**1. Introduction**

Facial eczema (pithomycotoxicosis) (FE) is a secondary-hepatogenous photosensitization disease of farm ruminants caused by the epipolythiodioxopiperazine mycotoxin sporidesmin A, present in the spores of the saprophytic fungus *Pithomyces chartarum* [1,2]. This fungus grows on dead vegetable matter at the base of the ryegrass-dominant pasture in all temperate worldwide zones during cloudy days with rain, with temperatures above 16 ◦C (optimal 24 ◦C), and relative humidity upper than 80% [3]. The portal circulation is the main route by which sporidesmin A enters to the liver following absorption from the intestinal tract [4]. In the biliary system, the sporidesmin A, characterized by the presence of an internal disulphide bridge, leads to the formation of toxic free-radicals that react with molecular oxygen to produce superoxide radicals [2,5,6]. The damaged liver is unable to remove a normal end product of chlorophyll metabolism, phylloerythrin (a photodynamic agent), from the blood for excretion in the bile [5]. When ruminants are exposed to sunlight, this pigment is responsible for unpigmented skin lesions [7]. FE was first recognized in New Zealand where it occurs more frequently [7]. However, this mycotoxic disease has also been reported in South Africa [8], Australia [9], the United States [10], France [11], Portugal [12], the Netherlands [13], Turkey [14], Uruguay and Argentina [3]. In Spain, the first and only reported outbreak occurred in the Basque Country [15]. In live animals, increases in the serum concentration of several enzymes such as gamma-glutamyltransferase (GGT) were found to be positively correlated to cholestasis and are indicators of bile duct damage in sporidesmin natural and experimental intoxication [1,7,12,16–19].

FE commonly affects sheep and cattle [7] while goats are more resistant to sporidesmin toxicosis than sheep [1]. The experimental sporidesmin toxicity in the rabbits has been demonstrated [17]. Although there was variation in susceptibility between individuals, the degree of liver injury and photosensitization appears to increase with both dose and length of time during which sporidesmin was administered [20]. The toxicity of pastures depends on the number of *P. chartarum* spores in the dead plant material and the toxicity of the particular *P. chartarum* strains [3]. Some authors reported that the most severe liver injury was due to the higher total consumption of spores caused by the combination of pasture spore concentration and dry matter intake [21]. Field observations reported that high spore counts (more than 40,000 spores/g of grass) of *P. chartarum* in grass samples can cause clinical signs [12], and there was a strong relationship between spore counts in ruminal content and severity of clinical signs [14]. In sheep, the clinical signs of photosensitization (erythema, edema and alopecia in unpigmented skin) appeared 14–18 days after intake of the toxin [1,20]. Jaundice [3] and loss of body weight in severe and chronic FE [7,20] are evident in this disease. In acute cases the liver is enlarged and shows a yellowish discoloration [10], and the gallbladder and extrahepatic bile ducts are distended [17]. In cases of longer evolution, there is liver atrophy and fibrosis [3] leading to liver cirrhosis [8,15].

The histological lesions include acute necrotizing cholangitis [12,17], bile stasis [1] and later on, bile duct hyperplasia and portal fibrosis [3,10,15]. Liver regeneration [7] and interlobular cirrhosis [8] have also been shown in the chronic phase of FE.

A lymphocytic infiltrate has been observed associated with the chronic hepatic changes of FE [8,10]. The main cell target in cholangiopathies are the epithelial cells lining the bile ducts (i.e., cholangiocytes), that are exposed to cytokines and inflammatory mediators produced by infiltrating lymphocytes, macrophages and activated myofibroblasts [22]. "Typical cholangiocyte proliferation" is usually found in acute obstructive cholestatic liver disease and "atypical ductular reaction" is commonly seen in human primary biliary cirrhosis [23]. Bile duct lesions have been described in subacute FE in cattle [24], but there are no studies about the type of ductular response in chronic FE cases or on how ductular cells interact with other cell types such as Kupffer and hepatic stellate cells (HSCs). Quiescent HSCs are resident perisinusoidal cells in the subendothelial space between hepatocytes and sinusoidal endothelial cells [25]. They are the primary site for storing retinoids (vitamin A) within the body. During liver injury HSCs proliferate and differentiate into contractile and matrix-producing myofibroblasts that generate progressive fibrosis and promoted a chemotactic activity for monocytes and lymphocytes, among others [25–27]. In liver pathology, ample evidence has been provided for an indirect role of macrophages in the development of fibrosis [25].

