HEPATOCELLULAR TOXICITY OF THE METABOLITE EMODIN PRODUCED BY THE COMMON BUCKTHORN (RHAMNUS CATHARTICA) IN GREEN FROG (LITHOBATES CLAMITANS) TADPOLES

The common buckthorn (Rhamnus cathartica), introduced to North America from Europe in the early 1800s, has become naturalized through most of north, east, and central North America (Knight et al. 2007; Kurylo et al. 2007). Because of its affinity for disturbed areas and its tolerance to both drought and flooded conditions, this invasive species has had a substantial effect on the local flora and fauna. Like other members of the Rhamnaceae family, the common buckthorn produces allelopathic substances to avoid competition from other plants as well as secondary compounds in their leaves, bark, and fruits to avoid predation by insects and other organisms (Knight et al. 2007). Among these compounds is the secondary metabolite emodin (Izhaki 2002; Tsahar et al. 2002), which has been shown to produce deformities and mortality in invertebrates (Roth et al. 2015), fish (He et al. 2012; Cui et al. 2014), and amphibian larvae (Sacerdote and King 2014; Chen et al. 2017).

Emodin belongs to the largest chemical group of natural quinones, known as anthraquinones, with a basic chemical structure of a tricyclic aromatic ring with two ketone groups in positions C₉, C₁₀ (1,3,8-trihydroxy-6-meth-ylanthraquinone). In mammals, emodin has been shown to have laxative, antiallergenic, antiinflammatory, antineoplastic, and antidiabetic uses (Abu Eid et al. 2017; Çandöken et al. 2017). However, research in mammals and other groups shows that exposure to high concentrations of emodin can result in oxidative stress in the liver resulting in inflammation, hepatocellular congestion, and necrosis as well as damage to lipids, DNA, and proteins (Lichtensteiger et al. 1997; Ma et al. 2016; Monisha et al. 2016; Abu Eid et al. 2017). Reports from fish (e.g., grass carp, Ctenopharyngodon idellus) noted that exposure to high concentrations of emodin resulted in an increase of reactive oxygen species, leading to the inhibition of oxidation enzymes, affecting the subcellular redox equilibrium and resulting in apoptosis of hepatic cells (Dong et al. 2016; Li et al. 2017). In fully aquatic amphibians, long-term exposure to compounds similar to emodin (e.g., atrazine, simazine) has been reported to result in the formation of nuclear and cytoplasmic inclusions in the liver leading to necrosis, atrophy, and vacuolization of hepatocytes (Zaya et al. 2011; Sai et al. 2016).

The effect of emodin in amphibian larvae liver function is not fully understood. Sacerdote and King (2014) reported mortality of western chorus frogs (Pseudacris triseriata) and African clawed frogs (Xenopus laevis) after exposure to emodin at an early developmental stage (stages 23–25; Gosner 1960). Tadpoles in early development stages died within 5 d of exposure to the toxin; however, the cause of death or the effect of the chemical on detoxifying organs such as kidney or liver was not reported. Moreover, their results were limited to the effect of emodin in early development tadpoles, when the kidney and liver are not fully developed (Altig and McDiarmid 1999; Ultsch et al. 1999).

We report changes in the livers of green frog (Lithobates clamitans) tadpoles exposed to high concentrations of emodin during advanced stages of development (stages 30–35; Gosner 1960). Because the function of the liver in tadpoles changes from hematopoietic during early stages to metabolic during later stages of development (Orkin 1996; Okui et al. 2016), late-stage tadpoles are expected to have a higher capacity for processing complex compounds such as emodin (and thus a greater chance for surviving exposure), but potentially suffering hepatic damage as seen for other groups (Chang et al. 2011; Li et al. 2017).

To determine the effect of emodin on green frogs (L. clamitans), we conducted controlled laboratory experiments exposing tadpoles in stages 30–40 (Gosner limb development stages, Gosner 1960) to 2.5 parts per million (ppm) concentrations of emodin. The concentration was chosen based on concentrations ranging from 0.017 to 2.007 ppm reported from field samples (Sacerdote and King 2014). Our trial was designed to test the morbidity and mortality of tadpoles exposed to the metabolite emodin for 21 d, which is a sufficient duration for hepatic metabolic detoxification and neutralization of toxins (Bingen and Kirn 1977; Medina et al. 2016).

