Abstract
Rayless goldenrod (Isocoma pluriflora) sporadically poisons livestock in the southwestern United States. Similarities with white snakeroot (Ageratina altissima) poisoning and nearly identical chemical analyses led early researchers to conclude that tremetol, a mixture of benzofuran ketones, is the rayless goldenrod toxin. The toxicity of these ketone toxins have not been fully characterized nor are the pathogenesis and sequelae of poisoning completely understood. The objective of the current study was to characterize and describe the clinical and pathologic changes of rayless goldenrod toxicity in goats. Fifteen goats were gavaged with rayless goldenrod to obtain benzofuran ketone doses of 0, 10, 20, 40, and 60 mg/kg/day. After 7 treatment days, the goats were euthanized, necropsied, and tissues were processed for microscopic studies. After 5 or 6 days of treatment, the 40-mg/kg and 60-mg/kg goats were reluctant to move, stood with an erect stance, and became exercise intolerant. They had increased resting heart rate, prolonged recovery following exercise, and increased serum aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and creatinine kinase activities. All treated animals developed skeletal myopathy with dose-related distribution and severity. The goats dosed with 20 mg/kg and higher also developed myocardial degeneration and necrosis. Although skeletal myonecrosis was patchy and widely distributed, the quadriceps femoris was consistently damaged, even in low-dosed animals. Myocardial lesions were most severe in the papillary muscles of 60-mg/kg—dosed animals. This indicates that goats are highly susceptible to rayless goldenrod poisoning, and that the characteristic lesion of poisoning is skeletal and cardiac myonecrosis.
Keywords
Introduction
Rayless goldenrod or southern jimmyweed [Isocoma pluriflora (Torr. & A. Gray) Greene (Asteraceae) previously Isocoma wrightii (A. Gray) Rydb and Haplopappus heterophyllus (A. Gray) S.F. Blake] is a toxic range plant that sporadically poisons livestock in Arizona, Colorado, New Mexico, and Texas (Fig. 1). Poisoning was first described in the early 1900s and was erroneously associated with bad water and called “alkali disease.” 10,11 Poisoned cattle and horses developed violent trembling that was most evident when animals were forced to move or were agitated. Most references lack explanation of why poisoned animals tremble, and they do not describe morphologic lesions that might be the cause. Even the current texts report that the histologic lesions of rayless goldenrod poisoning are “not distinctive” or are characterized by mild to moderate renal nephrosis and hepatic fatty change. 4,8
About the same time that Marsh described clinical rayless goldenrod poisoning in the Pecos, Texas, area, white snakeroot (Ageratina altissima var. altissima or Eupatorium rugosum) was found to be responsible for causing “trembles” in livestock and “milk sickness” in humans. 10,11 As the poisoning and lesions were similar, researchers assumed that the toxins were the same. In 1927, tremetol, a dark tar-like substance, was isolated from rayless goldenrod. 5,6 Recent work demonstrated that rayless goldenrod tremetol is a mixture of benzofuran ketones, including tremetone, dehydrotremetone, 3-oxyangeloyl-tremetone, and 3-hydroxytremetone. 7,9
Poisoning by rayless goldenrod usually occurs during fall and early winter when frosts may make rayless goldenrod more palatable, when other forages have been depleted, or when snow makes alternative forages less accessible. Poisoning has been reported in horses, cattle, pigs, sheep, goats, and humans. 4 Clinically, poisoning is seen as depression, anorexia, and inactivity followed by muscle tremors especially following exercise or activity. Some poisoned animals may develop tachypnea, tachycardia, with ascites and hydrothorax. As tremetol, or one of its metabolites, is excreted in milk, secondary poisoning of nursing neonates can occur with reduced, or without apparent, maternal toxicity. 4
The objectives of the current study are to characterize and describe the toxicity of rayless goldenrod with known quantities of benzofuran ketones in goats. Later, the authors hope to use this model to study the toxicity and toxicokinetics of individual rayless goldenrod ketones and to study the long-term sequelae of poisoning.
