs, respectively. Originally they rrere regarded as (+)- sn6 (-)- forms, respectively, of dihydro-7-hydroxy-L-hydroxynethyl-5H-pyrrolizine. Kadzierski & Buhler (1985, 19S6) sholred that the netaboLite from monocrotal_ine is racemic and concluded that the product from all heLiotridine and rerro- necine esters is the same (t)- forn. Johnson et al. (1978) painred the skin on the back of 16 fenale Swiss mice with dehydroretronecine, each dose equal!.ing 2O mg/kg body weight or about 5 Unol per mouse, once a week fot 4 weeks and then tlrice more after 6 months. Six mice developed skin tumours. Subcutaneous injections of the same dose yielded tumours in L3 out of 21 mice. Tnenty-eight out -L73-

of 55 mice given both topical applications and sc injections developed skin tr:nours. lrlhen the same study ltas rePeated on 34 BALB/c mice, onLy one skin tumour developed (Mattocks & CabraL, 1982). When 17 animals were given the same dose at more frequent intervals, e.g., 65 weekl.y doses, no tumours developed. Similar results were obtained in a study by Shumaker et al. (1976), already deseeribed in section 6.4.8.1, in which 39160 rate given repeated subcutaneous injections of dehydroretronecine developed rhabdomyosarcoinas at the site of injection. The compound aPPears to be a direct-acting carcinogen for rats and mice, though the susceptibility of various strains of mice varies. Peterson et al. (1983) gave 9 ip injections (60 76.5 ng/kg body weight) of dehydrohel.iotritline to rats over a 32-week period. A Large variety of tumours was produced. It was concluded that this metaboLite nay be responsible for the carcinogenicity of its parent PA.

6.4.8.4 Molecular structure and carcinogenic activity

Mattocks ( 1986) has reviewed the Present position concerning the relationship between molecular atructure and carcinogenic activity. Data available at pf,esent are not adequate for any strict correlation to be establ"ished between the molecular structure of PAs and the tyPes of tumoura produced by them in the rat. Ilowever, the connon determinants in the molecuLar structure of all carcinogenic PAs are that they are macrocycLic or "opentt diesters, in which the amino- alcohol moiety is retronecine, heliotridine, or otonecine. These are all esters of unsaturated necines and are caPable of being metabolized to pyrroLic estera in the mamnLian Liver. Studies on pyrroLic esters, the toxic metaboLic product of PAs, have yielded equivocal results. Monocrotaline has been sholrn to bind covalently to DNA, which is associated with the carcinogenic activity of the pyrrolizidine alkaloids (Robertson, L982). Dehydromonocrotaline, the prinary metabo- Lite of monocrotaline has been found to be an incomplete dermal carcinogen (Itooson & Grasso, 1976; Mattocks & Cabral, 1982), whereai the synthetic compound l"-nethyl-2r3-bistri- methylacetoxymethylpyrrole, which is chenicalLy similar to dehydtomonocrotaline, is clearLy carcinogenic (Uattocks & Cabral, L982). Dehydroretronecine (oun)' a second metabolite of monocrotaline an-d possibly other retronecine-based PAs, is aLso carcinogenic for the skin and is coasidered a proximate carcinogenic metabol,ite, since, unlike monocrotaline, it acts at the site of appLication directly and not at remote sites. -t74-

6.4.9 Antinitotic activitv

Literature on this phenomenon has been reviewed by Jago ( 1.969), DIcLean ( 1970) , and ltarrocks ( 1986) . The most characteristic feature of the chronic hepatotoxicity of the PAs ie the presence in the liver of megalocytes (Bull & Dick, 1959; llclean, 1.970), which are generally enlarged hepatocytes containing !.arge, hyperchromatic nuclei (section 6.4.1.5). These appear to be the result of a combined action of pAs on the hepatocytes, a stimulus to regenerate following parenchymal cell injury, and the powerful antinitotie action pAs of !lt" pyrrole rnetabolites of the (Mattocks; 19g6; Jago, This property has served as rhe basis for using a pA .L?69!.- :cancer (indicine-N-oxide) is a chemotherapeutic agent for (Letendre-et al., 1981, 1984). peierson (1t65) showed rhar the number of mitoses foLLowing partial hepatectomy lras reduced to 50% or less of normal valires by prior administration of hepatotoxic alkaloids, and that the effect r{ras: doae dependent. The hepatocytes seemed to continue to grow without dividing. The effect can be produced by a single sublethaL dose of the atkaloids (Schoental & Magee, 1957) or can be a cunulative effect of sna1l doses (Su11 & Dick, 1959). The lesion appears within a few weeks and may persist for the lifetine of the animal (Uattocks, 1986). It lras characteristically described in the liver.of the rat, but has also. been reported in a number of other animals, e.g., mouse, sheep, horse, and pig, and in some other organs, e.g., kidney and lung (ltcl.ean, L970). It has not been observed in human f.ivers (Tandon, H.D. et al., 1978; I"Iattocks, 1986), but it has been observed that cuLtured human fetal liver ceLls becooe enlarged when exposed to PAs (Armstrong et al., 1972) indicating a susceptibility to the antimitotic effect of pAs. The Leeion is not specific for PAs but has been reported to be produced by a number of toxins (ltcl,ean, 1970), including semisynthetic derivatives of PAs but not non-hepatotoxic pAs, such aa platyphyline (Jago, 1970). The ultrastructural features of negalocytea are controversiaL. However, consistent with other functional and metabotic features, the cell.s show morplological cheracters suggesting increased metaboLic activity lrith increased exchange of naterial between. the nucleus and cytoplasm. This unique reaction of the hepatocytes to pA has been used by Jago (1970) to develop a method for assessing hepatotoxicity and by Culvenor et al. (1976a) for screenirig O2 res for acute and chronic hepatotoxicity and pneunotoxicity. Mattocks (1985) suggested a scheme for the aniinitotic action of PAs in v1y11, consistent with observations of the phenomenon in aiGTil The pA is irreversibly netaboLized by the hepatic microsomes into a pyrrolic ester, which can be -175- hydrolysed to a pyrrolic alcohol. The l.atter is the agent tiiat inhibits nitosis. The more reactive prinary rnetabolite may also do this, by reacting vith tissue constituents to give products identical to pyrrolic alcohols, inhibiting mitosis. On the other hand, it can also produce acute injury to th'e ce1ls, which stimrlates regeneration. Thus administration of the PA or the pyrrolic ester can induce megaLocytosis to a much greater extent than a secondary netabolite alone. Antinitotic activity does not seem to be directly related to inhibition of DNA synthesis, since the latter recovers within a week while rnitotic inhibition continues for up to a period of 4 weeks (Peterson, 1965; Armstrong & Zuckerman, 1970). I"lattocks & Legg (1980) have shown that the LeveL of DNA synthesis is reduced in cells that do not divide, but is nor totally inhibited. Samrel & Jago (1975) investigated the position in the cell cycLe of the antimitotic action of trasiocarpine and of its pyrrolic metabolite, dehydroheLio- tridine. Their studies indicated that the alkaloid acts during the late S or early G2 phase of the ceLl cycle.

6.4.10 Imnunosuppression

Dehydrohel.iotridine (DnH), a pyrrolic metabolite' has a significant iumunosuppreasant activity on the primary response in young mice; when injected ip shortly before the antigenic stimuLus (Percy & Pierce, 1971). The secondary response to antigenic stirnuli; as measured by the reduction in the number of 7S and 19S specific antibody-synthesizing cell-s of the spleen, nas suppressed when DIIII rdas administered at the time of secondary stimulus', but not when it was given 24 or 36 h after the sntigenic stimulus. It was suggested that dehydroheliotridine selectively destroys or inactivates cel1s involved in the initial stages of anEigen recognition and processing.

6.4.11 Effects on mineral metabolism

Aberrations of mineral metabolism have been observed i-n several species of arrimaLs. The most notable among them relate to haemolysis and copper metabolism. Anaemia has been reported to occur in rats following PA poisoning (Schoentat & Magee,1959; Schoentalr 1963) and the kidney and liver show haemosiderosis (Ilayashi & Lalich, L967). Besides the haemolysis, PAs have been found to exert a direct inhibitory effect on haematopoiesis in the livers of new-born rats (Sundaresaa, L942). Disturbances of iron metabolism and haematopoiesis have also been demonstrated in rats fed Senecio .iacobaei and supplementary copper (sr,rick et a1., 198228-d -176- they have been found to develop raised copper levels in the liver and spleen, when fed on this plant (Swick et al., 1982b). Miranda et al. (L979) reported elevated levels of iron in the liver and spleen and of copper in the liver in rats fed on tansy raglrort. Studies vith radioactive iron also indicated a specific inhibitory effect of PAs on haematopoiesis. High copper levele in the liver associated with haemoglobinuria have been reported in sheep grazing on heliotrope (guLl et al., 1956), signs of disease closely resembling those of chronic copper poisoning. St. George- Grambauer & Rac (L962) reported a similar outbreak of fatal jaundice due to haemolytic crisis of chronic copper poisoning in shbep that had grazed Echium plantagineum ovet 2 or more seaaons. The pathological changes in the liver lrere indistinguishable from those of l{eliotropium, and the livers had a high copper content. Studies of Miranda et a1. (1981b) indicate that dietary copper can enhance PA hepatotoxicity in rats. It has been suggested that the hepatotoxic effects of some PAs may interfere with the excretion of copper (Bul1 & Dick, 1959; Barrington & GaLLagher, 1960). Sinilar effects have been observed in pigs fed Senecio (Itarding et al., L964). White et al . (1984) did iiftEserve any rise in hepatic copper levels in sheep fed Senecio jacobaea.

6.4.L2 Methods for the asseaament of chronic totoxicit Dneumotoxic v

Pyrrolizidine alkaloids produce acute as well as chronic liver damage. The acute effect is seen as extensive necrosis (SchoentaL & Magee, 1957), while the chronic effect in rats is manifested characteristically by the presence of greatly enlarged parenchynal ceLls (Schoental & llagee, L957; Bull & Dick, 1959), which persist long after a single expoaure (Schoental. E ltagee, 1959). The latter effect of megalocytosis nay nanifest without the liver ceLls going through the process of necrosis (Schoental & Magee, 1959). This property has been used by Jago (1970) to develop a method for the assessment of the rel-ative chronic hepatotoxicity of different alkaloids. It has since been used by other investigators (Culvenor et al., 1976e). It consists of the intraperitoneal administration of a single dose of between 0.025 and 3.2 umol of the alkaLoid per kg body weight in 0.2 ml aqueous solution to L4-dalrold suckling hooded l{istar rats of both aexes. The litters are randomized among the mothers, one day before the administration of the toxin, aad then weaned at 28 days. Aninals are killed 4 weeks after the injection. The relative acute toxicity is indicated by the dose 1evels that cause death within approximately a week and chronic hepatotoxicity -L77-

by those that produce hepatic megalocytosis within 4 weeks of the injection. With this method, it is not only possible to evaluate the hepatotoxicity of a given alkaloid but also to compere the hepatotoxic effects of different compounds in relation to each other on a molar basis. Chronic effects on the lungs can aLso be assessed by the same method. Culvenor et al. (L976a) found this method satisfactory for compounds of medium to high hepatotoxicity but faiLed to detect toxicity ia certain compounds of known, 1ow hepatotoxicity.

6.5 Effects on Wild-life

6.5 .L Deer

There is very f.ittLe infornation on the consumption of PA-containing pLants by non-domesticated animals, birds, and other wiLd-life. In one instance, the deaths of white-tailed deer in coastal marshes in Louisiana in 1967, was ascribed to the consumption of Crotalaria and/or Heliotropiun species in a period of feed scariTt-fl3@r et aL.i19-6fr-TEe aninals had thin and lratery bLood, abnornaL bone marrow, and serious atrophy of cardiac and mesenteric adipose tissue. In a 9-yearold doe, the liver I'ras dull and somewhat granular and evinced megalocytosis of the hepatocytes. In aoother 4-year-old animal, the liver appeared normal but with microscopic evidence of early megalocytosis, with considerabl.e vacuolization in the centrilobular hepatocytes. Plants of the genera Crotalaria and p!!g!1gg!g were abundant and rhere were some Senecio species. There were signs of ingestion of the p!.ants TfTf,Ea.... Senecio jacobaea (tansy ragwort) has been fed experimentally to bLack-tailed deer (Odocoil.eus- hemionus columbianus) in oregon to deternine the-ii-.usEE[tlbllity to poisoning (Dean & trIinward, L974). The ragwort was given ad libiturn together with different Levels of basaL ration (85Z, alfalfa, LOlt barley, 57" mol-asses) to captive deer. The ragrrort nas eaten, only when the basal ration was inadequate. One group, not given any basal ration for 6 days, began eating ragwort and consumed 5.4 kg dry weight per animal in 42 days. This represetted 247. of the animal body weight. The aninals did not show any toxic signs, and blood levels of SGOT and bilirubin were normal.

6.5.2 Fish

The effects of S. jacobaea alkaloids on rainbow trout ( SaLno airdneri ) fingerlings has been investigated (Hendricks et al. , 1981" Duplicate groups of 80 fingerlings lrere fed -178-

for up to L2 months on diets containing 20 or 100 mg alkaLoid/kg. The alkaloid conprised 9t7 jacobine, 3% jacazine, 2.52 senecionine, and,2.5it seneciphylline. The 100 ng/kg diet resulted in severe growth depression and mortality, which began at 3 - 4 months. Both leveLs of PAs produced severe hepatic lesions. The livers from these fish were shriveLLed, mottled yellow or whitish in colour, noduLar, fibrous, and sometimes haemorrhagic. MicroscopicaLly, there !'ras negalocytosis, severe fibrosis, and bile-duct proliferation. Characteristic veno-occlusive changes were seen in the centrilobular and hepatic veins, yhich, in the case of fish receiving the 100 rng/kg diet, appeared after only 2 months on the diet.

6.5.3 Insects

There are no reports of the toxicity of PAs for insects, but there is substantial Literature on the use of PAs by certain insect farnilies that have evoLved with the ability to store the alkaloids as defensive chemicals and to convert them into pheromones and other signalling chemicals (see recent reviews by Brovrn (1984) and Boppr6 (1986), and an earlier complementary paper by Edgar (1982)). In some species, such as moths of the fanily Arctiidae, the Larvae feed on PA-containing plants. In other families, such as Nymphalid butterflies of the sub-families Danainae and Ithomiinae, the Larvae of most species live on other plants, but the adult males seek out PA-containing species and contrive to ingest alkaloids from wilting, dead, or damaged plant material or from nectar. The alkaLoids so acquired have a functional role as defensive chemicals against predators and, in some species, are also converted into pheromones and other signaLLing chemicals involved in rnating. The alkaloid derivatives may be pyrrolic compounds reLated to dehydroretronecine.. or derivatives of the esterifying acids. In one Arctiid genus, Creatonotus, the aLkaLoids have a rnorphogenetic or hormonal effect, determining the size of the pheromone-disseminating organ. Thus, for some insect speciesl PA-containing plairts may be necessary for survival. -L79-

7. EFFECTSON MAN

The toxic effects of pyrrolizidine aLkaloids are principally on the 1iver. The toxic disease, produced by consuming PAs derived from certain Plantsr is called veno-occlusive diseaee (voD), the pivotal and pathognomonic lesion being the occlusion of the central and sublobular hepatic veins in the liver. The larger hepatic vein tributaries are characteristically unaffected in contrast with the findings in Budd-Chiari syndrorne (Bras, 1973).

7.L Clinical Features of veno-Occlusive Disease (vOD)

There are several good clinical accounts of the disease, rnostly in the earlier reports from Jamaica, in children (Jelliffe et a!.., 1954a,b), and adults (stuart & Bras, 1955)' which have been surunarized by Stuart & Bras (1957). Maksudov ( 1952) has described the clinicaL features anong chiLdren in outbreaks in the USSR, where it was called toxic hepatitis with ascites, and Ismailov (1948a,b), Mnushkin (1949, 1952)' and Zheltova (1952) have described them €rmong adults. Srivastava et al. (1978) also described the clinical findings among children in a large outbreak. ChiLdren seem to be particularly vulnerable as is evident from the report of sruart & Bras (1957) and that of Mohabbar er al. (l-976) concerning a large outbreak, though the disease is rare before the age of one or trilo years. Frequently, more than one menber of the family becomes affected (Stuart & Bras, 1957; !'tohabbat et al., 1976i Tandon et a1., 1976; Arora et a1., 1981) within days or weeks of each other. The disease generalLy has an acute oneet, characterized by rapidly developing and progressing synptoms of upper abdoninaL discomfort, dragging pain in right hypochondrium, ascites, and sometimes oliguria and oedema of the feet. Nausea and vomit.ing may be present. Jaundice and fever are rare. There is generally groaa, t,endet, smooth hepatomegaly often accompanied by massive pleural effusion, and sometimes slight splenomegaLy and minimal ankle oedena. Liver function tests may show only mild disturbance. The acute disease is associated with high mortality and a subacute or chronic onset nay lead to cirrhosis. Death often occurs after oesophageal. haeoorrhage. Stuart & Bras (1957) surrmarized the clinical data of 84 patients ranging in age from 6 months to 53 years. The highest. incidence occurred betlreen the ages of L and 3 years (392), and. 267" of patients belonged to the 3- to 6-year age group. Thus, children up to 6 years accounted for 657. of total cases. Although the VoD was relativeLy uncomnon at the 2 extremes of age, earLy infancy and adult life, mortality was -180- highest in these groups., being 60Z and 542, respectively. Ilepatomegaly and some degree of ascites ltere invariably present in acute cases; in 48 out of 64 patientsr ascites rtas acute enough to require paracentesis. In 38 patients, hepatomegaly nas grossly severe, reaching more than half way down to the umbilicus. Jaundice ltas relatively unconrmon. Anong the liver function testsr the most significant and = consistent changes were found in the serurrcholinesterase (t 2.67, O.O2 > P > 0.01) and serurralbumin levels (t = 2.82, 0.01 > P). Mortality rates associated with signs of parenchynaT liver danage were 747" wi.th cLinical jaundice or high leveLs of serurnbil,irubin, 622 witl:. dirninished serunF cholinesterase, and 58"/" with considerable anorexia and apathy. The mortality rates among casea with acute, subacute, or chronic disease rrere reported to be 277., I77., an.d 57%, respectively. Death was urostLy due to hepatocellular failure QL7") h the acute phase and haemetemesis in the chronic phase (75ll). More recent publications (Lyford et a1.' 1976; llcGee et al. , 1976; Tandon, B.N. et al., L977; Datta et al. , 1978arb) also describe the haemodynamic data of the hepatic blood flow and the results of portovenographic studies that suggest outflow tract obstruction in the liver at the post-sinusoidal level, and irregulariEy and obstruction/distortion of the hepatic venous radicles, respectively. The clinical course of the disease has been shom schematicall"y by Stuart & Bras (1957) (Fig. 12). However, the temporal relationships in the different phases of the disease are not precise, and their account' at best, represents a trend.

Comptete Clinical * Latentpetiodolyears recovery (48%) recovery (20"/.)

\ A. AcutevenoGlusive l/ (13olo) B. subacuteveno-mlusive c. chtonic veno-occlusive ---_> ---+ diseGe disease (sudden oiieae ipersisienr lepatomega$ina asciresl heriitomegaty)"orm (cinhosisol the liver)

Yr t\ / Fapiddeath(1e%) l\/; (increasingjaundice cholcemia) | Ctinicatimprovements 3 I ? Completerecovery

Irig.12. Clinical natural history of veno-occlusive disease of the 1iver. B and c nay be presen! with no clinical history of preceding illness. From: Stuart & Bras (1957). - 181 -

The onset of the disease may be sudden (acute) or insidious (subacute or chronic). The acute disease nay recover completely or result in death. A few patients may go on to the subacute phase, with almost none or very few symptoms, but a persistent hepatomegaLy. The patient may subsequently recover corrpletely, or mayr after or without apparent clinical inprovement, go on to the chronic phase of disease, nostly ending up in cirrhosis. Some patients riti.th the acute disease may go on to the subacute phase, even after clinical recovery, or as postulated by the authors, after a latent period of several years, progress to cirrhosis. Braginskii & Bobokhadzaev (1965) related an experience concerning the evolution of this disease that was sinilar to that observed in the USSR. About 50 - 60"/" of cases made a fulL recovery and 35"1 made a partiaL recovery with continuing hepatornegaLy. Abott 2% of cases deveLoped persistent ascites vith trloss of working capacitytt. It has been suggested that these cases may develop cirrhosis.

7.2 Salient. PathologicaL Features of Veno-Occlusive Disease

The pathological features of the disease at different stages have been described in detail (Terekhov, L939, L952; Dolinskaya, 1-952; Bras et al-. , L954; Bras & lrlatler, 1955; Bras & HilL, 1956; Stuart & Bras, 1957i StirLing et al., L962; Aikat et al., 1978r' Tandon, B.N. e€ al., 1-978; Taodon, Il.D. et al., L978). The following description of Bras & tr'latLer (1955) characterizes the morphological changes at different stages of the disease described by most investigators. Morphological features of the liver at autopsy in L9 patients from Jamaica who died at various stages of VOD have been described. The ages ranged from 10 months to 45 years. Fourteen patients were below the age of 14 years. Of the 14 patients, t had developed cirrhosis, 3 died of acute disease, and 7 had various Levels of fibrosis superimposed over acute VOD. In the acute stages, there r,ras acute centrilobuLar congestion. The centril-obular and sublobular veins showed different degrees of thickening of the wall and occtusion of, the lumen, nainly due to subintimal sweLling composed of loose reticular tissues and a few ce1ls including endothelial ce1ls, occasionaL histocytes, lymphocytes, and polymorphs, suggesting an acute exudative process. In addition, there was a small amount of fibrin. Organization of this exudate gradually led to collagenization and thickening of the wall. The centri- Lobular. congestion resulted in compression and even disappearance of liver ceLL cords. The reticular framework of the liver lobule was frequently. preserved but was ruptured at places. Clearly recognizabLe thrombi occurred sporadicall.y -182-

but were not coftmonly seen. The portal veins and hepatic arteries were normal. Stirling et al. (1962) in their description of the earLy lesion of VOD of the liver also emphasized that thrombosis of the hepatic veins was not an important histogenetic factor in the evolution of the lesion. In the later stages of the disease, hepatic venous occlusion was chronicaLly established. Perivenous reticular coLLapse and the resultant condensation and reduplication of the reticular framework resulted in non-portal fibtosis and later cirrhosis. Macroscopically, the cirrhosis Looked like Laennects cirrhosis. The cirrhotic process, which $ras non-portal to begin with, involved portaL tracts r^rhich also became incorporated in the connective tissue septa. The presence of oesophageaL varices lras a comnon finding. Extra-hepatic collateral vessel.s and congestive splenomegaly \,sere frequently seen. I'lithin the liver, intra- hepatic collateral circulation was estabLished with the coalescence of sinusoids. The larger hepatic veins and inferior vena cava did not shoro any thrombi or other pathological change. The subintimal sweLling observed in the hepatic veins was not seen in any other vessels in the body. The authors concluded that acute VOD gradualLy leads to Laennecrs cirrhosis, but, in the initial phases, it is non-porta1. No cLear dose or temporal relationships between the liver and parenchymal and vascular changes in the liver lobules are evident from the available human case reports. It i6 notable that megalocytosis, wtrich is the hall-nark of chronic PA loxicity in experimental animals and a morpho- logicaL manifestation of the antirnitotic effect of PAs, has not been observed in.human subjects (Tandon, H.D. et a1., 1978) (section 6.4.9). Tandon H.D. (personaL .communicatioa) has analysed all published and unpublished observations incLuding f.iver biopsies and autopsies derived from 3 outbreaks of VOD of which troo occurred in India (Tandon, B.N. et aL., L976, L9773 Tandon, R.K. et a1-., L976; Tandon, Il.D. et aI., L977) and one occurred in Afghanistan (Tandon, B.N. et aL., L978; Tandon, H.D. et a1., 1978). He comnented on the frequency with nhich the characteristic veno-occlusion may not be seen in the needle biopsy in the acute phase of the disease, though it is invariabLy observed in the autopsy materiaL. Ilaemorrhages may persist for irp to I year after subsidence of acute s)rmptoms. There is no significant inflamnation accompanying the veno- occlusive changes. Cirrhosis r,eas reported to have developed within 3 months in one patient, after an initial biopsy had shown co fibrosis (Stuart & Bras, L957). Brooks et al (1970) nade an ultrastructural- study of the liver in VOD in Jamaican children. Thev found extensive -183- damage in the sinusoidal epitheliurn resulting in entra-vasation of red cells in Dissers space and bettteen hepatocytes. The closed structure of the vessels, the absence of fenestrations, the existence of a basement membraner and the presence of collagen in the walL contribute to the resistance to celluLar debris and erythrocytes which track up Dissers space and tend to narrort the Lumen of the sinusoid where it enters the vein. Fibrin may be present in this location and occasionally in the sinusoids, but not within hepatic veins themselves. The venous block, therefore, does not appear to be the result of thrombosis. Pancreatic changes similar to those in Kwashiorkor were stated to be comonly observed (Bras & uill, 1956; Stuart & Bras, 1957).

7.3 lluman Case Reports of Veno-Occlusive Disease

Avail-abLe data on hunan cases are sumarized in Table 15t with the countries they have been reported from in alphabetical order. The first report of this disease in man and its reLationship with consumption of wheat flour contaminated by seeds of a plant of the genus Senecio was made by Albertjin i; 1918 in a report made to the g6ffi6Ent of South Africa, as quoted by Wilnot & Robertson (1920), who gave the first account of this disease in scientific literature from that country. It is possible that it may have existed even earlier, as its occurrence in farm animaLs had been described since the beginning of the century in veterinary l"iterature (sutL et al., 1968). WiLlnot & Robertson (1920) recorded the occurrence of tbread poisoningt in South Africa for a period of about 10 years, during which 80 cases had been observed, mostly resulting in death. The detailed account of clinical data and pathological findings in 11 cases was consistent rntith what is known as VOD today. The flour from which the bread was prepared was found to be contaminated with the fLower heads and seeds of Senecio spp. rsolated case @Es. are available in the literature (Ilashem, 1939; Wurn, 1939) describing changes in childrens' Liver characteristic of veno-occlusive disease; however, in these studies, no mention was made of possible etiological agents. A pattern of disease described in the Soviet Literature as dystrophy of the liver, and later as toxic hepatitis with ascites has been known in the Central Asian republics of the USSR especially Uzbekistan and Tadjikistan for a Long tine (Terekhov, 1952). The local inhabitants calLed the disease, characterized by rapidl"y filling ascitesr the Itcamel-be1lytt syndrome and are reported to have Long distrusted a weed with fine seeds, knorrn locally as ttKhannyktr, as a source of the -184-

a^9 o o 9€ q € !F .3 @ OF t-@6 d o o o tr Eaid !v 9d ts O O63v ! kv o 99N o 60 tr CoE . a6 o q 61 0d ! ! lrarvEO @4 A6

o o q\ .iO F Eo^ 0 o l.i! .t oF@o >d I o o 0 o , oE! q oo o q OvF F trI H o o o d oo o d !! 6A OE !h ) AE tso oo oo oo !!o u.i 9.d d @ 5! ,6 )o ao do root oo oo oo za oEdo qd 6r Od R5 O^ i! \! o -u bq @ogdc ;i€ o o o o o€ io J o o ! u.t+ !6 .S C O! i .i t co . O,d ! Ad ^O oEoc dE E | 9X.^tO oO d o! N d QO! Jd c q€ o E.i E sl rOFE ..d d o A ho d ga q5 F o uq ! so E OX cd .d rd .d !o dd od s! d dd a.i !6 p GA ts H! ,q oo oo q3 ! o

d o o -"1 Noo I H k al 6k g oo lo6 @! oo60 <@ stisI i$, ?htsxox .;l;l r^ or *t:l81.'l tsH x >, 5l p :l oo -' tolEl 5.lE d; tlEl - '!l$JflEl' * 556 qr oa E or Elol oco

ld o !o I q o ooc)! oo B^ .c S^ ao ,66 Cv^t{e EO Ot€ 6 6 00 @ o q oo H E .@ Co Na d.cd 5do or Ev 66v B t4 d d >v

d o o oq o !o ts .dq 6

@ o >u! oo oo o9 !o )o to t ./ !o q a.l oo o I EO d d ^ od fuO 'd I t.-\HHXU .t ilu I d I o oo d o d o oJ-*E P ! €!aFEO d\ ..t SSOO I E O O.a s@d EE!Nrc.d6ad Es JN x 6!o $5 E.i( d{oHde..-. gol E.d{E d o.iJi o .daEo u.d E O lo€d oogooiE !0o! o .iE --.dJ d! 'cEga !ox oox6ooEko! E H o O.-ll o O!6 > E OO OO o 5 o d I oo 6 | c6 o t->o ooG o F9zl&tSzlaiEoodoaB6

N€

kd til AS € o q E{ o

o it r .;o h.? p tc F oq eIa .: :6'E d ii s9 JEE do X ! dq lrx H O rA 060 O

.:t o € :lrOo NH pl i o il k dd I ,rr A3 to 3J -o N x *lel ot t6l cl rlEl 9l o - * o n $l3l El $- 8lH €l -186-

MOZE 6 .o. o E n ^ o !d .d3E6...dOdd 0 o i 'ii '^ d : -,iiS. f E.r.lS. f k do oda OHts OdA^.d..!€ -C OF.id@ o s9 ! drd dF € dNF€ dF o Nd di Ao q 6 C O tr 6F€ O .i.rOi O o )a E!dOUi d!d6.i!d !Fd.iuJ x H OvF oe H Ovd@ ov !Z dvo Ov

N o .-o o o\ ts to g o^ o ot o 'd o o t.dt r.'r o ! ohoo o E.i o !€ o o oo o , o,cr o NEA.d o Ov k h di h 5€ q

oq o t-.d O oo o o oc gloo 300 FO o o o-c c o o d o r.l !.dko .r@o ) ) 6F ts6o .!o o I O.rq./ E d!.d I z; a E d > ot d

o*] o 09 o 'x lz J o g E6 .^C .d I E! J' Uo Cl ococ o o os 4 06>!00 X C ET I t oi E doHc @ o o k.il .O O O 6O

q o o s.E h CF o cd o 9d E3 o s od dq J,! !F o qo d€ oo

dl !! -tl cl oo rl o .11 ! .rl oc !qt-, gl oo 6lcl z^ Htfl 8l o

.do9C o ! O! Od >d 9d o o U" H ! ! ox or hc E* !o d F.d .id '13 FA Ed oc *r 6q dE iik f.r !o H€ OdF Eh EO tq oc ooo oo o @o gdaoo oo

o^ u€ !q co N ! o @ o O.r + c k .1 H .i9 O o gr^ o 6 o n{,8 .\3-t.3 Y to <4 N O ts d }\v o.o fl

k! o ! oo ol troo .dl N6€ *3 d o 3d El ts oEo HI -L87-

{ EO ! !9 ? .3 . .S^ ts!^ .6'oJ 3T^ QF!F SFR EbR $gd JO!O tt6 'Ed!6p5 dllo

Ga g !qo €uq6 ! I o .d.d 6 o 'dlEo .tE, Eo N t.d 6 UE O UE o c.r I >a 6.i .-!i rr E .i oq E ol .-N.r O O .. Q td O dF d O I F | o O N.d .d O.dorl.- EOO.r!.d O 6 ! Q L o oE d d E ! o o o FFO6 F 6 !i O O C E ! O O !.il F O d O 6 E d ! !dFo .at !SEUdOCEO 5ok uFo.do t ci o d 6.A coo([email protected] o o.d E !5 0 0 0 o50d o€ oo6Ek.coodo 6io.ilo!3>\t 6 o: 0d.cFo!.c>o>

uo u x3 aq \ 60I O usd c ! E Ot C O.dd.d O d O N E !.d I @ dAqtE.ro I d a go u Q Od o N O tr3 r.d du+.iBq€

I o o o (, d 6. o o u !o o o o o !E o.r o do ! € o o -P*r o .:l ol .iB9lrlol ''l I o E r6l d ol o o 'F.rxo: @l 3.dl t3lts .i ol a o o r:iI ol ! as 901 o ti O'i.d @ o cq o

o €q u o ! H 90 3 9o o q6 d o 6' 6 €-t oo s.o o o6 5@ 5.d Aq td 3

! o oo OF dl | 6 o uo lo ol r o gN 'rl r >. FI :l a 'Il - -rh3 N rl5l i E!.d 6 !l ol o ! ts ;l @l v6 0 -188-

F d o !ts €o o€ o o .6 ts trd E oN c O@ ov o or o o 06 0 (,a OH >v BA qf, O .- O o\ E o^ O l.dt o60E qtsao .i ou o !.d o.d !o o ,roo 'no E3f- o€ E 3 !> O! o o .i O o Od 'EE c"'d Xt: : @, oi0 .36;.3 AE oo00! 3.3, !q p66 0 q ' 3 *'E3 E5 ? 3 , d'ots€ u od tr ogt o h k.d o o do t C A.i k O t H d o s ocQ o O rJ 6 Zt q O!.i ! O.d : 6: .l o>oiloq : >>€

9d :l .i I E! odod rI 3 oc ! Q o! E' il E o€ doFc tsO!

o o o hE 6.: r NO 9.d o EO odo \ 5F .::i o! il d !'5 E o5 EEE 6 oxp OO ! @'-! o .d 6G e o d o u o g I'r o o d itE ! tu 9E pd ifd ! sd o sh zcl g 'i: 51r o 9C ou .JO tro t t.J .d+ o o O! E >d o o €.r O oo ox EEHo uo o o9 Fp.co.o 9c :.: 3 d t Ea "?,d ! ts oto c o o @o 5d9d

! o o k o O@ d 0 9-o of I c o ot o

~~O^ ^o^ F JO ( trF o o@oo .d > a >3 E.-O !r.-O - Ea9^ tsd!d t^8^ >^d C a@ rvOO ov 6v a o i o o r.d o r d 9@ Ja! O6 Oo tsS6$ xo 'r . X6 k6tso !oE6 oi !Jtil .lr OJ aA@A @ 6;dv c olt o EvF{v AvHv~~

~~6 lc o N.d'd o gh o OH 5t" >H!O H ooHo o ts o!oF 't3e oq.c t 6.d5 d o i€ ! E~~

oooo ACEE oo oo o9 oo U.r ! .d !.d P.F )e aq Jq )o oo oo Qo oo dt dd dd od o tt - | €d E.r o-{.i6lOOql J i x ! qr o r.{ J or-O doOOO 6t @E a d o ! 5 | lFq Y.i ll a u dC aaal OiX \'r'"E u.d.i oa do 9p9>,'irEd JzO q!-;lF udos l- .>-id9p geazl -€49 @6dE .16 6 O rh qE J! @O . ! oF Eo .! ts+l 6;O€ dOO6+'iC $9OAn d d!o o di a.id3 d d€ o.uo

E OC E C.d o o .iE OO q -dO Gd E o o od.i € o .d o.il g E'r @ co ca E ha 'd!.dH .d o.il .c o o I No d O O AG ts d ha d O h./ O o ! o !9 q.r o o o .?6.9 o o oq 9XkO s AH d .d .t .d.i ! o! lo i 6 atq .d t.o o g o d oo F zlts o 6 o Fh; t i Et ; o 'Jt) 3l ola ol EI otH .9131 ;'l tsld ql *l 3tE st EI ilo otE .;t:l hl lE I'd

~~o o o kts 83 83 .lJ tiH ! q !q 6d CL c! ..!d.d6 E9 EP 6 6 6 n t ;E I h ! o o a oq oq I o ao~~

~~F 6 o ! o ots o to I {~~

~~I ots o UO O o H€ 9o sE d 9n F sl EIild : tr -190-~~

g^ d o,o d 6 F I o .$66n'd a € .06 {O dO E q6 ! d qd d o ql ,\ o^ !v !v €^ ts O tsN .dv + ! ! @ e q OF d @ r^ d o d o oh liao il. 60 ,d td do o hor Or td lri !L L:a c g€v b d av @rb @e av ,ii o\ .i€ EI o^ F@o €oo o t.d€ o I E I A € d O.d qxoo € qrHo.dorlHoE o ts.d o !.c€ o ! o ts o O € N kx p 0t€ O C:.d C ,t d./ ts OvF ! t o.d r k o d oN da o o.d ot Ood o 9O !oal !oo oc o TF FO oo o3 o Ht 5.t !lr s l'E !k !o ,o t! ,:'h 5.d O so oo I oE a qEl 2A do o 6a a: v

~~o J 96 .dtqs ocoE qo! Ja E 6€ F!OtsE: <9ed~~

~~o o~~

o s q a o E"OTEE :I 'i 'd .dg H H q .r .rH .rddtr^ a E E Eoid q 'q q # € ': €'qE o 9E '1 ': =a . frlg o f, I 33le o z& F A c i-'d =" o 'c5 .-9 -b=' *.5 OX E'E !o l,e 5l 9! ErE gr;EEg 9q fr if'd o P5'd €, o 65 . Sfr o 5o oo Oo ,s od @o =UH EE.EEiF.:.E8!.F 9!r

~~Olog.- i. :~~

~~oo F EsH E:n sr fr - h d I d 6 d @H~~

~~! O^ o d ss tro o o^ ! o o€ H oo o €v od~~

~~o\ E o^ o t.d€ o!oo ! .t o.d o , oEt Ov ts~~

~~o o o oc o ho o 5.d ts Po ts 60 o 2a o~~

~~o td .it Es~~

~~o 0p i c q€ rohc~~

~~@ I h o 6 d o~~

E h: o oo 0 .A !k ! (l ! o o d.dl o u o a Flt o ! .i.d otd x d 9qdlodu o o O.r Ot 6 € ! ts ! llslid o 9€ a E 9l>lo .d o Hlilo x ! o 2d 3: ="r,il&3 o o qE q .do o O! o >6 k U 9.d ! o)1 @ oll 'i: o o! o .d 9q 'a oq o

~~P cio oo =.f H ^l o~~

~~EIol I o ! .91 oI .itr o so a -o^ I oh <@ iI 0l d EI t FF El o o E33 A I58 gl* H 0 co ol -L92-~~

disease (Dubrovinskii, L952) (it is notable that, in rhe Afghan outbreak, wtrich occurred close to the area in Uzbekistan where the disease had been known, the toxic seeds were known as rrCharmakrr among the LocaL population). The ttcamel-bellytr disease was associated with the use of bread with a bitter taste (presunably caused by the nixing and grinding of toxic seeds with the wheat grains). Two waves of outbreaks occurred, the first between 193L-35 (terekhov, 1939) and the second between L945-46 (Dubrovinskii, L947; Ismailoy, 1948arb). In the first nave of outbreaks, approximately 1000 - 1500 subjects lrere estimated to have been affected with an overall mortaLity rate of 13 - l5Z. The age of the affected subjects ranged fron 3 to 50 years. In the second wave of outbreaks, 60 - 7O7"of the popul.ation in agricultural areas nere estimated to have been affected (IsnaiLov, 1-948a,b). Domestic animals aLso suffered from the di-sease (Dubrovinskii, L947). During the second wave of outbreaks, the disease vas found to have been caused by contamination of food crops with seeds of Heliot 1as (Dubrovinskii, L947; Khanin, 194TI confirrned by studies on a variety of experimental aninals using the suspect seed (Khanin, 1948; Kampanzev, 1952). Pathological features of the disease in children have been described by Dolinskaya (1952) and in all age groups by Terekhov (1939, L952). First an acute form of the disease was recognized, followed by hepatic cirrhosis, wtrich was found to result as a consequence of the initial disease (Isrnailov, 1948a,b; Sawina, L952). The disease has since been largely eradicated by the agricultural and public health measures taken. Selzer & Parker (1951), from South Africa, described l-2 cases, 10 of whon came from 3 fanilies that had eaten bread made of fLour of [imperfectly vinnowed wheatrr. Five cases who were autopsied showed characteristic occlusion of the centra!. vein of the liver Lobules. Attention on veno-occlusive disease as an entity followed a spate of reports, mostly from Jamaica, on the occurrence of this disease, mainly anong children (Bras et al. , 1954, 1961; Jelliffe et aL., L954a,b; Bras & Watler, 1955; Bras & Iti11, 1,956; Stirling et a1-., L962) but also among adults (Stuart & Bras, L955, 1957) . The youngest patient reported iras a 3-month-oLd infant (Stein & Isaacson, L962), but the disease nas stated even to have been observed in the newborn (Stein, L957)., Very often, severaL members of the faniLy rirere affected by disease, which presented prinarily as rapidly filling ascites, within a matter of days, sometimes accompanied by fever, and leading to hepatic failure. It was considered an important cause of cirrhosis among children in Jamaica (Jelliffe et al., L954a,b). On the basis of evidence - 1-93 - of an almost identicaL disease occurring among grazing animals (9i11, 1960), and a prevalent practice among the Jamaicans of using herbal infusions for treating a variety of aiLments as a home remedy, herbs, notably of the -Senecio and Crotalaria groups, were suspected of being a contffiFry factoli-IEEfr Hil1, 1956; Bras et al., l-957, 1961; Stuart & Bras, 1957). ttiLl (1960) gave a comprehensive account of current knowledge up to that tine of the world-wide distribution of seneciosis in man and aninals in which mention is made of this disease only in Egypt, the Federal Republic of Germany, South Africa, and the West Indies. Pyrrolizidine alkaloids lrere identified as the active element in the toxic factor in the pLant s . Ilowever, there are earlier reports of this disease from elsewhere. Bras (1973) cited a number of reports from Europe from 1905 to L949, including reports by Wurn (1939) and Teilum (1949), of an entity in infants and children that was call-ed Endophlebitis hepatica obliterans. Ile reported having studied some tissue sections of liver in Austrian children aged 7 months - 1 year, who had died with clinical and histological patterns of VOD. Some of these conformed to the Budd-Chiari syndrome whereas others were consistent vith VOD. The case of a 36-year-old rdom€tnreported by Teilun (1949) doeg not, however, faLl in the usual category of cases accepted as veno-occLusive, since thrombotic changes were also seen in the Larger hepatic veins and several, in the peripheral parts of the body, as well as the port.al vein. A conprehensive account of aLL avail-able published reports of thip disease to date fron different parts of the world is given in Table 15. A clear distinction between the acute and chronic effects of exposure is not always possible from the published data, since, in most patients, the history of onset of disease with intake of the toxic substance is. not geaerally volunteered or avaiLable, and so a precise temporal relation: ship is difficult to establish. Ilowever, this information is available irt a reasonabLy accurate form in more recent reports (Stillnan et al., L977i Datta et al.., 1978a,b; Fox et a!.., L978; Kurnanaet aL., 1985). cupta et al. (1963) fron India described 2 cases who had drunk some herbal infusions for the treatment of skin disorders. The liver biopsies confirmed changes typical of acute VOD. Ilowever, the herb was not identified and Bo analysis was made of the infusion. for aLkaloids. The cases of 59 children in Egypt who had syaptoms of rapidLy deveLoping abdominal distension. and hepatomegaly were reported by Safouh & Shehata (1965). Their ages ranged betweeD 1 and 12 years, 47 being beLow 4 years of age. A dietary survey carried out on 17 patients indicated ingestion of drinks nade by boiling some colrmon plant seeds, but these -L94- rrere not identified. Wedge biopsies of liver were performed in 6 patients and 16 postmortem examinations were made. In all the autopsy cases, there was thrombotic obLiteration of the main hepatic veins and their ostia. The central and sublobular veins were not uniformly thickened as they rtere frequently dilated or disrupted and some contained fresh thrornbi. There was centrilobuLar necrosis of the liver lobul.es. One patient developed decompensated cirrhosis in 3 years. The authors observed that the clinical and pathological features closely resembled those of VOD in Jamaica, but thrombosis !'ras an unusual feature of the latter disease. Al-Ilasany & Mohamed (1970) described a short outbreak in Iraq, occurring during a 3-month period, affecting 9 chil.dren, aL1 except one being below 12 years of age, and belongiag to 3 Bedouin families living outside Baghdad. One of the children died. Autopsy of this case and biopsies on the others shoved changes characteristic of VOD. Poisoning through the wheat flour contaminated by seeds of some PA-containing plant was suspected, but no anal-ysis of the food was carried out. Three of the Largest outbreaks of the disease have been reported from South Asia, two from the same site in central India (Tandon, B.N. et al"., L976, L977; Tandon, R.K. et a1., 1976; Krishnamachari et a1.., L977; Tandon, H.D. et aL., L977) and one from North-West Afghanistan (Tandon & Tandon, L9751, Mohabbatet a1".,19761,Tandon, B.N. et a1., L97B; Tandon, H.D. et a1,., 1978). The first Indian outbreak raras reported by Tandon, B.N. et a1. (1976) and Tandon, R.K. et al. (1976) and Tandon, B.N. et aI. (1977) and Tandon, H.D. et al. (1977). It occurred in a group of 5 tribal vil-lages in central India in L972-73. The epidemiology of the outbreak was later described by Arora et al. (1981). Out of a total population of 2060 in these villages, 71 househoLds with 366 members were investi- gdted. Among these, 39 cases had developed and L9 had died before cornnencementof the investigations. The incidence rate vras I.17" and the case fatality rate was 507". A11 cases occurred among 20 households. In many households, severaL members were affected. In one household, 4 out of 5 cases died. Six sick patients were investigated in detail with repeated liver biopsies. One patient died L7 months after onset and was autopsied. The clinical onset r^7ascharacter- istic of disease. Haemodynamic and radiographic studies suggested a combined post and perisinusoidal block. Liver biopsy studies showed features characteristic of acute centrilobular haemorrhagic necrosis with progressive fibrosis, hepatic veno-occlusion, and non-portal cirrhosis in one case who survived 17 months after the acute onset. The etiologicat factor of this outbreak was not established, though dietary contamination with pyrrolizidine alkaloids was considered. -195-

