Reactive Pyrrolic Metabolites Research Articles

Structure‐dependent relative toxic potencies of selected pyrrolizidine alkaloids

Pyrrolizidine alkaloids are naturally occurring secondary plant metabolites mainly found in plant families of Asteraceae, Boraginaceae, and Fabaceae. Chemically, Pas consist of a pyrrolizidine core bearing hydroxyl groups, the so‐called necine base, and mono‐ or dicarboxylic necine acids bound to the pyrrolizidine core via ester linkages. 1,2‐unsaturated PAs are hepatotoxic, genotoxic, and carcinogenic due to the highly reactive pyrrolic metabolites formed by cytochrome P450 monooxygenases (CYPs) primarily in the liver. The presence of PAs as frequent contaminants in the wide variety of food and feed products has to be considered a relevant safety issue.Based on the currently available data, the risk assessment of PAs was mainly approached using the two most toxic potent congeners, i.e., lasiocarpine and riddelliine. However, it is well recognized that toxicity is differing significantly between the congeners related to their structural features. The risk of PA‐containing products is indeed overestimated, and a comprehensive risk assessment should take these differences into account.After analyzing the data of many PAs, Merz and Schrenk derived interim Relative Potency (iREP) factors to present the differences in their toxicity between the sub‐groups of PA congeners concerning their structural features. The use of such iREP factors could probably provide a more scientific basis for PA risk assessment until sufficient experimental analysis of the toxicities of individual congeners is applied. To obtain a better understanding of the relationship between structure and toxicity of PA congeners and provide more evidence for further refinement of relative (toxic) potency factors, data of the in vitro cytotoxicity, genotoxicity, and mutagenicity of diverse individual PA congeners (lasiocarpine, monocrotalineφ, retrorsine, senecionine, seneciphyllineφ, echimidineφ, europineφ, heliotrineφ, indicine, and lycopsamine) has been generated in our project supported by Kooperation Phytopharmaka. Among them, lasiocarpine, retrorsine, senecionine, indicine and lycopsamine have been investigated on my part.Cytotoxicity was assessed using the Alamar blue assay in primary rat hepatocytes, HepG2 cells, and the HepG2 (CYP3A4) cell line. In HepG2 cells, none of the selected PAs exhibited cytotoxic effects, probably due to the lack of CYPs. In primary rat hepatocytes as well as in HepG2(CYP3A4) cells, a clear structure dependent cytotoxicity could be demonstrated. The role of CYP450 enzymes in metabolic activation was further confirmed using an inhibition assay. A kinetic assay analyzing 7‐benzyloxyresorufin‐O‐ dealkylation (BROD) was used for measuring the activity of CYP450 enzymes. Furthermore, the utilization of a glutathione‐reductase‐DTNB recycling assay indicated that glutathione might not play a critical role in PA‐induced cytotoxicity. A micronucleus test was used for determining the PA‐induced clastogenic genotoxicity. All selected PA congeners exhibited concentration‐dependent toxicity in the HepG2 (CYP3A4) cells. The relative potencies of PA congeners estimated from Alamar blue assay and micronucleus assay are generally consistent with the following ranking: lasiocarpine > senecionine > seneciphylline > retrorsine > heliotrine (?) echimidine > europine = indicine = lycopsamine = monocrotaline. The relative toxic potencies evaluated based on our findings were not completely consistent with the iREP classification previously reported by Merz and Schrenk. Monocrotaline in both assays exhibited considerably lower toxic potency. Echimidine, however, was more toxic than expected. On the other hand, mutagenicity was measured in Ames fluctuation assay with Salmonella typhimurium strains TA98 and TA100. None of the selected PA congeners up to 300 μM showed mutagenic effects despite metabolic activation with S9‐mix.

Correlation Investigation between Pyrrole-DNA and Pyrrole-Protein Adducts in Male ICR Mice Exposed to Retrorsine, a Hepatotoxic Pyrrolizidine Alkaloid

Pyrrolizidine alkaloids (PAs) have been found in over 6000 plants worldwide and represent the most common hepatotoxic phytotoxins. Catalyzed by hepatic cytochrome P450 enzymes, PAs are metabolized into reactive pyrrolic metabolites, which can alkylate cellular proteins and DNA to form pyrrole-protein adducts and pyrrole-DNA adducts, leading to cytotoxicity, genotoxicity, and tumorigenicity. To date, the correlation between these PA-derived pyrrole-protein and pyrrole-DNA adducts has not been well investigated. Retrorsine is a representative hepatotoxic and carcinogenic PA. In the present study, the correlations among the PA-derived liver DNA adducts, liver protein adducts, and serum protein adducts in retrorsine-treated mice under different dosage regimens were studied. The results showed positive correlations among these adducts, in which serum pyrrole-protein adducts were more accessible and present in higher abundance, and thus could be used as a suitable surrogate biomarker for pyrrole-DNA adducts to indicate the genetic or carcinogenic risk posed by retrorsine.

