Poisoning of sheep by Vernonia plantaginoides (Less.) Hieron in Uruguay

Abstract

Vernonia plantaginoides (Less.) Hieron, previously known as Vernonia squarrosa, is a rhizomatous subshrub with purple flowers that is prevalent in the natural grassland of Uruguay, Argentina, and southern Brazil. We report an outbreak of V. plantaginoides (yuyo moro) intoxication in sheep in Treinta y Tres Department, northeastern Uruguay. A total of 54 of 463 (12%) recently weaned lambs died 2–7 days after entering a natural pasture that had been invaded by sprouting V. plantaginoides. The first cases were found dead. Affected lambs showed marked jaundice, edema of the face, ears, and eyelids, and severe photodermatitis. At the autopsies of 3 lambs, the carcass was yellow, the liver was enlarged with a marked acinar pattern (“nutmeg liver”), and hemorrhages were observed on serous membranes. Microscopic lesions were characterized by diffuse periacinar hepatocellular necrosis and cholemic nephrosis. Three female lambs were experimentally dosed with the aerial parts of V. plantaginoides collected immediately after the outbreak. The lamb that was dosed once with 40 g/kg body weight died after 36 h with severe hepatic necrosis. The lamb dosed with 20 g/kg daily for 4 days showed clinical signs and microscopic lesions in the liver with multiple apoptotic hepatocytes in the periacinar zone. The third lamb, dosed with 30, 17, and 15 g/kg daily over 3 days, respectively, showed transient clinical signs and a rise in liver enzymes, but recovered, and no lesions were found postmortem. These results demonstrate that V. plantaginoides was responsible for severe field outbreaks of poisoning in sheep in Uruguay.

Introduction

Vernonia Schreb (family Asteraceae) is a large genus of perennial forbs and shrubs with red or purple flowers, distributed mostly in the tropics and subtropics of Asia, Africa, and America. ²² Some species in North America are given the name “ironweed” because of their stout stems that often persist throughout the winter. In South America, the genus is represented by ~350–400 species that grow mostly in southeastern Brazil, northern Argentina, Uruguay, Paraguay, and Bolivia. ⁷ The Vernonia genus has been reported to possess a plethora of bioactive compounds with medicinal properties, ³⁰ and some are edible and have nutritional value. ¹² A few—all in South America—are recognized as being toxic to livestock. Among these, Vernonia mollissima ⁹ and especially Vernonia rubricaulis^4,27 are reported to cause outbreaks of hepatotoxicity in cattle in central-western Brazil. Mortality occurs annually during the late dry or early wet season when the plants are sprouting and other green forage is scarce, when the stocking rate is high, or after moving naive cattle from other ranches. ¹⁷ Poisoning by V. rubricaulis has been induced experimentally in cattle and sheep,^4,20 whereas V. mollissima has been confirmed as hepatotoxic in cattle, sheep, goats, and rabbits.^13,25,29 However, cases of natural poisoning by V. rubricaulis and V. mollissima have never been reported in sheep (R Lemos, pers. comm., 2015), despite this species being, as are cattle, highly susceptible to both plants. ⁴ Another species of Vernonia, V. nudiflora, in Rio Grande do Sul, Brazil, has been experimentally proven to cause gastrointestinal disorders in cattle and sheep, but given its low palatability and toxicity, it is unlikely to cause spontaneous poisoning in livestock. ⁸

Vernonia plantaginoides (Less.) Hieron, previously known as V. squarrosa (Less.) Lessing, is a perennial, rhizomatous subshrub ~60–80 cm in height, with stout stems branched at the top, ascending linear leaves of 5–12 cm long and 1–6 mm wide, with revolute margins and white tomenta on the back, and purple inflorescences (Fig. 1). ⁵ Given its striking purple flowers, in Uruguay it is commonly known as yuyo moro (purple weed), or mio-mio moro (purple mio-mio), due to its similarity with Baccharis coridifolia (mio-mio). The plant grows in the hilly natural grassland of northeastern Uruguay and adjacent provinces of Argentina and southern Brazil. ⁵ The reserve compounds in the underground organs enable the rapid development of new shoots when the aerial parts die or are damaged by drought or frost, or are lost during the dormancy stage of the plant in winter, ¹ so it is one of the first plants to green after rain in spring and autumn, or as soon as the weather is hot and humid (i.e., practically anytime during the year in the temperate climate of Uruguay).

