Journal of Ethnopharmacology 78 (2001) 51 – 58 www.elsevier.com/locate/jethpharm The effect of Senecio latifolius a plant used as a South African traditional medicine, on a human hepatoma cell line V. Steenkamp a, M.J. Stewart a,*, S. van der Merwe b, M. Zuckerman c, N.J. Crowther a a Department of Chemical Pathology, South African Institute for Medical Research, Uni6ersity of the Witwatersrand Medical School, 7 York Road, Parktown 2193, Gauteng, South Africa b Department of Gastroenterology, Uni6ersity of Pretoria, Pretoria, South Africa c Department of Paediatrics, Coronation Hospital, Johannesburg, South Africa Received 10 April 2001; received in revised form 30 June 2001; accepted 17 July 2001 Abstract A number of traditional remedies used in South Africa contain pyrrolizidine alkaloids, some of which are hepatotoxic. We investigated the effect on human HuH-7 cells of Senecio latifolius DC., a plant that is a component of some traditional remedies and which is known to contain toxic pyrrolizidine alkaloids. Cells were also treated with extracts of a standard pyrrolizidine, retrorsine. The changes in the gross morphology of the cells were studied using light microscopy after haematoxylin and eosin staining. The cytoskeleton was investigated using fluorescence-labelled anti-b-tubulin antibody and the nuclear organisation was studied using fluorescence-labelled antinuclear antibodies. The plant extracts gave rise to dose-dependent gross morphological changes. At high doses, we observed necrosis and at lower doses, destruction of the cytoskeleton, nuclear fragmentation and apoptosis. Doses of less than the equivalent of 330 ng/ml retrorsine led to multinucleated cells with failure in spindle formation and clumping of nuclear chromatin. This latter finding suggests that chronic low-dose treatment with such traditional remedies could give rise to teratogenic and/or carcinogenic effects. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pyrrolizidine alkaloids; Retrorsine; Senecio latifolius; HuH-7; Apoptosis 1. Introduction Although the use of traditional medicine (muti ) is common in South Africa (Watt and Breyer-Brandwijk, 1962), the frequency and severity of adverse effects is difficult to demonstrate. Acute clinical hepatotoxicity is one of the most commonly reported effects (Hutchings, 1989). Many plants in South Africa contain pyrrolizidine alkaloids (PAs), the majority of which are non-toxic, but 24 species have been shown to be toxic, including Senecio latifolius DC., also known as Dan’s cabbage, groundsel or ragwort (Papetloana in SeSotho). S. latifolius leaves, as a paste are used by the Xhosa for the treatment of burns and wounds. The Zulu use a decoction of the root as an emetic and as an enema in * Corresponding author. Tel.: + 27-11-489-8551; fax: + 27-11-7172521. E-mail address: mikes@mail.saimr.wits.ac.za (M.J. Stewart). chest complaints. A decoction is also used for the treatment of venereal diseases (Watt and Breyer-Brandwijk, 1962). When used inappropriately as a remedy in humans, PAs have been shown to give rise to both acute and chronic liver disease, in particular veno-occlusive liver disease (VOD) (Willmot and Robertson, 1920; Weston et al., 1987). However, the reported scale of usage of these remedies in South Africa indicates that many patients ingest these remedies over long periods. PAs are excreted in breast milk and have been shown to affect breast-fed human children (Roulet et al., 1988). Most South African black babies are breastfed and the effects of low-dose intermittent exposure to PAs in humans are undocumented. Estimates of pyrrolizidine consumption by humans have been made (Culvenor, 1983) but are not applicable to South African practice in which the quantities ingested are not controlled. In addition, there are no published pharmacokinetics studies which link PA dosage to plasma concentrations in humans. 