Transfer and Metabolism of Pyrrolizidine Alkaloids 

from Jacobaea vulgaris Gaertn. in Ruminants 

 

 

submitted as a cumulative doctoral dissertation in fulfilment of the requirements  

for the degree of doctor rerum naturalium 

 

 

to the Faculty of Biology and Chemistry 

 

 

 Justus Liebig University Giessen 

 

 

 

presented by 

Julian Tänzer 

 

 

 

Berlin, october 2024 

 

 

 





Declaration of Authenticity  

I hereby declare that I have prepared the submitted thesis independently and without unauthorized 

external assistance, using only the aids mentioned in the thesis. All passages that are quoted verbatim or 

in meaning from published works, as well as all information based on oral statements, are clearly 

identified as such. In the investigations I conducted and mentioned in the thesis, I adhered to the 

principles of good scientific practice as laid out in the "Satzung der Justus-Liebig-Universität zur 

Sicherung guter wissenschaftlicher Praxis". In accordance with § 22 para. 2 of the general regulations 

for modularized study programs, I consent to a review of the thesis using plagiarism detection software. 

 

In preparing this work, ChatGPT (https://open.ai/chatgpt/) was used as a tool to stylistically improve 

existing sentences. 

 

 

                                                                                                                                                                                   

Berlin,                                                                                                                                   Julian Tänzer 

 

 

https://open.ai/chatgpt/


This work was conducted from november 2019 to august 2024 in the unit "Plant- and Mycotoxins" 

within the department "Safety in the Food Chain" at the German Federal Institute for Risk Assessment 

in Berlin. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1st Referee: 

Prof. Dr. Gerd Hamscher 

Institute of Food Chemistry and Food Biotechnology, Justus Liebig University, Giessen, Germany 

2nd Referee: 

PD Dr. Robert Pieper 

Department Safety in the Food Chain, German Federal Institute for Risk Assessment, Berlin, Germany 

 



   

 





   I 

Table of Contents 

List of Figures ........................................................................................................................................ II 

List of Tables .......................................................................................................................................... II 

List of Abbreviations ........................................................................................................................... III 

1. Abstract ......................................................................................................................................... V 

2. Zusammenfassung ....................................................................................................................... VI 

3. Introduction ................................................................................................................................... 1 
3.1. The PA-SAFE-FEED project ................................................................................................... 1 

3.1.1. Background .................................................................................................................... 1 
3.1.2. Project objective ............................................................................................................. 1 

3.2. Pyrrolizidine alkaloids ............................................................................................................ 2 
3.2.1. Structure ......................................................................................................................... 2 
3.2.2. Pyrrolizidine alkaloids from Jacobaea vulgaris Gaertn. ............................................... 3 

3.3. Metabolism and Toxicology .................................................................................................... 4 
3.3.1. Absorption, distribution and excretion .......................................................................... 4 
3.3.2. Ruminal metabolism ...................................................................................................... 4 
3.3.3. Liver metabolism ........................................................................................................... 6 
3.3.4. Toxicology for ruminants ............................................................................................... 7 
3.3.5. Toxicology for humans .................................................................................................. 8 

3.4. Exposure ................................................................................................................................. 8 
3.4.1. Exposure of ruminants to Jacobaea vulgaris Gaertn. .................................................... 8 
3.4.2. Exposure of humans ....................................................................................................... 9 

3.5. Feeding studies and carry-over ............................................................................................ 10 
3.5.1. Feeding studies ............................................................................................................. 10 
3.5.2. Carry-Over ................................................................................................................... 15 

3.6. Objective ............................................................................................................................... 16 

4. Summarizing discussion .............................................................................................................. 18 

5. References .................................................................................................................................... 25 

6. Cumulative part ........................................................................................................................... 39 
6.1. 1. Publication ........................................................................................................................ 39 
6.2. 2. Publication ........................................................................................................................ 60 
6.3. Other journal publications with the participation of Julian Tänzer ..................................... 86 
6.4. Conference contributions ...................................................................................................... 86 

 
  



   II 

List of Figures 
Figure 1: Pyrrolizidine ring system: basic structure of pyrrolizidine alkaloids. ...................................... 2 

Figure 2: Classification of pyrrolizidine alkaloids based on various characteristics such as the oxidation 

status of the nitrogen (red), the degree of esterification (yellow), and the structure and stereochemistry 

of the necine base (blue). ......................................................................................................................... 3 

Figure 3: Chemical structures from PA/PANOs in Jacobaea vulgaris Gaertn. ....................................... 4 
 

List of Tables 
Table 1: Feeding studies with cattle using Jacobaea vulgaris and Senecio species. ............................. 12 

Table 2: Feeding studies with sheep and goats using Jacobaea vulgaris Gaertn. and Senecio species. 14 

 

  



   III 

List of Abbreviations 
ADI Acceptable daily intake 

Adm. Administration 

BMDL Benchmark dose level 

bw Body weight 

DHP Dehydropyrrole 

DNA Deoxyribonucleic acid 

EFSA European Food Safety Authority 

HRMS High resolution mass spectrometry 

JKK Jacobaea vulgaris Gartn. 

MOE Margin of exposure 

ND No data 

NMR Nuclear magnetic resonance spectrometry 

PA/PANO Pyrrolizidine alkaloid (free base / N-oxide) 

PA Free base pyrrolizidine alkaloid 

PANO N-oxide pyrrolizidine alkaloid 

PA-SAFE-FEED Research project "Studies on the transfer of pyrrolizidine alkaloids into 

livestock"  

UHPLC Ultra high performance liquid chromatography 

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  



   IV 

  



   V 

1. Abstract 
Pyrrolizidine alkaloids (PA/PANOs) are compounds found in over 6000 plant species, including 

Jacobaea vulgaris Gaertn. (common ragwort). Upon hepatic activation, these protoxins exhibit 

hepatotoxic and carcinogenic properties. This study aimed to investigate the ruminal metabolism and 

tissue transfer of PA/PANOs found in Jacobaea vulgaris Gaertn. in ruminants to better assess the uptake 

of PA/PANOs by grazing livestock. 

This study investigated the ruminal metabolism of cyclic diesters. The nine major free base pyrrolizidine 

alkaloids (PA) and corresponding N-oxides (PANO) from Jacobaea vulgaris Gaertn. were examined in 

vitro with rumen liquid from cattle. The results confirmed that all PANOs were rapidly reduced to the 

corresponding PAs, and most PAs were swiftly metabolized. Compounds such as jacobine, jaconine, 

and senkirkine exhibited slow elimination, while jacoline remained almost stable. For the first time, it 

was shown that cyclic diesters are reduced in the rumen to 1,2-saturated metabolites. This reduction is 

crucial, as the 1,2-double bond is structurally necessary for the toxification of PAs in the liver.  

Analysis of ruminal fluids from in vivo feeding studies with cattle, sheep, and goats confirmed the in 

vitro results. Samples from these in vivo feeding studies also revealed a low transfer of PA/PANOs into 

the muscle tissue of the animals, with mostly jacoline, jacobine, and jaconine being detectable, 

consistent with the elimination rate observed in vitro in the rumen. Therefore, the risk of exposure to 

PA/PANOs through the consumption of meat from ruminants exposed to doses of PA/PANO similar to 

those in this study appears to be low. Moreover, the reported average levels of PA/PANO contamination 

in feed used in Europe are significantly lower than the doses used in this study. 

Interestingly, the metabolites identified in the rumen were not detected in the muscle tissue. A huge 

amount of these metabolites was detected in the feces of the animals, suggesting that these metabolites 

may not pass the intestinal barrier or are further transformed in other processes in the body of the animals. 

This study demonstrates that the rumen plays a crucial role in detoxifying PA/PANOs. This filtering 

function substantially reduces the uptake of PA/PANOs within the animal’s body, presumably decreasing 

the health threat posed by these substances to ruminants. Additionally, ruminal activities result in a low 

transfer of PA/PANOs into animal products such as meat, thereby lowering the risk of exposure through 

meat consumption. 

  



   VI 

2. Zusammenfassung 
Pyrrolizidinalkaloide (PA/PANOs) sind Verbindungen, die in über 6000 Pflanzenarten vorkommen, 

darunter Jacobaea vulgaris Gaertn. (Gemeines Jakobskreuzkraut). Nach hepatischer Aktivierung zeigen 

diese Protoxine hepatotoxische und kanzerogene Eigenschaften. Ziel dieser Studie war es, den 

Pansenmetabolismus und den Transfer ins Gewebe von PA/PANOs aus Jacobaea vulgaris Gaertn. bei 

Wiederkäuern zu untersuchen, um die Aufnahme von PA/PANOs besser bewerten zu können. 

In dieser Studie wurde der Pansenmetabolismus zyklischer Diester untersucht. Neun freie Basen (PA) 

und deren korrespondierenden N-Oxide (PANO), die vor allem in Jacobaea vulgaris Gaertn. zu finden 

sind, wurden in vitro mit Panseninhalt von Rindern untersucht. Die Ergebnisse bestätigten, dass alle 

PANOs schnell zu den entsprechenden PAs reduziert werden und die meisten PAs rasch weiter 

metabolisiert werden. Verbindungen wie Jacobin, Jaconin und Senkirkine zeigten eine langsame 

Eliminierung, während Jacolin nahezu stabil blieb. Zum ersten Mal konnte gezeigt werden, dass 

zyklische Diester im Pansen zu 1,2-gesättigten Metaboliten reduziert werden. Diese Reduktion ist von 

entscheidender Bedeutung, da die 1,2-Doppelbindung strukturell für die Toxifizierung der PAs in der 

Leber notwendig ist. 

Analysen von Panseninhalt aus in vivo Fütterungsstudien mit Rindern, Schafen und Ziegen bestätigten 

die in vitro Ergebnisse. Proben aus diesen in vivo Fütterungsstudien zeigten auch einen geringen Transfer 

von PA/PANOs in das Muskelgewebe der Tiere, wobei hauptsächlich Jacolin, Jacobin und Jaconin 

nachgewiesen wurden, was mit den in vitro beobachteten Eliminierungsraten im Pansen übereinstimmt. 

Daher scheint das Risiko einer Exposition gegenüber PA/PANOs durch den Verzehr von Fleisch von 

Wiederkäuern, die ähnlichen PA/PANO-Dosen wie in dieser Studie ausgesetzt sind, gering zu sein. 

Darüber hinaus liegen die durchschnittlichen PA/PANO-Gehalte in Futtermitteln, die in Europa 

verwendet werden, signifikant unter den in dieser Studie verwendeten Dosen. 

Interessanterweise wurden die im Pansen identifizierten Metabolite nicht im Muskelgewebe 

nachgewiesen. Eine große Menge dieser Metabolite wurde im Kot der Tiere gefunden, was darauf 

hindeutet, dass die Metabolite möglicherweise die Darmbarriere nicht passieren oder in anderen 

Prozessen des Körpers der Tiere weiter umgewandelt werden. 

Die vorliegende Studie zeigt, dass der Pansen eine entscheidende Rolle bei der Entgiftung von 

PA/PANOs spielt. Diese Filterfunktion reduziert die Aufnahme von PA/PANOs erheblich, wodurch 

vermutlich das Risiko durch diese Substanzen für Wiederkäuer verringert wird. Darüber hinaus führt 

die Aktivität des Pansens zu einem geringen Transfer von PA/PANOs in tierische Produkte wie Fleisch, 

wodurch das Risiko einer toxischen Exposition durch Fleischkonsum verringert wird.



The PA-SAFE-FEED project 

   1 

3. Introduction 

3.1. The PA-SAFE-FEED project 
This doctoral thesis was conducted within the collaborative project "Studies on the Transfer of 

Pyrrolizidine Alkaloids into Livestock" (PA-SAFE-FEED). Therefore, the project will be briefly 

introduced in the following. 

 

3.1.1. Background 

Grasslands are essential for providing forage to grazing livestock. Biodiverse grasslands, in particular, 

are highly valuable for supporting biodiversity, yet this diversity can also include potentially toxic plants. 

Pyrrolizidine alkaloid (PA/PANO) producing species, such as Jacobaea vulgaris Gaertn. (JKK, tansy 

ragwort) and various Senecio species, are such ecological valuable species in pastures, offering pollen 

and nectar for insects. However, the presence of ragwort in these ecosystems poses a challenge, 

especially for extensive and organic farming, where restrictions on weed control, fertilization, and 

mowing increase the risk to livestock being exposed to these plants containing toxic PA/PANOs. In 

northern Germany, Jacobaea vulgaris Gaertn. poses a crucial issue. These toxic plants not only 

potentially threaten animal health, with being on field and in feed, but also pose risks to consumers, if 

PA/PANOs enter the food chain via animal products. 

Differences in the sensitivity to PA/PANO toxicity have been published among ruminants with sheep 

being relatively resistant and no avoidance of PA/PANO plants during grazing has been observed. While 

cattle tend to avoid fresh ragwort if alternative vegetation is available, contaminated hay or silage is not 

rejected, increasing the risk of PA/PANO ingestion. 

Thus, reliable data for suitable recommendations and risk management strategies in Germany and the 

EU is requires. 