Rayless goldenrod (Isocoma pluriflora) is an erect, 30–120-cm-tall, bushy perennial that arises from a woody root stalk. It is unbranched or sparsely branched with sticky, alternate, linear leaves. Between 7 and 15 yellow flowers form heads that form flat clusters at the tips of the stems. It commonly grows in alkaline and gypsic soils in riparian zones along river valleys, drainage areas, or dry plains in southern Colorado, Texas, New Mexico, and Arizona.
Photomicrograph of the biceps brachii of a goat treated with rayless goldenrod (Isocoma pluriflora) to obtain a benzofuran ketone dose of 60 mg/kg per body weight for 7 days. Note the extensive edema (*), myonecrosis with clumping of sarcoplasmic proteins (arrowhead), and monocytic inflammation with nuclear proliferation (arrow). Hematoxylin and eosin. Bar = 60 μm.
Photomicrograph of a papillary muscle in the left ventricle of a goat treated with rayless goldenrod (Isocoma pluriflora) to obtain a benzofuran ketone dose of 40 mg/kg per body weight for 7 days. Note the focally extensive myonecrosis with clumping of sarcoplasmic proteins (arrowhead), and monocytic inflammation (arrow). Hematoxylin and eosin. Bar = 30 μm.
Photomicrograph of the liver of a goat treated with rayless goldenrod (Isocoma pluriflora) to obtain a benzofuran ketone dose of 60 mg/kg per body weight for 7 days. Note the expansion of hepatocyte cytoplasm with vacuolation (arrow) and eosinophilic debris (arrowhead). Hematoxylin and eosin. Bar = 30 μm.
Materials and methods
Plant material
Rayless goldenrod was collected in Pecos City, Texas (31°23′56″N, 103°29′58″W). The plant was taxonomically identified as rayless goldenrod (Isocoma pluriflora) and saved as vouchers 250012 and 250014 at the Intermountain Herbarium at Utah State University (Logan, Utah). The plant was air-dried, and a day before the beginning of the study, it was ground to pass through a 2.38-mm screen and thoroughly mixed to ensure homogeneity. Concentrations of the benzofuran ketone compounds (tremetone, dehydrotremetone, 3-hydroxytremetone, and 3-oxyangeloyl-tremetone) were determined using previously described techniques. 9
Animals
Fifteen, yearling, female Spanish goats weighing 29.3 ± 3.7 kg (mean ± standard deviation) were randomly divided into 5 groups with 3 animals per group. The animals were trained to lead and to run on a treadmill for 3 weeks before the start of the study. The day before the initial dosing, all animals were weighed, bled by jugular venipuncture, and exercised on a treadmill while their electrocardiograms (ECGs) were monitored and recorded. For 7 consecutive days, the goats were dosed intraruminally using an oral speculum and a 1-cm gastric tube with ground rayless goldenrod to obtain dosages of 10, 20, 40, and 60 mg benzofuran ketones/kg body weight (BW)/day. The control group was given similar volumes of ground alfalfa and/or grass hay using the same dosing method. Water and long-stem alfalfa hay were available ad libitum throughout the study. The animals were monitored and exercised, and serum and whole blood were collected daily. After 7 days of dosing, all the goats were euthanized and necropsied. At necropsy, samples of left lateral retroocular, tongue, masseter, superficial pectoral, triceps, intercostals, longissimus dorsi, semitendinosus, diaphragm, biceps femoris, biceps brachii, quadriceps femoris, gluteus medius, psoas major, adductor, and semimembranosus skeletal muscles were collected, attached to wooden tongue depressors, and fixed in 10% neutral buffered formalin. The heart was opened, cleaned with water, closely examined, fixed intact, and sectioned to examine portions of the right atrium, the right papillary muscle, the right free ventricular wall, the septum, the left atrium, the left papillary muscle, and the left ventricular free wall. Other tissues, including brain, spinal cord, lung, liver, right and left kidney, adrenal gland, urinary bladder, thyroid gland, lymph node, esophagus, rumen, omasum, abomasum, duodenum, pancreas, jejunum, ileum, cecum, and colon, were collected, fixed, and prepared for examination. Tissues were processed, sectioned, and stained using standard histologic techniques. Special stains of specific skeletal muscle, myocardium, and liver lesions included diastase-positive and diastase-resistant periodic acid—Schiff (PAS), Masson trichrome stain for collagen, and Congo red. Lesions were scored by distribution (percentage of tissue affected) and graded by the severity of lesion (0 = none, 1 = minimal [loss of striation and hypereosinophilia], 2 = mild [sarcoplasmic clumping with myocyte swelling], 3 = moderate [sarcomere disruption with focal monocytic inflammation and mild nuclear proliferation], 4 = severe [extensive inflammation, regeneration with fibrosis]).