Figure 1.

Flowering Vernonia plantaginoides (commonly known as yuyo moro) in northeastern Uruguay, March 2012.

Poisoning of sheep by V. plantaginoides has been known for decades by local people, farmers, and veterinarians in northeastern Uruguay. In some rural communities, folk stories commonly relate losses as high as 3,000–6,000 sheep in the 1960s during the shearing season in spring (August–November) or after weaning in summer or fall (January–March). A similar or identical local problem in sheep was investigated in the 1960s–1970s in the municipality of Uruguaiana, southern Brazil, close to the Uruguay frontier. ²⁸ The outbreaks were attributed to alecrim or Vernonia squarrosa (i.e., V. plantaginoides) because it was the main weed found in the affected farms and because the plant was experimentally lethal to sheep and cattle in doses of 30–50 g/kg body weight (bw); however, spontaneous cases were never seen, and the importance of the plant under natural conditions remained undetermined. We report an outbreak of spontaneous intoxication by V. plantaginoides (Less.) Hieron in sheep in northeastern Uruguay. The clinical findings and pathology from experimental administration of V. plantaginoides to sheep are also described.

Materials and methods

Spontaneous cases

The outbreak occurred in March 2012 on a mixed sheep and cattle farm situated on Las Pavas hill (latitude: 33°18′S; longitude: 54°94′W), Treinta y Tres Department, northeastern Uruguay. It affected a flock of 463 recently weaned crossbred Merino lambs (5–6 months of age) that were introduced into a large paddock of native pasture that had been invaded by vegetative and early flowering V. plantaginoides. Clinical, epidemiologic, and management data were collected during visits to the farm. Thorough autopsies and complete histologic evaluation were performed on 3 affected sheep (sheep 1–3), the first a few hours postmortem and the other 2 following euthanasia with intravenous (IV) sodium pentobarbital and exsanguination. Tissue samples, including from the central nervous system, were fixed in 4% buffered formaldehyde, processed routinely, and 5-µm sections cut and stained with Mayer hematoxylin and eosin (HE) for histology.

Experimental poisoning

The experiment was carried out according to the recommendations of the Uruguayan Honorary Commission of Animal Experimentation (CHEA). Fresh plants of V. plantaginoides were collected from the paddock where lambs died during the outbreak. In the laboratory, leaves and young stems were separated, chopped, and refrigerated at 4°C until dosing. Three female weaned lambs (4–6), 4–5 months of age, were used for experimental reproduction of the disease. Animals were kept grazing in a small paddock with water ad libitum during the experimental period and were given 5–7 days to acclimate to the test environment and to facilitate their subsequent clinical evaluation. Following 6–8 h of fasting (food but not water), the animals were weighed and the aerial parts of the plant were administered orally by hand. A single oral dose of 40 g/kg bw was administered to lamb 4 (Corriedale, 18 kg bw), whereas lamb 5 (crossbreed Merino, 23 kg bw) received 4 consecutive doses of 20 g/kg bw at 0, 24, 48, and 72 h (total dose of 80 g/kg bw), and lamb 6 (Corriedale, 24 kg bw) received 3 decreasing doses of 30, 17, and 15 g/kg at 0, 24, and 72 h (total dose of 62 g/kg bw). The lambs were bled, urine-sampled, and clinically examined twice a day prior to dosing to establish a baseline and every 4–6 h thereafter. Their behavior, heart rate (determined by auscultation), respiratory rates (abdominal-costal movements), and rectal temperature were recorded.