 

3.1.2. Project objective 

The PA-SAFE-FEED project aims to determine acceptable PA/PANO levels in feed from a consumer 

protection and livestock health perspective. The project focused on ruminants such as cows, sheep, and 

goats. Feeding studies were conducted to assess the health impacts of various PA/PANO levels in feed. 

The project also evaluated the transfer of PA/PANOs into milk and muscle tissues. Transfer was 

determined for both the total PA/PANO content and each individual PA/PANO. This enables conclusions 

about PA/PANO transfer into animal tissues when feed is contaminated with ragwort species other than 

Jacobaea vulgaris Gaertn. The project data should support organic farming goals, such as biodiversity 

conservation, biological production, and regional feed use. 

 



Pyrrolizidine alkaloids 

   2 

3.2. Pyrrolizidine alkaloids 
Pyrrolizidine alkaloids are secondary plant metabolites produced by 3% of all flowering plants (Wexler 

and Anderson, 2005). Notable representatives include plants primarily from the families Asteraceae, 

Boraginaceae, and Fabaceae (Hartmann, 1999). It is presumed that plants synthesize PA/PANOs as a 

defense mechanism against herbivores, particularly insects (Hartmann, 1999; Wink, 2019). However, 

these substances are also toxic to livestock and humans, and are among the most abundant plant toxins 

(Stegelmeier et al., 1999). In plants, pyrrolizidine alkaloids are predominantly present in their N-oxide 

form (Molyneux et al., 2011).  

 

3.2.1. Structure 

Depending on the genus, plants contain various types of PA/PANOs. All PA/PANOs possess a 

pyrrolizidine ring system, consisting of two five carbon rings with a nitrogen atom at position 4 (Figure 

1) (Häseler et al., 2016). Pyrrolizidine alkaloids are composed of a necine base and necic acid. 

Structurally, the pyrrolizidine ring system is present in the necine base, which features a hydroxymethyl 

group at position 1 and a hydroxyl group at position 7. These two groups are connected to necic acids 

via an ester bond (Wiedenfeld et al., 2008).    

 

 
Figure 1: Pyrrolizidine ring system: basic structure of pyrrolizidine alkaloids. 

Depending on the nitrogen configuration, PA/PANOs exist as free bases (PA) or N-oxides (PANO) (red 

box in Figure 2). Necine bases are further categorized based on whether they have a 1,2-double bond or 

a single bond. Additionally, the stereochemistry at position 7 and other characteristics play a role in 

determining the type of necine base (blue box in Figure 2) (Hartmann and Witte, 1995; Häseler et al., 

2016; Roeder, 1995). 

Based on the esterification type of the necine base and necic acid, PA/PANOs are classified into 

monoesters, open-chain diesters, and cyclic diesters. In monoesters, one of the two hydroxyl groups is 

esterified with a carboxylic acid. In open-chain diesters, both hydroxyl groups are esterified with 

separate carboxylic acids. In cyclic diesters, also known as macrocycles, the two hydroxyl groups are 

esterified with the same dicarboxylic acid (yellow box in Figure 2) (Hartmann, 1999; Hartmann and 

Witte, 1995; Roeder, 2000; Wiedenfeld et al., 2008). 

The esterified necic acids are carboxylic acids with up to ten carbon atoms, which can be branched and 

vary in their degree of saturation and attached functional groups (Hartmann, 1999; Rizk, 1991; Robins, 

1989; Wiedenfeld et al., 2008). This structural diversity leads to a large number of different PA/PANOs 



Pyrrolizidine alkaloids 

   3 

– approximately 650 pyrrolizidine alkaloids are known (Hartmann and Witte, 1995; Mattocks, 1986; 

Rizk, 1991; Roeder, 1995). 

 

 
Figure 2: Classification of pyrrolizidine alkaloids based on various characteristics such as the oxidation status of the nitrogen 

(red), the degree of esterification (yellow), and the structure and stereochemistry of the necine base (blue). 

3.2.2. Pyrrolizidine alkaloids from Jacobaea vulgaris Gaertn. 

Jacobaea vulgaris Gaertn. (syn. Senecio jacobaea L.) is a biennial plant that becomes perennial when 

damaged (Harper and Wood, 1957). As a pioneer plant, it quickly colonizes open areas and has low 

requirements (Cameron, 1935; McEvoy, 1984). It is one of the most common toxic plants on pastures 

(Fu et al., 2004; Stegelmeier et al., 1999; Wiedenfeld and Edgar, 2011). Originally from Great Britain, 

it has spread across Eurasia and has been introduced to America, Africa, Australia, and New Zealand 

(Harper and Wood, 1957; Mclaren et al., 2000). In these regions, it causes economic damage through 

losses of livestock, unusable forage, and control measures (Bull et al., 1968; Culvenor, 1985; 

Naranjo, 1987). 

In the genera Jacobaea (and also Senecio), only PAs that are cyclic diesters with a necine base of the 

retronecine or otonecine type are present (Kalač and Kaltner, 2021; Lu et al., 2021). Studies by Joosten 

et al. (2011) and Jung et al. (2020) identified up to 27 PA/PANOs in JKK, though only about ten occur 

in relevant quantities. Most of the PA/PANOs in JKK are of the retronecine type, with senkirkine being 

an exception as it is a otonecine type (Macel et al., 2004; Witte et al., 1992). The PA/PANOs can be 

further divided into two groups: those with and without angelic acid as a structural element. Jacoline, 

jaconine, and jacobine are PA/PANOs without angelic acid, while most other PA/PANOs in JKK contain 

this structure (Jung et al., 2020; Macel et al., 2004; Witte et al., 1992). Individually, the PAs can be 

differentiated by the positions of hydroxyl groups, epoxide groups (jacobine, erucifoline), and chlorine 

atoms (jaconine) (Figure 3) (Macel et al., 2004; Witte et al., 1992).  

 



Metabolism and Toxicology 

   4 

 
Figure 3: Chemical structures from PA/PANOs in Jacobaea vulgaris Gaertn. 

 

3.3. Metabolism and Toxicology 

3.3.1. Absorption, distribution and excretion 

The majority of studies conducted on this topic are performed on rats and other rodents. Orally ingested 

pyrrolizidine alkaloids are rapidly absorbed in rats and mice. Within less than 30 minutes PA/PANO 

concentrations were found in the plasma of the animals (Brauchli et al., 1982; Wang et al., 2011; 

Williams et al., 2002). Once ingested, the PANOs are reduced to the free base form, which is absorbed 

in the animal's intestine. This is known for rats and mice and rabbits (Mattocks, 1971; Powis et al., 1979; 

Yang et al., 2020). Via the portal vein the PAs are transported to the liver. There, among other metabolic 

reactions, the PAs are re-oxidized to PANOs (Wang et al., 2011; Williams et al., 2002). In section 3.3.3. 

a detailed discussion of the liver metabolism is provided. PA/PANOs also have been detected in the 

blood, kidneys, and lungs of rats and mice (Eastman et al., 1982; Estep et al., 1991). PA/PANOs are 

excreted from the body of rats via urine, bile, and feces (Estep et al., 1990, 1991). However, the 

PA/PANOs are also partially reabsorbed and thus enter the enterohepatic circulation, as observed in rats 

and mice (Candrian et al., 1985; Eastman et al., 1982; Estep et al., 1990). Excretion into the milk in 

general is only minimal (Eastman et al., 1982; Hoogenboom et al., 2011; Mulder et al., 2020). 

 

3.3.2. Ruminal metabolism 

The primary function of the rumen is the breakdown of plant material like cellulose and the de novo 

synthesis of proteins by the microbiome (Hungate, 1966; Van Soest, 1994). However, during grazing, 

animals are also exposed to a variety of toxic xenobiotics. The rumen microbiome can detoxify some of 

these (Aguiar and Wink, 2005; Craig et al., 1992; Loh et al., 2020). 

In most plants, PANOs constitute the largest proportion of the PA/PANO load and therefore enter the 

rumen of ruminants in extensive quantities. It has been demonstrated in vitro that the N-oxides (all 

configurations) are converted fast into the corresponding free bases in rumen liquid of sheep and cattle 

(Dick et al., 1963; Lanigan, 1970; Mulder et al., 2020). Several studies indicate the further elimination 

of PAs like heliotrine, supinine, intermedine, lycopsamine (monoesters) and lasiocarpine, echimidine 



Metabolism and Toxicology 

   5 

(open diesters) in in vitro incubation with rumen liquid of sheep (Culvenor et al., 1984; Dick et al., 1963; 

Lanigan and Smith, 1970; Russell and Smith, 1968). The evidence regarding cyclic diesters is less clear. 

Mulder et al. (2020) demonstrated a metabolic elimination of PAs from Jacobaea vulgaris Gaertn. plant 

material (cyclic diesters) incubated with rumen liquids of cattle. Wachenheim et al. (1992) showed such 

an elimination for jacobine and seneciphylline (in vitro studies using rumen liquid of cattle and sheep). 

Contrary, Aguiar et al. (2005) showed in vitro that senecionine is metabolically not eliminated in rumen 

liquid of sheep and cattle, while observing elimination for monocrotaline (both cyclic diesters). Also, 

Swick et al. (1983) concluded from a not observed detoxification of Jacobaea vulgaris Gaertn. plant 

material treated with rumen liquid from sheep, that no reductive metabolization of the free bases 

occurred. 

Apparently, open-chain diesters and monoesters are metabolized by the bacterium Peptostreptococcus 

heliotrinreducens, which, however, cannot metabolize cyclic diesters (Hovermale and Craig, 2002; 

Lanigan, 1976). To date, no microorganism involved in the metabolism of cyclic diester has been 

isolated (Lodge-Ivey et al., 2005). 

Some authors suggest that the microorganisms required for the metabolization of free bases are not 

ubiquitous, and therefore, animals do not naturally harbor these microorganisms in their rumen (Aguiar 

and Wink, 2005; Shull et al., 1976; Swick et al., 1983). Others believe PAs can be metabolized without 

prior exposure. Lanigan et al. (1970) and Culvenor et al. (1984) showed in their experiments that 

animals fed large amounts of PA/PANO-containing plants exhibited increased PA metabolism. This 

suggests that the microorganisms required for PA metabolization are present in most animals, but 

metabolism is increased upon exposure (Aguiar and Wink, 2005). The described differences in the 

ability to metabolize PAs are also reflected in documented feeding trials (see section 3.5.1).  

The first rumen metabolite of PAs was described by Dick et al. (1963). In an in vitro experiment with 

rumen liqiuid from sheep, 7α-hydroxy-1-methylene-8α-pyrrolizidine was produced from heliotrine. 

Other studies confirmed the formation of this 1-methylene metabolite for heliotrine, lasiocarpine, 

echimidine, and the cyclic diester monocrotaline (Hovermale and Craig, 2002; Lanigan and Smith, 1970; 

Russell and Smith, 1968). For PAs like those in Jacobaea vulgaris Gaertn., only a low formation of 

these metabolites is reported (Hovermale and Craig, 2002). Mulder et al. (2020) suggest that cyclic 

diester PAs are eliminated by cleavage of the ester bonds. 

In general, the described metabolites lack a 1,2-double bond, which is the prerequisite to express toxicity 

(see section 3.3.3). Culvenor et al. (1976) showed that the 1-methylene metabolite of lasiocarpine did 

not form pyrrolic esters in rat microsomes, and did not lead to signs of toxic effects in rats. Shull et al. 

(1976) also showed a reduced toxicity in rats for Jacobaea vulgaris Gaertn. incubated in rumen liquid 

of cattle. At the same time Shull et al. (1976) and Swick et al. (1983) did not see such a reduction when 

using rumen liquid from sheep.  

Depending on the study, authors offer different explanations for the varying sensitivity of ruminants to 

PA/PANO toxicity (see section 3.5.1). Authors who did not observe PA elimination in the rumen attribute 



Metabolism and Toxicology 

   6 

the different sensitivities to varying enzyme activities in the liver (Aguiar and Wink, 2005; Shull et al., 

1976; Swick et al., 1983). Others believe the reason lies in the ability to metabolite PAs in the rumen to 

non-toxic metabolites (Craig et al., 1992; Lanigan, 1976; Wiedenfeld and Edgar, 2011). It is likely that 

both organs contribute to this effect. Wachenheim et al. (1992) demonstrated that sheep rumen fluid 

contains more PA-metabolizing microorganisms and that PAs are eliminated faster compared to cattle. 

Simultaneously, the study by Craig et al. (1986) implies that liver metabolism also plays a crucial role: 

when the same dose of PAs was administered intravenously, cattle showed higher sensitivity than sheep.  

 

3.3.3. Liver metabolism 

The fundamental mechanism of PA/PANO biotransformation is similar in animals and humans 

(IPCS, 1988; Wiedenfeld and Edgar, 2011). Pyrrolizidine alkaloids absorbed in the intestine enter the 

liver via the portal vein and are bioactivated by liver enzymes. 1,2-unsaturated pyrrolizidine alkaloids 

are primarily metabolized through three pathways: hydrolysis by esterases (detoxification), oxidation to 

the corresponding N-oxide (detoxification), and oxidation to dehydropyrrolizidine (toxification) 

(Chen et al., 2010; Fu et al., 2004). The toxification process requires activation and relies on certain 

structural features of the PAs. The necine base must be unsaturated at the 1,2-position and esterified with 

a carboxylic acid at position C9 or C7. The esterified carboxylic acid must be at least singly branched 

and consist of a minimum of five carbon atoms (Allgaier and Franz, 2015; Roeder, 1995).  