Serum and electrocardiogram analyses
Hematology, serum biochemistries, and serum electrolyte analyses were performed using standard techniques with an automated hematology analyzer a and a biochemistry analyzer. b Reagents and methodology recommended by the manufacturers were used. Exercise tolerance and ECGs were carried out as previously reported. 7
Statistical analyses
All results from the current study were analyzed as a completely random design with 5 treatments. The model included treatments (0, 10, 20, 40, and 60 mg benzofuran ketones/kg BW), with individual animals nested within treatments. The hematology and serum biochemistry were analyzed as repeated measures over the multiday experiment. Animals were a random factor in the mixed linear model analysis using the procedures of SAS. c The primary independent variable in the present study was the dosage of benzofuran ketone compounds (mg/kg BW). The variance—covariance matrix was chosen by an interactive process wherein the best fit was based on the Schwarz Bayesian criterion. The unstructured and compound symmetry covariance models were often the best-fitting structures. Treatment differences were separated using predicted difference (PDIFF option in SAS c ) for significant interactions (P < 0.05) in the model. Preplanned, pairwise comparisons between treatment least square means were considered different at P < 0.05. Histologic scores were compared using analysis of variance (SAS c -GLM). Significant means (P < 0.05) were separated using Duncan's method.
Results
The ground rayless goldenrod contained about 0.21% benzofuran ketones that were composed of 12.5% tremetone, 32% dehydrotremetone, and 55.5% 3-oxyangeloyl-tremetone with no 3-hydroxytremetone. At this concentration, most of the doses given to the goats were between 150 g and 900 g or between 0.5% and 3% of their body weight.
After 4 or 5 days of dosing, 3 goats dosed at 40 mg/kg and 2 goats dosed at 60 mg/kg became reluctant to move, preferred to remain recumbent, and were exercise intolerant. When the affected goats were forced to stand, they would do so with post-like, straight legs, their back flexed in a humped-up position, with their tails flexed vertically. Often, the large appendicular muscles were swollen and firm. Rarely, affected animals developed prominent exophthalmia with minimally delayed pupillary reflexes. When standing, the goats would tremble and would quickly tire and lay down. Affected goats developed increased resting heart rate, decreased appetite, and when exercised, they had prolonged recovery times. More details of the exercise physiology and electro-cardiographic findings have been previously reported. 7 All of the goats dosed with 10 and 20 mg/kg BW were clinically healthy through the first 6 days of the study. However, on the final day, 1 goat dosed at 10 mg/kg and 2 goats dosed at 20 mg/kg had decreased exercise tolerance with increased resting heart rates and prolonged postexercise recovery times.
Hematology and serum enzyme changes
There were no dose-related hematology changes in any of the groups or times. However, several serum biochemical indicators of cellular damage (creatinine kinase [CK], aspartate aminotransferase [AST], alanine aminotransferase [ALT], and lactate dehydrogenase [LDH]) were increased (Table. 1). A dose by day interaction (P = 0.007) was clearly evident as the goats dosed at 40 mg/kg and 60 mg/kg differed on days 6 and 7. These increased activities were also different (P < 0.05) for those measured at previous times. Cardiac troponin-I concentrations were variable, and there were no significant group differences. Interestingly, 3 goats that were dosed at 40 mg/kg and that developed both gross and histologic heart lesions on day 7 had increased serum cardiac troponin-I concentrations of 0.11, 5.22, and 0.13 ng/ml (normal cardiac troponin-I concentrations are undetectable <0.02). Two of the 60-mg/kg—dosed goats with significant myocardial necrosis had increased cardiac troponin-I concentrations of 4.05 and 0.34 ng/ml (see Table 1).