Serum activities of alkaline phosphatase (ALP), aspartate aminotransferase (AST), and γ-glutamyl transferase (GGT) were determined using routine laboratory techniques at the Laboratory of Clinical Analysis, Veterinary Faculty, Montevideo, Uruguay. Urine samples were subjected to physical examination and semiquantitative determinations of protein, glucose, specific gravity, pH, bilirubin, urobilinogen, ketones, and hemoglobin using a strip reagent. ^a Clinical and biochemical data are presented as means ± standard deviation. A routine postmortem examination was performed on all lambs immediately following death or after euthanasia with an IV injection of sodium pentobarbital. Tissue samples were fixed in formalin and processed as above for the spontaneous cases.

Identification of the plant

Samples of the whole plant were sent for botanical identification to Dr. Eduardo Alonso, Department of Botany, Faculty of Chemistry, and to Dr. Ana C. González, Herbarium Bernardo Rosengurtt, Faculty of Agronomy, Montevideo, Uruguay.

Results

Spontaneous cases

Sixty lambs were affected and 54 died 2–7 days after entering the pasture. The first cases were found dead. Later, clinically affected animals showed depression, recumbency, and moderate to severe edema over the face, ears, and eyelids, with concurrent jaundice and congestion of the skin, sclera, and mucous membranes (Fig. 2). Cattle and horses in the same pasture were not affected.

Figure 2.

Marked edema and hyperemia of the face, ears, and eyelids can be seen in spontaneously affected lamb 2.

In severe cases, the swelling involved the submandibular region and extended down to the throat and upper neck. Most lambs died at this stage, but survivors showed photosensitization, with cracking and sloughing of the skin of the muzzle, face, eyelids, and the tips of the ears. At autopsy, the carcasses of cases 1–3 were icteric with yellow fluid under the skin and in the abdominal, pleural, and pericardial cavities. There were ecchymoses and petechiae on the serous membranes, mainly of the epicardium, tiny ulcers in the abomasal mucosa, and dried and mucoid content in the colon and rectum. The gallbladder was distended and edematous, the kidney cortex was pale yellow, and the urine appeared dark. The liver, which was consistently affected in all animals, was enlarged and had a diffuse mottled appearance with red, depressed areas intercalated with a pale yellow network of parenchyma (“nutmeg liver”; Fig. 3).

Figure 3.

The liver of lamb 2 is enlarged and has a marked acinar pattern (“nutmeg liver”) and edematous gallbladder.

Histologically, the changes in the liver were characterized by periacinar to midzonal coagulative and hemorrhagic necrosis, slight intrahepatic cholestasis, and swollen, vacuolated periportal hepatocytes. In cases 1 and 2, the hemorrhagic areas were confluent (bridging necrosis), whereas in case 3, the necrosis was centrilobular only. Marked pulmonary hyperemia, epicardial hemorrhages, and moderate cholemic nephrosis were found in cases 1 and 2. Chitinous body fragments or heads of the sawfly Perreyia flavipes (family Pergidae) larvae were not found in the reticulum, rumen, or omasum. Microscopic lesions were not found in any animals in the splenic white pulp or mesenteric lymph nodes, central nervous system, or other organs.

Experimental poisoning

Clinical signs

In lamb 4 (single dose of 40 g/kg bw), the heart rate (from 100 ± 17 predosing to 160 beats/min at the end of dosing), respiratory rate (79 ± 11 to 180 breaths/min), and rectal temperature (39.9°C ± 0.2°C to 40.9°C) increased during the administration of the plant (Fig. 4). Clinical signs were first seen 4–12 h after the end of dosing. The lamb stopped eating and appeared listless and indifferent to the environment. Respiratory rate dropped strikingly, and remained shallow and significantly slower than the predosing level until death at 36 h after ingestion (42 vs. 79 breaths/min, T = −9.4, p = 0.000, df = 9). By 12–16 h, the lamb appeared weak but was walking around. At 21–25 h, it was depressed and reluctant to move, with marked tachycardia (148–166 beats/min), cyanotic mucous membranes, and mild bloat. By 30–36 h, the lamb was in sternal recumbency with the head forward and the chin on the ground, and the heart rate and rectal temperature began to decline. It had opisthotonos, leg paddling, and muscle twitching, dying just before it could be euthanized.