Fasting augments pyrrolizidine alkaloid-induced hepatotoxicity.

Pyrrolizidine alkaloids (PAs) are phytotoxins widely present in various natural products and foodstuffs. The present study aims to investigate the effects of fasting on PA-induced hepatotoxicity and the underlying biochemical mechanisms. The results of hepatotoxic study showed that 15-h overnight fasting significantly exacerbated the hepatotoxicity of retrorsine (RTS, a representative toxic PA) in fasted rats compared to fed rats, as indicated by remarkably elevated plasma ALT and bilirubin levels and obvious liver histological changes. Further toxicokinetic studies revealed that fasting significantly enhanced cytochromes P450 enzymes (CYPs)-mediated metabolic activation of RTS leading to increased formation of pyrrole-protein adducts and thus decreased the in vivo exposure and excretion of both parent RTS and its N-oxide metabolite. Metabolic studies demonstrated that fasting induced enzyme activities of CYP1A2, CYP2B6 and CYP2E1 that participated in catalyzing RTS to its reactive pyrrolic metabolites. Moreover, fasting also dramatically decreased hepatic glutathione (GSH) content, which restricted the detoxification of GSH by neutralizing the reactive pyrrolic metabolite of RTS, further contributing to the enhanced hepatotoxicity. The present findings may have an impact on future PA toxicity tests with different dietary styles and/or risk assessment of metabolite-mediated toxins by considering the profound effects of fasting.

Pyrrolizidine Alkaloid-Induced Hepatotoxicity Associated with the Formation of Reactive Metabolite-Derived Pyrrole-Protein Adducts.

Pyrrolizidine alkaloids (PAs) with 1,2-unsaturated necine base are hepatotoxic phytotoxins. Acute PA intoxication is initiated by the formation of adducts between PA-derived reactive pyrrolic metabolites with cellular proteins. The present study aimed to investigate the correlation between the formation of hepatic pyrrole–protein adducts and occurrence of PA-induced liver injury (PA-ILI), and to further explore the use of such adducts for rapidly screening the hepatotoxic potency of natural products which contain PAs. Aqueous extracts of Crotalaria sessiliflora (containing one PA: monocrotaline) and Gynura japonica (containing two PAs: senecionine and seneciphylline) were orally administered to rats at different doses for 24 h to investigate PA-ILI. Serum alanine aminotransferase (ALT) activity, hepatic glutathione (GSH) level, and liver histological changes of the treated rats were evaluated to assess the severity of PA-ILI. The levels of pyrrole–protein adducts formed in the rats’ livers were determined by a well-established spectrophotometric method. The biological and histological results showed a dose-dependent hepatotoxicity with significantly different toxic severity among groups of rats treated with herbal extracts containing different PAs. Both serum ALT activity and the amount of hepatic pyrrole–protein adducts increased in a dose-dependent manner. Moreover, the elevation of ALT activity correlated well with the formation of hepatic pyrrole–protein adducts, regardless of the structures of different PAs. The findings revealed that the formation of hepatic pyrrole–protein adducts—which directly correlated with the elevation of serum ALT activity—was a common insult leading to PA-ILI, suggesting a potential for using pyrrole–protein adducts to screen hepatotoxicity and rank PA-containing natural products, which generally contain multiple PAs with different structures.

Developing urinary pyrrole-amino acid adducts as non-invasive biomarkers for identifying pyrrolizidine alkaloids-induced liver injury in human.