Selected mean serum biochemical data from control goats and goats dosed with rayless goldenrod (Isocoma pluriflora) to obtain benzofuran ketone doses of 0, 10, 20, 40, and 60 mg/kg body weight for 7 days.*
Different means (<0.05) between groups are indicated with superscript letters.
Estimates of normal range were determined as 2 standard deviations from mean values of control goats and pretreatment samples. These ranges are probably laboratory and assay specific.
Cardiac troponin-I concentrations below detection limits are reported as <0.02 ng/ml.
At necropsy, goats dosed with 40 and 60 mg/kg had swelling and pallor of nearly all skeletal muscles. The distribution of these lesions was patchy, and no distinct pattern or muscle groups were consistently damaged. However, the large appendicular muscles (semimembranosus, semitendinosus, biceps femoris, gluteus medius, quadriceps femoris, and triceps brachii) were affected most often. The hearts of most of the 40 and 60-mg/kg—dosed animals had soft, pale streaks throughout the myocardium. The liver of these animals was swollen and red. No significant gross lesions were found in the control goats or the 10 and 20-mg/kg—dosed goats.
Histologic studies found that all 6 of the 60 and 40-mg/kg—dosed goats developed extensive monophasic degeneration and necrosis of skeletal muscle (Fig. 2). The lesions were most severe (scores of 3 and 4 of 4) and widely distributed (involving between 30–50% of the myocytes) in the large appendicular muscles (see Table 2). The severity and distribution of the skeletal muscle degeneration and necrosis was related to dose (see Table 2). The degeneration was typically segmental muscular necrosis characterized by loss of striation, hypereosinophilia, clumping and disruption of sarcoplasmic contents, monocytic inflammation with macrophage phagocytosis of debris, and increased numbers of myocyte nuclei seen as prominent nuclear rowing (Fig. 2). In these foci of myocyte necrosis, there was minimal proliferation of fibroblasts with small amounts of collagen deposition. Many of the myofibrils were surrounded by edema. The 20- and 10-mg/kg—dosed animals had similar but less severe and more infrequent myocyte necrosis. In these animals, the skeletal muscle lesions were most often found in the quadriceps femoris, diaphragm, retroocular muscles, and tongue. Many of the larger appendicular and axial muscles from animals in these groups were histologically normal. No lesions were identified in the skeletal muscles of the control animals.
Histologic score of lesion severity and distribution in skeletal muscle of goats dosed with rayless goldenrod (Isocoma pluriflora) to obtain benzofuran ketone dosages of 0, 10, 20, 40, and 60 mg/kg body weight for 7 days.
The first number is the histologic score (mean and standard deviation). The histologic score = distribution × lesion severity or grade. This is followed parenthetically by the highest percentage of tissue involved and (after a slash, /) the highest histologic grade (0 = none, 1 = minimal, 2 = mild, 3 = moderate, 4 = severe). Means that are significantly different (P < 0.05) are indicated with superscript letters.
The myocardium of 2 of the 60-mg/kg—dosed goats and all 3 of the 40-mg/kg—dosed goats also had multifocal myocyte swelling, degeneration, and, rarely, necrosis (Fig. 3). These degenerative changes included loss of striation, sarcoplasmic hypereosinophilia, and coagulation and clumping of sarcoplasmic material. Affected myocytes are occasionally collapsed, making the nuclei appear more numerous. In all of these animals, the cardiac lesions were mild, with low distributions (<5%) when compared with the pathologic changes in the skeletal muscles. Although lesions were present in other portions of the heart, the ventricular papillary muscles often had the most extensive change (Fig. 3). Mild degenerative changes were consistently found in the atria of both the high-dosed groups (see Table 2). One of the 20-mg/kg—dosed goats also developed papillary muscle lesions of similar severity and distribution to the higher-dosed groups. Other than that 1 animal, the lesions in the hearts from the 20-mg/kg—dosed group were rare and consisted of mild myocyte hypereosinophilia and loss of striation. No histologic lesions were found in the sections of heart from the control goats and 10-mg/kg—dosed animals.
