Novel inflammatory cell infiltration scoring system to investigate healthy and footrot affected ovine interdigital skin

Ovine footrot is a degenerative disease of sheep feet leading to the separation of hoof-horn from the underlying skin and lameness. This study quantitatively examined histological features of the ovine interdigital skin as well as their relationship with pro-inflammatory cytokine (IL-1β) and virulent Dichelobacter nodosus in footrot. From 55 healthy and 30 footrot ovine feet, parallel biopsies (one fixed for histology) were collected post-slaughter and analysed for lesions and histopathological analysis using haematoxylin and eosin and Periodic Acid-Schiff. Histological lesions were similar in both conditions while inflammatory scores mirror IL-1β expression levels. Increased inflammatory score corresponded with high virulent D. nodosus load and was significant (p < 0.0001) in footrot feet with an inflammatory score of 3 compared to scores 1 and 2. In addition, in contrast to healthy tissues, localisation of eubacterial load extended beyond follicular depths in footrot samples. The novel inflammatory cell infiltration scoring system in this study may be used to grade inflammatory response in the ovine feet and demonstrated an association between severity of inflammatory response and increased virulent D. nodosus load.


INTRODUCTION
Lameness in sheep, caused by interdigital dermatitis (ID) and footrot, is widespread in the UK, having a major welfare and economic impact. Footrot is caused by Dichelobacter nodosus which occurs in both virulent and benign forms. Footrot is defined by separation of the hoof from the underlying structures and an accumulation of necrotic material, with varying degrees of severity (Beveridge, 1941;Thomas, 1962), this damage is believed to be mediated by immune pathology rather than by bacterial enzymes and toxins (Egerton, Roberts & Parsonso, 1969).
Footrot development is characterised by invasion of neutrophils and lymphocytes into the dermis and epidermis in response to bacterial invasion of the epidermis (Davenport et al., 2014;Egerton, Roberts & Parsonso, 1969). Early histological studies described footrot as a degenerative condition of the stratum granulosum and spinosum which results in cellular degeneration (cell ballooning), formation of micro-abscesses and vacuoles which coalesce and progress to cavities (Beveridge, 1941;Deane & Jensen, 1955;Thomas, 1962). A recent histological study of clinically healthy and affected feet showed a progressive increase in lymphocyte and neutrophil infiltration into the dermis and epidermis between healthy, ID and footrot samples (Davenport et al., 2014). In footrot samples, purulence was seen in areas of epidermal degeneration, necrosis and epidermal-dermal clefts. As in previous studies, cytoplasmic ballooning and nuclear condensation were observed in the stratum spinosum of the epithelium (Thomas, 1962), as well as areas of fibrosis indicating a chronic reaction to tissue damage (Egerton, Roberts & Parsonso, 1969).
The aim of this study was to perform a qualitative and quantitative analysis of histopathological features of the ovine interdigital skin comparing healthy and footrot affected feet. In order to evaluate inflammation in the epidermis and dermis a novel scoring system was developed and correlation of those scores with the expression of the pro-inflammatory cytokine IL-1β and D. nodosus load was investigated. In addition the depth of within tissue colonisation of eubacteria, D. nodosus and Fusobacterium necrophorum were examined in the context of hair follicle depth.

Ovine biopsies
Samples of ovine feet were obtained post-slaughter from an abattoir and assessed by two independent scorers for conformation and clinical conditions (healthy and footrot affected).
Conformation scoring was assessing the integrity of the sole and heel/wall of each digit: 0, undamaged sole and heel area with a perfect shape; 1, mildly damaged/misshapen sole and/or heel area of the digit (<25%); 2, moderately damaged/misshapen sole and/or heel area of the digit (>25% and <75%); 3, severely damaged/misshapen sole and/or heel area of the digit (>75%) (Maboni et al., 2016). Ovine feet were scored as described previously (Kaler et al., 2010) with healthy defined as an absence of any interdigital skin lesion and footrot as the presence of underrunning lesions. Roughage, faeces and mud were removed from the interdigital space of each foot and prior to biopsy taking the skin was wiped with 70% ethanol. Biopsies of 6 mm diameter (n = 85) were taken using a punch biopsy tool (National Veterinary Services, UK) in the interdigital space along the skin-hoof interface, of which 55 were from interdigital skin without signs of disease and 30 from feet with footrot.

