Lipopolysaccharides, cytokines, and nitric oxide affect secretion of prostaglandins and leukotrienes by bovine mammary gland during experimentally induced mastitis in vivo and in vitro

Domestic Animal Endocrinology 52 (2015) 90–99

Contents lists available at ScienceDirect

Domestic Animal Endocrinology journal homepage: www.domesticanimalendo.com

Lipopolysaccharides, cytokines, and nitric oxide affect secretion of prostaglandins and leukotrienes by bovine mammary gland during experimentally induced mastitis in vivo and in vitro K.K. Piotrowska-Tomala a, M.M. Bah a, K. Jankowska a, K. Lukasik a, P. Warmowski b, A.M. Galvao a, D.J. Skarzynski a, * a

Department of Reproductive Immunology and Pathology, Institute of Animal Reproduction and Food Research, Polish Academy of Sciences, 10-747 Olsztyn, Poland b Private Veterinary Clinic “Taurus”, 83-300 Kartuzy, Poland

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 September 2014 Received in revised form 3 March 2015 Accepted 4 March 2015

The aim of the study was to determine the effects of lipopolysaccharide (LPS), tumor necrosis factor alpha (TNF), interleukin-1-alpha (IL-1a), and nitric oxide donor (NONOate) on both in vivo and in vitro secretion of prostaglandin (PG)E2, PGF2a, leukotriene (LT)B4, and LTC4 by the bovine mammary gland. In the first experiment, tissues isolated from the teat cavity and lactiferous sinus were treated in vitro with LPS (10 ng/mL), TNF (10 ng/mL), IL-1a (10 ng/mL), NONOate (10
4 M), and the combination of TNF þ IL-1a þ NONOate for 4 or 8 h. PGE2 or PGF2a secretion was stimulated by all treatments (P < 0.05) excepting NONOate alone, which did not stimulate PGF2a secretion. Moreover, all factors increased LTB4 and LTC4 secretion (P < 0.05). In the second experiment, mastitis was experimentally mimicked in vivo by repeated (12 h apart) intramammary infusions (5 mL) of (1) sterile saline; (2) 250-mg LPS; (3) 1-mg/mL TNF; (4) 1-mg/mL IL-1a; (5) 12.8-mg/mL NONOate; and (6) TNF þ IL-1a þ NONOate into 2 udder quarters. All infused factors changed PGE2, 13,14-dihydro,15-keto-PGF2a, and LT concentrations in blood plasma collected from the caudal vena cava, the caudal superficial epigastric (milk) vein, the jugular vein, and the abdominal aorta (P < 0.05). In summary, LPS and other inflammatory mastitis mediators modulate PG and LT secretion by bovine mammary gland in both in vivo and in vitro studies. Ó 2015 Elsevier Inc. All rights reserved.

Keywords: Mastitis Prostaglandin Leukotriene Mammary gland Cow

1. Introduction The negative economic impact of acute coliform mastitis on the dairy industry demands new insights on the pathophysiology of the process [1]. Inflammation of the mammary gland may be associated with several bacteria, which invade the udder through the teat canal. Mastitis

* Corresponding author. Tel./fax: þ48 89 539 31 30. E-mail address: [email protected] (D.J. Skarzynski). 0739-7240/$ – see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.domaniend.2015.03.001

caused by coliform bacteria Escherichia coli is typically more severe and more frequently observed in high producing dairy cows around parturition and during early lactation [2]. Primary cells involved in the initial defense line are the resident mammary macrophages and the bovine mammary epithelial cells [3]. Pathogen-associated molecular patterns such as lipopolysaccharide (LPS) may be sufficient to elicit mastitis by E coli [2] and trigger an innate immune response with subsequent release of proinflammatory cytokines [4,5]. Several studies have suggested that levels

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

91

of tumor necrosis factor alpha (TNF), interleukin-1-alpha (IL-1a), and free oxygen radicals are significantly increased in milk and serum of cows during LPS-induced mastitis [6,7]. Differential regulation of local secretion of cytokines and other inflammatory mediators may result in acute mastitis, which causes different clinical symptoms [5,6]. As a result, reproductive performance in high production cows may be affected [1,8]. Arachidonic acid (AA) metabolites, such as prostaglandins (PGs) and leukotrienes (LTs), are enzymatically generated by prostaglandin endoperoxide synthase 2 and arachidonate 5 lipoxygenase during the inflammatory process. They act as major proinflammatory mediators and might play a critical role in the severity of mastitis in cows and their susceptibility to E coli [9,10]. Certainly, eicosanoids locally released during mastitis in the bovine mammary gland are important mediators of vascular permeability, chemotaxis, or hyperalgesia [11]. Our previous studies have shown that basic mediators of inflammation during E coli mastitis (TNF, IL-1a, and NO) may locally modulate the secretion of PG and LT by isolated bovine mammary epithelial cells; this response may be an important first-line defense mechanism for the mammary gland [12]. Consequently, we hypothesize that during inflammation of bovine mammary gland, locally produced inflammatory mediators TNF, IL-1a, NO, and AA metabolites act systemically and impact reproductive function. To address this question, we investigated: (1) the in vitro effect of inflammatory mediators on PG and LT secretion by bovine mammary gland tissues and (2) the in vivo effect of LPS and mastitis inflammatory mediators on PG and LT secretion in bovine mammary gland.

2%, ScanVet, Poland). Local epidural anesthesia was induced by injecting 4 mL of 2% procaine hydrochloride (Polocainum Hydrochloricum; Biowet Drwalew, Poland) between the first and second coccygeal vertebrae, and 4 catheters (Tomel Sp, Poland) were inserted for frequent collection of blood samples as follows:

2. Materials and methods

2.3. Collection of tissues from the teat cavity and lactiferous sinus of the bovine mammary gland

(1) (outer diameter (o.d.) ¼ 1.6 mm and inner diameter (i.d.) ¼ 1.2 mm) into the caudal vena cava via the coccygeal vein as previously described [13]; (2) (o.d. ¼ 2.1 mm and i.d. ¼ 1.6 mm) into the milk vein; (3) (o.d. ¼ 2.1 mm and i.d. ¼ 1.6 mm) into the jugular vein; (4) (o.d. ¼ 1.5 mm and i.d. ¼ 1.1 mm) into the posterior abdominal aorta through the coccygeal artery as previously described [13,14].