A second outbreak occurred at the same site in 1975 and was studied and reported independently by Krishnamachari et aL. (L977) and Tandon, B.N. et al. (L976, 1977). A totaL population of 486 was affected, 67 cases were reported, of whon 28 (462) had died. There was a strong faurily history (Tandon, B.N. et al., 1976). In a later survey, 108 patients were studied and the mortality rate rilas estimated to be 63"A (Tandon, B.N. et aL., 1977). This tine the etiologicaL factor was identified as the pLant Crotalaria nana Burm. which had been growing in the f ieids ofTTIl*-1F:Tfiin rniliare), rheir staple food crop. The seeds of this pTii-Effiffi-ixed with the cereal grain during harvesting. The toxic seeds contained pyrrolizidine aLkaloids that were identified as a macrocyclic ester closely sirnilar to monocrotaline. The total alkaloid content lras estinated to be 5.3 g/kg of seed, expressed as monocrotaline. The levels of contamination of the nillet with seeds were reported to be 0 - 3.4 g/kg in the uoaffected and 0 - 19 g/kg in the affected households (Krishnamachari et a1., 1977). A precise identification couLd not be mader but the same material, independently studied by Siddiqui et al. (1978arb), nere reported to contain 2 new alkaloids, cronaburmine and crotananine. The seed contained an aLkaloid level of 26 glkg. The levels of contamination of the millet rrere up to 20 glkg and the amount of alkaLoid ingested was estimated to be up to 40 rrngper day. The largest outbreak reported to date occurred in the Gulran district of Herat Province in northwest Afghanistan, close to the border of the USSR. Tandon & Tandon (1975) identified the pLant as being causative factor of the disease, which was surveyed and reported in detail by Mohabbat et aL. (L976). The outbreak, was estimated to have affected a population of approximately 35 000 in 98 vi1Lages. Examina- tion of 72OO inhabitants of the affected villages showed evidence of disease itr. 22.6%, which was more serious in L5Z. Thus, it was estimated that approxinateLy 8000 subjects suffered from the disease including 5000 who were seriously affected. A11 age groups were affected, but 461[ of. subjects rdere below 14 years of age. Ilowever, no sign of disease roas found in chiLdren beLow 2 years of age. A detailed report concerning the pathological materiaL obtained from 14 f-iver biopsies and 8 autopsies r^ras nade by Tandon, B.N. et aL. (1978) and Tandon, II.D. et aL. (1978). The rime inrerval between the onset of symptoms and the biopsy/autopsy lras not indicated. Pathological findings were characteristic, ranging from acute disease with characteristic veno-occLusion to non-portal cirrhosis, which was observed in 5 of the above 22 cases. The outbreak was ascribed to massive contamination of wheat, the staple food crop, foLlowing 2 years of drought, -196- nith the seeds of Heliotropiun popovii H. Riedl subsp. giLlianun H. Riedl, wtr@profusely among the wheat crop. The seeds contained pyrrolizidine alkaLoids at concentrations reported by 2 laboratories to be 7.2 and 13.2 - L4.9 e/ke, identified as mainly as the N-oxide of heliotrine (74%) (uattocks, personal conmrunication)r-and one or tno other compounds sirnilar in character to l-asiocarpine. SampLes of wheat from sevetal villages contained an average of 40 seeds/kg, i.e. , 0.037" by weight. It was estimated that an adult consumed at least 700 g flour/day, containing approximately 2 mg aLkaloid (based on a mean of the seed analyses). There is some uncertainty about the estimate, since Mohabbat et al. (1976) also stated that the samplet of the wheat fLour nere assayed and contained 0.186 - 0.5O7. alkaloid. This analysis and the 0.727" restLt for alkaloid in the H. popovii seed were from the same laboratory in Kabul (R.N. Srivastava, personal conrnunication), and together, imply 13 - 367" seed in the wheat. If correctLy reported, the result for the flour confLicEs vrith the estimated proportion of the seed in the wheat and can scarcely have been representative. Tandon & Tandon (1975) stated in their report of the survey during which the causative factor of the Afghan outbreak was discovered, that such cases had always been observed by Government physicians posted in this area in the past, sometimes in significant numbers, but oeither the population nor the physicians remembered that the disease had occurred in the form of an outbreak. There \ras no mention of VOD in the hepatic lesions observed by Sobin et al-. (1969) among the 121 specimens of liver obtained at medico-1-egal autopsies or by needle biopsies at Kabul, though 6.67" of the 89 autopsy specimens showed 'racute passive congestion with necrosistt. The ages of these subjects was not stated and the authors did not discuss the cause of the l-esion. Fol"lowing these outbreaks, a number of isolated cases have been described following the use of herbal medicines. McGee et al. (L976) reported a case from the United Kingdom of a 26-year-o1-d $roman who had consumed herbal tea containing pyrrol"izidine aLkaloids. The one sample examined contained on1.y trace amounts of aLkaloids. Neither the plant nor the atkaLoids were further characterized. The patient had been taking very large quantities of the tea for about 2 years and it is possible that some of the earlier batches may have been more heavily contaminated. Mat6 or Paraguay tea (Ilex species), which she was drinking, is stated 1o be a popfiii drink in .BrazLL and is not believed to contain pyrrolizidine alkaLoids. It has been stated (Huxtable, L980) that possibly she ingesEed the pyrrol-izidine alkal-oids from some other unidentified source. The ctinical" course of the disease -L97-

progressed rapidLy. Three biopsies carried out within one month-and the autopsy showed characteristic changes including centrizonal fibrosis, but no cirrhosis. It is notable that some involvement of muscular pu!.monary arteries was also seen. Lyford et a1. (1976) reported the case of a 35-year-old rtonan-from Ecuador, who had ingested herbaL tea prepared from Crgtalaria juncea. She had consumed L - 2 lilre-s of this infusion daiLy for 6 weeks, but no qualitative or quantitative analysis for pyrrolizidine atkaloids was made. She had had arthralgiaa for 3 years, for which she had received treatment with indomethacin and phenylbutazone. The liver biopsy showed characteristic changes of acute VOD from which she recovered conpletely as proved by a repeat biopsy carried out one year 1ater. The occurrence of acute disease lras reported in 2 infants in Arizona io the USA, aged 6 months and 2 months, respectively, following ingestion of infusions of a herb, Iocally called the GordoLobo yerba by the Mexican-American population and identified ae Senecio longilobus (Stillnan et a1.., L977; Fox er a1., 1978; nTffiF13F)Ithe pLant from which this infusion had been prepared was found io contain pyrroLizidine alkaLoids identified as riddeLline and N-oxides of^ -retrorsine, seneciphylline, and senecionine (ttExtab!.e, 1980), in a concentration of 3 g free alkaloid and 10.5 g (Stillnan I-oxides/kg et al., L977). Ir lras esrimated thatl during a period of 2 weeks, the 6-month-old infant received a total dose of betrdeen 70 and 147 ng of combined alkaLoid and N-oxide derivative. The liver biopsy showed characteristic features of acute disease, which had progressed to extensive central, portal, and sinusoidal fibrosis (Stilknan et al., L977). The chiLd subsequentLy deveLoped cirrhosis over the next 8-month period. The 2-oonth-old infant was administered an infusion of the same herb for 4 days, after which he became progressively more iLl and btuporous (Fox et a1., 197g). On admission he was diagnosed as a case of Reyers syndrome, but subs_equently developed jaundice and possibly ascites and died. The sanple of herb contained a concentration of alkaLoids of 15 g/kg. It lras estimated that the infant had received a total of 66 mg of mixed alkaloids over the 4-day period. At autopsy, extensive centrilobular haeroorrhagic necrosis of the fiver was seen, which is characteristic of ihe acute disease. Ilolrever, no occlusive lesions of the central vein of the LobuLes were described, and no obstructive lesions ltere seen in the larger hepatic veins or inferior vena cava. No mention was made of ascites. The basal ganglia showed kernicterus. Datta et a1. (1978 arb) reported 6 cases that occurred betneen L974 and L971. A11 of the patients had taken herbal medicines, identified in L case as lleliotropium eichwaldii, -198- which contained N-oxides of heliotrine. Two patients took the herb as an ext;ct of the whole plant, which contained an alkaloid concentration of 20 g/kE' for 20 and 50 days, respective!.y, and developed symptoms after a tine 1ag of 45 and 90 days, respectively. They were both estinated to have consumed 200 mg of heliotrine per day, the total alkaloid intake being 4 g and 10 g, respectively. They had taken the herb for treatment of epilepsy, and were admitted with acute onset of symptoms of abdominal pain, ascites, jaundice, hepatic encephalopathy, and gastrointestinal bLeeding, vhich suggested fulrninant viral hepatitis. They died within 2 - 12 weeks of the onset of symptoms. On1-y a brief description of the main autopsy findings in the liver was given, which indicated that there rtere changes characteristic of acute veno-occLusive disease of 1iver, including marked centri- Lobular haemorrhagic necrosis of the liver lobules, and occlusive lesions of the central and subLobular veins. It is interesting to note that both patients had also been on Long-term anticenvu!.sant phenobarbitone therapy. The remaining 4 patients had a chronic insidious onset of disease suggesting cirrhosis of the f-iver in 3, and alcoholic liver disease in one. One of the former had been taking some indigenous powder, presumably prepared from a herb, the indication and nature of which are not known. Ihe patient died from hepatic encephaLopathy. No detailed description of the autopsy findings was given, but it was stated that the central and subl.obular veins of the Liver showed chronic occlusive changes. The inferior vetra cava and large hepatic veins were patent. There was non-portal cirrhosis. A notabLe feature vas that the arsenic levels in the liver tissue were high (500 vg/kct nornal, 1 ug/ke). There is no nention of arsenic in the report on the analysis of the indigenous powder taken by the patient. Two of the remaining 3 patients with chronic disease, had taken herbaL nedicine for vitiligor and one for diabetes meLl-itus. The herb rilas identified as I{eliotropium eichwal"dii in the case of one of the vitiligo patients, who had taken it for 10 days and in rvhon the onset of symptoms occurred within 10 days. The herb was taken in the form of seed with an alkaloid content of L2 glkg. The daily intake sras estimated to be 500 mg of aLkaloid and the total intake, 5 g. This patient was admitted with a clinical diagnosis of cirrhosis of the Liver. The liver biopsy showed changes characteristic of acute VOD. Follow-up data are not known. The herb taken by the other 2 patients was not identified. The diabetic patient was being treated with oraL hypoglycaemic drugs and r,ras also known t.o be an alcoholic. The indigenous powder being taken by him contained a high concentration of arsenic (5 mg/kg). The results of haemo- dynarnic studies suggested hepatic venous outflow tract -199- obstruction of the intra-hepatic post-sinusoidal type in the smalLest hepatic veins. Liver biopsy showed characteristic centrilobular haemorrhages. Central veins could not be recognized. There were rniLd changes in the f.iver cells, but no alcohoLic hyaline nas seen. The Liver biopsy of the third patient showed characteristic features of acute disease with veneocclusion. Tno cases have been reported from China (Itou et a1., 1980). Both were adults who were takbn il1 after taking rnedicinal infusions prepared from Gynura segetum of the family Compositae (tribe, SLnecioneae). T6_e pd;ifng symproms and the cause of disease, as well as the pathological findings, were characteristic, except that one patient had jaundice and also portal vein thrombosis, which is not a usual feature of the disease. No chemical analysis of the plant was made and the alkaloid $tas not precisely identified. Furthermore, the total intake of alkaloid was not calculated. It has been stated that this rras the first report of such a case, but that it was possible that the disease might occur among adults more frequently r^rithout being reported. Ghanem & Hershko (1981) reported 3 cases of Arabs, one 3-year-oLd child and 2 adults, who were diagnosed as having VoD of the liver, on the basis of the clinical findings and morphological features of 1-iver biopsies. One more patient had occlusive lesions of both the sma1l and large hepatic vein radicles. They were among 29 patients with clinical features of hepatic vein thrombosis (HVT) found on a retrospective analysis of data from 9 major hospitals in Israel. Of these patients, 15 were Jews and 14 were Arabs. Notable features were that all Jewish patients r^7ere adults, whereas the rnajority of Arab patients were children below 10 years of age and primary HVT was 2.4 tirnes more cornnon among the l.atter. No analysis of the diet was made for PAs and their etiological role was suspected only on a presumptive basis. A survey of stored wheat grain in 9 villages shorrred that 2 samples were contaminated r^rith seeds of lolfu, belonging to the Graminae family, which r,rere found to contain 2 PAs (loline and norlo- line). However, these PAs are not known ro be hepatotoxic. The authors argued that, even though in a classical case of VOD there should not be thrornbosis of the larger hepatic vein radicles, the difference in the anatomical appearance of VOD and that of prinary HVT of the near-east type is not due to a different etiol-ogicaL agent but rather to a difference in the dose and rate of absorption of the ingested toxic compounds. A further report of the disease from Hong Kong, by Kumana et al. (1985), described it in 4 young Chinese women with psoriasis who took infusions of a herbal remedy, the toxic componeot of which has since been identified as Heliotrogig -200- lasiocarpun (Culvenor et al., 1986). They developed synptorog L9 - 45 days after starting the herbal treatment, and were examined 61 - 68 days after its initiation. The condition of patient No. 2, who continued taking the herb for 16 days after the onset of symptoms, deteriorated and she died of hepatic failure and was autopsied. The liver biopsies and autopsy confirmed the presence of acute disease in aLl patients. Patient No. 4 stopped taking the herb after 2L days, on account of a new rash. llhen assessed 77 days later, she had rnild hepatornegaly only. A detaiLed analysis of the al.kaloid content litas carried out for each case. The pyrrolizidine alkaloids were quantified as if senecionine based. The herb contained 0.42 g alkaloid/kg and 1.4 g N-oxide/kg. The daily intakes of aLkaloid base and N-oxide aere estiDated to b; L2 t l,ng and L8 t 4 .g, r?spectively. The respective cumrlative doses of alkaloid (base and ll-oxide) consumed by patients Nos 1, 2, and 3, up to onset- of sJrmptoms, nere calculated Co be 1350 mg over 45 days,900 rng over 30 days, and 570 over 19 days, respectiveLy. Patient No. 4 vho had irrefutabLe histoLogicaL evidence of disease but did not deveLop symptonatic disease, must have consumed 630 ng over 21 days. Patient No. 2, who died, was estimated to have taken a total amount of L380 mg alkaloid over 46 days. The estimated cumulative intake per kg body weight before the development of symptoms for patients Nos. l, 2, 3, and 4 was 26, !5, 12, and 15 ng/kg, respectively. It should be noted that, in patient No. 2, who died, the cunulative dose until the onset of synptoms riras the same as in patient No. 4, who was asynpto- matic. Moreover, the total intake by patient No. 2, was 23 ng/kg, which was lower than the intake by patient No.1, who survived. The authors compared the intake data of their patients with those of the Arizona chiLdren reported by stillnan et al. (L977> and Fox er aL. (1978). The 6-month-o1d baby, who survived but developed cirrhosis, and the 2-month-old baby, who died, are estimated by comparison to have taken cumulative doses of L2 - 25 and 11 mg/k1t respectively. The above data suggest marked variation in susceptibiLity among individual subjects. It is also known fron experimental animal studies that the young and new-born animals are particuLarly vulnerable (Jago, 1970). A case reported from the USA is ascribed to the consumption of comfrey (Sy*plZlgg sp.) powder, soLd as a digestive aid (nidker er-;1;-T985). A 49-year-old wonan presented with classical symptoms and signs of VOD. The haemodynamic data showed a hepatic vein wedge pressure of 3.07 kPa (23 rnmttg)with a sinusoidal pressure of 2.27 kpa (17 mmHg). Ilepatic venograms showed near obliteration of che smaller radicles of the hepatic veins during balloon distension of one of the intra-hepatic venous tributaries, and there -20L-

vras extra-vasation of the dye into the hepatic parenchyma. A porta-caval. shunt was carried out and the operative findings confirmed the presence of a post-sinusoidal block. The l-iver biopsy showed marked centrilobular necrosis and congestion with dilatation of the central veins and sinusoids' consistent with hepatic venous outflow tract obstruction. According to the clinical history, the patient had been a heavy consumer of food suppl-ements. Apart from several vitamins and mineraLs, she had been drinking 3 cups of camomile tea per week and for 6 months before admission had consumed I g/d'ay of a conrner- ciaLly availabl-e herbaL tea. For 4 months before admission, she had taken 2 capsules of lcomfrey-pepsintt pills with each meal. The herbaL tea and the piLLs ruere analysed for PAs. Pyrrolizidine alkaloids and their N-oxides were found, but the compounds \,rere not precisely ident-ified. On the basis of the snalysis of the PA content, the patient was estimated to have consumed a total of at least 85 mg of PA (HuxtabLe et a1., 1986) (14.1 ug/kg body weight per day, Ridker et al., 1985). The authors emphasized that the total PA consumption was relatively low. It lras possibLe that the patient had other sources of exposure and probably she had been consuming PA-cont,aining supplements for Longer than the periods stated by her in the clinical history. The latest is the report of a 13 year old boy from the U.K. who is stated to have developed symptoms of toxicity foll-owing administration of herbaL tea prepared from comfrey leaf (Syurphytum officinale) for treatment of inflarmnatory bowel disease for two or three years (Weston et al-.., 1987). The exact quantity of leaves consumed and frequency of administration were not known. The liver biopsy is stated to have shown a t'thrombotic variantil of veno-occLusive disease, though the inferior vena cava and the rnajor hepatic veins were patent on Doppler ultrasound and percutaneous phlebography. He had earlier been treated with predinisoLone and sulpha- sal-azine. The case is unusual in so far as the thrombosis of the central veins of the l"iver lobules, which is not a usuaL feature of veno-occl.usive disease of the liver.

~~7.4 VOD and Cirrhosis of the Liver~~

Jelliffe et al. (1954b) r,rere the first to dralt attention to VOD being an importaot cause of cirrhosis among Jamaican children. Prior to this, Hashem (1939), while reviewing the records of all cases of cirrhosis admitted to a childrenrs hospital in Egypt since 1933, described 3 cases of a special type of cirrhosis that was rare in adults. The clinical features and pathoLogicaL findings were similar to those in cases of VOD. They speculated that the cause was some metabolic toxins of gastroiatestinal origin. Royes (1948) -242-

suggested that the cirrhosis in the Jamaican chiLdren was very like the disease described from India and Egypt. The clinicaL and pathological features of l-00 cases of VOD among Jamaican children lrere described by Bras et al. (1954). Sixty-five of the cases rrere below L2 years of age. None of the 100 patients had cinhosis initially, but 5 showed occl-usive l-esions in hepatic veins and features of non-portaL fibrosis. Four of these 5 cases later developed cirrhosis. The authors concluded that VOD contributed to a substantial nunber of all cases of cirrhosis in this age group. Stuart & Bras (1957) studied 84 patients with VOD including 64 acute, 6 subacute, and 1-4 chronic cases. Tr,renty-three patients r,rere followed up, some for up to 5 years. Autopsy performed on 2L126 cases showed cirrhosis in 11 eases. Notable features lrere that 1 of the 6 cases of acute disease described in detail, deveLoped cirrhosis. Of the 2 cases with chronic diseaser l developed cirrhosis within 3 months of a 1-iver biopsy for acute disease, at which time the liver had shown hepatic venous occlusive lesions but no fibrosis. Autopsy findings were described by Bras & Watler (L955) in L9 patients aged 10 months - 45 years in different stages of the disease. Nine patients had cirrhosis that nas non-portal to begin r,rith but finally became indistinguishable from Laennecrs portaL cirrhosis. Rhodes (1957) studied the partern of liver disease among Jamaican children. A total of 193 f.iver biopsies was studied derived from 39 children who had one biopsy and 59 who had more than one at intervals of I areek - 3 years. Of the 14 autopsies on cases of VOD, LZ had cirrhosis. A notable feature rilas that the disease could occur asymptonaticaLLy with hepatomegaLy. The autopsy material from the University College llospital, Jamaica was analysed by Bras et al. (l-961). Of the 1560 autopsies, 28.52 cases concerned infants of less than one year old, mostly from poor, predominantly black families. Cirrhosis was seeir in 77 autopsies. ApproximateLy 3O% of. these 77 cases \irere diagnosed as post-VOD. The authors postulated that they might have resulted from the ingestion of -T;-tE6-Ifow-upCrotaLaria fulva or some other toxic substances. srudy of 61 parienrs who developed the disease in an outbreak in the USSR in 1958, 28 developed rrllepatoLeinaL syndromett (Braginski & Bobokhadzaev, 1965). Two of these cases, in whom the disease in the initial stages was not particularly severe, devel-oped cirrhosis within 4 years. Tandon, H.D. (personal cormnunication) has analysed the pathological data derived from the Afghan and Indian outbreaks on the basis of 61 Liver biopsies and 17 autopsies, including repeat biopsy studies on 15 pat.ients who r,rere followed-up for 1 month - 3 years after onset in the Lndian outbreaks. Three of the L1. patients, foll,owed up for 16 months or Longer for persist.ent cLinical evidence of disease, ended up with -203-

cirrhosis and 2 more had marked fibrosis with equivocal changes of cirrhosis in the biopsy. Notable features of the in study nere that the disease progressed to cirrhosis patients who were put on a normal diet, free from alkaloids, app"arance of symptoms of acute disease and did not receive"ft". any subsequent exPosure. Impact of alcohol- iotake was excluded. There was a poor correlation between the clinical and pathol-ogical severity of disease. Centrilobular haemorrhages, which are a sign of acut-e disease, ritere seen to were periisi for-r"1f. over one year in .patients-r some of- .whom lpprit"ttv At rieedle bibpsy, characteristic hepatic venous occlusions were not seen in nany patients in the acute phase of the disease, though they were seen in all livers at autopsy and most of them showed persistent centrilobular haenorrhages. Biopsy findings in cirrhotic livers were often not histolathoLogical-ly characteristic for any specific- forrn of cirrhosis. Featute! that might suggest the veno-occlusive etiol-ogy of cirrhosis at biopsy included paraseptal dilatetion of siriusoids and persistent haemorrhages or haenosiderin in the septa. rn- studies by Aikat et al' (1978) and Datta et al' (1978arb), 6 casei of VoD were reported following ingestion of herbal t.di"itt"" containing PAs. Four of the patients had symptoms of chronic disease and one of them developed non-portal cirrliosis. i)ne of the 2 infants from Arizona, USA, who suffered from VOD following the administration of PA-containing herbaL medicine, devel-oped cirrhosis during the- 8-month period followin! the applarance of acute symptons (Stillnan et'al' ' (1980) L977; Fox et aL., L978; Huxtable, 1980). Huxtable mentioned the death from cirrhosis of liver of a 62-year-old $rortran, who had consumed the same herb as the 2 infants for 6 months prior to her death. llowever, there rras no confirmation of the diagnosis of cirrhosis.

~~7.5 Differences between VOD and ffi(rcc)~~

people of A type-origin, of cirrhosis of Liver, peculiar to ttle Indian Indian Childhood Cirrhosis (ICc) ' has been ascribed to PA toxic etiology (Bras et aL., 1954i Rhodes, IgsD, owing to the observation of occlusive changes. in the aod sublobular veins of the liver., by some investi- ".otrrLgators (Radhakrishoa Rao, 1935; Prabhu, 1-940; Jelliffe et a1', tgSl1, Ranalingadwami & Nayak, 1969), though this is not a characteristis feature of the disease. The Presence of copper-positive granules in the hepatocytes (SalasPuro & Sipponen, 1976) has added to such a conjecture (Tanner & Poftmann, 1981), because of the reported aberration of coPPer -204- metabolisrn in experinental animaLs exposed to pAs (section o.4.rr). Ilowever, ICC and VOD are clinica!.ly and patho_ logicaLly distinct. rcc, confined to infants and chiidren, often affects siblings. Jaundice is a cortrnon sign and hepatospl-enomegaLy is a conmon feature. The disease is almost invariably fatal due to rapidly developing hepatocel-lular failure. Liver parenchymal changes are characterized by marked ball-ooning degeneration of hepatocytes, prominent deposits of alcoholic hyaline, severe cholestasis, and aggressive, pericellular- fibrosij (Nayak et a1. , lg6gir aLl features not characteristic of VOD: Moreover, occllsive changes in the hepatic veins are very rare and \{ere not obser_ved. by any member of the liver diseases subcorunittee of the Indian Council of Medical Research (1955), a critical study of the disease. "to-r"j"

~~7.6 Chronic Lung Disease~~

Ileath et al. (1975) reported the case of a l9-year-oLd African man who had died of congestive cardiac failure, and who r^ras suspected of having ingested a herbal remedy containing the seeds of Crotalaria I-aburnoides. Histopatho- logicaL examination of@characteiistic vasc_ular changes of severe primary pulmonary- hypertension. Powdered seeds of the plant were fed in the diei- to l,Iisrar albino rats for 60 days (ralle Lt). Characteristic features oj pu1-monary hypertension incLuding hypertensive vascul-ar changes in the lung, and right ventriculai- hypertrophy of the heart r^7ere produced in the animals showing- that tie seeds contained an agent capable of inducing pulmonary-theie hypertension in rats. Apart from this indirect evidence, was no proof of such a causal- relationship with the pulmonary hypertensive disease in the patient. A brief menrion.is made by McGee et a1. (1976) rrsomewhat of finding changes similar but rather more maturett (than those seen in the hepatic veins) involving some branches of the pulmonary artery in the l-ower Lobe of the l-eft l-ung, in a case of veno-occlusive disease of the !.iver caused by ingestion of PA-containing herbal teas. The alkaloids ltrere ttot further characterized. The changes were arso stated to be sirnilar to those produced in experimental animals by-involvement pAs. Apart from these 2 cases, there is no mention of of the pulmonary arterial system in any of the case reports available. The possibility - thar diet-mediated agents ,igtt induce pulmonary. hypertension in man has been discussed ai length by Fishman (1974). A parall.el was drawn with ttre epiaenr'ic ot pulmonary hypertension that occurred in Austria, tire Federal Republic of Germany, and Switzerland bet\^reen 1966 and 196g (Kay et a1., 197lb), which r^7assuspected of being caused by -205-

Aminorex, a compound that resembles epinephrine and amphetamine in chenical structure. Although the etiologicaL role of Aminorex could not be conclusively proved on epidemioLogical or experimental grounds, there continues to be 6 suspicion that agents taken by mouth can evoke pulmonary hypertension in rnan. Levine et al. (1973) reported the cases of 3 children aged 5 Llz - L3 years' and 11 months, respectively, with portal hypertension, nho developed progressive pulmonary hypettension resulting in cor pulmonale and death. In all 3 cases, there ntas evidence of extra-hepatic portal vein obstruct.ion confirmed at autopsy, and the s)rmptoms and signs of portal obstruction had appeared in early chiLdhood. They developed symptoms of car.dio- respiratory failure. Studies of pulmonary and cardiovascular function including haemodynamic studies of pulmonary circulation in 2 of the children suggested pulmonary vascular obstructive disease. At autopsy, no primary parenchymal lung disease was found. There were vascular changes of advanced pulmonary hypertension (plexiform lesions), but no evidence of thrombo-embolism was found. No factor tesponsible for pulrnonary vasoconstriction was identified. In one case, the Liver was stated to show coarse nodularity at autopsyr but the microscopic examination showed only patchy areas of portal fibrosis and regeneration. In the other 2 cases, there were only non-specific changes, with fibrosis in one. Ilortever' centrilobular congestion rras present in 2 cases. It is possible that some metabolites of toxic agents, such as PAs' which are metaboLized in the f.iver, rnight have blocked a metaboLic pathnay that ordinarily exerts a pulmonary antihypertensive effect, or that, by darnaging the liver, vasoactive substances, such as histamine, serotonin, and catechoLamines, might escape metabolic pathways to reach the Lungs and injure the pulmonary vessels. I{ortever' no such agents were identified. Kay et al. (L971b) made a p1-ea that, on the basis of the experimental data availabl-e, including the ability of several agents to produce pulmonary hypertension in experimental animals, and, in spite of the fact that pulmonary vascular disease has never been demonstrated in hunan cases of veno-occlusive dieease, careful enquiries shouLd be made on the possibility of patients presenting with unexplained puLmonary hypertension having ingested a plant product. A similar plea was made by lleath et al. (L975).

~~7.7 ?richodesma Poisoning~~

~~The disease rrOzhalanger encephalitis", which occurred in the Samarkand region of Uzbekistan, USSR in the per:iod 1942-5L, is believed to have been caused by contamination of -206-~~

food grain with the seeds of Trichodesna incanun (Shtenberg & orLova, 1955; Ismailov et al .,-TF6'F,-TfrEEGiEin 1.5 - 3.1,7" aLkaloids, mainly trichodesmine and incanine (Yunusov & Plekhanova, L959). CLinicaL features of the disease have been described by Ismailov et aL. (1970). This ourbreak differed from the others described above in that the primary sy.mptomatology lras extra-hepatic. Over 200 patients were affected, not including children bel-ow the age of 10, or the breast-fed infants of the affected mothers. Following exposure, there was a latent period of about 10 days, then vertigo and recurring headaches developed leading to nausea and vomiting. This was followed by generalized malaise, which progressed to delirium and loss of consciousness. Physical signs included pathological reflexes in 591l of the patients and paresis of the extremities and the faciaL nerve. Death was stated to have been caused mostl-y by respiratory depression. Of the 200 patients affected, 44 died. Auropsy findings were reJ-atively non-specific degenerative and necrotic lesions scattered in several organs including the central nervous system. No report of such a disease is available from outside the USSR.

~~7.8 Relationship Between Dose Level and Toxic Effects~~

In some recent human case reports, the PAs consumed have ben identified and estimates made of the daiLy and total- intakes (TabLe 16). The relationship between these intake Levels and the known toxicify of the alkaloids in rats has been discussed by Culvenor (1983) and Mattocks (1986). Discussion of the relationship between the dose level and toxic effects in human cases i" complicated, because the poisoning is generaLly due to a mixture of alkaloids found in naturally-occurring plant products, consumed as herbal remedies or food, and different plant species. There are wide differences in the acute toxicities of the al-kal,oids, which are the best available measure of comparative effects due to long-term intake as wel1. Furthermoie, estimates of intake can, at best, be approximate. When a large popul-ation is affected through the contamination of a food crop, as happened in the Afghan and Indian outbreaks (section 7.3), no precise estimates are possibl-e regarding the extent of contamination in different households, the amount of the contaminated grain consumed, the length of exposure resulting in the appearance of symptgms or signs of toxicity, or death, in the cases studied. There may also be various other contributing factors that are not apparent, e.g., food or cooking habits, on account of which no conclusive general,izations regarding the causative role of the toxic agent can be made. Reports on chemical analysis for the toxic agent and, hence, the amounts -207-

ingested, may not a1-waysbe reliable (refer to the controversy in the Afghan outbreak in section 7.3). In cases, such as the one reported by Ridker et al. (1985), when the total dose of PAs estimated to have been received by the Patient before the disease developed, was onLy a fraction of that of other alkaloids causing episodes of hurnan toxicity (Table 16), it is not certain whether that was the only toxic alkaloid agent that the patient was exposed to, or whether there were other contributing factors, particularly as the patient was stated to be a heavy consumer of herbs, vitamins, and natural food suppl-ements. Ln the known instances of human toxicity the principal a].ka1oidsinvo1vedarehe1iotrinefro'@,echimidine and related alkaLoids fron !gg!y939, riddelline and retrosine from Senecio longilobus, and crotananine and cronaburmine froor cr-oGGffible 16). Approxirnate acute toxicity valneT-G6ffira-ts) for these at-kaloid mixtures are 300, 500, and 50 mg/kg, respectively. For the mixture from C. nana, for wtrich there are no experimental data, the acute-ffity nas assumed to be simiLar to that of monocrotaline, 100 ng/kg. These relativities need to be taken into account in discussing dose-effect relationships for the PAs as a group. This has been done by discussing first the dose estimates for hel-iotrine, since it ruas the main aLkaloid in 1 epidemic and in 6 case reports (Table 16). Then in discussing the other a1-kaLoids, reference is made to a heliotrine-equivalent dose as well as the actual dose. The heliotrine equival-ent is;

tD5g of heLiotrine actual dose x LD56 of alkaloid mixture concerned The estimated daily intake in poisoning by heliotrine ranges from 0.033 mg/kg body weight in the Afghan epidemic (uohabbat et a1., L976), which after a period of about 180 days and a totaL intake of about 6 nglkg caused fatal-ities, to 3.3 rng/kg, which l-ed, in 2 cases in India (Datta et al., L978a,b), to death after 20 and 50 days and total intakes of 67 and 167 respectively. In between are 4 cases in llong Kong with mg/kg, 'intakes estimated daily ranging from 0.49 to 0.71 ng/kg (Kumana et a1., 1-983, 1985; Culvenor et al., L986). Three cases were non-fataL at total doses of LL - 27 rng/kg body veight and one was fatal at a total, dose of 23 ng/kg. These hel-iotrine cases imply that daily intakes are cumulative down to 0.033 ng/ke and may be fairly rapidly fatal above 0.5 ng/kg. Above a total dose of 6 - 15 rng/kg, VOD rnay become evident and sometirnes fatal. In the 2 cases due to riddelline and retrorsine in Senecio longilobus (stillnan et al . , 1977; Fox et aL., L978; nuG6'I4 198-DF'A estimated daity intakes were 0.8 - 0.17 and 3 ng/kg -208-

,j 9!N. N-ts 6 s - .J i ,i" a rd@dd 3 !ru x 6. ri t E; 3 oc .dN!!v t " t3 I '6:'i .g !u ..:ov.tarddor d> d ; i t3 ?6F€ o o r 6E 6 o o E r 6.c i:sH k 90 o g ! € I 't;it ! 5P*6FE;s .d! o g gEs E I $3rI3 E or! 90 oo OQ o !! B. a6 FO ol a !o oo 860 qql 90 o .dO - g! ! ^ u .d E0 Xq o 6d a 6 0 O dOA A oq o oo o o o o ooo o qo o0 t{o > !€ > € >!>> o h! a! oo o q F.q E lo ts I .o o ol .d € tsF oo n N O FOih I €€ N d .+ o dt 9g @ N o q!a Fo o B .o ) oH o o oi! 6 H n€ (, p o Q! E dU. Ja OF€ ! '.!.d o o F € oo $ o o oooo $ ooo 6 O 9O a O qo 6@tso 6 H 6 I $ T E}g o so o N oa o ilF E EME T q, 06 ! u ? .rkk o !oo o c>? { oFdo o 9^ org oo Q Oo. d ! o a qt o .o d 3 oo $ + 6€Oi O D truX 6 k6 d N6 d . {SdN N .d0d o d o -> .i Av o o o do 6 H !g @ > Hldd E6q O .r.O C! !l pl q6 6N E O € ON !OO'r o.'ol 9 66od I Eoo ! o oo d o' o @ J Q\ 6.JQF O nC 6 'd! t. d@oE O o oo € o o 605 dF oooo o H E ! o!d o o a.r 9 oioq ts ! trE o c H 56rC .l I Co tl El @d 3 .iB" t H9o k doG 3 , oo ts 3 @9ots o td o oo I O NN n il O NNNN \* @EO.t td .l .r d a! o ts -^ O O h 60tr @t E! d' t x a d d p qt 9 @t ) o | !! li d otd 3 6 €do J q€ E d sd o Q.d- FEO i oko o g d qdE o HE fiHHiE H a. oo 6 ,n dl iE;gigH€ ! E 3:3{E,;"o! x.d o a 5 5 3! €x 3! .i .F .rxEio.l! gg'JT;TEE8 E9id F 5 g E9 E f .io ozr: o EE'd SodsSAo ..'J;J;";,: HdopSdH

kd toq Nl !l olElol -209- body weight (equivalent to 3 and 1 mg heliotrine/kg, respect- iveLy, and the total doses were I-2 - 25 ar.d L2 mg/kg (equivalent to 72 - L50 and 72 mg heliotrine/kg, respectively). These leveLs are comparable with the highest reported intakes of heliotrine and, in infants, led to the rapid developntent of VOD and, in one case, death. In the epidernic due to mixed crotananine and cronaburmine in Crotal-aria nana (Tandon, B.N. et at., L976; Tandon, R.K. et al.S76l-ETEiamachari et a1., Lg71), rhe estimated daily intake of 0.66 mg/kg and the rotal intake of 40 mg/kg (equivalent to 2 and l2O ng heliotrine/kg, respectively) also corresponded to the highest intake of heliotrine. In the case of Symphyturnpoisoning (Ridker et al., .1985), the estimated dailyffil-f echimidi-ne and relared aLkaloidi was 0.015 mg/kg and the totaL dose 1.7 nglkg (equivalent to 0.009 and 1.0 mglkg heliotrine, respectivel-y). The dose levels are tower in equivalent terms than the lowest estimates in cases due to heliotrine by a factor of about 4 for dail-y intake and 6 for total- intake. The estimates $rere based on questioning of the patient and assay of the material concerned. It seems prudent to conclude that a daily intake of pyrrolizidine alkaloid as l-ow as the equivalent of 0.01 mg/kg hel-iotrine rnay lead to disease in humans. There is substantial overlap beEvTeen intake rates and total intakes for fatal and non-fatal poisoning. This presumably reflects the influences of a number of factots, such as individual sensitivityr aB€r nutritional status, and general health, but it i.s also due to the progressive nature of pytrolizidine toxicity and the effects of time. In the epidemics, in which some people died, only an estimated average intake is available and some who were alive at the time of investigation may have died later. Comparing the total intakes for human toxicity with the total doses up to death observed in the long-term administration of PAs to rats, 1.2 - 10.9 times the LD5g dose, equivalent to 36O - 327O ng heliotrine/kg (fable 10,-section 6.4.1.5), it is evident that human beings are more susceptible to the acute and chronic effects of the alkaloids than rats, sometimes markedly so. These considerations of the toxic effects in human beings of various intake levels could provide a basis for some assessment of the likely hazard from other types of exposure to PAs. For example, the consumption of comfrey root tea, estimated by Roitrnan (1981) ro contain 8.5 mg al-ka!-oid per cupr at the rate of 3 cups per day, or the ingestion of cornfrey leaf at the rate of one leaf per day, could lead to alkaloid ingestion rates of 0.40 and 0.016 mg/kg. These rates are respectively, much greater than, and equal to, the lowest daily rate causing veno-occlusive disease. Lower leve1s of exposure arising from such sources as the consumption of milk -2LO- from cows eating PA-containing plants or of honey derived from such plants, seems unlikely, in practice, to cause acute or subacute liver disease. Ilowever, care should be exercised. In an experimental situation in which covs erere fed Senecio jacobaea, the milk was reported to contain up to OTBFnE Effiiffi/titre. A 30-kg child drinking 0.5 litre/day of this rnilk could ingest 0.014 rng/kg alkaloid, equivalent to 0.028 rng heliotrine/kg (assuning an LD59 of 150 mg/kg for S. jacobaea aLkaloid). This is above the lowest daily rate leading to veno-occlusive disease and the lowest estimated total toxic dose woutd be achieved in 36 days. This level of conuaminacion of milk is undoubtedly extreme and there is no knowledge of any contamination of comnercial milk supplies. Iloney derived from Echiun plantagineum has been reported to contain up to 1 mg alkaloid/kg (Culvenor et al., 1981). A 30-kg child consuming 30 g/day of honey (a high eonsumption rate) would ingest 0.001 rng alkal-oid/kg body weight. The lowest estimated total toxic dose (L.7 mg comfrey alkaloid/kg, very similar to -FEchium alkaloid) would be achieved in 1700 days. Although seems f-ikely that consurnption of contaminated rniLk and honey would lead to acute or subacute liver disease, the possibility remains that they may contribute to chronic liver disease or liver tumours. The possibiLity of carcinogenic effects due to lonfterm exposure to PA-containing p1-ants has been discussed by Culvenor (1983). Some of the PAs invoLved in instances of human poisoning have been found to be carcinogenic in experimental animals (labLe 13). Data from some of the significant experimental studies lrere surmlarized by Culvenor (f983) with approximate estimates of PA dosages administered ro rats in terms of rng/kg body weight per day (Table 17). The dose rates that r^rere carcinogenic for rats (table 17) ranged frorn 2 to 6 mg/kg per day for an initial period atd 0.2 - 3 mg/kg per day for a remaining period of about L2 months, except in one study in which a dose of L0 rng/kg per day $ras used. It. can be seen that, in all- except tr^to instances of human poisonings sumarized in Table 16, the estinated daily rates of intake ranging fron 0.015 to 3.3 rng/kg body neight per day are within cLose range of those known to induce tumours in rats. In other reports, the consumption rates are above and below this range. Epidemiological studies to assess the carcinogenic role of PAs for man are not available. In countries with a high incidence of primary liver cancer, it is possible that PAs nray have an additive effect with those attributed to aflatoxin (Newberne & Rogers, 1973) and hepatitis B virus. The total- evidence now available r^Tarrants Long-term studies of the survivors of poisoning outbreaks, especiall-y where a substantial number of peopl-e were affected, as in the Afghanistan outbreak. -zlt-