Pyrrolizidine alkaloids (PAs) have been found in over 6000 plants worldwide and represent the most common hepatotoxic phytotoxins. Currently, a definitive diagnostic method for PA-induced liver injury (PA-ILI) is lacking. In the present study, using a newly developed analytical method, we identified four pyrrole-amino acid adducts (PAAAs), namely pyrrole-7-cysteine, pyrrole-9-cysteine, pyrrole-9-histidine, and pyrrole-7-acetylcysteine, which are generated from reactive pyrrolic metabolites of PAs, in the urine of PA-treated male Sprague Dawley rats and PA-ILI patients. The elimination profiles, abundance, and persistence of PAAAs were systematically investigated first in PA-treated rat models via oral administration of retrorsine at a single dose of 40 mg/kg and multiple doses of 5 mg/kg/day for 14 consecutive days, confirming that these urinary excreted PAAAs were derived specifically from PA exposure. Moreover, we determined that these PAAAs were detected in ~ 82% (129/158) of urine samples collected from ~ 91% (58/64) of PA-ILI patients with pyrrole-7-cysteine and pyrrole-9-histidine detectable in urine samples collected at 3 months or longer times after hospital admission, indicating adequate persistence time for use as a clinical test. As direct evidence of PA exposure, we propose that PAAAs can be used as a biomarker of PA exposure and the measurement of urinary PAAAs could be used as a non-invasive test assisting the definitive diagnosis of PA-ILI in patients.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00204-021-03129-6.

Quantitation of DNA reactive pyrrolic metabolites of senecionine – A carcinogenic pyrrolizidine alkaloid by LC/MS/MS analysis

High hydrostatic pressure processing (HPP) is a non-thermal pasteurization technology which has already been applied in the food industries. Besides maintaining the food safety and quality, HPP also has potential applications in the enhancement of the health benefits of food products. This study examines the current progress of research on the use of HPP in the development of health foods. Through HPP, the nutritional value of food products can be enhanced or retained, including promotes the biosynthesis of γ-aminobutyric acid (GABA) in the food materials, retains immunoglobulin components in dairy products, increases resistant starch content in cereals, and reduces the glycemic index of fruit and vegetable products, which facilitates better control of blood glucose levels and decreases calorie intake. HPP can also be utilized as a hurdle technology in combination with existing processing technologies for the development of low-sodium food products and the maintenance of microbial safety, thereby lowering the risk of triggering cardiovascular disease. Additionally, HPP can be used to enhance the diversity of probiotic food products. Appropriate sporogenous probiotics can be screened and added to various high-pressure processed food products as a certain bacterial count is still retained in the products after HPP. As HPP causes physical damage to the structures of food products, it can also be used as a synergistic extraction technology to enhance the extraction efficiency of functional components, thereby reducing extraction time. By applying HPP in the extraction of functional components from food waste, the production costs of such components can be effectively reduced. This study provides a summary of the mechanisms by which HPP enhances the health benefits of food products and the current progress of relevant research. HPP possesses huge potential in the development of novel health foods and may provide an abundance of benefits to human health in the future. © 2019

Metabolic Activation of Pyrrolizidine Alkaloids – Generation of Multiple Reactive Metabolites Leading to DNA Adduct Formation and Potentially Liver Tumor Initiation

Pyrrolizidine alkaloids (PAs) are toxic and carcinogenic phytochemicals widespread in the world. Metabolic formation of dehydro‐PAs and 6,7‐dihydro‐7‐hydroxy‐1‐hydroxymethyl‐5H‐pyrrolizine (DHP) has long been considered the principle metabolic activation pathway. We determined that a set of DHP‐DNA adducts are generally formed from the metabolism of a series of carcinogenic PAs in vitro, in cells and in vivo. The level of DHP‐DNA adducts well correlated with liver tumor potency for a set of PAs. We have proposed that this set of DHP‐DNA adducts is a potential biomarker of PA exposure and PA‐induced liver tumor initiation. Our present studies have found that, in addition to dehydro‐PAs and DHP, there are another seven reactive pyrrolic metabolites that are capable of binding to cellular DNA to produce the same set of DHP‐DNA adducts. These DNA reactive metabolites are: 7‐glutathione‐DHP, 7‐cysteine‐DHP, 7‐N‐acetylcysteine‐ DHP, 1‐CHO‐DHP, cysteinyl‐[2′‐S‐7]‐1‐CHO‐DHP, cysteinyl‐[3′‐N‐7]‐1‐CHO‐DHP, and 1‐cysteinylimino‐DHP. These pyrrolic metabolites are enzymatically and/or chemically formed from the reaction of the corresponding amino acids (and glutathione) with either dehydro‐PAs or DHP. Also, these pyrrolic metabolites are interconvertible, apparently ascribed to their unique reactive bifunctional alkylating character. These overall results represent the first time that concerning the metabolic activation of PAs: (i) there involve the formation of multiple DNA reactive pyrrolic metabolites; and (ii) metabolism of carcinogenic and potential carcinogenic PAs all generate the identical DHP‐DNA adducts. [This abstract is not an official U.S. Food and Drug Administration (FDA) guidance or policy statement. No official support or endorsement by the U.S. FDA is intended or should be inferred. The authors declare no competing financial interest.]