The hydrolysis by esterases represents a crucial step in the detoxification of PAs. This nonspecific 

esterase activity primarily occurs in the liver, but also in other tissues. The resulting necine bases and 

necic acids are considered non-toxic (Chen et al., 2010; Mattocks, 1986; Roeder, 1995). 

The formation of pyrrolizidine alkaloid N-oxides (PANOs) in the liver represents a second detoxification 

pathway (Mattocks, 1971). Their polarity and ability to conjugate facilitate their excretion. However, it 

is possible that these N-oxides are reduced back to their corresponding PAs in the liver, thereby 

becoming available for toxification processes again (Chen et al., 2010; Mattocks, 1986). Otonecine-type 

PAs cannot be detoxified through the formation of N-oxides because a methyl group is conjugated to the 

nitrogen of the necine base (Lin et al., 2000). 

Pyrrolizidine alkaloids of the retronecine and heliotridine types are hydroxylated at position C3 or C8 

by cytochrome P450 monooxygenases and spontaneously react to dehydropyrrolizidine alkaloids 

through dehydration (Allgaier and Franz, 2015; Chen et al., 2010; Cooper and Huxtable, 1996; 

EFSA, 2011). These unstable dehydropyrrolizidines can further react in three ways. 1) The cleavage of 

the ester at position C7 leads to the formation of an aromatic pyrrole, which reacts with nucleophiles 

such as DNA or proteins. This is crucial for the toxicity of PAs, as the adducts with cellular components 

explain their genotoxic, mutagenic, and cytotoxic effects (Allgaier and Franz, 2015; Fu et al., 2004; 

Roeder, 1995). These bound pyrrolizidines can also be re-released and become reactive again (Allgaier 

and Franz, 2015). The aromatic pyrroles can also be scavenged by nucleophilic glutathione. 2) The 

formed aromatic pyrrole reacts with water leading to the formation of dehydropyrroles (DHPs). They 



Metabolism and Toxicology 

   7 

also react with nucleophiles such as DNA, proteins, or glutathione. Due to their higher stability, 

dehydropyrroles can migrate to other tissues (Allgaier and Franz, 2015; Yang et al., 2016). 3) 

Dehydropyrrolizidines can also directly react with glutathione, leading to detoxification 

(Yang et al., 2016). 

Otonecine types are converted to dehydropyrrolizidines by cytochrome P450-dependent 

monooxygenases through oxidative N-demethylation. After demethylation, a ring closure occurs, 

forming a hydroxylated PA that can also react to dehydropyrrolizidines (Chen et al., 2010; Roeder, 1995). 

1,2-saturated pyrrolizidine alkaloids can also form dehydropyrroles; however, these are stable and do 

not react with nucleophiles (Mattocks and White, 1971). 

 

3.3.4. Toxicology for ruminants 

In ruminants, primarily sub-acute and chronic effects are observed for PA/PANOs (Kalač and 

Kaltner, 2021; Wiedenfeld and Edgar, 2011). The following symptoms and effects are derived from 

studies involving various animal species, breeds, PA/PANO plants, and administration methods, which 

influence the absorption, metabolism, and mode of action of PA/PANOs. However, the symptoms are 

comparable, thus a general overview is provided here. 

Liver and other organ damage result in typical clinical manifestations of PA/PANO poisoning. This 

syndrome, known as "Winton Disease" or "Pictou Cattle Disease” is characterized by symptoms such 

as loss of appetite, depression, ataxia, diarrhea, and wandering behavior (Anjos et al., 2010; 

Barri et al., 1984; Damir et al., 1982; Johnson et al., 1985; Molyneux et al., 2011; Wiedenfeld and 

Edgar, 2011). Additional symptoms include weight loss, reduced milk production, lethargy, hunched 

back, irritability, unpredictable behavior, tenesmus, abdominal distension, and dyspnea 

(Anjos et al., 2010; Barri et al., 1984; Damir et al., 1982; Dickinson et al., 1976; Johnson et al., 1985; 

Molyneux et al., 2011; Wiedenfeld and Edgar, 2011). A few hours before death, jaundice and 

hemoglobinuria have been described (Anjos et al., 2010). 

Liver damage can be assessed by measuring enzyme parameters in the blood of the animals. Liver injury 

leads to the release of enzymes into the blood, indicating the action of hepatotoxins in general and thus 

also PA/PANOs (Ford et al., 1968). However, the specific enzyme parameters depend on the overall 

condition of the animal, including health, age and sex (Johnson and Smart, 1983). Commonly examined 

enzymes related to PA/PANOs include aspartate aminotransferase (AST) and γ-glutamyl transferase 

(GGT), which are general indicators of liver damage. Elevated levels of these parameters in the blood 

can indicate PA/PANO-induced toxicity (Anjos et al., 2010; Baker et al., 1991; Craig et al., 1986; 

Culvenor et al., 1984; Goeger et al., 1982a; Johnson, 1982; Johnson et al., 1985; Johnson and 

Molyneux, 1984; Johnson and Smart, 1983; Maia et al., 2013; Molyneux et al., 1991; Mortimer and 

White, 1975; Nobre et al., 2005; Ohlsen et al., 2022). Some studies also consider other enzymes 

(Damir et al., 1982; Dickinson, 1980; Dickinson et al., 1976; Ford et al., 1968).  



Exposure 

   8 

Pathological examinations of animals reveal primarily liver damages. Fibrosis (Anjos et al., 2010; Damir 

et al., 1982; Dickinson et al., 1976; Jago, 1969; Johnson et al., 1985; Johnson and Molyneux, 1984; 

Molyneux et al., 1991), necrosis (Anjos et al., 2010; Dickinson et al., 1976; Johnson et al., 1985; 

Johnson and Molyneux, 1984; Mattocks, 1986; Thorpe and Ford, 1968), hyperplasia (Jago, 1969), and 

hemorrhages in the liver are observed (Johnson et al., 1985; Mattocks, 1986). The liver appears swollen 

and discolored (Johnson and Molyneux, 1984; Molyneux et al., 1991) and sometimes fatty (Damir 

et al., 1982). Overall, liver cirrhosis (Jago, 1969) or veno-occlusive disease (VOD) can occur (Damir et 

al., 1982; Johnson et al., 1985; Thorpe and Ford, 1968). At the cellular level, megalocytosis (Damir et 

al., 1982; Dickinson et al., 1976; Jago, 1969; Thorpe and Ford, 1968) and karyomegaly (Jago, 1969) are 

observed.   

Other organs also show hemorrhages, hyperplasia, and obstructions (Damir et al., 1982; Johnson et al., 

1985; Johnson and Molyneux, 1984). Copper poisoning (Anjos et al., 2010; Damir et al., 1982), changes 

in the central nervous system (Anjos et al., 2010; Johnson et al., 1985), and pulmonary edema (Damir 

et al., 1982) are also associated with PA/PANO poisoning in ruminants. 

 

3.3.5. Toxicology for humans 

The toxic effects of pyrrolizidine alkaloids in humans are similar to those observed in ruminants and 

occur in acute, subacute, and chronic forms (Kalač and Kaltner, 2021). Acute poisonings manifest 

through symptoms such as abdominal pain, bloody necrosis, hepatomegaly, ascites, and diarrhea 

(Koleva et al., 2012; Wiedenfeld et al., 2008; Wiedenfeld and Edgar, 2011). Subacute exposure often 

leads to veno-occlusive disease (VOD) of the liver, which damages the sinusoidal endothelial cells and 

hepatocytes around the central vein (Wiedenfeld et al., 2008). Chronic exposure to PAs can result in 

chronic VOD, causing necrosis and fibrosis, eventually leading to liver cirrhosis (Chen and Huo, 2010; 

Wiedenfeld et al., 2008). 

Unlike in ruminants, the genotoxic and carcinogenic effects of chronic PA exposure are concerns in 

humans. Studies in rats have demonstrated that PAs can induce tumors in the liver and other organs 

(Fu et al., 2004; IARC, 1983, 1987; NCI, 1978; NTP, 2003).  

 

3.4. Exposure  

3.4.1. Exposure of ruminants to Jacobaea vulgaris Gaertn. 

Livestock is confronted with Jacobaea vulgaris Gaertn. through its growth on pastures used for grazing 

or the production of forage. Whether grazing animals selectively avoid or consume ragwort is a topic of 

ongoing debate. Generally, it is assumed that the plant is not selectively avoided in hay and silage and 

is thus ingested (Kalač and Kaltner, 2021). Thorpe and Ford (1968) observed that cattle rejected pellets 

with a high content of ragwort, a finding also noted by Goeger et al. (1982b) in sheep. Cattle are believed 

to graze on ragwort in the field under certain conditions, e.g. feed shortages, but otherwise consciously 

avoid it (Brumme, 2015; Gilruth, 1905; Johnson and Molyneux, 1984; Naranjo, 1987). However, young 



Exposure 

   9 

plants in the seedling or rosette stage cannot selectively be avoided and are consumed as part of the 

forage (Brumme, 2015; Stegelmeier et al., 1999). Numerous studies indicate that sheep graze on ragwort 

in large quantities voluntarily (Brumme, 2015; Cameron, 1935; Gilruth, 1905; Goeger et al., 1982a; 

Harper and Wood, 1957; Naranjo, 1987; Ohlsen et al., 2022; Schmidl, 1972; Sharrow and Mosher, 1982). 

Although no studies have confirmed that goats graze on Jacobaea vulgaris Gaertn., reports of poisoning 

incidents indicate that they come into contact with the plant (Anholt and Britton, 2017). Also they are 

known to actively consume Senecio inaequidens (Sánchez Valdés et al., 2022).  

In the field, the abundance of ragwort and thus the exposure to PA/PANO can be controlled through 

management practices. After harvesting, ensiling forage provides an effective method for reducing 

PA/PANO levels in feed (Gottschalk et al., 2015; Kalač and Kaltner, 2021; Klevenhusen et al., 2019). 

Jimenez et al. (2013) found that during the ensiling of ragwort, primarily jacoline, jacobine, and jaconine 

remain while the other PAs were eliminated. No reduction in the PA/PANO content is reported during 

the production of hay or pellets (Kalač and Kaltner, 2021; Kaltner et al., 2018; Wiedenfeld and 

Edgar, 2011). A European survey found average levels of 0.29 mg/kg of PA/PANOS in roughage and 

forage (EFSA, 2011). 

 

3.4.2. Exposure of humans 

Humans are exposed to PA/PANOs through a variety of foods, particularly plant-based products such as 

grains, salads, and spices. Documented cases of human intoxication by PA/PANOs are largely associated 

with these foods. An overview of such incidents is available in several publications (EFSA, 2011; 

Wiedenfeld and Edgar, 2011). Additionally, honey, eggs, herbal medicines, and dietary supplements 

contribute to human PA/PANO exposure (BfR, 2007, 2020; Gottschalk et al., 2020; Kaltner et al., 

2020, 2020a; Mulder et al., 2015, 2016, 2018; Roeder, 1995; Wiedenfeld and Edgar, 2011). The 

exposure to PA/PANOs through milk is considered low or negligible (Dusemund et al., 2018; Klein 

et al., 2024; Mulder et al., 2015, p. 201, 2018). Studies with extensive sampling of commercially 

available milk have detected only trace amounts of PA/PANOs (0.03 – 0.30 µg/L). No PA/PANOs have 

been detected in meat samples from retail markets (Huybrechts and Callebaut, 2015; Klein et al., 2024; 

Mulder et al., 2015, 2018). 

Since PA/PANOs are genotoxic and carcinogenic substances, no acceptable daily intake (ADI) can be 

established. To assess potential risks, the margin of exposure (MOE) is used for such substances. The 

European Food Safety Authority (EFSA) has established a benchmark dose lower confidence limit 

(BMDL10) of 237 μg/kg bw/day as a reference point for calculating the MOE from observed 

concentration data and consumption data (EFSA et al., 2017). Klein et al. (2024) determined MOEs for 

milk based on their data, which exceed 10,000 notably, suggesting no need for regulatory action.  

 



Feeding studies and carry-over 

   10 

3.5. Feeding studies and carry-over  
As early as the beginning of the 20th century, certain cases of intoxication in ruminants were linked to 

the consumption of Jacobaea vulgaris Gaertn. (Gilruth, 1903). Since then, numerous cases have been 

documented. An overview can be found in various sources (Bull et al., 1968, p. 198; Mattocks, 1986; 

Molyneux et al., 2011; Panziera et al., 2018). Given that these cases often lack precise information on 

the amount of plant material ingested by the animals and the PA/PANO content in the plants, these 

reports only allow for rough estimates of the toxicity of the plant and their contained PA/PANOs. 

However, the emergence of this PA/PANO issue prompted targeted feeding studies to better understand 

the toxicity (see section 3.5.1). Over the course of the century, in addition to cases of poisoning in 

ruminants, human intoxications by PAs were also observed (Wiedenfeld and Edgar, 2011). This led to 

investigations into whether and to what extent PA/PANOs are transferred to animal-derived foods (see 

section 3.5.2). 