~~@@ tsr oo~~

~~oo !! aro !!F^ .d .d N o o a qqooo EFi^dr HH@v@v FH FH>r600 og|dduo0d 99evdqd F!9h& oodx-! OO-.!q6O a, -1.1altsorkdt dl o I o E E€ :8 f; * ei oE o Eq€55>qriuE 3 Zza&ozG~~

~~o0 q o .io o rk SS@OOO o oJ NNdNts6A p ts!do odoE ms€€NOO o 4!>' N N d d il .1. E O!~~

~~d z~~

~~@ o pl^ .i o ox !d 6.E o do Oq roq ! !! F} AF d@do o E! o $,j* o-:t .dtro o e a xo NNHOONF06 o Or S ! 6g o d *3* 6 &>\ o *-9- o o Q.r S H <5E ":"n 1!- ev do o~~

d H @ot o o€qo-o * o f E ,3 f # - , 2 drH I 0 ji 6 F o ! ' a E : 9g { 3{. 3.t g o + x ig n sg. .: '> I a @ dN"3 N '3; o '€r .E ."s i .$: E "-: o :- iF t E -63 a3 6 o - .ff tq.9 -S bf .f ;" uo H o * B sTflB o :_E - E u Eo Bts *t N o .o -55 eB ts d ,} N$h 6 0g N_+q 6N 6P9!odO

~~o o~~

~~d o o k s o d E -2L2-~~

~~I 6^ i+$ 66 ood6 o lrrroo OvFNr €ooo rtdt o Cddv odd F. ! !9aili €oodd6 dddp!! k odsooo o !!! cEEooo s 33 r 5 I I ! EEEFF! OOO.d.i!d qqo!d!d5 6~~

~~6$NO$+ o \NdNiNN \ B r$O@in\ d- o o~~

~~xp ; tb -o g ! qq o' oo 3 o.i ^ oq !4 o..:*o o "-t lotosots; - t€ 'i - ,38 'P6; !Q -i 8f,n *fS 6.d o"i od o€ ts uo~~

F o ao x cl iul .i !91 '!tu 6 Fl r rif p ts d OC \O O O + !Q O>. Hts q! * il c dE -< - b o .F ! ! gk yq a Q x> o o o u Eu o c 9t O: ii> .i lrol i {.3 tB Og m! o N B N \3 d- - i3 E .:f, \ ^o s .:n ;fE"E ;E .?: .?: g gg '!S !o 'E: .ois" .i*'^ ^'a @ o o aco {.Jt {-9* !1 ;o o Oo O.ds dO @OO @O; Hd! _. ugt Eg5 (Oq.ON-NO ';3 o 5 N$n r$il 9r!d > o,5 o JO o '490 ts o.d.i a 9! UJ .iFx 6 6.P o O.d- g ! ds-a ,l€ utrc o oot ts A@@ slsl -2t3-

~~7.9 Pyrroliz.idine AlkaLoids as a Chernotherapeutic Agent for Cancer~~

The PA, indicine N-oxide derived from Heliotropium indicum, a widely used inligenous drug in Ayurvediffi{ has been found to have an antitumour activity and has been used in clinicaL trial"s as a chemotherapeutic agent for l-eukaemia(Letendre et al.,1981,1984; Cook et al., 1983) and soLid tumours (Kovach et al., L979a,b; NichoLs et al., 1981; Ohnnmaet a1., t982; Taylor et a1., 1983). Dosing scheduLes typicaLly nere 5 consecutive intravenous doses of 0.15 3 g/m" body surface area (approximateLy 2.5 - 5 ng/kg body weight) repeated at 4- or 6-week intervals (Kovach et a1., L979a; Letendre et al., 1981; Ohnumaet aL., L982). Hepatic toxicity, as judged by increased SGOP levels, was infrequent and rnild. llowever, subsequent trials with this agent have indicated more serious hepatotoxicity. Io a more recent report by the same vorkers (Letendre et a1., 1.984), 5 out of 22 casee of refractory acute leukaemia, treated with indicine N-oxide, had severe hepatotoxicity, presumably induced by the drug. One of these patients had been treated for 18 months with rnethyl,-testosterone, 4 months prior to receiving indicine N-oxide., Synptons of severe hepatocelLular faiLure appeared in 3 patients after the initiaL course of treatment. This occurred after 4 daily doses of 3 g/m2 surface area in one patient and after 5 daily doses of 3.75 g/n2 surface area in 2 patients. Two other patients vho had received 3 g/-" surface area daily for 5 days developed hepatocellul-ar failure after the second course of treatment, one at 3.3 glm2 and the other at 3.75 g/m2 surface area, daily, for 5 days. In each patient, the onset of hepatic disease was rapid and the course was downhill. Livers of 4 of these patients examined at post-mortem sholred severe centrilobular vascular congestion with necrosis of parenchymal cells, and, in one patient, a few subLobular veins were found to be occluded. Miser et aL. (l-982) reported severe hepatotoxicity in 4 of 45 children treated with indicine -Similarly,N-oxide for refractorv !.eukaemia or advanced solid tumours. Cook et aL. (1983) reported the case of a 5-year-oLd child with acute myeLoid Leukaemia who deveLoped severe hepatic failure within 3 days of starting the treatment. Autopsy showed massive hepatic necrosis. However, it shouLd be noted that no hepatic failure was reported'in more than Loo adults with soLid tumours, treated with the same agent (Kovach et aL., L979arb; NichoLs et a1., L98L; Taylor et 61. , L983) . No hepatotoxic effects rirere reported by Ohnuma et al. (1982) amoflg 37 patients who received this drug for soLid tumours. The najor toxic effect was myelosuppression (Kovach et a1., 1979b). -2L4-

~~7.10 Prevention of Poisoning in Man~~

At present, prevention of poisoning can be achieved only by reducing or eliminating ingestion o,f the alkaloids. The two effective procedures are control of PA-containing pl.ants in agriculturaL areas and educational prograrmnes directed to the populations at risk. The control of pLant populations for this purpose has been carried out onLy in Uzbekistan, USSR, following the epidemics of human disease due to contamination of grain by seeds of EeliotTopium lasiocarpur.n and Trichodesma ineanum. The following measures were introduced and have been effective in preventing further outbreaks :

1. A state standard \das set for the quality of seed grain, which must be certified by a State Seed Inspectorate. Current standards prohibit the sowing of wheat, rY€, barley, or oats contaminated by seed of Heliotropium lasiocarpum or Trichodesma incanum.

~~2. A state standard was set for the quality of grain stored for food. The f.irnits for Ilel ioEropium lasiocaroum and Trichodesma incanum seeds are 0.2% and zero, respectively.~~

3. Agricultural (agritechrrical) measures to ensure minimum contamination of crops and harvested grain, including specification of the most suitable methods and tirning of cultivation, use of clean seed for sowing, weediag of crops prior to maturing of the grain (towards the end of May), and mechanical cleaning of grain.

~~4. Methods for monitoring 1-evels of contamination of flour, bread, and similar products.~~

5. Publication of educational booklets describing the biological, environmental, and morphological characteristics of the toxic weeds, their pathways of distribution and the causes of the toxicoses experienced.

~~6. Promotion of weed control by governmental authorities and provision of legislation to enforce the control measures.~~

In other countries, the control of some PA-containing weeds in crops is practised by cul-t.ivation and herbicide treatment, in order to maximize yiel-d and the general quality of the gtain. In pastures, animal management and herbicide treatments are used to increase pastures and reduce poisoning of animals. Specific treatment methods differ according to the plant species and the circumstances. General references v,7erenot available to the Task Grouo. -2t5-

In AustraLia, where Heliotropium europaeum and Echiurn pLantagineum are widespreail weeds in vheat-growing areas but where nornal agricultural practices prevent all but occasional minor contamination, relevant tolerance standards for wheat delivered at storage sil-os are not specific. Ileliotropiun ggggry seed is rarely seen and would form part of the "unmillable naterial-" the seed component of which caa be up to 1% of the volume of the wheat. Seed of Echium plantagineum is occasionally seen in delivered grain at Levels of up to 10 seeds per half litre, the tolerance level for this seed fraction being 50 seeds per half 1itre. -2L6-

~~8. BIOLOGICAL CONTROL~~

BioLogical control methods have been investigated for severaL .PA-containing pLant species, notabLy Senecio jacobaea, He1iotropiumeuropaeum.Echiurnp1antagineu'.-aiFichodes'a lncanum. The effectiveness of such methods is variable and fr?l-iesults may be conf ined to certain regions where favourable conditions exist. For example, in control- programnes against S. jacobaea in Australia, Canada, New Zealand, and the USA, using 3 insect species, results varied from virtually nil to nearly 100% control (Julien 1982). The effects of the introduction of the cinnabar moth (Tvria iacobaea L.) on S. jacobaea in these countries havea;;; summarized as in Table L8.

~~Table 18. Results of the attempted control of Senecio jacobaea (ragwort) with the cinflabar moth-~~

~~Country or region Resul t~~

~~Austral ia Establishment precluded by Predation' parasitism and disease~~

~~Western Canada Moth populations stabilized below that required for cortrol~~

~~Eastern Canada Establishment and subsequent notable reductions in ragwort levels~~

~~New Zealand Marginal establishnent, moth population linited by predation and parasitism, little inpact on ragwort~~

~~USA I{idespread establishnent, ragwort levels sometimes reduced at localities near the linits of its distribution~~

Several agents are being tested in Australia for the control of 11. europaeum and one spec.ies, a flee beetle Longitarsus ;;i3-i;F-has been reieased (Julien, L982; Delfosse, 1985). There are good prospects in this country for the biological control- of Echium plantagineum and 2 other Echiuur spp., r^rith 8 insect species approved for release; when legaL restrictions are f.ifted (Delfosse & Cu1Len, 1985a,b). Preliminary studies have been made on the biological control of Arnsinckia and other Senecio species (..g., Pantone et al., 1e85)--- Given adequate funding, PA-containing plants are a suitable target for bioLogicaL conErol. -2L7-

~~9. EVALUATION OF I{I'I.,fANHEALTTI RISKS AND EFFECTS ON THE ENVIRONMENT~~

~~9.1 lluman Exposure Conditions~~

~~9.t.L Reported sources of human exposure~~

~~The two main sources of exposure of human beings to toxic PAs that have Led to najor outbreaks of poisoning with high mortal-ities as well- as to individual cases in several countries are: \~~

~~(a) the contamination of cereal grains, such as wheat and rniLlet , rnrith the seed or other parts of pLant s containing alkal-oids; and~~

~~(b) the consumption for medicinaL or dietary purposes of herbs containing the aLkaloids, either as the pLant itself or as infusions.~~

Consumption of contaminated grain is more likely to occur in regions where food is in short supply, and particularly when drought favours infestation of the grain crop by PA-containing weeds. A qualitative field test for detecting the presence of toxic pyrrolizidine alkaloids in plant materials, using sirnple Laboratory methods, is now available (section 2.2.2.5).

~~9.1.2 PLant species invol,ved~~

Pl-ant species containing toxic PAs occur throughout the world and are known in 47 genera in 6 plant farnilies. As many as 6000 species are potentially PA-containing. The most important genera responsible for human and animal disease are Senecio and other genera.of the tribe Senecioneae (fanily Edffiitae), crotalaria (farniLy Leguminosae) and Heliotropiurn, rri-cjqg

~~9.1.3 Modes and pathways of exposure~~

~~9.1.3.1 Contamination of grain crops~~

Large outbreaks of poisoning have occurred through contamination of wheat crops in Afghanistan, India, and the USSR. In particular, 3 species of Boraginaceae (tt.ti.tr.pi"t l-asiocarpum, H. popovii, and @) are well adapted to vigorous growth under the cl-imatic conditions in which wheat is usuaLly gro$rn. Contamination can be effectively cootroLled in wheat produced using modern harvesting techniques and grain seed that is inspected and controll-ed for r4reedseed contamination, but control of contamination is more difficult where these conditions cannot be met. Contamination of staple food grain is of particular concern, since entire populations are exposed, and, control may not be possible, if the peopLe are not aware of the hazard that PA-containing needs present.

~~9 .L.3.2 llerbal medicines~~

Herbal preparations containing PAs are used as tonics, treatments, preventatives, and food supplements. Such usages are so r.ridespread that they are nearly universaL. Many are traditional, whiLe others reflect a rejection of, or lack of access to, standard health care services (section 3.3.2). Veneocclusive disease was first recognized as a clinical entity in Jamaica as a result of the medicinal use of PA-containing herbs prepared from Crotalaria. CrotaLaria- containing herbs have also been @le Er-human poisonings in Barbados, Equador, and other locations in the West Indies. Heliogggglgg herbs have been reported to cause poisoning in tt-iffii!-and=nave India. Symphytuur and Senecio- containiig herbs liven rise to cZ-reports in ttre-. Other reports of PA poisoning are knovm in which the herbs used were not botanicaLl"y identified. -PA-poisoning has been associated with both home-prepared and comibrciaL!.y available herbs, the Latter including prescriptions by herbaliets (Westonet al., 1987). Various other genera of PA-containing plants in the farnilies Boraginaceae, Compositae, and Leguminosae are also -2L9- widely used as herbs. No case rePorts are availabl-e for these genera. Reported cases of PA poisoning due to the use of the herbs are geographically widespread, but few in number. However, the scale on which PA-containing plants are used as herbs, the typicalLy delayed effects of long-term exposurer and the difficuLties of diagnosis led the Task Group to conclude that there is every indication of under-reporting of intoxications from the use of such herbs. Symphytum root preParations, in particular, represent a hazarifr-ai?l-Ertain user groups are routinely exposed to Levels of Symphytum alkaloids that are higher than those at which intoxications have been reported. The risks associated rnrith the use of PA-containing herbs are accentuated by the difficuLties of controLLing this use.

~~9.1.3.3 PA-containing plants used as food and beverages~~

Some PA-containing pl-ants are used for food or the making of beverages in many countries, including developed countries. The following species are knovrn to be used (though many other plants are also probably used in this way): Caca!11lgErrb11!, Symphytum species, Ligql-aria .d.entata, @, -GFio-6,rrche1lii,vu!vrrErtrr, ruqvv9eruErlor y^vrvLerr Ef"ei-l-e"is. l.ht* laburnifolia,labrrrnifolia- C.c- oumila-pumila, C. recta.recta, and C. retusa. No Tnt-;'ETo-t- is ZEiTdS"[f on Effi t to wh i6GETT-f f erent types are consumed(section 3.3.3).

~~9.1.3.4 Other foods contaminated by PAs~~

SeveraL species of Boraginaceae are nectar and pollen sources for bees. Echium pl-antagineum, in particular, is a widespread weed in sffi-E6GEffilffi substantial source of honey containing a low level of alkaloid. Senecio species are also visited by bees and yield alkaloid:Gn-taining honey, though Senecio-derived honey is not knovtn to be produced in quantitt Tor saLe. Thus, some regional and l-ocal popuLations are exposed to a low-level intake through the presence of PAs in honey, and surveillance may be desirable in countries producing honey (section 3.3.4). Under experimental conditions, PAs are transmitted from the feed of dairy cotts and goats into the milk. Some PA-containing species, such as Senecio jagobeee, S. lautus, and Echium planiagineum, are weeffiaTft-T-asturffi-n some counffi by cattle under certain conditions. There are no published reports of alkaLoid in mil-k supplies for human consumption (section 3.3.5). -220-

The Task Group was not aware of any reported cases of pyrrolizidine toxicity that had been ascribed to either honey or dairy products. No infornation was avaiLable to the Task Group on the possible presence of alkaloids or their metabolites in meat from animaLs that had consumed pA-containing plants shortly before slaughter. The results of metabolic studies in rats have indicated that the alkaloid is rapidly cleared from the body and, therefore, the Levels of pAs in meat are expected to be very low. However, there is no information on the possibility of alkaloid accumulating in storage sites.

~~9.1.4 Levels of intake~~

Reliable estimates of !.eveLs of intake of pAs, especialLy in outbreaks of disease caused by the contamination oi cereal crops with the seeds of toxic plants, are extremely difficult to make. Sanpling of .the contaminated grain may not be strictly representative, since the extent of the contamination may vary in different sites and households, as is evident from the estinates of PA intake in the Indian and Afghan outbreaks reported in section 7.3. Furthermore, no accurate record is possible of the amount of contaninated food consumed over an uncertain length of tirne. No records of the levels of toxic PA intake are avaiLable in the earlier reports of human toxicity. Where available, estimates of intake in outbreaks caused by the contamination of staple food sa6ps have been made on the basis of random sampl.ing of the contaminated grain in food stores, and rough estimates of daily consuarptiJn by average adults. Food-on-the-pLate analyses have not been made. The estimated lengths of exposure, and hence the amount of totaL intake, are also approximate. The contamination of cereal grains with the seed of PA-containing plants has caused rnajor epidenics of human poisoning, though, in the two instances where estinates of alkaloid intake are avai!.abLe, the intakes were lower than in some exposures due to the use of herbaL medicines. The estimated intakes are surmnarized in Tabl-e 16. In an outbreak in India, millet contaminated lrith CrotaLaria nana seed had an average alkaloid conrent of 0.5 g/k;-;aE;-;;irnated dai!.y intake by the population was 0.66 rng/kg body weight. In a Larger outbreak in Afghanistan, due to the seeds of He+otropium_pgpoyii in wheat, the leve1 of contamination rras probably variable; representative sampLes of wheat contained alkaloid at. O.O4 glkg. The estimared daily intake rras 0.033 rng/kg body weight. These intakes, sustained for periods of approximately 2 and 6 rnonths, respectively, resuLted in typical acute and subacute veno-occlusive disease. -22L-

The highest intake rates have been associated riTith the use or misuse of medicioal herbs and have resulted in acute liver damage and death. Ln two occasionsr the consumPtion of ttelio- tropium eichwaldii as a treatment for epilepsy led to aa esiTmaEEd-inta[ of 3.3 ng/kg body weight daily for 20'or 50 days, and the use of extracts of Senecio longilobus as t"ai"it for young children led to estii?Finta-lffiffi and 0.8 - " L.7 rng/kg body weight. The highest intake led to extensive liver necrosis. However, it is possible that, in the above case of poisoning by Heliotropiurn eichwaldii, the ioxicity was eniranced due io ;imffin of phenobarbitone, which has a potentiating effect on the microsomal enzymes in the liver ceLls that convert the PAs to toxic metabolites. The use of Heliotropiun lasiocarpum as a component of a herbal treatrnent@ somewhat lower daily intake rates of O,49 - 0.71 mg/kg body weight in 4 patients' who, after periods of L9 - 46 days, developed veno-occLusive disease. The patient with the longest intake period and a total intake of 1.4 g alkaLoid or 23 mglkg body weight died. The Lowest ingestion rate leading to a case of veno- occlusive disease was also due to medicinal herbal treatment orr more specificalLy, to the use of a digestive aid containiog a preparation of comfrey root. Cormnercial herb and food supplement preparations containing comfrey leaf or root are on sale in many countries. Limited assays of one comfrey- pepsin preparation prepared from comfrey root indi-cated a PA content of. 2.9 glkg. Another preparation made from cornfrey leaf contained up to 0.27 g alkaloid/kg. The consumption of these preparations led to an estimated daily intake of 0.015 mg/kg body weight. Veno-occl-usive disease was diagnosed after a 4- to 6-month period. The consumption of Syrnphytumofficinale (comfrey) and S. x uplandicum (Ru;sian courffif food, infusi6i{ F-other preparations is widespread, though the full extent cannot be estimated. A high 1evel of consumption as salad appears to be about 5 - 6 leaves per day and consumption as comfrey tea probably reaches a sinilar level. Liniced assays indicate that the average alkaloid content of the leaf is about 1 mg/leaf, the concentration being higher in the younger, smalLer leaves. The alkaloid intake from comfrey Leaves couLd therefore vary from a Low value, up to about 6 mg/day; or 0.1 mg/kg body weight for an adult, an intake within the range producing veno-occlusive disease. However, the Task Group noted that some people claim to have consumed comfrey at such a rate without suffering any disease. Overall, the estimates of intake of PAs by human beings (Table 16) indicate that ingestion rates above 0.015 ng/kg body weight for the mixture of echimidine and related _222_

alkaLoids in comfrey nay lead to acute or subacute Liver disease. If expressed in terms of equivalent doses of heliotrine (section 7.8), the estimated totaL doses in the known outbreaks or cases of veno-occlusive disease riog.- tro, I to 167 mg/kg body weight. There is r.ittle real dif"ference in^the.ranges of estimated total doses in non-fatal cases-ii-_ (1 _ 120. rnglkg_-body weight) and those Leading to a..tt L67 mgl,"g). These figures, when compared with the total LethaL dose of several pAs in rats, i.e., - (equivalent L.2 10.9 tirnes the LD

~~9.2 Acute Effects of Exposure~~

~~9.2.L Acute liver disease~~

AlL cases of human intoxication in reported accounts have tl:, t":::-phase of the disease, ihe dorninanr symptom l:::-i"oe-rng rapidly filling ascites. The disease can affect large subpopulations 1nd, in one study, up to 22.6% of the population was affected. Children appear to be the most vuLnerable group and nortality can be high at the extremes of age. the- Liver is principal- ll" target organ. In the acute stage of the disease, the liver shows a characteristic centrilobular haemorrhagic necrosis, which in man is accompanied by occlusion of the hepatic veins. However, characteristic veno-occlusive lesions, seen in the centra!. veins of hepatic lobules, rnay not always be evident in the needle biopsy - 223' examination of the liver, but are always aPParent on examination of the autoPsy material.

~~9.3 Chronic Effects of ExPosure~~

~~9.3 .l- Cirrhosis of the tiver~~

There is evidence that the administration of a singLe dose of PA to experimentaL animals of, a single ac:tt:.episode.of illness in man, fol-lowing brief consumption of PA-con?inxng herbs or PA-contaminated food, may lead to progresslve cnronac liver disease resulting in cirrhosis. Cirrhosis may also be a consequence of long-term l-ow-dose administration of PAs to rrt"l aninaLJ and possibly aLso of long-term low intake of""p.tit PAs by human beings, though there is no proof of the in latter. airrhosis resulting from the toxic effects of PAs the advanced stage, m:ly not be distinguishable from that resulting from other causes (sections 6.4.1.5 and 7 '4) ' The Group did not find any evidence suggesting that PAs. Task -",Te " causative f;ctor of the specific disease, Indian childhood Cirrhosis (section 7.5).

~~9.3.2 Mutagenicity and teratogenicity~~

Several PAs, PA-derivatives, and related compounds have been shown to produce chromosome aberrations in plants and (.salmoneLla severaL cell culture systems, mutagenic effects iiAnetstt), sister chromatid exchanges, and other te-s-fffiiE i"."tog"ttl-" and fetotoxic effecti in experimentaL animals (sections 6.4.5r. 6.4.6t 6.4.7). Chromosomal aberrations have been reported iir the blood cells of children suffering from veno-ocClusive disease, beLieved to have been caused by fuLvine. The Task Group was not artare of data on the teratogenic/fetotoxic effects of PAs on human beings and was unable to evaluate the potential for these effects in PA exPosure.

~~9.3.3 Cancer of the liver~~

A reLatively large number of people have been exposed, in the past, to PAs and have suffered acute and chronic toxic effects. Ilowever, no information is available on the long- term follow;up of these populations' to ascertain whether this type of .*pos.rt. could havl resulted in an increased incidence ol 1in"t cancer or other tyPes of cancer. Because of this lack of knowledge, it is not possibLe, at Present' to make an evaluation of the cancer risk due to PAs. I{owever, various PAs have been shown to be carcinogenic for experimental- animaLs, which inplies that a potential cancer risk for human beings should be seriously considered. -224-

pAs 0f several evaluated for carcinogenicity by IARC (I976, 'rsufficient 1983), rhere is or lirniied for the carcinogenicity in experimental animals (IARC,".rid"rr"L" lg76) of monocrotaline, retrorsine, isati.dine, lasiocarpine, peta_ sitenine, senkirkine, and of extracta of the pA-containing plants Petasites j=aponi_cum, TussiLago farara, S)rmphytum g$ j.sinat@s-, @_g"***I ilffii,,m- Japonlcum, and Selsg.o cannabifolius. These studFwere", carrled out marnly on rats, with few studies on mice or (section hamsters 6.4.9). The carcinogenicity data obtained with other PAs are difficult to evaiuate, iec.rr"e of the lirnited number of treated animals and the lack of adecuate numbers as controls. The main target organ is the 1iver, where liver cel1 tumours and haemJngioenlotheliar, ..r"oo,"" were observed. In some instances, tumours in extra-hepatic (1ung, tissues pancreas, intestine)-were aLso observed, namely with monocrotaline, retrorsine, and lasiocarpine. Some pAs, for example, retrorsine, have been shown to be carcinogenic after a single dose. The pyrro!.ic metabolites have also been shown to be carcinogenic for rats. 1t may be recalled that several of the pAs involved in human poisoning include the above compounds. It is notabLe that the dose rates that have been effective in inducing tumours in rats, mostly equivalent to 0.2 - 3 rng/kg body weight per day (table t7), are roughly similar in ,ag-niirrde to estirnated intake rates (0.49 - 3.3 rng/kg body weighi per day) (Table 16) in several episodes of human toxicity.- Comparisbn of the total- intakes resulting in human toxicity wiih the Eotal doses to death observed in the chronic toxicity studies on rats indicates that human beings are more susceptible (section 7.8) and suggests that human beings may survive for sufficient time to develop cancer after only a briet exposure at this Level or a longer exposure at a markedly lower ieveL. A more quantitative assessment is not possible on the basis of the available informationp and the Task Group stressed the need for appropriate epiderniological studies.

~~9.3.4 Effects on other organs~~

Substantiated reports of pA-induced exEra-hepatic injury in man are lirnited to Trichodesma intoxication, in ,fri"t, symptoms and signs were prE?ililffi neuroLogical.. The range of - organs affected by other pAs in experirnental and farrn animals suggests that exposure of hurnan beings to other pAs may also carry the potential for extra-hepatic injury. There are extensive reports of experimentJl studies in which PAs have been demonstrated to produce the characteristic vascular changes of primary pulmonary hypertension and consequent right ventricul"ar hypertrophy of the heart in rats and non-human primates (section 6.4.2). Susceptibility is age -225- dependent, veanling rats being more vulnerable thao older animaLs. There is only circumstantiaL evidence of PA-induced pulmonary vascular disease in one patient (section 7.6), but judging by the experimental evidence available, it is possible that human beings may be susceptibLe to PA-induced cardiopulmonary changes. In the opinion of the Task Group, the neurologicaL involvement which is a dominant feature in PA-intoxicated horses and is al.so seen in cows and sheep, cannot be explained solely as a consequence of liver damage. Central oervous system lesions have been demonstrated in sheep, pigs, and rats. Distribution studies of the radiolabell-ed metabolite, 3tl-dehydroretronecine, show increasing accumuLation of radioactivity in the brain with tirne. Trichodesma alkaloids, structura!.Ly reLated to rnono- crotEffi-ra neurotoxic agents. Trichodesna toxicosis in man has been report,ed only from ffiA-ffi together rnrith several studies on experimental animaLs. Detailed reports on the pathological findings vrere not avaiLable to the Task Group, but the infornation avaiLable indicated that the central nervous system lras the primary target organ (sections 6 .4.3 and 7 .7) . Stomach and intestinal lesions have been shown in PA- exposed sheep, mice, cons, and rats. Distribution studies with radiolabel.led pyrroLes showed a high retention of radio: activity in the stomach, consistent with the acid-sensitive nature of the pyrroles. In rats, pymolic metabolites are secreted in high concentrations in the bi1e. Kidney changeb foLlowing to PA administration have been shown in mice, pigs, horses, sheep, and monkeys. Pyrrol"izidine metabol-ites have been found covalently bound to kidney DNA in rats. Urinary excretion is a rnajor route of excretion of netabolic products of PAs in rats. There is no evidence of involvement of organs other than the liver and central nervous system ascribed primarily to PA toxicity in any of the published human case reports. It is possible that under some circumstances, other major organ systems may also be at risk. As bioactivation of PAs has been demonstrated onLy in the Liver, the risk of damage shouLd be expected to be lower in the organs.

~~9.4 Effects on the Environment~~

~~9.4.I AgricuLture~~

In some countries, PA-containing weeds densely cover areas of up to thousands of square kilometres. Their adverse effects include the covering of pastures, additionaL costs in agricultural production, and the poisoning of farm anirnal"s. -226-

The toxicity of PAs for farm animals, incLuding sheep, cattle, horses, pigs, goats, and poultry, which has been the inspiration for much of the investigation of PA toxicity. In AustraLia, for example, lleliotropiun and Echiurn plantagineum cause the d Ls aniiEffi fs€-?ffi 6-7).

~~9.4.2 l,lild-life~~

By contrast, littLe is known about the consumption of PA-containing plants by wild-1ife, or of their individual sensitivities. The death of deer in Louisiana has been ascribed to eating Heliotropium or !@ig species, and an experimental study has shown that the rainbow trout (Salmo gairdneri) is sensitive to Senecio jacobaea aLkaTo-ffi

There is no information on the effects of the alkaLoids on field rodents or other seed-eating mamals. and birds that rnight be expected to consume seeds of PA-containing plants and to suffer toxic effects.

~~9.4.3 Insects~~

Many species of insects, such as some moths of the farnily Arctidae and butterfLies of the sub-farnilies Danainae and TTfrii-inae, have become dependent on PA-contaii-ing plants, using the alkaLoids as defensive chemicals and derivatives of them as pheromones and other signalling chemical.s.' Thus, compLete elimination of PA-containing p1-ants in a region might l.ead to a marked reduction in the Loca1 population of insects of this type (section 6.5.3).

~~9.4.4 Soil and water~~

There have not been any studies on the fate of PAs when the plants in which they occur wilt and age. If alkaloid is leached into soil or water, it is probably readiLy degraded by microorganisms since, as a base and ester, it is subject to oxidative and hydrolytic reactions. -227-

~~REFERENCES~~

AFZELIUS, B.A. & SCIiOENTALtR. (1967) The ultrasrrucrure of enlarged hepatocytes induced iri rats with a single oral dose of retrorsine, pyrrol-izidine (Senecio) alkaloids. J. uLtra- struct. Res. , \: 328-345.

AIKAT, B.K., BIIUSNURMATII,S.R., DATTA, D.V., & CITHUTTANI,p.N. (1978) Veno-occlusive disease in north-west India. Indian J. Pathol. Microbiol. , 4: 2O3-2L1. AL-HASANY, M. & MOHAMED,A.S. ( r.970) Veno-occlusive disease of the liver in Iraq. Arch. dis. child. , 45: 722-724.

~~ALLEN, J.R. & CARSTENS,t.A. (1968) Veno-occLusive disease in rhesus monkeys. Am. J. vet. Res., 29: L68L-L694.~~

~~ALLEN, J.R. & CARSTENS, L.A. (1970) Pulmonary vascular occlusions initiated by endothelial lysis in monocrotaLine- intoxicated rats. Bxp. mol . Pathol-. . , f|: 159-171.~~

ALLEN, J.R. & CARSTENS,L.A. (1971) Monocrotaline induced Budd-Chiari syndrome in monkeys. Arn. J. dig. Dis., !1: 111-121. ALLEN, J.R. & CHESNEY, C.F. (L972) Ef fect of age on development of cor pulmonale in non-human primates fol!.owing pyrrolizidine alkaLoid intoxication. Exg.'_91. Pallol., !: 220-232

~~ALLEN, J.R., CIIILDS, c.R., & CRAVENS,W.W. (1960) Croralaria spectabilis toxicity in chickens. Proc. Soc. Exp. Biol. Med., LO4: 434-436.~~

~~ALLEN, J.R., LALICII, J.J., & SCHMUTTLE,S.M. (1963) Crora- l"aria spectqbills induced cirrhosis in turkeys. Lab. rnffi Tzz 5 L7.-~~

*-""r", J.R., CARSTENS,L.A., & OLSON, B.E. (1967) Veno- occlusive disease in Macaca speciosa monkeys. $|-$!I$, 5A: 653-667 |

~~ALLEN, J.R., CARSTENS,L.A., & KATAGIRI, c.J. (1969) Hepatic veins of monkeys with veno-occLusive disease-sequentiaL ultrastructural changes. Arch. Pathol., 87; 279-289.~~

~~ALLEN, J.R., CHESNEY, C.F., & FRAZEE, W.J. (1-972) Modifications of pyrrolizidine alkaloids intoxication resulting from altered microsomal enz)rmes. ToxicoL. appl. PharmacoL., 23: 47Q-479. -228-~~

~~ALLEN, J.R., IISU, I.-C., & CARSTENS,L.A. (L975) Dehydro- retronecine induced rhabdomvosarcomas in rats. @_8, 35: 997-I0O2.~~

~~AMES, M.M. & POhlIS, c. (1978) Determination of indicine N-oxide and indicine in plasma and urine by electron-capture gas-liquid chromatography. J. Chromatogr., t6t 5L9-526.~~

~~ARMSTRONG,S.J. & ZUCKERMAN,A.J. ( 1970) Production of pyrroles from pyrrolizidine alkaloids by hurnan embryo tissue. Nature (Lond.) , 2j2 569-570.~~

~~ARMSTRONG,S.J., ZUCKERMAN,A.J., & BrRD, R.G. (L972) Induction of morphol-ogical changes in human embryo l-iver cells by the pyrrol-izidine aLkaloid lasiocarpine. Br. J. exp. Pathol., 53t 145-149.~~

AROM, R.R., PYARELAL, GI{OSH, T.K., UATI{UR, K.K., & TANDON, B.N. (1981) Epidemiology of veno-occlusive disease in tribal population of Madhya Pradesh and Bihar. J. conmun. Dis. (India), 13: 147-151.

ARSECTLER.{[NE,S.N., ctNATTLAKA, A.A.L., & PANABOKKE,R.G. (1981) Studies on medicinal plants of Sri Lanka: occurrence of pyrroLLzidine alkaloids and hepatotoxic properties in some traditionaL medicinal herbs. J. Ethnopharrnacol., L. L59-L77. ASADA, Y., FURIIYA, T., & MUMKAMI, N. (1981) Pyrrolizidine japonica. alkaloids from Ligularia Planta l.{ed., 9r 202-203. ASADA, Y., SltrRAISHr, M., TAKEUCHI, T., OSAWA,Y., & FURUYA, T. (1985) Pyrrolizidine alkaLoids from Crassocephalurn crepidioides. Planta Med., 51: 539-540.

~~BARNES,J.u., MAGEE,P.N., & SCIIOENTAL,R. (1964) Lesions in the Lungs and livers of rats poisoned vith pyrrolizidine aLkaLoids ful.vine and irs N-oxide. J. P3g!9!-Bac.!95!9!., E: 521-53L.~~

~~BERRY, D.M. & BRAS, c. (1957) Venous occlusion of the liver in Crotalaria and Senecio poisoning. North Arn. vet., 323-w-- 8:~~

~~BHATTACHARYA,K.J. (1965) Foetal and neonatal responses to hepatotoxic agents. J. Pathol. Bacteriol., 90:151-L61.~~

~~BICCIII, C., DrAltATO, A., & CAPPELLETTI,E. (1985) Determina- tion of pyrroLizidine alkaloids in Senecio inaequidens D.C. by capil-lary gae chromatography. -l . cttrffi@0. -?29-~~

~~BICK, Y.A.E. ( 1970) Comparison of the effects of LSD, heliotrine, and X-irradiation on chromosome breakage and the effects of LSD on the rats of cel1 division. Nature (r,ond.) 226t LL65-1167.~~

BICK, Y.A.E. & CULVENOR,C.C.J. (1971) Effects of dehydroheliotridine, a metabolite of pyrroLizidine alkaloids on chromosome structure and cell division in cuttures of animal cel-ls. Cytobios, 3;, 245-255.

BICK, Y.A.E. & JACKSON,W.D. (196e) Effects of the pyrrolizidine alkaloid heliotrine on ce11 division and chromosome breakage in cultures of leucocytes from the marsupial Potorus tridactylus. Aust. J. bio!-. Sci. , 2Lt 469-48L.

~~BICK, Y.A.E., CULVENOR,C.C.J., & JAGO, M.V. (1975) Cornpar- ative effects of pyrroLizidine aLkal-oids and related compounds on leukocyte cultures from Potorus tri@.!{5. Cytobios, 14: 151-160.~~

~~BIF.ECKA, I{., FROHLICH, M.I,'I., HULL, L., & CHASKES' M.J. (1980) Pyrrolizidine alkaloids of Hef@, from Mexico and adjacent USA. Phytochemistry, Ig:EF4I6|-~~

~~BIRECKA, H., CATALFAMO,J.L., & EISEN, R.N. (1981) A sensitive method for detection and quantitative determination of pyrrolizidine aLkaloids. Phytochernistry, 20; 343-344.~~

~~BLACK, D.N. & JAGO, M.V. (1970) Interaction of dehydroheliotridine, a metabolite of heliotridine based pyrrolizidine alkaloids, nith natural and heat denatured RNA. Biochem. J., 118: 347-353.~~

~~BOHLMANN,F., KLOSE, I.I., & NIKISCH, K. (1979) [Synthesis of dehydroheliotridine and 3-oxy-dehydroheliotridine. I TetraLedr. Lett., L979: 3699-3702 (in German).~~

BOHLMANN,F., ZDERO,C., JAKUPOVTC,J., GRENZ,M., CASTRO,V., KING, R.M., ROBINSON,H., & VINCENT, L.P.D. (1986) Further pyrrolizidine alkaloids and furoeremophilanes from Senecio species. Phytochemistry, 15: 1151-1159.

~~BOPPRE, M. ( 1986) Insects pharmacophagously utilizing defensive pl-ant chemicals (pyrrolizidine alkaloids). Naturwiss enschaften, 73t 17-26.~~

~~BMGINSKII, B.M. & BOBOKHADZAEV,I. (1965) [Hepatospleno- megaly against the background of heliotropic toxicosis.l Sov. Med., 28: 57-60 (in Russian). -230-~~

~~BRAS, G. (1973) Aspects of hepatic vascular diseases. In: GalL, E.A. & Mostofi, F.K., ed. The liver: International Academy of Pathology monograph, Balt@ and t'lilkins, pp. 406-530.~~

~~BRAS, c. & HILL, K.R. (1956) Veno-occlusive disease of the - liver essentiaL pathology. Lance'1r, 2z L6l-163.~~

~~BRAS, c. & WATLER, D.C. (1955) Further observations on the norphoLogy of veno-occlusive disease of the f.iver in Jamaica. I'lest Indian med. J., Lt 2OL-2I1.~~

~~BRAS, G., JELLTFFE, D.8., & STUART, K.L. (L954) Veno- occlusive disease of the liver with nonportal type of cirrhosis occurring in Jamaica. Arch. Pathol"., Z: 285-300.~~

~~BRAS, G., BERRY,D.U., & GYORGI,P. (1957) Plants as etiological factor in veno-occlusive disease of liver. @!, 1: 960-962.~~

~~BRAS,G., BROOKS,S.8.11., & I^IATLER,D.C. (1961) Cirrhosis of tiver in Jamaica. J. PathoL. Bacteriol., 82t 503-512.~~

~~BMUCIILI, J., LUTHY, J., ZWEIFEL, U., & SCIILATTER,C. (1982) Pyrrolizidine alkaloids from Symphytun officinale L. and their percutaneous absorption in rats. Experientia (Basel), 38: 1085-1087.~~

~~BRIGGS, L.H., CAUBIE, R.C., CANDY, 8.J., 0TDONoVAN,G.I{. , RUSSELL,R.A., & SEELYE,R.N. (1965) Alkaloids of New Zealand Senecio species. Part II; senkirkine. J. Chen. Soc., CL965: 2492-2498.~~

BRINK, N.G. (L982) Somatic and teratogenic effects induced by heliotrine in Drosophilia. Mutat. Res., &: L05-111. BROOKS,I^l.E.It., I'IILLER, C.G., MCKENZIE,K., AUDRETSCIT,J.J., & BRAS, c. (1970) Acute veno-occlusive disease of the liver.' Fine structure in Jamaican children. Arch. Patho1., 89: 507-529.