2.2. Mammary gland tissue collection Mammary glands (n ¼ 30) were collected postmortem at a local abattoir, from cyclic cows in the luteal phase of the estrous cycle (day 8–12). Milk samples were collected from selected animals and taken to the laboratory for bacteriologic tests (data not shown). Health of the udders (cows free from clinical mastitis), as well as the phase of the estrous cycle, was confirmed by veterinary inspection. The whole mammary gland was collected within 5 min after death, washed, and kept on ice during transport to the laboratory.

2.1. Animals and surgical procedures All animal procedures were approved by the Local Animal Care and Use Committee, University of Warmia and Mazury in Olsztyn, Poland (Agreement No. 83⁄2009). Normally cycling Polish Holstein–Friesian cows (4–6 lactations) were used for this study (n ¼ 36). There was no history of mastitis in any of the cows used. The animals were culled by the owners from 2 dairy herds because of their low level of milk production. The cows were examined and scored generally for rectal temperature, general attitude, respiratory and heart rates, and specifically, for udder quarter swelling and pain (data not shown). The cows used in this study were free of major mastitis pathogens, diagnosed negative by bacteriologic examinations with a quarter foremilk somatic cell count below 100,000 cells/mL (data not shown). To carry out the in vivo experiments, the estrus was synchronized using an analog of PGF2a (dinoprost, Dinolytic; Pharmacia and Upjohn, Belgium) injected twice with an interval of 11 d, as previously described [13]. The onset of estrus was taken as day 0 of the estrous cycle. During the experiment, animals were premedicated with xylazine (25–30 mg/animal intramuscular; Xylavet

Tissues representing mammary gland mucosa were collected from the teat cavity and lactiferous sinus of bovine mammary gland and used in this study. Udders were separated into quarters, which were cut along the teat canal. Briefly, strips from the teat cavity and lactiferous sinus were washed 3 times in sterile phosphatebuffered saline containing 20-mg/mL gentamicin (G1397; Sigma–Aldrich, USA). The tissue was then cut into small pieces (30–50 mg), washed again in phosphate-buffered saline with 20-mg/mL gentamicin, and finally placed in a glass culture tube (12  75 mm) containing 2 mL of Dulbecco’s Modified Eagle’s medium (DMEM) and Ham’s F-12 phenol red free culture medium (DMEM/F12, D-2906; Sigma–Aldrich), supplemented with 0.1% bovine serum albumin (A9056; Sigma–Aldrich), 10,000 U/mL penicillin G, 10-mg/mL streptomycin, and 25-mg/mL amphotericin B (antibiotic antimycotic solution, A5955; Sigma–Aldrich). Tissues were preincubated in a water bath with shaking at 38.5 C under 5% CO2 for 1 h and then incubated in the same conditions for 4 or 8 h. Time and doses of LPS, TNF, IL-1a, and NONOate in vitro and in vivo treatments were previously determined (data not shown).

92

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

2.4. Experimental procedures 2.4.1. Experiment 1: In vitro effect of LPS and inflammatory mediators of mastitis on PG and LT secretion by bovine mammary gland tissues After preincubation, culture medium was replaced by fresh DMEM supplemented with 0.1% bovine serum albumin and 10,000 U/mL penicillin G, 10-mg/mL streptomycin, and 25-mg/mL amphotericin B (antibiotic antimycotic solution, A5955; Sigma–Aldrich), and tissues (n ¼ 5/treatment) were treated 4 or 8 h with (1) no exogenous factors (control); (2) LPS from E coli 055: B5 (10 ng/mL, L2880–100 MG; Sigma– Aldrich); (3) TNF (10 ng/mL, recombinant human TNF, HF-13; Dainippon Pharmaceutical Co, Ltd, Osaka, Japan); (4) IL-1a (10 ng/mL, recombinant human IL-1a: HL-18; Dainippon Pharmaceutical Co,); (5) NONOate (10
4 M, 82150 Spermine NONOate, Cayman Chemical Co, Ann Arbor, MI, USA); and (6) TNF (10 ng/mL) þ IL-1a (10 ng/mL) þ NONOate (10
4 M). After incubation, the culture medium was collected in tubes with 5-mL EDTA and 1% acetylsalicylic acid solution (pH 3), and frozen at
20 C until determination of PG and LT by enzyme immunoassay (EIA). 2.4.2. Experiment 2: In vivo effect of LPS and inflammatory mediators of mastitis on PG and LT secretion by bovine mammary glands At day 8 of the estrous cycle, cannulae were inserted into the following vessels: (1) caudal vena cava, (2) milk vein, (3) jugular vein, and (4) posterior abdominal aorta. In the next day, animals were divided in the 6 groups (n ¼ 6/group), and immediately after milking, the following treatments were injected into the right udder quarters (anterior and posterior): group 1, sterile saline (5 mL); group 2, 250-mg LPS in 5 mL of saline; group 3, 1-mg/mL TNF in 5 mL of saline; group 4, 1-mg/ mL IL-1a in 5 mL of saline; group 5, 12.8-mg/mL (10
4 M) NONOate in 5 mL of saline; and group 6, TNF (1 mg/mL) þ IL1a (1 mg/mL) þ NONOate (12.8 mg/mL), all in 5 mL of saline. Twelve hours after the first treatment, animals were subjected to the second injection. The intramammary administrations were performed at
12 h and 0 h of experiment. Blood samples were collected at
12, 0, 1, 2, 4, 8, 12, 18 and 24 h of experiment into sterile 10 mL tubes containing 100 mL of 0.3-M EDTA and 1% acetylsalicylic acid, pH 7.4. After centrifugation (1,500g, at 4 C, 10 min), plasma was stored at
20 C until further analysis. 2.5. Eicosanoids determination 2.5.1. PGE2 determination Concentrations of PGE2 in plasma samples or in culture media were determined by EIA as previously described [13,15]. Samples were measured in duplicate. The PGE2 standard curve ranged from 0.07 ng/mL to 20 ng/mL, and the effective dose for 50% inhibition (ID50) of the assay was 1.25 ng/mL. Intra-assay and interassay coefficients of variation (Cv) were 6.9% and 9.7%, respectively. 2.5.2. 13,14-Dihydro,15-keto-prostaglandin F2a (PGFM) determination Concentration of PGFM in plasma samples was determined by EIA as previously described [13]. Samples were

