Dietary monoglyceride supplementation to support intestinal integrity and host defenses in health-challenged weanling pigs

Abstract

Frequent incidence of postweaning enterotoxigenic Escherichia coli (ETEC) diarrhea in the swine industry contributes to high mortality rates and associated economic losses. In this study, a combination of butyric, caprylic, and capric fatty acid monoglycerides was investigated to promote intestinal integrity and host defenses in weanling pigs infected with ETEC. A total of 160 pigs were allotted to treatment groups based on weight and sex. Throughout the 17-d study, three treatment groups were maintained: sham-inoculated pigs fed a control diet (uninfected control [UC], n = 40), ETEC-inoculated pigs fed the same control diet (infected control [IC], n = 60), and ETEC-inoculated pigs fed the control diet supplemented with monoglycerides included at 0.3% of the diet (infected supplemented [MG], n = 60). After a 7-d acclimation period, pigs were orally inoculated on each of three consecutive days with either 3 mL of a sham-control (saline) or live ETEC culture (3 × 10⁹ colony-forming units/mL). The first day of inoculations was designated as 0 d postinoculation (DPI), and all study outcomes reference this time point. Fecal, tissue, and blood samples were collected from 48 individual pigs (UC, n = 12; IC, n = 18; MG, n = 18) on 5 and 10 DPI for analysis of dry matter (DM), bacterial enumeration, inflammatory markers, and intestinal permeability. ETEC-inoculated pigs in both the IC and MG groups exhibited clear signs of infection including lower (P < 0.05) gain:feed and fecal DM, indicative of excess water in the feces, and elevated (P < 0.05) rectal temperatures, total bacteria, total E. coli, and total F18 ETEC during the peak-infection period (5 DPI). Reduced (P < 0.05) expression of the occludin, tumor necrosis factor α, and vascular endothelial growth factor A genes was observed in both ETEC-inoculated groups at the 5 DPI time point. There were no meaningful differences between treatments for any of the outcomes measured at 10 DPI. Overall, all significant changes were the result of the ETEC infection, not monoglyceride supplementation.

Infection with enterotoxigenic Escherichia coli resulted in clear detriments to weanling pig health and performance, and these adverse effects were not mitigated by supplementation with a combination of butyric, caprylic, and capric fatty acid monoglycerides.

Introduction

The Gram-negative bacterium enterotoxigenic Escherichia coli (ETEC) is a common cause of postweaning diarrhea (PWD) in young pigs. ETEC can be further classified based on small projections known as fimbriae which allow it to bind to enterocytes. In the case of PWD, the most common strains of ETEC are either F4- or F18-fimbriated. These adhesions partially determine host specificity as F4 typically infects neonatal pigs while F18 almost exclusively impacts weanling pigs (Fairbrother et al., 2005). Frequent incidence of this disease and its associated rates of morbidity and mortality can result in significant economic losses for producers, making it a common target for interventions that may mitigate its adverse effects. Previously, low levels of in-feed antimicrobials helped minimize ETEC infection and subsequent postweaning diarrhea; however, due to concerns about emerging antibiotic resistance, the use of such drugs in this manner is now closely regulated (US FDA, 2017). Consequently, different nutritional interventions to improve both growth performance and immune function have become an increasingly important area of research within the animal industry.

Monoglycerides are one potential intervention that may promote host defenses and overall performance through various mechanisms of action. Monoglycerides consist of a fatty acid esterified to a glycerol molecule, and this physical property elicits some of their health benefits including the ability to disrupt the phospholipid membrane of bacteria (Jackman et al., 2020). In addition, many fatty acids have an unpleasant smell that is reduced in the monoglyceride form, improving the palatability of diets in which these additives are included (Jackman et al., 2020). Fatty acids can be classified into short-, medium-, and long-chain fatty acids based on the length of their hydrocarbon chain. One example of a short-chain fatty acid is butyrate, which provides energy and promotes proliferation, differentiation, and maturation processes in intestinal epithelial cells (Guilloteau et al., 2010). Butyrate is also involved in maintaining the function of the epithelial barrier and has demonstrated anti-inflammatory effects in the intestinal mucosa, both of which are crucial during an infection such as PWD (Wang et al., 2012; Yan and Ajuwon, 2017; Feng et al., 2018). Medium-chain fatty acids, such as caprylic and capric acid, have demonstrated similar abilities to enhance barrier function and promote growth (Hanczakowska et al., 2011; Wang et al., 2018; Gebhardt et al., 2020). Additionally, medium-chain fatty acids have the ability to disrupt pathogen membranes, particularly when delivered in the monoglyceride form (Jackman et al., 2020). That being said, the results of these studies are inconsistent, which necessitates further investigation of the potential benefits of fatty acid glycerides, especially during a disease challenge. As such, the objective of our study was to assess how supplementation of a monoglyceride combination of butyric, caprylic, and capric fatty acids would influence growth performance, inflammation, and intestinal integrity of weanling pigs challenged with F18 ETEC. We hypothesized that both the infection status and monoglyceride supplementation would affect the immune response outcomes and overall performance of these pigs.