0378-8741/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 8 7 4 1 ( 0 1 ) 0 0 3 2 1 - X 52 V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 While the relationship between the chemistry of individual pyrrolizidines and their purported efficacy have not been elucidated, that between pyrrolizidine chemistry and hepatotoxicity is well documented (Leonard, 1960; Warren, 1978; Huxtable, 1979; Mattocks, 1986; Hirono, 1987). The alkaloids are classified into four major groups: macrocyclic diesters, open diesters, monoesters and necine bases (Kim et al., 1993). In plants, PAs occur as free alkaloids and as N-oxides, both of which may be toxic. Three conditions are essential for the hepatotoxicity of PAs: 1. A 1 –2 double bond in the necine base. 2. Esterification of the hydroxyl group in one or more positions. 3. A branched carbon chain in at least one of the ester side chains (McLean, 1970). The degree of toxicity of PAs correlates with the complexity of PA esterification, with cyclic esters as the most hepatotoxic. In cattle, the N-oxides have been shown to be as toxic as the free bases (Molyneux et al., 1991), but it is disputable whether this is so in humans (Huxtable, 1980). PA salts are readily absorbed from the gastro-intestinal tract into the portal circulation (De Waal, 1941) and some PAs may be absorbed through the skin (Schoental, 1955). There are two major routes of PA metabolism; one involves N-oxidation via the mixed function oxidase system (Powis and Wincentsen, 1980) and the other involves dehydrogenation to dehydroalkaloids or pyrroles also by means of cytochrome (CYP) P450 (Williams et al., 1989; Miranda et al., 1991). The dehydroalkaloids and pyrroles are highly reactive (Cheeke, 1988) and the dehydroalkaloids have short half-lives, of the order of seconds, in aqueous media (Glowaz et al., 1992; Yan and Huxtable, 1994). The toxicity and metabolism of PAs are markedly influenced by the availability of glutathione and taurine to which they link to form non-toxic excretory products (Yan and Huxtable, 1995; Lin et al., 1998). The effects of pyrrolizidines may differ when they are administered as concoctions of plant material rather than the pure chemical. There are few in vitro studies, which have confirmed possible therapeutic or protective effects of plants containing pyrrolizidine alkaloids. Liu and Ng (2000) have shown that extracts of S. scandens have anti-oxidant activity, but as Hammond et al., 1998 have commented, ‘‘the biological activities of most herbal remedies have yet to be confirmed in the laboratory’’. In vitro studies with a primary hepatocyte culture/DNA repair test have shown the toxic pyrrolizidine, lasiocarpine, to be genotoxic (Williams et al., 1980), the toxicity being related to the metabolites (Mattocks and Legg, 1980). Lasiocarpine has also been shown to cause chromosomal aberrations and inhibition of RNA synthesis (Reddy et al., 1968) and other PAs have been shown to be mutagenic in vitro (Yamanaka et al., 1979). Animal studies with chronic low-dose administration of PAs have indicated that these give rise to a variety of malignancies (Cook et al., 1950; Schoental, 1968; Svoboda and Reddy, 1972) and to a variety of non-hepatic pathologies including lesions to the lung, and kidney (Hayashi and Lalich, 1967). PAs are known to cross the placenta (Bull et al., 1968) and teratogenicity has been demonstrated in foetal rats whose mothers were fed lasiocarpine (Green and Christie, 1961). It is of interest that immunosuppression has been reported in mice treated with PAs (Deyo, 1991). We had previously observed apoptotic cells in histological specimens from the liver of a patient who had ingested a traditional remedy which was thought to contain toxic PAs. As a complement to our clinical studies, we wished to investigate the in vitro effects of extracts of these plants on the morphology of a human hepatoma cell line. In particular, we were interested in the possible effects of low-dose administration. The purpose of this study was to observe the effects on the gross morphology, cytoskeleton and nuclear components of HuH-7 cells after treatment with extracts of Senecio latifolius and to compare the effect with those obtained using a standard toxic pyrrolizidine (retrorsine). 2. Methods 2.1. Plant specimens Senecio latifolius DC. was accessioned to the Pretoria National Herbarium as number 9411000.231. 