Histological sample preparation
Preparation was as previously described (Davenport et al., 2014), in brief all samples were fixed using 10% (v/v) neutral buffered formalin; however to obtain optimal tissue integrity healthy biopsies were fixed for 48 h at 4 • C and footrot biopsies for 24 h at room temperature (fixation and processing were optimised using two fixatives (10% v/v neutral buffered formalin verses 4% v/v paraformaldehyde), two incubation temperatures (room temperature verses 4 • C) and two incubation times (24 h verses 48 h). An extended tissue processing protocol was used (60 min in dH 2 O, 4 h in 50% ethanol, 4 h in 70% ethanol, 16 h in 90% ethanol, 4 h in 100% ethanol, 4 h in Xylene) followed by embedding into paraffin wax (2 h, 60 • C). Paraffin wax embedded tissues were soaked in 10% (v/v) ammoniated water and 6 µm thick sections were cut from each block by microtome (RM2255; Leica, Wetzlar, Germany). Serial sections were mounted on polysilinated microscope glass slides (Menzel Gläser Polysine R ; Thermo-Scientific, Darmstadt, Germany) and dried at room temperature overnight.

Tissue staining
Paraffin sections were heated at 60 • C for 5-10 min, incubated in xylene twice for 5 min each and rehydrated in 100% ethanol, 90% ethanol, 70% ethanol then twice in dH 2 O for 5 min each. Protocols were optimized for haematoxylin and eosin (H&E) and Periodic Acid-Schiff (PAS; Sigma-Aldrich, Dorset, UK). In brief H&E stained samples received 2.5 min in haematoxylin, 15 s in 1% acetic industrial methylated spirits, 15 s in ammoniated water and 4 min in eosin. The PAS stain included immersion in periodic acid-Schiff solution for 15 min, haematoxylin for 3 min. Following each staining protocol sections were dehydrated through an ethanol series.

Image capture and analysis
Each observer was blinded to the sample identification to avoid subconscious bias. Images were captured using a Leica CTR500 microscope (Leica Microsystems, Darmstadt, Germany) with bright field light. For each sample, three sections approximately 400 µm apart were analysed. At 40× magnification, five non-overlapping photos were taken from each section from both the epidermis (1,275 photomicrographs analysed) and the dermis (1,245 photomicrographs-two fragmented dermal tissue sections were excluded) of H&E stained sections and from the dermal-epidermal junctions of PAS stained sections (1,275 photomicrographs analysed). In addition each sample was viewed at 5× magnification and photomicrographs were merged in order to visualise entire sections.

Tissue and cell scoring
For each tissue sample, 15 photomicrographs were scored for inflammation. The maximum score of five non-overlapping images from each of three slides per block was determined as the score for that tissue sample. A score of 0 represented no leukocytes in photomicrograph, score 1 occasional infiltration of single leukocytes visible, score 2 focal infiltration of leukocytes, score 3 coalescing leukocytes-individual loci could not be distinguished, and score 4 diffuse infiltration of leukocytes throughout the field of view. Examples of the epidermal and dermal scoring system are shown in Figs. 1 and 2, respectively. The scoring system was validated using two independent scorers blinded to sample identification.  Parakeratosis was defined as the retention of nuclear remnants in the stratum corneum (Brady, 2004). Grading in the interdigital skin stratum corneum was based on a three ordinal scale criteria as: the absence of parakeratosis (score 0), the presence of focal parakeratosis (score 1) and diffuse parakeratosis (score 2) (Figs. 3A-3C).  Micro-abscesses were defined as the aggregation of inflammatory cells and cellular debris, confined by fibrotic tissue walling. Grading was according to skin layer (intracorneal, sub-corneal and dermal (Figs. 3D-3F)) and based on a nominal grading scale: presence/absence from five images throughout each tissue section. In order to determine the area of ballooned cells per cross sectional area in the epidermis, photomicrographs were taken, saved in a Tagged Image File Format and uploaded into the analysis software Image Pro 6.3 (Media Cybernetics, Rockville, MD, USA). Each image was calibrated for the measurement of ballooned cells area in absolute values (µm 2 ). Ballooned cells were identified and manually selected using the software to trace around each ballooned cell or group of cells. Percentage area of ballooned cells per photomicrograph calculated (total area of ballooned cell(s)/total area of epidermal cross section × 100, for example see Fig. 4A).
Congestion was defined as presence of dilated blood capillaries with visible red blood cells (erythrocytes) within the dermis and haemorrhage was defined as observed blood cells outside capillaries in tissues (Figs. 4B-4C). For both features, 15 photomicrographs were analysed per tissue with a nominal grading scale: presence/absence. Basement membrane integrity was assessed using PAS stained sections scored for basement membrane disruptions at the dermal-epidermal junction as follows: score 0, no basement membrane disruption identified; score 1, focal basement membrane disruption identified; score 2, multiple basement membrane disruption identified (Figs. 4D-4F).
The IL-1β expression and D. nodosus quantification data have been published previously (Maboni et al., 2017a).