 

3.5.1. Feeding studies 

Table 1 and 2 summarize feeding studies with cattle, goat and sheep involving Jacobaea vulgaris Gaertn. 

and Senecio spp. plants. There are considerable differences and missing information regarding the 

following aspects: animal species, breed, number, sex, and age of the animals, health status (including 

pregnancy and body weight), administered plant species, PA/PANO content of the plant, mode of 

administration, duration and frequency of administration, and measured endpoints. Generally, the cited 

studies tend to use female animals.  

Despite the differences in the study designs and the resulting difficulties in directly comparing the 

studies, similar effects can be observed in the studies. Some effects, however, are inconsistently reported 

or have only been investigated in individual studies. To enable a rough comparison, table 1 and 2 display 

the different trials with key information. For the reported doses, several major assumptions had to be 

made in some cases, as relevant data is missing in these studies.  

Within each study and under similar conditions, it becomes evident that the dose of administered 

PA/PANOs influences the animals’ health. This is manifested in stronger symptoms and increased 

mortality rates at higher doses (Ford et al., 1968; Goeger et al., 1982a; Johnson et al., 1985; Johnson 

and Molyneux, 1984; Maia et al., 2013). However, dose-dependence is not clearly observed in all studies 

(Mortimer and White, 1975; Nobre et al., 2005), and in one case, even an inverse relationship was 

reported (Johnson and Smart, 1983). 

The method of PA/PANO administration and the duration of exposure also shows effects across various 

studies. Generally, animals demonstrate higher tolerance to PA/PANOs when plants are incorporated 

into their feed, mimicking realistic consumption conditions. For instance, cattle that received the same 

dose throughout the day exhibited no symptoms compared to those given a single bolus 

(Johnson et al., 1985; Johnson and Molyneux, 1984). The form of administration (pure or within plants) 

seems to play a minor role in single-dose studies (Molyneux et al., 1991). Additionally, animals that 



Feeding studies and carry-over 

   11 

ingested PA/PANOs continuously throughout the day could tolerate the amounts longer without 

symptoms or fatalities compared to those given a single dose per day (Johnson and Molyneux, 1984). 

There can also be delays between the last PA/PANO intake and the onset of symptoms. Molyneux et al. 

(1988) and Johnson and Smart (1983) observed this in cattle, while Dickinson et al. (1980) reported 

similar findings in goats. New pregnancies and heat stress were triggers for the symptoms in these cases 

(Dickinson, 1980; Johnson and Smart, 1983).  

Also, animal weight, age, and overall health play a role. Johnson and Molyneux (1984) demonstrated 

that lighter cattle exhibited problems at a certain dose, while heavier animals did not. This study also 

indicated that older laboratory animals coped better with PA/PANOs, however, this was not observed in 

sheep (Johnson and Molyneux, 1984; Mortimer and White, 1975). Johnson and Molyneux (1984) 

assume increased cellular activity in younger animals leading to enhanced formation of PA-toxins.  

Although pregnant animals may suffer from prior or current PA/PANO intoxication, Johnson and Smart 

(1983) concluded from complication-free births that the fetuses are not harmed. This suggests either 

detoxification by the mother or that reactive metabolites do not cross the placenta. Calves and kids that 

consumed milk from mothers exposed to doses of 9.4 ± 5.6 mg/kg bw/day or 16.0 mg/kg bw/day of 

PA/PANOs also showed no adverse effects (Dickinson, 1980; Dickinson et al., 1976). Liver damage 

caused by PA/PANOs appears to be potentially reversible (Goeger et al., 1982a). 

In Section 3.3.2, the adaptation of animals to PA/PANOs was discussed. For example, Culvenor et al. 

(1984) demonstrated this adaptation in the rumen fluid of animals from feeding trials with Echium 

plantagineum. Anjos et al. (2010) showed adaptation to monocrotaline in Crotalaria seeds in a feeding 

trial with sheep. Animals initially receiving lower doses of monocrotaline over an extended period coped 

better with a single dose of 342.0 mg/kg bw compared to animals that had not previously been exposed 

to monocrotaline and received a single dose of 205.2 or 273.6 mg/kg bw. 

  



Feeding studies and carry-over 

   12 

  
Ta

bl
e 

1:
 F

ee
di

ng
 st

ud
ie

s w
ith

 c
at

tle
 u

sin
g 

Ja
co

ba
ea

 v
ul

ga
ris

 a
nd

 S
en

ec
io

 sp
ec

ie
s. 

 
 

 
 

 
 

 
 

 
 

Ch
an

ge
s 

 
So

ur
ce

 
Pl

an
t 

Sp
ec

ie
s 

An
im

al
 

# 
Ad

m
. 

ty
pe

 
D

ur
at

io
n 

ad
m

. [
d]

 
D

os
e 

[m
g/

kg
 b

w
/d

] 
Ab

so
lu

t d
os

e 
[m

g/
kg

 b
w

] 
D

ea
d 

 
an

im
al

s 
Ti

m
e 

to
 

de
at

h 
[d

]*
* 

Cl
in

ic
al

 E
nz

ym
e 

ac
tiv

ity
 Pa

th
ol

og
ic

/ 
hi

st
ol

og
ic

 
As

su
m

pt
io

ns
 

Th
or

pe
 

et
 a

l. 
(1

96
8)

, 
Fo

rd
 

et
 a

l. 
(1

96
8)

 

J. 
vu

lg
ar

is
 

ca
tt

le
 

1 

A3
 

38
 

6.
5 

± 
2.

6 
* 

24
8 

± 
10

0 
* 

1 
49

   
++

 
++

 
++

 
25

0 
± 

25
 k

g 
bw

; 0
.2

-0
.4

%
 P

A/
PA

N
O

 a
m

ou
nt

 
in

 D
M

 p
la

nt
 

2 
33

 
4.

3 
± 

1.
8 

* 
14

4 
± 

58
 *

 
2 

60
   

++
 

++
 

++
 

1 
38

 
2.

2 
± 

0.
9 

* 
83

 ±
 3

3 
* 

1 
55

   
++

 
++

 
++

 
1 

12
8 

2.
2 

± 
0.

9 
* 

27
9 

± 
11

2 
* 

1 
16

6 
  

+ 
++

 
++

 
M

or
tim

er
 

an
d 

W
hi

te
 

(1
97

5)
 

J. 
vu

lg
ar

is
 

ca
lf 

3 
B1

 
77

 
3.

6 
± 

1.
5 

* 
18

2 
± 

88
 *

 
3 

50
   

++
 

N
D

 
++

 
sp

ec
ifi

ed
 b

w
 ±

 1
0%

; 0
.2

-0
.4

%
 P

A/
PA

N
O

  
am

ou
nt

 in
 D

M
 p

la
nt

 
3 

77
 

2.
1 

± 
0.

9 
* 

88
 ±

 4
7 

* 
3 

42
   

++
 

N
D

 
++

 

Di
ck

in
so

n 
et

 a
l. 

(1
97

6)
 

J. 
vu

lg
ar

is
 

ca
tt

le
 

4 
B3

 
35

 
9.

4 
± 

5.
6 

17
8 

± 
29

 
4 

35
   

++
 

++
 

++
 

 

Jo
hn

so
n 

et
 a

l. 
(1

97
6)

 
J. 

vu
lg

ar
is

 
ca

tt
le

 
2 

B2
 

57
/3

0 
9.

9 
± 

3.
9 

38
6 

± 
14

9 
2 

74
 ±

 3
5 

N
D

 
++

 
++

 
0.

2-
0.

4%
 P

A/
PA

N
O

 a
m

ou
nt

 in
 D

M
 p

la
nt

 
2 

75
 

3.
9 

± 
1.

8 
29

2 
± 

13
8 

2 
75

 ±
 1

 
N

D
 

++
 

++
 

2 
50

/8
4 

3.
0 

± 
1.

4 
15

9 
± 

14
3 

2 
90

 ±
 5

1 
N

D
 

++
 

++
 

Jo
hn

so
n 

et
 a

l. 
(1

98
2)

 
J. 

vu
lg

ar
is

 
ca

lf 
1 

B1
 

18
 

3.
0 

  
54

 
  

0 
15

0 
  

++
 

+  
++

 
 

Jo
hn

so
n 

et
 a

l. 
(1

98
3)

 
J. 

vu
lg

ar
is

 
ca

tt
le

 
7 

B2
 

15
 

2.
3 

  
33

,8
 

  
4 

18
8 

± 
10

1 
++

 
+ 

++
 

 
7 

B2
 

15
 

2.
0 

  
29

,7
 

  
0 

   
 

- 
- 

- 
 

 
 

 
 

 
 

 
 

 
 

 
  

 
 

 
 

Jo
hn

so
n 

an
d 

M
ol

yn
eu

x 
(1

98
4)

 
S.

 d
ou

gl
as

ii 
ca

tt
le

 

2 

B1
 

16
 

5.
0 

  
80

 
  

0 
   

 
+ 

+ 
+ 

 

4 
18

 
8.

0 
  

14
4 

  
1 

44
9 

  
+ 

+ 
+ 

 

2 
16

 
10

.0
 

  
16

0 
  

0 
   

 
+ 

+ 
+ 

 

3 
20

 
10

.0
 

  
20

0 
  

2 
46

   
++

 
+ 

++
 

 

4 
10

/1
0 

10
.0

 
  

20
0 

  
3 

22
0 

  
++

 
+ 

++
 

 

8 
15

 
13

.0
 

  
19

5 
  

8 
61

   
++

 
++

 
++

 
 

2 
14

 
15

.0
 

  
21

0 
  

2 
38

   
++

 
++

 
++

 
 

2 
15

 
18

.0
 

  
28

5 
  

1 
45

   
++

 
++

 
++

 
 

2 
2 

40
.0

 
  

80
 

  
2 

3 
  

++
 

++
 

++
 

 

5 

A2
 

10
0 

2.
0 

  
20

0 
  

0 
   

 
N

D
 

- 
- 

 

5 
10

0 
4.

0 
  

40
0 

  
0 

   
 

N
D

 
- 

- 
 

5 
10

0 
6.

0 
  

60
0 

  
0 

   
 

N
D

 
- 

- 
 

3 
20

 
10

.0
 

  
20

0 
  

0 
   

 
N

D
 

- 
- 

 

4 
20

 
20

.0
 

  
40

0 
  

0 
   

 
N

D
 

+ 
- 

 

4 
20

 
30

.0
 

  
60

0 
  

0 
   

 
N

D
 

+ 
- 

 

4 
12

 
30

.0
 

  
36

0 
  

0 
   

 
N

D
 

- 
- 

 

 



Feeding studies and carry-over 

   13 

  

Ta
bl

e 
1 

co
nt

in
ue

d:
 F

ee
di

ng
 st

ud
ie

s w
ith

 c
at

tle
 u

si
ng

 J
ac

ob
ae

a 
vu

lg
ar

is 
G

ae
rtn

. a
nd

 S
en

ec
io

 sp
ec

ie
s. 

 
 

 
 

 
 

 
 

 
 

Ch
an

ge
s 

 
So

ur
ce

 
Pl

an
t 

Sp
ec

ie
s 

An
im

al
 

# 
Ad

m
. 

ty
pe

 
D

ur
at

io
n 

ad
m

. [
d]

 
D

os
e 

[m
g/

kg
 b

w
/d

] 
Ab

so
lu

t d
os

e 
[m

g/
kg

 b
w

] 
D

ea
d 

 
an

im
al

s 
Ti

m
e 

to
 

de
at

h 
[d

]*
* 

Cl
in

ic
al

 E
nz

ym
e 

ac
tiv

ity
 Pa

th
ol

og
ic

/ 
hi

st
ol

og
ic

 
As

su
m

pt
io

ns
 

Jo
hn

so
n 

et
 a

l. 
(1

98
5)

 
S.

 ri
dd

el
lii

 
ca

tt
le

 

4 
A2

 
20

 
20

.0
 

  
40

0 
  

0 
   

 
- 

- 
- 

 

4 
20

 
40

.0
 

  
80

0 
  

1 
15

8 
  

+ 
+ 

+ 
 

4 
20

/2
0/

20
 

30
.0

 
  

60
0 

  
1 

40
0 

  
+ 

- 
+ 

 

4 
B4

 
20

 
10

.0
 

  
20

0 
  

0 
   

 
- 

- 
- 

 

4 
20

 
20

.0
 

  
40

0 
  

0 
   

 
- 

+ 
- 

 

4 
20

 
20

.0
 

  
40

0 
  

3 
13

8 
± 

14
8 

++
 

++
 

+ 
 

4 

B2
 

20
 

10
.0

 
  

20
0 

  
0 

   
 

- 
- 

- 
 

4 
20

 
15

.0
 

  
30

0 
  

4 
57

 ±
 1

8 
++

 
++

 
++

 
 

4 
20

 
20

.0
 

  
40

0 
  

4 
32

 ±
 6

 
++

 
++

 
++

 
 

4 
20

 
25

.0
 

  
50

0 
  

4 
54

 ±
 2

1 
++

 
++

 
++

 
 

2 
5 

60
.0

 
  

30
0 

  
2 

27
 ±

 4
 

++
 

++
 

++
 

 

M
ol

ly
ne

ux
 

et
 a

l. 
(1

99
1)

 
S.

 ri
dd

el
lii

 
ca

lf 

3 
B1

 
20

 
45

.0
 

  
90

0 
  

3 
29

 ±
 8

 
++

 
++

 
++

 
 

3 
D1

 
20

 
4.