Quantitation of DNA reactive pyrrolic metabolites of senecionine – A carcinogenic pyrrolizidine alkaloid by LC/MS/MS analysis

Pyrrolizidine alkaloids (PAs) are carcinogenic phytochemicals, inducing liver tumors in experimental rodents. We previously determined that (±)-6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHP), 7-glutathione-DHP, 7-cysteine-DHP, 7-N-acetylcysteine-DHP, and 1-CHO-DHP are DNA reactive pyrrolic metabolites potentially associated with PA-induced liver tumor initiation. In this study, we developed an LC/MS/MS multiple reaction monitoring (MRM) mode method to identify and quantify these metabolites formed from the metabolism of senecionine, a carcinogenic PA, by mouse, rat, and human liver microsomes, and primary rat hepatocytes. Together with the chemically prepared standards of these metabolites, this represents an accurate and convenient LC/MS/MS analytical method for quantifying these five reactive pyrrolic metabolites in biological systems.

Development of a two-layer transwell co-culture model for the in vitro investigation of pyrrolizidine alkaloid-induced hepatic sinusoidal damage

Pyrrolizidine alkaloids (PAs) are hepatotoxic and specifically damage hepatic sinusoidal endothelial cells (HSECs) via cytochrome P450 enzymes (CYPs)-mediated metabolic activation. Due to the lack of CYPs in HSECs, currently there is no suitable cell model for investigating PA-induced HSEC injury. This study aimed to establish a two-layer transwell co-culture model that mimics hepatic environment by including HepaRG hepatocytes and HSECs to evaluate cytotoxicity of PAs on their major target HSECs. In this model, PAs were metabolically activated by CYPs in HepaRG hepatocytes to generate reactive pyrrolic metabolites, which react with co-cultured HSECs leading to HSEC damage. Three representative PAs, namely retrorsine, monocrotaline, and clivorine, induced significant concentration-dependent cytotoxicity in HSECs in the co-culture model, but did no cause obvious cytotoxicity directly in HSECs. Using the developed co-cultured model, further mechanism studies of retrorsine-induced HSEC damage demonstrated that the reactive pyrrolic metabolite generated by CYP-mediated bioactivation in HepaRG hepatocytes caused formation of pyrrole-protein adducts, reduction of GSH content, and generation of reactive oxygen species in HSECs, leading to cell apoptosis. The established co-culture model is reliable and applicable for cytotoxic assessment of PA-induced HSEC damage and offers a novel platform for screening toxicity of different PAs on their target cells.

Primary and secondary pyrrolic metabolites of pyrrolizidine alkaloids form DNA adducts in human A549 cells

Humans and animals can be exposed to carcinogenic pyrrolizidine alkaloids (PAs) through consumption of plants commonly found in many parts of the world. Although the liver is the primary target organ for carcinogenic PAs, they have also induced lung tumors in rodents. Hepatic cytochrome P450 activity converts PAs into dehydro-PAs that can be hydrolyzed to dehydropyrrolizidine (DHP); these reactive pyrrolic metabolites can produce four characteristic DNA adducts associated with PA-induced liver tumor initiation in laboratory animals. We reported recently that these four DNA adducts are also formed when 7-glutathione-DHP (7-GS-DHP) or 7-cysteine-DHP is incubated with calf thymus DNA. Here we showed that the four characteristic DNA adducts were formed when human A549 brochoalveolar carcinoma cells were treated with three dehydro-PAs (dehydroriddelliine, dehydromonocrotaline, or dehydroretronecine) or with 7-GS-DHP or 7-cysteine-DHP. For comparison, two parent PAs (riddelliine and monocrotaline) and 7,9-di-glutathionine-DHP were studied. No DHP-DNA adducts were detected with these incubations, confirming that A549 lung carcinoma cells do not express cytochrome P450 enzymes required for metabolic activation of PAs. Our results show that primary and secondary pyrrolic metabolites of carcinogenic PAs produce characteristic DHP-containing DNA adducts in A549 lung cancer cells, suggesting that they are DNA reactive metabolites.