~~BROWN,K.S. (1984) ChemicaL ecology of dehydropyrrolizidine alkaloids in aduLt Ithorniinae (Lepidoptera: NynphaLidae). Rev. Bras. Biol,. , 4,!: 435-440.~~

~~BRUGGEMAN,I.M. & VAN DER IIOEVEN,J.C.M. (1985) Inducrion of SCEs by some pyrrolizidine aLkaloids in V79 Chinese hamster ceLls co-cultured with chick embryo hepatocytes. Ig@, L42t 209-212. -23L-~~

BRUNER, L.tr., CARPENTER,t.J., I{AMLOW, P., & ROTI{, R.A. (1986) Effect of a mixed-function oxidase inducer and inhibitor on monocrotaline pyrrol-e pneumotoxicity. Toxicol. appl. Pharmacol., 85t 416-427.

BUCKMASTER,G.l,l., CITEEKE' P.R., ARSCOTT' G.H., DICKINSON' E.D., PIERSON, M.L., & S1IULL, L.R. (L977) ResPonse of Japanese quail to dietary and injected pyrrol-izidine (Senecio) aLkaloid. J. anirn. Sci. , 9(6). L322-1325.

BUIL, L.B. (1955) The histoLogical evidence of liver damage from pyrrolizidine alkaloids. Aust. vet. J., ll: 33-40. BULL, L.B. & DrCK, A.T. (1959) The chronic pathological effeets on the liver of the rat of the pyrrolizidine al-kal.oids heliotrine, lasiocarpine, and their N-oxides. J. Pathol. Bacteriol., 78: 483-502.

BL'LL,L.B. & DrCK, A.T. (1960) The function of total dose in th production of chronic 1etha1 disease in rats by periodic injections of the pyrrolizidine alkal-oid heliotrine. Aust. J. exp. Biol- . med. Sci. , !!: 515.

BULL, L.B., DICK, A.T., KEAST,J.C., & EDGAR,G. (1956) An experimental investigation of the hepatotoxic and other effects on sheep of consumption of Ileliotropium euroPaeum L. Heliotropium poisoning of sheep. Aust. J. agric. Res., 7: 28L-332.

~~BULL, L.8., DrcK, A.T., & MCKENZTE,J.S. (1958) The acute effects of heliotrine and Lasiocaroine and their N-oxides on the rat. J. Pathol. Bacteriol.. 75: 17-25.~~

~~BlrLL, L.B., CULVENOR,C.C.J., & DrCK, A.T. (1968) The pyrroLizidine alkaloids, Amsterdam, North Holland Publishing Co.~~

~~BURGUERATJ.A., EDDS, G.T., & OSUNA,O. (1983) rnf luence of selenium on aflatoxin B1 or Crotalaria toxicity in turkey poults. Am. J. vet. Res.,44t L7I4-L717.~~

~~BURNS, J. (1972) The heart and pulrnonary arteries in rat fed on.[ !aig@.. J. Pathol., q: 87-194.~~

~~BUTLER;W.H. , MATTOCKS,A.R. , & BARMS, J.M. ( 1970) Lesions in the Liver and lungs of ratg given pyrrole derivatives of pyrrolizidine alkaloids. J. Pathol., f99: 169-175. -232-~~

CAMPBELL, J.c. (1956) An investigation of the hepatotoxic effects in the fowl of rag$rort (Senecio jacobaea tinn) with special reference to the induction of liver tumours with seneciphyll-ine. Proc. R. Soc. Edinburgh, W 111-130.

CAMPBELL, J.G. ( 1957a) Studies on the influence of sex hormones on the avian l-iver. II. Acute liver damage in the male fowL and the protective effect of oestrogen, as determined by a liver function test. J. Endocrinol. 15: 346-350.

~~CAI'{PBELL, J.c. (1957b) Studies on the influence of sex hormones on avian 1iver. III. Oestrogen-induced regeneration of the chronical-ly damagedliver. J. Endocrinol., 15: 351-354.~~

CANDRIAN, U., LUTIIY, J., GRAF, U., & SCHLATTER,CH. (1984a) Mutagenic activity of the pyrroLizidine alkaLoids seneci- phyLl ine and senkirkine in Drosophila and their transfer ioto rat mil"k. Food chern. ToxicoffiD3-225.

~~CANDRIAN, U., l,UTHy, J., SCHMID, p., SCHLATTER, CH., & GALLASZ, E. (1984b) Stability of pyrrolizidine alkaloids in hay and siLage. J. agric. food Chen., 32: 935-937.~~

CANDRIAN, U., LUTI{Y, J., & SCIILATTER,CH. (1985) In vivo 3tt) binding of retrooecine-labelled ( , seneciphyllinfriiE 3H-senecionine to DNA of rat Liver, 1-ung and kidney. Chem.-biol. Interact. , 542 57-69.

~~CARILLO, L. & AVIADO, D.M. (1969) Monocrotaline induced pulmonary hypertension and g-chlorophenylalanine. Lab. Invest., 20: 2L3-218.~~

~~CARSTENS,L.A. & ALLEN, J.R. (l-970) Arterial degeneration and glomeruLar hyaLinization in the kidney of monocrotaline intoxicated rats. Am. J. Pathol. , 6O: 75-92.~~

~~CHALMERS,A.H., CULVENOR,C.C.J., & SMTTH, L.t{. (1965) Characterization of pyrrolizidine alkaloids by gasr thin 1ayer, and paper chromatography. J. Chromatogr., gz 270-277.~~

~~CHEEKE, P.R. & GORMAN,c.R. (L974) Influence of dietary protein and sulphur amino-acid levels on the toxicity of Senecio jacobaea (tansy ragwort) to rats. Nutr. Rep. Iot., 9: ,Y t-zu t .~~

~~CIIEEKE, P.R. & PIERSON-GOEGER,M.L. (1983) Toxicity of Senecio and pyrrolizidine alkaloids in various laboratory animaLs and avian species. Toxicol. Lett., 18: 343-349. -233-~~

~~CI{EN, K.K. (1945) Pharmacology. llepatotoxic alkaloid.. @ Rev. Physiol. , Z: 695-697.~~

~~CI{EN, K.K., IIARRIS, P.N., & ROSE' C.L. (1940) The action and toxicity of platyphylline and seneciphylline. J. Pharmacol. gl:g-Thel, 68: 130-1-40.~~

~~CIIESMY, C.F. & ALLEN, J.R. (1973a) Resistance of the guinea pig to pyrroLizidine aLkal.oid intoxication. Toxicol. apPl. Pharmacol., 26: 385-392.~~

~~CHESNEY,C.F. & ALLEN, J.R. (1973b) Endocardial fibrosis associated with monocrotaLine induced pulmooary hypertension in non-human primates (uacaca arctoides). Am. J. vet. Res.' 34:1577-1581.~~

CIIESMY, C.F., IISU, I.C., & ALLEN, J.R. (L974) Modifications of the in vitro metabolism of the hepatotoxic pyrrolizidine a Lta L o i dT.onoc r o t a 1 ine . Res. Cosmun. chem. PathoL. Pharmacol., 8z 567-57O.

~~CIIOPRA,R.N., ed. (1933) Indigenous drugs of India, Calcutta, The Art Press.~~

~~CLARK, A.M. (1959) Mutagenic activity of the alkaloid (Lond.)' heliotrine i" @!!!.. Nature !Q!: 731'732.~~

~~CLARK, A.M. (1"975) Naturally occurring mutagens. $!g!-Res., 322 361-174.~~

cooK, 8.A., SINNIIUBER,.J.R., THOMAS,P.J., OLSON, T-A., SILVERMAN,T.A., JONES' R., I',HITEHEAD'V.U.' & ROYMANN'F.B. ( 1983) Hepatic faiLure secondary to indicine N-oxide toxicity. A pediatric oncology group study. @, 522 6L-63.

~~COOK, J.Vi., DUFFY, E., & SCIIOENTAL'R. (1950) Primary liver tumours in rats following feeding with alkaLoids of Senecio jacobaea. \]1!11]'1]]!, 4: 4O5-4LO.~~

~~cRArc, A.M., SHEGGEBY,G., & wrcKs,c.E. (L984) Large scale extraction of pyrrolizidine alkaloids from tansy ragrtort (Senecio jacobaea). Vet. hurn. Toxico1.. 26: 108-111.~~

~~CULVENOR, C.C.J. (1968) Tumour inhibitory activity of pyrrolizidine alkaloids. J. pharm. Sci., 5Jr LLL?-LLL7.~~

~~CULVENOR,C.C.J. (1978)) Pyrrolizidine alkaloids: occurrence and systematic inportance in angiosperms.- Bot. Notiser' !]!: 473-486. -234-~~

~~ctLVENOR, C.C.J. (1980) ALkaloids and human disease. I n R.L. smith, & Bababunmi, 8.A., ed. Toxico 1n the t rSt London, Taylor & Francis Ltd., pp. I~~

~~CULVENOR,C.C.J. (1983) Estimared intakes of pyrrolizidine alkaloids by humans. A comparison with dose rates causing tumours in rats. J. Toxicol. environ. llealth, 11: 625-635.~~

~~CULVENOR,C.C.J. (1985) Pyrrolizidine alkaloids: some aspects of the AustraLian involvement. Trends pharmacol. Sci., 6: 18-22.~~

CULVENOR, C.C.J. & JAGO, M.V. (1979) Carcinogenic planr products and DNA. In: Grover, P.L., ed. Chemical carcinogens and DNA, Boca Raton, Florida, CRC press, voffiT6-1-!!- cuLvENoR, c.C.J., DOWNING,D.T., EDGAR, J.A., & JAGO, M.V. ( 1969) Pyrrolizidine alkaloids as alkylating and antimitotic agents. N.Y. Acad. Sci., 163: 837-847. cuLvENoR, c.c.J., EDGAR,J.A., SMITH, L.W., & TWEEDDALn,H.J. (1970a) Dihydropyrrol-izines. tII. Preparation and reactions of derivatives related to pyrroLizidine alkaloids. Aust. J. Chem., 23: 1853-1867. cuLvENoR, c.C.J., EDGAR,J.A., SMITH, L.W., & TWEEDDALE,It.J. (1970b) Dehydrolizines. IV. Manganese dioxide oxidation of 1,2-dihydropyrrolizidines. Aust. J. Chem., Er L869-L879. cuLvENoR, c.c.J., EDGAR, J.A., JAGO, M.V., OUTTERIDGE,A., PETERSONJ.8., & SMITII, L.W. (1976a) llepato- and pneumotoxicity of pyrrolizidine al-kaloids and derivatives in relation to molecular structure. Chern.-bio1. Interact., 12: 299-324

~~CULVENOR,C.C.J., EDGAR, J.A., SMITII, L.W., & IIIRONO, I. (1976b) The occurrence of senkirkine in Tussil"ago farfara. Aust. J. Chem., 2i 229-233.~~

CULVENOR,C.C.J., CLARKE,M., EDGAR,J.A., FRAttN, J.L., JAGO, M.V., PETERSoN,J.8., & SMITH, L.W. (1980a) Structure and toxicity of the alkaloids of Russian comfrey (Syrnphytun x uplandicum Nyman), a medicinal herb and item of EGEiiliffi (sasei), ETp-'e-ii6Eia !!: 377-389. cuLvENoR, c.c.J., EDGAR,3.A. r FMHN, J.L., & S}4ITH, L.V,I. ( r.98ob) The alkaloids of Symphytum x uplandicum (Russian comfrey) . Aust. J. Chem., 33:1105-1113. -235- cttLvENoR, c.c.J., EDGAR, J.A., & SUrTtt, L.!il. (1981) Pyrrolizidine alkaloids in honey f"o* $\!gp]39!3g!g t. J. agric. food Chem., 29t, 958-960. cuLvENoR, c.c.J., JAGO, M.V., PETERSON,J.E., SMrrIl, L.w. ' PAYNE, A.L, CAI'{PBEIL,D.G., EDGAR,J.A., & FRAtl'N,J.L. (-1984) Toxicity of Echium plantagine-um (Patersont s curse). I. Marginal toxic effects of Merino wethers from long-term feeding. Aust. J. agric. Res., 35.. 293-304. cuLvENOR,c.c.J., EDGAR, J.A., SMITII, L.$1., KIJUANA,C.R., & LIN! H.J. (1986) Ie!qtf!piq4--!99i99gll Fisch and Mev r.qenElfr-ec as cause offi quc Lu 4 'Eluar tea. Lancet, Lt 978,

~~DANN, A.T. (1960) Detection of N-oxides of the pyrrolizidine aLkaloids. Nature, 4730: 1051.~~

DANNINGER,T., HAGBI,IANN,U., SCHl,lrDT, V., & SCHOENHOEFER,P.S. (1983) Toxicity of pyrrolizidine aLkaloid-containing medicinaL plants. @[g, 128; 289-303. DATTA, D.V., KHUROO,M.S., MATTOCKS,A.R., AL(AT, B.K., & CHHUTTANI,P.N. (1978a) Veno-occlusive disease of l-iver due to lleliotropium plant used as medicinal herb (report of six casea lrith review of literature). J. Assoc. Phys. India, 26(5)t 383-393.

~~DATTA, D.V., KIIUROO,M.S., MATTOCKS,A.R., ATKAT, 8.K., & CHIIUTTANNI,P.N. (1978b) Herbal medicines and veno-occlusive disease in India. Postgrad. med; J., 54; 5L1-515.~~

DAVIDSON,C.S. (1963) ?lants and fungi as etiologic agents of cirrhosis. New Engl. J. Med., 4t LO72-L073. DAVIDSON,J. (1935) The action of retrorsine on ratrg liver. J. Pathol. Bacteriol., 40: 285-295.

~~DEAGEN, J.T. & DEINZER, M.L. (L977) Improvement in the extraction of pyrrolizidine alkaloids. !!99!!3, 40: 395-397.~~

~~DEAN, R.E. & WINWAFO,A.Il. (1974) An investigation into the possibility of tansy ragwort poisoning of black-tailed deer. J. wildl . Dis., p: 166-169~~

~~DETNZER,l.{.L., TIIOMSON,p.A., BI'RGETT,D.M., & ISAACSON,D.L. (L977) Pyrrolizidine aLkaLoids3 their occurrence in honey from tansy ragwort (S. jacobaea). science, 195: 497-499. -236-~~

DEINZER, M.L., THOMSON,p.A., GRTFFIN, D., & DICKINSON, E. ( 1978) Sensitive analytical method for pyrroLizidine alkaloids. The nrass spectra of retronecine derivatives. Biomed. mass Spectrom., 2: L75-L79.

DBLFOSSET E.S. (1985) Re-evaluation of the biological control program for Heliotro in Australia. In: Delfosse, 8.S., ed. Proceed sof VI International. ium on Bio icaL Control of Weeds 9-25 Vancouver Ca , Ottawa Agr Cure Canada, pp. 73

~~DELFOSSE,E.S. & CULLEN, J.M. (1985a) CSIRO division of Entomology Submission to the enquiries into biological control of Echium plantagineum L., Patersonts curse/Sal"vation Jane. Plan@.~~

DELFOSSE,E.S. & CULLEN, J.M. (1985b) Echium planragineum: cataLyst for confLict and change in AustEllT--GT-D-fE6G, 8.S., ed. Proceedings of the VI International Svmposium on niolosical iii;6ii6jil Uanada, Ottawa Agriculture Canada pp. 249-292.

DELORME,P., JAY, M., & FERRY, S. (L977) phytochemical inventory in indigenous Boraginaceae: study of the aLkaloids and polyphenolic cornpounds (anthocyanic and flavonoid compounds). Plant. M6d. Phytoth€r., Lt: 5-11. (in french)

~~DE WAAL, H.L. & VAN ThtISK, p. (1964) The chemical composition of four Senecio species from Kruger National Park. Koedoe (South Africa), 7'. 40-42.~~

~~DICK,-A.T., DANN,A.T., BULL, L.B., & CULVENOR,C.C.J. (1963) Vitarnin B12 and the detoxication of hepatotoxic pyrroLi- zidine alkaloids in rumen liquor. Nature (Lond.), lg!, 207-20g.~~

~~DICKINSON, J.O. (1980) Release of pyrrolizidine alkaloids into roilk. Proc. West. pharmacoL. Soc.r 23:377-379.~~

~~DICKINSON, J.O., COOKE, M.P., KING, R.R., & MOHAMED,P.A. (L976) MiLk transfer of pyrrol"izidine alkaloids in cattle. J. Am. Vet. Med. Assoc., 169: LL92-I196.~~

DIMENNA,c.P., KRICK, T.p., & SEGALL,H.J. (1980) Rapid high performance f.iquid chromatography isolation of monoeaters, diesters, and macrocyclic diesters and pyrroLizidine alkaloids frorn Senecio jacobaea and Amsinckia intermedia. J. chromat6ffffiZ7F478. -237 -

DOLINSKAYA,K.N. (L952) [Pathomorphology of heliotropic hepatic dystrophy in children.l In3 MiLenkov, S.M. & Ki;haikinr-Y., ;d. Icollection of scientific papers on Toxic trtt 1ng oft vers l"ty ral As , pp. 165-172(ir Russian).

DREIFuss, P.A. (1984) A study of the mass spectra of pyrrolizidine alkaloids by negative ion chemical ionization and ns/urs: the identification of a monoester pyrrolizidine alkaloid in Eupatorium rugosum. Diss. Abstr. In!: 8., 45: '1183-1184.

DRMR, It.E. & }'IATTOCKS,A.R. (1984) The toxic effects in rats of some synthanecine carbamate and Phosphate esters analogous to hepatotoxic pyrrolizidine alkaLoids. Chen.-biol. !E5g:9:, 51: 201-218. DUBROVTNSKTT,S.B. (1946) [About the alinentary toxicosis caused by heliotrope.l J. Sov. Prot. Ilealth. 6z L7-2L (in Russian).

DUBRoVINSKII, S.B. (1952) [The etiology of toxic hePatitie with ascites.I Inr Millenkov, S.U. & Kizhaikin, Y., ed. Collection of scientific s on Toxic titis with scites tl ra , PuDtlSn].ng I{ouse ot t University of Central As a' PP. 9-25 (in Russian).

EASTMAN, D.F. & SEGALL, II.J. (L982) A nen pyrrolizidine atkaloict metaboLite, l9-hydroxysenecionine, isoLated from in vitro mouse hepatic nicrosornes using high perforrnance f.iquft chromatography. Drug Metab. Disp. , g. 696-699.

EASTMAN,D.F., DIMENNA,G.P., & SEGALL, It.J. (1982) Covalent binding .of tno pyrrolizidine alkaloids, senecionine, and seneciphylline to hepatic macromolecules and their distribution, excretion and transfer into nilk of lactating ri"e. \ gL_!!g., L0: 236-240. EDGAR, J.A. (1932) Pyrrolizidine alkaloids sequestered by Solomon Island Danaine butterflies. The feeding preferences of the Danainae and Ithomiinae. J. ZooL. (Lond.), 196: 385-399.

EDGAR, J.A. (1985) Gas chromatography of pyrrolizidine aLkaLoids. In: Seawright, A.A., Hegarty, M.P., James, L.F.' & Keeler, R.F., ed. Plant toxicologv. Proceedings of the AustraL ia USA onoug Plants sium. Brisbane ustra 1SDane, sonous Plants pp. -238-

~~EISENSTEIN, D. & IIUXTABLE, R.J. (1979) Approaches to the treatment of pyrrol izidine toxicosis. In: Cheeke, P.R., ed. Proceedi OI I"Um rolizidine (Senecio)~~

~~EL DAPiEER,S.M., TILLERY, K.F., LLOYD, }I.H., & HILI,, D.L. (1982) Disposition of indicine -N-oxide in mice and monkeys. Cancer Treat. Rep., !9: 183-186.~~

~~EMIIEL, M.w. (1948) Crotalaria poisoning in carrle. J. Am. Vet. Med. Assoc., 113: 164.~~

~~EMMEL, M.W., SAlilDERs,D.A., & ItENtEy, I,{.w. ( 1935) Crotalaria spect.abiLis seed poisoning in srrine. J. An. Vet. Med. Assoc 862 43-49.~~

EVANS, J.V., pENc, A., & NTELSEN, C.J. (rg7g) The gas chromatographic rnass spectrometric analysis of the new antitumour drug indicine-N-oxide utilizing a novel reaction accornpanying trimethysilylation. Biomed. mass Spectrom., 6; 38-43.

EVANS, J.V., DALEY, s.K., MCCLUSKY,G.A., & NIELSEN, C.J. (L980) Direct quantitative analysis of indicine N-oxide in cancer patient samples by gas chromatography using the internal standard heliotrine N-oxide including a nass spectraL comparison of their trinethyTsilyl derivatives. Biomed. mass -@_, 7z b5-73.

~~FARRINGTON,K.J. & GALIAGIIER, C.fl. (1960) Complexes of copper nith some pyrroLizidine alkaloids and with some of their esterifying acids. Aust. J. biol. Sci., t3: 600-603.~~

~~FISH, l.{.S., SIiXEIEY, C.C., JOHNSON,N.Ir{., IAWRENCE,E.p., & I{ORNING,E.C. (1956) Chemical and enzymic rearrangements of NrN-dinethylaminoacid oxides. Biochem. Biophys. Acta, L2, Tg6-tgt.~~

~~FISHMAN, A.P. (L974) Dietary puLmonary hypertension. Circ. -L' 35: 657-660.~~

~~FORSYTII, A.A. (1968) British poisonous plants, London, Her Majesty's Stationery Office. -239-~~

~~FOXr D.W., HART, M.C., BERGESON,P.S., JARRETT, P.B. ' STILI..},IAN, A.8., & HUXTABIJ, R.J. (1978) Pyrrolizidine ( Senecio) intoxication ninicking Reyer s syndrome. J. Paediatr., 93: 980-982.~~

~~FRAI{N, J.L., CULVENOR, C.C.J., E MTLLS, J.A. (1980) Preparative separation of the pyrrolizidine aLkaloids, intermedine and Lycopsamine, as their borate complexes. J. Chromatogr., L95i 379-383.~~

~~FURMANOhIA,l{., GUZEWSKA,J., & BELDOWSKA,B. (1983) Mutagenic effects of aqueous extracts of Symphytun qfficinale L. and of its alkal.oidal fractions. J. appT. toxicdl., 9! 127-130.~~

~~FURIIYA, T. & IIIKICHI, M. (1971) AlkaLoids and triterpenoids of Symphytumofficinale. Phytochenistry' l-Q: 2217-2220,~~

~~FUSHTMT,K., KATO, K., KATO, T., MATSUBAM,N., & HTRONO,r. (197S) Carcinogenicity of flower stalks of Petasites japonicus Maxirn in mice and Syrian golden hamsters. Tox:!co!;-lggL' !: 29L-294.~~

~~GADELLA, T.!il.J., KLIPITUIS, 8., & HUIZING, Il.J. (1983) Cyto- and chemotaxononical studies on the sections of$cinaLia and coeruLea of the genus Symphytum. Bot. IleLv., p: L69-L92.~~

GANEY,P.E., FINK, G.D., & ROTH, R.A. (1985) The effect of dietary restriction and aLtered sodium intake on the cardiopulmonary toxicity of monocrotaline pyrroLe. Toxicol. appl. Pharmacol., 13,. 55-62.

~~GARDINER, M.R., ROYCE, R., & BOKOR, A. (L965) Studies on Crotalaria crispata, a newLy recognized cause of Kimberley horse disease. J. PathoL. Bacteriol., 89; 43-55.~~

~~GHANEM,J. & ITERSIIKO,C. (1981) Veno-occl-usive disease and primary hepatic vein thrombosis in Israeli Arabs. Isr. J. med. Sci., 17: 339-347.~~

~~GIIODSI,F. & WILI, J.A. (1981) Changes in pulmonary structure and function induced by monocrotaline intoxication. Am. J. Physiol., 240: ttL49-Il155.~~

~~GILRUTII, J.A. ( 1903) IIe tic cirrhosis affect horses and trllinton cattle (so-called d WC Eon, Nen Ze and Department of Agriculture, Pp. 22 279 ( 11th annual report). -240-~~

~~GOEGER,D.E., CIIEEKE, p.R., SCHI,\ITZ,J.A., & BI'HLER, D.R. ( 1982) Toxicity of tansy rag\nort (:.r949_j=g9!9.") to goats. An. J. vet. Res., 432 252-254.~~

~~GOEGER,D.E. r CIIEEKE,p.R., RAMSDELL,H.S., NICIIOLSON,S.S., & BLHLER,D.R. (1983) Comparison of the toxicities of Senecio jacobaea, Senecio vulgaris and Senecio glabellus irfffi'; Toxicol. Lett., 15:~~

~~GONZALEZ,A.G., DE LA FUENTE, c., REINA, il., & LOYOLA, L.A. ( 1986a) Pyrrolizidine alkaloids from Senecio- phiLLipicus and Senecio illinitus. Planta Med., 52; 160.~~

GONZAIEZ,8., GARCIA, R., LEMUS, I., & ERAZO, S. (1gg6b) Pharmacological studies on senecionine, a pyrrolizidine alkaloid from Senecio fistu!.osus poepp.-- Ex less. (huaLtata). An. n. Acad. rar5lEi)IFi52.

~~cRASEs, P.J. & BEKER, S.G. (1972) Veno-occlusive disease of liver. the A case from Venezuela. Arn. J. Med., ll: 51I-516.~~

~~GRAY, A.I., NIC, A.N., TSAOIR, 8., & I,iATERMAN,P.c. (f983) Ilepatotoxic alkaloids and allantoin in Symphytum tuberosurn. J. Pharm. Pharmacol., {Suppl): 13.~~

GREEN,C.R. & CIIRISTIE, c.S. (1961) Malformarions in foetal rats induced by the pyrrolizidine alkaloid, heliotrine. Br. J. exp. PathoL., !!: 369-318. GREEN,M.II.L. & UURIEL, I.I.J. (1975) Use of repair-deficienr strains of Escherichia coli and liver microsomes to detect and characterise-5frE-Aarnage caused by the pyrrotizidine al-kaloiils heliotrine and monocrotaline. Mutat. Res., 28: 331--336.

GREEN,C.8., SEGALL,H.J., & BYARD, J.L. (1981) Merabolism, cytotoxicity, and genotoxicity of the pyrrolizidine alkaloid senecionine in prinary cultures of rat hepatocytes. ToxicoL. appL. PharmacoL., q: 176-185. clrENcERICH, F,p. (I9j7) Separation and purification of multiple forms of microsomal cytochrome P-450. J. biol_. Cheg;-r 252:3970-3979.

~~GUENGERICII,F.P. & I4ITCHELL,M.B. (1980) Metabolic acrivarion of modeL pyrroLes by cyt,ochroare P-450. Drug Metab. Disp., B: 34-38. -241-~~

~~GUIDUGLI, F.H., PESTCHANKER,M.J., DESAI.,MERON,M.S.A., & cIoRDANo, o.S. (1986) l-hydroxypLatyphyllide, a norsequiter- pene lactone from Senecio gilliesiano. Phytochemistry, D: L923-L926.~~

~~GUNER, N. (1986) ALkaloids of lleliotropium suaveolens. J. nat . Prod . , 49'. 369.~~

~~GUPTA, P.S., cUpTA, G.D., & SHARMA, M.L. (1963) Veao- occlusive disease of the l"iver. Br. med. J., L: 1184-L1-86.~~

IIABS, It. (1982) Senecio numorensis fuchsii. Carcinogenic and mutagenic activity of an alkaLoidaL extract from a phytotherapeutic drug. Dtsch. Apoth. Ztg, Vt 799-804. HABS, II., HABS, M.t MARQUARDT,H., RoDER, E., SCHMAIIL,D., & WIEDEMELD, !t. (L982) Carcinogenic and mutagenic activity of an alkaloidal extract of Senecio numorensis spp. fuchsii. Arzneimittelforsch, 32: L44-L48.

~~IIAGGLUND,K.M., L'EMPEREUR,K.M., ROBY,M.R., & STERMITZ,F.R. ( l-985) Latifoline and latifoline N-oxide from }lackel-ia fLoribunda, the western fal-se forget-m-e-not. {:--ne!-. -E94., 48: 638-639.~~

IIAMMOUDA,F.M., RIZK, A.M., ISMAIL, S.I., ATTEYA, S.Z., GHALEB,It.A., MADKOUR,M.K., POI4IAND,A.E., & IIOOD' G. (1984) Poisonous plants contaminating edibLe ones and toxic substances in plant foods. Part 3: pyrrolizidine al.kdloids from lleliotropiusr digynurn Forssk (= H. luteum, poir). tharmazie, 39: 703-705.

~~HARDTNG,J.D.J., LEWTS,c.r DONE,J.T., & ALLCROFT,R. (1964) ExperimentaL poisoning by Senecio jacobaea in pigs. Pathol. Vet. , 1.' 204-220.~~

~~ITARRIS, C., REDDY, J., CItrGA, M., & SVOBODA,D. (1969) PoLyribosornaL disaggregation by Lasiocarpine. Biophys. Acta 180: 587-589.~~

~~IIARRIS, P.N. & CHEN, K.K. (1970) Development of hepatic tumours in rats following ingestion of @!1!g!!$_. Cancer Res., 30: 2881-2886.~~

~~I1ARRIS,P.N., ROSE,C.L., & CIIEN, K.K. (L957) llepatotoxic and piarrnacoLogical properties of heliotrine. Arch. Pathol., 64: L52-t57. -242-~~

~~HARTMANN,T. & ZIMMER, M. (1985) Organ-specific distribution and accumulation of pyrrolizidine alkafoids during the life history of 2 annuaL senecio species. J. plant PhysioL., V. 67-80.~~

~~HASHEM,M. (1939) Etiology and pathology of types of liver cirrhosis in Egyptian children. @, 22: 3L9-354.~~

~~IiAYASHI, Y. (1966) Excretion and alteration of monocrotaline in rats after a subcutaneous injection (abstract). Fed. Proc., 25:688.~~

~~HAYASIII, Y. & LALICII, J.J. (1967) Renal and pulmonary alterations iaduced by a singLe injection of monocrotaline. Proc. Soc. Exp. Biol. Med., V, 392-396.~~

~~HAYASIII, Y. & LALICII, J.J. (1968) Protective effect of mercaptoethyl.arnine and cysteine against monocrotaline intoxication in rats. Toxicol. appl. Pharrracol. , Lzi 36-43.~~

~~IIAYASIII,y., lluSSA, J.F., & LALICH, J.J. (1967) Cor pulmonaLe in rats. Lab. Invest., 16: 875-881.~~

~~ItAYASttI, Y., KATO, M., OTSUKA,H. (L979) Inhibitory effects of diet reduction on Monocrotaline intoxication in rats. Toxicol. Lett., 3: 15L-155.~~

~~IIAYASIII , Y. , SIITMADA, M. , & KATAYAI'IA,H. ( L977) Experimental insulinoma in rats after a single administration of monocrotaline. loxicol. Lett., L: 4L-44.~~

HAYASHI, Y., KOKUBO,T., TAKAHASHI,U., FURUKAWA,F., OTSI'KA, H. , & IIASIIII,IOTO, K. ( 1984) Correlative rnorphological and biochenical studies on monocrotaline-induced pulmonary alterations in rats. Toxicol. Lett. , 4t 65-77. HAYES,M.A., ROBERTS,E., & FARBER,E. (1985) Initiation and selection of resistant hepatocyte noduLes in rats given the pyrrolizidine alkaloids lasiocarpine and senecionine. Cancer Res., 45: 3726-3734.

~~IIEATII, D. , SHABA, J. , WTLLIAMS, A. , SMrTIl, p. , & KOMBE, A. (1975) A pulmonary hypertension-producing plant from Tanzania. Thorax, 30: 399-404. -243-~~

IIENDRICKS,J.D., SINNI{UBER,R.O., IIENDERSON,l.{.D., & BITHLER, D.R. (1981) Liver and kidney pathology in rainbow trour (-9g1gg_gg!d""9) exposed to dietary pyrrolizidine (Senecio) alkaLoids. Exp. noL. pathol., 35: 170.1g3.

~~llILLr K.R. (1960) Worldwide distribution of seneciosis in nan and animals. Proc. R. Soc. Med.. 53: 28L-282.~~

~~HILLIKER, K.S. & ROTtt, R.A. (1984) Alreration of mono- crotaLine pyrrole-induced cardiopulmonary effects in rats by hydrallazine, dexamethasone or sulphinpyrazooe. Br. J. Pharmacol., 82: 375-380~~

HILLTKER, K.S., BELL, T.c., & ROTH,R.A. ( 1983) Monocrotaline pyrrol*induced pulmonary hypertension in fawn-hooded rats with pLateLet storage pool deficiency: 5 hydroxytryptamine uptake by isolated, perfused lungs. Thromb. llaemostasis (stutrgarr), 50: 844-847.

~~HILLIKER, K.S., BELL, T.c., LORIMER,D., & ROTlt, R.A. (1994) Effects of thrombocytopaenia in monocrotaline pyrroleinduced pulmonary hypertension. physiol., Am. J. W H747-II753.~~

~~HIRONo, I., SIIIMIZU, M., FUSHIMI, K., MORI, ll., & KATO, K. (1973) Carcinogenic activity of petasites japonicus Maxirn, a kind of colrsfoot. cann uonogr. cai?6ffi81@%28.~~

~~HIRONO,I., MORI, H., & CULVENOR,C.C.J. (L976) Carcinogenic activity of coltsfoot, Tussilago farfara L. Gann Monogr. Cancer Res., fl: 125-129.~~

HIRONO, I., MORI, H., YAMADA, K., ttIMTA, y., HAGA, M., TATEMATSU,H., & KANIE, S. (L977) Carcinogenic activity of petasitenine, a nevr pyrrolizidine alkaloid isolated from Petasites japonicus Maxim. J. Natl Cancer Inst., 58: L155-1157.

~~HIRONO, I., 'symphytuur MORI, H., & HAGA, M. ( 1978) Carcinogenic activity of of f ic inaLe. J. Natl Cancer Inst.. 61: 865-869.~~

HIRONO,I., HAGA, M., FUJII, M., MATSUUM, S., MATSUBAM,N., NAKAYAMA,M., FURLryA, T., IIIKICUI , M., TAKANASITI, H. ! UCHIDA, E., HOSAKA, S., & UENO, I. (L979a) Induction of hepatic tumours in rats by senkirkine and symphytine. J. Natl Cancer Inst., 63: 469-472. -244-

HIRONO, I.T MORI, H" IIAGA, M.' FUJII, M.' YAMADA, K., TAKAI{ASHI, H., UCHIDA, E., IIOSAKA, S., UENO, I., MATSUS}ILMA, I., UMEZAVA, K., & SHIRAI, A. (1979b) Edible plants containing carcinogenic pyrrolizidine alkaloids in Japan. In: MilLer, B.C. ET AL., ed. Naturall occurr carcl mu and modulators of carc Ba E€r ryland, University Park Press, pp. 7

~~IiIRONO,r., I'ENO, r., ArSO, S., YAMAJI, T., E HAGAT M. ( 1983) Carcinogenic activity of Farfasium aponicum and Senecio cannabifolius. Cancer Lett.,~~

HOET,P., ASTIGUETA,M., & FRISQUE, A.U. (1981) Phytochemical study of Crotalaria nitens. @, 12: 34-35. II00PER, P.T. (L972) Spongy degeneration in the brain in relation to hepatic disease and arunonia toxicity in domestic animals. Ve1!.:_Reg., 90: 37-38.

~~IIOOPER, P.T. (1974) The pathoLogy of Senecio jacobaea poisoning of mice. J. Pathol., LL3'. 227-230.~~

~~H0OPER, P.T. (1975a) Spongy degeneration in the centraL nervous system of domestic animals. Part I3 morphology. Acta neuropathol. (Berlin), 3l: 325-334.~~

IIOOPER, P.T. (1975b) Spongy degeneration in the central nervous system of domestic animaLs. Part III: occurrence and pathogenesis - hepatocerebral disease caused by hyper- armnonaemia.Acta neuropathol. (Berlin), 31: 343-351".

~~IIOOPER, P.T. (1975c) Experimental acute gastro-intestinal disease caused by the pyrrolizidine alkaloid, lasiocarpine. J. comp. PathoL., Q: 341-349.~~

HOOPER, P.T. (1978) Pyrrolizidine allcaLoid poisoning- pathology vith particuLar reference to differences in anirnal and plant species. In: Keeler, R.F., Van Kampen, K.R., & James, L.F., ed. Effects of poisonous plants on l-ivestock, New York, Academic Preffi

~~IIOOPER, P.T. & SCANLAN, I,I.A. (1977) Crotalaria retusa poisoning in pigs and poultry. Aust. vet. J., 53: 109-114.~~

~~HOOPER,P.T., BEST, S.M., & MURRAY, D.R. (L974) llyper- armronaemia and spongy degeneration of the brain in sheep affected with hepatic necrosis. Res. vet. Sci., 16: 2L6-222. -245-~~

~~IIOOSON,J. & GRASSO,P. (1976) Cytotoxic and carcinogenic reaponse to monocrotaLine pyrrole. J. Pathol., M: L2I-L27.~~

~~HOU JINQUI, XIA YUDIN, YU Z\{ANS}II, AN YAN, & TANG YIXAI (L980) [Veno-occl.usive disease of the liver with report of 2 (in cases.l Chung ttua Nei ko Tsa Chih, !Z: 187-191 Chinese).~~

~~HsU, I.C., ALLEN, J.R., & CHESNEY,C.F. (1973) Identification and toxicological effects of dehydroretronecine, a metabolite of monocrotaLine. Proc. Soc. Exp. Biol. Med., !44: 834-838.~~

HSU, r.C., SHUI.,IAKER,R.C., & ALLEN, J.R. (L974) Tissue distribution of tritium 1abe1led dehvdroretroneeine. Chem.- biol. Interact., !: 163. HUIZING, Il.J., DE BOER, R., & UALINGRE,T.M. (1981) preparative ion-pair high performance liquid chromatography and gas chromatography of pyrrolizidine alkal-oids from comfrey. J. Chromatogr. , U: 257-262.

~~IIURLEY, J.V. & JAGO, l.{.V. (L975) pulmonary oedema in rats given dehydromonocrotaLine: a topographic and electron- microscopic study. J. Pathol., ll7: 23-32.~~

~~ITUXTABLE,R.J. (1980) Herbal teas and toxins: novel aspects of pyrrolizidine poisoning in the United States. Perspect. BioL. Med., !: L-L4.~~

~~ITUXTABLE, R., PAPLANUS, S., & LAUGHARN, J. (L977) ThC prevention of monocrotaline induced right ventricuLar hypertrophy. Sg!, 71: 308-310.~~

TIUXTABLE,R.J., CIARAT{ITARO,D., & EISENSTEIN, D. (1978) The effect of a pyrroLizidine alkaloid, monocrotaLine and a pyrrole, dehydroretronecine, on the biochemical functions of pulnonary the endothelium. Mol. Pharmacol., !!: 1189-1203.

~~I{UXTABLE,R.J., LUTHY, J., & ZWEIFEL,V. (1996) Toxicity of comfrey-pepsin preparations: Letters to the editor. New Engl. J. Med., 315: 1095.~~

IARC (L976) Pyrrolizidine alkaloids. In: Some naturally occurring substances, Lyons, International Agencl-Ei6FFE36iffi @42- (rARc Monograph ott th. Evaluation of Carcinogenic Risk of Chernicals to Manr Vot . 10). -246-

~~IARC ( 1983) Some food additives, feed additives and naturally occurr z-Z+S (IARC l"lonograph on the Evaluation of Carcinogenic Risk of Chemicals to Ilumans, vol.3L).~~

~~INDIAN COUNCIL 0F MEDICAL RESEARCIT (L955) Infantile cirrhosis of the liver in India. Iodian J. ned. Res.. 432 723-747.~~

ISMAILOV, N.I. (1948a) [on clinical features, etiology and pathogenesis of heliotropic dystrophy (toxic hepatitis nith (in asiites).1 c1in. lled., !t 28 Russian). ISMAILOV, N.I. (1948b) [Heliotropic toxicosis (toxic hepatitis with ascites),1 Tashkent, Academy of Sciences of @nRussian).

~~ISMAILOV, N.I., MADZIIrDOV,N.U., MAGRUPOV,A.r., MAKIIKAI.IOV, c.u., & MUKMINoVA,s.c. (1970) [Clinical signs, diagnosis and treatment of Trichodesma t AGph;'6;milT rashkent, I-{~~

~~JAGO, il.V. (1969) The development of the hepatic negalocytosis of chronic pyrroLizidine aLkaloid poisoning. An. J. Pathol., 562 405-422.~~

~~JAGO, M.v. (1970) i rnethod for the assessment of the chronic hepatotoxicity of pyrrolizidine alkaloids. Aust. J. exp. Biol. med. Sci., 48: 93-103.~~

~~JAGO, M.V. (1971) Factors affecting the chronic hepatotoxicity of pyrroLizidine alkaloids. J. Pathol., L05: 1-11.~~

JAGO, M.V., LANIGAN, G.W., BINGLEYT J.B., PIEREY, D.W.T., WHITTEN, J.H., & TITCIIEN, D.A. (1969) Excretion of the pyrrolizidine al"kaloid heliotrine in the urine and bile of sheep. J. Pathol., 98: 115-128.

~~JAGO, M.V., EDGAR, J.A., Sl.{rTll, L.W., & CULVENOR,C.C.J. (1.970) MetaboLic conversion of heliotrine based pyrrolizidine alkaLoids to dehydroheliotridine. Mo1. Pharmacol., 6: 402-406.~~

~~JELLIFFE, D.B., BRAS, G., & STUART, K.L. (1954a) Veno- occlusive disease of the liver. Paediatrics, L4: 334-339.~~

~~JELLTFFE,D.8., BRAS, c., & STUART,K.L. (1954b) The cLinical picture of veno-occlusive disease of the liver in Janaican children. Ana. trop. Med. Parasitol., 48: 386-396. -247-~~

~~JELLIFFE, D.B.t BRAS, G., & MUKIIERJEE,K.L. (1957) Veno- oeclusive disease of the Liver ard Indian childhood cirrhosis. Arch. dis. child. 322 369-385.~~

~~JOHNSON,A.E. (1,975) Changes in caLves and rats consuming miLk fron cows fed chronic lethal doses of genecio jacobaea (tansy ragwort). fun. J. vet. Res., 37:107-110.~~

JOltNSoN, A.E. (L979) Toxicity of tansy ragwort to cattle. Ini Cheeke, P.R., ed. Proceedings of the Synpogium on Pyrrolizidine (Senecio) Alkaloidsi Toxic Poisonous PLant Control Measures, Corvallis usA Fe , Oregon, Nutrition Research Inst tute, pp.