2.2. Specimen preparation S. latifolius stems and leaves were dried, cut and ground to a fine powder using a pestle and mortar. For treatment of cells, 1 g of the powdered material was extracted by suspension in 10 ml boiling (distilled) water and infusing for 15 min. The suspension was centrifuged and the supernatant filtered through Whatman number 1 filter paper and then filter-sterilised using a 0.22 mm filter (Waters Corporation, Milford, MA, USA). For extraction of PAs, powdered plant material was extracted using the method of van Wyk et al. (1992). Briefly, 1 g of powder was extracted twice with 0.05 M sulphuric acid, passed through a glass column containing alkalinised celite, extracted with dry dichloromethane and evaporated to dryness. 2.3. Screening and quantitation of pyrrolizidine alkaloids The residue was screened for total alkaloids using the V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 colorimetric method of Birecka et al. (1981) which relies on the reaction of the alkaloids with methyl orange followed by the release of the ion pair with sulphuric acid and measurement of the orange colour at 525 nm. Toxic PAs were estimated using the method of Mattocks (1968) in which the pyrroles are reacted with Ehrlich’s reagent to give a magenta-coloured compound, which may be analysed quantitatively at 555 nm. All measurements were carried out in a 1 cm light path quartz cell using a Cintra 5 spectrophotometer (GBC Scientific Equipment, Melbourne, Australia). The yield of PAs obtained from 1 g dried material was 1.75 mg total alkaloid. 2.4. Standard Synthetic retrorsine, 12,18-dihydroxysenecionan11,16-dione (b-Longilobine) was purchased from Sigma Chemical company (St Louis, USA). 2.5. Cell culture The human hepatoma cell line, HuH-7, was obtained from Dr Patrick Arbuthnot (Department of Molecular Medicine and Haematology, University of the Witwatersrand). The cells were kept in continuous culture and maintained in Dulbecco’s Modified Eagle’s medium (DMEM), supplemented with 5% foetal calf serum (FCS), penicillin (50 units/ml), streptomycin (0.05 mg/ ml) and glutamine (2 mM) which were purchased from Gibco (BRL, Gaithersburg, MD, USA). Aseptic techniques were applied throughout and all experiments were carried out in a laminar airflow cabinet. All solutions were sterilised using a 0.22 mm filter (Waters Corporation, Milford, MA, USA). 2.6. Viability study Viability studies were carried out at the beginning and end of each time period using the microtitre tetrazolium (MTT) assay and flow cytometry with propidium iodide stain (Ware, 1985). 2.7. Morphological studies HuH-7 cells were seeded on to coverslips in 6-well multiplates (Nunc) at a density of 90 000 cells per well. After growth to near confluence, the medium was changed to 3 ml containing concentrations of retrorsine ranging from 33 ng/ml to 33 mg/ml (0.94 nM–0.94 mM) or S. latifolius extract containing the equivalent of 13.5 mg of total alkaloid. Concentrations greater than this had been shown in a concentration study to be lethal to the cells. Control cells were treated with fresh medium containing no PAs. A time-dependent study over 72 h was carried out. Cells were stained with haematoxylin 53 and eosin (H&E) using standard procedures (Kiernan, 1990) and were viewed and photographed using light microscopy. 2.8. Indirect immunofluorescence The effects of S. latifolius extracts on the cytoskeleton were studied using mouse monoclonal antibodies (Sigma Clone TUB2.1) against b-tubulin. Binding to the b-tubulin was visualised using biotinylated goat anti-mouse IgG (Fab specific), which was rendered fluorescent with ExtrAvidin FITC (Sigma) alone or with the addition of 0.1 mg/ml 4,6-diamino-2-phenylindole (DAPI) (Sigma). This procedure highlights the cytoskeleton in interphase and spindle formation in metaphase. The effect on nuclear morphology was studied by treatment with a human antinuclear antibody, which recognises a single 34– 36 kDa fibrillar protein, fibrillarin. A goat anti-human IgG conjugated with FITC was used as the secondary antibody. Photography was carried out using 1600 ASA film on a Nikon Optiphot microscope equipped with an episcopicfluorescence attachment and an excitation– emission filter with an average wavelength of 425 nm for FITC, and 400 nm for DAPI. A similar series of experiments were carried out using the standard PA, retrorsine at a concentration of 330 ng/ml. 2.9. Detection of apoptosis using flow cytometry (Sherwood and Schimke, 1995) Cells, cultured in 6-well multiplates and exposed to S. latifolius extracts or retrorsine for 24 