Nitrosative tissue damage and apoptotic cell death in kidneys and livers of naturally ethylene glycol (antifreeze)-poisoned geese

Abstract

Three naturally ethylene glycol (EG)-intoxicated geese were investigated for pathological changes, nitrosative tissue damage and apoptotic cell death. Severe degeneration of kidney tubular epithelium and congestion of kidney and liver tissues were observed. Immunohistochemical staining for inducible nitric oxide synthase and nitrotyrosine revealed strong immunoreactivity with both antibodies in kidney and liver tissues compared with the weak immunostaining in the control animals. In both tissues of the EG-intoxicated geese, erythrocytes were also highly immunoreactive with nitrotyrosine antibody. A high degree of apoptotic cell death was present in the kidney tubule epithelium of EG-intoxicated geese. Some apoptotic cells were also observed in the liver. These results show that nitrosative tissue damage and apoptotic cell death takes place in kidney and liver during EG intoxication in geese.

Lésion tissulaire par nitrosation et mort des cellules par apoptose dans les reins et les foies d'oies naturellement empoisonnées par de l'éthylène glycol (antigel)

Trois oies naturellement intoxiquées par de l'éthylène glycol (EG) ont fait l'objet d'une recherche des lésions anatomopathologiques, des lésions tissulaires par nitrosation et de la mort des cellules par apoptose. Une dégénérescence sévère de l'épithélium tubulaire des reins et une congestion des tissus des reins et du foie ont été observées. La coloration immunohistochimique pour la synthase inductible de l'oxyde nitrique et pour la nitrotyrosine a révélé une forte immunoréaction avec les deux anticorps dans les tissus des reins et du foie comparée à l'absence d'immunocoloration chez les témoins. Dans les deux tissus des oies intoxiquées par l'EG, les érythrocytes ont également été très fortement immunoréactifs vis-à-vis de l'anticorps nitrotyrosine. Un niveau élevé de mortalité cellulaire par apoptose a été noté dans l'épithélium tubulaire des reins des oies intoxiquées par l'EG. Quelques cellules apoptotiques ont également été observées dans le foie. Ces résultats montrent que les lésions tissulaires par nitrosation et la mort des cellules par apoptose prennent place dans les reins et le foie durant l'intoxication des oies par l'EG.

Nitrosative Gewebezerstörung und apoptotischer Zelltod in Nieren und Lebern bei mit Äthylenglykol (Frostschutzmittel) vergifteten Gänsen

Drei mit Äthylenglykol (ÄG) vergiftete Gänse wurden auf pathologisch-anatomische Veränderungen, nitrosative Gewebezerstörung und apoptotischen Zelltod hin untersucht. Es wurde eine hochgradige Degeneration des tubulären Nierenepithels sowie eine Nieren- und Leberkongestion festgestellt. Bei der immunhistochemischen Färbung auf induzierbare Nitritoxidsynthase und Nitrotyrosin zeigte sich in den Nieren- und Lebergeweben dieser Tiere im Vergleich mit der schwachen Immunfärbung bei den Kontrolltieren eine deutliche Immunreaktion mit beiden Antikörpern. Bei den ÄG-vergifteten Gänse ließen in beiden Organen auch die Erythrozyten eine hochgradige Immunreaktion mit dem Nitrotyrosin-Antikörper erkennen. Außerdem konnte im tubulären Nierenepithel hochgradiger apoptotischer Zelltod beobachtet werden. Einige apoptotische Zellen traten auch in den Lebern auf. Diese Ergebnisse belegen, dass eine ÄG-Intoxikation bei Gänsen zu nitrosativer Gewebezerstörung und apoptotischem Zelltod führt.

Daño tisular oxidativo y muerte celular apoptótica en riñones e hígados de gansos envenenados de forma natural con etilenglicol (anticongelante)

Se estudiaron las lesiones, el daño tisular oxidativo y la muerte celular por apoptosis en tres gansos intoxicados de forma natural con etilenglicol (EG). Se observó degeneración grave del epitelio tubular renal y congestión en riñón e hígado. La tinción inmunocitoquímica para la sintasa de óxido nítrico inducible y para la nitrotirosina mostraron una inmunoreactividad intensa con ambos anticuerpos en tejidos de riñón e hígado en comparación con la débil inmunotinción observada en los animales control. Los eritrocitos de los dos tejidos de los gansos intoxicados con EG también fueron altamente inmunoreactivos con el anticuerpo frente a nitrotirosina. Se observó un grado elevado de muerte celular por apoptosis en el epitelio de los túbulos renales de los gansos intoxicados con EG. También se observaron algunas células apoptóticas en el hígado. Estos resultados muestran que el daño tisular oxidativo y la muerte celular por apoptosis tiene lugar en el riñón e hígado durante la intoxicación por EG en gansos.