measured in duplicate. The PGFM standard curve ranged from 32.5 pg/mL to 8,000 pg/mL, and the effective dose for ID50 of the assay was 315 pg/mL. Intra-assay and interassay Cv were 7.6% and 10.4%, respectively. 2.5.3. PGF2a determination Concentration of PGF2a in culture media was determined by EIA as previously described [16]. Samples were measured in duplicate. The PGF2a standard curve ranged from 0.195 pg/mL to 50 pg/mL, and the effective dose for ID50 of the assay was 1.82 pg/mL. Intra-assay and interassay Cv were 4.9% and 5.4%, respectively. 2.5.4. LTB4 determination Concentrations of LTB4 in plasma samples or culture media were determined using commercially available ELISA kit (# 520111, LTB4 EIA Kit; Cayman Chemical Company, USA) according to the manufacturer’s instructions. Samples were measured in duplicate. The LTB4 standard curve ranged from 1.96 pg/mL to 1,000 pg/mL, and the effective dose for ID50 of the assay was 2.5 pg/mL. Intraassay and interassay Cvs were 4.1% and 6.2%, respectively. 2.5.5. LTC4 determination Concentrations of LTC4 in plasma samples or culture media were determined using commercially available ELISA kit (# 520211, LTC4 EIA Kit; Cayman Chemical Company, USA) according to the manufacturer’s instructions. Samples were measured in duplicate. The LTC4 standard curve ranged from 0.98 pg/mL to 500 pg/mL, and the effective dose for ID50 of the assay was 1.85 pg/mL. Intraassay and interassay Cv were 4.9% and 7.4%, respectively. 2.6. Statistical analysis In experiment 1, data are shown as the mean  standard error of the mean of values obtained in 6 independent experiments, each performed in triplicate. Levels of PG or LT were expressed as ng/g tissue. All values were presented as fold of basal output. Data from experiment 1 were analyzed using 1-way analysis of variance tests followed by a Newman-Keuls Multiple Comparison Test (GraphPad Prism, version 5.00; GraphPad Software). P < 0.05 was considered significant. In Experiment 2, analyses of PG and LT in plasma samples were performed using two-way analysis of variance tests followed by a Bonferroni Multiple Comparison Test (GraphPad Prism version 5.00, GraphPad Software) as previously described [13,17]. P less than 0.05 was considered significant. 3. Results 3.1. Experiment 1: In vitro effect of LPS and inflammatory mediators of mastitis on PG and LT secretion by bovine mammary gland tissues The concentration of 10 ng/mL for LPS and cytokines and 10
4 M for NONOate produced the most consistent results in preliminary experiments, after 8 h treatment for PG and 4 h for LT.

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

Fig. 1. Effects of LPS (10 ng/mL), TNF (10 ng/mL), IL-1a (10 ng/mL), NONOate (10
4 M), and the combination of TNF (10 ng/mL) þ IL-1a (10 ng/mL) þ NONOate (10
4 M) on PGE2 and PGF2a secretion by bovine mammary gland tissues. The exposure time for PG was 8 h. Data are the mean  SEM of 6 separate experiments. All values are expressed as n-fold change from control. The concentrations of PG in the controls were (A) PGE2: 143.1  9.3 ng/g tissue; (B) PGF2a: 56.1  3.9 ng/g tissue. Asterisks indicate significant differences (P < 0.05) among treatments. IL, interleukin; LPS, lipopolysaccharide; NONOate, nitric oxide donor; PG, prostaglandin; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

Treatment of bovine mammary gland tissue with LPS showed increased secretion of PGE2 (P < 0.01; Fig. 1A) and PGF2a (P < 0.001; Fig. 1B). Both TNF and IL-1a treatments also augmented the output of PGE2 and PGF2a (P < 0.001; Fig. 1A, B, respectively), whereas NONOate exclusively

Fig. 2. Effects of LPS (10 ng/mL), TNF (10 ng/mL), IL-1a (10 ng/mL), NONOate (10
4 M), and the combination of TNF (10 ng/mL) þ IL-1a (10 ng/mL) þ NONOate (10
4 M) on LTB4 and LTC4 secretion by bovine mammary gland tissues. The exposure time for LT was 4 h. Data are the mean  SEM of 6 separate experiments. All values are expressed as n-fold change from control. The concentrations of LT in the controls were (A) LTB4: 312.2  34.9 pg/g tissue; (B) LTC4: 848.9  19.6 pg/g tissue. Asterisks indicate significant differences (P < 0.05) among treatments. IL, interleukin; LPS, lipopolysaccharide; LT, leukotriene; NONOate, nitric oxide donor; PG, prostaglandin; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

93

stimulated PGE2 (P < 0.001; Fig. 1A). In addition, the secretion of both PGE2 and PGF2a was increased after treatment with the combination of TNF þ IL-1a þ NONOate (P < 0.001; Fig. 1A and P < 0.05; Fig. 1B, respectively). Regarding LT production, LPS treatment stimulated both LTB4 and LTC4 (P < 0.001; Fig. 2A, B, respectively). Both LT were increased by TNF (P < 0.01; Fig. 2A and P < 0.001; Fig. 2B). Moreover, IL-1a and NONOate augmented production of LTB4 (P < 0.05; Fig. 2A, respectively) and LTC4 (P < 0.001; Fig. 2B, respectively). Finally, tissue treatment with TNF þ IL-1a þ NONOate caused the same effect and increased LTB4 and LTC4 production (P < 0.05; Fig. 2A, B, respectively). Summarizing, PGE2 and PGF2a secretion was stimulated by all treatments, with the exception of NONOate alone, which did not change the secretion of PGF2a in bovine mammary gland tissue. Moreover, LTB4 and LTC4 secretion was stimulated by all treatments.