Tadpoles were collected from a pond at the University of Wisconsin-Waukesha field station with no common buckthorn trees surrounding it or in the vicinity. Tadpoles were acclimated for 10 d in 20-gal tanks at Carroll University's aquatics laboratory. Forty tadpoles of similar size and developmental stage were randomly selected and divided between the treatment and control groups and placed individually in 1-L tubs filled with 0.5 L of dechlorinated aged water and arranged in a randomized block design in 122×244 cm shelving units. The metabolite emodin used for the trials was prepared from powdered commercial stock (AdooQ Bioscence, Irvine, California, USA) and diluted in dechlorinated water to the desired concentration. Because of its partial insolubility in water, we used stirring bars for 10 h or until no precipitate was detected. Tadpoles were fed commercial, high-protein fish pellets daily at a ratio of 3% of body mass, which is sufficient for normal growth and development (Wilbur 1977). The food requirement was calculated based on the body mass of a separate sample of five nonexperimental tadpoles, treated identically to the controls, which were weighed before every feeding and the food amount calculated. To ensure high water quality, 100% water changes were conducted every 3 d during the experiment (emodin concentrations were maintained with every water change). The room was kept at 21 C (70 F), with a 12:12 light:dark light cycle.

Tadpoles were monitored twice daily for survival and morbidity. Dead individuals were removed from their containers as soon as possible to avoid tissue decomposition. Dead tadpoles were necropsied and the liver was removed and placed into 10% neutral buffered formalin for histopathology; spleen and kidneys were frozen for possible future testing. Tadpoles that exhibited clinical signs consistent with emodin intoxication, such as dull skin colors, mouth gaping and dyspnea, reddened skin, excess mucus, disorientation, or seizures for >24 h during the experiment were humanely euthanized and necropsied. At the end of each trial, all surviving individuals were humanely euthanized by immersion in benzocaine (100 g/L ethanol stock solution) at a 500 mg/L dosage (Leart et al. 2020) until cessation of ventilating and movement was identified. Euthanized individuals were necropsied as described above. The research protocol was approved by the Carroll University Institutional Review Board (no. 17-023), which functions as the instructional animal care and use committee.

Tadpole livers were placed in 10% neutral buffered formalin. The fixed livers were routinely processed into paraffin blocks, sectioned at approximately 5 µm, placed onto glass slides, stained with H&E (Wittekind 2003), and examined using light microscopy for histopathologic evaluation. The livers were carefully examined for signs of metabolic detoxification and/or hepatic injuries consistent with emodin metabolism including fibrosis, necrosis, atrophy, vacuolization of hepatocytes and inflammation, and presence of “stones” in the gallbladder (Chen et al. 2014; Cui et al. 2014). To determine differences in mortality between treated individuals and controls, we conducted a Kaplan-Meier survival analysis with a Breslow generalized Wilcoxon test. All analyses were performed with SPSS version 26 (IBM Corp., Armonk, New York, USA).

We found 85% mortality of tadpoles at advanced developmental stages (stage 30–40; Gosner 1960) when exposed to 2.5 ppm concentrations of emodin (Fig. 1). Mortality during the 21 d of the experiment was significantly higher in subjects exposed to emodin than in control animals (χ²=31.67; 1 df; P<0.01). All emodin-exposed subjects (including the survivors) displayed clinical signs including dyspnea (gasping for air), erythema, lethargy, and short-term erratic movements (spasms). Gross necropsy revealed enlarged nodular livers and accumulation of yellow or gold material, resembling emodin powder, within the gallbladder (Fig. 2B), liver (Fig. 2A), and biliary system (bile duct; Fig. 2B). No significant changes were seen in control animals (Fig. 2C).

Histopathologic analysis of animals exposed to emodin showed bridging fibrosis throughout the liver (Fig. 2D) with marked bile duct proliferation (Fig. 2C, H). Masson's trichrome and Alcian blue (pH 2.5) revealed collagen and mucin associated with the fibrosis (Fig. 2E, F). The gall bladder and bile ducts generally contain flocculent material, interpreted as the “yellow material” noted grossly (Fig. 2G, H).