~~JOHNSON,A.E. (1982) FaiLure of mineral-vitamin supplements to prevent tansy rag\rort (Senecio jacobaea) toxicosis in cattle. An. J. vet. Res., 43: 718-723.~~

JoHNSON,A.E. & l,loLYNEUx,R.J. (1984) Toxicity of thread leaf groundseL (_S""""i"-_a""glgg:i var. Longilobus) to cattle. Am. J. vet. Res., !2r 26-3L. JOIiNSON,A.E., MOIYNEI'X,R.J., & sTuART,L.D. (1985) Toxicity of Riddelrs groundsel (Senecio riddelli) to cattLe. An. J. Ig!:-&a, 46: 577-582. JoHNsoN, !f.D. (L981) Mechanisn of in vitro acute toxicity of dehydromonocrotaline a netaboLite ifrflrrolizidine alkaloid monocrotaline. I9IjS.@, 1: 107-108.

~~JOI{NSON,W.D., ROBERTSON,K.A., POUNDS,J.G., & ALLEN, J.R. (1978) Dehydroretronecine-induced skin tumours in mice. J. NaEl Cancer Inst., 61: 85-89.~~

~~JULIEN, M.ll., ed. (1982) Biological control of weeds. A world catalog of gents an nstitute iological Control.~~

~~JUNEJA, T.R., GUPTA, R.L., & SAMANTA,S. (1984) Activation of monocrotal-ine, fulvine and their derivatives to toxic pyrroles by some thiols. Toxicol. Lett., 21; 185-189.~~

KAMPANZEV,N.N. (1952) [Experirnental heliotropic hepatitis.l In: Milenkov, s.M. & Kizhaikin, Y., ed. IcoLLection of scientific papers on Toxic Hepatit,is wittr Ascites,-TrIE's=hGE ilbliEhing ttouiE -E the Un:-versity of Central Asia, pp. L65-L72 (in Russian). -248-

~~KAY, J.I.{. & HEATH, D. (1966) Observation on the pulmonary arteries and heart neight of rats fed on CrotaLaria spectabilis seeds. J. PathoL., 92: 385-394.~~

~~ro] , J.l.{. & IIEATII, D. (1969) Crotalaria spectabilis: the p.uLngary hypertension p1ant, Springfield, Illl;;ft Charles C. Thomas, p. 38.~~

~~KAY, J.M. , HARRIS, p., & HEA,TH, D. ( 1967a) puLmonary hypertension produced in. rats by ingestion of Crotalaria spectabilis seeds. Thorax, 22: L76-L79~~

~~KAY, J.l{., GILLUND, T.D., & IIEATH, D. (1967b) Masr cells in the lungs of rats fed on Crotalaria spectabilis seeds. Arn. J. Pathol-., 5L: 1031-L044.~~

~~KAY, J.M., HEATH, D., SMITH, p., BRAS, G., & SUMMERELL,J. ( 197la) FuLvine and pulmonary circulation. ]@, 262 249-26t.~~

KAY, J.M. , SMITH, P., & HEATII, D. ( 197I-b) Aminorex and the pulmonary circulation. Thorax, 262 262-27O. 'J.M., KAY, KEANE, P.M., & StYAMA, K.L. (1985) pulmonary hypertension induced in rats by monocrotaline and chronic hypoxia is reduced by p-chlorophenylalanine. Respiration, 47: 48-56.

~~KEDZIERSKI,B. & BUHLER,D.R. (1985) Configuration of necine pyrroles - toxic metabolites of pyrroLizidine alkaloids. Toxicol-. Lett., 25: 115-119.~~

~~KEDZIERSKI, B. & BUIiLER, D.R. (1986) The formation of 6,7-dihydro- 7-hydroxy-l--hydroxyurethyl-5H-pyrrol iz ine , a metabol ite of pyrrolizidine alkaloids. Chem.-bio1. Inrteraqq., 57:217-222.~~

~~KHANIN, M.N. (1948) [The erioLogy of toxic hepatitis with ascites.l {rcfr. Pathol. USSR, 1: 42-47 (in Russian).~~

KIM, u.L. & JONES, L.P. (1982) Protective effects of butylated anisole, ethoxyquin and disulfiran on acute pyrrolizidine alkaloid poisoning in mice. Res. Comn. chem PathoL. Pharmacol. , 36.. 34L-344.

KTRFEL, A., WILL, c., WTEDENFELD,H. & RODER, E. (1980) -9 ( laR, 6bR,LoR, 1lR) , 15-dioxo-10-hydroxy-l-0 , 11 , 13-rrimerhyl- la, 2 , 3, 6b- te trabydro-5ll-pyrrol iz ino- [ l-a, 6b, 6a, c ], B-d ioxa- 15-c is- tride cene,C1gH25NO5. Cryst. Styruct. Comm., 9: 353-361. -249-

KNIGnT, A.P., KrMBEPJ,rNC,C.V., STERMITZ, F.R., & ROBY, lt.R. (1984) Cynoglossun officinale (houndrs tongue) - a cause of pyrrol.izidine aLkaloid poisoning in horses. J An. Vet. Med. Assoc., 185: 647-650.

~~KOEKEMOER,M.J. & I^IARREN,F.L. (1951) The occurreoce and preparation of the N-oxides. An improved method of extraction of the Senecio alkafoids. J. Chen. Soc., C1951: pp. 66-68.~~

~~KOLETSKY,A., oYAsur R., & REDDY,J.K. (1978) Mutagenicity of the pyrroLizidine (sengcip) alkaloid, Lasiocarpine in the (Abstract). salmonella/microsome TeFt.EL. Invegt. , Y. 352~~

~~KoVACII, J.S., MOERTEL,C.G., & IIAHN, R.G. (L979a) A phase I study of indicine-N-oxide. Proc. Am. Assoc. C4ncer Res.r 20: 357.~~

KOVACH,J.S., AMES, M.M., POWIS, G., MOERTEL,C.G., HAHN, R.G., & CREAGANE.T. (1979b) Toxicity and pharmacokinetics of a pyrrolizidine alkaloid, indicine-N-oxide, in humans. Cancer Res. , 39'. 4540-4544.

~~KI{AUS, c., ABEL, G., & sCHIMlmR, o. (1985) [Studies on the chromosome darnaging eftect of some pyrrolizidine alkaloids in (in human Lymphocytes in vitro. I Planta Med. , I1: 89-91 Gerrnan)~~

KRISITNAMACIIARI,K.A.V.R., BHAT, R.V., KRIS}INAMURTHY,D., KRISI{NASWAMY,K., & NAGARAJAN (1977) Aetiopathogenesis of endernic ascites in Sarguja district of Madhya Pradesh. Indian J. med. Res., 652 672-678.

~~KT'HARA,K., TAKANASIII,H.T IIIRONO, I', FURIryA, T., & ASADA, Y. (L980) Carcinogenic activity of clivorine, a pyrrol-izidine alkaloid isolated fron Ligularia dentata. Cancer LqtE., 10: LL7-L22.~~

~~KU},IANA,C.R., NG, M., LrN, H.J., KO, W., WU, P.c., & TODD, D. (f983) Ilepatic veno-occtusive disease due to toxic aLkaloid herbal tea - Letter to the editor. Lancet, L0 December: 1360-1361 .~~

KUMANA,C.R., NG, M., LIN, Il.J., KOr l,tr.r WU' P.C.r & TODD' D. (L985) llerbal tea induced hepatic veno-occl.usive disease: quantification of toxic alkaloid exposure in adults. g!, 4: 101-104. -250-

~~KUROZITUI, T., TANAKA, K., KIDO, M., & SHoYAMA,Y. (1983) Monocrotaline-induced renal lesions. Exp. rooL. Pathol., 39: 377-386.~~

~~LAFRANCONI, W.M. & IilIXTABLE, R.J. (1983) Changes in angiotensin-converting enzyme activity in lungs danaged by the pyrrolizidine alkaloid, monocrotaline. Thorax, 38: 307-309.~~

~~TAFRANCONI, W.M. & ITUXTABLE,R.J. (1984) tlepatic metabolisrn and pulmonary toxicity of monocrotaline using isolated perfused liver and 1ung. Biochern. PharmacoL., 332 2479-2484.~~

~~DtHAl,rEL, R.C., BRENDEL,K., & HUXTABLE,R.J. l$ll4scoIl,_-I^r.M.,_(1984) Differenriation of the iardiac and pulnonary- toxicity of monocrotaline, a pyrrolizidine aLi~~

~~LAFMNCONI, I,l.u. , 0HNITMA, S. , & HUXTABI,E, R. S. ( 1gg5) Biliary excretion of noveL pneumotoxic metabolites of the pyrrolizidine alkaloid monoerotaLine. Toxicon, 23: 903-992.~~

~~LALICE, J.J. & EHRHART, L.A. (L962) Monocrotal.ine induced pulmonary arteritis in rats. J. atheroscler. Res., 2: 482-492.~~

~~LALICI{, J.J. & MERKOVi,L. (1961) puLmonary arteritis produced t_1.: by il. feeding Crotal,aria specrabilis. L€b. Invest., 10; 744-750.~~

~~1AN9II -^(1984) Pyrroliiidine aLkaloids (edirorial). Lancer, lz 2Ol-202.~~

~~LAI\IGLEBEN,D. & REID, L.M. (1985) Ef fecr of .-merhylprednisolone on monocrotaLine-induced pulmonary vascular ili.sease and right ventricular hypertrophy. Lab. Invest., 52: 298-303.~~

LANIGAN, c.W. (1971) Metabol-isn of pyrroLizidine alkaloids in the ovine rumen. III. The competitive relationship between heliotrine metabolism and methanogenesis in rumen fLuid in vitro. Aust. J. agric. Res., 2Z:, LZ3-13O.

LANIGAN, c.W. (l-972) Merabolism of pyrrolizidine alkaloids in ovine rumen. IV. Effects of chloral hydrate and halogenated methanes on rumen methanogenesis and alkaloid metaboLism in fistulated sheep. Ausr. J. agric. Res., 3j: 1085-1091. LANIGAN, c.W. & SMITH, L.W. (1970a) 'aLkaloids Metabolism of pyrrolizidine in the ovine rumen. I. Formation of 7 ehydroxy-lrmethyl-8epyrroLizidine from heliotrine and lasiocarpine. Aust. J. agric. Res., 21-: 493-5O0. -25L-

LANTGAN, G.!I. & sulTlt' L.t'I. (1970b) Metabolisn of pyrrol"i- zidine alkaloids in the ovine rumen. II. Formation of 7 rhydroxy-1a-methyl-8c-pyrroLizidine from heliotrine and l,asiocarpine. Aust. J. agric. Res., 22: 123-L30.

~~IANIGAN, G.!{. & WIIITTEM, J.tt. (1970) Cobalt PelLets and Ileliotropium europaeum poisoning in penned sheep. Aust. vet. J,, 46.. L7-2L.~~

LANrGAlt, G.W., PAYNE, A.L., & PETERSON,J.E. (1978) Anti- methanogenic drugs and lleliotropium euroPaeum poisoning in penned sheep. Aust. J. Agric. Res., 29: I28L'L292. tAws, L. (1968) Toxicity of Crotal"aria mucronata- to sheep. Aust. vet. J., 44: 453-455.

LETENDEE,L. , Sl,lITltSON,W.A., GILCHRIST, G.S., BURGERT,E.O. r Jr, tlOGtAND, C.H., AI'IES, M.M., POWIS' G., & KOVACII, J.S. acute ( 1981) Activity of indicine-N-oxide- in refractory I-eukemia. !gg, Nz 437-441.

~~LETENDRE, L., LUDWIG, J., PERRAULT, J., SUITIISON, W.A.' & KOVACIi, J.s. (1984) Ilepatocellular toxicity during the treatment of refractory acute Leukemia with indicine-N-oxide. 9 ' 542 1256-t259.~~

~~LEVINE, O.R., TIARRIS, R.C., BLANCE, W.A., & MELLINS, R.B. (1973) Progressive pulmonary hypertension in children with portaL hypertension. J. Pediatr., 83: 964'972.~~

~~LIN, J.J., LIU, c., & SVOBODA,D.J. (1974) Long-term effects of af!.atoxin 81 and vocal hepatitis on marmoset l-iver; a prelininary report. Lab. Invest., 3O: 267-278,~~

LUTIIY, J., ZWETFEL,U., SCIILATTER,C., & BENN, M.H. (1980) IPyrrolizidine a].kaloids in coltsfoot (_fqi!]"go t".f."" 1,.) of various sources.l llitt. Geb. Lebensn. ttyg-- 71: 73-80 (in German with Englistr srrffifr-

~~LUTHY, J., Z!|ErFEL, U., KARLHUBER,B., & SCHLATTER,C. (1981) Pyrrolizidine alkaloids of Senecio alpinus L. and their dltecrion in feedstufts. .r. agffiil 29: 302'305.~~

~~LUTHY, J., IIEIU, TH., & SCHLATTER,CH. (1983) Transfer of [ 3H] pyrolizidine aLkaloids from Senecio vglgari-q L. and metabolites into r,at r,nilk and tissffi, lZ: 283-288. -252-~~

LUTIIY,.J., BRAUCIILI, J., ZI,IEIFEL, U., SCHMID, p., & SCIIIATTER, CH. (L984) [Pyrrolizidine alkaloid in arzneipflanzen der Boraginaceen: Borago officinalis L. and pul.nonarii officinalis L. I qbarr. 6g1;-Ej;;:EzF'i46 ( in ceffi

~~LYFORD, C.L., VERGAVA,G.G., & I{OELLER, D.D. (1976) Ilepatic veno-occLusive disease originating in Ecuador. Gastro- enteroLogy, 70: 105-108.~~

~~MCCOMISH,M.r BODEK, I., & BMUFMAN, A.R. (1980) quantitation of the antineoplastic agent indicine-N-oxide in human plasma by differentiaL pulse polarography. J-. pharnacol-. Sci., 69: 727-729.~~

MCCOYr.J.W.,ROBY, M.R., & STERMITZ, F.R. (1983) Analysis of plant alkaloid nixtures by amonia chemical ionization mass spectroscopy. J. nat. Prod., 46: 894-900. ryCGEE:J.0rD., PATRICK, R.S., I,ilOOD,C.B., & BLIIMGART,L.H. (1976) A case of veno-occlusive disease of the liver in Britain associated with herbaL tea consumption. J. c1in. Pathol., 29: 788-794.

~~MCGRATII, J.p.M., DttNcAN, J.R., & MUNNELL, J.F. (1975) CrotaLaria spectabilis toxicity in swiae: characterization of the renal glomerular lesion. J. comp. pathol ., g(2): 185-194.~~

~~MCLEAN, E.K. (1969) The earLy sinusoidal lesion in experimental veno-occlus ive disease. Br-:_j:_g.:_ESg}., 19: 223-229.~~

~~MCLEAN, E.K. ( 1970) The toxic actions of pyrrolizidine (Senecio) alkaloids. PharmacoL. Rev. 222,429-483.~~

~~MCLEAN, E.K. (L974) Senecio and other pLants as Liver poisons. Isr. J. med. sci-T107 $6-440.~~

~~MCLEAN,E.K. & HILL, K.R. (1969) portaL hypertension in acute experimental veno-occulsive disease of the liver in rats. Br. J. expf Pathol-., fl: 37-4L.~~

~~MCLEAN,E.K. , BRAS, G., & CYORGY,P. (1964) Veno-occlusive lesions in Livers of rats fed Crotalaria fulva. Br. J. exp. Pathol., 45: 242-247. -253-~~

uAKSITDOV, A.M. (1952) [tleliotropic toxic hepatitis with aecitee.l tn: Milenkov, S.M. & Kizhaikin, Y., ed. lco]"lection of scientif ic papers , on Toxic llgPati.tis yilh As-cileg, I Ta;Etena; Fublishing House of the University of Central Asia' pp. 103-116 (in Russian).

ITIARTIN, P.A., TIIoRBURNE,M.K., IIUTCIIINSON' S., BRAS, G., & MILLER, C.C. (L972) Preliminary findings of chromosome studies in rats and humans with veno-occlusive disease. Br. J. exp. Pathol., 2, 374-380. MATTOCKS,A.R. (1961) Extraction of heat-labile alkaloids (tond.)' from plants. Nature 4: l28L-L282.

~~MATTOCKS,A.R. (1967a) Spectrophotometric determination of unsaturated pyrrolizidine alkaloids. Anal. Chem., 392 443-447.~~

~~MATTOCKS,A.R. (1967b) Detection of pyrroLizidine alkaloids on thin-layer chronatograms. J. Chromatogr., 27: 505-508.~~

~~MATTOCKS,A.R. (1968a) Toxicity of pyrrolizidine alkaloids. Nature (Lond.) , 2!, 723-728.~~

~~MATTOCKS' A.R. (1968b) Spectrophotometric determination of pyrrolizidine alkaloids: some improvements. AnaL. Chem., 401 1749.~~

~~MATTOCKS, A.R. ( 1969) Dehydropyrrolizine derivatives from unsaturated pyrroLizidine alkaloids. J. Chem. Soc. , c] 969: 115 5-1 162.~~

MATTOCKS, A.R. ( 1971a) Synthetic compounds nith toxic properties simil-ar to those of pyrrolizidine alkaloids and Eheir pyrrolic metabolites. Nature (Lqnd_l' 232: 476. 'MATTOCKS, A.R. (197lb) The occurrence and analysis of pyrrolizidine alkaloid N-oxides. Xenobiotica, L: 45L-453.

~~MATTOCKS, A.R. (197lc) HePatotoxic effects due to pyrrolizidine alkaloid N-oxides. Xenobiotica' L: 563-565.~~

~~MATTOCKS,A.R. (1971d) A fieLd test for N-oxides of unsaturated pyrrolizidine alkaloids. TroP. Sci. ' g, 65-70.~~

~~MATTOCKS,A.R. (L972a) Toxicity and metabolisn of Senecio al-kaloids. In: Ilarborne, J.B. ' ed. Phytochemicq! egelegy., London, NewYork, Academic Press, pp. 179-200. -254-~~

~~MATTOCKS, A.R. (L972b) Acute hepatotoxiciry and pyrrolic metabolites in rats dosed with pyrrolizidine alkaloids. Chen.-bioL. Interact. 5z 227-242.~~

~~MATTOCKS,A.R. (1973) Mechanisms of pyrrolizidine alkaloid toxicity. PharmacoLogy and the future of man. In: proceedings of the 5th lra"rr?.igr.l Coogl."t pp.114-123.~~

MATTOCKS,A.R. (1977) Tissue distribution of radioactivity in rats given tritiated analogues of hepatotoxic pyrroLiziiine aLkaloids. Xenobiotica, !. 665-670. MATTOCKS, A.R. (1980) Toxic pyrrolizidine alkaloids in comfrey. Lancet, 22 November: 1136:1L37.

MATTOCKS,A.R. (198La) Liver cell enlargement in rats given hydroxynethyL pyrroles analogous to pyrrolizidine alkaloid metaboLites, foLLowed later by the hepatotoxin dirnethylnitros- amine. Toxicol. Lett., g:201-205.

~~MATTOCKS,A.R. (198lb) Relation of structural features to pyrrolic metabolites in livers of rats given pyrrolizidine aLkaloids and derivatives. ghen.-biol.. Inleqse!., 35:301-310.~~

~~MATTOCKS,A.R. (1981c) A sinple preparation of dehydroretronecine using potassium nitrosodisulphonate. Chem. Ind., 7: 251.~~

~~MATTOCKS, A.R. ( 1982) Hydrolysis and hepatotoxicity of retronecine diesters. Toxicol. Lett., 14:111-116.~~

~~MATTOCKS, A.R. ( 1986) Chernisrry and toxicology of pyrrolizidine alkaloids, Lottdo~~

MATTOCKS,A.R. & BrRD, r. (1993) Pyrrolic and N-oxide metaboLites formed from pyrroLizidine alkaloids by hepatic microsomes in vitro: relevance to in vivo hepaEotoxicity. chem.-biol. ffiiE, !!t 2og-222.

~~MATTOCKS, A.R. & CABRAL, J.R.p. (Lg7g) Effecrs of some pyrrolic and dehydropyrrolizine esters on mouse skin: a prel.iminary study. Tumori, 652 289-293.~~

~~MATTOCKS,A.R. & CABML, J.R.p. (1982) Carcinogenicity of some pyrrolizidine alkaloid metaboLites and analogues. Cancer Lett., 17: 6L-66. -255-~~

MATTOCKS,A.R. & DRMR, tt.E. (1983) A comparisoo of the pneumotoxicity of some pyrrolic esters and sinilar compounds analogous to pyrrolizidine aLkaloid metabolites given intravenously to rats. IgIiggLSy, 27: L59-L77.

~~MATTocKs, A.R. & DRIVBR' lt.E. (1937) Toxic actions of senaetnine, a new pyrrolizidine alkaloid, in rats. Toxicol. Lett., 38: 315-319.~~

~~MATTOCKS,A.R. & JUKES, n. (1987) New improved field test fof, toxie pyrrolizidine aLkaloids. J. nat. Prod., 50: 161-166.~~

MATTOCKS,A.R. & LEGG, R.F. (1980) Antimitotic activity of dehydroretroneeine, a pyrtoLizLdine alkaloid metatoliter afld sone analogous compoundb-in a rat liver parenchymaL cell Line. Chem.-bioL. Interact., 30: 325-336.

~~MATTOCKS,A.R. & WHITE, I.N.tt. (1970) Estimation of metabolites of pyrrolizidine alkaloids in anirnal tissues. Anal. Biochem., 38: 529-535.~~

~~MATTOCKS,A.R. & WI{ITE, I.N.It. (1971a) The cooversion of pyrrolizidine alkaloids to dihydropyrrolizine derivatives by rat-liver microsomes in vitro. Chern.-biol. Interact. ' 3: 383-396.~~

~~MATTOCKS,A.R. & WHITE, r.N.It. (1971b) Pyrrol ic metabo Lites from non-toxic pyrrolizidine alkaLoids. Nat. new Bio1. 23lt 114-115.~~

~~MATTOCKS,A.R. & WHTTE, I.N.It. ( 1973) Toxic effects and pyrroLic metabolites in the Livet of young rats given Ehe pyrrolizidine alkaLoid retrorsine. Chem.-biol. Interact., 6: 291-306.~~

~~MATTOCKS,A.R. & WHITE, I.N.Il. (1976) The distribution of [ 3It]-synthanecine A bis-n:ethylcarbamate and its metabolites in the rat. Chem.-bio1. Interact., fl: 173-1.84.~~

~~MEYRICK, B.o. & REID, L.M. 0979) Dpvelopment of pulmonary arterial- changes in rats fed Crotal-aria sPectabilis. Am. J. Pathol., 942 37-5O.~~

~~UEYRICK, B.0. & REID, L.t{. (1982) crotalaria-induced pulmonary hypertenqion: uptake of lf-thyrnidine by the cells of the pulmonary circulation and alveolar waLls. Arn. J. Pathol. L06: 84-94. -256-~~

~~r,rrlKowsKY,A.S. (1985) The synthesis and anti-tumour activity of monocyclic anaLogsof indicine-N-oxide. Diss. Abstr. Idt., 45(3): 844.~~

~~MILLER, W.C., RICE, D.L., KREUSEL,R.c., & BEDROSSIAN,C.W.M. (L978) Monocrotaline model of non-cardiogenic pulmonary edema dogs. in J. appl. Physiol. , E: 962-965.~~

MIRANDA,C.L., BUHLER,D.R., & CHEEKE,p.R. (1979) The effecr of tansy ragwort consumption on mineral metabolism in the rat. In: Cheeke, P.R., ed. proceedi of rhe s pyrrolizidine (Senecio) ALkaloidsl Toxicity. Met Poisonous Plant Control Measures Corva z5-z+ .[-eDruarv L9 l9 . Pp. or-o4.

MIRAI{DA, c.L.., CARPENTER,II.M. , CIIEEKE, p.R. , & BUIILERTD.R. (L981a) Effect of ethoxyquin on the toxicity of pyrrotizidine alkaloid monocrotaline and on hepatic long metaboiism in nice. Chem.-bioL. Interact., 37:, 95-L07.

~~MrPd,NDA, C.L. , HENDERSoN,U.C. , & BITHLER, D.R. ( lgglb) Dietary copper enhanees the hepatotoxicity of Senecio -iacobaea in rats. pharnacoL., Toxicol. app!.. $(3): 4LB-E--~~

MIRANDA, C.L., PGED, R.L., CHEEKE, p.R., & BUHLER, D.R. ( L98lc) Protective effects of butylated hydroxyanisole against the acute toxicity of monocrotaline in mice. Toxicol. appl. Pharrnacol.. 59: 424-430

MIMNDA, C.L., BUIiLER,D.R., RAI!{SDELL,H.S., CHEEKE,p.R., & SC-HMITZ,J.A. (1982a) ltodifications of chronic hepatoroxiciry of pyrrolizidine (_gggg&) alkaioids by butylarld hydroxyanisole and cystein"[email protected]_ t"tt._, LOz L77-LBZ.

MIRANDA, C.L. , ITENDERSON,M.C., BIIIILER, D.R., & SCHIIITZ, J.A. (1982b) Comparative effects of antioxidants on the toxicity of mixed pyrrolizidine aLkaloids from Senecio iacobaea i; mice. J. Toxicol. environ. Itealth, 9(5-6): gg:-9g9.

MIRANDA,C.L., HENDERSON,M.C., REED, R.L., SCIIMITZ,J.A., & BUIILER, D.R. (1982c) Protective action of zirc against pyrrolizidine alkaLoid-induced hepatotoxicity in rats. J. Toxicol. environ. Ilealth, lz 359-366.

~~MIROCIINIK,M.F. (1938) [Clinical course and etiopathogenesis of . t-o."].c hepatires with ascites. I uedical Institrg erapeu@ Tashkent, Medical Literatufe Pn6Tishers. -257-~~

MISERTJ.S., MISER, A.W., SMTTHSON,W.A., COCCIA,P.F., AMES, M.l'1., DAVIS, D.M., IIUGHESTC.S. r & crLCItRrST, c.S. (1982) A phase I trial of indicine-N-oxide in childhood naLignancy. Proc. Am. Soc. CLin. hg9L, 1.. 137.

~~MNUSHKIN,A.S. (1949) [Some materials to clinical features and pathogenesis of toxic hepatitis with ascites.l Sov. lled., L: 8-9 (in Russian).~~

UNUSTIKIN,A.S. (L952) [the clinicaL features, pathogenesis and treatment of toxic hepatitis vith ascites.l In: MiLenkov, S.M. & Kizhaikia, Y., ed. tC.tt."ti"" Toxic ltepatitis with Ascitesrl Tashkent,"t ""i.PubLishing llouse of the University of Central Asia, pp. 91-98 (in Russian).

~~MOHABBAT,O., SRMSTAVA, R.N., YOUNOS,M.S.,. SBDIq, G.G., MENZAD, A.A., & ARAI.{, c.N. (1976) Ail outbreak of hepatic veno-occlusive disease in north-western Afghanistan. :@!, 7 August: 269-271.~~

~~MOLTENI, A., I"IARD,W.F., TSTAO,C.-S., PORT, C.D., & SOLLIDAY, N.1I. (1984) Monocrotalineinduced pulmonary endotheliaL dysfunction in rats. Proc. Soc. Exp. Biol. ued., W, 88-94.~~

~~MOLYNELIX,R.J. & ROITMAN, J.N. (1980) Specific detection of pyrroLizidine aLka'Loids on thin-layer chromatograms. J. Chromatogr., E; 4L2-4L5.~~

MOLYNEUX,R.J., JOIINSON,A.E., ROITMAN, J.N., & BENSON,lt.E. (1979) Chernistry of toxic range plants. Determination. of pyrroLizidine alkaloid content and composition in Senecio species by nuclear magnetic reaonance spectroscopy. J.:Efc: f ood Chem., lt 494-499.

I'iORI, tl., SUGIE, S., YOSHIMI, N.r ASADA, y., FURUYA, T., & WILLIAMS, G.M. (1985) Genotoxicity of a variety of pyrroLi- zidine alkaloids in the hepatocyte prinary cuLture-DNA repair testusingrat'mouse'andhamsterhePatocytes.@-, 45:, 3L25-3129.

~~I'tllNIER, R. (L953) Separation of alkaloids from tlreir N-oxides by paper chrornatography. BuLl. Soc. Chin. Biol. , Yr 1225.~~

~~NARDITN.8., CIMMLER,U.C., & tErTE, B.c. (1990) Antinitotic action of integerrimine in rats. Rev. Bras. Genet. 3 z .387-392 . '258'~~

NATIONAL CANCER INSTITUTE (f978) Bioassay of lasiocarpine for possib!.e carcinogenecity, nethe@ Cancer Institute, US .Department of lteaLth, Education and !'felfare (Report of Carcinogenesis Testing Programe, Division of Cancer Cause and Prevention).

~~NAYM, N.C., SAGPGIYA,K., & RAMALINGASWAMI,V. (1969) Indian childhood cirrhosis - the nature and significance of cytoplasnic hyaLine of hepatocytes. Arch. Pathol., 88: 631-637.~~

~~NET,UBERNE,P.l,t. (1968) The influence of a Low isotrope diet on response of maternal and fetal rats to lasiocarpine. Cancer Res., 283 2327-2337.~~

~~NEWBERNE,P.M. & ROGERS,A.E. (1973) Nutrition, monocrotaline and aflat,oxin 81 in Liver carcinogenesis. Plant Food Man. L: 23-31.~~

NEWBERNE,P.M., IiILSON, R., & RoGERS,A.E. (1971) Effects of a low-Lipotrope diet on the response of young male rats to the pyrrolizidine alkaloid monocrotaline. Toxi,icol:___ap.gl: Pharmacol., L8z 387-397.

NEWBERNE,P.M., CHAN, I,l.C., & ROGERS,A.E. (1974) Inf luence of light, riboflavin and carotene on the response of rats to the acute toxicity of aflatoxin and .monocrotaline. Toxicol. appl". PharmacoL. , 282 2OO-2O8,

NrcHoLs, w.c., MoERTL, C.G., RUBTN, J., SCIII'LT, A.J., & BRITTEL, J.c. ( 1981) Phase II trial of indicine-N-oxide (Nsc132319) in patients with advanced colorectal- car;inoma. Cancer Treat. Rep., 65: 337-339.

NIWA, H., ISHII^IATA,H., & YAMADA,K. ( 1985) Isolation of pecasitenine, a carcinogenic pyrroLizidine aLkal-oid from japonicum. Farfugium J. nat. Prod., !q: 10083-1004. NOLAN, J.P., SCIIEIG, R.L., & KLATSKIN, G. (1966) DeLayed hepatitis and cirrhosis in weanling rats foLlowing a single smal1 dose of the Senecio alkaloid, lasiocarpine. Arn. J. Pathol., 49: 129-L5L

~~OHNUMA,T., SRTDHAR,K.S., RATNER, L.H., & HOLLAND, J.F. (1982) Phase I study of indicine-N-oxide in patients with advanced cancer. Cancer Treat. Rep., 66(7): L509-1515. -259-~~

OHTSUBO,K., ITO, Y., SAITO, I't., FURIIYA, T., & IIIKICIIIT l,t. (L977) with Hypertrophy-rats of pulmonary arteries and arterioles cor puLmonale in induced by seneciphyLLine, a pyrroli- (Basel)_, zidine alkaloid. Experientia 3: 498-499.

~~ol,soN, J.I,{.r HACKER,A.D., ALTIERE, R.J., & GTLLESPIE, M.N. (1984) Polyamines and the development of monocrotaline- induced puLnonary hypertension. Arn. J. Physiol., 247: n682-H685.~~

OLSON, J.W., ATKINSON,J.8., IIACKER,A.D., ALTIERE, R.J., & GITLESPIE, M.N. (1985) Suppression of polyanine biosynthesis prevents monocrotaline induced pu.lmonary oedema and arterial rnedial thickening. Toxicol . appl. PEry!., 81: 91-99.

oRD, M.J., HERBERT,A., & uATTOcKlS,A.R. (1985) The ability of bifunctional and monofunctional pyrrole conpounds to induce sister chromatid exchange (SCE) in human lymphocytes and mutations in SalmoneLla typhimurium.. Mutat. Res., !2t 485-493,

PANTONE,D.J., BRO!iN, S.M., & I'IOMER.SLEY,C. (1985) Biological controL of fiddleneck. s!!!91g!1!g:i5,2. 4-9. PECKHAM, J.C., SANCSTER, L.P., & JONES, o.n. (L974) Crotalaria spectabiLis poisoning in swine. J. An. Vet. Med Assoc., 165: 633-638.

~~PEDERSEN,E. (1975) PyrroLizidine aLkaLoids in Danish specles of the farniLy Boraginaceae. Arc\ Pharn. Chem. Sci. Ed., 3: 55-64.~~

~~PERCY, J.J. & PIERCE, A.E. (197L) Imrunosuppressive activity of the pyrrolizidine alkaloid metabolite dehydroheliotridine. fryfggl, 2I: 273-28o.~~

~~PERSAUD,J.V. & HOYIE, D.A. (1974) Pregnancy and progeny in rats treated with the pyrrolizidine alkaloid fulvine. Erc&- Pathol.., 9.' 59-53.~~

~~PESTCIIANKER, M.J. & GToRDANO, O.S. ( 1986) Pyrrolizidine al.kaLoids from f ive Senecio soecies. J. nat. 3 .' 49t 722-723.~~

~~PESTCHANKER,M.J., ASC}IERI, M.S., & GIORDANO,O.S. (1985a) Uspallatine, a pyrrolizidine aLkaloid from Senecio uspallatensis. Phytochemistry, 24; L622-L624. -260-~~

~~PESTCHANKER,l.!.J., ASCHERI, M.S., & GIORDANO,o.S. (1gg5b) Pyrrolizidine alkaloide from Senecio subulatus and glandulosus. S. Planra Med., l!: 165-ffi~~

~~PETERSON,J.E. (1965) Effecrs of pyrrolizidine alkaloid lasiocarpine N-oxide on nuclear and ce1l division in the Liver of rats. J. Pathol. Bacteriol.. 89: 153-171.~~

~~PETERSON,J.E. & CULVENOR,C.C.J. (1983) plant and fungal toxins. In: KeeLer, R.F. & Tu, A.T., ed. @ toxins, New York, Marcel Dekker, Vo!.. I, pp. 637-681.~~

~~PETERSON,J.E. & JAGO. M.V. (1980) Comparison of the toxic effects- of dehydroheli6tridine and heLiotiine in pregnant rats and their embryos. J. Pathol., 131": 339-355.~~

~~PETERSON, J.E. & JAGO, M.V. (1984) Toxicity of Bchium plantagineum (Patersont s curse) : pyrrol"izidine alkaloid poisoning in rats. Aust. J. agric. Res., g: 305-316.~~

~~PETERSON,J.E., SAMUEL,A., & JAGO, M.V. (L972) pathological effects of dehydroheliotridine, a metaboLite of heliotridine- based pyrroLizidine alkaloids in the young. rar. J:_!E$g!-1-, 1.07: L07-189.~~

PETERSON,J.E., JAco, M.v., REDDY, J.K., & JARRETT, R.G. (1983) Neoplasia and chronic disease associated rnrith the prolonged administration of dehydroheliotridine to rats. J. Natl Cancer Inst.,j9: 381-386.

~~PBTRYT T.W., BOVilDEN,G.p., HIIXTABLE, R.J., & SIPES, I.C. (1984) Characterizatiotr of hepatic DNA darnage induced in rats by the pyrrolizidine alkaloid monocrotaline. Cancer Ris., 44: 1505-1509.~~

~~PETRY, T.W., BOI'IDEN,G.P., BIIILER, D.R., & Slpgs, I.c. (1986) ceootoxicity of the pyrrotizidine alkaloid jacobine in rats. Toxicol. Lett., 32: 275-281.~~

~~PTERCY, P.L. & RUSOFF, L.L. (1946) Crotalaria spectabi_Lis poisoning in Louisiana livestock. J. Arn. Vet. Med. Assoc., 108:69-73.~~

~~PTERSON, M.t., CHEEKE, p.R., & DTCKINSON, E.O. (1977) Resistance of the rabbit to dietary pyrroLizidine ( Senecio ) alkaloid. Res. Comun. qhem. PathoL. Pharmacol., 16; 5ffiffi: -26L-~~

rIl PIETERS, L.A.C. & VLIETINCK, A.J. (1985) Quantitative Fourier transform nuclear magnetic resonance spectroscotric analysis of mixtures of pyrrolizid:ine alkaloids from Senecio vuLgaris. Freseniusi Z. Anal . Chem., 4z 355-358. PIETERS, L.A.C. & VLIETINCK, A.Jt. (1986) Comparison of high-performance Liquid chromatography with rH nuclear magnetic resonance spectrometry for the quantitative analysis of pyrrolizidine alkaLoids from Sienecio vulgaris. J. liq. Chromatogr., 2: 745-755.

~~PLESTINA, R. & STONER, Il.B. (L972) Pulmonary oedema in rats given nonocrotaline pyrrole. J. Pathol., LO6z 235-249.~~

~~POOL, B.L. (1982) Genotoxic activity of an alkaloidal ext,ract of Senecio nemorensis spp.- -svsEns-t fuchsii in Salmonella tvphimurium anclffiFi6--?6f i - I e ttffi r oxi coJ6Ei, 34-35'i=355.~~

~~POWIS, G., AMES, M.M., & KOVACII, J.S. (1979) MetaboLic conversion of indi.cine-N-oxide-3564-3570. to indicine in rabbits and humans. Cancer Res., 39:~~

~~PRABIIU,M.B. (1940) Infantile cirrhosis of liver. Indlan J. Pediatr., 7: L21-134.~~

~~RADHAKRISHNARAO, M.v. (1935) Hist,opathology of the liver in infantile biliary cirrhosis. Indian.I. med. Res 23t 69-9A.~~

~~RAIAGOPALAN, T.R. & NECI, R.K.S. (1985) ALkaloids from Doronicum pardalianches Linn. Indian J. Chem., 248, 882.~~

RAKIETEN, N., GORDON,B.S., BEATTYTA., COONEY,D.A., DAVIS, R.D., & SCIIIEN, P.S. (1971) Pancreatic islet cell tumours produced by the combined actiorr of streptozotocin and nicotinanide. Proc. Soc. Exp. Bio1. Med., B7t 280-283.

~~RAMALINGASWAI'I,V. & NAYAK, N.C. (1969) Liver disease in India. In: Popper, It. & Schaffner, F., ed. !@__ig_fiIgg diseases, New Yorkr.London, Grune and Stratton, pp. 222-235.~~

RAI'ISDELL,A.S. & BIIHLER, D.R. (1981) Itigh perfornance liguid chromatographic analysis of pyrrolizidine (Senecio) alkaloids using a reversed phase styrenedivinylbenzenffi columr. J. Chromatogr., 4: 154-158. RAO, M.S. & REDDY, J.K. (1978) M:rLignant neoplasms in rats fed lasiocarpine. Br. J. Cancer, yr 289-293. -262-

~~MPPAPORT, A.M., KNOBLAUCIT,M., SI'MMERSKILL' J. ' & BRAS, G. (1967) Experimental veno-occlusive disease. Gastroenterolgg, 542 L64.~~

~~RATNOFF,O.D. & MIRICK' C.S. (1949) Influence of sex uPon the lethaL effects of an hepatotoxic alkaloid, monocrotaLine. Bu1L. Johns llopkins tlosP., 843 507-25.~~

~~RECIIICIGL, M., Jr, ed. (1983) CRC llandbook oa naturallY occurring food toxicants' Boca naton, ftofida, CRC Press, 266 pp.~~

~~RESCH, J.F. & MEINWALD,J. (1982) A revised structure for acetylheliosupine. Phytochenistry' 2lz 2340-243L.~~

~~RIIODES, K. (1957) Two tyPes of liver disease in Jamaican chiLdren. West Indian med. J. ' 9r L'29.~~

RIDKER, P.M., OHKI]I,TA,S., MCDERMOTT,W.V., TREY, C., & I{UXTABLE, R.J. (1985) Itepatic ' veneocclusive disease associatid with consupption of pyrroLizidine alkaloid containing dietary suPpLements. Gastroenterology' 4, 1050-1054.

RIZK, A.M., HAMMOUDAT,F.U., rSMAIt, S.I., GIIALEB' Il.A. ' MADKOUR,M.K., POHIAND, A.E., & WOOD' G. (1983) Poisonous plants contaminatiag edible ones and toxic substances in plant ioods. I. Alkaloids from Senecio desfontainei. Ej@., 54: Ll5-L21.

~~ROBERTSON,K.A. (19S2) Alkylation of N2 in deoxyguanosine by dehydroretronecine, a carcinogenic metabol"ite of the pyrro1izidinea1ka1oidmonocrota1ine.@,!z.8-L4.~~

ROBINS, D.J. (1982) The pyrrolizidine alkaloids. In: Herz, W., Grisebach, H., & Kirby, G.W., ed. Progress in the chemistry of organic natural products, Vienna, New Yorkt Bp-inget-Verlag, pp. 115-203.

~~RODER,E., WNIDENFELD,H., & STENGL,P. (1981) Pyrrolizidine alkaloids senecionine and retrorsine from Senecio inaequidens. Planta Med., 4]-: 412'4L3.~~

RODER, E., WTEDENFELD,H., & KNOZINGER-FISCIIER (1984a) PyrroLizidine alkaloid from Senecio abrota4ifoLius sPp. abroatanifoLiusr spp. abrotanifoLius, and var. tiroliensis. Pl-anta Med.. 41: 4L2-413. -263-

~~RODER, 8., IdTEDENFEID, H., & BRITZ.KIRSTGEN, R. ( 1984b) PyrroLizidine alkaloids from Senecio cacaliaster. Phyto- chemistry, 23: L76L-L763.~~

ROGERS, A.E. & NEI'IBERNE,P.M. (f971) Lasiocarpine factors infl"uencing its toxicity and effects in liver ce1l division. ToxicoL. appl. Pharmacol., 18: 356-366. noiru*u, J.N. (1981) Comfrey and liver damage. @9!, 25 ApriL:944.