Figure 4.

Changes in respiratory rate, heart rate, and rectal temperature in lamb 4 experimentally challenged (time = 0) with a single dose of 40 g/kg body weight of Vernonia plantaginoides. Bar = administration period; arrow = sternal recumbency; cross = death.

Lamb 5 (4 daily doses of 20 g/kg bw) became listless and partially anorexic 6–8 h after the first dose. The following morning, the lamb appeared bright and alert, with normal appetite, but signs worsened and became more apparent after the second dose. After the third dose, the lamb gradually developed mild facial and ear edema, reddening of eyelids, engorged episcleral vessels, and scant serous ocular and nasal discharge. After the last dose, depression deepened, the heart rate was elevated, respiration was rapid, there was rumen atony, and slight abdominal distension was observed; the lamb remained standing for long periods with the head lowered, or lying down with the head turned against its side, although it stood when prompted. Rectal temperature was not altered. The animal was euthanized 24 h after the last dose. Lamb 6 (3 decreasing doses) had only mild clinical signs. After the first and second doses, the lamb appeared somewhat depressed and inappetent, had a respiratory rate of 40–45 breaths/min, moderate hyperthermia (40.4–40.6°C), and tachycardia (144 beats/min). Appetite was regained by 72 h and thereafter appeared clinically normal. It was euthanized 24 h after the last dose.

Serum and urine biochemistry

Figure 5 depicts the changing serum enzyme activities of lamb 4. Predosing activities of AST, GGT, and ALP were within published reference intervals: 258 ± 10.2, 50 ± 2, and 290 ± 18 U/L, respectively. ¹⁶ By 12 h post-dosing, there was a dramatic rise in the serum activity of AST (1,710 U/L) and GGT (197 U/L), reaching 9,160 and 345 U/L at 30 h, respectively, when the lamb was recumbent. Changes in the activity of ALP were similar to those for AST and GGT, but of lower magnitude, reaching 7 times (2,840 U/L) the basal activity 6 h before death. Mild elevation of GGT occurred in lamb 5 at 12–24 h after each dosing (59–62 U/L), whereas AST (380 U/L) increased in lamb 6 at 31 h after the first dose (6 h after the second dose), returning to the pre-dosing activity (141 ± 21 U/L) by 44 h. The urine of lamb 4 was bilirubin positive at 16 h after dosing and strongly positive at 25 and 30 h. Proteinuria was first detected by 16 h (150–1,000 mg/L), whereas pH dropped markedly from 8 to 5 just before death. In lambs 5 and 6, proteinuria was detected 24 h after initial dosing and remained positive until death.

Figure 5.

Changes in activity of the enzymes AST, ALP, and GGT in serum from experimental lamb 4 (single dose of 40 g/kg body weight of Vernonia plantaginoides). Bar = administration period.

Pathology

Macro- and microscopic lesions of the 3 experimental cases were similar to those of the spontaneous cases. The liver of lamb 4 (40 g/kg) was swollen with rounded edges and had a marked “nutmeg” appearance; on cut surface, it had irregular, depressed, bright-red areas that were regularly intermingled with a network of yellow bulging parenchyma (Fig. 6). The gallbladder was distended and edematous. There was mild jaundice, multiple subendocardial and subepicardial petechial hemorrhages, and a large amount of yellow fluid in the pericardial sac and less in the pleural and abdominal cavities. Kidneys were pale and moist with a yellow pelvis, and the lungs had severe congestion and edema, frothy fluid on cut surface, and dense foam filling the tracheal and bronchial lumens. In lamb 5, the liver was pale and had a slight, barely visible, red mottled appearance on cut surface. There was scarce pleural and pericardial serous effusion, and dry content in the colon and rectum. The only recognized gross abnormality in lamb 6 was a slightly pale liver.

Figure 6.

The cut surface of the liver of lamb 4 has a nutmeg appearance.