h, were harvested by removing the tissue culture medium and adding 1 ml 0.25% trypsin containing 1 mM EDTA in phosphate buffered saline to the wells for 5 min at 37 °C. The detached cells were dispersed by gentle ‘pipetting’ of the solution through a syringe needle. The number of cells present per well was determined using a haemocytometer to ensure a minimum of 0.5–1.0 ×106 cells. Cells in suspension were stained using indirect immunofluorescence procedures: 2×106 cells were first incubated with purified Apo 2.7 mouse monoclonal antibody (clone 2.7A6A3) (Beckman Coulter, USA), which binds to the 38 kDa mitochondrial membrane protein which is exposed in cells undergoing apoptosis, or the relevant mouse isotypic control (purified IgG1, clone 679.1Mc7) at room temperature, using the recommended dilutions. Cells were then washed and incubated with the secondary immunoglobulin, goat anti-mouse F(Ab) 2 Fluorescein (Beckman Coulter, USA). In order to detect all cells that bind the Apo2.7 antibody, permeabilisation is necessary. For this purpose, saponin (Sigma) was added during each incubation period at a final dilution of 0.3% and during each wash at 0.1%. Analysis of 10 000 54 V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 cells/sample was carried out using a Coulter Epics XL flow cytometer (Beckman Coulter, USA) with a 488 nm argon laser. A positive control was prepared by irradiating a cell culture with 100 cGy for 2.30 min and then incubating the cells for a further 24 h. 2.10. Estimation of DNA content using flow cytometry After incubation for 24, 48 or 72 h, cells were treated with 1ml acid pepsin, incubated for 5 min at 37 °C, and then for a further 2 min at room temperature while being stirred. The solution was filtered through a 50 mm nylon mesh to remove debris and 1 ml PBS was added to neutralise the pH, followed by 6 ml of DAPI solution (2 mg/l). This was allowed to stand overnight at 4 °C. Flow cytometry was carried out using a PASIII flow cytometer equipped with a high pressure 100 W mercury lamp, a UG-1 excitation filter, dichroic mirror TK-420 and a GG 435 barrier filter (Partec, Munster, Germany). DNA histograms of at least 10 000 cells were plotted. The diploid cell population was used as an internal reference standard for the identification of aneuploid clones. The coefficient of variation of the assay was 92.1%. Normal cells and a normal dividing cell stained with DAPI are shown in Fig. 1H. After treatment with an extract of S. latifolius, apoptotic cells can be clearly seen in Fig. 1I and that of treatment with retrorsine in Fig. 1J. This shows lagging of the chromosomes during attempted cell division. When stained for antinuclear antibody, localisation of fibrillarin in the fibrillar regions of the interphase control nuclei is shown in Fig. 1K. Following treatment with an extract of S. latifolius, prominent segregation of fibrillar and granular components in the nucleoli was observed (Fig. 1L). 3.4. Flow cytometry (apoptosis) Experiments were carried out in triplicate. Untreated cells showed minimal apoptosis (1.9890.38%) (SEM). Positive controls (irradiated) showed up to 279 6.9% apoptosis. Cells treated with the extract of S. latifolius for 24 h showed 14.293.4% apoptosis. A concentration study showed that cells treated with retrorsine at 3.3 mg/ml (final concentration) showed 16.996.0% apoptosis, while those treated with higher concentrations showed less apoptosis and considerable necrosis. 3.5. Flow cytometry (DNA content) 3. Results 3.1. Viability In both treated and untreated cells, there were 989 0.70% viable cells remaining at the end of the 24 h treatment period. 3.2. Morphological studies Control cells showed normal morphology after 24 h (Fig. 1A). Cells treated with S. latifolius extracts showed blebbing, hypercondensed nuclear chromatin and nuclear fragmentation with small pyknotic nuclei, loss of cytoplasm and loss of cell– cell interaction. Retrorsine, at concentrations of 330 ng/ml and above, induced similar alterations in cell morphology (Fig. 1B– E). 