ROITMAN, J.N. (1983) Ingestion of: pyrtoLizidine alkaloids: a health hazard of globaL proportions. In: Finlay, J.tr{. & Schwass, D.8., ed. Xenobiotics in foods and feeds, Washington DC, Amer i c an cherni c aT-6ffi!!f 365:578-GTB-G i e s ) .

ROSE, A.L., GARDMR, C.A., MCDONNELL,J.D., & BULL, L.B. (1957a) FieLd and experinental. investigation of rrwalk-aboutrl disease of horses (Kimberl-ey horse disease) in northern Australia3 CrotaLaria poisoning in horses. Part II. Aust. vet. J., 33; 49-67:-

ROSE, A.L., GARDNER,C.A., MCDoI$NELL,J.D., & BULL, L.B. (1957b) Fiel-d and experimental inLvestigation of rrwalk about" disease of horses (Kirnberl-ey horse disease) in northern Australia. Part I. Aust. vet. J., 33t, 25-33.

~~RosE, C.L., HARRIS, P.N., & CHEN, K.K. (1959) Some pharmacological action of supinine and lasiocarpine. J. Pharmacol. exp. Ther. , L%t L79-L84..~~

~~RosE, E.F. (L972) Senecio species: toxic plants used .as food and nedicine in the Transkei. s. Af:r. Med. J., 46: 1039-1043.~~

~~ROSENFIELD,I. & BEATH,o.A. (1945) risstue changes induced by Senecio riddellii. An. J. cLin..Pathol.. 15: 407-4L2.~~

~~ROTH, R.A., DOTZLAF,L.A., BAMNYI:, B., & HOOK,J.B. (1981) Effect of monocrotaline ingestion on liver, kidney and Lung of rat. Toxicol . appL. PharmacoL., q: 193-203.~~

~~ROYES, K. (1948) Infantile hepirtic cirrhosis in Jamaica. Caribb. ned. J.. 10: 16-48.~~

~~SAFOIIII, M. & SHEHATA, A.H. ( 1965) Itepatic vein occlusion disease of Egyptian children. J. ?e,diatr. , 672 4L5-432. -264-~~

~~ST. GEORGE-GRAMBAITER,T.D. & RAC, R. (L962) Hepatogenous chronic copper poisoning in sheep in south Australia due to consumption of Echiun plantagineum L. (Salvation Jane). Auet. vet. J.,38:288-293.~~

~~SALASPURO, M. & SIPPONEN, P. (L976) Demonstrarion of an intracellular copper-binding protein in orcein staining in long standing cholestatic Liver diseases. Gut, 17: 787-790.~~

SAMUEL,A. & JAGO, M.V. (1975) LocaLization in the cell cycle of the antimitotic action of the pyrroLizidine alkaloid, lasiocarpine, and its metabolite, dehydroheliotridine. Chern.-biol.. Interact., 10: 185-197.

~~SANDERS, D.A., SIIEALY, A.L., & EMMEL, M.W. (1936) The pathology of Crotalaria spectabilis Roth poisoning in cattle. J. Am. Vet,. Med. Assoc., 9g: 150-156.~~

~~SAVIN, I.c., (1983) [Research of the influence of heliotrin on the microsornal rnono in biologicaL chenistry) (in Russian).~~

SAWINA, K.I. (1952) Pathological anatomy of atrophic hepatic cirrhosis. Arch. Pathol. , !!: 65-7O. SCIIOENTAL,R. (1959) Liver lesions in young rats suckLed by mothers treated with the pyrrolizidine (Senecio) alkaloids, lasiocarpine, and retrorsine. .1. p.th"1]-T6E.lig1., 772 485-504.

~~SCIIOEtiltAL,R. (1961) Iterbal medicines and liver disease. J. exp.'meil. Sci. , Lr L26-L35.~~

~~SCHOENTAL,R. (1963) Liver disease and Inaturalt hepato- toxins. Bull. World Health Organ., 29t 823-833.~~

~~SCHOENTAL,R. (1968) Toxicology and carcinogenic action of pytrolizidine alkaLoids. Cancer Res. , 2Bz 2237-2246.~~

SCHOENTAL,R. ( 1970) Hepatotoxic activity of retrorsine, senkirkine and hydroxysenkirkine in newborn rats, and the role of epoxides in carcinogenesis by pyrrolizidine aLkaloids and aflatoxins. Nature (Lond.), 227'. 40L-4O2.

SCHOENTAL,R. (1975) Pancreatic islet-cell and other tumours in rats given heliotrine, a monoester pyrrolizidine a1kaLoid, and nicotinamide. Cancer Res., 35: 2020-2024. -265- scHoENTAt, R. & BBNSTED,J.P.M. (1t963) Effecte of whole body irradiation and of partial hepatclctomy on the liver lesions induced in rats by a single dose of retrorsine, a pyrrolizidine (Senecio) alkaloid. Br. J. Cancer' yr 242-251. sclloENTAL, R. & CAVANAGH,J.B. (1972) Brain and spinal cord tumours in ra ts treated with pyrrolizidine alkaloids. J. Natl Cancer Inst., 49; 665-67L.

~~SCHOENTAL,R. & COADY, A. (1968) The hepatotoxicity of some Ethiopian and east African plants including some used as tiaditional medicines. East Afr. ned. J., 45: 577-579.~~

~~SCHoENTAL, R. & IIEAD, !{.A. (195:t) PathologicaL changes in rats as result of treatment with rronocrotaline. Br. J. Cancer, 9z 229-237.~~

SCITOENTAL,R. & IIEAD, M.A. (L957) Progression of Liver lesions produced in rats by temporrary treatment with pyrrolizidine (_9S""S1g) alkaLoids, and the effects of betaine and high casein diet. Br. J. Cancer, !][_: 535-544. scl{oENTAL, R. & MAGEE, P.N. (L957) Chronic Liver changes in rats after a single dose of Lasiocarpine, a pyrroLizidine (Senecio) aLkaloid. J. PathoL. Bacrgeriol. , 74: 305-3 19.

SCIIOENTAL,R. & MAGEE,P.N. (1959) Further observation on the subacute and chronic liver changes rats after a singLe dose of various pyrrolizidine (sengc:!q,) alkaloids. J. Pathol . Bacteriol.,' 78: 47L-482.

~~SCHOENTAL,R. & MATTOCKS,A.R. (19160) Ilepatotoxic act-ivity of semisynthetic analogues of pyrr,rLizidine. alkaLoids. Nature (Lond.), 185: 842-843.~~

scuoENTAL, R. , HEAD, M.A. , & PEA,COCK,P.R. (L954) Senecio alkaloids: primary liver tumoursi in rats as a result of treatment with (1) mixture of alkaloids from 9:-jgg!33 l'i"' (2) retrorsine, (3) isatidine. Br. J. Cqnger' 8:458-465.

SCHOENTAL,R., FOIiLER,M.E., & COADY' A. (1970) Islet ce11 tumours of the pancreas found in rats given pyrrolizidine alkal"oids from AmsincLia intermedia Fisch and Mey and from Hel iotropiun supfriii-Gffi?ilfr z 2127 -2131.

~~SEGAIL, II.J. (1979a) Reversed phase isolation of pyrrolizidine aLkaloids. Liq. ChrorIE!.€1, 2z 429-436. -266-~~

~~SEGALL, n.J. (1979b) Preparative isolation of pyrroLizidiae alkaloids derived fron Senecio vulgaris. Liq. Chromatogr., 2: 1319-1323.~~

~~SEGALL, H.J., DAT.LAS, J.L., & IIADDoN, w.F. (1984) Tno dihydropyrrolizine aLkal,oid metaboLites isolated from nouse hepatic microsomes in vitro. Drug Metab. Disp., y, 68-7L.~~

~~SEGALL, H.J., WILSON, D.W., DALLAS, J.L., & HADDON,W.F. (1985) Trans-4-hydroxy-2-hexenol: a reactive metabolite from the macrocyclic pyrrolizidine alkaloid senecionine. Sciegllir 229: 472-475.~~

~~SEGER, C.L., NEWSoN,J.D., ROTII, E.E., & HUTCIITNSON,W.R. (1969) Chronic toxic hepatitis in deer frorn a Louisiana coastal marsh. Bul1. Wildlife Disease Assoc. Proc. Ann. Conf., pp.295-6.~~

~~SELZER, G. & PARKER, R.c.F. (1951) Senecio poisoning exhibiting as Chiarirs syndrome: a report oI--12-EEses. An. J. PathoL., 27: 885-907~~

~~SENER, 8., TEMTZER,It., TEMIZER, A., & KAMKAYA, A.E. (1996) High performance f.iquid chromatographic determination of aLkaLoids in Senecio vernalis. J. pharn. Belg., 41; 115-1.17.~~

~~SHARMA,R.K., KIIAJURIA, c.S., & ATAL, C.K. (1965) Thin layer chromatography of pyrrolizidine alkaLoids. J. Chromatogr., 19: 433-434.~~

~~SHERLOCK,S. (1968) Diseases of liver and bilia tem, 4rh ed., Oxford, Edinbuigh, s, 250 pp.~~

~~SHIENBERG,A.I. & ORIOVA, N.V. (1955) [The question of rhe tt0zhalangar etiology of the so-ca1led Encephalitisrt.] Vopr. Pitan., L4: 27-3L (in Russian).~~

~~SHttLL, L.R., BUCKMASTER,c.W., & CIIEEKE, p.R. (1976) Facrors influencing pyrrolizidine (Senecio) aLkaloid metabolism: speeies, liver sulphydryls anF& fermentation. J. anim. Sci., 43: L247-L253.~~

SHI'MAKER,R.C., ROBERTSON,K.A., HSU, I.C., & AI,I,EN, J.R. (L976) Neoplastic transformation in tissues of rats exDosed to monocrotaLine or dehydroretronecine. J. Natl Cancer Inst., 562 787-789. - 267 .-

(1978a) SIDDIQI , M.A. , SURI, K.A. , SIIRI , M.P., & ATAL, C.K. Novel pyrrolizidine alkaLoid from Crotalaria nana. Phytochemistry, L7 z, 2L43-2144. (1978b) srDDIQr, M.A.r SI,RI, K.A. , suRr, o.P., & ATAL, C.K. Genus Crotalaria. Part 34. Cronaburmine, a new pyrrolizidine a Lkat off%i--dro t aL ar ia nana Burm. Indian J. Chem., 168: 1132-1133.

~~Sl{ITIl, L.W. & CIILVENOR' C.C.J. (1981) Plant sources of hepatotoxic pyrroLizidine alkaloids. J. nat. Prod., Lt L29-L52.~~

SMITI{, T.E., WEISBACI{, I{., & UT'ENFRTENDS. Q962) studies on the mechanism of monoaminooxidase: metabolism of N'N-dinethyltryptarnine-N-oxide' !,!-dinethylttyptarnine and Biochemistry, 13 137. 5OBIN, L.H., FETRAT, M.E., & .ANtlER, M.A. (1969) ttepatic lesions in Afghanistan. @!Ir. Med., 21-'. 27-29. SRIVASTAVA, R.N., MOIIABBAT,O., GHANI, A.R., & ARAM' G.N. (1978) Veno-occlusive disease, of the Liver. lqlian. !" g, 15: L43-t46.

SRITNGBOONMEE'S. & MASKASAI{E,C. (1981) A preliminary study on the toxicity of Crotalaria jrlgggg to cattle. Sattarilaphaet !9, 9: 91-107. STEIN, lI. (1957) Veno-occLusive disease of Liver in African children. Br. med. J., 29.|g: L496'L499-

STEIN, H. & ISMCSON, C. (L962) Veno-occ!'usive disease of the Liver. Br. med. J.' .1@2 372'374. SfENMARK, K.R., MORGANROTII,M.L., REUTGTO, L.K., VOELKEL' N.F., MURPIiY,R.C., IIENSON,P.U., MATI{rAS,14.M.' & BSEVES,J. ( 198i) Alveolar inflarnmation arrd arachidonate metabolism in monocrotalin*induced pulmonary hypertension. Ao. J. Physiol., 248: H859-n866.

~~STILLMAN, A.8., IIUXTABLE, R.J" CONSRoE,P., KOIINEN, ?.' & sMITtt, s. (1977) IlePatic veno-occLusive disease due to pyrroiizidine poisoning in Arizona. Gastroenterology, l2: 349-352.~~

~~STIRLING, G.A., BRAS' G.' & URQIIHART,A.E. (1962) The early lesions in veno-occlusive disease of the liver. Arch. dis' child., 37: 535-538. -268-~~

~~STOYEL, C. & CLARK, A.M. (1990) The transplaceatal micronucleus test. lrtutat. Res. , 742 393-39g.~~

~~STuART, K.L. & BRAS, c. (1955) Clinical observations on veno-occlusive disease of the liver in Jarnaican adults. Br. med. J., 2: 348-352.~~

~~STUART, K.L. & BRAS, c. (1956) Veno-occlusive disease of the liver in Barbadoe. trlest Indian med. J., 5: 33-36~~

~~STUART,K.L. & BRAS, c. (1957) Veno-occlusive disease of the liver. Q. J. Med., 26t 291-3L5.~~

~~STYLES, J., ASItBy, J., & MATTOCKS,A.R. (19S0) Evaluation in vitro of several pyrrotizidine alkaloid carcinogens: observF tion on the essential pyrroLic nucleus. Carcinogenesis, 1: 16r.-164.~~

~~SUFFNESS,M. & CORDELL,c.A. (1935) Antirumour alkaloids. In: Brossi, A., ed. The alkaloids, New york, Academic press, pp.1-347.~~

suGITA, T., HYERS, T.M., DAUBER,I.M., WAGNER,w.t{., ucMURTRy, I.F., - & REEVES, J.T. (1983a) Lung leak precedes righi ventricular hypertrophy in monocrottline treated rats. J. appL. Physiol. , \z 37L-374.

SUGITA, T., STENI\,IARK,K.R., WAGNER, W.W., HBNSON, p.M., HENSON,J.8., HYERSTT.M., & REEVES,J.T. (l9S3b) Abnormal alveolar ceI-1s in monocrotaline induced pulmonary hypertension. Exp. lung Res. , 5z 2OL-215.

~~SUNDARSSoN,A.E. (L942) An experimental s tudy of placental perrneability to cirrhogenic poisons. J. Pathol. Bacteriol 54: 289-298.~~

~~svdBODA, D. & REDDY, J.K. (1972) Malignant tumours in rats given lasiocarpine. g}8-gl|. Cancer Res. , Lt~~

~~SVOBODA, D. & REDDY, J.K. (L974) Lasiocarpine induced, transplantable squamous ce11 carcinoma of rat skin. J. NatL Cancer Inst., 53; 1415-1418.~~

~~svoBoDA,D. & SOGA,J. (1966) Early effects of pyrrolizidine alkaLoids on the fine atructuf,e of rat liver cells. Am. J. Pathol., 48: 347-373. -269-~~

SWICK, R.A., CIIEEKE, P.R., & BITIILER, D.R. (1979) Factors affecting the toxicity of dietar:y tan8y raglrort to rats. In: Cheeke, P.R., ed. Proceedings of the SynPosium on Pyrrolizidine (Senecio)-I1lifffi Poisonous PLant Control Measur:es. Corvallis, Oregon, USA' Inst itute. shtlcK, R.A., CIIEEKE, P.R., GOEGER,D.8., & BT'IILER, D.R. ( 1982a) Effect of dietary Serneciolnec1() lacoDaeaiacobaea and injected -!------s- Senecio alkaloids and nonocrotEiSE-;-ffiE pics. J-.an^in. EEi- -::., 1411-1416. SWICK, R.A., CHBEKE,P.R., & BIIIILER, D.R. (1982b) Subcellular distribution of hepatic copper, zirl.c and iron and serum ceruloplasmin in rats intoxicated by oral pyrroLizidine (Senecio) alkaloids. J. anim. Sci.:-, 55(6)t L425-L430.

sl,llcK, R.A. , CI|EEKE, P .R. , PATTON, N.M., & BtHLER, D.R. (L982e) Absorption and excreti,on of pyrrolizidine (@19) aLkaloids and their effects on nnineral netabolism in rabbits. J. anisr. Sci. , Dr L4L7-L424.

SI,IICK, R.A. , CHEEKE, P.R., MIIRANDA, C.L., & BUIILER' D.R. (1982d) The effect of consurnption of the pyrrolizidine alkaloid containing plant Senecio jacobaea on iron and copper metabolism in thJ rat. FT;Effi,-f=nviron. Health, L0: 7 57-7 68.

~~swlcK, R.A., CHEEKE, P.R., RAl.'tSDEtL,H.S., & BI'HLER, D.R. (1983) Effect of sheep rumen fermentation and methane inhibition on the toxicity of Seqeeio jacobaee. J. anim. Sci., 56: 645-651.~~

~~TAKAI{AS!ir, H. , I}MEDA, M., & EIRONO, I. (1980) Chromosomal aberrations and mutation in cultured mamaLian cells induced by pyrroLizidine alkaloids. llutat. Res., 78: 67'77.~~

~~?AKEOKA, O., ANGEVINE, U., & LAt:tCIt, J. (1962) Stirnulation of mast ceLLs in rat fed various chemicals. Am. J. Pathol.,40; 545-554.~~

~~TANDON,B.N.r TANDON,II.D., TANDON,R.K., NARENDRANATHAN,I,{., & JOSnr, Y.K. (1976) Epidenic of veno-occlusive disease io eentral India. Lancet, 7 August: 21L-272. -270-~~

TAIiIDON,8.N., TANDON, H.D., KOSIIY, A., NARENDMNATHAN,M., JOSHI, Y.K., TANDON,R.K., BilARGAVA,S., RAJANI, M., BIIATIA, M.L., MANCHANDA,S.C., & KASTURI, T.E. (1977) Epidenio- logical, clinical, biochemical and haemodynamic study of veno-occlu8ive disease of the Liver due to Crotalaria alkaLoids in India. J. A11 Ind. Inst. Med. Sci., 3: 16ffi-

~~TANDoN, B.N., TANDON,H.D., & MATTOCKS,A.R. (1978) Srudy of an epidemic of veno-occLusive disease in Afghanistan. Indiatr J. rned. Res., 68: 84-90.~~

~~TANDON,n.D. & TANDON,B.N. (1975) idemic of liver disease - Gulran District. Ilerat Province --UFfTce stan. A -or the Eastern M;aii..-;;;;;; G"'Jil"."ot reporr No. EM/Mc/oCDl 001/RB),~~

~~TANDON,H.D., TANDON,8.N., TANDON,R., & NAYAK, N.C. (1977) A pathoLogical study of the liver in an epidemic outbreak of veno-occlusive disease. Iqdian J. med. Res. 652 679-684,~~

~~TANDON, II.D., TANDON, B.N., & MATTOCKS, A.R. (1979) An epidernic of veno-occlusive disease of the liver in Afghanistan. Am. J. Gastroenterol., 72:607-6L3.~~

~~TANDON, R.K., TANDON, B.N., TANDON, H.D., BHATIA, M.L., BHARGAVA,S., L[, p., & ARORA, R.R. (L976) Srudy of an epidenric of veno-occlusive disease in India. Gut, 173 849-855.~~

~~TANMR, I'{.S. & PORTMANNTB. (1981) Indian childhood cirrhosis. Arch. dis. Child. , &i 4-6.~~

~~TAYLOR,S., BELT, R.J., HMS, C.D., & HoOGSTMTEN,B. (1993) Phase I triaL of indicine-N-oxide on t$ro dose schedules. Cancer,51: L988-1991.~~

~~TEILIIM, c. (L949) Endophlebitis heparica oblirerans. Acta pathoL. microbiol., ?9i L47-166.~~

~~TEMIZER, A., 0NAR, A'.N., SENER, 8., TEMIZER, ll., & KAMKAYA, A.E. (1985) Determination of alkaloids by differential pulse polarography. I. Senecio alkaloids. J. Pharsr. Belg., 4O(2)z 75-78.~~

~~TEREKIIOV, c.N. (1939) [Pathomorphology of toxic hepatitis with ascites.l In: Mirochnik, M.F., ed. [Functional, diagnostic and patholosical chanses of toxic hepati-ETE--ilffi~~

~~Technology and Socio-economic Literature of Uzbekistan, pp. 145-200 (in Russian). -271-~~

TEREKItov, c.N. (1952) [ttre pathomorphology of alimentary toxic dystrophy with ascites.l In: Milenkov, S.M. & Kizhaikin, Y., ed. ICollection of scientific papers on Toxic Hepatitis with Ascites rT Tas , PubII.s Ilouse of the Un verslty o Central Asia, pp. 148-164 (in Russian).

~~TIIORPE, E. & FORD, E.J.It. (1968) DeveLopnent of hepatic lesions in caLves fed with rag\4rort (Senecio jacobaea). J. conp. PathoL., J9: 195-205.~~

~~TITTEL, G. , IlINz, H., & WAGNER, II. (L979) Quantitative determinat.ion of pyrrolizidine alkaloids io !ylnP!y!.tj$ by HPLC. Planta Med., 37: L-8.~~

TUCttl/iEBER,B.r KOVAKS,K.' JAGO, M.V.' & BEAULIEU' T. (L974) Effect of steroidal and non-steroidal microsomaL enzyoe inducers on the hepatotoxicity of pyrrol-izidine aLkaloids in rats. Res. Comun chem. Pathol. Pharrndcol., 7: 459-480.

TUCKER, A., BRYAIIT, S.8., FROST, H.H., & MlGAttY, N. (1983) Chenical sympathectomy and serotonin inhibition reduce monocrotaline-induced right ventricuLar hypertrophy in rats. Can. J. Physiol. PharnacoL., 6Lt, 356-362.

~~TURNER,J.lt. & LALICE, J.J. (1965) ExPerimental cor puLnonale in the rat. $h.&!bl, !22 4o9-4L8.~~

~~VALDMA, E., SONNAD,J., HAYASIII, Y., & LALICH, J.J. (1967a) ExperimentaL interstitiaL puLmonary oedema. AngioLogy, L8: 378-383.~~

~~vALDrVrA, E., LAtrCH, J., HAYASHT,Y., & SONNAD,J. (1967b) ALterations in pulmonary alveoLi after a single injection of monocrotaline. Arch. Pathol. , 84: 64-76.~~

~~vAN DER WATT, J.J., PURCITASE,r.F.H., & TUSTIN, R.C. (1972) The chronic toxicity of retrorsine! a pyrrolizidine alkaloid, in vervet monkeys. J.Pa@!., IOTz 279-287.~~

~~VrLr,ARROEL,L.8., TORRES,R.G., NAVARRo,J.l'l., & FAJARDO, V.M. (1985) Senecionine and seneciphyll"ine from Senecio patagonicus. Fitoterapia, $:, 250-25L.~~

~~VISCONTINI, M. & GILITOF.SCITAUFELBERGER,I{. (1971) Synthesis of (1) dehydroheliotridine. Ilelv. chim. Acta, 54:449-456.~~

~~WAGENVOORT,C.A., DTNGEMANS,K.P., & TOTGERING,G.G. (L974a) Electron microscopy of pulmonary vasculature after application of fulvine. Thorax, 29: 5Ll-52L. -272-~~

~~WAGBNVOORT,C.A., WAGENVOORT,N., & DIJK, It.J. (1974b) sffect of fulvine on pulmonary arteries and veins of the rat. Thog, 29: 522-529~~

~~WAGNER,Il., NEIDHARDT,U., & TITTEI, c. (l9gl) TLC and HPLC anaLysis of pyrrolizidine N-oxide- aLkaloids of Slmphyti radix. Planta Med., 4L: 232-239.~~

~~WAKIM,K.G., HARRIS, p.N., & CHEN, K.K. (1946) The effecrs of pharmacoL. senecionine on the monkey. J. exp. Ther. r gT:3g-45,~~

~~rfAtKER, K.n. & KTRKLAND,p.D. (1981) Senecio lautus toxicitv in cattLe. Aust. ver. J., 572 L-7.~~

~~I,{ATT, J.M. & BREYER-BRANDWIJK,M.c. (L962) Medicinal and plalts ryi:onoug of s ern and .astertt Afri"Eli'ii6lilGl t ondon~~

~~WEHNER,F.C., TIIIEL, p.G., & VAN RENSBURG,S.J. (1979) Muta- genicity of alkaloids in the Salnooella/microsome system. Mutat. Res., !.j1: 187-190.~~

~~wESTON, C.F.M., COOPER,B.T., DAVTES, J.D., & LEVINE, D.F. (1987) Veno-occlus,ive disease of the liver secondary to ingestion of comfrey. Br. med. J., !$: 183.~~

~~WIIITE, I.N.tt. (1976) The role of liver gJ.urathion in rhe acute toxicity of' retrorsine to rats. Ctren.-biol. Interact., L3z 333-342.~~

~~WHITE, I.N.H.~~

~~WIIITE, I.N.H. & UATTOCKS,A.R. (1971) Some facrors affecting the conversion of pyrroLizidine alkaLoids to N-oxides and to pyrrolic derivatives in vitro. Xenobiotica, l_: t03-505.~~

WIIITE, I.N.H. & MATTOCKS, A.R. (L972) Reacrions of dihydropyrrolizidine with deoxyribonucteic acid in virro. Biochen. J., LBt 29L-297. tIt{ITE, I.N.II., I{AT?OCKS,A.R., & BUTIER, W.H. (1973) The conversion of the pyrrolizidine alkaloid retrorsine to pyrrolic. derivatives in vivo and in vitro and its acute toxicity to various anTilil-!?cies. ffi6il]6E1 . Inreracr., 6: 207-2t8. -273-

WHITE, R.D., KRUMPERIT{AN,P.H., CHEEKE, P.R., & BUIIIER, D.R. (1983) An evaLuation of acetone extracts from six plants in the Amesmutagenicity test. Toxicol. Lett., 15: 25-31. wItITE, R.D., SWICK, R.A. , CIIEEKE, P.R. (1984) Effect of dietary copper and molybdenum on tansy ragwort (Senecio Jacobaea) toxicity in sheep. @: J, Y9!:-39a, 45: 159-161. m{O (1980) Inventory of medicinal plants used in the different cou @3).

WICKRAI.{ANAYAKE,p.p., ARBoGAST,B.L., BITHLER,D.R., DEINZER, M.L., & BURLINGAI'{E,A.L. (1985) Alkylation of nucleosides and nucleotides by dehydroretronecine: characterization of covalent adducts by liquid secondary ion mass spectrometry. J. Am. Chem. Soc., L07: 2485-2488.

~~WIEDENFELD,H. (1982) Two.pyrroLizidine alkal-oids from Gynura scandens. Phytochemistry, 2Lt 2767-2768.~~

~~WIEDENFELD,11., PASTEWKA,U., STENGL,p., & ROEDER,g. (1981) On the gas chromatographicaL determination of the pyrrolizidine alkaloids of some Senecio species. Planta Med., 41: L24-L28.~~

~~I/'IIEDENFEID,Il., RODER,8., & ANDERS,E. (1985) Pyrro Lizidine alkaloids from seeds of Crotalaria scassellatii. Phyto- chemistry, 24: 376-378.~~

~~WILLIAMS, A.O., EDINGTON, c.M., & OBAKPONOVI,IE,P.C. (1967) llepatocellular carcinoma in infancy and childhood in Ibadan, western Nigeria. Br. {. Cancer, 21(3): 474-482,~~

~~I,'IILM0T, F.C. & ROBERTSON,c.I'1. (l-920) Senecio disease or cirrhosis of the liver due to Senecio loG-n-ing. Lancet, 23 October: 848-849.~~

~~I'IITTIG, M. & STEPHEN,C. (1964) uodification of microsomaL lipid peroxidation and drug metaboLism by cytopLasmic copper. Res. Coarnun. Chem. Pathol. Pharmacol., 44: 477-493.~~

~~I'IONG, R.Y. & ROITMAN, J.N. (1984) Structure and absolute configuration of (+)-doronine-benzene (1:1). Acta Crystallogr., Sect. C: Cryst. Struc. Comnun., C4O: 163-f66-.~~

~~WW, H. ( 1939) [A cluster of endophlebilis hepatica obliterans in the age group of suckling infants. I ftin. (in Wochenschr., !9. 1527-1531 German). -274-~~

YAMANAKA,H., NAGAO,M., SUGIMUM, T., FURI'YA' T.' SIIIRAI, A., & MATSUSIIIMA, T. (1979) Mutagenicity of Pyrrolizidine alkaloids in the Saknonella/marnmal.ian microsome test. Itutat. Res., 68: 2LL-2L6.

~~YULDASHEVA, L.N. & SULTAIIOVA, R.G. (1983) foxidative reactions in rat Liver tissue under conditions of chronic (in heliotrine hepatitis.l Vopr. Med. Khin., 2, 81-85 Russian).~~

YUNUSOV,S.YU. & PLEKHANOVA,N.V. (1959) [The alkaloids of Trichodesma incanum. The structure of incanine and (in , 2. 677-684 Russian)' ZHELTOVA, L.I. (1952) [ttre clinical course of toxic hepatitis.l In: MiLenkov, s.l{. & Kizhaikin, Y., ed. ICollection of scientific papers on Tgxic ltepqtitig witE n nt university of Central Asia, pp. 76-90 (ia Russian). -275-

~~d d rl OrO d fls o .o os N O o $ls h {{o{ E~~

$ o d ; dE€ .$s G'^ d .^ : o s N 6F E fis € o E I E sS t E$ ^ ^E h d v Ev g Nd Fl It : 'Afi 'Efr i; 'H3*Hssaa S' :s^^ e $ fl: : fl t 3 t fr sl rX SEHHH :a:d H d fl g ! .d EEE 13 EI$$ $fi$$ $EIE E $ $ fi H x T H 6 E a a Fl UI d H. q 4 o co FI H E E. o d ,\ d a o 6 i # fi , ef;t n inii . i o r o E t N ,, l$$g-ffi$E:€t$'E;cgA'i I $l E ! H€ rra* , rrnr T qH g H t' FE g F s fi I E$Fr$:sFfi : fi$iE H .fr'H '{ Hli fr!frrefr fi E[eH arcHH B r 7 d E ggH;gggH: H N H3 E 3: fi $H:;f; I g ! H orE sl X

~~JA n H (a n,aHfEfrn$n!$g|$$$$$$$$Fg R H -276-~~

~~F{ Fl E o\ dNoidNoo E{ oo { r oro rg OOOOd 60 o < 44(){ o ilSo El <~~

> .E€6 cE,tg Ea* l @ ^td d ^ C A Oi6 N N rd ^ rO F|o lO N 1o rU O dO F F \O N ts \OF F F IO .N 0 -co or or 6^@ or oto o or o! 6 H ^ O d d a on6 d ^dd^d d d aX E s Od N rO dtod $vv{.v 6l! a : E F: .d-eI fi A 5 :3; E v-g'uvi'.c8,!sBs;$; 6^;^ Hg 3 F6-3-5t S (n ?.8 t a; s 9Ei39u E U, j i *,d6 E .;utjt o9tga.c g$g$f;EEf;EEi$B$E E$E:$3E$fi'lEIH EE il fr HE E'iAs $ ? - fifi!.!,$Hi.fi,fl fi I * I r,g.* $fi n E TIe f, ildgfr f;i { Hf; fiBEf; g,g g $ fi firu f,fiEE!$ $,frEfiE EcnFFgt#s f;E s F !i3 f,FgggE f;i55sg g E E *fiHgs

~~xfr x $EEEE*fi d 9l F{ J p gFtfi$$$rY r+ H -277 -~~

~~iO Or or or o36 dr FO F'O gn ooo m co or}ro @o @co olN E-d aa I (E o s Esss El*d Al'4 ss~~

fi$ g;lgaH 5Fae$ g eo s^Fifi s sf. ;FHH;; 55F Sgsl* 6?6?E5E $st;E E it? €1?$3siHi$;E :5:;le E ;dE 5ii;i 6'IdI;; n*ngdg;;;;8$E E H3 t g fii3i, ' $' $E$ $ g ggsf,$ggi$EggIEifi!fir$n$gEEgf;fifi$iirrrE{

~~o ql tf;*f*$F,E'dg r, gfiAfiE t, * fi iEiH o t{ tl qq 'i€eE dd pn$f 'd -d UIF xx E $g$EFEF '{4 E& -278-~~

~~!-{~~

a a$ :5- od h-d -dt o tOa -h6,{' =.o o - t fr, gS$..ft'?--^ Bh EflF-5F ^ i.r e 3s$qi$SFF=o3a- f, I :::833ig gE+t- g xigg€ 'A F $ -r-r-i . .r, =E j .3 g U ::: >,?C: Cd .6_2; , ? i: 6 a s :33 $ *iE ;i l;f;I 1 i t iit : a E E lll ggg$ gg;gEt i H H $i g E ; ; sg $g ; ggg,, ,sflfi $i*8,f;$ n Er fi$IB{$iHfifi 3gF.Hdtgfi€i$f fi,€HrHE$g E fiE flE fiE E c#Eiec ! si*s {fil $gi!i**H idEI EEf; i $glf; $I i#i. gi f;q

~~o q +Jg o d U' .g fsr e f; Hgfr frs H fi $ gE s fiE ! r Hd.H B f; €f I 6d E E EE$frfr g t -279-~~

~~ri+ oro N N o o'o o'lo NN os r r o\o 5 b,t oo * { B- En d d o ofrf;l < o I SN &l' H .< o~~

~~ro r o) i6 TH tEE et gg R b B f :.s B a S$ E 3E e $ s s$ 3 H 35;;==i5s$i -'Eh IF'91o- qE'E i 1 8" ri ??r$ssf;E:i$iis*EE S S i I o ' l;; d i; qqssHEEtHgg$-H"E'f HE 3i;;s*6a*&6s s E E f;~~

~~-!P li €E s\ IF * eE EEs €e g g ro j .Hed :i * E g E E;:$oHHEleEfir.fiEEa*EEE€tf;HgItcci$g $ fl frlHfiEggF:glfrn:[EE€ E *s $e gsES:gfifi$$:gHHg3 fi $ g~~

~~'[e4,8 'g 'fi .fi s,4"H f;f,f; As*H EgEEiEABfrE!Ef,$F*g -280-~~

~~o (, ON d{. .ro t: SR oe ari 8.o 8{ -9j *RRB N ,( < El d El{ d~~

o1 E€j ! $5 s ;; t sFe ls _1BH$g_ qq q Eiru.i;s$ i i ip ltiq i fr;;: ig$EtEslI:t:;: IsiEsEfilnH;-gT$ifi}**E $$ fi I$ $Fi rss#ii$g nss$Ef;sflIfifr$$$ ##E s ; *E ,*E .$ n fu$s,*f*iu{u iuu$eti i u IEE!€Huj,jEH,j,f,fH{EffEE $;$Friii$gf,fi f#igF f,Hfl ggggf,HgI gEgas$?HfiN'E

~~~~o .E sr o I +) o E 6 p o -e lr ?1 8 r -28L-~~~~

~~~~\o\o () oo m+$'d-nr..9q br:ocr-oroo-d' o oo. R'*R 4EflHHEAE dEEESn$hSSSd F EEt d$d '{'<4~~~~

ss Eh :: 3 .EF flg $6rB* i B s$s Hg ; eH*eu ss$freeiue-s'E HE $ n H F E EEgSiiEiiiEii;iE$fs Esri, g i-::: gellffiligE;Ef;lggigir Egg$E$;fififi Ti fi dI ? Hi q ? ${? g, s , fiq ; .s i f :F g g ss.EF*gHfi*Esl.E *fiH.r9;il-c$ns:qEsg$Cfi$-{f; $$ee i,isa{ E gl*fis$$gEf,figg$€€$Ff;$E fi{f;ii e HruE*.aIF;f;{i$fifiEEEEH{lqH[E F **,S'!;;;& &'Eca*aE;aF;;#=#$tiE ;;f A .f;f;FE; g sl I =g v,6 Ii fliF I HN

~~~~E .d66eFf, ?t c.1"1 E EP9 -282-~~~~

~~~~,{tgi52!^"'g q .^ o! ot o\N rN oi to n, sdrnto o E\oqqoq d I :'S .i .l .iA a:N . @ 6 o6-6i o EE-E- t4EdEh~~~~

pe5 xo gltdii. 5E E T ,E T tg e $ 9:EfiH YzEs .\ rd drrH-- F . *H *H 16 /1 + "* ", _ _83hfi .P-* s€.t., $j.{.iE9 d e 3-9-g x$9; s flH I $ $ ishth id3'd-i-.:,o"; e 61 I."-E* 1t'{t d 'd 1 lijig ljij : I l $fi'jsr€EEEf;f;$f; Eli$fii Ei$fi gEr E i i

~~~~3fi ii$ .flgH fr E Ei E*q- : d tftufir,E*gf; $sgsfir figfifi{ is E $i, ffig!$FE$lE'ss nrfsrf mfifrgnu$fi gtnr~~~~

~~~~I qcoo o 'Eso d o c o ;'e 'a ,t{ u€ flg 'cF,g o !! E oo 5E dd a lq rr. de EEu(, -283-~~~~

~~~~Croru.toto o{@lo rtd o r ea 9N d@os$oN ord -r*oo< o{.' N oo~~~~

$ s* :. ! $ qg -f;igE.t,..t--.{ Hu x$a:xs Hi $ds€ins*tt H$*f,HiEfi€ :I ::Hi E :.9=fr.-i:,gt::;fHgjtl$xi ;;Es€E s$sEn*i $;ffil$sffiEl ii$$iH*E;$ fg*u i n s {frnqfrfrg$gd',s i .E**u d s$sf;; EE Ei,* ;E$u {ftE,i. iglfr$';i;Erfi g !E,sa;BpEH$ ElEg,E$a igglrFEu gfi$59$€8istfingtB EE Uci Ud A flii Ad oc;HS g'fiAd Eiiiiiii f

~~~~oHo.fi o .F 'E .F8 drue.4a g d.d6!8 ". u E E HE d6cil i o .i u as fl E -284-~~~~

~~~~rr90() c,sir..6idPg$'3i$oR3SSdSh hh hE (on o ilil E d hd*d E{ d F tiilti88o~~~~

frs g 6 H:hh^ il qF.rtllL : $I€ fi 1 '3 & , $c' F r -s IS i EE^.ns H HE Fd ' E'- e n;-:€$ EHi *noo H i.E€fi ii h 6lEErsH ;$:nnFxr!I il $i fi{f iis! ; tEilF; Eilg$g$nE$$$tfigE a$$g s;HsHIg$ E*f, $i ,, qi ;s Ee$ d f !H E;;r;fi;niH EEfi ri $ddE :c t,, !f;dEs2 ff EE $E 9.. I {9.,EFE i$F :figr .E:Frls sE irEe fi; fiH $g E; $tnE:d'fisfis3is$i$frefii#;*Hl.