We found that green frog tadpoles (Lithobates clamitans) at advanced developmental stages experienced high levels of mortality and liver damage when exposed to emodin at concentrations of 2.5 ppm. Histologically, the changes in the liver are consistent with those previously reported in cases of hepatic toxicity, where an accumulation of toxins within the liver resulted in oxidative stress, death of hepatocytes, and ultimately fibrous replacement within areas of hepatocellular loss (Stara et al. 2012; Li et al. 2018). Histologic analysis of the livers of exposed individuals did not show obvious hepatocellular necrosis; however, fibrosis and bile duct proliferation were observed, and we believe this resulted in the secondary obstruction of the portal triads in the biliary system. This obstruction may have occurred when the flocculent material noted within the bile ducts impeded flow into the liver and gallbladder (Fig. 2D, E, F). It is also possible that a combination of factors contributed. Given the hepatocellular swelling and accumulations of flocculent material within the gall bladder and bile ducts, it is likely that flow was impeded within the portal triads, further limiting the detoxification function of the liver (Medina et al. 2016; Liu et al. 2018). The neurologic signs observed (erratic movements) suggest a possible hepatic encephalopathy, although brain tissue was not examined to confirm this possibility.

Destruction of liver cells and the production of scar tissue (fibrosis), as seen in the challenged animals, is commonly associated with interference in the blood supply of the liver, which decreases the oxygen supply and increases the pressure of the blood entering the liver (portal hypertension). The decrease of oxygen reaching the liver cells exacerbates the oxidating stress produced by the reaction of the cytochrome P450s with endogenous liver cells, leading to hepatocellular congestion and necrosis (Medina et al. 2016; Yang et al. 2018).

As an anthraquinone compound, emodin is metabolized in the liver mostly through chemical reactions such as oxidation during phase I metabolism, when the compound is biotransformed to several hydroxylation metabolites and catalyzed by cytochrome P450s (Qin et al. 2016; Jiang et al. 2018). However, there are reports that reactive metabolites such as the ones produced during phase I metabolism can react to endogenous proteins from the liver, causing toxic effects such as oxidative stress and hepatoxicity (Park et al. 2005; Jiang et al. 2018). Gross and histopathologic changes seen in the liver of our treated green frog tadpoles demonstrated that the exposure to the chemical has deleterious effects on the liver.

Our findings help to understand the effect of complex compounds such as emodin on the liver function of amphibians and poses the question of how its effects will vary among species and stages of development, as liver function in tadpoles varies significantly through development. Because the tadpole's liver shifts from functioning as a hemopoietic organ initially to metabolic functions later, the capacity of an individual to metabolize complex compounds such as emodin will vary based on its stage of development and liver metabolic capacity, with younger tadpoles being potentially more susceptible to liver damage or death when exposed to the chemical, as described by Sacerdote and King (2014).

This potentially higher susceptibility of individuals at early developmental stages could have severe conservation implications. Amphibians inhabiting areas where the common buckthorn is widely distributed will be exposed to the chemical at early stages of development, especially early breeding–species such as chorus frogs (Pseudacris spp.), cricket frogs (Acris spp.), and spring peepers (P. crucifer) that inhabit areas where thawing of the winter snowpack could accelerate decomposition of buckthorn leaves deposited the previous fall; ergo, boosting the concentration of the chemical in the water and increasing mortality rates of larvae. This increase in early larvae mortality will diminish the rates of recruitment of these species, affecting the overall community. Thus, the potential persistence of the metabolite in the environment, as well as the wide distribution of this invasive plant in areas such as the Midwest, could represent a significant threat, particularly to early breeding–species, potentially contributing to local declines of populations in areas where the buckthorn is currently established. More research describing the concentrations of residual emodin persistent in the environment and the effect of the metabolite on the liver function of fast-developing species is needed. Moreover, the management of this invasive plant species should be considered, including the proper disposal of removed plants, avoiding or reducing the reintroduction of the chemical into the environment.

We would like to thank the people that assisted in one way or another in the completion of this research; Frank Gorichanaz, James Zelten, Maddeline Gubka, Millar Minahan, Alex Goyette, David Leaders, and Jacob Lee for assistance in data collection and field and laboratory work; Monika Baldridge for assistance and guidance in the preparation of samples; and Susan Roskopf for her invaluable assistance in the running of the controlled experiements. We also thank the University of Tennessee Histology Laboratory for slide preparation and special staining. This work was supported by Carroll University faculty development grants and research grants from TriBeta Biology Honor Society.