The main objective of this study is to characterize the inflammatory infiltrate, as well as the presence of HSCs and their distribution in damaged hepatic tissue in natural cases of FE from an outbreak that occurred in Northwestern Spain. Additionally, the acute and chronic liver lesions found in these cases of the disease, with special reference to the ductular reaction, are characterized, mainly considering their relationship with the presence of inflammatory cells, as a possible indicator of the role that the immune response can play in the development of liver damage in this toxicosis.

#### **2. Materials and Methods**

#### *2.1. Ethical Information*

Experimental animals were not used in this work. An observational study was performed with the blood samples obtained during regular veterinary clinical services and with *post mortem* tissue samples that are routinely collected after the death of animals.

#### *2.2. Case History and Clinical Observations*

The outbreak of FE occurred in La Mata, Grado, an inland municipality in the Principality of Asturias, in the north-west of Spain, between the end of September and the beginning of October 2003. This estate belonged to Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA), Asturias. According to official data of the Asturian Society of Economic and Industrial Studies and the Meteorological Territorial Center of Asturias, the summer climate of 2003 was dry and very warm, with temperatures higher than 30 ◦C in the inland [28]. The autumn was warm (with temperatures above 19 ◦C in September and the first 19 days of October) and very rainy with a total monthly rainfall of 141.4 and 219 L/m<sup>2</sup> in October and November, respectively. The affected sheep were grazing on two plots (named 1B and "mixtures") at 50 m altitude. The plot 1B, of a 3.2 hectare (ha) of land, was sowed in 2001 with perennial ryegrass (*Lolium perenne* L. var. Tove), 30 kg/ha; hybrid ryegrass (*Lolium boucheanum* var. Kunth), 12 kg/ha and white clover (*Trifolium repens* var. Huia), 3 kg/ha. The plot mixtures of 2.5 ha was sowed in 1993 with perennial ryegrass (*Lolium perenne* L. var. Phoenix), hybrid ryegrass (*Lolium boucheanum* variety Dalita), and white clover (*Trifolium repens* var. Huia), at the same doses. In the plot 1B 28 crossbreed female sheep and 11 ewe lambs were grazing, of which 7 (6 adult and 1 ewe lamb) showed clinical signs of photosensitivity (pruritus, erythema and alopecia) on the face and ears. In the plot mixtures 2 adult sheep (over a total of 24) and 3 ewe lambs (over a total of 11) developed similar clinical signs, according to the information supplied by the practitioner. The four most affected animals had been treated symptomatically with Alergia-N (Pfizer), an antihistamine drug (cyprohetadine chlorhydrate, chlorphenamine maleate), and Penbex (Industrial veterinaria, S.A-INVESA), for treatment of secondary bacterial infections caused by germs sensitive to the association penicillin-dihydrostreptomycin sulfate, both by intramuscular route. Initially, the face skin lesions recovered when the animals were removed from the pastures for some days into the shade, but began to appear when the sheep returned to the plots and were exposed to sunlight. In the following 6 months, animals involved in the FE outbreak 4 sheep and 1 ewe lamb (plot 1B) and 1 sheep and 2 ewe lambs (plot mixtures) showed lower weight gains, wool loss and crusting in the dorsum of the head, nose and ears were observed.