3.2. Experiment 2: In vivo effect of LPS and inflammatory mediators of mastitis on PG and LT secretion by bovine mammary gland tissues–plasma collections from the (1) caudal vena cava, (2) milk vein, (3) jugular vein, and (4) abdominal aorta

(1) Samples collected from the caudal vena cava presented an increase in PGE2 concentration after intramammary infusions with LPS (blood collections between 2 and 12 h; P < 0.05; Fig. 3A), TNF (collections between 8 and 14 h; P < 0.001; Fig. 3A), IL-1a (collections between 1 and 24 h; P < 0.001; Fig. 3A), or TNF þ IL-1a þ NONOate (collections between 2 and 8 h; P < 0.05; Fig. 3A). Plasma PGFM concentration was increased in samples collected at 18 h after intramammary infusions with TNF (P < 0.05; Fig. 3B), NONOate (collections between 1 and 2 h and 12 and 24 h; P < 0.05; Fig. 3B), and TNF þ IL1a þ NONOate (collections between 0 and 2 h and 18 and 24 h; P < 0.05; Fig. 3B). Regarding LTB4, infusions with TNF þ IL-1a þ NONOate increased its concentration in plasma samples collected at 0 h (P < 0.05; Fig. 3C). Finally, infusions with NONOate increased the LTC4 concentration in plasma samples collected between 0 and 4 h (P < 0.05; Fig. 3D). (2) Plasma samples collected from the milk vein showed an increase in PGE2 concentration after intramammary infusions with LPS (collections between 8 and 18 h; P < 0.05; Fig. 4A), TNF (collections between 0 and 24 h; P < 0.05; Fig. 4A), IL-1a (collections between 0 and 24 h; P < 0.05; Fig. 4A), and NONOate (collections at 18 h; P < 0.05; Fig. 4A). Level of PGFM in plasma was increased after intramammary infusions with LPS (collections between 1 and 4 h and at 24 h; P < 0.01; Fig. 4B), TNF (between 0 and 24 h; P < 0.05; Fig. 4B), NONOate (at 2, 8, 24 h; P < 0.05; Fig. 4B), and TNF þ IL1a þ NONOate (at 2, 12, 24 h; P < 0.05; Fig. 4B). Considering LT, concentration of LTB4 was elevated in plasma samples collected at 0 and 8 h after infusions of LPS (P < 0.05; Fig. 4C) and IL-1a at 1 h (P < 0.05; Fig. 4C).

94

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

Fig. 3. Effect of repeated intramammary infusions of LPS (250 mg), TNF (1 mg/mL), IL-1a (1 mg/mL), NONOate (12.8 mg/mL; 10
4 M), or TNF (1 mg/mL) þ IL-1a (1 mg/mL) þ NONOate (12.8 mg/mL; 10
4 M) on concentrations of PG and LT in plasma samples collected from the caudal vena cava. Means (SEM) are shown. Asterisks indicate significant differences (P < 0.05) from the control value. IL, interleukin; LPS, lipopolysaccharide; LT, leukotriene; NONOate, nitric oxide donor; PG, prostaglandin; PGFM, 15-keto-prostaglandin F2a; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

Furthermore, LTC4 concentration was increased after administration of LPS in plasma samples collected between 0 and 2 h and 8 and 12 h (P < 0.05; Fig. 4D), TNF (at 12 h; P < 0.05; Fig. 4D), IL-1a (between 2 and 4 h and at 12 h; P < 0.05; Fig. 4D), NONOate (between 0 and 2 h;

P < 0.05; Fig. 4D), and TNF þ IL-1a þ NONOate (at 4 h; P < 0.05; Fig. 4D). (3) Samples collected from the jugular vein showed an increase in plasma PGE2 concentration after infusions

Fig. 4. Effect of repeated intramammary infusions of LPS (250 mg), TNF(1 mg/mL), IL-1a (1 mg/mL), NONOate (12.8 mg/mL; 10
4 M), or TNF (1 mg/mL) þ IL-1a (1 mg/mL) þ NONOate (12.8 mg/mL; 10
4 M) on concentrations of PG and LT in plasma samples collected from the milk vein. Means (SEM) are shown. Asterisks indicate significant differences (P < 0.05) from the control value. IL, interleukin; LPS, lipopolysaccharide; LT, leukotriene; NONOate, nitric oxide donor; PG, prostaglandin; PGFM, 15-keto-prostaglandin F2a; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

95

Fig. 5. Effect of repeated intramammary infusions of LPS (250 mg), TNF (1 mg/mL), IL-1a (1 mg/mL), NONOate (12.8 mg/mL; 10
4 M), or TNF (1 mg/mL) þ IL-1a (1 mg/mL) þ NONOate (12.8 mg/mL; 10
4 M) on concentrations of PG and LT in plasma samples collected from the jugular vein. Means (SEM) are shown. Asterisks indicate significant differences (P < 0.05) from the control value. IL, interleukin; LPS, lipopolysaccharide; LT, leukotriene; NONOate, nitric oxide donor; PG, prostaglandin; PGFM, 15-keto-prostaglandin F2a; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

with LPS (collections between 2 and 4 h; P < 0.05; Fig. 5A), TNF and IL-1a (between 0 and 24 h; P < 0.001; Fig. 5A), and NONOate (between 1 and 18 h; P < 0.05; Fig. 5A). Regarding PGFM, it was increased after infusion of LPS (samples collected at 2, 8, 12 h; P < 0.05; Fig. 5B), TNF (collections at 12 h; P < 0.05; Fig. 5B), NONOate (at 4 h; P < 0.05; Fig. 5B), and TNF þ IL-1a þ NONOate (collections between 2 and 18 h; P < 0.05; Fig. 5B). With respect to LTB4, its concentration was augmented after administration of LPS (collections between 0 and 1 h and at 12 h; P < 0.05; Fig. 5C) and IL-1a (collections between 0 and 1 h and at 4 h and at 12 and 24 h; P < 0.05; Fig. 5C). No significant changes in LTC4 concentration were observed after infusions with any of the stimulators in plasma samples collected from the jugular vein (P > 0.05; Fig. 5D). (4) Plasma samples collected from the abdominal aorta showed increased PGE2 concentration after intramammary infusions with LPS (samples collected between 2 and 4 h and at 12 h; P < 0.05; Fig. 6A), TNF (collections between 1 and 24 h; P < 0.05; Fig. 6A), IL-1a (collections between 0 and 24 h; P < 0.05; Fig. 6A), and TNF þ IL-1a þ NONOate (collections at 0 h; P < 0.05; Fig. 6A). The level of PGFM in plasma samples increased after infusions with TNF (collections between 18 and 24 h; P < 0.001; Fig. 6B), NONOate (at 0 h; P < 0.01; Fig. 6B), and TNF þ IL-1a þ NONOate (collections between 0 and 1 h and between 4 and 24 h; P < 0.05; Fig. 6B). Considering LTB4, its concentration was increased after LPS administration (collections at 0 h and between 2 and 18 h; P < 0.05; Fig. 6C), TNF