Bacterial localisation
Tissue samples were briefly incubated in 70% ethanol followed by overnight incubation in 30% (w/v) sucrose and embedded in OCT (VWR International, Oud-Heverlee, Belgium). Alternating thick (40 µm) and thin (9 µm) transverse sections were sectioned from the dermal layer across biopsies into the epidermis. The cryostat blade was cleaned with 70% ethanol prior to each thick section in order to prevent potential bacterial contamination. Thick sections intended for DNA extraction to determine bacterial abundance were preserved in 0.5ml RNAlater R (Sigma-Aldrich, St. Louis, MO, USA) at room temperature and incubated overnight prior to DNA extraction. Corresponding alternate thin sections underwent H&E staining. DNA was isolated using the QIAamp cador R kit (QIAGEN, Hilden, Germany) as described previously (Maboni et al., 2016). Bacterial load was quantified using quantitative PCR as described previously for total eubacteria (Strub et al., 2007), D. nodosus (Frosth et al., 2012 and F. necrophorum (Frosth et al., 2015). In order to compare bacterial localisation and load across the same depth of skin layers across different samples, 11 bin groups representing skin depths from 0 to 2,200 µm (range 200 µm/bin) were created. N = 4 healthy and five footrot samples.

Statistical analysis
Statistical analyses were performed on Graphpad Prism version 6 for Windows. Resulting data were presented as frequencies and percentages. Categorical data within and between clinical conditions were compared by Fisher's exact and Chi-square tests while continuous data were analysed by Student T -test or Kruskal Wallis test, dependant on data distribution. Statistical data on DNA samples was carried out using Pearson correlation. Analysis was taken as significant when p ≤ 0.05.

Ethical approval
Ethical approval was obtained from the School of Veterinary Medicine and Science Ethics committee, University of Nottingham (ethical approval number: 796 130128).
We have shown previously that mRNA expression of IL-1β, similar to inflammation scores, were comparable in healthy and footrot affected tissues in parallel samples from the same feet (Maboni et al., 2017a). Here we analysed inflammatory scores, parakeratosis and presence of micro-abscesses in the context of IL-1β mRNA expression. IL-1β expression was significantly higher in the presence of diffuse parakeratosis in healthy tissues with no difference in footrot affected tissues (Figs. 7A-7B; n = 41 healthy and 23 footrot, p < 0.01). When comparing relative IL-1β expression with the inflammation scores they mirrored to an extent (Fig. 8).
Since footrot pathology is mediated by host inflammatory responses whilst the disease is initiated by D. nodosus, a virulent D. nodosus load against host inflammatory cells was compared in healthy (n = 30 for both epidermis and dermis) and affected tissues (n = 18 epidermis and 17 dermis). Inflammatory scores increased as virulent D. nodosus load increased (Fig. 9). In the epidermis, the virulent D. nodosus load was significantly higher (p ≤ 0.0001) in tissues with an inflammatory score of 3 compared to scores 1 and 2 respectively (Fig. 9B, Fig. S1).

