Eighty-two sow mammary tissue specimens exhibiting gross lesion(s) detected during meat inspection were acquired from slaughterhouses in Incheon and Gyeonggi (South Korea), in 2015 and 2016. Swab samples were obtained from 20 tissue specimens associated with suppurative exudates after incision. The samples were taken from sows which had experienced parturition more than 6 times (approximately 4 to 5 years old). The sows had been slaughtered because of their low reproduction rate due to poor milk secretion caused by the prolonged presence of mammary lesions; regardless, they did not exhibit any signs of systemic illness. Tissue specimens were fixed in 10% neutral buffered formalin, trimmed, embedded in paraffin, and sectioned (4 μm thick) per routine laboratory protocols [17]. From each sow, 1 or 2 sections of the mammary tissue specimen were stained with hematoxylin and eosin and histologically analyzed to categorize mastitis and hyperplasia in the lesions.

To characterize the lesions, the expressions of ER-α, ER-β, and the progesterone receptor (PR) in mammary hyperplastic lesions and the cluster of differentiation (CD) 3 and CD79a, as well as IL-1α, IL-1β, IL-6, and IL-8, in inflammatory lesions were analyzed by immunohistochemical assay using specific primary antibodies for each of these investigated targets. Briefly, sectioned tissues were dried for 1 day, followed by dewaxing in xylene, and then rehydrated through a succession of graded ethanol baths. Tissue sections were treated for 20 min at room temperature with 3% hydrogen peroxide in phosphate-buffered saline (PBS; pH 7.4, 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, and 2 mM KH₂PO₄) to inhibit endogenous peroxidase activity. Subsequently, heat-induced epitope retrieval was performed by boiling the section submerged in Tris-EDTA (pH 9.0; for PR, CD3, IL-6, and IL-8) or citric acid (pH 6.0; for ER-α, ER-β, CD79a, IL-1α, and IL-1β) in a microwave oven; the sections were then cooled in cold water. For each step, 3 PBS washes were performed. Normal goat serum (5%) or normal horse serum (2.5%, for IL-6) diluted with PBS were used as blocking agents, and primary antibodies were sequentially applied as described in Table 1. After washing off unattached primary antibodies with PBS, tissue sections were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (DAKO REAL Envision kit; DAKO, Denmark) for 40 min or biotinylated anti-goat serum for 15 min and conjugated with streptavidin-HRP (R.T.U. Vectastain kit; Vector Laboratories, USA) for 30 min (for IL-6) at room temperature. To detect immunolabeled proteins, diaminobenzidine was used as a visualizing agent. Tissue sections were then counterstained with Gill's hematoxylin, dehydrated in ethanol, cleared with xylene, and coverslipped. To confirm the specificity of the primary antibodies, normal mammary tissue and spleen and tonsil tissues from sows were used as positive controls for the target hormone receptors CD3/CD79a and IL-1α/IL-1β/IL-6/IL-8, respectively. Isotype-matched immunoglobulins were used as negative controls.

Nuclear expressions of ER-α, ER-β, and PR were considered to be positive when present in >10% of mammary gland epithelial cells [18]. For those factors expressed in the cytoplasm of inflammatory cells, including CD3, CD79a, IL-1α, IL-1β, IL-6, and IL-8, semi-quantitative morphological analyses were performed using the image analysis software ImageJ (version 1.50b) downloaded from the National Institutes of Health website (http://rsb.info.nih.gov/ij). One image at 40× (for CD3 and CD79a) or 5 images at 200× (for IL-1α, IL-1β, IL-6, and IL-8) magnification were acquired at the site of inflammation in each section. The mean of the assessed integrated density (the sum of the multiplied values of an area and the average gray value of the pixels in the image) divided by the tissue area in each image was calculated.

Isolation of causative bacteria was achieved by streaking the 20 swab samples on blood agar and incubating those in various environments viz. aerobic, anaerobic, and 5% carbon dioxide for >18 h at 37°C. Subsequently, the 16S ribosomal RNA (rRNA) genes of the isolated bacteria were amplified by performing polymerase chain reaction (PCR) with 27F/1492R universal primers under the following cycling conditions: 94°C for 5 min (pre-denaturation); 30 cycles at 94°C for 1 min (denaturation); 55°C for 1 min (annealing); 72°C for 1 min (extension); and 72°C for 10 min (final extension) [19]. The sequences of the PCR products were analyzed to identify species by using BLAST as provided by the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). Homology was accepted at the genus level of the isolated bacteria at the criterion of > 97% identity [20].

Pearson's correlation analyses were performed to validate the relevancy between immunoreactivity of the investigated cytokines in the inflammatory lesions; p < 0.05 was considered to indicate a statistically significant difference. SPSS version 17.0 (IBM, USA) was used for statistical analysis.

Eighty-two specimens were obtained from pigs that had experienced more than 6 parturitions and exhibited macroscopic, focally swollen, and raised mammary lesions with or without erosion or ulceration. Histopathological findings included mammary gland hyperplasia (n = 46), mammary gland hyperplasia with duct epithelial hyperplasia (n = 3), mastitis (n = 32), and mammary gland hyperplasia with mastitis (n = 1) (Table 2). Mammary gland hyperplasia was diagnosed when the number of epithelial cells composing the mammary gland was increased (Fig. 1A). Ductal hyperplasia was diagnosed in cases in which the number of ductal epithelial cells had increased and had accumulated in the forward center of the duct. All samples diagnosed with mastitis exhibited more than 1 central bacterial focus with inflammatory cells in its vicinity consisting mostly of lymphocytes, macrophages, and neutrophils (Fig. 1B).

Samples of hyperplastic lesions (n = 50) were immunohistochemically analyzed with ER-α, ER-β, and PR antibodies to determine the role(s) of these factors in the emergence of hyperplasia. None of the ER-α- and ER-β-stained samples were confirmed to be positive (Fig. 2A and B). By contrast, expression of PR was evident in > 10% of the mammary gland tissue in all examined samples (Fig. 2C).

Analyses of T- and B-cell distributions were performed using primary antibodies against CD3 and CD79a, respectively (Fig. 3A and B). Among the 32 mastitis samples, 24 (75%) exhibited more B cells than T cells. Expression of cytokines was observed in the cytoplasm of inflammatory cells in all mastitis samples (Fig. 4); however, some epithelial cells also exhibited these cytokines. Although the level of expression varied among samples, significant correlations were identified (Table 3). In samples with an increased presence of T cells, IL-1β expression was upregulated (p < 0.01). By contrast, the B-cell distribution was not significantly correlated with the expression of any cytokines. Among the cytokines, IL-1β and IL-6 had a significantly positive relationship (p < 0.01). Although significant correlations were not detected, T-cell presence exhibited a positive trend with IL-6 expression, but negative associations with IL-1α and IL-8 expressions. The B-cell levels demonstrated positive correlational trends with IL-1α and IL-6 expressions, but negative trends with IL-1β and IL-8 expressions.

Sequencing of the 16S rRNA gene was examined to identify the causative bacteria (Table 4). Of the 20 samples, 19 bacterial isolates (demonstrating > 97% identity) were allocated to 8 genera, with Pseudomonas spp. being the most frequently detected (n = 10), followed by Escherichia spp. (n = 2), Psychrobacter spp. (n = 2), and 1 isolate each of Aeromonas spp., Proteus spp., Rahnella spp., Staphylococcus spp., and Yersinia spp.