(collections at 2 h; P < 0.001; Fig. 6C), NONOate (collections at 0 h; P < 0.05; Fig. 6C), and TNF þ IL-1a þ NONOate (collections at 0 h; P < 0.05; Fig. 6C). Finally, intramammary infusions with NONOate also elevated LTC4 level in plasma samples (between 1 and 2 h; P < 0.05; Fig. 6D). 4. Discussion The present study describes both in vitro and in vivo actions of LPS and inflammatory mediators on local eicosanoids secretion by bovine mammary gland. To the best of our knowledge, we describe for the first time the in vitro stimulatory effect of LPS, TNF, IL-1a, and NONOate on PGE2, PGF2a, LTB4, and LTC4 secretion by bovine mammary gland tissue. These findings evidence that mediators of E coli-induced mastitis may locally modulate the production of AA metabolites by bovine mammary gland tissue. Moreover, using an in vivo experimental model for intramammary infusions [18] of cytokines, we describe the influence of repeated infusions of LPS, TNF, IL-1a, and NO on eicosanoids plasma concentrations in samples collected from the caudal vena cava, the milk vein, the jugular vein, and the abdominal aorta. Intramammary infection is often associated with breakdown of the blood–milk barrier, resulting in local and systemic effects. Intramammary infusion of LPS has been used to mimic bacterial invasion and subsequent inflammatory response in bovine udder [7,18,19]. Moreover, previous studies demonstrated that TNF, IL-1a, and NO are crucial factors involved in local inflammatory response [20]. In addition, aforementioned factors have been reported to

96

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

Fig. 6. Effect of repeated intramammary infusions of LPS (250 mg), TNF (1 mg/mL), IL-1a (1 mg/mL), NONOate (12.8 mg/mL; 10
4 M), or TNF (250 mg) þ IL-1a (1 mg/mL) þ NONOate (12.8 mg/mL; 10
4 M) on concentrations of PG and LT in plasma samples collected from the abdominal aorta. Means (SEM) are shown. Asterisks indicate significant differences (P < 0.05) from the control value. IL, interleukin; LPS, lipopolysaccharide; LT, leukotriene; NONOate, nitric oxide donor; PG, prostaglandin; PGFM, 15-keto-prostaglandin F2a; SEM, standard error of the mean; TNF, tumor necrosis factor alpha.

be increased in blood serum and milk during natural and experimental coliform mastitis infection [7,18,21]. TNF is one of the first cytokines released during coliform mastitis and is mainly secreted by macrophages. Initial pathogen triggering of innate immune receptors such as Toll-like receptors leads to the activation of TNF production by macrophages. Macrophage activation also leads to nitric oxide (NO) production via inducible nitric oxide synthase, promoting pathogen lysis [22]. In the present study, administration of TNF was considered as a trigger for inflammatory response via macrophage activation and also induction of polymorphonuclear leukocytes migration into udder [6,7]. Furthermore, the role of this cytokine in physiological and pathologic regulations of reproductive functions has been widely characterized in different species [13,18,23,24]. Based on our previous in vivo studies [18], we speculated that after intramammary infusion of TNF, the elevated concentration of this cytokine could locally mediate autocrine and/or paracrine secretion of eicosanoids. However, other contributions for TNF plasma level, such as Kupffer cells in the liver or macrophages in extramammary organs should not be excluded. The effect of IL-1a on PG and LT secretion by epithelial cells from bovine mammary gland was previously described [12]. Both IL-1a and IL-1-beta messenger RNA expressions have been detected in normal bovine milk cells [25] and in epithelial cells from bovine mammary gland infected with E coli [26]. Besides acting as an immune mediator, IL-1a is one of the main cytokines governing local regulation reproductive events, playing roles in maintenance of corpus luteum (CL) and regulating the local PGE2:PGF2a ratio in bovine endometrium during the estrous cycle [27,28].

Previous studies have indicated that NO is also an important mediator of the inflammation present in bovine mastitis and is considered as indicator of disease progression, within the course of inflammatory response in the udder [29]. Indeed, different reports evidenced the interaction between NO and PG, suggesting that NO enhances PG synthesis by direct activation of prostaglandinendoperoxide synthase 2 [30]. On the other hand, NO is also involved in the process of functional and structural luteolysis in cattle [14,23]. Along with cytokines, eicosanoids locally released during mastitis [31,32] are important mediators and modulators of inflammation. It is well known that PG are involved in the regulation of the estrous cycle and pregnancy maintenance in different species [24,33,34]. PGF2a released from the uterus in a pulsatile manner has been shown to cause regression of the ruminant CL [33]. In contrast to PGF2a, PGE2 is known to be a luteotrophic agent [35]. With regard to LT, their involvement in leukocyte recruitment and cytokine production have been described [36,37]. Alternatively, LTs are suggested to mediate PG action in uterus and ovaries [38]. LTB4 plays a luteotrophic role in opposition to LTC4, which has been reported to be a luteolytic agent [39]. Thus, we considered that eicosanoids locally produced in mammary gland during mastitis might also affect function of other organs such as the reproductive tract. We addressed the present question by measuring peripheral (blood samples collected from jugular vein) and local (blood samples collected from milk vein, caudal vena cava, and abdominal aorta) changes of AA concentrations, in response to intramammary infusions of LPS, TNF, IL-1a, NONOate, and the combination of TNF þ IL-1a þ NONOate.

