All horses utilized in this study were housed at the North Carolina State University (NCSU) Teaching Animal Unit. A total of 12 horses of mixed breed and gender were used in this study and included quarter horses, thoroughbreds, Tennessee walking horses, and breed crosses. Ages of horses were varied and ranged from 5 to 15 years. All horses were kept on pasture under similar conditions and received no medications for the duration of this study. All procedures were approved by the NCSU Institutional Animal Care and Use Committee.

Leukocyte-rich plasma (LRP) was obtained as previously described (35). Briefly, 30–60 cm³ of heparinized equine whole blood was collected via jugular venipuncture and placed into 15 mL conical polypropylene tubes in 10 mL aliquots. Tubes stood for 1 h at room temperature to allow erythrocytes to settle out of solution, and the LRP was collected from above the red blood cell layer.

For pretreatment of leukocytes, various concentrations of misoprostol (Cayman Chemical, Ann Arbor, MI, USA) or vehicle control (0.05% DMSO) were applied to LRP 30 min before stimulation with 100 ng/mL LPS from E. coli 055:B5 (Sigma Aldrich, St. Louis, MO, USA). For posttreatment of leukocytes, the same concentrations of misoprostol or vehicle control were applied to LRP either 30 or 60 min following LPS stimulation.

Leukocyte-rich plasma was stimulated at 37°C with 100 ng/mL LPS, or vehicle control for LPS (sterile PBS), for the times indicated. Cell viability was assessed via trypan blue exclusion before and after misoprostol and LPS treatment and was routinely >95%. For cytokine protein analysis (multiplex bead immunoassay), equine LRP was diluted 1:1 in phenol red free RPMI media supplemented with 100 U/mL penicillin and 100 µg/mL streptomycin (Sigma Aldrich) before stimulation. This antibiotic media dilution was performed to prevent bacterial contamination of cultures over the 18-h incubation period. For shorter incubation periods that were required for leukocyte mRNA analysis, LRP was not diluted in antibiotic media before stimulation.

Leukocyte-rich plasma from five separate donor horses (n = 5) was utilized for cytokine protein analysis. LRP from each horse was treated with misoprostol, and either before or after stimulation with LPS (as described earlier), for 6 or 24 h. Cell supernatant was collected and stored at −80°C (for no longer than 3 months) until analyzed.

All samples from each of the five horses were assayed in triplicate. Analysis of cytokines and chemokines secreted by equine leukocytes was carried out using an equine-specific Milliplex^® Map Magnetic Bead Panel (EMD Millipore, Billerica, MA, USA) per the manufacturer’s protocol. To gain a clearer understanding of the endogenous cytokines influenced by misoprostol, 23 different cytokines and chemokines were initially assayed. These cytokines were chosen as they comprised a mixture of mediators of innate and adaptive immunity, and therefore more clearly defined the role of misoprostol in different arms of the immune response. Cytokines/chemokines included in the analysis were as follows: interleukin-1α (IL-1α), IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12 (p70), IL-13, IL-17A, IL-18, TNFα, interferon γ (IFNγ), IFNγ-induced protein 10 (IP-10 or CXCL10), granulocyte-stimulating factor, granulocyte monocyte-stimulating factor (GM-CSF), growth-regulated protein (GRO or CXCL1), monocyte chemotactic protein-1 (or CCL2), fibroblast growth factor-2 (FGF-2), eotaxin (or CCL11), fractalkine (or CX3CL1), and RANTES (or CCL5).

Briefly, supplied 96-well assay plates were washed using kit wash buffer before use. Background (assay buffer), standard, and control wells were loaded onto the plate and diluted 1:1 with serum matrix. The serum matrix for this kit was diluted 1:1 with RPMI media before use to ensure consistency with the experimental samples. Next, 25 μL of each sample and assay buffer were plated in sample wells, and each sample was plated in triplicate. Plates were covered and incubated on an orbital shaker overnight at 4°C in the presence of magnetic beads coated with fluorescently labeled capture antibodies for each analyte. Beads were then washed and incubated with biotinylated detection antibodies, followed by addition of streptavidin phycoerythrin. Beads were washed again before resuspension in drive fluid and sample analysis using a Luminex MagPix^® instrument and Luminex xPONENT^® software (Luminex, Austin, TX, USA).