5 
Rd

 
90

 
  

0 
   

 
- 

+ 
- 

 
3 

D1
 

20
 

40
.5

 R
dN

 
81

0 
  

3 
44

 ±
 5

 
++

 
++

 
++

 
 

3 
D1

 
20

 
45

.0
 R

d+
Rd

N
 

90
0 

  
3 

50
 ±

 1
4 

++
 

++
 

++
 

 
 

 
 

 
 

 
  

 
 

 
  

 
 

 
 

Fl
et

ch
er

 
et

 a
l. 

(2
01

1)
 

S.
 

br
ig

al
ow

en
si

s 
ca

lf 
N

D
 

N
D

 
42

 
2.

5 
 

10
5 

 
0 

 
 

- 
- 

- 
 

 
 

 
 

 
 

  
 

 
 

  
 

 
 

 
A1

: V
ol

un
ta

ry
 in

 th
e 

fie
ld

; A
2:

 P
la

nt
 in

 fe
ed

; A
3:

 P
el

le
ts

 m
ad

e 
fr

om
 p

la
nt

s 
an

d 
an

im
al

 fe
ed

; B
1:

 O
ra

l a
dm

in
is

tr
at

io
n 

of
 th

e 
pl

an
t; 

B2
: O

ra
l a

dm
in

is
tr

at
io

n 
of

 th
e 

pl
an

t i
n 

sl
ur

ry
; B

3:
 A

dm
in

is
tr

at
io

n 
of

 th
e 

pl
an

t v
ia

 
fis

tu
la

; B
4:

 O
ra

l a
dm

in
is

tr
at

io
n 

of
 p

la
nt

s 
in

 g
el

at
in

; C
1:

 O
ra

l a
dm

in
is

tr
at

io
n 

of
 p

la
nt

 e
xt

ra
ct

; D
1:

 A
dm

in
is

tr
at

io
n 

of
 p

ur
e 

su
bs

ta
nc

e 
vi

a 
fis

tu
la

 –
 /

: B
re

ak
 d

ur
in

g 
th

e 
Ad

m
in

is
tr

at
io

n;
 *

: A
ss

um
pt

io
ns

 w
er

e 
m

ad
e 

in
 

or
de

r t
o 

be
 a

bl
e 

to
 d

et
er

m
in

e 
ad

m
in

is
te

re
d 

do
se

s;
 *

*:
 A

ft
er

 fi
rs

t a
dm

in
is

tr
at

io
n;

 -:
 N

o 
ch

an
ge

s;
 +

: S
m

al
l c

ha
ng

es
; +

+:
 S

ev
er

e 
ch

an
ge

s;
 A

dm
.: 

Ad
m

in
is

tr
at

io
n,

 b
w

: B
od

y 
w

ei
gh

t; 
DM

: D
ry

 m
at

te
r;

 N
D:

 N
o 

da
ta

. 
 



Feeding studies and carry-over 

   14 

  

Ta
bl

e 
2:

 F
ee

di
ng

 st
ud

ie
s w

ith
 sh

ee
p 

an
d 

go
at

s u
si

ng
 J

ac
ob

ae
a 

vu
lg

ar
is 

G
ae

rtn
. a

nd
 S

en
ec

io
 sp

ec
ie

s. 

 
 

 
 

 
 

 
 

 
 

Ch
an

ge
s 

 
So

ur
ce

 
Pl

an
t 

Sp
ec

ie
s 

An
im

al
 

# 
Ad

m
. 

ty
pe

 
D

ur
at

io
n 

ad
m

. [
d]

 
D

os
e 

[m
g/

kg
 b

w
/d

] 
Ab

so
lu

t d
os

e 
[m

g/
kg

 b
w

] 
D

ea
d 

 
an

im
al

s 
Ti

m
e 

to
 

de
at

h 
[d

]*
* 

Cl
in

ic
al

 
En

zy
m

e 
ac

tiv
ity

 
Pa

th
ol

og
ic

/ 
hi

st
ol

og
ic

 
As

su
m

pt
io

ns
 

G
ilr

ut
h 

(1
90

5)
 

J. 
vu

lg
ar

is
 

sh
ee

p 
2 

A2
 

16
8 

14
.1

 ±
 6

.4
 *

 
23

63
 ±

 1
07

6 
**

 
0 

   
 

- 
N

D
 

- 
50

 ±
 1

0 
kg

 b
w

; 0
.2

-0
.4

%
 P

A/
PA

N
O

- 
am

ou
nt

 in
 D

M
 p

la
nt

 

M
or

tim
er

 
an

d 
W

hi
te

 
(1

97
5)

 
J. 

vu
lg

ar
is

 

la
m

b 
10

 
B1

 
11

2 
9.

0 
± 

3.
0 

* 
10

08
 

± 
47

5 
**

 
1 

N
D

 
 

++
 

+ 
+ 

0.
2-

0.
4%

 P
A/

PA
N

O
 a

m
ou

nt
 in

 D
M

 p
la

nt
 

sh
ee

p 

8 

B1
 

11
2 

9.
0 

± 
3.

0 
**

 
10

08
 ±

 4
75

 *
* 

0 
 

  
++

 
+ 

+ 

0.
2-

0.
4%

 P
A/

PA
N

O
 a

m
ou

nt
 in

 D
M

 p
la

nt
 

4 
14

0 
9.

3 
± 

3.
5 

**
 

52
8 

± 
18

40
 *

* 
3 

17
5 

 
++

 
N

D
 

++
 

4 
28

 
9.

3 
± 

3.
5 

**
 

23
99

 ±
 1

27
7 

**
 

0 
   

 
+ 

N
D

 
+ 

4 
56

 
9.

3 
± 

3.
5 

**
 

47
89

 ±
 2

54
9 

**
 

1 
35

   
+ 

N
D

 
+ 

Cr
ai

g 
et

 a
l. 

(1
98

6)
 

J. 
vu

lg
ar

is
 

sh
ee

p 
3 

C1
 

20
 

9.
8 

  
19

6 
  

0 
   

 
- 

- 
+ 

  

O
hl

se
n 

et
 a

l. 
(2

02
0)

 
J. 

vu
lg

ar
is

 
sh

ee
p 

63
 

A1
 

16
3 

34
.0

 ±
 1

4.
0 

**
 

21
42

 ±
 8

82
 *

* 
0 

   
- 

- 
- 

40
 ±

 5
 a

nd
 6

0 
± 

5 
kg

 b
w

; 2
0%

 D
M

; 
0.

2-
0.

4%
 P

A/
PA

N
O

 a
m

ou
nt

 in
 D

M
 p

la
nt

 
16

3 
50

.0
 ±

 1
9.

7 
**

 
31

50
 ±

 1
24

1 
**

 
0 

 
 

- 
- 

- 
 

 
 

 
 

 
 

 
  

 
 

 
 

 
 

 
Di

ck
in

so
n 

(1
98

0)
 

J. 
vu

lg
ar

is
 

go
at

 
4 

B3
 

12
5 

16
.0

 
  

20
00

 
  

3 
13

3 
± 

30
 

++
 

+ 
++

 
  

G
oe

ge
r 

et
 a

l. 
(1

98
2)

 
J. 

vu
lg

ar
is

 

ki
d 

2 
A3

 
38

1 
33

.0
 ±

 1
3.

7 
**

 
12

54
5 

± 
48

74
 *

* 
1 

66
8 

  
N

D
 

- 
+ 

0.
2-

0.
4 

%
 P

A/
PA

N
O

 a
m

ou
nt

 in
 D

M
 p

la
nt

 
1 

15
5 

8.
8 

± 
4.

1 
**

 
13

63
 ±

 6
42

 *
* 

0 
   

 
N

D
 

- 
- 

go
at

 

1 

A3
 

38
8 

10
.9

 ±
 5

.1
 *

* 
42

24
 ±

 1
99

1 
**

 
1 

73
9 

  
N

D
 

- 
++

 

0.
2-

0.
4 

%
 P

A/
PA

N
O

 a
m

ou
nt

 in
 D

M
 p

la
nt

 
1 

15
2 

24
.5

 ±
 1

1.
5 

**
 

37
17

 ±
 1

75
2 

**
 

1 
50

3 
  

N
D

 
- 

++
 

2 
16

2/
11

4 
13

.0
 ±

 5
.0

 *
* 

17
90

 ±
 7

75
 *

* 
0 

   
 

N
D

 
- 

- 
1 

43
 

5.
7 

± 
2.

7 
**

 
24

4 
± 

11
5 

**
 

0 
   

 
N

D
 

- 
- 

 
 

 
 

 
 

  
  

 
  

 
 

 
 

H
ip

pc
he

n 
et

 a
l. 

(1
98

6)
 

S.
ve

rn
al

is
 

go
at

 
2 

B3
 

10
7 

15
.3

 ±
 4

.4
 *

* 
16

38
 ±

 4
75

 *
* 

0 
  

N
D

 
N

D
 

+ 
50

 ±
 1

0 
kg

 b
w

; 0
.5

-0
.7

%
 P

A/
PA

N
O

 
am

ou
nt

 in
 D

M
 p

la
nt

 
 

 
 

 
 

 
  

  
 

  
 

 
 

 
A1

: V
ol

un
ta

ry
 in

 th
e 

fie
ld

; A
2:

 P
la

nt
 in

 fe
ed

; A
3:

 P
el

le
ts

 m
ad

e 
fr

om
 p

la
nt

s a
nd

 a
ni

m
al

 fe
ed

; B
1:

 O
ra

l a
dm

in
is

tr
at

io
n 

of
 th

e 
pl

an
t; 

B2
: O

ra
l a

dm
in

is
tr

at
io

n 
of

 th
e 

pl
an

t i
n 

sl
ur

ry
; B

3:
 A

dm
in

is
tr

at
io

n 
of

 th
e 

pl
an

t v
ia

 fi
st

ul
a;

 
B4

: O
ra

l a
dm

in
is

tr
at

io
n 

of
 p

la
nt

s i
n 

ge
la

tin
; C

1:
 O

ra
l a

dm
in

is
tr

at
io

n 
of

 p
la

nt
 e

xt
ra

ct
; D

1:
 A

dm
in

is
tr

at
io

n 
of

 p
ur

e 
su

bs
ta

nc
e 

vi
a 

fis
tu

la
 –

 /:
 B

re
ak

 d
ur

in
g 

th
e 

Ad
m

in
is

tr
at

io
n;

 *
: A

ss
um

pt
io

ns
 w

er
e 

m
ad

e 
in

 o
rd

er
 to

 b
e 

ab
le

 
to

 d
et

er
m

in
e 

ad
m

in
is

te
re

d 
do

se
s;

 *
*:

 A
ft

er
 fi

rs
t a

dm
in

is
tr

at
io

n;
 -:

 N
o 

ch
an

ge
s;

 +
: S

m
al

l c
ha

ng
es

; +
+:

 S
ev

er
e 

ch
an

ge
s 

; A
dm

.: 
Ad

m
in

is
tr

at
io

n,
 b

w
: B

od
y 

w
ei

gh
t; 

D
M

: D
ry

 m
at

te
r;

 N
D:

 N
o 

da
ta

. 
 



Feeding studies and carry-over 

   15 

Lethal doses 

Many of the feeding studies presented in table 1 and 2 administered doses that led to the death of the 

animals. Modern feeding studies focus on more realistic doses and aim less to induce severe diseases. 

Studies that used higher dosages allow the estimation of the lethal dose of animals exposed to Jacobaea 

vulgaris Gaertn. In cattle, a total intake of 0.05-0.2 kg JKK/kg bw over several days is considered lethal, 

whereas for sheep and goats, it is estimated to be 1.25-4.04 kg JKK/kg bw (Goeger et al., 1982a; IPCS, 

1988). The varying lethal doses indicate that cattle, sheep, and goats exhibit different sensitivities to 

Jacobaea vulgaris Gaertn. Hereinafter these three species are categorized along with others according 

to their sensitivity (Goeger et al., 1983; Hooper, 1978; Hooper and Scanlan, 1977; Maia et al., 2013; 

Mortimer and White, 1975):  

 

sheep = goat = rabbit > mice > rat > cattle > chicken > pig 

 

The differences in sensitivity are likely due to variations in digestive systems and differing rates of 

PA/PANO metabolism in both the rumen and liver. This explains why non-ruminants can be less 

sensitive or similarly sensitive to PA/PANO compared to ruminants. (Duringer et al., 2004; 

Lanigan, 1972; Peterson and Jago, 1984; Shull et al., 1976; White et al., 1973). 