3.3. Indirect immunofluorescence Control cells stained for b-tubulin are shown in Fig. 1F. The effect of S. latifolius extracts on the b-tubulin components of the cytoskeleton is shown in Fig. 1G. In the unexposed cells, normal b-tubulin is dispersed evenly throughout the cell. In contrast, treated cells show abnormal cytoplasmic tubulin network and spindle-associated b-tubulin with clumping of the tubulin. Metaphase cell spindle formation appeared to be abnormal with uneven chromosome distribution. Experiments were carried out in triplicate. Results for control cells and cells treated with an extract of S. latifolius for 24 h are shown in Fig. 2. It can be seen that in the controls, 119 1.5% of diploid cells were in G2 phase compared with 2492.2% in the S. latifoliustreated cells, indicating a blockage in cell division. No tetraploid cells were present; however, the number of cells counted may not have allowed for detection of a small number of these. 4. Discussion There have been many suggestions that PAs may contribute to carcinogenesis in humans, but none of these have been proven. Chan et al. (1994) observed that karyomegaly, cytomegaly and cytoplasmic vacuolation occurred in hepatocytes treated with the PA riddelliine, but the mechanism was not fully elucidated. In this in vitro study, we have shown that extracts of the traditional remedy, S. latifolius and standard retrorsine at a concentration of 4.5 mg/l gave rise to either micronuclei, apoptosis or abnormal multinucleate cells in HuH-7 hepatoma cells over a period of 24 –72 h. At higher concentrations, the cells become necrotic. No apoptosis was observed in any of the control cultures. The formation of micronuclei containing abnormal numbers of chromosomes and/or fragmented DNA can V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 55 Fig. 1. Photomicrographs of cells after 24 h treatment. H&E stain (400 × ) (A) Control HuH-7 cells showing normal cell morphology and a dividing cell in metaphase. (B) Treated HuH-7cells showing a cell undergoing micronuclei formation (white arrowhead) containing hypercondensed chromatin (white arrow). The black arrowhead indicates cytoplasmic blebbing. (C) Treated HuH-7cells. Nuclei of different sizes can be seen (black arrowheads). In the centre, advanced late stage apoptosis with apoptotic body formation is shown (black arrow). (D) Treated HuH-7cells. Micrograph showing a single treated cell with well-formed micronuclei. (E) Treated HuH-7 cells showing a multipolar dividing cell. Indirect immunofluorescence: b-tubulin (1000 × ). (F) Indirect immunofluorescent preparation showing b-tubulin distribution in normal cells. (G) Treated HuH-7cells showing destruction of b-tubulin in the cytoplasm. DAPI stain (1000 × ). (H) Micrograph of control cells with DAPI-stained chromosomes. The white arrow indicates a dividing cell with prominent chromosomes. (I) Micrograph of cells treated with an extract of S. latifolius. Blebbing and fragmentation of the nucleus is evident. (J) Micrograph of retrorsine-treated cells (330 ng/ml). The white arrow indicates lagging chromosomes in a dividing cell. Indirect immunofluorescence: antinuclear antibody (800 × ). (K) Control showing nucleolar morphology. (L) Treated cells showing a giant cell with multiple segregated nucleoli (white arrowheads). 56 V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 be an intermediate step in the development of apoptosis (Seegers et al., 1989), but Sherwood and Schimke (1995) have pointed out that apoptotic cells may be confused with those containing micronuclei resulting from aberrant mitosis since both show decreased DNA content. The precise mechanism by which apoptosis is induced in the HuH-7 cell line by PAs is not known, since HuH-7 cells express a mutant form of the tumour suppresser gene, p53 (Hsu et al., 1993) that makes them insensitive to anticancer drug-induced apoptosis. Thus PAs must activate p53-independent apoptotic pathways in this cell line (Muller et al., 1997). The cytoplasmic protein bcl-2, a well-recognised inhibitor of apoptosis (Adams and Cory, 1998), which is thought to act by blocking free radical formation (Pourzand et al., 1997), is also not expressed in HuH-7 cells. Therefore, lack of bcl-2 in HuH-7 cells may make them more susceptible to agents that lead to intracellular formation of free radicals (Thannical and Fanburg, 1995; Huang and Chou, 1998). Pyrrole active radicals not only affect DNA, but can cause damage to proteins. There are two groups of proteins of concern in this study: those