Data analysis was conducted using Milliplex Analyst^® software (EMD Millipore). A minimum count of 50 beads per well was used for inclusion in analysis. Outliers were removed from any triplicate with a coefficient of variation (CV) > 15% and the mean of the remaining duplicate samples was used for analysis. Values that fell below the lower limit of detection (LLOD) of the assay were assigned the lowest detectable concentration for that analyte, as determined by the Analyst^® software. Cytokines/chemokines in which >50% of sample values fell below the LLOD were not analyzed. Mean fluorescence intensity data using a five-parameter logistic standard curve was used to calculate analyte concentration. One horse was removed from analysis due to high CV across all samples and presence of multiple outlying analyte values at baseline and upon LPS stimulation at 6 and 24 h (via extreme studentized derivative method, α = 0.05).

All RNA isolation and DNAse materials were obtained from Qiagen (Valencia, CA, USA). Following stimulation, LRP was briefly centrifuged, and supernatant was removed to isolate equine leukocytes. RNA was then isolated from the leukocytes using an RNeasy Mini Kit with QIAshredder column homogenization per manufacturer protocol. An RNase-free DNAse set was used to perform two DNase steps, one on-column and one following elution of RNA. Following the second DNase digestion, RNA cleanup was performed using the RNeasy Mini Kit. RNA was then quantified using a Nanodrop Spectrometer. Equal quantities (up to 10 ng) of RNA from each sample were used for cDNA synthesis using the Superscript III First-Strand Synthesis System (Invitrogen, Thermo Fisher Scientific, Grand Island, NY, USA) per the manufacturer’s protocol with random hexamers (50 ng/μL).

Of the proteins assayed via multiplex bead immunoassay, production of TNFα, IL-6, and IL-1β proteins was most significantly altered by misoprostol treatment. As these cytokines are also significant pro-inflammatory mediators in vivo, they were chosen for further interrogation of the effect of misoprostol pretreatment on pro-inflammatory cytokine mRNA expression in equine leukocytes. In addition, because cAMP-elevating agents can elicit differing effects on cytokines at the mRNA and protein levels, we chose to evaluate the effect of misoprostol on mRNA production of a cytokine that was not altered at the protein level, based on our multiplex bead immunoassay. Thus, IL-8 was chosen for mRNA analysis, as it is a critical pro-inflammatory mediator of the innate immune response that appeared unaffected by misoprostol treatment at the protein level.

Real-time PCR was performed using a BioRad MyIQ Single Color Real-Time PCR Detection System (BioRad, Hercules, CA, USA) using Taqman primers and probes (Invitrogen, Thermo Fisher Scientific). Primers and probes were obtained from Invitrogen’s proprietary equine-specific gene expression assay database and have been validated by the company. Invitrogen identifies the NCBI target sequence used to design the primers and probes, the 25 base pair region of probe binding, and predicted amplicon size (36, 37). In preliminary experiments, all PCR products were run on a 2% agarose gel and visualized using EZ Vision Three-DNA Dye (Amresco, Solon, OH, USA) to verify specificity of the product. ID numbers of Taqman gene expression assays used were as follows: TNFα, Ec03467871; IL-6, Ec03468678; IL-1β, Ec04260296; IL-8, Ec-3468860; and β2M, Ec03468699. PCR samples were run in triplicate. Each well contained equal quantities of cDNA (between 1 and 10 ng), 1× Taqman primers and probes, 1× Taqman gene expression master mix, and RNase-free water up to 25 μL. PCR cycle conditions were carried out per the manufacturer’s protocol as follows: 50°C for 2 min, once; 95°C for 10 min, once; 95°C for 15 s, followed by 60°C for 1 min (with real-time data collection enabled), 40 times.