Bacterial localisation in ovine interdigital skin
Eubacterial D. nodosus and F. necrophorum DNA was quantified across the entire skin depth. Eubacterial load in healthy samples (n = 5) was similar throughout the tissue depths but peaked at depth of 201-400 µm (Fig. 10). In footrot samples (n = 4) eubacterial load progressively decreased from the outermost skin surface to a depth of 1,000 µm, eubacterial load was then similar from 1,200-2,200 µm. In footrot samples, D. nodosus was quantified across depths of 601-2,200 µm, whereas it was detected throughout all levels of the healthy samples. F. necrophorum in healthy samples was detected in deeper skin tissues (601-2,200 µm) but was present throughout every levels in footrot samples (Fig. 10). The distribution pattern of eubacterial load in ovine interdigital skin samples were investigated with respect to depth of skin sections in the context of depth of follicles in skin. A relatively low eubacterial load of range of 0.4-19.6 pg/section was observed across healthy samples. In comparison, footrot samples eubacterial load was higher in four out of five samples with peak values of 1,218 pg/section observed in the superficial skin depth (≤500 µm; Fig. 10). Localisation of eubacterial load did not extend beyond follicular depths (2,119-2,710 µm) in healthy samples while in footrot samples, eubacterial load extended beyond follicular depths (1,460-2,699 µm; Fig. 10, Fig. S2).

DISCUSSION
In this study, histopathological features of the ovine interdigital skin were analysed and compared between healthy and footrot affected feet using a novel scoring system. In addition, these features were compared to levels of IL-1β and virulent D. nodosus. Of note in this study was the severity of inflammatory cell infiltration that was found to be similar between healthy and footrot affected. Inflammation in healthy samples may have been caused by unfavourable ground conditions of the pasture due to wet weather. Under those condition previous studies have also noted an increase in increase the prevalence of footrot and interdigital dermatitis (Beveridge, 1941;Emery, Stewart & Clark, 1984;Graham & Egerton, 1968;Wassink et al., 2003;Wassink et al., 2004). Alternatively, inflammatory cell infiltrations could be an indication of subclinical disease not yet progressed to visible signs of ID and footrot. Previous studies have qualitatively described mild inflammatory cell infiltration in healthy ovine feet (Angell et al., 2015;Davenport et al., 2014). To the best of our knowledge, this is the first study to quantitatively examine inflammatory cells in the ovine feet using a new scoring system. Similar investigations in cattle also found inflammatory cell infiltration in apparently healthy hooves (Mendes et al., 2013;Tarlton et al., 2002). However, Tarlton and colleagues attributed the inflammatory response observed to the changes in collagen expression and keratinization of epidermal laminae caused by intensive breeding regimes during the peripartum period (Tarlton et al., 2002). Interestingly, bovine digital dermatitis, which is driven by inflammatory response in its pathogenesis, is also characterised by inflammatory cell infiltration (Mendes et al., 2013;Refaai et al., 2013).
Histological lesions including cell ballooning, parakeratosis and micro-abscesses were similarly observed in both conditions and in areas associated with pathology including inflammation in the skin (Refaai et al., 2013). Cellular ballooning in the epidermis is an indication of host cell response to stimuli and may be preceded by keratinocytes hyper-proliferating to remove pathogen infected areas (sloughing) (Edwards, Dymock & Jenkinson, 2003). Hyper-proliferation of keratinocytes without a corresponding rate of differentiation results in parakeratosis. This has been shown to occur in low grade inflammation caused by microbial presence such as F. necrophorum (Egerton, Roberts & Parsonso, 1969). Parakeratosis may trigger the release of pro-inflammatory cytokines such as IL-1β (Chang et al., 1992) which in-turn mediates the recruitment of inflammatory cells.
skin. In contrast, in a similar polymicrobial disease of bovine feet, digital dermatitis, no association was found between treponemes and hair follicles (Evans et al., 2009).

CONCLUSIONS
Data presented in this study showed for the first time that there are no consistent differences in the level and range of histological lesions examined between healthy and footrot affected feet. Interestingly, the novel inflammatory cell infiltration scoring system developed and validated in this study mirrored the pro-inflammatory cytokine IL-1β and confirmed an association between severity of inflammatory response and increased virulent D. nodosus load.