Introduction

Ethylene glycol (EG) is a colourless, relatively non-volatile, sweet-tasting viscous liquid that is the main ingredient of most brands of antifreezes used in automobile radiators. It is commonly used as an industrial solvent. There have been frequent reports of dogs (Urman & Kahraman, 1982; Thrall et al., 1984; Rowland, 1987; Connally et al., 1996; Khan et al., 1999; Goicoa et al., 2003; Keller & Goddard, 2005) and cats (Thrall et al., 1984; Hamlin, 1986; Khan et al., 1999) with toxicosis due to EG ingestion, although cases of toxicosis have also been reported in other species such as cattle (Crowell et al., 1979), pygmy goat (Boermans et al., 1988), polar bear (Amstrup et al., 1989), and raccoon (Foley & McBurney, 2002). EG toxicosis due to accidental or intentional ingestion has also been described in humans (Bobbitt et al., 1986; Baud et al., 1988). The first report of EG toxicosis in poultry was made by veterinarians in the German army in 1940 (Eberbeck & Hemmer-Halswick, 1943). This study was followed by demonstration of calcium oxalate crystals in various organs of chickens (Hutchison & Dykeman, 1997; Radi et al., 2003), ducks (Stowe et al., 1981), pigeons (Urman et al., 1982) and geese (Riddell et al., 1967).

EG-intoxicated animals usually present neurological signs of incoordination, convulsions and ataxia (Riddell et al., 1967; Stowe et al., 1981; Radi et al., 2003). Death can occur within 5 to 48 h due to metabolic acidosis depending on the amount of EG ingested and the species of animal, with up to a 100% mortality rate (Moriarty & McDonald, 1974; Szabuniewicz et al., 1975; Barton & Oehme, 1981). Postmortem examination of animals usually reveals variably congested and enlarged kidneys. Histopathological findings include degeneration and necrosis of renal tubule epithelium and presence of calcium oxalate crystals in the lumen of renal tubules, and occasionally in the cerebral vessels (Hutchison & Dykeman, 1997; Foley & McBurney, 2002). Although there are many reports describing these findings, none of them described the extent and level of tissue degeneration due to EG toxicosis. Therefore, the aims of the present study were to study the pathological changes, to examine for evidence of oxidative tissue damage, and to investigate the type of cell death occurring in tissues of a flock of geese naturally poisoned with EG.

Materials and Methods

Three dead geese presented for necropsy to the Department of Pathology, University of Kafkas were examined. According to the owner, there were a total of 11 dead geese in a flock composed of 22 animals. All had shown difficulty in standing and walking, and were depressed with their heads down. Death occurred less than 1 day after the signs were seen. The geese were free ranging and had been seen in the yard of a local auto repair shop before they showed the clinical signs, hence EG toxicosis was suspected.

Routine necropsy was performed on the geese, and to observe the pathological changes histological sections were prepared from formalin-fixed paraffin-embedded tissue blocks. Haematoxylin and eosin and, to reveal calcium oxalate crystals, von Kossa and periodic acid Schiff stains were performed.

Immunohistochemical investigations

Immunohistochemical localization of inducible nitric oxide synthase (iNOS) and nitrotyrosine were investigated in sections of kidney and liver tissues of the EG-poisoned geese. As a control, five healthy geese obtained from Kafkas University, Animal Research Farm were sacrificed, and the tissues were processed in the same way as described above. For immunohistochemistry, 4 µm sections were cut from the formalin-fixed and paraffin-embedded tissue samples, deparaffinized with xylene, and rehydrated with an alcohol series. After blocking the endogenous peroxidase activity with a 0.3% hydrogen peroxide solution, antigen retrieval was performed by microwave treatment of the sections with 10 mM citrate buffer, pH 6.0. The sections were pretreated with 1% normal goat serum in phosphate-buffered saline (PBS) to prevent non-specific antibody binding. Primary antibodies against iNOS (rabbit anti-iNOS polyclonal antibody; Lab Vision, Fremont, USA) and against nitrotyrosine (rabbit anti-nitrotyrosine polyclonal antibody; Calbiochem, Darmstadt, Germany) were then applied to tissue sections for 1 h in 10 µg/ml and 6.5 µg/ml concentrations, respectively. Biotinylated anti-rabbit antibody and streptavidin peroxidase solution were consecutively applied onto tissue sections for 30 min each with three 5-min PBS rinses between the applications. Diaminobenzine with 0.3% H₂O₂ solution was used as the chromogen to visualize the antigenic localizations of both molecules. Finally, the sections were counterstained with Mayer's haematoxylin and coverslipped for microscopic observation.