All samples were assayed in triplicate. Outliers with an SD greater than 0.167 were removed from any triplicate, and the mean of the remaining two wells was used for analysis. Fold change in mRNA level was determined using the ΔΔCt data analysis method. β2M was chosen as the housekeeping gene in preliminary equine leukocyte experiments using established procedures. Briefly, β2M was selected by evaluating expression stability of three common housekeeping genes (β2M, GAPDH, and β-actin) using protocols established by Radonić et al. ΔΔCt values for all three genes were calculated under different stimulatory conditions as previously described (38), with values closest to 0 indicating the least amount of change in gene expression. Of the genes evaluated, β2M was the most stably expressed in equine leukocytes in our system and thus was used as the housekeeping gene in all real-time PCR experiments.

Data were analyzed using SigmaPlot Version 12.0 (Systat Software, San Jose, CA, USA). Numbers of donor horses (n) utilized for each experiment were based on adequate power of statistical comparison (α = 0.05). Each experimental sample was assayed in triplicate for both protein (multiplex bead immunoassay) and mRNA (RT-PCR) analysis. Data were assessed for normality via the Shapiro–Wilk test (p < 0.05), and non-normal data were log transformed to achieve normality for analysis (noted in figure legends). Data that required log transformation included TNFα mRNA fold change at the 2-h time point (Figure 1A), TNFα pg/ml at 6 h (Table 1), and fractalkine and IFNγ pg/ml at 24 h (Table 1). Data were back transformed for presentation. Data were analyzed via one-way RM ANOVA with Holm–Sidak multiple comparisons post hoc testing, or paired t-test where noted. Log transformation did not achieve normality or equality of variance for IL-1β protein values at 6 or 24-h time points (Table 1); therefore, raw median IL-1β values were analyzed using Friedman’s RM ANOVA on Ranks with Student–Newman–Keuls multiple comparisons post hoc testing. For comparison, all data are presented as mean ± SEM.

Reports have demonstrated that cAMP-elevating agents elicit differing and even diverging effects on cytokines during early and late immune responses (39, 40). Thus, we evaluated the effect of misoprostol treatment on cytokine secretion in LPS-stimulated equine leukocytes at both an early (6 h) and late (24 h) time point. For this study, a multiplex bead immunoassay was used to simultaneously evaluate the effect of LPS stimulation and misoprostol treatment on 23 different cytokines. As shown in Table 1, 100 ng/mL LPS significantly increased secretion of almost all cytokines and chemokines evaluated at 6 h, including TNFα, IL-6, IL-1β, IL-8, IP-10, IL-10, IL-4, IFN, IL-2, IL-18, IL-5, fractalkine, G-CSF, GRO, MCP-1, and eotaxin. Of these analytes, only TNFα, IL-6, IL-1β, IL-8, and IL-10 remained significantly increased at 24 h. Seven of the analytes (FGF-2, IL-13, IL-17A, IL-1α, GM-CSF, RANTES, and IL-12p70) were below the limit of detection for this assay.

The results of misoprostol treatment on LPS-induced cytokine protein levels are shown in Table 1. Misoprostol treatment both before and after LPS stimulation resulted in a statistically significant decrease in TNFα and IL-6 protein following 6 and 24 h of LPS stimulation. IL-8 protein levels were unaffected by misoprostol treatment. Interestingly, misoprostol pretreatment led to a significant increase in IL-1β protein at 6 h of LPS stimulation, while misoprostol posttreatment had no effect on IL-1β at the 6-h time point. Furthermore, misoprostol posttreatment led to a statistically significant decrease in IL-1β protein at 24 h of LPS stimulation, while misoprostol pretreatment had no effect on IL-1 β at the 24-h time point.

In addition to the above cytokines, misoprostol posttreatment significantly inhibited IP-10 protein production after 6 h of LPS stimulation. We did observe a greater than twofold decrease in IFNγ protein production by misoprostol pre- and posttreated leukocytes at 24 h (Table 1), but this was not statistically significant. No other cytokines or chemokines were significantly affected by misoprostol pre- or posttreatment.