associated with the mitochondrial permeability pore, which are involved in one of the primary mechanisms of apoptosis, and tubulins, acting both as components of the nuclear spindle and as structural proteins in the cytoplasm. The flow cytometry studies with the use of APO2.7 antibody gave confirmation of the presence of apoptosis (O’Brien et al., 1995) and specifically the involvement of the mitochondria in this process. Both S. latifolius extracts and retrorsine gave rise to aberrant spindle formation in dividing cells but we found, in addition, major effects on cytoplasmic tubulin at all concentrations, with severe derangement of the intracellular architecture. This is a new finding and we postulate that the mechanism for this is binding of pyrrole radicals to the thiol groups in the tubulin, leading to derangement of the tertiary structure. The studies of DNA content confirmed a significant decrease in G0/G1-phase of the cell cycle in PA-treated cells and a two-fold increase in the proportion of cells blocked in G2 phase of the cell cycle. The G2/M block in cell cycle progression suggests that this checkpoint in the cell cycle may be the most important for DNA damage and repair. The immunofluorescent studies of nucleolar morphology showed segregation of the fibrillar and granular components, which is also indicative of DNA damage. One prerequisite for G2/M arrest is p34cdc-2 kinase activation and although this was not investigated in this study, we propose that apoptosis in the PA-treated HuH-7 cells may be p34cdc-2-dependent. The other possible explanation is that apoptosis is due to the effect of the PA metabolites on microtubulin and spindle formation. This is problematical since inhibitors of microtubulin formation such as taxol and vinblastine Fig. 2. DNA content by flow cytometry. (A) Control cells. (B) Cells treated with an extract of S. latifolius. V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 activate p53-dependent apoptotic pathways (Tishler et al., 1995), which should not be active in HuH-7 cells. Abnormal spindle formation does not invariably lead to apoptosis since failure by the cell to detect abnormal spindles or damage to DNA can lead to duplication of the abnormal DNA. In this study, we have found both multinucleate cells and apoptosis, suggesting that more than one mechanism is active. This study was designed to confirm that the known adverse effects of standard pyrrolizidines could be reproduced using an extract of a pyrrolizidine-containing Senecio species used as a traditional remedy. This is important, since other reports have indicated that there are Senecio species, which may be active as antioxidants (Hammond et al., 1998; Liu and Ng, 2000). In summary, these results illustrate that exposure of HuH-7 cells to extracts of S. latifolius or low-dose retrorsine leads to two possible changes. The most common is apoptosis; however, where this fails, polyploid and aneuploid cells containing abnormal DNA are seen. Although one cannot extrapolate directly from these in vitro studies to the in vivo situation in humans, the findings of destruction of cytoplasmic tubulin and apoptosis in a hepatocyte cell line, add additional information to the unclear picture of the mechanism of the toxicity of pyrrolizidine alkaloids in humans. Acknowledgements We wish to acknowledge the Department of Physiology, University of Pretoria, for facilities made available to us to do the indirect immunofluorescence work. We thank Mrs Pepita Bianchi of the Department of Surgery at the University of the Witwatersrand for her assistance with the flow cytometry, and Professor W. van Heerden of the Department of Oral Surgery, University of Pretoria, for the DNA flow cytometry. This work was supported by grants from the Research Endowment Fund of the University of the Witwatersrand and South African Institute for Medical Research. References Adams, J., Cory, S., 1998. The bcl-2 protein family: arbiters of cell survival. Science 281, 1322 – 1326. Birecka, H., Catalfamo, J., Eisen, R.N., 1981. A sensitive method for the detection and quantitative determination of pyrrolizidine alkaloids. Phytochemistry 20, 343 – 344. Bull, L., Culvenor, C.C.J., Dick, A.S.T., 1968. The pyrrolizidine alkaloids; their chemistry, pathogenicity and other biological properties. North Holland Publishing Co., North Holland. Chan, P., Mahler, J., Bucher, J.R., Travlos, G.S., Reid, J.B., 1994. Toxicity and carcinogenicity of riddelliine following 13 weeks of treatment to rats and mice. Toxicon 32 (8), 891 –908. 