In situ TUNEL staining

The DeadEnd™ Colorimetric TUNEL System (Promega, Madison, USA) was used to demonstrate apoptotic cells. Briefly, 4 µm sections of formalin-fixed paraffin-embedded kidney and liver samples of EG-poisoned and control geese were cut and processed through xylene and an alcohol series. The sections were rinsed in 0.1 M PBS, and placed in PBS-buffered proteinase K solution for 30 min. Following rinses in PBS, the sections were placed in equilibration buffer containing 200 mM potassium cacodylate, 25 mM Tris–HCl, 0.2 mM dithiothreitol, 2.5 mM cobalt chloride, and 0.25 mg/ml bovine serum albumin for 10 min and then incubated with reaction buffer containing biotinylated nucleotide mix and terminal deoxynucleotidyl transferase for 1 h in a humidified chamber at 37°C. To stop the reaction, the samples were incubated in 1 M sodium citrate solution at room temperature for 15 min. Sections were then rinsed twice in PBS and endogenous peroxidase activity was blocked by 3% H₂O₂ for 5 min. The sections were then incubated with a streptavidin horseradish peroxidase solution for 30 min. After rinsing with PBS, peroxidase activity in the samples was revealed with a solution of diaminobenzidine/H₂O₂. Finally, the sections were rinsed in distilled water and counterstained with 0.1% methyl green for 10 min. Following rinses in distilled water and butanol, the sections were placed in two changes of xylene for 2 min, and were coverslipped.

Results

Postmortem examination of the EG-poisoned geese revealed that the animals were in good body condition with plenty of adipose tissue deposits. Congestion and oedema of the kidneys, livers and, in one animal, pancreas were the only gross abnormalities observed. Microscopic findings were similar in all three geese. Histological examination of haematoxylin and eosin-stained kidney sections revealed severe congestion and moderate to severe tubular degeneration and necrosis. Light-yellow crystals, which were angular and often irregular, arranged in sheaves or rosettes shape, were observed in the lumen of numerous renal tubules. They were located mostly in the cortex, although some were also observed in the medulla. The morphology of renal tubular epithelium was altered with many cells swollen and containing vacuoles that were indicative of degenerative changes. Many sloughed-off tubular epithelial cells and eosinophilic proteinaceous casts were observed in the lumen of numerous tubules. Damaged renal tubules were often not associated with the calcium oxalate crystals. Bowman's spaces were distended throughout the kidney sections (Figure 1a). The crystals were identified as calcium oxalate by von Kossa stain (Figure 1b) and periodic acid Schiff stain. In the liver, mild to moderate degeneration of hepatocytes was observed. Degenerate hepatocytes were often vacuolated and presented a cloudy appearance. Severe vascular and sinusoidal congestion was also present. There were no observable histopathological findings in other organs.

Figure 1.  1a: Pale-yellow calcium oxalate crystals in renal tubular lumens of an EG-intoxicated goose. Severe congestion, and tubular degeneration and necrosis are evident. Haematoxylin and eosin, ×370. 1b: Black-stained calcium oxalate crystals. von Kossa, ×185.

Immunohistochemical staining of the kidney tissues for iNOS revealed a light diffuse staining on the tubule epithelium in the control animals (Figure 2a). There was moderate to strong immunostaining for iNOS in the kidneys of EG-poisoned geese. Immunoreactivity was observed only on the tubule epithelium and there was virtually no staining in the glomeruli and other cellular components of the kidneys (Figure 2b). In the liver of control geese, a light diffuse immunostaining was present. In liver sections of EG-poisoned geese, strong immunoreactivity against iNOS antibody was observed on the hepatocytes throughout the sections; however, staining intensity varied from cell to cell with some cells having less immunoreactivity.