Taken together, these results indicate that misoprostol pre- and posttreatment of LPS-stimulated equine leukocytes inhibits secretion of pro-inflammatory TNFα and IL-6 proteins and exerts differential effects on IL-1β cytokine production. Interestingly, our data show that the anti-inflammatory cytokine IL-10, which has previously been documented to be increased by misoprostol treatment in human leukocytes (27), was unaffected.

To further interrogate the effect of misoprostol on cytokine production by LPS-stimulated equine leukocytes, we next sought to determine whether misoprostol treatment altered levels of cytokine mRNA in LPS-stimulated equine leukocytes. To address this objective, we first determined the kinetics of LPS-induced cytokine mRNA expression in our equine LRP model. TNFα, IL-6, and IL-1β mRNA production peaked after 2 h of stimulation with 100 ng/mL LPS and then quickly declined by 6 h of treatment (Figures 1A–C). In contrast, IL-8 displayed a biphasic pattern of mRNA production, peaking first at 2 h before declining and peaking again at 18 h (Figure 1D). This pattern has also been reported in human whole blood following LPS stimulation (41). Because all cytokines displayed a peak in mRNA following 2 h of LPS stimulation, this time point was chosen for initial analysis of the effects of misoprostol pretreatment on LPS-induced cytokine mRNA levels in equine leukocytes.

Elevation of intracellular cAMP elicits different effects on cytokines at the mRNA and protein level, indicating a potential role for cAMP in transcriptional and posttranscriptional cytokine regulation (39, 40). In this study, misoprostol pretreatment led to a concentration-dependent decrease in TNFα, IL-6, and IL-1β, but not IL-8 mRNA, in equine leukocytes following 2 h of LPS stimulation (Figure 2). TNFα mRNA was most potently inhibited by misoprostol at concentrations ≥1 µM (maximum inhibition to 10.3% of control; Figure 2A), which is in agreement with previous reports in human peripheral blood mononuclear cells (27). Comparatively, IL-6 and IL-1β mRNA levels were significantly decreased at ≥10 μM misoprostol (maximum inhibition to 33.1 and 64.5% of control, respectively; Figures 2B,C). In contrast, IL-8 mRNA levels were unaffected by misoprostol pretreatment (Figure 2D). Considering the opposing effects of other cAMP-elevating agents on IL-6, IL-1β, and IL-8 mRNA in other equine leukocyte models (14, 15, 18), our data suggest that misoprostol exerts a uniquely inhibitory effect on IL-6 and IL-1β mRNA production in LPS-stimulated equine leukocytes.

In many clinical situations, veterinarians institute anti-inflammatory therapy following, not preceding, an inflammatory insult. Therefore, we were interested in determining the effect of misoprostol on cytokine mRNA levels in leukocytes that had already been stimulated with LPS. To determine this, we evaluated the effect of 30- and 60-min misoprostol posttreatment on TNFα, IL-6, and IL-1β mRNA levels at both early (2 h) and late (6 h) time points.

At 2 h LPS stimulation, misoprostol 30 and 60-min posttreatment resulted in statistically significant decreases in TNFα (29.5 and 33.7% of control, respectively) and IL-6 mRNA (73.3 and 57.8% of control, respectively) (Figure 3A). In addition, misoprostol 60-min posttreatment significantly diminished IL-1β mRNA production (76.9% of control) (Figure 3A). In contrast, at 6 h LPS stimulation, misoprostol 30 and 60-min posttreatment only decreased IL-6 mRNA to 41.9 and 48.2% of control, respectively, and had no effect on mRNA levels of TNFα and IL-1β (Figure 3B). From these data, we conclude that misoprostol attenuates the initial equine leukocyte pro-inflammatory response to LPS, even after intracellular signaling has already been induced. However, these effects are varied after 6 h of LPS stimulation.