 

3.5.2. Carry-Over  

Some feeding studies were conducted to investigate the transfer of PA/PANOs from Jacobaea vulgaris 

Gaertn. into tissues of animals used for food production. Studies in which milk from exposed cattle and 

goats was fed to calves, kids and rats showed that PA/PANOs partially transfer into the milk. However, 

because the toxicity of the milk was low, the transfer is considered minimal, even with high doses of 

Jacobaea vulgaris Gaertn. administered to the mother (Dickinson, 1980; Goeger et al., 1982b; 

Johnson, 1976). Dickinson et al. (1976) confirmed in two studies that transfer of PA/PANOs into the 

milk from cattle and goats administered with Jacobaea vulgaris Gaertn. was low, with 0.1–0.2% and 

only jacoline detectable in the milk. The authors also showed, that calves, kids and rats fed with this 

milk were not harmed. Based on this, the authors assessed the toxicity of jacoline as low (Dickinson, 

1980; Goeger et al., 1982b). This low transfer rate into ruminant milk has been confirmed by many other 

studies (Candrian et al., 1991; Deinzer et al., 1982; Hoogenboom et al., 2011; Mulder et al., 2020; 

Panariti et al., 1997). These studies also confirmed that when Jacobaea vulgaris Gaertn. is fed, the main 

PA in the milk is jacoline. Studies with further Senecio species that produce a jacoline free PA-profile 

could demonstrate that otonecines are predominantly detectable in the milk (Hoogenboom et al., 2011; 

Mulder et al., 2020). The observed low transfer into milk is consistent with concentrations measured in 

commercial samples (Klein et al., 2024; Mulder et al., 2015, 2018).  

There are currently few studies on the transfer of PA/PANOs from feed into the meat of ruminants. 

Fletcher et al. (2011) reported that feeding Senecio brigalowensis to weaned calves resulted in a low 



Objective 

   16 

transfer, primarily of otonecine alkaloids, into the muscle tissue. However, concentrations or transfer 

rates were not specified (Fletcher et al., 2011).  

For evaluating the carry-over, it is important to note that both meat and milk are usually processed before 

consumption. De Nijs et al. (2017) for example showed that during microbial fermentation during cheese 

and yoghurt production PA/PANO concentration was reduced which can further alter the PA/PANO 

profile and content. 

 

3.6. Objective 
Pyrrolizidine alkaloids, naturally occurring toxins found in plants such as Jacobaea vulgaris Gaertn., 

possess hepatotoxic and carcinogenic properties that can affect both humans and animals. Jacobaea 

vulgaris Gaertn. spreads rapidly and extensively. Considering the health risks associated with 

contaminated grazing lands and feed, and known cases of livestock intoxication, authorities monitor the 

spread of this plant with concern. 

Previous animal studies have primarily focused on high concentrations of PA/PANOs and their severe 

to lethal effects. However, studies on the carry-over into animal products like milk are scarce, and for 

meat, almost non-existent. This gap in data complicates the formulation of adequate recommendations 

for animal and consumer protection for the use of grassland that is infested with PA/PANO producing 

plants. 

In this context, the PA-SAFE-FEED project was initiated to determine, through in vivo feeding studies, 

the dose-dependent health effects of PA/PANOs on cows, sheep, and goats, and to investigate the transfer 

of these substances and their metabolites into animal products. Additionally, in vivo and in vitro 

experiments aimed to elucidate the kinetics and metabolism of these substances. The goal was to collect 

data that can be used for risk assessment and competent authorities to manage grazing land with 

PA/PANO-plants and feed contaminated with PA/PANOs. 

The tasks of this doctoral thesis included analytical and evaluative support for the in vivo studies as well 

as the conduct and analysis of in vitro experiments to identify PA/PANO metabolites. Other studies 

within the PA-SAFE-FEED project were conducted by project partners at Ludwig-Maximilians-

Universität, the Max Rubner Institute, and the Friedrich-Loeffler-Institute. 

Since pyrrolizidine alkaloids are metabolized in the rumen of animals, this work aimed to simulate the 

rumen in vitro using rumen fluid from fistulated cattle, to observe the behavior of PA/PANOs and to 

enable the identification of potential metabolites. To determine the extent to which PA/PANOs and their 

metabolites are resorbed by the animals in the gastrointestinal tract, Ussing chamber experiments with 

various bovine epithelial tissues from the gastrointestinal tract were planned. Given that PA/PANOs are 

converted into toxic compounds in the liver of humans and animals, the toxification of PA/PANOs from 

Jacobaea vulgaris Gaertn. and their ruminal metabolites should be investigated using microsomes from 

cattle, sheep, and goats.  



Objective 

   17 

High-resolution mass spectrometry should be used to analyze in vitro and in vivo samples, providing 

structural information about the metabolites and ensure the in vivo relevance of these metabolites. 

Analyzing PA/PANO and metabolite contents in various tissue types (blood, milk, muscle, liver, bile, 

urine, feces) from in vivo experiments should lead to a better understanding of the distribution and fate 

of PA/PANOs in the animals. Overall, the data generated in this dissertation aimed to elucidate the 

mechanisms leading to the transfer of PA/PANOs from Jacobaea vulgaris Gaertn. into the tissues of 

ruminants



Summarizing discussion 

   18 

4. Summarizing discussion 
Pyrrolizidine alkaloids are plant toxins found in over 6000 plant species, including Jacobaea vulgaris 

Gaertn. After bioactivation in the liver, they exhibit hepatotoxic and carcinogenic properties harmful to 

both humans and animals. It is known that ruminants consume these plants either by choice or because 

they lack the ability to selectively avoid them. Leading to potential health problems for livestock and 

due to carry-over into animal products also for humans. Studies suggest low carry-over of PA/PANOs 

into milk and meat, but as the overall recovery for these substances within these studies is low, 

PA/PANOs fate in the animal body still is poorly understood. This raises questions about the absorption, 

distribution, metabolism, and excretion (ADME) of these compounds in ruminants and this study aimed 

to account for these topics. Findings of this doctoral thesis have been published and the two respective 

publications are accessible in chapter 6. The PA-SAFE-FEED project in which this thesis was conducted 

is still ongoing. Thus, many of the data and results have not yet been published or fully finalized. 

Therefore, the following discussion integrates, analyzes, and contextualizes the published as well as the 

unpublished results. 

Previous studies showed that PA/PANOs are metabolically eliminated in the rumen. For mono- and open 

diesters also the identification of formed metabolites is described. The metabolism of cyclic diesters in 

rumen is poorly understood (Loh et al., 2020). Therefore, the aim of this work was to investigate, the 

rumen metabolism of cyclic diesters, which are produced by Jacobaea vulgaris Gaertn. and Senecio 

species. Nine PAs and their corresponding PANOs were studied in vitro using bovine rumen content 

(publication 1). It was demonstrated that all PANOs being rapidly and quantitatively transformed into 

their corresponding free bases, consistent with the results of Mulder et al. (2020). We also confirmed 

further elimination for most of the free bases but observed different elimination rates, with jacobine, 

jaconine and senkirkine being eliminated more slowly, while jacoline even remained comparatively 

stable throughout the experiment.  

For the first time, ruminal metabolites of cyclic diesters from Jacobaea vulgaris Gaertn. were identified 

in our study. The first transformation step included the reduction of the double bond in the necine base 

to 1,2-saturated structures. This molecular change was proposed by the high-resolution mass 

spectrometric determination of respective sum formulas and supported by fragmentation patterns 

characteristic for saturated PAs. This finding aligns with studies observing 1,2-saturated metabolites for 

mono- and open diesters in the rumen (Dick et al., 1963; Hovermale and Craig, 2002; Lanigan, 1970; 

Russell and Smith, 1968). The reduction is a crucial step in the ruminan metabolism of pyrrolizidine 

alkaloids since the 1,2-double bond is structurally necessary for PA bioactivation in the liver. In a second 

step the metabolites were further reduced. The molecular formula changed by obtaining two hydrogens 

during these second reduction. Since the fragmentation pattern could not be used to determine the 

location of the reduction, the structure of these metabolites remains uncertain. Nevertheless, we found, 

that these second-step metabolites were primarily identified for PAs with an angelic acid structure in 



Summarizing discussion 

   19 

their necic acid, such as erucifoline, retrorsine, riddelliine, senecionine, seneciphylline, and senkirkine. 

This suggests that the second reduction affects the carbon-carbon double bond in the necic acid part of 

the molecule.  

Platyphylline, a commercially available 1,2-saturated PA, is theoretically identical to the first-step 

metabolite of senecionine that we have postulated (publication 1). Unfortunately, it was found to have a 

different retention time, suggesting additional structural modifications during metabolism. These 

changes do not appear to affect the molecular formula and, as a result, are not detected by HRMS 

(unpublished date). Mulder et al. (2020) predicts an ester cleavage for cyclic diesters, which we could 

neither confirm nor exclude on basis of fragmentation patterns of metabolites. Experiments with hepatic 

microsomes from cows, goat, human and rats have shown that the ruminal metabolites were not 

metabolized further by liver microsomes suggesting that a bioactivation during hepatic metabolism 

towards pyrrolic metabolites will not occur (unpublished data). In addition, we showed that the identified 

ruminal metabolites are not detectable in the blood, milk nor muscle. Our studies showed that most of 

the metabolites were excreted in the feces, with only trace amounts found in other tissues. This suggests 

that these metabolites are poorly absorbed in the intestines (unpublished data). A full structure 

elucidation, for example with nuclear magnetic resonance spectrometry, could help explain why PAs are 

adsorbed and their ruminal metabolites not. 

Demonstrating a high recovery for the ruminal in vitro experiment it can be assumed that all quantitative 

relevant ruminal metabolites have been identified. The investigation of the bovine rumen fluids from 

the in vivo experiments of the PA-SAFE-FEED project confirmed the quantitative relevance of the found 

metabolites. Also, the other findings were verified, as no PANOs were detectable in the rumen fluids, 

whereas PAs, particularly jacoline, jacobine, and jaconine, were detectable (publication 1). In addition, 

rumen fluid samples from sheep and goats were analyzed for PA/PANOs and their metabolites. Despite 

general concentration differences attributed to different experimental designs and physiological 

differences, similar metabolic processes were observed (publication 2). The extensive and rapid 

metabolism of PA/PANOs in the rumen probably explains the low levels of these compounds in the 

animals' muscle tissues.  

Only small amounts of PA/PANOs were detected in the muscle of the animals from the PA-SAFE-FEED 

experiments, consistent with earlier studies (publication 2) (Fletcher et al., 2011). However, this work 

is the first that determined the transfer parameters for the individual PA/PANOs from Jacobaea vulgaris 

Gaertn. into muscle tissue of ruminants. Interestingly, mostly jacoline, jacobine, and jaconine were 

detectable in all muscle samples from the three studied animal species. This correlates with in vitro 

results showing that these three PAs were eliminated more slowly in the rumen. However, other PAs, 

like senkirkine, erucifoline, retrorsine and seneciphylline were also found in small amounts in the in 

vivo rumen fluids of the three species. Since these PAs were also detectable in the plasma (unpublished 

data) but were under the limit of detection in the meat from sheep and goat, rumen metabolism probably 

is not solely responsible for the selective PA transfer into muscle. A study on laying hens, fed with 



Summarizing discussion 

   20 

Jacobaea vulgaris Gaertn., showing also transfer of selected PAs like jacoline into muscle, supports the 

hypothesis, that also the liver metabolism influences PA transfer into meat (Mulder et al., 2016). It is 

extensively proven that pyrrolizidine alkaloids like the ones in Jacobaea vulgaris Gaertn. are 

metabolized during the biotransformation in the liver of species like huma, rats, cattle and sheep 

(IPCS, 1988; Wiedenfeld and Edgar, 2011). Geburek et al. (2020) demonstrated that individual PAs are 

metabolized at different rates in experiments with rat and human microsomes, with jacoline being barely 

eliminated. In own studies with bovine and sheep microsomes, we also observed such differences 

(unpublished data), also suggesting that the liver metabolism has an effect on the selective transfer from 

PAs into muscle meat of cattle, sheep and goat. 

With PA concentrations in meat from sheep and goats that were prematurely withdrawn from the PA-

SAFE-FEED feeding study due to severe health reactions, and incorporating data from the dose-finding 

study conducted prior to this feeding study, we were able to demonstrate that the PA content in the 

muscle meat of the animals decreased rapidly and that almost no PAs were detectable after 48 hours 

(publication 2). However, this data is not suitable for precise kinetic evaluations, and further 

investigations are required to provide a more detailed analysis. 

Interestingly, PA metabolites identified in the rumen were found only in trace amounts in muscle, milk, 

and plasma samples, yet were present in extensive quantities in feces (unpublished data). This suggests 

that these metabolites are unable to cross the intestinal barrier, which is unexpected given their proposed 

structural similarity to PAs. Own Ussing chamber experiments with epithelia from different parts of the 

ruminal gastrointestinal tract, showed comparable permeability for PA/PANOs and their rumen 

metabolites (unpublished data). Buchmueller et al. (2022) showed that in human Caco-2 cells, 

representing the intestinal barrier, echimidine and intergerimine (open diester and monoester) passed the 

barrier less efficient compared to monocrotaline, senkirkine, senecionine or retrorsine (cyclic diesters) 

(Buchmueller et al., 2022). As mentioned earlier we found indications that the cyclic diesters are 

structurally changed in the rumen of the animals. Although the fragmentation patterns of the metabolites 

do not indicate a cleavage in the necic acid, it cannot be ruled out that some kind of cleavage occurs 

leading to steric hindrance, probably leading to a different behavior in the intestinal tract. Further studies 

are needed to fully understand these processes. 