57 Cheeke, P., 1988. Toxicity and metabolism of pyrrolizidine alkaloids. J. Anim. Sci. 66 (9), 2343 – 2350. Cook, J., Duffy, E., Schoental, R., 1950. Primary liver tumours in rats following the feeding with alkaloids of Senecio jacobaea. Br. J. Cancer 4, 405 –410. Culvenor, C.C.J., 1983. Estimated intakes of pyrrolizidine alkaloids in humans: a comparison with doses causing tumours in rats. J. Toxicol. Environ. Health 11, 625 – 635. De Waal, H.L., 1941. Toxic principles of Senecio-species. Farming in S. Afr. 16, 69 –70. Deyo, J., 1991. Tier-2 studies on monocrotaline immunotoxicity in C57BL/6 mice. Toxicology 70 (3), 313 – 325. Glowaz, S., Michnicka, M., Huxtable, R., 1992. Detection of a reactive pyrrole in the hepatic metabolism of the pyrrolizidine alkaloid monocrotaline. Toxicol. Appl. Pharmacol. 115, 168 – 173. Green, C., Christie, G., 1961. Malformations in foetal rats induced by the pyrrolizidine alkaloid heliotrine. Br. J. Exp. Pathol. 42, 369 – 378. Hammond, G.B., Fernandez, I.D., Villegas, L.F., Vaisberg, A.J., 1998. A survey of traditional medicine plants from the Callejon de Huaylas Department of Ancash, Peru. J. Ethnopharmacol. 61 (1), 17 – 30. Hayashi, Y., Lalich, J., 1967. Renal and pulmonary alterations induced in rats by a single injection of monocrotaline. Proc. Soc. Exp. Biol. Med. 124, 392 – 396. Hirono, I., 1987. Pyrrolizidine alkaloids. Naturally occurring carcinogens of plant origin. In: Hirono, I. (Ed.), Toxicology, Pathology and Biochemistry, vol. 2. Elsevier, Amsterdam, pp. 25 –51. Hsu, I., Tokiwa, T., Bennett, W., Metcalf, R., Welsh, J., Sun, T., Harris, C., 1993. p53 gene mutation and integrated hepatitis B viral DNA sequences in human liver cancer cell lines. Carcinogenesis 14, 987 – 992. Huang, Y., Chou, C., 1998. Bcl-2 blocks apoptotic signal of transforming growth factor-beta in human hepatoma cells. J. Biomed. Sci. 5, 185 – 191. Hutchings, A., 1989. A survey and analysis of traditional medicinal plants as used by the Zulu, Xhosa and Sotho. Bothalia 19 (1), 111 – 123. Huxtable, R., 1979. New aspects of the pharmacology and toxicology of pyrrolizidine alkaloids. Gen. Pharmacol. 10, 159 –167. Huxtable, R., 1980. Herbal teas and toxins: novel aspects of pyrrolizidine poisoning in the United States. Perspect. Biol. Med. 24, 1 – 14. Kiernan, J., 1990. Histological and Histochemical Methods. Pergamon Press, London. Kim, H.Y., Stermitz, F.R., Molyneux, R.J., Wilson, D.W., Taylor, D., Coulombe, R.A., 1993. Structural influences on pyrrolizidine alkaloid-induced cytopathology. Toxicol. Appl. Pharmacol. 122, 61 – 69. Leonard, N., 1960. Senecio alkaloids. In: Manske, R. (Ed.), The Alkaloids, vol. 6. Academic Press, New York, pp. 37 – 121. Lin, G., Cui, Y., Hawes, E.M., 1998. Microsomal formation of a pyrrolic alcohol glutathione conjugate of clivorine. Firm evidence for the formation of a pyrrolic metabolite of an otonecine-type pyrrolizidine alkaloid. Drug Metab. Dispos. 26 (2), 181 – 184. Liu, F., Ng, T.D., 2000. Antioxidative and free radical scavenging activities of selected medicinal herbs. Life Sci. 66 (8), 725 – 735. Mattocks, A., 1968. Spectrophotometric determination of pyrrolizidine alkaloids — some improvements. Anal. Chem. 40, 1749 – 1750. Mattocks, A., Legg, R., 1980. Antimitotic action of dehydroretreonecine, a pyrrolizidine alkaloid metabolite, and some analogous compounds in a rat liver parenchymal cell line. Chem. -Biol. Interact. 30, 325 –336. Mattocks, A.R., 1986. Chemistry and Toxicology of Pyrrolizidine Alkaloids. Academic Press, Orlando, FL. 58 V. Steenkamp et al. / Journal of Ethnopharmacology 78 (2001) 51–58 McLean, E., 1970. The toxic actions of pyrrolizidine (Senecio) alkaloids. Pharmacol. Rev. 22, 429 – 483. Miranda, C.L., Reed, R.L., Guengerich, F.P., Buhler, D.R., 1991. Role of cytochrome P450IIIA4 in the metabolism of the pyrrolizidine alkaloid senecionine in human liver. Carcinogenesis 12, 515 – 519. Molyneux, R.J., Johnson, E., Olsen, J.D., Baker, D.C., 1991. Toxicity of pyrrolizidine alkaloids from Riddell groundsel (Senecio riddellii ) to cattle. Am. J. Vet. Res. 52, 146 – 151. Muller, M., Strand, S., Hug, H., Heinemann, E.M.0, Walczak, H., Hoffmann, W.J., Stremmell, W., Krammer, P.H., Galle, P.R., 1997. Drug induced apoptosis in hepatoma cells is mediated by the CD95(APO-1/Fas) receptor ligand system and involves activation of wild type p53. J. Clin. Invest. 