#### *2.3. Biochemical Assay and Statistical Analysis*

Blood serum samples were taken from the jugular vein from 73 sheep (11 with and 62 without skin lesions) during the chronic stage of the episode (six months after the FE outbreak). Serum concentrations of γ-glutamyltransferase (GGT), alkaline phosphatase (ALP), aspartate aminotransferase (AST), albumin and total protein (TP) were determined on a multianalyser (Cobas Integra 400, Roche diagnostics). The results of the serum parameters analyzed were reported as mean, standard deviations and range (minimum and maximum), calculated using routine descriptive statistical procedures. The Kolmogorov– Smirnov test was used to assess normality of data. Non-parametric statistical methods were used to compare groups. Mann-Whitney U test was employed to compare the exposed animals that not showed cutaneous clinical signs with exposed sheep that presented skin lesions. *p*-values of less than 0.05 were considered statistically significant. All the statistical analyses were performed with the R software version 3.6.1 (R Development Core Team, R Foundation for Statistical Computing, Vienna, Austria, December 2019).

#### *2.4. P. chartarum Spore Counts*

A total of 36 grass samples from every field were randomly collected to count *P. chartarum* spores/g on 23 November 2003. Each grass sample was taken at least 10 m apart. The grass was cut 1 cm above the ground, avoiding taking soil, and cut into pieces of approximately 4 cm in length. From this mixture, 15 g of grass was taken and 150 mL of water was added, and then the mixture was homogenized for 3 min, in order to release the spores in the water. The *P. chartarum* spores present in each sample were identified and counted in an aliquot of wash water, using a Fusch-Rossental chamber. Finally, depending on the volume of water investigated, calculations were made to express the results as number of spores/g of grass. These grass samples were analysed in the Department of Animal Health of NEIKER- Basque Institute for Agricultural Research and Development.

#### *2.5. Animal Cases and Pathological Examination*

One adult sheep (plot 1B) with clinical signs suffering the acute phase of the disease, two adult animals (1 of the plot 1B and 1 of the plot mixtures), two months after the FE outbreak, and two adult sheep of the plot 1B six months after the FE outbreak, all of them belonging to the Galician breed, were examined in this study. Two sheep of a different estate were selected as reference and healthy animals. All of them were submitted to the Pathologic Diagnostic Service of the Veterinary Faculty of León over a seven-month period (November 2003–May 2004). In November an adult sheep with clinical signs in the acute phase of FE was humanely euthanized in the flock and were submitted for necropsy. Only when the pathological findings were discussed with the clinician, there was enough evidence to consider sporidesmin toxicosis as a possible cause of the liver and skin lesions. Two affected sheep and two other healthy control sheep were examined at slaughter in January and different tissue specimens (liver, skin, kidney, lung, heart) were submitted for histopathology. Finally, in May, two adult alive sheep were submitted for necropsy and euthanasia was performed by intravenously injection of barbiturate (T61; Intervet International, Madrid, Spain), after xylazine (Bayer, Leverkusen, Germany) subcutaneous administration, followed by exsanguination after the severing of carotid arteries and jugular veins according to our institution guidelines. Complete necropsies were performed in 3 sheep, one submitted during the acute, and two during chronic phases of FE. After gross examination, representative tissue samples (skin, liver, kidney, spleen, mesenteric lymph nodes, intestine, pancreas, adrenal glands, heart, lung, skeletal muscle and brain), were collected in all necropsied animals. All samples were fixed in 10% neutralbuffered formalin, processed routinely and embedded in paraffin wax. Sections (4 μm) were cut, mounted on glass microscope slides and stained with haematoxylin and eosin (HE), Masson Goldner trichrome for collagen, AFIP (Armed Forces Institute of Pathology) method for lipofuscin, Hall´s bilirubin stain, Perls´ Prussian blue stain for ferric pigments and acid rubeanic method for copper. Two livers with lesions consistent with steatosis

and macronodular cirrhosis belonging to sheep of plot 1B that died in January and May, respectively, were not included in this study due to severe *post mortem* alterations (autolysis and putrefaction).