The liver of lamb 4 had severe centrilobular hemorrhagic necrosis with marked vacuolation of midzonal and periportal hepatocytes (Fig. 7). In the lung, there was alveolar hyperemia, interstitial and alveolar edema, and discrete air bubbles in the alveolar lumens. The kidneys had marked hydropic degeneration of tubular epithelial cells, mainly in the proximal convoluted tubules, mild interstitial edema, and proteinaceous fluid in the tubular lumen and Bowman space. Marked microscopic lesions were found in the central nervous system. There were multiple perivascular hemorrhages in the cerebral cortex, particularly at the junction of the gray and white matter, and less frequently in the thalamus, midbrain, medulla oblongata, and obex. Moderate swelling of astrocytes and dilation of the pericapillary spaces were evident in the subcortical white matter. No microscopic lesions were found in other organs. In lamb 5, there was dissociation of hepatic laminae, microhemorrhages around the hepatic venules, vacuolated hepatocytes, and multiple apoptotic hepatocytes in the periacinar and midzonal zones. There was pulmonary hyperemia and moderate hydropic degeneration of renal tubules. No significant histologic lesions were found in lamb 6.

Figure 7.

Centrilobular hemorrhagic necrosis is observed around a centrilobular vein (upper right corner) in the liver of lamb 4 (40 g/kg body weight). Hematoxylin and eosin. 100×.

Identification of the plant

All of the collected plants were identified as V. plantaginoides (Less.) Hieron. Voucher specimens are kept in the herbarium of the Faculty of Chemistry, identified as MVFQ 4381.

Discussion

Clinical signs, gross and histopathology of the liver, and the presence of young defoliated plants of V. plantaginoides in the affected paddock strongly suggested intoxication by this plant. Feeding experiments showed that the plant was hepatotoxic and lethal to lambs, thus confirming that V. plantaginoides was the cause of the spontaneous disease. The outbreak was short and severe, killing 54 lambs in only 5 days, which is in line with old local stories describing devastating mortalities associated with yuyo moro (V. plantaginoides) in northeastern Uruguay, and alecrim (V. squarrosa) in southern Brazil. ²⁸ The reason why such a historic and severe disease was not diagnosed and reported earlier is unknown; possible explanations include the limited access of veterinarians and farmers to a diagnostic laboratory (the East Regional Laboratory was founded in 1989), the short duration of the outbreaks, poor monitoring of extensively managed flocks, and the secondary role of sheep in the farming economy. The marked decline in sheep numbers and value after the collapse of the wool price in the 1990s may have also contributed to the delay in diagnosis.

The present outbreak was epidemiologically and pathologically almost indistinguishable from the poisoning caused by the South American sawfly P. flavipes. Both diseases occur as a point epidemic with many animals found dead within a few days, while clinical signs of jaundice and photosensitization appear late in the outbreak, and the main autopsy finding is a “nutmeg liver” due to severe, diffuse, centrilobular hepatocellular necrosis.^10,26 Furthermore, the present outbreak was diagnosed in a county with a highly significant geographical clustering of sawfly poisoning in Uruguay (Dutra F, et al. Descriptive statistics and spatiotemporal analysis of bovine hepatotoxic diseases diagnosed in Uruguay. Proceedings of the 8th International Symposium on Poisoning Plants, 2009 May 4–8, João Pessoa, Paraíba, Brazil), which can be explained by the fact that V. plantaginoides and P. flavipes larvae coexist on the same hilly grassland ecosystem of northeastern Uruguay and southern Brazil.^22,23 Although the present outbreak occurred in late summer to early autumn and sawfly poisoning occurs mostly in winter (June–September), ¹⁰ both diseases may overlap in time because cases of P. flavipes poisoning can occur in early autumn ¹⁰ and V. plantaginoides (V. squarrosa) poisoning is traditionally reported in August–November during shearing. ²⁸ Thus, differential diagnosis between these diseases must rest on a careful postmortem examination, looking for chitinous fragments and larval heads with typical sensory hairs in the ruminal content ²⁴ and, harder to notice, lymphocytolysis in the spleen and other lymphoid organs, which is widespread in sawfly poisoning^10,26 and minimal or absent in V. plantaginoides. Hepatogenous photosensitization in sheep grazing plants containing lithogenic saponins produces rather similar clinical signs (“yellow big head”) both in eastern Uruguay (Heliotropium ocellatum, unpublished data) and southern Brazil (Brachiaria spp.), ²¹ but this possibility was ruled out by the microscopic absence of crystalloid material in the liver and bile ducts. There were no Cestrum parqui, Wedelia glauca, Xanthium spp., or other common local hepatotoxic plants in the affected paddock; nor did the lambs have access to clover pastures or any other type of feed, ruling out chronic copper poisoning.