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

4.1. Plasma samples collected from the caudal vena cava Interestingly, PGE2 concentration was increased by all treatments, with the exception of NONOate alone. In contrast, PGFM concentration was increased after intramammary infusions with NONOate and the combination of TNF þ IL-1a þ NONOate. It has been previously described that TNF modulates the ratio between luteolytic PGF2a and luteotrophic PGE2 in uterus, which is determinant for estrous cycle regulation [13,17]. Moreover, other studies suggested that high concentrations of TNF observed during some chronic inflammatory processes [7,18] might result in an unbalance on prostanoids metabolism, inverting the proportion between PGF2a and PGE2, resulting in CL lifespan prolongation. In fact, the relative proportion between both PG syntheses may be more relevant than absolute levels of each individual PG. Moreover, it has been previously reported that direct administration of IL-1a into the uterus during late luteal stage of the estrous cycle significantly stimulates PGE2 output and prolongs bovine CL lifespan [27]. This might suggest that in vivo IL-1a may play a luteoprotective role. Regarding LT, we observed changes in LTC4 concentration in samples from the caudal vena cava in response to NONOate infusions. Similarly, it was shown that intraluteal administration of NO donor in vivo stimulated both LTC4 and PGF2a [40]. Accordingly, intravenous administration of endotoxin to cyclic animals caused the release of PGF2a and NO, which have been reported to differently interfere with reproductive processes [1,8,41]. After all, the present results showed changes in the eicosanoids concentration in blood samples from the caudal vena cava that may modulate the ratio of both PG and LT and has been shown to influence bovine female reproductive tract functions. 4.2. Plasma samples collected from the milk vein The blood samples from the milk vein were collected to assess the local eicosanoids concentrations in blood outflow from the mammary gland. Similar to the results from the caudal vena cava, PGE2 concentration was increased by TNF and IL-1a infusions. Pezeshki et al [10] and Vangroenweghe et al [42] suggested that PGE2 concentration in milk increased after intramammary challenge of E coli, and this PG production was related to bacterial growth and level of inflammatory response. Considering PGFM, all examined factors promoted its secretion, with exception of IL-1a. Concerning the local cellular mechanisms regulating LT production in animals with healthy vs inflamed mammary glands, low concentrations of LTB4 has been associated with healthy mammary quarters; however, LTB4 was drastically increased in acute mastitis [43]. Surprisingly, Pezeshki et al [10] did not find changes in LTB4 concentration in milk from cows with experimentally induced E coli mastitis. However, Boutet et al [43] and Rose et al [44] found that experimental induction of mastitis with bacteria other than E coli increased LTB4 concentration in bovine milk. We found slight changes in LTB4 concentration in blood samples collected from the milk vein after IL-1a and LPS infusions. Generally, the present LTB4 level may be related to an auto–paracrine role of LTB4 and/or to increased metabolism of LT in the mammary gland.

97

Importantly, intramammary infusions with LPS significantly increased LTC4 concentration. It has been reported that the main sources of LTC4 are immune cells such as basophils, eosinophils, macrophages, and vascular endothelial cells or platelets [45–47]. Therefore, one of the effects of LT should be tissue swelling, as a way to increase vascular permeability. Our results support previous reports, suggesting that during experimental mastitis, AA metabolites are locally produced in bovine mammary gland. As a consequence, triggered inflammatory response may be spread to other organs. Therefore, changes in eicosanoids concentrations observed after intramammary infusions may be a cause for pathology in mammary gland and other organs. 4.3. Plasma samples collected from the jugular vein Peripheral changes in eicosanoids concentration were considered regarding their putative actions on the hypothalamus–pituitary–ovary axis. We demonstrated that the intramammary infusions of cytokines and NONOate significantly increased PGE2 concentration, whereas PGFM concentration was significantly elevated by the combination of TNF þ IL-1a þ NONOate in blood samples collected from jugular vein. Considering LT, we observed a significant increase in LTB4 secretion after intramammary infusions with LPS and IL-1a. Interestingly, in blood samples collected locally from the milk vein, infusions with the same reagents affected particularly LTC4 concentration. LTB4 is known to induce infiltration and sequestration of neutrophils in bovine mammary gland [48]. Increased levels of LTB4 after LPS infusion can result from the activation of local leukocytes. Furthermore, it was previously reported that LTB4 is the primary effective chemoattractant during LPS-induced mastitis [5], and high concentrations of LTB4 play an important role in PMN migration into the mammary gland [49]. Thus, the source of LTB4 could also be the recently recruited immune cells such as neutrophils, which represent most of the responding cells during early stages of mastitis. Besides this, other cells such as vascular endothelial cells and epithelial cells of the respiratory system [50] should also be considered, suggesting an influx into the mammary gland of LTB4 from other organs. Taken together, these findings suggest that changes in AA metabolites concentrations in the jugular vein may result not only from inflammatory process in bovine mammary gland but also from generalized inflammation in other organs. 4.4. Plasma samples collected from the abdominal aorta Likewise, by measuring the concentration of AA metabolites in the abdominal aorta, we were able to study the hypothetical impact of inflammatory mediators and eicosanoids locally secreted in mammary gland on reproductive tract. We detected changes in PGE2 concentration after intramammary infusion of TNF and IL-1a, whereas PGFM was significantly increased by the combination of TNF þ IL1a þ NONOate. Furthermore, LTB4 concentration was significantly increased after infusion of LPS. Waller et al [51] reported that LTs mediate the response to LPS because

98

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99

intramammary infusion of LT biosynthesis inhibitor diminished LPS-induced mammary gland recruitment of neutrophils. With regard to LTC4, its concentration increased after infusion of NONOate. Finally, involvement of LTC4 in vascular permeability has been already discussed. Changes in eicosanoids concentrations observed in blood samples from the abdominal aorta suggest their putative interference with functions of reproductive organs. Despite the fact that PG may act at several sites in the reproductive system, a modulatory effect of LT on in vivo PG secretion by the bovine reproductive tract has been described [39]. LTs are found to be auto and/or paracrine factors, and their action in the bovine reproductive tract depends on LT type. LTB4 seems to have a luteotropic role, whereas LTC4 has been considered luteolytic in vivo [39]. In fact, LTs produced in the CL tissue, influence PG function and serve as an important mediator during estrous cycle or early pregnancy. In this line, we consider that LT produced locally in bovine mammary gland during mastitic episodes may influence bovine reproductive tract function. In conclusion, we have shown that different mediators of mastitis, ie, LPS, cytokines, and NO, regulate in vitro and in vivo secretion of eicosanoids in bovine mammary gland. Briefly, LPS, TNF, and IL-1a mainly influenced PGE2 concentration, whereas the combination of TNF þ IL-1a þ NONOate increased PGFM. Besides LPS, mainly IL-1a increased LTB4 concentration, whereas NONOate stimulated LTC4 secretion. Moreover, intramammary infusions of LPS and other inflammatory mediators induced changes in AA metabolites concentrations in plasma samples of blood vessels with different dynamics and intensity. Interestingly, TNF (alone or in combination with other factors) appears to be the main modulator of both luteotropic and luteolytic PG sampled in the milk vein and abdominal aorta. Regarding the LT secretion, LPS significantly increased luteolytic LTC4 in the milk vein (between 0 and 2 h) and luteotropic LTB4 in the abdominal aorta (between 2 and 18 h), respectively. This may indicate that luteolytic factors from the mammary gland enter the systemic circulation, being thereafter bioconverted in different organs like lungs before reaching the reproductive tract. In conclusion, AA metabolites produced locally in bovine mammary gland enter the blood stream and may reach the reproductive tract and potentially modulate its functions. Acknowledgments This study was supported by a research grant of the Polish Ministry of Sciences and Higher Education (grant N308600439). Katarzyna K. Piotrowska-Tomala was supported by the European Union within the European Social Fund (DrINNO2, Olsztyn, Poland). None of the authors has any financial or personal relationships that could inappropriately influence or bias the content of the article. References [1] Huszenicza G, Jánosi S, Gáspárdy A, Kulcsár M. Endocrine aspects in pathogenesis of mastitis in postpartum dairy cows. Anim Reprod Sci 2004;82-3:389–400.