Figure 2.  2a: Light and diffuse immunostaining for iNOS in kidney of a control goose, ×185. 2b: Moderate immunostaining for iNOS on renal tubular epithelium in an EG-intoxicated goose, ×185.

Immunohistochemical staining for nitrotyrosine revealed light immunoreactivity in the kidney tubule epithelium of the control geese (Figure 3a). In the EG-poisoned geese, immunostaining for nitrotyrosine was moderate to strong in the renal tubular epithelium. There was also a strong immunoreactivity on the erythrocytes in these sections (Figure 3b). Light diffuse immunostaining was present in the livers of control geese. In the livers of EG-poisoned geese, moderate to strong immunoreactivity was observed on the hepatocytes. As in the kidney sections, erythrocytes were also strongly immunoreactive against nitrotyrosine antibody.

Figure 3.  3a: Light immunostaining for nitrotyrosine in kidney of a control goose, ×185. 3b: Moderate to strong immunostaining for nitrotyrosine in renal tubular epithelium of an EG-intoxicated goose. Also note that there is a strong immunoreactivity on erythrocytes, ×185.

A large number of kidney tubule epithelia were shown to be apoptotic by in situ TUNEL staining in the EG-poisoned geese (Figure 4a). In tubules where staining took place, all of the tubule epithelial cells were mostly stained. There were also many tubules with one or more stained cells. Apoptotic cells were more common in the cortex than the medulla and there were no apoptotic cells in the glomeruli. While calcium oxalate crystals were often not associated with tubules with apoptotic cells, some tubules with apoptotic cells contained the crystals. In the liver, some apoptotic cells, more often in the periportal region, were observed (Figure 4b). There were no observable apoptotic cells in the kidney and liver tissues of the control animals.

Figure 4.  4a: Numerous apoptotic cells in the renal tubules of an EG-intoxicated goose. Note that while apoptotic cell death occurs in all cells of a tubule, single or a few apoptotic cells are also present in a single tubule. In situ TUNEL staining, ×370. 4b: Some apoptotic cells located in the periportal region of liver of an EG-intoxicated goose, In situ TUNEL staining×370.

Discussion

EG (antifreeze) poisoning is an occasional finding in animals. It is most commonly observed in cats and dogs, which can more easily access antifreeze that is frequently spilt on the floor of a garage or an auto-repair shop. While there are reports of accidental EG poisoning in many species, none of them investigated the extent of oxidative tissue damage and the type of cell death occurring in the kidney and liver, which are the organs most commonly affected by EG intoxication.

In the present investigation, EG poisoning was diagnosed based on the anamnesis and the histopathologic observation of calcium oxalate crystals in the renal tubules of geese. Degeneration of renal tubular epithelium and hepatocytes together with severe congestion in both organs were other prominent findings. These histopathological observations are consistent with previously reported cases of EG toxicity in both poultry and other species (Riddell et al., 1967; Stowe et al., 1981; Urman & Kahraman, 1982;Hamlin, 1986; Hutchison & Dykeman, 1997). Calcium oxalate crystals have also been occasionally reported in cerebral blood vessels, duodenum, and marrow cavities of the tracheal ring of various species (Riddell et al., 1967; Crowell et al., 1979; Stowe et al., 1981; Urman & Kahraman, 1982; Foley & McBurney, 2002; Radi et al., 2003). These findings seem to be variable from case to case, probably due to the amount of EG ingested and the time when the necropsy was performed after the first intake. In the present case, calcium oxalate crystals were not observed in these organs. On the other hand, severe congestion was observed in the pancreas of a goose.

EG exerts its toxic effect through the metabolism to glycolaldehyde, glycolate, glyoxylate, and oxalate by the catalytic action of alcohol dehydrogenase in liver. Later, calcium binding to oxalate forms calcium oxalate crystals that are located especially in the kidneys, where they cause tissue degeneration. However, we and others observed that the degenerate kidney tubules were not always associated with calcium oxalate crystals (Urman & Kahraman, 1982; Boermans et al., 1988). In addition, the degree of renal necrosis was not always related to the amount of crystals formed (Goicoa et al., 2003). Therefore, other mechanisms such as direct cytotoxic effects of the hippuric or glyoxylic acid metabolites have been suggested to involve in the tissue damage (Clay & Murphy, 1977; Jacobsen et al., 1982; Thrall et al., 1984). These formed aldehydes may inhibit oxidative phosphorylation and respiration, and hence lead to oxidative tissue damage.