~~~~E o E .AE tc 9X o UI ..{ .'E $ e .FT .d R.r'Y gEH 'd'nE -doEEF { E d.XU' r{lJl T o 9g{J ss 3 d HHH -285-~~~~

~~~~sE h$ E q Eae s5p s-s:q ^s:q o oo .,Hood Eiud4iliis"Hd.,rcr,oi" &l-{ E EEo< "HH";dH~~~~

~~'j i I -* 1::_ xq.u gq[-_-[ gE E g i t t{ i$Eifiqll:llE s*gi;*-$r$e;qfr E€ €€*; tBEffirIEHfiuatgfigg$fi[$eE€iH$,$ EEEng rfl^ fif; s e$€ qi{rtt i * I;! tattsgfigslsislsitl$f,s liiss geq1lg~~

~~~~o T E n D~~~~

~~~~H -286-~~~~

~~~~co@ 6N i$ ! NE suaneinn or or oro dd oc.,~~~~

Il E I sii d FI s h *$ 6 H* *$ E *N o E |n rl6 o oo! -' UE En -- i* dd H A' SE.g gfi tJ L' L' o * I -^ I 6 @ (, roco o qqEtE I o! > --to\o! > $$,i;:$g EI il .o .6 nii E d Hd € |!€ lUtdra i d l, tt I'x .a >-E xc56 n$E .CO !ttE.{ $frElE$;lEE n$gg$ $ Ig CI H$ Hgff

~~~~1E dE $;slrgfigi $stg* a ar* gssls~~~~

~~~~I t{ o H s.H 999 E 6u -o.|g|! $HE rfDb -287 -~~~~

~~~~B U noF6 or6 {!t \o oooN ilNrq NroNoii rornui oo60 r\ mo fl lqtrttrtro ilAE{EEEhil<.rESS L?"~~~~

~~~~o o r} o ll a n.! q $. nq=* $$$, s q $ $;fli=ixn$$F$$. iii 5E $s $$E t # ; rF;;:;;:r'rEiiif ii Ii iii fl EI *E$clg$g€AftgfsFggE$ gg$ fr~~~~

~~~~5 s ,$$$g , tH,:I: .E EtgErg f i $slsfiFEftEa$Es$figfiggi$ g~~~~

~~~~ClB e.$fi e $ dP.4.d ,: t\ -ltrt{l{t{ xeq ':: 'EeB,6teE.ES.q 3H .S Dil E'r.r.i -288-~~~~

fi fifi"fi3"8"frfr4{c,: BHfi$" $eH$"$"$nfiilfifi. $$$g i i is i r s r Ii ; g*tu3 tt{ fl3$u6$ag g.{if .$$ dE ff,'-f,$xjlgu.EEfrfrifi*EE*sgl $: $. ; h n*or.rEx,,8 lui i i *ilEti;:tFn nEss#fE$#n$$$f;$s I rixllEIggES;l;gfi3hi f,EE $H$g#fifrri;t; ;;gFIErll,.. $fli;u.i:d€EHfl, n lg{ gsisigsggfi$ss$galigfig$ gigi lrggr

~~~~E o.I 'c COH 'd .5 ,s fr fi dH H E EE 5 s gs -289-~~~~

~~~~o roFaoo|oF{ o F N d OIN O d NNNONNdN c' 6 N N N dN N E ooEE-En-fi8 E oo o dd dccr o I S dS{E~~~~

~~~~E-$6 eE I s E , $ Hi E BE$ ill - g sl*$,s $E$ - $ -x;$ i :: n'f;HH$x $ a$ E se $*EI $ t B1SE T $fi$$::ri; f ii EtT ErEi i T iTf;g gf,f,EEsg;$rH $!i Ef;$ iEi € i f;II$E~~~~

~~~~EX flt c r Itr*isH EE. flgHl;.lfi{n fi flfi $.n.n16 ru $$EaE{aifi r$.d 3Et: H: 1E=* rE gg do lB3gln$H$$Fg $fi nFf,Hf;HfiEgfr~~~~

~~~~.H .H.fr. .f; s$$$E$gnfi -290-~~~~

~~~~('l q aogqiocpg'ot .9lojo n'oo * 6 N +soo'.r .oo iloHttBH5frfi$ililil5ilig f; "$fi il ii fifitts FH~~~~

d $ lp ; $;.[ $: ifr g IE$: p' 8",$$:s g$ $ sisr'H$ h. *efiigE s*lg.$# $$H;g$- q $i:6 $ ffiI'H i $ ulaH;ng''lltnlii,[E i II' lI Eggl [ifi ,3' # flHH i E j srfru ! f;{E;#r$;.,i,Eo3 d$ o{f;$ d F HEH tl: f;{fif;$* $llgf, rr,lggfi,*l$ fil $ Ffi $gg$i$f, $fi

~~~~E'i g f;g .H.gf;H r ns fi $$ E f; fi && & & -29L-~~~~

~~~~!lFl a 9Qodoil{ ro o ooF{N@F oo$dNoElQo cl@FF Qr: N hh{r)$no- * - @ iri 66inan06a\ - !oot.o-{d.q|4|4A' - o!cqo!{- - d d o il d d d El El{ d d d d d d d., d { d' d fi d d El,< E El~~~~

~~~~HH E Eg € *EE d nE-.: H Eg6 P E:it H ;EE E-. 8Edeg i i$: e E-'.,-*E*xi;5 EEid$g* $ -;i;q${i{ieffi-l-fi; $ql igfigali E;~~~~

E ri i. g 3 tg it i ilie .E ri Es $.:t i , xs d dE;E ri,; dfi fig,i H $d*u'E* ; ig ess'::ie Bb i, o gFfrBes fi$ f,H€EHfig'd6 6,EE I$ ?u E$HIEI€EEb.ie$ddtstsH .Fs?# .sA$fr A,e EaE ts dddsiqid d d ddddddddd dd dd d qide'i q; -292-

~~~~{oo6t tt o @o o Fo ao@ ao @ o' torog eo'o 66N+ rl N \0r |o FN dFl F{ r{ N oaN Nnq dd-ild 8l{il< S^o~~~~

Fd \ooE E$ ;1 q;R*$ E* aa.of g $ g*nE 3ni ;- ; eE:tBB B E :^ s$nnsg :i;€ E:;inain:Efii 3i i I EH€irl*i fi$EE$fiEfi gtgfitliEE EE EEg isEtEn*

~~~~!eH qqd: ii fi;$t$ BEiiaf,fufuE iasigstti,tflttuttsrsfi~~~~

~~~~g x~~~~

~~~~r{ n gr{ ts .xJ( & x -293- f;$$$f;EFH$fi$"$"Hodu Fno,( o HHfiH"~~~~

$$$ * E ;;I E € $$$ i$ i' FF d d d *t .< n hh .fr..E.f! . . o- - I Egi*,* *fis g ^8^tpplg E$EnE6.tlE :$: Ei :: fi. s B $ $EEfi,r lg;gE$E$IIfiEIffiEE fi;IIIf;$fifig $t= E$ li -E i E ss t;, n$ ! gH g€g *: Hs$iHn !9,eug ;er$rnuFfis I Ie**$isfiEsc fig;$gEg$$$$fi $u#u

~~~~q tr fiff;fr '4,fl $ $$$$ -294-~~~~

r}{F rdr. cr ol,o'.o. o oi,< o ,rio oi.r ooor'.{ r'o 6 o ro +'rn o',-{ a'-r - o60 aN- {${ - ${(l(l rJl|nloro${rli0roro N{FF{rl{trlF nsE El cia (J ElEl 4 d B - d c d - d * d d - d - d <.ro il d d - d - d * d

~~~~'i' ! $$ ri; $ !g$; .{f;$ ,BnEE$f *$ i ii ${.- i$g ;*i'Ef,*uE**;H,:g$ E{H-}$:HH$gg H*Esr lf;EE$llgffiagffcifi[[nl$fl$$llE3[ I$E$l ad~~~~

~~~~E: !cE gg € Erc f; .g '3 . ".;11 iii , E en, 1 g g-Efrl9nfl;fl3ut; dBEd t i $: *,*i,:9f; fi i $Eflf,Hfi € $ $EEfii g? fi figtl#$ttg$ qi Ec;dEH I'EEdE dc'isi dsi ,/i d sidv;dsi d d s; -295-~~~~

~~~~o'o .r o @ r- .io ds u o n ao ooilo ,o c.t'co n r. rraoo 3 .oO\O|coN l')6 d dNN { nooNst d*dd<~~~~

~~~~! I tgs fr-.,)8 H X ,c. H :5 HE . s$$ H 9s . e $; sp',HE; g ,1*E.3 s.i t** ; g fiT;$i gi gi- Hdes 8r -H6+in R:qi i ; tgf,trf;iEq; 6n$lFiEl;uEd :rtlseiE frBHfi$E$EHEH$$flfEfiEg$fi$E!EiIifgHf;E~~~~

gE !s llf, **ld.€ -t.** EI[tH, n t* ,cio Sol.8.9 fg€gEfiEt ,E ,5 fi#i[flg -BE friE.t;[g;,fl;i$gggfiofr sl fi8 EfiHgEgg.EfrlEHFEEflAI frg$jffif;HEEE{IHBHgf,fiEHflE.flf;HsEffiHEgdjde-^-eeiHE E vi ddsiddd!i uis;ddduisidd,idddsid dsidddd s; -296-

~~~~- f9 lo--t-logOF s{Ne {oo,ri.q\oFoor A |4\9|r)€rolo FF.N GartFGFiai(ir {fi~~~~

~~~~E =Hf$ .;g; g $$* t"H ;g-i$tg tigf; .F. l$fi$,*a$lgiffii$f,fl;tuIiiElE~~~~

~~~~E 4 .t f, E; * € - E-*, ll-g l,fifi, . *fsE H ; #r${fle##If,$B{t#$$rfr$lnl dsidd dqi dqi qidddddd '; ';6H S g g $$$$$$tu -297-~~~~

~~~~|Jl rO F ts 6 O Fl O O $ Ot O N F{ CO tO ol drOF O 39 o N N { { $ ro g,) rn |'} |') \o (o ro { \o 3 r)ro |o € 6 d d d d d',d d d d dd H d d d d dd d d tl~~~~

ro|! \t gi{ 6 .qEBol $$d d EE*,?$ -$nf,, @d HH,H*HHl,{iI * o. j qgE* .'dE (6 ;; $;il;33: ifl F$d tts t lJoa Crrt l,Eb fifi9 HEftffiI$; te Iffa l*[gffagfl$ d'd Hd tl j ? ? d.,: 5.s e5..e a f fi. ds I-ffHff [-fig* fi;sg! **E * H ; $$*o fifii!fffifiiff $sEE$#$!fl i Ft 3E g$ s sF E EE € f; $s H $ $$HE fl f, f,o fi -298- g.s$n;g.nRfi "$Hf;'E$npis$HHf; fi.fi "$sH"fi il*g.nfi "

g$ * g h- *63 $$*^g E. .$i- *$. i $ ;litllafulig;gg' g'fl{ffiifu*fti*,'i E'i $'s E $ .,s?ii I s, .l H: g E;sir' jf* I !# o,'i,g fi#;il-fi fl3, g*;o*fi 'fi,d, $n frE$gHfEfgEEf Eag$EEfifi$$ ds;dd d €Es d uiuids;d ngEA .; ,; ddddddd stdd dd ui -299-

~~~~F{ o'g to,o'r- 69 m,o r.t'to F F ao o co ro ro'o a o r.'r- < {.roF m ooOoOm{{Ol@N$SSOr)IOto\OiOrO!t rOFF o < &l< s ^ fl fl . fl s s d - d ff Hds g d. 8l$ $ * d' S'El ,{{ (,~~~~

~~~~h' *$.*E .i : g $ a ." i,Hh:*i.Iu$;ge*s' E : 8 ;E dH:c$s?iu* :{i rlEE$[$EEii;€.$EAfiftEif$,I!fr;lsaEn if,snu[:g ifir~~~~

~~~~..{..3 t E' 'Hl,$I q: d r gt rF$tE; ! in.d;t nE* itt$ti -tHHE?H"Elr~~~~

~~~~H$sannnrfi Bnfifrf;f, flfia fi$ {fitb..~~~~

~~~~.dE gE E d u) -300-~~~~

~~~~U) Hilf;""f;H$4<~~~~

~~~~Hq ; $_i ; - :q.$ t l$s€E* *.:::;$$iffiiiiili$u$*;ifl Efi $€ :: iigg$g$$#n$;$$EfrEEr-El#gng$i g $$ Hf,,fl F $q i$gq;", {,fi- * ;; Hij~~~~

~~~~;{ - * $Ees,Etgfil,;"i dH o H.H.H'j ' f,i,lu$sEi*s3g i $ f; fi $$Ftg$$$$gfrff*t$g;fflffflfi$$$ Hfr~~~~

~~~~E e.H .g ,g d ,g Ena$$ -301-~~~~

~~~~o 10 o d CD cOdN r Fm Oo'o - ,4do n oo{ d NoOrONd OFI rrO(o r| f\F o NNd d rNN O 06o il {dddEhE dd HEI^ O'~~~~

b., 8,. d n $ 8 ; 8F - iHH$s 1 E i: s. S€F{r{.9^8;8hghH 'd EE,sEi.'xEg$d sI;.E.h $n,E g lxiis : g.$g; $x i iTEl*;gEIilili $r;fi iEf;T;i iH *.EXer* $$$xgE$Esss $EHf;$$Eif;Fnfigfi If; HEEEi; H d, aE 6F , .i E,a gsEig$ra igfi$Blsg tsg $giI$ f*llgs

~~~~ol 9E .EE lt fr€ e c.d gEg. .d|!.9 'i HCg, s, 5Aa (,)(n oaAo flF$ -302-~~~~

~~~~o6dN rC < I dmdts F{ 300d: hdddJ ooolo u t a fi $fi$fiooo o a'a t~~~~

I 3 33 d @ o 06 o Js dd 9l - .o8-S,* ..8 E ;Esi'! o' c EE8* E 888 .S... r d rtrrdd do\ots D o t r-..R.9-Ef 8.583 0, F{ fl$ F{.OdFl fl s EnB'*3I?i 3E* E fld rd i h;i;.Etl.E s l3t i do' tEaE$ t E€ IgHEE Iifi3fi:gglgggg; .{! H Fd UI6 gggEfififi EEEfrEg EI fi ffifi E frE g T H fififrEflflf; fl fffl$ fr d#vN .{ .d 6 NFI N .d{aE t{ .d ? Ffi v RN rJ g q o so 5 iF $ri !i *qrffi t ud. g8 EE I (l_ 4fi* efifli.f, d A :$ N5 ta s8 *$g fi6n8fffi.8 sH (!F u IJ .".8' Eq a 6 b.9 s$ nb6 .isfiE fi .gt6 gfisssgt$fig$'$g$$loo' gE R€fi D ils.E .d.d J f6E Etrtr

~~~~ggPilflil ,9flfl p 9l p g&& .tJ s'3 s q E FE H a l;,: g3H q g p!)D UI x6 -303-~~~~

~~~~I o{ o~~~~

~~~~6a oo o 6l d! o o o Its o 6o ts r@ @ s! o o @ di d @dOa N d .s$ { NdEO o E E EE O d<(),A~~~~

~~~~o oo o o !.idq 6 otsoo ! o o oo!o o ts03d 6 3 o o~~~~

~~~~F ts~~~~

HX HO XO o za d 3! o o tP d I C t6O .d rl 9.J E xtr o o o .r c d x6 .i.i o .AE E.aCCC e .r hxd !gE a o Et 3E'd'E: 3 H t9A g3'E o ts ! o 6 6 6 r 6 uE d s r? o o o o o k ! ! o .r c \d o o.i ! !! sk.d.i.r Q c 6_d.i tloE o d o o { aoa h o | ;d k tr ho rE .Er6qa a o F!! .rdo

~~~~o o o~~~~

~~~~a a~~~~

! ts ql o al ol E'l N 'tl 6 .a el 3 AO | tsl ol ( @F tsl st ol o tJl ol .dl dl q o ts -'l nl l,l sl dl ts I ol EI o Fl 6 ol q l!l ol .dl ti I ql ;l t o 'l ol F tal -I o tol Bl a .!l -. ! o t!l ';l ';t a € ol x o EI l=' ol o ol o FI 9? F .o ol ol o q d li^ rl !t ts !l I & c 6l d I 13 al@< t o -304-

~~~~\t^ 6f oo~~~~

~~~~o o ot~~~~

~~~~oh o3 h! 6t oc< rl'l @ rA rA rq O<~~~~

~~~~! ts o o t L o~~~~

~~~~d o OF clol06 d.A E o ok oook o o 3 3c €€€s o~~~~

h I !o oo E.i 6o oo .dk oo o ^ g; .EE'Eg.:3 b .EEgb g.I c o a oE q o .dS- H .t.d39 c Ao .d o.d d O € g coE! o ao 06 Ooo o Oo ts Eo o o u o o o xo o Q ! o L a o o d o ! d ctr EE oo vt .d.do.rcEo .i.rQ XdXEE q! >.oE .d C CoO 'q tii.d./ du.d .do .JuoQ E co.n ! E c H hi E€ .il o r o E o E E E o€ E xE 9S.i.iE g€.r >uhso s6 h6o ds d.dco.dil.i tsaee.i FoE !0!oE s.d !.io .A6 od.-EdiE !!AtdE Ha5 ooOrts 60 osA CA r.iCi{Xh.d oo.do.i oo./ oGooQ o o .do o.Q.io!!o r.lo"dE! ! o! c t oo! 5H o k o ao oq6pooo E>,o.r0 EX9 l-.rXC oO I Oh ,>r XESCqos 'dd 'dd Q O O O N6!d.i Ov FvH od H S o.t d d!

~~~~o E~~~~

~~~~@ .o EO €.i ; q '4r j .-l o el €k .:l . .; o !o Nl! @ ;tEtrr d it; IE F cl 6 *l H !t: 6 iltl H olc o~~~~

~~~~:^t': = @< o o :t; i k il fl$ o -305-~~~~

~~~~\' 6l o $ oN 6 o O\+ -? o I .o 0 6 O d Q @ N @iO \+ A ts S o o -+ o o o o$\+ $ s \+o o F E F E FO TA fAE€ 4 'O € A~~~~

~~~~id d6 .i .d ! FK s oo sd diHd ! ooso O d d -6 - A d d .d .du.i g .d ! 6 ooko o h ts oF o ! o E O O Oq O O O A 33SB B d d k6 h s H~~~~

o o6 .d! o o.i 6 g .:x d o o i t.; o t Q u a i I p .iOQ.ilOOOOO.dO o orc qo io o! Fo.iE tro kq o oc cs.r .aF o dd da .rci;{ .ido o.i F.A 'i ts O P.r tr F.d .i x O.dO X A.i E il P .F! !ut tu .J u! Ji tutn au'A qf i d a a a o d I o o O O o Od O @ d F .d O .rG -ao od 'a a 5 q@ ocxs oda Eo !.i C 6.i .d.r p O 6 .il d .e.i! C .r.i ! .r.d c! cEi rt o trd us itq6 iEo kd E O >XO O o E >h S I O oO I O o.i .r O ad oot Eo s oo iF SodF !oo >g

~~~~E I~~~~

€. F ! EI ,r El 6. JI EI ul ,l ol P !l @ ot !l 3 it .d ol ol .it al sl < dl cl il !t rl Ft ql il ol .rt ol o 'il El .dl .il tsl tt ol 4l ul ot 0l .:t ,3t El o ol . trl v El :l tlr sl s ilt .:t .:td @l€ rl >16 J el =l El.: El "lEI EI"l EIFI ts 5 r:l ol ol al al al il* o L:l !:l ot x al o olo ol 11 91 9lo H it dl it dl :to 6 ri ::i @t. @l o at cl Old Ol 31 3l 8lj Et tt dtr ! .n hi >i xla hl al3 ol a 51 5l ,3ir { o -306-

~~~~o o~~~~

~~~~@ a^6 H^ il6 n trN @a d@ s60- soil od J d N os a F € Od { $60Ev6\+46r^6 6 6Ev €EF < O F O OE E~~~~

~~~~ddi q dd r!60 6 6 o ao o 6 d .d tfl.dd ti .d .4 .d.il d .r .i ts F !O ! L ! !k Q ts h o o o! o c q qlo E al o o doB ( 6 q E6 I d 6~~~~

F E 'n; o F o oO o C o o oo / q oq ta ! 'it .il '!.5C.i O.il.i ! : .5i.: g qo ij E6>;€too.: o I I gb- dodidiidccc ep !ee s 'il'iEHp H':. i 'i3'rH5e.:.1e'HsiiiH€fiH$Egil:i'i.fr .!€ i'E! i t 355558tE.5 .E'Eqiq 3 3 i'6.t 'dH t3 tt.?uB-P*sgggBBEB E': :3.8'i : :9'd :.:'i';g"IjHr':::=gg9'j u3 _EH3"r .g -3d.r i -gi -p.1Ei3iH-93.9"e3333

~~~~.d M~~~~

E 3 ol o u!l HIEI FI o 9t Ht.ll El . 6 al ?l'r E N ^El,r i Z GI 8lil v'il.dl o tsl o ! @l ol Ol :i d s tr dtd €l !l ql il o ulol o :l d o Flol Sl 51.:l3 F r' ol:\l dl q .ilol dldl tl o cl I >l d ot 3el5l -l : tsl stElI o ql ; l.9l X ':l 'at .ildl c':lrl : d tot ! d utEt ol il ol:El o o sl ztv E o 30.1" = 5 -307-

~~~~d o 6~~~~

~~~~o o s I 16-3^ oi I a € o€ - >@ gNql + I @ 6 oi N o ?E 66 Ev a@o a I o e nr F N (5- PF €F € F E E rA A N G O~~~~

~~~~d dlit .r! 9q q o o o o ( o o 6 I J .d.d .t .d !O o p oo 9!l o o o h k k do lHH € f t € 3$ S S~~~~

~~~~I t x~~~~

! H , Eo q .: 'ilg 'tg I 'i u.d .i3 t:r o ! kc 6 'de .. H .9 gi g !,69 o o .H q$. t eif H'*,,s f.H g q.H*'1 .H'[",* i !si3 isi !riiiBii Hi:E!iiI"E ", IIii .j : .; p'.s.ji F, :l.i F.Fa€i r .: i.!.E.5d.i i EF..a .s 3j; H I-{i 3j; f !i,! g .EH:f f!i !.t H!

o o o € a .dt cl +l =l EI el Elfl !l sl 5l !l .dJ ht;t 5t el 6l .ilil .il| al o olel ol 6l €lAI I s :1,; :IE F .Elxl =l 5l .il5l ol o F..lt a slElsl sl ;t . .9151 .91 !l ol! .el .il ./l o ! c 6l:l ;l rl ol =rv =l -l #l'd c -308-

~~~~@ o~~~~

~~~~.; ad @ o! do o- r!a .tr9 ot 6 F € $ o 5 dNo d i { F$ r F r F € g €O@ a 6 @ FE 6 E O n Fa < €Ee o o ca~~~~

~~~~o oq HO o tt d i a a i d s6 o o 6 6 O - 6 6 .d.d .il ! .i .i q .d !!i € .i .i k ts .ts o ts F 6 ! ooo o H F o o o o o o o o o o.c o o o a d o o o o r s ! kB o d 6~~~~

~~~~o oo ss d.d ts!~~~~

~~~~.io~~~~

~~~~tsq >! ooo~~~~

tsF! ooo o o o ii/ o E t o E o ooo c .r o o I .a t g.n g qg444g.t g g o tr d tro o o ts .d d o ! o g 'r ! o.d o d.r xxx.d ! .d Ed ! ts ts.rFs E d Fsdu C5!diits6 h ! .r O s u ! ts.d o .d O u O.{ d .i O S O O O u O u d qO O o of Eo d O OOqd troE@@@oO O C o.d .i .i .i o.i.r o o 'i.i o.i .d.d./ d d d.i.r .d .i ts o dd Ao ts o i o HE AiE 6 d dd o d E t( o O OOJO 5 X oSto Joo | | | O d O o oi S E EE@a O i !dOo otQNNF,Ca 5 O

i o.il ! Fl ol oo o v kl riq dr ErAl 3 d 5l I o .6 ElEl- I t .i .dt 3t Et o .& ol.rt pl tsl o o .Jl ol dlll .dl. 'dlol6l E dl !l 9l ?l:E Xl !iltl ;I-l T qI ! 6bl X"let E16l€l 6l d it ot JI st.rt Fl o ol alEl !ldlel rl 2 :l o El =1.11El*lel ol o al 5ldl )l fl EI E .dl .dldl .rl r.rl .it .e 9l Al otil ol d al sl ol ol4l ol>ol sl t o tsl !l llo !l tsl !l !l . ulvul rl ol olp ol ol o ol o .l .rt .ito .it..il .i| d *l dlt ilodl "1 cl al't olo ol o ol o o qt

~~~~o -309-~~~~

~~~~d 6 o i d o. il !Q o 6l o .io g^ oa o a 'd6 0 50 0 .dd o 6NN F i O @o r €a€ @r @ o ldv C 66 a o F€OO EFo €~~~~

~~~~ot oa ! HO 6~~~~

~~~~iddd 6Od66OOC 60 .i .d .d .i€ .d * i .A ( ! ts F tso k O O ! ho o o o oo o & s o OE A d d G So o 3 F 6 o3~~~~

~~~~.=9 9~~~~

h t" 3 o O tsO O 'd tr de c € d .J ho ri .d ! tro O O d .d.r o rd d ':.:co o o ! o tr a P .5 , .5'dfl ,9 b3 a o .r O Ooi HOtr OO! .dO tr! o.rQC qO al o o tr o E o o o tr.i c a o t ! .d o o € tr.d .n E F ts .i C 'diu il,Jq .i.di O o E ! d.i.i O q.i S u.i !xO q!O O!q OS 6i hOPF as I o di 6iJ! .i6d oS.i il.i o x !.d>O 46 A o E.d Coo !co !cE >,€ Ar ooEo ot E .i .J ! .i O O .d.d! d.il.r p o o o ts g.o .d.d o ! d E d c ts! F !t ! o h o @ o o F o d E F q d o E o I o '/ o.i 6 a.d o o h d I o h x o o i q G o r E > oi o o > ov d o F 4 o E E o o

s?l e E .11 H o o s '11 o |f il 1l ol ts k 6l o o ql .dl 6l E sl .dl ol ^!l 4l . !l ol ! ql tlE 9t ol o ! trt ot F >.1 6 9l E tsl clc !l I "rl o ol olts ol 5l iil d o!t t4 I ol F - cl 6l ,cl 3l ,dt .il^ ol @ 6l o I Hlr HI >]| ( sl olc o ol EI dl O il ilo hl o t{ al! o ol! o o ol E E *l =14 qt6 ; ts 'r'tF ': o .dl @-rl rl:, J E o s - 310 -

~~~~@ 6 o d €6 6E oo DO O N6 N o6o{ 6Fd O 6€ n o6 @ oFo6 6O6E ('v F EF @ EF4€F F€89~~~~

~~~~_i ue q oo .d oo u b ," g t j-i k id d dd - - q . .:..i-Eq dq d d s 6 s6o. .i .r.i ,d tu .i .r .d .d.;ilq HH H HilH Hi H H H HH€"q~~~~

~~~~o o~~~~

-E H b e e E'e E g5 'H, 'r 'HOO.dO iE E ;'d " p ", ,*EE e *€ **E 'is '$'!sE 'd'roi€'dd'dE'too**':-1.9g.Eg.Hr'i,t .F!.:S 5 i,:'d.XS'! .:I".E !EEs E Ei $'EsE '!E ,YE5H FEIfH;fFtES; 'rBiiEiii E !F fl'!i': l': :* 3.!rHsiH-s3':s" B Bi BBi B B gB'ii HBHtr;'lhi?$j

.o ; N o oo AJ @ @l a q 5l o Od=) .il .r @o ol El F d € @ dl ( 1t .dt s o o .il G !l @l ! sf d dl c sl ol a Pl ^l cl !l o iO ol 6l J Slsl:l ! kt !t ts tsl o rsl 6l €l .il ol dl 2 ol tr4 olcltrto ol ll a.l OE dt 'dt ol 'I ol d ol !l rl a bl ol O hl ol !l t =t 7t 7 "l o !t ! "lololE ut -l Gl ol dl t CI GI E ol tsl Fl !l 6 ot ot I ol ol ol F CI EI E X o! ot ol o il .i?t Et :t .e dl | .rl ./t o ol ol o >l 4IttElo ol ol o ts o - 311-

~~~~O^ NN @@ o oo € 6 .l s 6 o d .€ os !io o osr 6l o s oo o du o 6 tr^ d^ O F io i o N@ oo o@ >r ts .o .do !o N EO 6 O - o no60 d:H 6r O ts 6o otv gd d ()vdgo o F< .4rA~~~~

~~~~D! u! oo o o o oo o o o ts! F ! !~~~~

~~~~-l dil i d d6 o 6 d o o d 'drr .d.ru! ! .i u il ! d ! .i ( !F o ts ts O O ! O O O O O k oo o o o o o o o E o ,c o o sS ! s ( ts ! o k B h I F d~~~~

~~~~gH o .i o H o~~~~

~~~~q s rt !~~~~

! oo o vtr q oE t ctu o d . €8, E g 3,.:'E3r,3 . oo .iooo qo.ictso.rhtrccod oo od o o cF gEcq 9ta.a oc5d.i.i.lF.i occ Qc cotr ,d.d F./.r.r .f,.r F ad./ Od E€il.i o C,i.d ,i.r .l c.J E€ aEu! !Eof q!OXSOh€F .il!€ !! O.d! ?'d33EEir3.ErE3 t}''d 'EF'EtE AE .HElg'E 3.E O.r.f, q.J.-.i.d.iu oo o od o E d.i .i d .d€H oo qi.cE EEqa !sooosdEo 6EE tE tdF o..l sooo >ooO.ioO I hCt Xh Cxo ot Odh Xq ddOO aQd& >Odrd./roE OAe Oo OSo d./

~~~~oh oq az 3:El - o Et ol ;l| o 'dl ol cl 'ilel Et ^ ,l .:l ! ol ul 3t El o ! Ol .-l o st bt il o dl ul E bl il 0 ut al al xl"l H tl "l~~~~

~~~~e !l !l 51 5l El o >! >-! ;.l El "J a atattJ .i >! >l F,I RI FJ a @al ::l ol @l ol c A o -3L2-~~~~

~~~~6 -N Od o d / od o a od q -^ ii E^ o$ o@ _ E@ !o N o @o o -q 6 oo I Ei o r or - d .il J gv d + ilr diSvi 4 q Qo FFAoe~~~~

~~~~-d E.i ots uo od ll 'SO- q G u .r .r€u E !, s ts tsoo o ts 9 0 0 000 a o 6 00L a d~~~~

oo ! lF o .il'd Et .:o gc O ts 9.d ooo '= o od tr tr *x I c 'r o uou.i!al.i E.d C'A q .iil d ,ti 'Fd'dCCtrou.roXc o o o& o e E i id Odrhrx 9d+.r.i.iOo O tO OO o E >'-a E >E = xh!!ei o o 'E or.? to'? sEs:EeifE:E f;s".8 €HE.: :E.E .? !ooEFE,q9! a uts .no d $ rg E aE 3'{{3}hE3 e E3 X.t A E i.3 3 i.g

ol ol d EI B ;l ilI z ; dl r El ol ol 'dl ft .Ht El E, .91 .Il (l sl EI al trl g'l Fl sl fi'l qt .el Fl .11. ':l .:l 3l :l il il blg =l .sl -l 6l 'il dl *l ti o ft "l . 3l ts F :l El El gl" "l 31 3l o ts ql ol ol^ 6l dl il ol oll olq .'11 H € ul hl xl o €l 3l €1. tl , ;l ;l I 6l fl 6lE o*l S 31 3l o ';l 'il 'il; :l 'd gl € @l pl 3l j o ts Fr FI Hrv

~~~~n'l ioo0 €oo 6HZ 4~~~~

~~~~.dc isq 6l "ut! o !of o O^ o@4 o N € A6 H [email protected] o A a i N N o o N oi i-/id:t~~~~

~~~~ts a~~~~

~~~~ds 600 !!'i!!ii s ooroooo oa oooootg d6 Fts6FF}}~~~~

G ooo .ic.d.d O u Ooo .i O o o OO OU Eo! dca e c c t cc Fq .t c o d,J.r.i .r .i .J o.t o .id .dF o q E.;i Aldt Fts .dE E Oo .dl.:ECO .E S'd S'i 'd.I?oso./oi!c g .5.5 .58 .13 f;3X.: $ . r !i o.;c o .d) .ej og o 6 ! ! ! o o { o o I F o 2 3 t 33 56! :.9:: ; F E HH 3.9t 3 ,q aq .h.- o o! osqq .i h .d OqO O !'d ltd

'nlol :l jl a 7l 9l Ll . d .61 ol';ial.i J ol F .!lJ ol .dl '!d Fl E ot 3l cl a tl^ o d dl a E cl !l Jl E o !l .il .dl rl o E trl vA x ul ol dl d o ol . d ol ol ht 6 dl al dl€ J t ol tsl El ot . El :lolald gl kl! cl . cl .rt ol I ol s lo =lt Jl ol ol El !l ! ! ol! ./to cl al al-l al ol ts d ql ! .rl o llo !l 6l il ol 6 6 c llb ol crl !l el ol El A ts O !l ilo ",,| o Ol! qlt -:l *l .:l 6l El e o dl ilo lE l,- I cl @l al ) o al ql €ld ql! cl 'rl ol ol o ol o >,1 , .JIt .rl - .41 rl Ol .rl a dl - 3l ' :1.: -l= Jl .'l ol Cl t qt u ule il! cli ol 6l ltl ol o 'J cl! olo Jl-- Ol .l sl !l ! a ?la cli cl.- cl bl !l ol o ol o €lr cl! rlel al ol ol A H EI e o a -3t4-

~~~~: 8i o oi @ od6 33 G 6q d v 6 d dd 6 - 6 N.d l.d, d6d SF.- d tt 3 NNlo.9 Oo - - oo c~~~~

~~~~i R ni } a sH 'S { d ;q$ d rr iE H H ;i sJ n.g.EjiE'A F E () a < 0 € F E € F()< o~~~~

~~~~k o o .d as s 'q .. 'a { .o o F! : 9 ur* oo o 3 6 b 8 8P: d6 s F o 6 i iEE~~~~

oo trotr E '5E o s i ;.i .E ,.i .E .. E E e . cd il o Pg 't T g H i 'ie.E_ iF HI .i,q ':: .HHl, .p f ., .E''i", 'EA2? .:l* ci. t.i: sH r6 i' .Hi !E {.i! H;9.i He!l o iso i :irtr c.A !i fi { :i{ g:i *fl$ :$ E5 3s'iE 3AF .r5 si E fl.i =r sJ3. hs

1!g eEt **: it j j il ., ! El j ii d fl il jl nl Er ErEr €l:, € ,3 nt qt it El Il glfl +l il c; o g] s ?El rl ;l El El flfi| $l Fl €l F 3:l sl3:l :l :l El ll:l er n 3lFl ; ;l !l' il ;l ;l rl rl gl Ir :r .:r :l:r ,rx"'ll.: FrFl ;r!l 5l gl Flgj jl 5l 5l -F -315-

~~~~-t 6@ i60 oo €@@ d- ooo ddd -d !rh ooa odF dtd ot oqo trd'k 6l {!u o € o 6^ ooo o dd s os o>,>, o ! H o 'i o -EB { 6u! o .d 6 o o o od o d n d i $~~~~

~~~~I ts o~~~~

~~~~ddiiiHd 6 6 d oo d s s .d .i .d d.d .i ! 'd 'd e ts ! o! ! o ts F d o o o tso o o o o d o o !o d ts s d~~~~

o o g: o .;{.ioo oooi, tc t o o oc-tr.; oii.r o"i - o; oa.r.;.ru o Erfa o trFO o .i oE.r o oE.i€cNs Oc .r.di F .r'iE d c'dD I c.ds.d.d.dd tr.r tro x 'i tr H.d o ./sd .r .rldF9! 5 .ilts o6! + ot{J .i L6o F kSoSOSS tsd'i!O l.! .r9,F 3 9!E I 9!Ei!i: 9i 3 8.i ! 3 e',i C ,;Os .r .l@@@@@o .i@ FOs I tr!|C o d.i.d d i.J.i.i.r.d o r.J 0 o il o o F o o odi o odiddil E oi q e a o o'il o

N N s I 9o 6l tstr rl a6 o .: -l ZF X =l E H j s j ,x !l :l ql EIEl sl sl *;l Et ;'11 il o st .il E tl S .!t zl tsl ! !t ! :l ;l .91sl Fl ;l fl d tr ol ! ol;l oldl o@l tl I o ol e dl ol ql h9l ;l .dl .rl ;1 :l t! !l !-.1 5l .o at c j:l EI jl s tsl d ;l o ;l ;l x r fl a 9l ; al 6l al .d Fl c! at 0pl o 001 !l !l ! ./t .il1il o ol ; :l 3l 3l o 5 - 316 -

~~~~q $ s € 6~~~~

~~~~6d o6 o $9 uo o oj o i! ! ^o o- - or No E o \+$ s$ o $) \+N N Ji ii & dFl d$~~~~

~~~~<4 F< FO~~~~

~~~~i d li I J il i d 6 ss s d 06 6 0 oo 'i E ./ .r .r .i .r r.i .il ! o ts ! ts ! F oF ts o o ! o a o o o Eo o Q o 6 s $ 6 d 3!d 6 "dB 3l~~~~

~~~~O^ !o~~~~

~~~~C.r 'd€ x!~~~~

Ho i vE o o o o'E o gJ o 0 .: 3 .E qo V .i O F E O OO i .d O :Oo ...1!Oo.r iloF 9...1 o.,.1 E EE a e E Odtrd tr.i E C a r C.F C F G tr .i O.i.d h .r .d C \.d.r Oo'r./o h.rq o.dO O.dO Ctr.dd &OO!OCO .d:'E.d !6ldc .cto qca Edo o.ddo GcEgdod @_cior .i! !.i.d qts.: .d O ts .i Ok .itrdq .d.d.i O.i.i,d F.t.Jd oo.ioo X.d0 trcS ECS oOo.il oNPbrod O.rOd sold6 !J,9 9o@ Qo@ o9!o 9loOoooo kOAs u O F u ! 6 C C k ! d ts Lil C O d 3 E o O s o E a u a E 6 O O O O O d g O O O t O O J O!.t ts O O o tr 6 O S O iD O.d ACoAO Aoo !!p 9€p oO!O o.-.q.d.-o.E tsooh

ol .31 sl ol 6l ';lbq .:l dl ^ol ol d.1l ql rl ,]l o d ol .d.nl 3 El !9 o s v ol a old q E ol 5 Jl> olEl 'dl J^ ol 9t !t d o !16l o olfl ol Fr q q ol slzl 'a 'Eltlt.it )l b9l ':l ol ^ol >lc s olql !l fl d !l.dl dl trl ts ! 6l dlo . o 'il ! G '-l il*t9 ! elEl al F !l!l ol dl o ol ol '- ? otot st 6l s vel ot5 ^ tlt | | !l llc' o o o'rl 'dlo- .31€l .:l .el 6 0l o dlo- c 31. 31 3l o D !La'; q o cl olo = o Btf 6t 6t l Algv a of o 6l @l

~~~~o -3L7-~~~~

~~~~oo @@ oo di~~~~

~~~~h.. d dd d6 - dl o s 9!! o No o n 6 600 I €9 l+d d di o 600 ts o d o 6 @€ r o 6 q d@ o $6h6oo6F h 4 A{ d d r a&dt a N i i di o 6 pA Fa €OO e m F E EO~~~~

~~~~99 !E d~~~~

~~~~d d dd ! o d s dd a 66 a .d .il .i.d .j.lH o o ts ts !ts H kH6 5 0 0 00 a ooo B 6 G 60 S da~~~~

o s oooo d -q)EoAAo .A6.dCtil.dF qd ,J / O OO .J d Ald .i tril q g a d E E E i o o I trd E o o .F \ c 'i c > .r Q.r.r ! tr d h 'd x .r E E..t e.i S dnlt 'd 'd .d E E!qh.r o-ao !l a o.ido jl o@ o r ooEHo .d,i! O! .i .iFr@ F Ets h .i .d.d'iOts 339 $''j 3 33'ig 3'dg I 3 33€y! CeD ! d tr dO oo o du ! tr tr Eou! o o o d o o o ! di o 3 o o q o o o C o o6! .iO o Ooqq tr Ots ! o oo.t'rl

~~~~b @ ?ol~~~~

:J EII ; .3t o. C El i @ I I = 5l d F o Hl.Hl,,5 :il .if .a ) ol Ol E O q !l !l € 5 ot ot ot . ! ol sl Jl J ol s ct At ot :l *|:l*kr Hl€l il fl .rl ol ol d ! dt ct !t rt ot o ! 5l 'dl dl ol .il a AIil;|;I el ul Jl tsl EH E !l ! ts oElclcl=l5ld o cl al sl cl sl ! tlttl cl cl cl ol ol o O .-t .rl .it .dt .rt .d :1!l:hl :lst n '-'. el ol ul ol o ! a, ol ct ol ot o 'r =L =l cl trl cl t r-. '_l at ot ol ot o c .-.; al @l a o .:l @l Hlsil!' 3l; o - 318 -

~~~~5 q o:+ N od @ OO 6~~~~

~~~~dd ! 66 q o UU dc OO o .its d !o H$€ o .co ooo s oi € zdd dcv I~~~~

~~~~o o !~~~~

~~~~di a 66 0 d 'i!i d .i I tsk o F o oo ! o do 3 6~~~~

ol c ox I 'i tro o o .ik I o ox i\ c O-c i(. o tr;d 6 0 k ot i o ! 5 tst ! osc 0 so >,O al O ! d 6r ! o. .cE 'd'd.:CEdi.;6 '. .: T L'd .9H 6 o gPBggH c'deE eqe!' l3 .do .:1.n9!qqiac'rE-.io 6bib r.r od ik66d,E

~~~~o6Ei rE[fr:FI EEii!3 n si o i u q i Sg Pg: AA l, [,'it I q.dE o *!t | | o I x | | i i i! d o o -.o6 g ooo~~~~

; o o A. go ! o oa drJ 6 @l a al .dt o kl .d bt ! st o trl .r 3t r .rt B ot o ol ot al 'rlol ts sl o :l5l a .il ! 6l nl ! 'dl!l O rt o ! ol.rl i sl tr ft {l E Otlt c al a !t al F o !l!l ) dl 6 qt o Pldl F ol u cl ldl "l o O l.rl O ol q ol ol o .il !l .d .dl q olul o p ol o ol ol c:l o c'l o ol ol .r cli t o o Olc G ol a ol ol o o 4l> a atl .tl o1l q -319-

~~~~d €N o o / 6~~~~

~~~~d dl r! o 9o o s o o k 6 o k o@ood € QN o 699rr O ON r FF aaaad E) <$ dE o ra< €rq~~~~

~~~~H 6 dda I o ao6 o 666 .t .i q.d o ! ol !! o ts ! 6k aoooo o o qo a os ds 3 o 6 dd~~~~

o o d o.5 do O.iOOOO d tr do c E F c c.i.i'do ! od o >.i oi o oi o o od F o Ei O Eld FdOC EddO 6 CHOO! qC o ./ h o q .d ad .d xc./ .d x.i d .i q.i >c tr o F.d ! 3E€.:'; EIF 8e'd: 6€:'d t.5:€'di.!'68 .il.i.r E! .dOq .r.i!il .d.ilr! .dCiJ'd!!X tr.d € O o A O O.1 > o U O O O 9 O O o O O O O.rd I A 3gE3i 93.; EP!E PE€! E3EEXT3, iE oOouq OOO.i OO.rO 0u.ioOOE .-ooou Ood 9O o9 a ook! oqH! @oko!o6 Ho

g o o o .t, A^^ o d .919 . ts|6 I 04 !l= '31 ,j ,t .il a , a ot ol al o li l.i €t Ft ./l al oli ol .il .i l kl !lol 3 tt ot 6l o !l Hl il .ql ! Hl 31.13ct dt ol ! ol sf i I.i ol 3l Hl 3l dtotS at El ol al o .il ot9tt kt >J ol .il o ol olil! ol ul !l EI I H ol ol ol ol ol H .il .rl .rl 'dl € ol "ldl" ol ol ol ! ol olc o ot ot ol ol .; El dl> i rl cl trl rr ol olo o ot ot ol A o bl @l @l @l o $l A o -320-

~~~~N @ st ^o 66 I o~~~~

~~~~E33 9odr@ E^ i r id! Eg ilo o g €6 6 F o € iN oo .J .d od g !'. r 6 $ r N @@ 6@ € O Fv d d il d +d ai r E <€F€E E 4 €E F O € I~~~~

~~~~il il dd il c o06 s d6 d ri q.d .d .d tsd .J oots F 6k F ! k oo o oo o o o € €H /6 d 6 d~~~~

~~~~o T .Y 9 F o >'Oo E' ' .: 3 .5 .He .H s r !e qEi,iF Hi:! E,: H,Hi:EEi.He :H F fr'E'i=e3 i sa';: 5.:: E.:r€x s€:'t i * $s sggil F ugi sfiii$'i'i:F & i fi !rfrHi frHrrsBEEs:i E!;fi:fififii .;~~~~

j. -! n! oEg E ol .: ij ! bl ?l .e! . 'nl . O o HJI J nl 'l A > ol F F:1l .rr 'etu:l 91 9 5l Er 'rl 3- .?l o .il 5 !l 3l 'tl .rl='l a dl';11 il E o .> ;l .:l gl .: 'tlFl ,jil .dl o 9l trl ,l ol tl ! 'ir Er 6 '5t E :l 91 5l FtHt plbl F oororcrSlSll l'lpl rl J ol !l !l !l ol c Ulhl bl ,.:l ql"' o .91 .91 .:l .:l .el .: 'lli 'dlol 'dlol r !l 3l st slil i E o ol il il ; l-3 sl sl 3l o d.;l ;l ;l ;l ; ;t; 31 3l o I -321-