Materials and Methods

All animal care and experimental procedures described herein were approved by the University of Illinois Institutional Animal Care and Use Committee.

ETEC screening procedures

Before the start date, procedures for F18 ETEC screening were implemented at both the herd and litter levels to ensure that pigs enrolled in the study did not pre-carry ETEC before inoculation. Approximately 4 wk before the study began (i.e., day postinoculation [DPI] −28), fecal swabs were collected using sterile polyester-tipped applicators from nursery pigs on six different commercial farms located in Illinois and Missouri. Within each farm, two pigs from each of 15 litters were sampled. Validated quantitative polymerase chain reaction (qPCR) procedures to detect F18 ETEC were run on pooled samples from three litters. Briefly, 100 to 120 mg of fecal sample was homogenized in 750 µL of sterile 1× phosphate-buffered saline (PBS). A total of 100 µL of this homogenate was used for nucleic acid extraction using a Biosprint 96 Work Station (Qiagen, Hilden, Germany). Real-time PCR was performed on the ABI 7500 FASTSystem (Thermo Fisher Scientific, Waltham, MA). The reaction mixture contained 1× PerfeCTa FastMix II Low ROX (Quantabio, Beverly, MA 01915), 0.5 µM each of the forward and reverse primers, 0.25 µM of probe, and 5 µL of DNA in a total reaction volume of 20 µL. The challenge E. coli strain used for inoculating the animals served as a positive control. The following thermal cycling profile was used for all amplifications: an initial denaturation at 95 °C for 30 s, followed by 45 repeating cycles of 95 °C for 10 s and 60 °C for 30 s. A cycle threshold (Ct) < 34 was considered positive for F18 and thus unacceptable. The farm from which all pigs were ultimately sourced was chosen based on proximity, herd size, herd health, and having acceptably low/undetectable levels of F18 ETEC. The selected farm was located in Illinois and was negative for porcine respiratory and reproductive syndrome, porcine epidemic diarrhea virus (PEDV), porcine deltacoronavirus (PDCoV), and transmissible gastroenteritis virus (TGEV). Additionally, no modified live oral E. coli vaccines were administered to sows or weaned pigs on this farm.

A second screening was performed on this farm 6 d before each group of pigs was selected (DPI −19) to ensure that F18 ETEC levels remained low or undetectable in litters from which the specific pigs used in each trial would be sourced. At this time, two pigs from each of 24 litters were swabbed and qPCR was run on pooled samples from 2 litters. In this screening, F18 ETEC was undetectable in all pools, so litters were chosen based on overall health and if there were at least five barrows and five gilts available within the litter.

The third screening occurred 3 d after arrival to the animal facility (DPI −4) to confirm no major ETEC break within selected pigs due to postweaning and travel stressors. The qPCR testing of pooled fecal swabs from five pigs in each of 16 pens confirmed that all pigs were negative for F18 ETEC before experimental inoculation.

Animal husbandry and allotment

At weaning (approximately 21 d of age), 80 barrows and 80 gilts were sourced from 16 dams ranging from parity 2 to 8 within a commercial herd (PIC 1050 × PIC 337 genetics; Grand View Farms, LLC, Carthage, IL). These pigs were then transported to Carthage Innovative Swine Solutions Veterinary Research Facility (CISS-VRF; Champaign, IL), which is a biosafety level-2 containment facility with four isolated production rooms and five pens (3.35 m × 1.83 m) per room. Four pens in each room were used to group-house pigs at a stocking density of 10 pigs per pen (five barrows and five gilts) for the study duration.