The hepatocytes of the 20-, 40-, and 60-mg/kg—dosed goats were swollen with large, eosinophilic, cytoplasmic aggregates and rare cytoplasmic vacuoles (Fig. 4). Often the swelling obliterated the sinusoids and obscured the cord structure. The eosinophilic material was positive for PAS and negative for Congo. In several of the 60-mg/kg—dosed animals, there were rare, patchy, centrilobular necroses. In these animals, the adjacent central veins were also dilated and filled with blood. No significant histologic lesions were detected in liver sections from the control and 10 mg/kg—dosed animals. No significant lesions were detected in the kidneys or other tissues from any of the control or treatment groups.
Discussion
Rayless goldenrod is a known toxic plant and has a history of sporadically poisoning livestock for more than 100 years. 4,10 Various toxins have been described in rayless goldenrod. 3,5,6,9,16 In addition to the previously mentioned tremetol, 5,6 several monoterpenes, sesquiterpenes, 16 and essential oils 3 have also been described. In vitro studies using a tremetol mixture suggests that at least 1 component of this mixture or of its metabolites inhibits oxidative phosphorylation, most likely by inhibiting the tricarboxylic acid cycle. 15 As poisoning is reported to cause a variety of clinical and histologic disease in different species, additional work is needed to determine which compounds are toxic and in what species.
The sporadic nature of rayless goldenrod poisoning makes predicting the risk of poisoning (and when poisoning may occur) difficult. It has been suggested that plant doses of 1–1.5% of an animal's body weight over a 1–3-week period are toxic. 8 In the current study, doses of 10 mg/kg benzofuran ketones or about 0.5% of the body weight for 7 days produced histologic lesions in skeletal muscle. This suggests that a goat would have to ingest about 150 g (approximately 10% of the daily diet) for 7 days. Plant collections have widely different toxin profiles, and as has been suggested in the past, certain plant populations have varying toxin concentrations. Some populations may even be nontoxic. 13 The quantitative high-performance liquid chromatography analysis of the benzofuran ketone compounds in rayless goldenrod and white snakeroot will be useful in identifying toxic populations or subspecies and could help to better predict the risk of poisoning by the rayless goldenrod species as well as by white snakeroot. 9
Even though it has been suggested that the benzofuran ketones are the toxic components of rayless goldenrod and white snakeroot, the toxicities of the individual benzofuran compounds have not been studied. A previous study demonstrated that in vitro microsome-activated tremetone is cytotoxic, but dehydrotremetone was not. 2,3 This suggests that an “activated” tremetone metabolite is responsible for white snakeroot toxicity. However, this has not been confirmed experimentally.
In the present study, serum enzyme activities were found to be significantly changed and were different from control values between days 5 and 7 for CK, LDH, AST, and ALT (Table. 1). These changes directly correlated with the clinical and histologic findings because goats that had clinical disease and extensive histologic findings also had increased serum enzyme activities. Although the mean cardiac troponin-I concentrations were not different, individual animals had elevated amounts. Troponin-I is a subunit of a regulatory protein of the thin filament of striated muscle. Cardiac troponin-I is unique from other forms of striated muscle, and for nearly a decade, it has been used as a sensitive and specific biomarker of myocardial damage in many species. 1 There is evidence that some immunoassays for human cardiac troponin-I are poor indicators of bovine cardiac troponin-I. 14 However, other studies have been shown to be ultrasensitive in measuring cardiac troponin-I in various species. 12 Additional comparison is needed to determine whether these cardiac troponin-I concentrations correlate with the degree and severity of myocardial damage. The specificity of this particular assay and how well it cross-reacts with caprine cardiac troponin-I also needs to be determined.
The histologic skeletal myonecrosis was monophasic and acute, with minimal nuclear proliferation and fibrosis (Fig. 2). The minimal fibroblast proliferation and collagen accumulation were likely early fibrosis; however, they may have been due to myocyte loss and collapse and to aggregation of the supporting connective tissue stroma. This interpretation is difficult because only a single duration was examined. Longer exposure durations and recovering animals are likely to develop much more severe inflammation and fibrosis. Additional studies are needed to better document this progression and to better compare these lesions with those caused by other nutritional and toxic etiologies.