The role of formation of pyrrole-ATP synthase subunit beta adduct in pyrrolizidine alkaloid-induced hepatotoxicity.

Pyrrolizidine alkaloids (PAs) are one of the most significant groups of hepatotoxic phytotoxins. It is well-studied that metabolic activation of PAs generates reactive pyrrolic metabolites that rapidly bind to cellular proteins to form pyrrole-protein adducts leading to hepatotoxicity. Pyrrole-protein adducts all contain an identical core pyrrole moiety regardless of structures of the different PAs; however, the proteins forming pyrrole-protein adducts are largely unknown. The present study revealed that ATP synthase subunit beta (ATP5B), a critical subunit of mitochondrial ATP synthase, was a protein bound to the reactive pyrrolic metabolites forming pyrrole-ATP5B adduct. Using both anti-ATP5B antibody and our prepared anti-pyrrole-protein antibody, pyrrole-ATP5B adduct was identified in the liver of rats, hepatic sinusoidal endothelial cells, and HepaRG hepatocytes treated with retrorsine, a well-studied representative hepatotoxic PA. HepaRG cells were then used to further explore the consequence of pyrrole-ATP5B adduct formation. After treatment with retrorsine, significant amounts of pyrrole-ATP5B adduct were formed in HepaRG cells, resulting in remarkably reduced ATP synthase activity and intracellular ATP level. Subsequently, mitochondrial membrane potential and respiration were reduced, leading to mitochondria-mediated apoptotic cell death. Moreover, pre-treatment of HepaRG cells with a mitochondrial membrane permeability transition pore inhibitor significantly reduced retrorsine-induced toxicity, further revealing that mitochondrial dysfunction caused by pyrrole-ATP5B adduct formation significantly contributed to PA intoxication. Our findings for the first time identified ATP5B as a protein covalently bound to the reactive pyrrolic metabolites of PAs to form pyrrole-ATP5B adduct, which impairs mitochondrial function and significantly contributes to PA-induced hepatotoxicity.

Metabolic Activation of Pyrrolizidine Alkaloids Leading to Phototoxicity and Photogenotoxicity in Human HaCaT Keratinocytes

Pyrrolizidine alkaloids, produced by a large number of poisonous plants with wide global distribution, are associated with genotoxicity, tumorigenicity, and hepatotoxicity in animals and humans. Mammalian metabolism converts pyrrolizidine alkaloids to reactive pyrrolic metabolites (dehydropyrrolizidine alkaloids) that form covalent protein and DNA adducts. Although a mechanistic understanding is currently unclear, pyrrolizidine alkaloids can cause secondary (hepatogenous) photosensitization and induce skin cancer. In this study, the phototoxicity of monocrotaline, riddelliine, dehydromonocrotaline, dehydroriddelliine, and dehydroretronecine (DHR) in human HaCaT keratinocytes under ultraviolet A (UVA) irradiation was determined. UVA irradiation of HaCaT cells treated with dehydromonocrotaline, dehydroriddelline, and DHR resulted in increased release of lactate dehydrogenase and enhanced photocytotoxicity proportional to the UVA doses. UVA-induced photochemical DNA damage also increased proportionally with dehydromonocrotaline and dehydroriddelline. UVA treatment potentiated the formation of 8-hydroxy-2′-deoxyguanosine DNA adducts induced by dehydromonocrotaline in HaCaT skin keratinocytes. Using electron spin resistance trapping, we found that UVA irradiation of dehydromonocrotaline and dehydroriddelliine generates reactive oxygen species (ROS), including hydroxyl radical, singlet oxygen, and superoxide, and electron transfer reactions, indicating that cytotoxicity and genotoxicity of these compounds could be mediated by ROS. Our results suggest that dehydropyrrolizidine alkaloids formed or delivered to the skin cause pyrrolizidine alkaloid-induced secondary photosensitization and possible skin cancer.