[2] Burvenich C, Van Merris V, Mehrzad J, Diez-Fraile A, Duchateau L. Severity of E. coli mastitis is mainly determined by cow factors. Vet Res 2003;34:521–64. [3] Rainard P, Riollet C. Innate immunity of the bovine mammary gland. Vet Res 2006;37:369–400. [4] De Schepper S, De Ketelaere A, Bannerman DD, Paape MJ, Peelman L, Burvenich C. The Toll-like receptor-4 (TLR-4) pathway and its possible role in the pathogenesis of Escherichia coli mastitis in dairy cattle. Vet Res 2008;39:5. [5] Bannerman DD. Pathogen-dependent induction of cytokines and soluble inflammatory mediators during intramammary infection of dairy cows. J Anim Sci 2009;87:10–25. [6] Shuster DE, Kehril Jr ME, Stevens MG. Cytokine production during endotoxin-induced mastitis in lactating dairy cows. Am J Vet Res 1993;54:80–5. [7] Hoeben D, Burvenich C, Trevisi E, Bertoni G, Hamann J, Bruckmaier RM, Blum JW. Role of endotoxin and TNF-alpha in the pathogenesis of experimentally induced coliform mastitis in periparturient cows. J Dairy Res 2000;67:503–14. [8] Huszenicza G, Jánosi S, Kulcsár M, Kóródi P, Reiczigel J, Kátai L, Peters AR, De Rensis F. Effects of clinical mastitis on ovarian function in post-partum dairy cows. Reprod Domest Anim 2005; 40:199–204. [9] Giri SN, Chen Z, Carroll EJ, Mueller R, Schiedt MJ, Panico L. Role of prostaglandins in pathogenesis of bovine mastitis induced by Escherichia coli endotoxin. Am J Vet Res 1984;45:586–91. [10] Pezeshki A, Stordeur P, Wallemacq H, Schynts F, Stevens M, Boutet P, Peelman LJ, De Spiegeleer B, Duchateau L, Bureau F, Burvenich C. Variation of inflammatory dynamics and mediators in primiparous cows after intramammary challenge with Escherichia coli. Vet Res 2011;42:15. [11] Craven N. Chemotactic factors for bovine neutrophils in relation to mastitis. Comp Immunol Microbiol Infect Dis 1986;9:29–36. [12] Piotrowska-Tomala KK, Siemieniuch MJ, Szóstek AZ, Korzekwa AJ, Woclawek-Potocka I, Galvao AM, Okuda K, Skarzynski DJ. Lipopolysaccharides, cytokines, and nitric oxide affect secretion of prostaglandins and leukotrienes by bovine mammary gland epithelial cells. Domest Anim Endocrinol 2012;43:278–88. [13] Skarzynski DJ, Piotrowska KK, Bah MM, Korzekwa A, WoclawekPotocka I, Sawai K, Okuda K. Effects of exogenous tumour necrosis factor-alpha on the secretory function of the bovine reproductive tract depend on tumour necrosis factor-alpha concentrations. Reprod Domest Anim 2009;44:371–9. [14] Skarzynski DJ, Jaroszewski JJ, Bah MM, Deptula KM, Barszczewska B, Gawronska B, Hansel W. Administration of a nitric oxide synthase inhibitor counteracts prostaglandin F2-induced luteolysis in cattle. Biol Reprod 2003;68:1674–81. [15] Skarzynski DJ, Siemieniuch M, Pilawski W, Woclawek-Potocka I, Bah MM, Majewska M, Jaroszewski JJ. In vitro assessment of progesterone and prostaglandin E2 production by the corpus luteum in cattle following pharmacological synchronization of the estrus. J Reprod Dev 2009;55:170–6. [16] Skarzynski DJ, Miyamoto Y, Okuda K. Production of prostaglandin F2a by cultured bovine endometrial cells in response to tumor necrosis factor-a: cell type specificity and intracellular mechanisms. Biol Reprod 2000;62:1116–20. [17] Skarzynski DJ, Bah MM, Deptula K, Woc1awek-Potocka I, Korzekwa A, Shibaya M, Pilawski W, Okuda K. Roles of tumor necrosis factor-alpha of the estrous cycle in cattle: an in vivo study. Biol Reprod 2003;69:1907–13. [18] Blum JW, Dosogne H, Hoeben D, Vangroenweghe F, Hammon HM, Bruckmaier RM, Burvenich C. Tumor necrosis factor-a and nitrite/ nitrate responses during acute mastitis induced by Escherichia coli infection and endotoxin in dairy cows. Domest Anim Endocrinol 2000;19:223–35. [19] Werner-Misof C, Pfaffl MW, Bruckmaier RM. Dose-dependent immune response in milk cells and mammary tissue after intramammary administration of lipopolysaccharide in dairy cows. Vet Med 2007;52:231–44. [20] Okada H, Ohtsuka H, KonNai S, Kirisawa R, Yokomizo Y, Yoshino T, Rosol TJ. Effects of lipopolysaccharide on production of interleukin1 and interleukin-6 by bovine mammary epithelial cells in vitro. J Vet Med Sci 1999;61:33–5. [21] Nakajima Y, Mikami O, Yoshioka M. Elevated levels of tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) activities in the sera and milk of cows with naturally occurring coliform mastitis. Res Vet Sci 1997;62:297–8. [22] Parameswaran N, Patial S. Tumor necrosis factor-a signaling in macrophages. Crit Rev Eukaryot Gene Expr 2010;20:87–103.