Upon arrival, pigs were individually weighed (initial body weight [BW] 6.45 ± 1.08 kg) and allotted to one of three treatment groups based on weight and sex. While litter of origin was not a factor for allocation, no pen was composed of pigs from which the majority were from the same litter. Both the infected control (IC) and monoglyceride-supplemented (MG) treatment groups had a total of 60 pigs, which were randomly assigned to one of three rooms where they all would ultimately be inoculated with ETEC. A fourth, separate room was used to house a total of 40 sham-inoculated pigs that represented the uninfected control (UC) treatment, per biosecurity protocols.

Experimental treatments

There were three experimental treatments used in this study with two experimental diets and two states of infection. A common basal diet formulated to meet all the nutrient requirements for nursery pigs was produced at the University of Illinois Feed Technology Center (Table 1; NRC, 2012). This basal diet served as the control diet for both the UC and IC treatment groups. The third MG treatment group was ETEC-inoculated and fed the basal diet supplemented with monoglycerides of butyric, caprylic, and capric fatty acids (BalanGut LS P, BASF, Florham Park, NJ). This test article was a powder containing 43% to 49% active ingredients added directly on top of the basal diet at 0.3% (3 kg/MT; i.e., no space reserved for the test article in the final formulation). This inclusion rate was based on the recommended minimum dosage for suckling and weanling pigs for this commercial product. After a week-long acclimation period, during which pigs received their assigned dietary treatments, pigs were orally inoculated on each of three consecutive days with either 3 mL of a live F18 ETEC culture (3 × 10⁹ colony-forming units [CFU]/mL, isolate number: UI-VDL #05-27242, Carol Maddox, College of Veterinary Medicine, University of Illinois, Urbana, IL) or 3 mL of PBS (sham-inoculation). Actual F18 ETEC concentrations used in this study were 2.73 × 10⁹, 2.90 × 10⁹, and 3.17 × 10⁹ CFU/mL for each consecutive inoculation day (0, 1, and 2 DPI, respectively). This dosage was determined based on prior research that used this same ETEC strain as a health challenge for weanling pigs (Song et al., 2012; Wojnicki et al., 2019; Smith et al., 2020). As confirmed by traditional PCR, this strain of F18 ETEC produced heat-labile toxin, heat-stable toxin b, enteroaggregative E. coli heat-stable enterotoxin, and Shiga-like toxin 2. The first inoculation day was designated as 0 DPI, and all study outcomes reference this time point.

Table 1.

Ingredient and calculated composition of basal diet¹

Item
Ingredient, g/kg
Corn	495.4
SBM	200.0
Dried whey	172.0
Fish meal	100.0
Choice white grease	15.0
Ground limestone	5.00
Sodium chloride	5.00
Vitamin-mineral premix2	3.00
Choline chloride, 60%	0.70
l-Lysine HCl	2.50
dl-Methionine	0.50
l-Threonine	0.50
l-Tryptophan	0.40
Calculated composition
ME, kcal/kg3	3,423
CP, g/kg	224.1
SID amino acids, g/kg
Lysine	15.2
Methionine + cysteine	7.90
Tryptophan	2.70
Threonine	8.50
Valine	9.60

¹All pigs received allotted treatment diet starting at −7 DPI. The monoglyceride supplement was added directly on top of the basal diet at 0.3% of the final formulation.

²Vitamin-mineral premix (JBS United, Sheridan, IN) included the following per kilogram complete diet: vitamin A (retinyl acetate), 11,128 IU; vitamin D3 (cholecalciferol), 2,204 IU; vitamin E (dl-α tocopheryl acetate), 66 IU; vitamin K (menadione nicotinamide bisulfite), 1.42 mg; thiamin (thiamin mononitrate), 0.24 mg; riboflavin, 6.58 mg; pyridoxine (pyridoxine hydrochloride), 0.24 mg; vitamin B12, 0.03 mg; d-pantothenic acid (d-calcium pantothenate), 23.5 mg; niacin (nicotinamide and nicotinic acid), 44 mg; folic acid, 1.58 mg; biotin, 0.44 mg; Cu (copper sulfate), 10 mg; Fe (iron sulfate), 125 mg; I (potassium iodate), 1.26 mg; Mn (manganese sulfate), 60 mg; Se (sodium selenite), 0.3 mg; and Zn (zinc oxide), 100 mg.

