Red mark syndrome of trout (Oncorhynchus mykiss; Walbaum, 1792): Histopathological scoring and correlation with gross lesions

Abstract Red mark syndrome (RMS) is a skin disorder affecting rainbow trout (Oncorhynchus mykiss). The present work aimed to correlate the gross skin lesions affecting 46 fish sampled from farms surveyed for RMS with their microscopic features, identifying histological parameters that may be suggestive of disease progression. Skin lesions were grossly included in one of three categories (types I, II and III) according to the progressive degree of severity. Histological parameters and anti‐proliferating cell nuclear antigen (PCNA) tissue immunoreactivity were semi‐quantitatively assessed. In the dermis, PCNA‐positive lymphocytes, fibroblasts and endothelial cells were indicative of active phlogosis. A significant increase in PCNA‐immunoreactive lymphocytes, from gross type I to type III cases, was found only in the hypodermis. The histological parameters significantly associated with the gross lesion severity were progressive loss of the epithelium and scales, recruitment of inflammatory cells in the stratum compactum, loss of architecture of the stratum compactum, perivascular and perineural granulomatous inflammation and increase in lymphocyte infiltration of the muscular layer. In the type II and type III categories, inflammation in the hypodermis and muscle displayed a granulomatous pattern, reinforcing the hypothesis of an immunopathological mechanism. The morphological diagnosis of “deep chronic dermatitis associated to panniculitis and myositis, characterised by lympho‐histiocytic and granulomatous reaction” is suggested.

. European RMS strongly correlates with two similar rainbow trout diseases reported in the United States, namely U.S. strawberry disease (USSD) and U.S. Rush (La Patra et al., 1994;Lloyd et al., 2008;Olson et al., 1985). Furthermore, RMS shows morphological similarities with a warm water form of strawberry disease (SD) previously described in Europe (Fleury et al., 1985;St-Hilaire & Jeffery, 2004). It is currently believed that RMS and USSD are the same disease (Oidtmann et al., 2013). The transmissible nature of the disease has been demonstrated by experimental cohabitation trials (von Gersdorff Jørgensen et al., 2019;Verner-Jeffreys et al., 2008), and the involvement of a Midichlorialike organism (MLO) (Cafiso et al., 2015;Metselaar et al., 2020) has been hypothesized. Intracytoplasmic microorganisms, resembling Rickettsiales, were observed by transmission electron microscopy (TEM) within macrophages, fibroblasts and erythrocytes in the skin of RMS-affected rainbow trout in Italy and Slovenia Galeotti, Ronza, et al., 2017). In a recent study, a strong positive correlation was demonstrated between the presence of MLO and RMS-affected fish, further supported by immunohistochemistry (Metseelar et al., 2020). Early clinical signs are characterized by pale yellow/grey flat patches with central reddening, mainly on the fish flank, or opaque raised patches, with increased mucus production (Schmidt-Posthaus et al., 2009;Verner-Jeffreys et al., 2008). The lesion evolution includes an increase in size, elevation and redness of the wounds. The diameter of the lesions ranges from 5 mm to several centimetres; they appear round or ovoid, elongated, with progressive scale loss. Ulceration is not a common finding (Ferguson et al., 2006;Noguera, 2008;Verner-Jeffreys et al., 2006. The histology of the affected skin reveals a mild-to-severe dermatitis, with intact epidermis and massive infiltration of inflammatory cells, such as lymphocytes, plasma cells and macrophages (involving the stratum spongiosum [SS], the scale pockets [SP] and the stratum compactum [SC]). A severe infiltration of the same cells is consistently observed in the hypodermis and underlying muscle Galeotti, Ronza, et al., 2017;McCarthy et al., 2013;Schmidt-Posthaus et al., 2009;Verner-Jeffreys et al., 2008). Most of the investigations described in the literature have aimed to understand the aetiopathogenesis of the disease, whereas relatively few papers have focused on a detailed analysis of its histological features in order to both define a classification of the lesions during their evolution and further understand the pathogenesis of this disease, which is still far from being clear. The present work aimed to describe the development of the lesion from the histological point of view, starting from the early stages to the fully advanced lesions, and to correlate the scoring of skin histological patterns with selected macroscopic features of the lesions.