Hepatotoxicity and Tumorigenicity Induced by Metabolic Activation of Pyrrolizidine Alkaloids in Herbs

In the recent decades, the use of herbal products has been rapidly growing in the Western countries. While their use in many cases causes adverse effects, to date, safety issues of herbal products have not been adequately addressed. It is rarely determined whether the non-purported bioactive constituents in the herbs and the metabolites of the bioactive components can lead to adverse effects. In this review, we discuss, using pyrrolizidine alkaloids (PAs) as an example, the hepatotoxicity and tumorigenicity induced by metabolic activation of herbal components and by herb-herb and herb-drug interactions with other herbal ingredients and synthetic drugs. PAs are constitutively produced by plants as the secondary metabolites. There are more than 600 PAs and PA N-oxides identified in over 6000 plants, and more than half of them exhibit hepatotoxicity. Toxic PA-containing plants grow in many geographical regions worldwide, rendering it highly possible that PA-containing plants are the most common poisonous plants affecting livestock and humans. PAs require metabolic activation mediated by cytochrome P450 enzymes to generate reactive pyrrolic metabolites that react with cellular proteins and DNA leading to hepatotoxicity and genotoxicity. PAs can also modulate both phase I and phase II metabolizing enzymes, which may alter the metabolic fate of endogenous and exogenous chemicals. Alteration and/or competition of the metabolizing enzymes by PAs upon the co-administered herbal medicines or drugs can potentially result in serious clinical and toxicological consequences through decreased pharmacological activities or increased toxic effects.

The Comparative Metabolism of the 4 Pyrrolizidine Alkaloids, Seneciphylline, Retrorsine, Monocrotaline, and Trichodesmine in the Isolated, Perfused Rat Liver

Despite their similarity in structure, pyrrolizidine alkaloids (PAs) vary in their LD50s and in the organs in which toxicity is expressed. We have examined whether there are differences in the metabolism of certain PAs that are associated with these quantitative and qualitative differences in toxicity. Isolated rat livers were perfused with one of four PAs (seneciphylline, retrorsine, monocrotaline, and trichodesmine) at 0.5 mM for 1 hr, and the pyrrolic metabolites determined that were released into perfusate and bile or bound in the liver. The proportion of the PA removed by the liver varied from 93% for retrorsine to 55% for trichodesmine. However, trichodesmine-perfused livers released the greatest amount of the dehydroalkaloid into the perfusate. These reactive pyrrolic metabolites appear to be largely responsible for the toxicity of PAs. Over the course of a 1-hr perfusion, dehydroalkaloid release varied fourfold among the PAs examined. Seneciphylline and retrorsine significantly increased bile flow. Highest concentrations of PAs in bile were achieved at 30-40 min perfusion. Conversion of dehydroalkaloid to the conjugate 7-glutathionyl-6,7-dihydro-1-hydroxymethyl-5H-pyrrolizine (GSDHP) is a detoxification reaction. GSDHP release into bile varied from 80 nmol/g liver for trichodesmine to 880 nmol/g for retrorsine. Release of the less toxic hydrolytic product of dehydroalkaloids, 6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine, was also determined. Bound pyrroles in liver are probably an indication of heptatoxicity. At the end of perfusion these varied from 55 nmol/g for monocrotaline to 195 nmol/g for retrorsine. The chemical form of the bound pyrroles is a 7-thioether conjugate of 6,7-dihydro-1-hydroxymethyl-5H-pyrrolizine. No 7,9-dithio conjugate was detected, indicating that only monoalkylation has been found. These differences in metabolic pattern reflect differences in reactivity of the initially formed dehydroalkaloid and can account for the toxicological differences between the parent PAs.

Recovery of the pyrrolic nucleus of pyrrolizidine alkaloid metabolites from sulphur conjugates in tissues and body fluids

Reactive pyrrolic metabolites formed when pyrrolizidine alkaloids are given to rats can alkylate soluble and tissue-bound thiol groups. Pyrrolic thioethers thus formed are relatively stable, and may persist in tissues for long periods. A simple procedure has been developed for recovering the nucleus of the pyrrolic metabolite from such S-binding, whether in solution or attached to solid tissues, in an easily identifiable form. The thioether bond was broken by silver nitrate and the pyrrolic moiety allowed to react with ethanol or methanol to form an ethoxy or methoxy derivative. The chemical basis of the procedure was established by model experiments on a preparative scale, but for small scale recovery from tissues, pyrrolic ethers were extracted and identified by TLC, HPLC, capillary GC and mass spectrometry. Because the pyrrolic derivatives thus formed were easily recognised and unrelated to any physiological compound, the recovery method described, especially when applied to blood samples, provided a way to monitor animals for previous exposure to toxic pyrrolizidine alkaloids.