The metabolic processes occurring in the rumen have a substantial impact on the toxicity of PA/PANOs 

in animals. When ruminants consume plants that produce PA/PANOs, their livers are exposed to a 

different PA load compared to monogastric animals. PAs that are rapidly metabolized in the rumen tend 

to be quickly processed in the liver as well (publication 1, Geburek et al. 2020). Therefore, with 

eliminating these reactive PAs the rumen metabolism results in a PA load that is less reactive during 

hepatic metabolism. As a result, ruminants encounter a PA load with reduced toxic potential, as PAs that 

are less metabolically active in the liver produce fewer pyrrolic metabolites, which are associated with 

genotoxic and carcinogenic effects. This might give an explanation for the lower sensitivity of ruminants 

to the toxicity of PA/PANOs. However, there are also differences in sensitivity among ruminant species. 



Summarizing discussion 

   21 

Studies have shown that sheep and goats can tolerate higher amounts of Jacobaea vulgaris Gaertn., and 

therefore higher levels of PA/PANOs, compared to cattle ((Goeger et al., 1982b; Mortimer and 

White, 1975)). In terms of ruminal metabolism, one could speculate a higher ruminal transformation 

rate of PA/PANOs for animals that are more robust, hence for sheep and goat. However, our studies did 

not provide data for a clear explanation for the varying sensitivities among these ruminants. In the PA-

SAFE-FEED study, sheep and goats were given higher doses of PA/PANOs compared to cattle. Despite 

this, sheep showed lower concentrations of PA/PANOs and their metabolites in rumen fluid than cattle, 

whereas goats had higher concentrations (publication 2). It is important to note that comparing 

PA/PANO concentrations and their metabolites in the rumen is challenging due to anatomical and 

physiological differences, such as relative rumen volume, saliva production, and rumen passage rates of 

ingested material. An adjusted study design that accounts for these factors might help clarify the role of 

the rumen concerning the different sensitivities. Other studies have estimated that goats and sheep have 

two to three times the number of PA metabolizing bacteria compared to cattle, suggesting higher 

metabolic rates in the rumens of these more robust species (Wachenheim et al. 1992). Craig et al. (1986) 

showed that sheep cope better with intravenously administered PA/PANOs compared to cattle, 

indicating that different levels of liver activity in the animals may also play a role for their sensitivity. 

Own experiments showed that bovine microsomes barely eliminated the studied PA/PANOs, while 

sheep microsomes partially metabolized them (unpublished data). As increased elimination is associated 

with increased formation of toxic pyrrolic metabolites, it remains unclear why sheep showed a higher 

resistance during the study of Craig et al. (1986). Further research is needed to better understand the 

precise mechanisms and differences between species. 

Our study identified jacoline, jaconine, and jacobine as the predominant pyrrolizidine alkaloids present 

in the meat samples. Similarly, Knoop et al. (2024) reported that these three PAs were the primary 

compounds transferred to cattle milk in the PA-SAFE-FEED feeding studies. Compared to the original 

plant material, the animal's body appears to reduce the overall PA/PANO burden, effectively filtering 

out PAs such as senecionine, seneciphylline, erucifoline, riddelliine, and retrorsine, since some of these 

filtered compounds are known to be activated in the liver, forming toxic pyrrolic metabolites. Even 

though jacobine, which is transferred to the muscle and meat of the animals, also was found to produce 

significant amounts of pyrrolic metabolites, while jacoline did not form these metabolites (Geburek 

et al., 2020). Nevertheless, in general the toxic PA burden is reduced by the animal body raising the 

question of whether such PA burdens should be considered less concerning. Currently, the classifies all 

PAs as equally potent regarding their toxicity. However, there are studies showing different hepatic 

reactivity, that is why several studies propose to use potency factors for PAs (Allemang et al., 2018; Frei 

et al., 1992; Haas et al., 2023; Lester et al., 2019; Merz and Schrenk, 2016). Some feeding studies also 

have shown that calves or rats fed milk from cattle that consumed high levels of Jacobaea vulgaris 

Gaertn. experienced little to no adverse effects. This raises the question if jacoline, the primary PA 

transferred to milk, may be less toxic than other PAs in the plant (Johnson, 1976; Miranda et al., 1981). 



Summarizing discussion 

   22 

Contrary, Goeger et al. (1982b) observed adverse effects in rats fed contaminated milk and those fed 

equivalent amounts of Jacobaea vulgaris Gaertn., suggesting that the pyrrolizidine alkaloids PAs 

transferred to the milk are as potent as those found in the plant. 

Beside filtering reactive PAs, we showed that the metabolic activities in the ruminants also lead to 

transfer of only small amounts of PAs into meat. The PA concentrations found in the meat of animals 

receiving the highest PA/PANO dosages in the PA-SAFE-FEED study were used to calculate a 

theoretical MOE (EFSA, 2005). This calculation utilized a BMDL10 of 237 μg/kg bw/day as the 

toxicological reference point for PAs and considered estimated consumption of cattle and sheep meat 

(EFSA, 2011; EFSA et al., 2017). Goat meat was not considered due to its low consumption. For cattle 

and sheep meat, MOEs were found to be around or above 10,000 (publication 2). It has to be considered 

that animals in the PA-FEED-SAFE trial received their PA/PANO administration once or twice a day, 

probably influencing the transfer. In a natural scenario, animals consume PA/PANOs gradually 

throughout the day with their feed, probably leading to more efficient metabolism. Therefore, even if 

the animals ingest PA/PANOs at levels as high as the highest dosage used in our study, the MOE should 

remain above 10,000. Moreover, with average PA/PANO concentrations of 0.29 mg/kg in forage and 

roughage, and even lower levels in silages, the PA/PANO contamination in feed across Europe 

theoretical results in lower doses than those used in our study (Bolechová et al., 2015; EFSA, 2011; 

Gottschalk et al., 2015; Mulder et al., 2009). Consequently, margin of exposure values below 10,000 

are questionable. According to the risk assessment concept, an MOE of 10,000 or higher is considered 

to be of low concern from a public health perspective and is deemed a low priority for risk management 

actions. Additionally, it is important to consider that muscle tissue measured in the PA-SAFE-FEED 

study may undergo processing or aging before being consumed, which could potentially lower the PA 

concentration further. 

When evaluating our feeding study, it is crucial to consider several factors to accurately translate the 

findings to husbandry conditions. Research by Johnson et al. (1984, 1985) highlights that the toxic 

effects of PA/PANOs are influenced by the mode of administration. Their studies demonstrated that 

toxicity varies depending on whether PA/PANOs are given as a single bolus or distributed throughout 

the day. In natural settings, ruminants consume their feed gradually, allowing for continuous rumination 

and digestion over an extended period. This results in the gradual absorption of PA/PANOs, unlike the 

typical single-dose administration in feeding studies. Johnson et al. (1984, 1985) concluded that the 

rapid absorption from a one-shot bolus leads to more pronounced toxic effects. In contrast, the slow 

intake of PA/PANOs in the rumen likely facilitates more effective metabolism and detoxification. 

Therefore, studies involving ruminants must take these effects into account and design their experiments 

accordingly. Theoretically, the form in which the PA/PANO bolus is administered could also influence 

toxicity. For instance, administering the dose as intact plant parts might result in a slower release 

compared to using plant extracts, where the PA/PANOs are already dissolved. However, the literature 

does not provide clear evidence to support this hypothesis (Johnson and Smart, 1983).  



Summarizing discussion 

   23 

In the PA-SAFE-FEED feeding studies, PA/PANOs were administered as extracts via a stomach tube 

once per day. Based on the observations of Johnson et al. (1984, 1985), we assume that the chosen 

method of administration allowed more PA/PANOs to pass unmetabolized through the rumen compared 

to ingestion of whole plants. Consequently, it is hypothesized that if the chosen doses had been 

administered under natural feeding conditions, they would have resulted in less health effects and 

transfer rates. Despite these considerations, feeding studies must balance multiple requirements. From 

an analytical perspective, using an extract is advantageous as it enables the creation of a homogeneous 

mixture, ensuring that all animals receive identical amounts and compositions of PA/PANOs during the 

feeding trial. Achieving such homogeneity with Jacobaea vulgaris Gaertn. plants would not have been 

feasible. Additionally, administering the dose as plant material could have posed challenges for small 

ruminants due to the volume required, potentially impacting their feeding behavior and acceptance. 

However, it was subsequently found that using extracts also presented problems. The extraction process 

likely altered the PA/PANO profile. The extract contained more free bases than are typically found in 

plants, which usually have a PA/PANO ratio of about 0.1, whereas the extract had a ratio of 2.0. Since 

PANOs are very quickly reduced to PAs in the rumen of animals, this difference can be neglected. 

However, the extract also contained higher concentrations of jaconine compared to plants, which could 

have potentially influenced the transfer and toxicological effects. 

In general, animal studies are difficult to compare. Even under similar conditions, differences in 

observations can occur (Johnson and Smart, 1983). For example, two studies that used comparable 

PA/PANO doses over a similar period of time as the PA-SAFE-FEED study, reported much stronger 

symptoms, including death. This was observed even in one study where PA/PANOs were administered 

throughout the day as plant pellets (Johnson and Smart, 1983; Thorpe and Ford, 1968). 

Currently, maximum levels of PA/PANOs in foodstuffs are regulated only for certain plant-based foods 

and dietary supplements. According to Regulation (EU) 2023/915, these foods must be tested for 35 

specific PA/PANOs. This selection primarily includes PA/PANOs that are prevalent in various plant 

species from different families, aiming to cover the PA/PANO contamination in foods with a manageable 

number of individual substances. Contaminations with Jacobaea vulgaris Gaertn. and Senecio plants 

can be detected by monitoring PA/PANOs such as retrorsine, senecionine, and seneciphylline, along 

with their corresponding N-oxides. If efforts are made to protect consumers from PA/PANO 

contamination in animal-derived products, the existing scope would not be useful because the activities 

of the rumen and liver lead to an altered PA-profile.  

In conclusion, it can be debated whether ruminants can be fed contaminated feed or grazed on pastures 

containing Jacobaea vulgaris Gaertn. Two aspects need to be considered: animal welfare and consumer 

protection. According to §3 of the animal welfare act, it is prohibited to feed an animal with substances 

that cause pain, suffering, or harm. In the highest dosage group of the PA-SAFE-FEED cattle experiment 

enzyme levels increased over the four weeks of the trial, indicating the beginning of liver damage 

(Knoop et al., 2023). Knoop et al. (2023) suggest that liver damage could also occur in animals from 



Summarizing discussion 

   24 

lower dosage groups when exposed to PA/PANOs over extended periods. As a result, the authors 

recommend that cattle exposure to PA/PANOs be avoided. In the PA-SAFE-FEED study, sheep and 

goats exhibited more pronounced reactions, with several animals needing to be excluded from PA/PANO 

dosing prematurely (unpublished data). Interestingly, Ohlsen et al. (2022) reported that all sheep 

remained healthy, despite potentially higher doses and longer exposure times. This raises the possibility 

that under real-world conditions, equivalent doses used in the PA-SAFE-FEED project may not lead to 

the same level of harm or suffering in animals. Additionally, it is important to consider that PA/PANO-

producing plants, such as Jacobaea vulgaris Gaertn., are seasonal, raising questions about the feasibility 

of chronic exposure. For instance, the highest dosage in the PA-SAFE-FEED cattle trial corresponds to 

the consumption of 20-30 Jacobaea vulgaris Gaertn. plants per day per animal. 

Evaluating the risk to ruminants from PA/PANO exposure is complex, given the multiple factors 

involved, such as animal health, ecological considerations, and other interests. It is crucial for decision-

makers to carefully examine the data, including those generated in this doctoral thesis and the PA-SAFE-

FEED project, and to assess it objectively. This will enable informed, beneficial decisions to be made 

that prioritize both animal health and consumer safety. 

Most of the scientific research on pyrrolizidine alkaloids, particularly on ruminal metabolism and 

feeding studies, was conducted between 1960 and 1990. Several factors may explain why only a few 

groups continued to focus on this topic after that period. One likely reason might be, that with the 

knowledge gained during those decades, improved forage and grassland management practices were 

implemented, reducing the incidence of PA intoxication in ruminants. As a result, scientific attention 

shifted to other areas of interest. The decline in feeding studies can also be attributed to stricter ethical 

standards, which have made it more challenging to obtain approval for such experiments. However, it 

remains intriguing that research on ruminal metabolism was not fully extended to cyclic diesters PAs. 

These compounds are present in some of the plants crucially affecting the health of grazing animals, and 

thus, they should have been of considerable interest. 

However, the present study, could elucidate the fate of these pyrrolizidine alkaloids from Jacobaea 

vulgaris Gaertn. ingested by ruminants. It was found that highly toxic PAs are primarily metabolized 

into non-toxic compounds in the rumen of these animals. Consequently, the transfer of these PAs in the 

tissues of the animals, such as meat, is low. Given the minimal carry-over, it can be assumed that, 

considering the observed contamination levels of feed with pyrrolizidine alkaloids in Europe, there is a 

low risk to consumers from PA/PANOs in animal-derived food products.  