K.K. Piotrowska-Tomala et al. / Domestic Animal Endocrinology 52 (2015) 90–99 [23] Skarzynski DJ, Jaroszewski JJ, Okuda K. Role of tumor necrosis factor–alpha and nitric oxide in luteolysis in cattle. Domest Anim Endocrinol 2005;29:340–6. [24] Galvão A, Valente L, Skarzynski DJ, Szóstek A, PiotrowskaTomala KK, Rebordão MR, Mateus L, Ferreira-Dias G. Effect of cytokines and ovarian steroids on equine endometrium function: an in vitro study. Reprod Fertil Dev 2013;25:985–97. [25] Peli A, Scagliarini A, Britti D, Boari A. Detection of proinflammatory and regulatory cytokines in bovine milk using RT-PCR. Vet Res Commun 2003;27:779–81. [26] McClenahan DJ, Sotos JP, Czuprynski CJ. Cytokine response of bovine mammary gland epithelial cells to Escherichia coli, coliform culture filtrate, or lipopolysaccharide. Am J Vet Res 2005; 66:1590–7. [27] Majewska M, Woclawek-Potocka I, Bah MM, Hapunik J, Piotrowska KK, Tasaki Y, Acosta TJ, Okuda K, Skarzynski DJ. Is interleukin-1alpha a luteotrophic or luteolytic agent in cattle? Reproduction 2010;139:665–72. [28] Nishimura R, Bowolaksono A, Acosta TJ, Murakami S, Piotrowska KK, Skarzynski DJ, Okuda K. Possible role of interleukin-1 in the regulation of bovine corpus luteum throughout the luteal phase. Biol Reprod 2004;71:1688–93. [29] Bouchard L, Blais S, Desrosiers C, Zhoa X, Lacasse P. Nitric oxide production during endotoxin-induced mastitis in cow. J Dairy Sci 1999;82:2574–81. [30] Goodwin DC, Landino LM, Marnett LJ. Effects of nitric oxide and nitric oxide-derived species on prostaglandin endoperoxide synthase and prostaglandin biosynthesis. FASEB J 1999;13: 1121–36. [31] Anderson KL, Kindahl H, Petroni A, Smith AR, Gustafsson BK. Arachidonic acid metabolites in milk of cows during acute coliform mastitis. Am J Vet Res 1985;46:1573–4. [32] Peter AT, Clark PW, Van Roekel DE, Luker CW, Gaines JD, Bosu WT. Temporal changes in metabolites of prostanoids in milk of heifers after intramammary infusion of Escherichia coli organisms. Prostaglandins 1990;39:451–7. [33] McCracken JA, Custer EE, Lamsa JC. Luteolysis: a neuroendocrinemediated event. Physiol Rev 1999;79:263–323. [34] Okuda K, Miyamoto Y, Skarzynski DJ. Regulation of endometrial prostaglandin F(2alpha) synthesis during luteolysis and early pregnancy in cattle. Domest Anim Endocrinol 2002;23:255–64. [35] Milvae RA, Hinckley ST, Carlson JC. Luteotrophic and luteolytic mechanisms in the bovine corpus luteum. Theriogenology 1996;45: 1327–49. [36] Williams KI, Higgs GA. Eicosanoids and inflammation. J Pathol 1988; 156:101–10.

99

[37] Bottoms GD, Adams HR. Involvement of prostaglandins and leukotrienes in the pathogenesis of endotoxemia and sepsis. J Am Vet Med Assoc 1992;200:1842–8. [38] Niswender GD, Juengel JL, Silva PJ, Rollyson MK, McIntush EW. Mechanisms controlling the function and life span of the corpus luteum. Physiol Rev 2000;80:1–29. [39] Korzekwa AJ, Acosta TJ, Miklewicz M, Okuda K, Lee SH, Skarzynski DJ. Leukotrienes affect secretory function of ovarian cells in vitro. Reprod Domest Anim 2010;45:288–96. [40] Bogacki M, Jaroszewski JJ, Skarzynski DJ, Hansel W. The influence of nitric oxide on cyclooxygenase and lipoxygenase products of arachidonic acid metabolism in the bovine corpus luteum in vivo. Biol Reprod 2002;66(Suppl 1):325. [41] Kindhal H, Odensvik K, Bekana M, Kask K. Prostaglandin release as a mediator between infections and impaired reproductive performance. Reprod Domest Anim 1996;31:441–4. [42] Vangroenweghe F, Duchateau L, Boutet P, Lekeux P, Rainard P, Paape MJ, Burvenich C. Effect of carprofen treatment following experimentally induced Escherichia coli mastitis in primiparous cows. J Dairy Sci 2005;88:2361–76. [43] Boutet P, Bureau F, Degand G, Lekeux P. Imbalance between lipoxin A4 and leukotriene B4 in chronic mastitis-affected cows. J Dairy Sci 2003;86:3430–9. [44] Rose DM, Giri SN, Wood SJ, Cullor JS. Role of leukotirene B4 in the pathogenesis of Klebsiellapneumoniae –induced bovine mastitis. Am J Vet Res 1989;50:915–8. [45] Bigby TD, Meslier N. Transcellular lipoxygenase metabolism between monocytes and platelets. J Immunol 1989;143:1948–54. [46] Lewis RA, Austen KF, Soberman RJ. Leukotrienes and other products of the 5-lipoxygenase pathway. Biochemistry and relation to pathobiology in human diseases. N Engl J Med 1990;323:645–55. [47] Lam BK, Penrose JF, Freeman GJ, Austen KF. Expression cloning of a cDNA for human leukotriene C4 synthase, an integral membrane protein conjugating reduced glutathione to leukotriene A4. Proc Natl Acad Sci U S A 1994;91:7663–7. [48] Levy BD, Clish CB, Schmidt B, Gronert K, Serhan CN. Lipid mediator class switching during acute inflammation: signals in resolution. Nat Immunol 2001;2:612–9. [49] Salmon JA, Higgs GA. Prostaglandins and leukotrienes as inflammatory mediators. Br Med Bull 1987;43:285–96. [50] Luo M, Lee S, Brock TG. Leukotriene synthesis by epithelial cells. Histol Histopathol 2003;18:587–95. [51] Waller KP. Modulation of endotoxin-induced inflammation in the bovine teat using antagonists/inhibitors to leukotrienes, platelet activating factor and interleukin 1 beta. Vet Immunol Immunopathol 1997;57:239–51.