An important factor in the outbreak reported herein was the access of hungry or stressed and recently weaned lambs to an area of the paddock where large amounts of V. plantaginoides were sprouting. In those conditions, they ate large quantities of young plants in a short period of time. It would appear that V. plantaginoides is harmful at a certain stage of growth and that certain climatic and management factors are necessary for the disease to occur. It is apparently consumed by sheep only when fleshy, basal leaves are developing, and becomes rapidly unpalatable when it is growing or is fully grown and fibrous leaves and stems predominate. This is supported by the characteristic experience that the disease stops spontaneously after a few days or weeks, and also suggests that it would be a good practice to avoid grazing for a time (e.g., 15–30 days) or to move sheep to another pasture as soon as the disease appears. In the present outbreak, only sheep were affected despite cattle also grazing in the same paddock when the disease occurred. This difference in susceptibility could be caused by different grazing behavior, because cattle—in contrast to sheep—tend to avoid the steep or hilly areas where V. plantaginoides predominates, or prefer different plants in the pasture or, more likely, because cattle have a greater and more variable threshold level of toxicity than sheep. ²⁸ Thus, both species may be at risk, and although the plant is quite bitter tasting and typically is not eaten except under dire circumstances, sheep are most likely to eat the sprouting plants and are at greatest risk.

The experimental toxic dose of V. plantaginoides in the present study was relatively high, in line with a previous report. ²⁸ A single dose of 40 g/kg was lethal in 36 h, whereas 35 g/kg or 20 g/kg bw were found to be poisonous after the first administration, but not lethal even when administered for 3 or 4 consecutive days, suggesting a noncumulative effect of the plant. Thus, large quantities of plants sprouting at once are necessary to attain toxic levels in the invaded pasture in addition to hungry or naive animals and a high stocking rate in order for the outbreak to occur. In contrast, other closely related hepatotoxic plants, such as V. rubricaulis ⁴ and V. mollissima, ¹³ are lethal to cattle and sheep in single doses of only 2–10 g/kg and 10–20 g/kg bw, respectively.

The toxic compound of V. plantaginoides is unknown. Phytochemical investigations of Vernonia spp., both in America and Africa, have resulted in the isolation of numerous classes of cytotoxic sesquiterpene lactones with potent biological activities.^18,30 Glaucolides, hirsutinolide derivatives, and related sesquiterpene lactones have been isolated from South American poisonous V. nudiflora and V. mollissima.^3,6 However, despite the hundreds of terpenoids isolated from the Vernonia genus, only a limited number have been tested for toxicity, and it is not certain which phytochemical is responsible for the hepatotoxic effects. There are reports of hepatotoxicity in mice by alcohol extracts of V. amigdalina, ¹⁴ and in rats by extracts of V. colorata. ¹⁵ In contrast, the terpenoid fraction of the plants also seems to be hepatoprotective and is able to reduce acetaminophen- and carbon tetrachloride–induced hepatotoxicity in rats.^2,11 However, it has been shown in rats that the hepatic enzymes significantly increase in a time- and dose-dependent manner in both the low- and high-dose groups when compared with the control group. ¹⁹ This seems to indicate that when consumed in large quantities, the plants may indeed elicit hepatotoxicity, as occurred in the present outbreak.