| Case study and sampling
The disease was identified and surveyed in five rainbow trout farms (namely A, B, C, D and E) located in Northern Italy, throughout the spring-autumn period from 2011 to 2015, when the water temperature was between 9 and 10°C. The percentage of affected fish in the farms was 10%-15%, and their mean weight was around 500 g.
Fish did not show alteration of behaviour or other signs of disease except for the cutaneous gross changes ascribable to RMS; mortality was absent. Forty-six symptomatic fish were sampled and killed by an overdose of MS-222 (300 mg/L); the number of individuals collected from each farm was as follows: A = 10, B = 10, C = 10, D = 8 and E = 8. Samples of skin, heart, spleen, kidney, liver, gill, brain and gut were included for the histological analysis. A single skin lesion was collected from each symptomatic fish, and then subjected to gross classification and processed for histology. For those individuals presenting multiple skin lesions, the most severe were selected as representative for this purpose.

| Gross classification criteria for RMS skin lesions
To classify the skin lesions of RMS-affected rainbow trout on the basis of gross appearance, the authors employed grading criteria as described by Galeotti et al. (2013) with some modifications. The included criteria for descriptive methodology were chosen considering a high number of skin lesions observed throughout the last years in Italian, Slovenian and Bosnian rainbow trout farms with full-blown RMS episodes. The five macroscopic parameters employed in this study to grade RMS skin lesions were as follows: size, colour variation, exudate, visible scales and erosion. Size ranging from few mm to several centimetres in diameter and considering the major diameter for the oval or elongated lesions. Colour variation from pale grey-whitish to bright red haemorrhagic patches. Presence of exudate: mild (sharp demarcated pale areas with slightly raised scales),

| Histology
Tissue samples obtained from skin/muscle, brain, gill, liver, heart, spleen, kidney and gut were fixed in 4% neutral buffered formaldehyde at 4°C overnight. After fixation, samples were equilibrated at room temperature, processed by an automatic histoprocessor (TISBE tissue processor, Diapath) and embedded in paraffin (Paraplast Plus, Diapath). Serial 5μm sections were obtained by a programmable microtome (Reichert-Jung 2050) and then stained with haematoxylin and eosin and Van Gieson Weigert trichrome. The specimens were examined with a light microscope (Leica DMRB), and images were acquired with a Nikon system. A single observer (MG) scored the histopathological parameters on skin samples, using semi-quantitative criteria. For each parameter, the score ranged from 1 to 4, as follows: 1 = absence of the lesion, 2 = mild lesion, 3 = moderate lesion and 4 = severe lesion. The parameters evaluated are shown in Table 2.

| Proliferating cell nuclear antigen (PCNA) immunohistochemistry
Twenty skin/muscle samples, representative of the scored lesions (6 macroscopically scored as type I, 5 as type II and 9 as type III), were evaluated with immunohistochemistry using an anti-PCNA antibody to assess the cell proliferation rate within the lesion. Four-micronthick formalin-fixed, paraffin-embedded sections were deparaffinized for 30 min in Solvent Plus (Carlo Erba) and then hydrated in a series of graded alcohols. Subsequently, endogenous peroxidase activity was blocked with 3% hydrogen peroxide in distilled water for 30 min at room temperature; sections were then treated in citrate buffer (pH 6.0) for 10 min in a microwave oven (750 W) for antigen retrieval. Non-specific binding was blocked by incubating the sections with 5% normal goat serum and 1% bovine serum albumin (BSA) and phosphate-buffered saline (PBS) (blocking solution) for 1 hr at room temperature, and then, sections were incubated overnight at 4°C with mouse monoclonal anti-PCNA (Clone PC10, Cell Signaling Technology) diluted 1:4,000 in blocking solution. A horseradish peroxidase (HRP) polymer conjugate SuperPicture™ Detection Kit (Zymed, Invitrogen) was applied to the sections for 10 min at room temperature to visualize antibody-antigen binding. The immunohistochemical reaction was developed with 3,3-diaminobenzidine solution (Sigma-Aldrich). The slides were counterstained with Papanicolaou haematoxylin, dehydrated in a series of graded alcohols and cleared in xylene before mounting.