³Metabolizable energy and standardized ileal digestible (SID) amino acid values were calculated using NRC (2012). Analyzed crude protein determined as CP = (N × 6.25).

Abbreviations: CP, crude protein; ME, metabolizable energy; SBM, soybean meal; SID, standardized ileal digestibility.

Growth performance

Both individual pig weights and feeder weights were recorded on −7, 0, 5, and 10 DPI and used to calculate average daily gain (ADG), average daily feed intake (ADFI), and feed efficiency (G:F). Pig morbidity and mortality did occur, which necessitated the removal of pigs throughout the study period, so ADFI and G:F were corrected for said removals. Additionally, a feed delivery error took place on DPI 8. Due to this, ADFI and G:F values for the MG group during 5 to 10 and −7 to 10 DPI were estimated as appropriate. Pen served as the experimental unit for all growth performance outcomes, while individual pig was the experimental unit for all other outcomes.

Rectal temperatures and fecal analyses

Rectal temperatures were measured on both 5 and 10 DPI. Fecal samples were individually collected from six random pigs per pen on 0 DPI. Of those six pigs, a random subset of three pigs was assigned to an additional blood collection on 4 DPI and fecal and tissue collections on 5 DPI. The remaining three pigs were used for blood collection on 9 DPI and fecal and tissue collections on 10 DPI (48 pigs total on each sampling day). As such, pigs used for either 5 or 10 DPI collections had a baseline value from 0 DPI for both dry matter (DM) and bacterial counts. On the respective collection days, a representative mass of feces was collected and dried in a 105 °C oven for a minimum of 48 h to determine DM content and serve as an objective measure of fecal consistency over the course of the ETEC infection. If a minimum of 0.02 g feces could not be collected on the designated days, these values were excluded from the data set as it proved difficult to accurately determine DM content on such a small sample.

A separate fecal sample from these collection days was snap-frozen in liquid nitrogen and submitted to the University of Illinois Veterinary Diagnostic Laboratory to quantify total bacteria (rrs), total E. coli (gadAB), and total F18 ETEC (fedA) via established qPCR procedures described earlier. For each gene, all real-time primers and probes were designed in-house using PrimerQuest Tool (Integrated DNA Technologies, Coralville, IA) as described in Table 2.

Table 2.

Analytical details regarding quantification of bacterial gene expression¹

Target	Gene	Primer/probe sequences	Amplicon. bp
Total bacteria	rrs	5ʹ-TTGGTAGCGTATCAGCAACTAC-3ʹ	99
		5ʹ-CCAATTCCGTTTGGTGTAACAG-3ʹ
		/5Cy3/TGGCACTGTAGGAGACACCATTCA/3IAbRQSp/
Total E. coli	gadAB	5ʹ-CAGGCAAACCAACGGATAAAC-3ʹ	131
		5ʹ-TCGGGTCCATAAACAACTGAC-3ʹ
		/5Cy5/TAAATTCGC/TAO/CCGCTACTGGGAT/3IAbRQSp/
F18 ETEC	fedA	5ʹ-ACGGCCGCAAGGTTAAA-3ʹ	88
		5ʹ-GTGGATGTCAAGACCAGGTAAG-3ʹ
		/5SUN/ACAAGCGGT/ZEN/GGAGCATGTGGTA/3IABkFQ/

¹Real time primer/probe sets are listed in the following order: forward primer, reverse primer, and probe.

Abbreviations: bp, base pairs; ETEC, enterotoxigenic E. coli; fedA, F18 fimbrial subunit protein; gadAB, glutamate decarboxylase A and B genes; rrs, ribosomal RNA gene.