  



References 

   25 

5. References 
 

Aguiar, R. and Wink, M. (2005), “Do naïve ruminants degrade alkaloids in the rumen?”, Journal of 

Chemical Ecology, Vol. 31 No. 4, pp. 761–787. 

 

Allemang, A., Mahony, C., Lester, C. and Pfuhler, S. (2018), “Relative potency of fifteen pyrrolizidine 

alkaloids to induce DNA damage as measured by micronucleus induction in HepaRG human liver cells”, 

Food and Chemical Toxicology, Vol. 121, pp. 72–81. 

 

Allgaier, C. and Franz, S. (2015), “Risk assessment on the use of herbal medicinal products containing 

pyrrolizidine alkaloids”, Regulatory Toxicology and Pharmacology, Vol. 73 No. 2, pp. 494–500. 

 

Anholt, H. and Britton, A. (2017), “Presumptive chronic pyrrolizidine alkaloid poisoning in 2 pygmy 

goats due to ingestion of tansy ragwort (Jacobaea vulgaris) in southwestern British Columbia”, The 

Canadian Veterinary Journal = La Revue Veterinaire Canadienne, Vol. 58 No. 11, pp. 1171–1175. 

 

Anjos, B.L., Nobre, V.M.T., Dantas, A.F.M., Medeiros, R.M.T., Oliveira Neto, T.S., Molyneux, R.J. 

and Riet-Correa, F. (2010), “Poisoning of sheep by seeds of Crotalaria retusa: Acquired resistance by 

continuous administration of low doses”, Toxicon, Vol. 55 No. 1, pp. 28–32. 

 

Baker, D.C., Pfister, J.A., Molyneux, R.J. and Kechele, P. (1991), “Cynoglossum officinale toxicity in 

calves”, Journal of Comparative Pathology, Vol. 104 No. 4, pp. 403–410. 

 

Barri, M.E., Adam, S.E. and Omer, O.H. (1984), “Effects of Crotalaria saltiana on Nubian goats”, 

Veterinary and Human Toxicology, Vol. 26 No. 6, pp. 476–480. 

 

BfR. (2007), “Salatmischung mit Pyrrolizidinalkaloidhaltigem Greiskraut verunreinigt - Stellungnahme 

028/2007 des BfR vom 10. Januar 2007”, available at: 

https://bfr.bund.de/cm/343/salatmischung_mit_pyrrolizidinalkaloid_haltigem_geiskraut_verunreinigt.

pdf (accessed 20.09.2024). 

 

BfR. (2020), “Aktualisierte Risikobewertung zu Gehalten an 1,2-ungesättigten Pyrrolizidinalkaloiden 

(PA) in Lebensmitteln: Stellungnahme Nr. 026/2020 des BfR vom 17. Juni 2020”, available at: 

https://bfr.bund.de/cm/343/aktualisierte-risikobewertung-zu-gehalten-an-1-2-ungesaettigten-

pyrrolizidinalkaloiden-pa-in-lebensmitteln.pdf (accessed 20.09.2024). 

 

 



References 

   26 

Bolechová, M., Čáslavský, J., Pospíchalová, M. and Kosubová, P. (2015), “UPLC–MS/MS method for 

determination of selected pyrrolizidine alkaloids in feed”, Food Chemistry, Vol. 170, pp. 265–270. 

 

Brauchli, J., Lüthy, J., Zweifel, U. and Schlatter, C. (1982), “Pyrrolizidine alkaloids from Symphytum 

officinale L. and their percutaneous absorption in rats”, Experientia, Vol. 38 No. 9, pp. 1085–1087. 

 

Brumme, S. (2015), "Jakobsgreiskraut (Senecio jacobaea) als Giftpflanze im norddeutschen Grünland: 

Bewertung von Verbiss und Aufnahme durch Nutztiere bei Beweidung" (Master's Thesis), Hochschule 

Bremen. 

 

Buchmueller, J., Kaltner, F., Gottschalk, C., Maares, M., Braeuning, A. and Hessel-Pras, S. (2022), 

“Structure-dependent toxicokinetics of selected pyrrolizidine alkaloids in vitro”, International Journal 

of Molecular Sciences, Multidisciplinary Digital Publishing Institute, Vol. 23 No. 16, p. 9214. 

 

Bull, L.B., Culvenor, C.C.J. and Dick, A.T. (1968), "The pyrrolizidine alkaloids", Amsterdam, North 

Holland Publishing Co. 

 

Cameron, E. (1935), “A study of the natural control of ragwort (Senecio jacobaea L.)”, Journal of 

Ecology, Vol. 23 No. 2, pp. 265–322. 

 

Candrian, U., Lüthy, J. and Schlatter, C. (1985), “In vivo covalent binding of retronecine-labelled 

[3H]seneciphylline and [3H]senecionine to DNA of rat liver, lung and kidney”, Chemico-Biological 

Interactions, Vol. 54, pp. 57–69. 

 

Candrian, U., Zweifel, U., Luethy, J. and Schlatter, C. (1991), “Transfer of orally administered 3H-

seneciphylline into cow’s milk”, Journal of Agricultural and Food Chemistry, Vol. 39 No. 5, pp. 930–

933. 

 

Chen, T., Mei, N. and Fu, P.P. (2010), “Genotoxicity of pyrrolizidine alkaloids”, Journal of Applied 

Toxicology, Vol. 30 No. 3, pp. 183–196. 

 

Chen, Z. and Huo, J.-R. (2010), “Hepatic veno-occlusive disease associated with toxicity of 

pyrrolizidine alkaloids in herbal preparations”, The Netherlands Journal of Medicine, Vol. 68 No. 6, pp. 

252–260. 

 

Cooper, R.A. and Huxtable, R.J. (1996), “A simple procedure for determining the aqueous half-lives of 

pyrrolic metabolites of pyrrolizidine alkaloids”, Toxicon, Vol. 34 No. 5, pp. 604–607. 



References 

   27 

Craig, A., Blythe, L. inda, Lassen, E. and Slizeski, M. (1986), “Resistance of sheep to pyrrolizidine 

alkaloids”, Israel Journal of Veterinary Medicine, Vol. 42 No. 4, pp. 376–384. 

 

Craig, A.M., Latham, C.J., Blythe, L.L., Schmotzer, W.B. and O’Connor, O.A. (1992), “Metabolism of 

toxic pyrrolizidine alkaloids from tansy ragwort (Senecio jacobaea) in ovine ruminal fluid under 

anaerobic conditions”, Applied and Environmental Microbiology, Vol. 58 No. 9, pp. 2730–2736. 

 

Culvenor, C.C.J. (1985), “Pyrrolizidine alkaloids: some aspects of the Australian involvement”, Trends 

in Pharmacological Sciences, Vol. 6, pp. 18–22. 

 

Culvenor, C.C.J., Jago, M.V., Peterson, J.E., Smith, L.W., Payne, A.L., Campbell, D.G., Edgar, J.A. 

and Frahn, J.L. (1984), “Toxicity of Echium plantagineum (Paterson’s Curse). 1. Marginal toxic effects 

in Merino wethers from long-term feeding”, Australian Journal of Agricultural Research, Vol. 35 No. 2, 

pp. 293–304. 

 

Culvenor, C.C.J., Edgar, J.A., Jago, M.V., Outteridge, A., Peterson, J.E. and Smith, L.W. (1976), 

“Hepato- and pneumotoxicity of pyrrolizidine alkaloids and derivatives in relation to molecular 

structure”, Chemico-Biological Interactions, Vol. 12 No. 3, pp. 299–324. 

 

Damir, H.A., Adam, S.E.I. and Tartour, G. (1982), “The effects of Heliotropium ovalifolium on goats 

and sheep”, British Veterinary Journal, Vol. 138 No. 6, pp. 463–472. 

 

Deinzer, M.L., Arbogast, B.L., Buhler, D.R. and Cheeke, P.R. (1982), “Gas chromatographic 

determination of pyrrolizidine alkaloids in goat’s milk”, Analytical Chemistry, Vol. 54 No. 11, pp. 1811–

1814. 

 

De Nijs, M., Mulder, P.P., Klijnstra, M.D., Driehuis, F. and Hoogenboom, R.L. (2017), “Fate of 

pyrrolizidine alkaloids during processing of milk of cows treated with ragwort”, Food Additives & 

Contaminants: Part A, Vol. 34 No. 12, pp. 2212–2219. 

 

Dick, A.T., Dann, A.T., Bull, L.B. and Culvenok, C.C.J. (1963), “Vitamin B12 and the detoxification 

of hepatotoxic pyrrolizidine alkaloids in rumen liquor”, Nature, Vol. 197 No. 4863, pp. 207–208. 

 

Dickinson, J.O. (1980), “Release of pyrrolizidine alkaloids into milk”, Proceedings of the Western 

Pharmacology Society, Vol. 23, pp. 377–379. 

 

 



References 

   28 

Dickinson, J.O., Cooke, M.P., King, R.R. and Mohamed, P.A. (1976), “Milk transfer of pyrrolizidine 

alkoloids in cattle”, Journal of the American Veterinary Medical Association, Vol. 169 No. 11, pp. 

1192–1196. 

 

Duringer, J.M., Buhler, D.R. and Craig, A.M. (2004), “Comparison of hepatic in vitro metabolism of 

the pyrrolizidine alkaloid senecionine in sheep and cattle”, American Journal of Veterinary Research, 

Vol. 65 No. 11, pp. 1563–1572. 

 

Dusemund, B., Nowak, N., Sommerfeld, C., Lindtner, O., Schäfer, B. and Lampen, A. (2018), “Risk 

assessment of pyrrolizidine alkaloids in food of plant and animal origin”, Food and Chemical 

Toxicology, Vol. 115, pp. 63–72. 

 

Eastman, D.F., Dimenna, G.P. and Segall, H.J. (1982), “Covalent binding of two pyrrolizidine alkaloids, 

senecionine and seneciphylline, to hepatic macromolecules and their distribution, excretion, and transfer 

into milk of lactating mice”, Drug Metabolism and Disposition: The Biological Fate of Chemicals, 

Vol. 10 No. 3, pp. 236–240. 

 

EFSA. (2005), “Opinion of the Scientific Committee on a request from EFSA related to a harmonised 

approach for risk assessment of substances which are both genotoxic and carcinogenic”, EFSA Journal, 

Vol. 3 No. 10, p. 282. 

 

EFSA. (2011), “Scientific opinion on pyrrolizidine alkaloids in food and feed”, EFSA Journal, Vol. 9 

No. 11, p. 2406. 

 

EFSA. (2017), “Risks for human health related to the presence of pyrrolizidine alkaloids in honey, tea, 

herbal infusions and food supplements”, EFSA Journal, Vol. 15 No. 7, p. e04908. 

 

Estep, J.E., Lamé, M.W., Morin, D., Jones, A.D., Wilson, D.W. and Segall, H.J. (1991), 

“[14C]monocrotaline kinetics and metabolism in the rat”, Drug Metabolism and Disposition: The 

Biological Fate of Chemicals, Vol. 19 No. 1, pp. 135–139. 

 

Estep, J.E., Lamé, M.W. and Segall, H.J. (1990), “Excretion and blood radioactivity levels following 

[14C]senecionine administration in the rat”, Toxicology, Vol. 64 No. 2, pp. 179–189. 

 

 

 

 



References 

   29 

Fletcher, M.T., McKenzie, R.A., Reichmann, K.G. and Blaney, B.J. (2011), "Risks from plants 

containing pyrrolizidine alkaloids for livestock and meat quality in Northern Australia", pp. 208–218 in 

Riet-Correa, F., Pfister, J., Schild, A and Wierenga, T., "Poisonings by plants, mycotoxins and related", 

United Kingdom, CABI Publishing. 

 

Ford, E.J.H., Ritchie, H.E. and Thorpe, E. (1968), “Serum changes following the feeding of ragwort 

(Senecio jacobea) to calves”, Journal of Comparative Pathology, Vol. 78 No. 2, pp. 207–218. 

 

Frei, H., Lüthy, J., Brauchli, J., Zweifel, U., Würgler, F.E. and Schlatter, C. (1992), “Structure/activity 

relationships of the genotoxic potencies of sixteen pyrrolizidine alkaloids assayed for the induction of 

somatic mutation and recombination in wing cells of Drosophila melanogaster”, Chemico-Biological 

Interactions, Vol. 83 No. 1, pp. 1–22. 

 

Fu, P.P., Xia, Q., Lin, G. and Chou, M.W. (2004), “Pyrrolizidine alkaloids — genotoxicity, metabolism 

enzymes, metabolic activation, and mechanisms”, Drug Metabolism Reviews, Vol. 36 No. 1, pp. 1–55. 

 

Geburek, I., Preiss-Weigert, A., Lahrssen-Wiederholt, M., Schrenk, D. and These, A. (2020), “In vitro 

metabolism of pyrrolizidine alkaloids – metabolic degradation and GSH conjugate formation of 

different structure types”, Food and Chemical Toxicology, Vol. 135, p. 110868. 

 

Gilruth, J.A. (1903), "Hepatic cirrhosis affecting horses and cattle (“so-called Winton Disease”): A 

report on its nature, cause, treatment, distribution, etc.", Wellington, Department of Agriculture. 

 

Gilruth, J.A. (1905), “Hepatic cirrho