| Statistical analysis
Score data of skin histological variables, listed in Table 1, and score of PCNA positivity were tested for normality using the Shapiro-Wilk W test. Because the data distribution was not normal, the correlation of the scored variables with the three types of gross lesions was performed using the non-parametric Spearman rank-order correlation test (CSS Statistica, Statsoft). A p value <.05 was considered significant.

| Gross skin lesions
Lesions were allocated to one of three categories as follows: type I = 7 lesions (Figure 1a, b), type II = 19 lesions (Figure 1c, d) and type III = 20 lesions (Figure 1e, f). The macroscopic examination of visceral organs did not reveal lesions apart from a slightly increased spleen volume in 18 out of 46 (39%) sampled trout. These spleens also showed a rough surface, and upon dissection, they showed congestion.

| Histology
The histological observation of skin sections allowed describing the main features of each gross lesion type, as follows:

| Type I
Mild gross skin lesions. The epidermis was always intact, only rarely showing areas with slight intra-epithelial lymphocyte exocytosis.
The single constant observation was a mild-to-moderate infiltration of lymphocytes and monocytes in the SS of the dermis, involving also the area surrounding the SP. The dermal SC in some samples (5 out of 7, 71%) appeared with a mild-to-moderate lymphocyte infiltration, whereas the hypodermis revealed a moderate infiltration of lymphocytes and monocytes ( Figure 2).

| Type II
Moderate gross skin lesions. There was moderate inflammation involving all the skin layers from the epidermis to the underlying muscular tissue (Figure 3a). The epidermis was often present or partially missing. In some specimens (7 out of 19, 37%), a multifocal epidermal necrosis or epidermal oedema with acanthosis was evident ( Figure 3b). Multifocal mild lymphocyte intra-epithelial infiltration was observed (Figure 3b). The hypodermis showed panniculitis with a moderate-to-severe infiltration of lymphocytes, plasma cells and macrophages (Figure 3h).
A mild fibrosis was also detectable. The underlying muscular tissue showed moderate-to-severe infiltration of lymphocytes and plasma cells, with frequent mitoses. Muscle inflammation was mainly localized beneath the areas of panniculitis and mainly involved the myosepta. In a few cases, both hypodermal tissue inflammation and muscular tissue inflammation showed a granulomatous pattern.
Inflammation was sometimes observed in deeper areas, along the intramuscular septa and up to the peritoneum.

| Type III
Severe gross skin lesions. A severe inflammation involving all the skin layers from the epidermis down to the underlying muscular tissue was always present (Figure 4a). The epidermis was absent in 10 out of 20 (50%) skin sections due to erosion and ulceration of the epithelium (Figure 4a). When the epidermis was present, necrosis and severe intra-epithelial lymphocyte infiltration were observed.
At the dermo-epidermal junction, cells necrosis, capillary congestion and haemorrhages were observed. In the case of skin ulceration, the lesion involved the epidermis, the SS and partially the SC.
The SS contained numerous necrotic cells and a severe inflammatory infiltrate composed of lymphocytes, macrophages and scarce neutrophils. Vascular congestion was always severe, accompanied by moderate haemorrhages and slight oedema. In some samples, the vessels displayed thrombosis (Figure 4c). Scales were absent due to complete resorption, and in some specimens, an early process of regeneration was detected. The SPs were numerically reduced or lacking, replaced by areas of chronic inflammation (recruitment of lymphocytes, plasma cells, macrophages and multinucleated giant cells) (Figure 4b). The SC displayed a severe cellular infiltrate including monocytes, macrophages, lymphocytes and plasma cells. Lesions affecting the internal organs were observed in trout with type II and type III lesions. Among the internal organs, histological lesions were observed in the heart, spleen, kidney and liver. In 19 out of 46 examined fish, heart lesions included pericarditis (17 fish, 37%) and myocarditis (2 fish, 4%), both with lymphocyte infiltration.
In the spleen, hyperplasia of the white pulp associated with perivascular histiocytic cell proliferation was observed in 18 out of 46 fish (39%). The liver showed foci of lymphocyte infiltration and lymphohistiocytic vasculitis in 13 out of 46 specimens examined (28%). In 14 out of 46 kidney specimens (30%), a mild vacuolar degeneration of tubular epithelium was observed. No pathological changes were observed in the gill, brain and gut samples.