~~~~d! €6 d d o~~~~

~~~~!d ! ! F ov o o . 9l v5 6l Eo a Q o d; d 64 o c6 /! o^ .c.o o; d@ t€ 06 o n i N@ €€ ts o ij go 6 Q 6 'r!o E o "oRllo a o a o.; bi @ @ + aA @ 6 9d o @ ac| c)v ; d d 9v d d av ql~~~~

~~~~ts d q o dd ! ddo odd ./'ii ! i ,i ! 'd gE 6 k!o o o ! o lr oo oos o .c o o o o! oGB ! B.id F S FA~~~~

~~~~6 a .g s.g s ss.i :Eqp o iq i i*i ::e ii iq i e .HiiB .EI E 5Eg il[ F.E.Ei .E I H'lErE €F EEg ggg s! 3s 3e E P rq* iH ?E o ss s tstd .is gfi EI g E .EIf E$ {.g~~~~

~~~~d a6 g o od !t: .d3~~~~

~~~~Fa3 d€o d6d~~~~

~~~~t-dv . 6 € d . od p € * -o M 6 dN a06~~~~

~~~~E o N ! O O qFFO g 6 o- o 6 6666 P':\ d d6 .il d rddN trg AQ F E d@ - NO O O - *q ? d Eo rl o ts e o o $eoo \i er o ad a .d s n N 6 6o66 I @ 9v E A i d S I iildd n d~~~~

~~~~id idiid dddd 6E60dGSqt .d .d rJ .i idr.d.i.i.d !l!ts !9!OtshHk ooo.coooo ts-gs $ NOS36d6d~~~~

.H. J.C og.Io'dotr .i ji g3 .:g 3 :3 .i H* 'Fq H?E sf;H'd3g!g.H'E:H i:E .E.E€iP i.g l'! 6"*'; !''3 6.i XEIF'd €3gps.t'j'd"BB iE i .i'a F H!.r H.d.do .do o.r@ d ts .d.r,r.J.JE !k rg ! EH,lE$E !Egl' gg g!?f E EE?Elg'*i 33 fl 3sfl 36,9 fls3.: 3.: .:gg:. fl gs.g.g.s:gg

~~~~F ; q~~~~

)d j o z F.dA ol 3 al ol E ,t a .rl 5t ql j .dl ol ol J ol dt rt !l ! ot !t .rt c :l ql 5l ) ";l .rt .t vt E $l ot .it ot tr =l el ./t rt @t o at FI -l ,l El F F *I '-i '-l 'rl '.rlsl ol o ol ol c o ^l 'a! 'ii"l ':l ':11 .:l O 9i :l :?l o a tr 3l ol o al :l 6l ! ol @l oat .:l al ql -323-

.i r€ 6 ! €@ o oon ! !d d o ae& ov qtl o c N c c9 \t o i c a 9, 6E I o N d^ N C^ G d d E € 06 o I o - do . .-D@ u F ts F @ O l:6 io OrOO o0 €dO o oo oo o o o od di 9€Fd o.d 66.+ I NN NN N d N Fv NN NdEv tu(, rrr o OF €F o ro 6 < raF €roo < e€lA

~~~~o uo GF 6 o d- €adi .Oq66 6rd 66 oo Jq g E .r .i .r 9dd 6 ro oo h F ! ! tsts tH ooo s o !s o o Q o oo oo o !d a o 6 O o d q6 ds H33~~~~

~~~~@ o ts o~~~~

o oa .i ad u O .i.d A E & I H 6 .i !k Q J .d.r i ! o o o o tl o o o s .d Q F I F F cc CO O C jl tr 'd .il oo .dg a .t c rd r oi 6a i.d oh o i o oo Eo r croxx odddX c iod o Fo ,:C h X h .d>iXX qXd.dO .d Xtr.r ..1d ts.r E E t cS.ioo .dthuo ci !.d.d sx rl a q o oa3!! oaEd! o ood h! Q! .r .i .r .r.d!OO H.dAhO .r .rHil .'.{ o @fr o o o o o.;{ xo o o >. o ts o o o o j!o oJ o o O oOg!S tsOliA O OtsE Fo utr F c tr cqct I uc60d c d!€ ot- trO o 0 O oOo6ci OOdot O OA.d OOf .d@ O q 6 6@6dr ts@Q@O O OFI

~~~~ts o~~~~

o : o X Y b ol qoG d 6U ol ol {fi|-61 ot 3 €l .:l Fl ol 6; 9lrl t .dl a Hl .el el 'rl .r el El :l3l c ^.il .il >l a ol .rl ;t .?l !.rl dl ol c I s sl .it it a 3l .el f l:l E r'11 ol al I El sl u'Fl -t rt '11 st $| F a 6l ol ol orl Jl 3l y -l jlEl -l *l ol ol qlSl ol ol i.:l .:l el .g 'il .rl 'dl 'dl c ol ol ul o ol ol ! ol cl ol o $t; 3t sl Glh Pt sl K..:l .:l .tl .! ol ol olo @l @l oF/ 3t 3t A = c I -324-

~~~~on @€ oo~~~~

~~~~dvt. 66.c .€ ud!o o60 g!cd; E0co E^ 6^ O es >r E€ d@ o 16 Eo 6 N O to o o N O dN Od NO or 6 d o or H H a d TN Nd NN a! N Fv d N > Ev N N O {. 6.{. Nd Nd~~~~

~~~~!!s ooo ooo tskts~~~~

~~~~dtiF AA i dJ qqoodo od d dct i.idq 'il.1 .d! .d'd ItFOko6 !k !o hI oo oo oo $ $ 3€ 3 € .9 dd 6k o6~~~~

X€ E.r +o o O! o! o.d Ed o o aEoo o oa E o '40tG C a@ .i c g o E ./ .iO .- Q.il >x r .i o E O O I O A At d ql ae xx d E c nci .iQc qlc Eo.i dQ ri.i a o h .d .d c o.i E.i .i oE o x.il tr ! rd E t c hl ooo ac ooo c !Eo O! Oo O ir O Aq KQ @.id oo 9@a .do06c .dOlt./O!r !oO .d.d O.dts Ctr.r€.d o o 60 o s o u0 s o ts o o ts o s a).i o.r c ql o dd Q o o oo Eu C u C ctr s3 sEc tg ei.r 3t958 o tr oo o c o oo EO 300 to oo, !Eodo O.A dd q .A O ! H F q o G Fp o O @q€

k 'dl a rl {l dl 3l o ol .91 -91 ,l d ol Hl .ol d H >' d h ri o s d F o "11 ol o al ol ol 'dl ol ol o al al €l il ol 'rl 6t €l 6l il 2l $l o kl ol -11 >l 3l 6l 5l o .Jl kl ts ; El FI r.| !l d ;l 9l EI F{ o al !l o "tl ot :l o o El EIEI sl trl ts El I g .:l ol ol :l al o .il .e| .el .:l .nl .il a ol ot qt . o ! ot o ol old ! El !l !l 9l Fl o ! a ol ot 5t 6l o cl @l @l @l sl sl 31,54 $t -325-

NN o€ 60 ddd. q/ €(! N..! Nddo{ l' oqo .- h 6l F!!ON o nNoooO q q N N !^ / d^ N O C d6 d € qlN q ' , P d@ N O €@ ts . € o o N N iO o d € N c o G6 0 N N 6 N .i .i .ri o o + o o N N i N F b >v d N ilN (, o & o E ro ra () o < rq € qtA

~~~~o .t o F h o o 6 i a-dOOdO od d .l!!.iid.ild s tsooFooFo st ts ! qaood&o! oo o o qHF6BBd} id d~~~~

o ooooo d FEtACo .d.dt I d d O Oi O O/ O d O^ .i I i F 6A E lA CO d trc. o tO a h h .d .d h .i h .d c > .dv c ! E 5 E F CE t dE d.d .c C .i .rrr O O A O O O f o O O jl .i A O O d O o ! tr tr .il .d .r.d .il .r.d o ! .a .a Q ts tr F ! :5 3 3 E 33 3 33'dfl 3 3'j e gi oO u C tr Cq A CF J! Q Co ! 06 o .do q O O OO O oO 6O O O.r o oO o €.d a o o 6 6 @ o 6 a! 6 o 3 k c o

cl :t 'il ; ^ ! ;l o Et*[l o o :ts FV F ; J 5 I to6G -o l-. >:i > 6 ka * 2 @ o ;, -: I 6 @ .c 'drt*53,t sl € ; o a,J i a ol .El*, i, gl .:l .91 .dl ir ir : o ql o.l .il o ';t .jl aEr f;rsr f;r :ru!l tl ol nl Hl s !t .it ol fl ! .il ot .ct [kE]E B o{ ol o.l 5€l flflf,| ol ol ol H e .dl .rl .il ;:l:l:l:liFrl:l o ot ot ol x o ot ot ol ot ot ol A @ @f ot 6l o r;l;lililgEglil ; 6 -326-

~~~~d 6@r 6 nod NNd ! o o jtsi6@ oo\+$o S^ NNNNd 6^ EO Ee r@ s6 o QooN o + 6 I N N EO @ o oo.J.$ \+ S o 6 o $ od s Fv N NNNN N N N d d N rOv d m e&Ea Fa € Fa E F F4E~~~~

~~~~d!! 600 'doo !g! o FTdHi.s d-d(i6d ss .d ,! !q!s ! .j .r .j .i .dtd! I I o$o9 o ts k ! h ItsO I O OOOO O o o O o ooo F ts tsd H ! ts O O 6 6 o6!~~~~

! . g.HgHi H .H .H F .H .H 3 := =:i 5 : s g="39 g.H s .He" Heg i .i E€'fi:: q.dGo tde .d.d!!d F€ Fe$.fl :.HgH E s .i.JtsE€'; i n e A.- Otd .d k6 .i g fE ?gBi Ig 'gg g g gEEFf !Efi Ei ! s tH tBgs Hs Hl. H g ggEHggf;I gfi g

ii a ! . ol Jogll -.: Il ;;31 iEt ; iEiSl ; €sl * 6rt.9oJR ;i'ut 6 'il k ot 9t (, il €l 3l?l or .; 3l al sl sl3.l 'dl a =l il il 3l =l gl!t ;,j it fil!l "l 61 3l o -.?l fl il; Fl il il gl gl: flfl It gl d E5.Ft+t jitfl jilfl !,1 *l El 8l El:l fl ts q|rlol )l dl J il il rl 3.1 EII H.l 3.1el Sil el tsl '.el .91 .el .:l ,91- .31.9i .91 .91fi1 .91 .el ol . ol ol X r El sl El El t El sli Elsl HI* gsl oP @lHI olHI o tl sl sl sl;sl sl il 3 -327-

~~~~.9 i 6@ 6O KF ! ov OJ oooo co o 6006 k dc NdNN o^!q N6 0t a .;@ ! ts O .i. -@ ONc) O $ 6 Eo I o o o \t o no 46N s 6 6 oi 6 old N I NN dN$ d N 6i Frv N to() o Fa fa n€ iloF n a4~~~~

~~~~I d A i;d sd6 oo o(o o .d .l !.t d.J* o h F ots ho o!o o o o0 s 6 !{ s; 'd3 9ts~~~~

? a E d so oo o.i o o tri o-d E A .dv C'dd 'd q d .rd x o o o oi o.d ro qt dd_a c o oc o c cl iF o >dO tr OiXA .d d tr.i tr '.! .t- ./HO ,c.dd .d ch_cx .d .d .i.r .t oo I cs cqc A!.r @ .iE diJ 4 o 4il q C I O O A O 2.r .d ! i ts o a.i d r ! Fi k 'd.d 'd .d.r .i.d E o.d o O q :\ o d O O O o O E H o o 0 o I O gt3 I !H3& E X X€ r gR 3 BH gEH oo! o oioo .i o o.i a a7 o oo oo@ ooo F oood ! ! gh ! FH 6 o o oo,

~~~~i'e FP (JC)0 b o~~~~

!!E F oo .6.6 d!!>a o J/.ts!!'d E a ol O O 6 o :d A 3l F Fra i, F o 3 .ir .rr .il 'r 5li b or .9 5l Ol.d .iljl .il o .rl ts of; o ^El 3l Hl* =19 d R c ol o ! el ol !l . ol ol Fl d F €l! ! ql Fl Elt El El.dl ol o El d F Fl !l olo !l !l.dl !l 3 6l t ? O ol Al tlq .rl .rlol tl a 9l o o Fl !l !l !l tslJil Fl ts ol o | | | tLt t ol ol ol . ol 016l ol o ol o o.dl .il .ilE .il .rlol .dl .i 'dl .i c ol ol oli ol ol ol o o 3t 3 .: EI Pl EF gl gl,: 8l g Et t I 1l ol ol= ol 016 ol o o ol 0 cil @l' 6l @l> 6l a 6 @ta 3.ot 7 -328-

~~~~.; q c !o ooo a -Fa N ||N() J-q N s c6 €. 6^ .d N € N S3 N C^ 0 6 tN ao sO - 16 @ o oo € tsF o o @ d6 6 € 6o a 9+ od € cd d N TN N Ni &v N Ev E e E rq |4 €O~~~~

~~~~?_.tl ddddF 'a-lolq6dGdd .j a .d ri q.r u .j .i ;i : I F ! 6k o ! H ; E f, 3 $ .s3 g 3 S~~~~

~~~~' .H ,H g .H .iE ; - € i! i:'i=;ifii f'fi i:Eai s;; ;. ;iut i? fr;*iiH:g' ;E fr l;: FlHfif Hfiil~~~~

~~~~sh F~~~~

.dj o qu .do: t .f ; E ts ol ! EX3*3d ot o ..EvoC! '51!l ts ol jld 3 il ol i, ; Esr :, ./l !l jl .Hl €l gt il :l gril ;l il o el .:l '11 ! !t ul El :l 3l il bl 5l el ! cl tsl tr !t ol ts o ol al o 4t al Hl IJ fil{l Elr $t Hl tl ;l ;l ;l ;ls 3l 3l at^ ol E C .:l- .rl .:l c !l-: ol .:l .:l .:l ! '-l= 'rl El El !l !l :ld :l 3l '-lt cl =t) ! .rl jl jl; jl gl o L "t" ;l sl E c -329-

~~~~ts. od dd~~~~

~~~~dd!i cr!60.!' @ o!!H! o ooooo x dl o dcccc o rc T E G^ E 0 ! d^ d^ ! Q S^ o E€ EP o{r E9 o s d@ d@ !o r@ ts o o .c6 J:o Oo J:o o rro a N od od oil od N Ng C .+ €v tQv Av Ov o (,o~~~~

~~~~! o o I I -.i ! 6d o 006a d.i.i!.i o otshok oo !oooo A s6 € 3ddts(~~~~

gg3 0.i rd 'i .53 3 3 06 F q ge ! ! o! ! ! i"qI j e j FXh.i-.h g 3!..5 .: = EH.: .r o.d G o o x .d'- o o 'd8SIE o.I c d .c o 9 d o 3 I .3 PEI d 000!6 o o o9o o cFklo o o tslo s E 'o o'i9t8A A 5 tiA !ctrtsuFttlo-.tslll Ov O.d.rO .iilOi6d O d X q iad o .!, .E.F:.EEE&{b..+aq.E qf lc trLxc.drch.dx.d .ac ? o! .F .?.?.? o.; o iji.i o 5 sd od o J o !,1 o o o o !l g .a c .i o Qh F s,d o o o o H.d q | ts- o I o Eg Ei"AHEE!HEHf;Eg f; q EE f;gg 33 C.:?3E53,rttt gg I : EB J.aJ-

A ..d d.'Q ou ts 3 0l u I ./l o al :t jl :l El El ;l El 'il o !l -3t .et 5l .:t E trI u cl Il 6^1 5.1 !l .il ts .:t Pt ft fil Hl o -! ;l 51 5l :l ;l I H ol ol ol ol ol .d I .-l .d | | .dl .il .il i ul q !1 !:l a 31 31 31 31 3l 5l 51 5l El 6l o al ol @l ol @l @l o

~~~~a -330-~~~~

~~€ x ^.o a o od o .i @@ u oF ,9 o v.v qF d HN oq $ o.;d4 H N o o o !! n €F o ts no N H NN oo H !^ N N!^ o9 $ O.i o - u6 >r! o @ ss nR R.s3 NF @ Od t O Nd Ni sav sd advstl FM €E €E 6< oo <(.,~~

~~~~ErE g q6 '^ kts II 5E .E -e i'i !: ao Hr ri,j c E Et t ad 33 s$ € € s ilj j~~~~

~~~~ss.H g, P snsi:EiHq*.is .H ! .i.Ell;:t.E E ,lu fififi;HgHfifi figHHfr{gi! Hf;*rH {r .E$Hs~~~~

~~~~o .T~~~~

d .o E. dd J Y rl 'f sl F o ql @l 6l .dl .dl a al dl @l Ll dtEt ql[l d@ gl ol UO sl o dt tsl ot @l OI a tsl ol rl .TI ts oo .dt I -l o> ol fl ol Pt >l B .dl ol o ol"l ol ol H JI ol ol ol .it .il"TI ol 0t dl o | @l El E OJ ol ol ol @l 4l ht al &u' @l ol Hl

~~~~{ a -331-~~~~

~~~~.J 9@ @€ o NN N o F o@ od 6NOr o €@ @oo6 oo $$ NN NNNN e €raeo~~~~

~~~~s 4 6d d €E.i€€ FF ootsoo oo ooooq tu 6d do@~~~~

~~~~to doc @ oC .d I d.i ! o .i! o 6 C oo g o E! I d o o9 0 h dogd c S .AeA c d d u o.d€ d~~~~

o.Pg I F 6iElrF I i.trl el oo o d d.d o . o rrop iil E o o C.rlE C sd.dltsltr!l .i q g .dd .d o d rr h.i d d otltJ I o.d .acEo 'dcEo a ts 6 X N o E E 6 t't'tsd o EE ! o u Q.d c.iiddd o > ! .r'i 6 F.i ts.d ! Q o ! d o ! x * x hd d o ! ts 0 q € O O ! aJ !q o ts d o ts E oE oG E ! Q ! 'il ! oo>,ts !kooooG@ !, dd d U! @! 6O.trr o dd od od J9OO9A!tr ro ? go o d I h € o d 6 0 d o d tr .d.11 0 i !6 o bES* EEI$TSUS E il[ 5 .E E.t A,a

.e dq JO E! o F ; '3.r E X E 'dI ' . 6 . .6 u^91 o .; 3! 8:, B ! o EAHI ol 9l d il !l ;t !,*slil EP fl xl E o sl :l €E f;l.ilthl.il tl ol dl 6 o I H 'nlsl d d HI H o Gl i|3l;E|3l ;l dl il$ o s il o !l c c ol .J ll E ol a o il ;|;sls sti5 Jl @ o -332-

~~~~N N6-t @;@ o6 .l F6d. 'od l!sa -6*0o ; ,, _ € €! ;i 6 r Er F^ R-- R' . o oo oo tr ote'iir8sfB;,iEoooo e 5 d .id .d= .o N o 9r) oi oAE @ IC Av N av ;i *F dNo 6 n e o a 6 o € F() nrqE~~~~

~~~~6 €t I oo T r i q I Eo ro..9 o oo o @I u ; 3 b b3 333 o ! o d @ o o 6E~~~~

oo I 'oo.il ..9 s.F .: P P 3 tGt :'d -.'d ts ! .d .d .r oQoo!d.d 'ii c gtcoE d d HE!3'de t t P t .d.d.i.dEo606 ! ? = 6 !.d ts E u ! .i u o @ oedo o./o i! gEEAfS iqi i ! :q'd60 FEh ! e 9.d!!! 0 69 gBBfHs ts Htglo oco FE E Ei.f E o 906(.r E6E g;5 .drts!! o!o ue b.e.:b35e e ! ! sus! E @E

; H ! .'H E< q o FuE . o> xd o dl o o FFcI @ o E .:B :!t = ol ! .gl o ol ql 6l €31 "io, EI !l 6 .Pl pbl 6l or 3l ol T hl 'gl 'dt .il .:l h -rol .)l EI EI !l o ! k! =l ! rt ct El Il iljl €l sl , 9l ol dl FI otl ol dl €l ,:l!l 6l ul F o 6l cl .rl '-l cls kl .:I TI Ll ! t.i iI JITI o al cl k ,!ls ! il a ai -l :t . dl c ul !l iil< *l *l *1,!l*l 6l c a! T] !l ./ ci ^l tt!l L! !l* ol 5 :i ol9 sl slilr sl 3l

~~~~I -333-~~~~

~~~~o d r o~~~~

~~~~6l € o o € 0 q ts @r N O+O6'F @oo o od d i r @ r Nd o 6 N O oo o non E o rA lA ta Cots|~~~~

~~~~ts o B k o d q~~~~

~~~~d ^d !.d€ € 9.d t € tq!t 99 oFo o ok o o os ! o o o o o- o oo o o oo o o o o60 0 od o 6 0d 6 @ 6~~~~

~~~~I ts~~~~

~~~~! d 60 !q 6 o./ OH a osoa*"tB..1.. .i .a q c t c . "'d a He" Hi.EHP .Hsst.Ht i .He .FE fi'!E.EFit.H ! H'€'{hEE H 6 ffs€F f'HI gEgIgH ;H'i€I E E a sEE$ O .:Hg ss'i!.i E EEgFb e e E b,i5.:~~~~

= d! Bfi 'E 3l :jt .:t .11 ! [l ,; rl :l jj at dt (l flH El*,l1 El 3t !l Fl o ;l -31'r 33t''l '-l5l a F ts J 'itol 'ilol 'it:l 'dttl:l5l fllltl fl o cJl Gl ol 6l o il il il !1,!l!l o cl cl sl olc d o !l !l !l llo .i !t !l 9l el€ E OI OI ul alo slsls sl v -334-

~~~~@ o j ,.: q6~~~~

~~~~d !! q .soo I N o S E^ E^ qo o o N.$O @ € q @ 6A @@€ ots t - € NN O d i NN N O o o 6NR R i9 =9 os E o o NOo O p.v Av EA O 14 0 lA ela < rq E aQE E C () o~~~~

~~~~oo oO 0o !€€ € E€ @o i rg tt t 3 ooo o oo € oo .c o ooo oo BB , o oo ! 6 o~~~~

~~~~o H F ?'o Qd .: .i 6.t3 tg s H tr.H u.d HE* #* * f f F.93 85 E "",5 :oH oo a HgH 93 E I otr oa o doo Ek ! E€€EE$fiEFfi:gg~~~~

~~~~3: .J 8.8 3 o~~~~

oo F @ .rl .i ii 6 f , iflif|,fi,f, +* ur o il 2l q 91 9 5 ':: ol gl a frl i, tllrillltl srHllr o *lIr sl ul 'clql .it .1 s EI EI E el!l 6l a :l !l il fllH Ft il iliHl ! :litllii*l f Hlol H FI !t !t !t it{r!tErst ll.it*l !l ol ; !l ! *l il fl ! !l rl I OI jbst o il il 3lsls ;ls5 il il sl -335-

~~~~6 ! o~~~~

~~~~6l o r o N o tr^ N q €€ H I F@ o 06 d s { $ oo o N Nd O .td o 6 d i 60 o oo o Bv o o o o oo ql & F O'A A~~~~

~~~~d o k H o 6 o o B ! o- o .d !€ g €€.r E € E9 6 ! oo o oo! oo o o oo o oo o ooo ao o o oo d 6@ 6 6AS oooo~~~~

o g ots EK g g B ' i' 3' 3 Pg cg .n .l o .f, - b E .d- .i .d.C Ed o d i .i Od ,d O d.d A e E O.i! o d d C.i O tCn h E qi q .i9l trfO ! ! O.r o ! ./.d.i O .i !.d ! ! o .i I t ! e 9.lEs 3 HHH g r 3€ I b€ rg.e t s399 3 ==i g I 8t I S'€ 5!b E Cuu! q .d.ilO O ! EO E !'i d9O e es3fl e *il8 b s E* e .:5 €b3

7 -; o o El o !* ! al q I rc .glql rr 6 il3 d qlo !lHl fil Bt El NI o 'd sl ol btq! q eil3l s El $ ;fil?l fl il e ol ol ot o tr tst kt 'l o n1{l;l 3l el s otl o .!l (l d ot 6 F 'at .il F tsl ! ql I H o dl o jt illilill il ill ;l 6 *l .s @ dl cl d 6l d o ul ul ul ! gl gl tr ol ol c rl'i €l ts sl f; a OI U 8l.e 31 3t 3l a sStlol -336-

~~~~N o ts o ! @ {^ -!@ s o6 rF xo o N,OO Ed SN sod xs a&v 49 mFt EE~~~~

~~~~! o o o H~~~~

~~~~qo o o !k I o T T TT oo o o 6 a6 qo ! o~~~~

~~~~I N x o I o o N I~~~~

H l.d o dF ho ! 6 4q I o o I c EE 3e .Z ot o ie*'l r€ F .E,.Hs .H.H 'ro E.'iX'4 !! oo ueq.e c E E do .ii_!? g; .E3 b'E.er .: .9 o o d9 ;EBt s.i 36 rlsg I H 0 l'e :1F+l' 9gl i'n !6i.i ; E E k€ 6!o-* 53 .1! .:g3g g g o dc 'p o o 6 o h 6sl o 'Eilb o it o o @ ': o o # x:$li-ol- : q o E H li o o : q ii ..BtHl . +;1,?,E o o ! Al - & ! o

~~~~'@ o d.d d6d E6 .9 d9o! doo ol q o .+!d66 o o x^ ts^ l^ oO dd oO F . .i o@ .o o d@ o€ o .{' o H tso 5o F6 F € o s + $ 5 .ild Or .rr $ s o o o o Ev =v Fv o o o raEa~~~~

~~~~ts d~~~~

~~~~o diid doodc6 o 9.dHdi.i't.r F o!60'0!!l oorosE"ooo E A a6q-ll'd60 o~~~~

~~~~o x o o !~~~~

I o tr X I o oo irl o@ HQ ON H c. X6 oo Hdo o! q tu z6 o o_ d td c ts >\o A! 6 .d d Ec it o g - !'i 6 E <@ E O O€ o O o O o'i !l C tr N o E.d C C s E o E E o rl o.d E c hN .d.d !.n c .i d €h o .r x'r d9 1> .i sq 3l o o€ I c oi o o o d o o o s ol dF .d ./ !o qE ro o dd .: 6 O E ! € H O O xo O o O og I o >h .r ts u G .d 9€ d 3l r dO O ! .ch d! ok F EC dl i il E I O o O I 5 ! .d.i g .B I rl G 6 @ r o d ,c ts 6 o H4{ ol 9l ! .rl @ al o j:l dl -l o E 6 ! ts j .91 h o xl = @ jjtS ;l ol EI B sl o lil il El ! it al :l >; I :l ;oJ st o !r ;l 6l gl ! 9l B q d qt Fl gl N .il gl ul 'il o :rl 6r 'o d ol a olj ol olilEt ,Bt q dl 'i sl 7l kl -t -l gl -t .ol =l =l I .:l 3l t .dl cq al at al :l F ol 6l .;l al e Hl o .l o P{ il HI ot o !t 6 ot 6t F ts il il ol 'il d d cl 0 cl .il dl it o .; qt o !t .JI H El 3l EIH O El 6 .it .il 9l OI ' GI E _ trl rd trl sl o o ol ol .dl *$l kt r<

~~~~j s 3 N~~~~

~~~~€€F ddh x BFB HS n H ido 66€EF'Q€€ ,ets|rq~~~~

~~~~datdiqC,io';? 'i 'n .al a6 .A .d q .; tr.t! kko ooo H H H H $"q H x qd!~~~~

~~~~o .: ooB~~~~

~~~~o qoo."E'do q .:. .lr. q EG E I q e .r! E o E X tr ! .i !u s h od0 q1 .eF EXA ;'E H; i fr i, O O ed ,c .d oao 6xd H ? !!! H T,: ; E€ € $ q '1 tl tsq 6 d rdo '; ',; 'i! oid E 3 i a OAE~~~~

dl ^il| >.rl I * -? x:l .; x 6 q ol .? *fl = .? .':B. i Fl 4ol B B HvHtvdl v.o o aqy t;i!'31 E.a ,ril Alr t.: I OJ sl :il i Jlo o 6 0l :l ':l.{ll til @ trl ol 6 1l o o o.l al cl o Bl E!l ;ti sl ol ol .r ol tl Xl s EI ol .rl €l .d :IE N ^qt il p gl 9l ol ':l 41.; FI .ilt dt .r il .Il'trcril ;I;[:FIHHJ o ! col ol tl El 6 -91 o'o E il ':l I | !l.X !1.! it .il sl :] 6t ot :tck *lfH Hla ol 3l s lo vol F .il ;l ;li E|SI 3r," ?to :l .:l =lt I trl ilo ol rl . . i]il :t 5t iti 3 hl3* htc tst o ol-c +l {l! fi , flE! g .el ;t :l 3t. 3 -:l EE 3 -lE 'l o or.: o ileo rl Mc qto E sl El bl El*' 5 Hl Hl* x SlPh .: 3l*E ot o ol dt et. o Fl !t. I El.dE o !l .E € dl- tl 9t. o i:l alo F Ft ht= €lov

~~~~,o @ o6 !d~~~~

~~~~BT od €o o"l Fdo! qe9o E^ 6 0 .in ! q gF o € o o o d N !o o @ 6 H r E a n 6 I N € n od € € 6 r F o ooo.SootvooJOOot o d fafaanrarq~~~~

~~~~! o o ! F d~~~~

~~~~dil dJidd 6006d 6('64O .ddi.i.d .J .i aA 'd 'd q kookts KFtsHH ottoo ooooo tu dFB66 6drlaE~~~~

~~~~'0 0~~~~

6 H s oalooo c9ddE o rr o .il O O .J O.J .i O O.d d o c t c Ei n trd A I ci o O d C,J d ./ O.r O r.ir d d .r .do c h ,dc * E Eu tr xcx >. c @ { .d Z' H.; ]d .J .i o .r -Z.eZ' .c .;{ t trd O q 6k A k OC O OkO f O O oli ! e dd o td o s fl hEg I g b oo ; b 3.8 bt E l[".i H dod q k F Es o o6 3 E EP ,e 3 $& 3 i.8F E 3 E Ed 3l 'ElEI ;. :jl J ot .dlq .if . El 9", ;El al . ol I . O A Sl 1gl .9 3 B 3 bq d ol n ! BE o ,ttl olo.jj* q i"r"rE . @l 0l 5l N d,€l3l flElf;l il*E"r *l *l ;t. ;l o fl .9t 6tH 5t. El d €l ;l!l jl $fragl*l cH,Eli il H 3€lfl H sjl !l 3l f| il il o x a ;l:;t;l ;l ;lqt;h ;l s jl gl jl o jl jlii gEjl o tjl il $l $|i{l#l o -340-

~~~~@60 €N H!O a hNEiNOF o OqEFnF @ iiigi i i N @6d q $ d €3 3 H $* : R S tsF@ n d d qv I N o o d o o ooo F O F < F € rArA O F G €Fan~~~~

~~~~o o 6 - d d a t s id il d d .! { q d 6 d 66 6 6 6 ':. .i .n .d .i .d .il .r .d {o@ .r .d dd H F H ts F ! ql KK F K F Hoo I o o o o oo 6 it o o d d d 6 G d6 6 6 6 H€€~~~~

i:r O !B FSFP!'- u'i c o .i 99 o9 0! .d ? F.'i F.'i : N x.r x.r N -i :t:6 : HK l. ts !g ! t ! O i oH oF k .i.iOOO Et et '5.t '.5 sgiFjF I 'Eq.dq.jdoed '! H HTi€I.E g '! i''a'! E P * sg !, t t 'd 'i U Eq s'd € : silf$g E E S 6 od d x F F tr E g H ! c q >F E ! EIFLH.l t P d I d ii ilk 6 o i rh&_i&j F i I i ,t 3 E'ds 3 t odi

;

~~~~o. j;~~~~

o,d hJ ." r ilp .i o ol 3 gg o iJ, ol o tl 3 t . *Elf .dl ,ll 'n 9l N tl o dl inl F ;, ;l;l :r ;l $';l .il rl ul E nl d s glil tlll $l EI F E:I:IE i1 rl :l Il ql :l a hl ! nl nlTl hl El.il ol ol o ilf,lil o .rl.il.i HI o ol ol o il .:3t 3t B UI I ol ol o ol 9@l6l@ ;l ;lilllilslgs|;'l 3l I .3 -341-

~~~~o o o € o o~~~~

~~~~ro N $$6 + 6 \+ F 6 0 0 @o \tso + \t -+ oo re Fa ro faro~~~~

~~~~o€t ooo ooo oo ii.-r sdoos .j€€.;;;€.jd{99 kooFoootsooo!l 3 33$ € € gS 933 3 g '€ €i~~~~

d,, at @r dr €ll xO ,x tF i>r T T .E.H T T B €€u t.Hu .Hi .H il ji .i'ifrgitSi i$si.Ftt.,tHi F FF f ;Eifsli':'iFi + fiis *.* isjs Fiib$IilAIX Ht F ifir"r.: s .eH ifiit H.ibtFibgFlH i H ,F, i?it5 X.: F.e F i?rE5.I 5.i Xt€+i*€fr€* F € $ EI Fti*i j FtF[ixp]jFiii] i t i ii

k H o ..: AO 6 !i o6 o 4: d6 H; Fp .i6 t ti doJFl rd 6l dl = a di F .dt ol 6H F @l .i o @l cl dl rll ii.i , .dl ! 'd.l ol @ 6l El .i a N ol E 5t 3l gl ot d !l rF;d ;lsl olx o fl€l xl !! d 9t 9l ;ol A El trl El 3{ ;l: d I s ts 61 6 s dl sl jl€ .dt .t 'nt 'il .el .Il 3 cl ! H o ot 6 d at 6l jl s dld il .il+ E 6l s d !l !t !l :l sl sls , ol o :t xl tsl ! H ;t El 9lo o ol ol 8l8l 8lEE o -342-

~~~~H a ! @ N6 o @ d^ _o o xS !6 =l o \t6 € N nd F€ 4 o+ 9 o oo o o 66 60 6 O6 Cv ooooo-o6 s6 n rQ mo ,o 6 ,Qe €Fa ra FN~~~~

~~~~o €i ;.g €€€EEEEEE"9.9 .9 3 .g.g .9.9 .9 ,'r€E~~~~

oo .& J" Er lok i,ro oo 5"E'" .! s 66 b.5f.: P s lr9 q9 !i .r qcOo E \x ?'{3,'i 3. Pg B. I . o g'i .E.H ld ooo !il5 o @s riIi t t i E.HoE fcE Hi;i .E EE .E *.? xsG i :-{ :-{ Eu Ft I oo ots@ iaih € EE i. e ,.:E.Ee &! 3A €l il dl !l .!l sl .il .il El !* :l 3t ; a d "i d j Il E?i ! j..: ! .:l slNI al fi r 6lr r,!i *, s s: sl .rl Fl &:s N xl 6l Hl fl Il ir Hl ;l ol .dl :l;l sl 0 tsl BI it ltEiEt ;l il ,$l I il Hl cl o (t !l llgl El I jl gl ol 3r!r fl .tl :r Er il !l Elll €.:l .:11' tt :t .it .Ft .:l FI fr .rl €l it fl

~~~~&l g|:|*| d't ll5El lilE sl fl f|H "; 5 -343-~~~~

~~~~o o o~~~~

~~~~dl d60 o ooo .$+o o I €No$oo$oooooooo o ooooooooooo0606 o -1. $ + o o s o o o o o o o o o & ca lq e Fq la n Fq fa e o € aQ !q fa,rl~~~~

~~~~I 6~~~~

~~~~ooooooooooooooqt iddfl/ildilddilddd d ooooooooooooooo A €€€€EE f € €€€€E E f~~~~

o rbEr.oa k J H jr, F F{ Fr Fl .rH Fr Flrr o q o o o o o o0 o oo a d g E a trF c E a tr c9 c 9c E .d 'd .i E ./ .r .i .r .d .d .i.d !t .d,i .il o @ A 6q @ 6 @ 6 60 6 OO o o i r O A > aA o O A A rO r AA .: E .9 e 2 2 b _ 93 P 3 P 3 38 3 89 o .d .d o o o a oo dt o o o oo o oo o & t d 6 d , 6d s 6 6 s G6 o o6 > d t a d d E rr r i d d ii d di ts d a d o G F 6 a d o d a Go 6 a6 3 E S f"'^ € 3 tr€ 6. 5o € € €" € "€ ;

~~~~dq e .;; a; o! ol qg >F&EP;i.:t 'r& o '.:;3or El gl 5l .3 Elti' El o ",Il 0 ? E 6':lHlElili~~~~

N 'r ':lsl €l:J El g o .i, El x 3l :ll Gl ol dl ol o tlf,rltttlil 6 Eil Fl El .:l .:l .:l .:l .: .:l .:l .:l .:l .:l .:l gl o ;:l 'rl:l 'il*l clsl sls o E'il jt sl sl el sl s HlflHlfl81frls H&tjt rt il rt it ! a A fl o ;::l El :l *l fl fl E EIEI EI EI EI EI E ! -344-

~~~~qloF€@o@ @ oo Iro{$\+$S*q9oooo -+ s.$ FEAe€oq FATA~~~~

~~~~d o~~~~

~~~~aloooooo o =da.iHd :o9oooo o r€f € €f € B o.o~~~~

~~~~ts rEg o. .3 s q q s q sfi sfi .:.:E EEE.g.E; g'EJ O EP gF ,IE .:.1'e:a LE .EXc.n'F.:€ .EsG.i 'E:t E' i q ; !E ;, a Elli Efii .F E o.! i !! I t ;Err sEff s a 3 iH 3 H i:H3 j:gH 3~~~~

.T o q o ,i ;, er .: €l FiQ o EI EI q ll H Uli i 5 'rl!l ita tl ll .: E .c 6r N 'ilo '9 Hl 3l ;5;.is!l !l Ft! o ol tl dr et 6l ;l Elo E nl ,l Hl ,:l ::l .:l .Fl .:l a '61 oq dl H .I Il 'ilUl 9l €l 3l ;IT s'l .El .:l & Yl! H : 3'l fl ;lilIl5l *l 5l! H 3 SlFlol;l 'l el 3 N 3.:l :l *l sl sl tl al i .ile * rr Er Er:r El ?l El EI .c 4l! o !elflJlgl ElJlEl El E 3ls i.: o5 -345-

~~~~6l o o o F.: N 'Jn o -t \t ss s$ o €ra ,aa NE~~~~

~~~~ts o~~~~

ol ! d o El 'd!qqq ( tsosdd 9'Xt ooooo o dl eFiri H AI o il qillg'l o p -Pl 6 i .d u El .t e'st.dl NilI 601 3b d hl o I hl d.d .il N^ Rel .dl -*l dv .. ol oro trd !>rd $l*l =l o h"qi,8 il d qc otr ts I o h o F.d q d.4 o o . Jrlolol d t l.a c g./ g d{ F .9Et ; rl'*l .3 t :E5 :85 E .tnl o o.ic > o!6 ooo !qo t ul I C o o X d O d E 6d I cl o O dI€I o.d.i h o otd odd oE.r EEC! F !dX Ed5 l.9d 's p .d.i.r o I o ho o h0 olH x :lEl ooo! h ddN 9iN Elsd ''rrl OooO 5 dOd 6St Sl9{ .3 o StE6 | d.iO r.do glH.r E -!l !!u.i d O!D AlP !l ! U^Ol r.rl q@ El oq 9l svOl

do^ c9 o o@ fE o >6 u Erd €^ o tv ooHd tc-.i.N 5o, 610 o ts oo dblE6rX I,CO Er Hl'I €l E o'il! o ql 'nl ?l 'ild o !6> 5l 31 3 'il a gl; N !J !t .rt cl El a o 'il 'lt Pl "lb o oo I oq 6 *l .el .11, .1ll a da ts .dl ol dl! nld "11dl 90!! . &l 8li ooo o rt:l a at !tBlE €J Eto jltl o oo g g E dM 6 [ iltl d FlEl d .rlEl dlQHl': ;. dlxHl'; illHl c Ff u ll ah-E alo tul < TAO

~~~~sd @a dl~~~~

~~~~Ottrer tLtles available ln ttre ENVIRONI{EIUAL HEALU CRITER.IA series (eontlnued):~~~~

44. Mlrex 45. CanphechLor 46. GuideLines for the Study of Genetlc Effeets in Human Populatlons 47. Sunmary Report on the Evaluation of Short-tern Tests for Carcinogens (Collaboratlve Study on In Vitro Tests) 48. Dlnethyl Sulfate 49. Acrylanide 50. Trlchloroethylene 51. Gulde to Short-tern Tests for Detectlng Mutagenic and Carcinogenlc Chemicals 52. Toluene 53. Asbestos and Other Natural Mineral- Flbres 54. Annonia 55. Ethylene Oxide 56. Propylene Oxide 57. Principles of Toxicokinetlc Studies 58. Selenium 59. Prlnclples for EvaLuating Health Rlsks fron Chemicals Durlng Infaney and Early Childhood: The Need for a Speclal Approach 50. Prlnclples and Methods for the Assessment of Neurotoxicity Associated Wlth Exposure to Chemlcals 61. Ghronium 62. L,2-Di.chloroethane 63. Organophosphorus Insecticides - A General IntroductLon 64, Carbamate Pestlcides - A General Introduction 55. Butanols - Four Isomers 66. Kelevan 57. Tetradifon 58. Hydrazine 69. Magnetic Fields 70. Principles for the Safety Assessment of Food Addltives and Contaminants in Food 7L. Pentachlorophenol 72. Principles of Studies on Diseases of Suspected Chenical Etiology and Their Prevention 73. Phosphine and SeLected Metal Phosphldes 74. Diaminotoluenes 75. Toluene Dilsocyanates 76, ltriocarbanate Pesticides - A General Introduction 77. Man-made Mineral Fibres 78. Dithiocarbanate Pesticldes - A General Introductlon 79. DlchLorvos