In the current study, the degree of tissue damage was investigated immunohistochemically using antibodies against iNOS and nitrotyrosine. Increased immunoreactivity against both antibodies in the renal tubular epithelium and hepatocytes of the EG-poisoned geese indicated that a high level of free radicals plays an important role in the oxidative tissue damage in these animals. iNOS is known to involve in the catalytic formation of nitric oxide from l-arginine. Nitric oxide plays a key role in the formation of peroxynitrite, by reacting with some oxygen free radicals such as superoxide (Beckman & Koppenol, 1996). Peroxynitrite, which is highly cytotoxic, can either cause protein nitration or be parted to yield NO₂ or NO₃ that might further cause DNA damage (Ischiropoulos et al., 1992; Nordberg & Arner, 2001). Therefore, nitrated tyrosine residues can be detected by anti-nitrotyrosine antibodies, and hence can indirectly show nitric oxide formation and nitrosative tissue damage.

In the congested kidneys and livers of EG-poisoned geese, erythrocytes were observed with strong immunoreactivity against nitrotyrosine antibody. However, no immunostaining against iNOS was observed on these cells. Therefore, tyrosine nitration could not take place due to iNOS expression in these cells. There are two possible explanations for strong nitrotyrosine immunostaining on erythrocytes in EG-poisoned geese. Self-nitration of haemoglobin by peroxidative activity of this protein might be the reason for the presence of nitrated proteins in erythrocytes as it has been shown that in vitro exposure to peroxynitrite results in protein nitration in these cells (Denicola et al., 1998). EG poisoning has been shown to cause haemolysis (Crowell et al., 1979; DePass et al., 1986), and somehow the breakdown of haemoglobin during EG poisoning might trigger the protein nitration. Another explanation for strong immunoreactivity against nitrotyrosine antibody on erythrocytes might be release of nitric oxide from the endothelium during the intoxication and further free pass of it to these cells (Gladwin et al., 2004). Nitric oxide then might cause protein nitration (Quijano et al., 2005).

Calcium oxalate crystals in renal tubules are known to block the tubules, causing necrosis of the tubular epithelium (Stowe et al., 1981; Hutchison & Dykeman, 1997). In haematoxylin and eosin-stained sections, many necrotic tubular epithelia were also observed in the EG-poisoned geese. We also tested for apoptotic cells in the renal tubules of these animals. Surprisingly, many renal tubules revealed a high degree of apoptosis. In most sections, all of the tubular epithelium was observed as apoptotic, although single or few apoptotic cells in a tubule were also numerous. Calcium oxalate crystals were observed to not always be associated with the tubules that had a high degree of apoptosis, as in the case of tubules with necrosis. Some apoptotic cells in the portal areas of liver were also determined in these animals. Apoptosis in renal tubules of EG-poisoned animals was not reported previously. Concurrent existence of apoptotic and necrotic cell deaths might be explained by the presence of multiple mechanisms. Direct physical damaging effect of calcium oxalate crystals on tubular epithelial cells might be one way of causing degeneration and death of these cells. This type of effect results in necrotic cell death. Cytotoxic effects of hippuric or glyoxylic acid metabolites might be another, and probably the most effective, mechanism for necrotic cell death. Peroxynitrite, a potent oxidant, is an important inflammatory mediator (Thérond et al., 2000). As it was demonstrated indirectly by immunohistochemical staining in this study, a high degree of peroxynitrite forms in the kidney tubule epithelium as a result of EG poisoning, most probably due to oxalic acid metabolites. The presence of increased iNOS immunoreactivity was also a confirmation of a high level of peroxynitrite production. It seems these cytotoxic effects are somehow involved in the apoptotic process. NO₂ and NO₃ molecules produced from the peroxynitrite are potential candidates for apoptotic DNA damage, and hence require further studies.

In conclusion, the extent and level of tissue damage were investigated in three naturally EG-poisoned geese. Nitrotyrosine and iNOS immunoreactivities, as markers of reactive nitrogen species, were demonstrated in the kidney and liver tissues in these animals. Severe renal tubular necrosis and the coexistence of high degree of apoptosis was an unexpected finding. The concurrent presence of large amounts of necrotic and apoptotic cells might be explained by the existence of different mechanisms in the process of cell death, and requires further investigation.