| Data analysis by statistics
The correlation between the single histological scored parameter and the three types of gross lesions, assessed with the Spearman rank-order correlation, revealed several significant results but with different conclusive considerations (Table 2). A significant positive correlation (score increase) from type I to type III gross lesions was apparent with the following histological parameters (supplementary graphs a-f): progressive loss of the epithelium and scales, progressive recruitment of inflammatory cells in the SC of the dermis, loss of regular architecture of the SC of the dermis and increase in granuloma-like reactions surrounding the small vessels and muscular fibres (highest R value in Table 2). The above-mentioned variables thus acquired a higher score parallel to the progression of the macroscopic lesion. For several other variables, the results showed lower values in the type I group compared with the other groups, in which type II was similar to type III. These variables did not provide any indication of the progression of the lesion from type I to type III gross features; if scored with lower values, they only offered an indication of an early and mild gross feature.

| Immunohistochemistry
Several PCNA-positive cells were found in the epidermis (lymphocytes) (Figure 5a, b) and in the dermis (lymphocytes, fibroblasts and endothelial cells) (Figure 5c), but only PCNA-positive lymphocytes ( Figure 5d) were found in the hypodermis. A significant increase in PCNA-positive lymphocytes from gross type I to gross type III cases was revealed only in the hypodermis (supplementary graph g); no other association was apparent.

| D ISCUSS I ON
The clinical evaluation and the differential characterization of the skin lesions examined in the present study were based on the RMS case definition in rainbow trout proposed by Oidtmann et al. (2013).
Adding to the current state of knowledge, the results of the present work contribute particularly with the observations of type II and type III lesions; these data highlight that panniculitis and myositis may display a granulomatous pattern surrounding vessels as well as the nerves and muscle fibres. This pattern, when observed in the hypodermis, can be defined as panniculitis referring to the development of severe granulomatous reactions in the proximity of blood vessels and nerve fibres. This condition is consistently associated with severe myositis of the underlying tissue.
Furthermore, in some samples, the chronic inflammatory reaction also reaches the peritoneal serosa, with adhesions developing between the serosa and the gut wall. Furthermore, the SS of type III lesion in some samples shows vessels with occlusive thrombosis and early processes of scale regeneration, thus demonstrating that this regeneration occurs not only in fully healing injuries but also in ongoing injuries.
At the internal organ level, the most relevant microscopic lesions were observed in the heart and spleen. The pericarditis observed in 17 out of 46 trout affected by RMS represented the most prominent lesion, characterized by an intense lymphocytic infiltration of the pericardium. Similar findings were described by Ferguson et al., (2006), who reported heart lesions in 40% of SD-affected fish (20% of which showed diffuse myocarditis) and by Verner-Jeffreys et al., (2008), who reported epicarditis and cardiomyositis in 20% of the examined fish with RMS. In the spleen, hyperplasia of the white pulp associated with perivascular histiocytic cell proliferation was observed in 18 out of 46 individuals (39%), as also described by Metselaar et al., (2020). In 13 out of 46 trout examined, the liver showed foci of lymphocyte infiltration and vasculitis. The observation of liver focal necrosis was also recorded in RMS-affected fish by Noguera et al., (2008) and Verner-Jeffreys et al., (2006). Mild tubular degeneration in the kidney was observed in 14 out of 46 specimens, similarly to findings by Noguera et al., (2008) and Verner-Jeffreys et al., (2006), whereas a moderate-to-severe diffuse histiocytic proliferation of the interstitial haematopoietic tissue was also reported by Metselaar et al. (2010) and Metselaar et al. (2020).In the gill, brain and gut, no pathological changes were observed. All the lesions in the internal organs were recorded in trout showing macroscopic lesions classified as type II or type III, suggesting a systemic compromise.
The observations made in this study and the statistical data analysis allowed us to survey the evolution of the histological lesion and In conclusion, in the light of our experience with the histological observation of skin lesions from 46 RMS-affected trout, classified in three gross developmental stages, and based on the statistical analysis of the histological scores, we suggest an improvement of the initial first accurate descriptions of the "Strawberry disease" skin lesion "…focal, non-suppurative dermatitis with ulceration and extensive infiltration of mononuclear inflammatory cells" (Olson et al., 1985), proposing a new morphological diagnosis as "…deep