99, 403 – 413. O’Brien, M., Healey, S., Raney, S.R., Hurst, J.M., Avner, B., Hanley, A., Mies, C., Freeman, J.W., Snow, C., Koester, S.K., Bolton, W.E., 1995. Discrimination of late apoptotic/necrotic cells (Type III) by flow cytometry in solid tumours. Cytometry 28, 81 –89. Pourzand, C., Rossier, G., Reelfs, O., Borner, C., Tyrrell, R.M., 1997. Over-expression of bcl-2 inhibits UV-mediated immediate apoptosis in rat 6 fibroblasts: evidence for the involvement of bcl-2 as an antioxidant. Cancer Res. 57, 1405 – 1411. Powis, G., Wincentsen, L., 1980. Pyridine nucleotide cofactor requirements of indicine N-oxide reduction by hepatic microsomal cytochrome P-450. Biochem. Pharmacol. 29, 347 – 351. Reddy, J., Harris, C., Svoboda, D., 1968. Inhibition by lasiocarpine of RNA synthesis, RNA polymerase and induction of tryptophan pyrrolase activity. Nature 217, 659 – 661. Roulet, M., Laurini, R., River, L., Calame, A., 1988. Hepatic venoocclusive disease in a newborn infant of a woman drinking herbal tea. J. Pediatr. 112, 433 –436. Schoental, R., 1955. Kwashiorkor-like syndrome and other pathological changes in rats as a result of feeding with Senecio alkaloids (Isatidine). Voeding 16, 268 – 285. Schoental, R., 1968. Toxicity and carcinogenic action of pyrrolizidine alkaloids. Cancer Res. 28, 2237 – 2246. Seegers, J., Aveling, M.-L., van Aswegen, C.H., Cross, M., Koch, F., Joubert, W.S., 1989. The cytotoxic effects of estradiol-17b, catecholestradiols and methoxyestradiols on dividing MCF-7 and HeLa cells. Steroid Biochem. 32, 797 – 809. Sherwood, S., Schimke, R., 1995. Cell cycle analysis of apoptosis using flow cytometry. Methods Cell Biol. 46, 77 – 97. Svoboda, D., Reddy, J., 1972. Malignant tumors in rats given lasiocarpine. Cancer Res. 32, 908 –913. Thannical, V., Fanburg, B., 1995. Activation of an H2O2-generating NADH oxidase in human lung fibroblasts by transforming growth factor beta. J. Biol. Chem. 270, 3034 – 3038. Tishler, R., Lamppu, D., Park, S., Price, B., 1995. Microtubule-active drugs taxol, vinblastine and nocodazole increase the levels of transcriptionally active p53. Cancer Res. 55, 6021 – 6025. van Wyk, B.-E., Verdoorn, G., Schutte, A., 1992. Distribution and taxonomic significance of major alkaloids in the genus Podalyria. Biochem. Syst. Ecol. 20 (2), 163 – 172. Ware, C., 1985. Protocol for a colorimetric assay to determine cell viability using a Biorad microtitration plate reader. BIORAD Bull. 1203, 1 –4. Warren, F., 1978. The pyrrolizidine alkaloids. In: Manske, R. (Ed.), The Alkaloids, vol. 12. Academic Press, New York, pp. 246 – 324. Watt, J., Breyer-Brandwijk, M., 1962. The Medicinal and Poisonous Plants of Southern Africa. Livingstone, London. Weston, C.F.M., Cooper, B.T., Davies, J.D., Levine, D.F., 1987. Veno-occlusive disease of the liver secondary to the ingestion of Comfrey. Br. Med. J. 295, 183. Williams, D.E., Reed, R.L., Kedzierski, B., Dannan, G.A., Guengerich, F.P., Buhler, D.R., 1989. Bioactivation and detoxification of the pyrrolizidine alkaloid senecionine by cytochrome P-450 enzymes in rat liver. Drug Metab. Dispos. 17, 387 – 392. Williams, G., Mori, H., Hirono, I., Nagao, M., 1980. Genotoxicity of pyrrolizidine alkaloids in the hepatocyte primary culture/DNA repair test. Mutat. Res. 79, 1 – 5. Willmot, F.C., Robertson, G.W., 1920. Senecio disease or cirrhosis of the liver due to Senecio poisoning. Lancet 2, 848 –849. Yamanaka, H., Nagao, M., Sugimura, T., Furuya, T., Shirai, A., Matsushima, T., 1979. Mutagenicity of pyrrolizidine alkaloids in the salmonella/mammalian microsome test. Mutat. Res. 68, 211 – 216. Yan, C., Huxtable, R., 1994. Quantitation of the hepatic release of metabolites of the pyrrolizidine alkaloid monocrotaline. Toxicol. Appl. Pharmacol. 127 (1), 58 – 63. Yan, C., Huxtable, R., 1995. Relationship between glutathione concentration and metabolism of the pyrrolizidine alkaloid, monocrotaline, in the isolated perfused liver. Toxicol. Appl. Pharmacol. 130 (1), 132 – 139.