In the goats in the current study, higher doses (>1.0% of BW) were required to develop cardiac lesions (Fig. 3). It has been suggested that horses are more susceptible to the cardiotoxic effects of white snakeroot and rayless goldenrod. 4,8 It may be that higher doses over shorter exposure durations are required to be cardiotoxic. Certainly, more information is needed to determine whether there are species differences in susceptibility. Interestingly, little is known about the sequelae of poisoning. As subclinical poisoning is likely, especially in nursing neonates, it may be that the effect of poisoning is much larger than current estimates. Maternal exposure and subsequent fetal or neonatal cardiotoxicity may result in permanent heart damage that could permanently impair subsequent production and performance.
At these doses and durations, the hepatocellular swelling and eosinophilic aggregates seen in the liver of many medium- and high-dosed goats appear to be degenerative lesions. The hepatocellular swelling did not seem to progress or to become more severe with increasing dose. As these changes were found in animals with and without cardiac lesions, they are probably not related to impaired cardiac function. The special stains suggest the eosinophilic material was composed of aggregates of glycogen, cellular organelles, and debris. Ultrastructural studies and recovery studies to better characterize these lesions are underway by the authors. Several of the 60-mg/kg—dosed goats also had mild, patchy, hepatic, centrilobular degeneration with minimal central vein dilation. The patchy distribution and the lack of correlation with the extent or severity of the cardiac lesions suggest these lesions are not related to generalized heart failure. In any case, the 60-mg/kg—dosed goats did not develop the extensive hepatocellular necrosis or renal nephrosis reported previously in goats fed white snakeroot (Reagor JC, Jones LP, Ray AC, et al.: 1989, Eupatorium rugosum, white snakeroot poisoning in goats. In: Proceedings of the Third International Symposium on Poisonous Plants, p. 98, July 24–28, Logan, UT) or rayless goldenrod. 4 A likely explanation is that this rayless goldenrod sample lacks specific hepatotoxic benzofuran ketones or the proper ketone ratio to produce this response.
One goat from the 60-mg/kg—dosed group appeared to be relatively resistant to poisoning because she was only mildly affected and developed relatively mild histologic lesions. Others have also reported that with repeated doses, animals may become tolerant, 4 which is confusing because others have suggested that the effects of these benzofuran ketones are cumulative. 4,8 An alternative explanation is that tolerance may also be related to physiologic states, such as nutritional status and antioxidant potential. Research continues to define the toxicokinetics of the rayless goldenrod toxins to determine the effect of time, duration, and physiologic state on poisoning.
In conclusion, the current study demonstrated that the primary lesion and cause of the clinical trembles in rayless goldenrod intoxication in goats is skeletal muscle degeneration and necrosis. In goats, relatively small amounts of rayless goldenrod (about 0.5% of BW) given for 7 days will produce clinical disease and histologic skeletal muscle degeneration and necrosis. As these subtle lesions are most consistently seen in the quadriceps femoris and diaphragm at this low dose, these may be the best tissues to examine. However, if there are clinical signs of poisoning and lesions are severe, the large appendicular muscles, such as the triceps brachii, biceps femoris, quadriceps femoris, and adductor, are likely to show significant histologic change. Higher doses produced myocardial lesions. Many questions of rayless goldenrod poisoning remain unanswered, and additional studies are needed to determine the relative toxicity of individual benzofuran ketones, to better define the specific pathogenic mechanism of their toxicity, and to determine the effect and long-term sequelae of poisoning.
Acknowledgements
The authors thank Ed Knoppel, Joseph Jacobson, and Katie Lott for their assistance in animal care and laboratory technical expertise. The authors also thank Dr. Kerry Rood, Dr. Steve Hooser, and Lynette Harris for their constructive review of this manuscript. This research was partially supported by the Utah Agriculture Experiment Station, Utah State University and approved as journal paper 8170.
Footnotes
a.
Beckman Coulter Inc., Fullerton, CA.
b.
Hitachi 7180, Hitachi High Technologies Inc., Pleasanton, CA.
c.
Version 9.1 for Windows, SAS Institute Inc., Cary, NC.