Blood sampling to assess intestinal barrier integrity

On 4 and 9 DPI, three pigs per pen that were preselected for fecal and tissue collections were also subjected to the dual sugar absorption test to assess intestinal integrity. A lactulose–mannitol (L:M) solution was prepared by dissolving powder forms (≥98% purity) of lactulose (CAS 4618-18-2; MP Biomedicals, Santa Ana, CA) and mannitol (M4125; Sigma-Aldrich, St. Louis, MO) in ultrapure water (Synergy R System, MilliporeSigma, Burlington, MA) so the final solution contained 30% lactulose and 3% mannitol on a weight/volume basis. The L:M solution was stored at 4 °C and allowed to warm to room temperature before oral dosing. Dose volume was calculated based on BW with the aim that each pig receives 500 mg of lactulose/kg BW and 50 mg of mannitol/kg BW. Both the concentration and the dosed volume of the L:M solution were chosen based on previous literature and pilot studies conducted within and outside our lab (Zhang and Guo, 2009). Each of the three selected pigs was orally administered the L:M solution using plastic syringes with volumes ranging from 1 to 10 mL. As needed, a combination of syringe sizes was used to draw up each pig’s designated dose volume to the nearest tenth of a milliliter. Approximately 4 cm of flexible tubing was attached to the threaded syringe tip to allow for better delivery of the solution, and animals were visually monitored to confirm the solution was properly ingested. One hour after L:M oral dosing, blood was collected from individual pigs via the jugular vein into a 6-mL evacuated tube containing sodium heparin using a 21 G needle. Samples were then centrifuged at 1,160 × g for 15 min at 4 °C and plasma was aliquoted into microcentrifuge tubes and stored at −20 °C before analysis.

Before analysis by high-performance liquid chromatography (HPLC), samples were first thawed at 4 °C for 24 h before being brought to room temperature. Plasma was subsequently combined with 10% trichloroacetic acid in a 9:1 ratio, and samples were vortexed and incubated for 30 min at room temperature to allow for protein precipitation. Samples were once again frozen at −20 °C, thawed at 4 °C, and brought to room temperature before centrifugation at 16,100 × g for 15 min to ensure efficient removal of intact proteins. To quantify the concentration of lactulose and mannitol in the plasma, a procedure similar to that described by Bao et al. (1996) was followed. The HPLC system (model ICS 5000; Dionex, Sunnyvale, CA) included a gradient pump module, an eluent degassing module, and a pulsed amperometric detector with a gold working electrode. Samples were automatically injected using a 25-µL sample loop into an anion-exchange column (250 × 4.0 mm inside diameter, particle size 8.5 µm, pellicular resin; model CarboPac MA-1, Waltham, MA) with an associated guard column. The lower detection and quantification limits for both lactulose and mannitol were <0.1 and 0.2 µg/mL. The upper end of the linear range was 80 µg/mL for lactulose and 40 µg/mL for mannitol. Lactulose, mannitol, and glucose were used as the standards for the HPLC analysis with concentrations of 10 to 80, 5 to 40, and 1,000 µg/mL, respectively. After analysis, a valley-to-valley skimming process was applied to the chromatograms to accurately quantify lactulose, which had peaks that were partially masked by the preceding glucose peak.

Intestinal samples to assess antibody expression and neutrophil infiltration

Before tissue collection, three of the preselected pigs that previously underwent blood and fecal collections were anesthetized with an intramuscular injection of a telazol–ketamine–xylazine solution (50 mg of zolazepam reconstituted with 2.50 mL ketamine [100 g/L] and 2.50 mL xylazine [100 g/L]; Fort Dodge Animal Health, Overland Park, KS) at 0.06 mL/kg BW. After sedation, pigs were euthanized through cardiac puncture using sodium pentobarbital (Euthasol Euthanasia Solution CIIIN; Patterson Veterinary Supply, St. Paul, MN) at a dosage of 1 mL/4.5 kg BW. The ileocecal junction was located, and tissue and mucosal samples were collected from the distal ileum of the designated three pigs on both 5 and 10 DPI. One tissue sample was flushed with PBS, and mucosal scrapings were collected and snap-frozen in liquid nitrogen. These mucosal samples were homogenized in PBS and used to quantify porcine secretory immunoglobulin A (sIgA) using the Pig IgA ELISA kit (assay range: 1.37 to 1,000 ng/mL, catalog number E101-102, Bethyl Laboratories, Montgomery, TX) and measure total protein via the Bradford assay. An additional ileal tissue sample was fixed in 10% neutral buffered formalin and later transferred to 70% ethanol pending histological analysis. All histological slides contained two representative cross sections from the ileum of each pig. Slides were subsequently used to quantify myeloperoxidase (MPO), a marker of neutrophil infiltration, via immunohistochemistry (IHC) staining. Sections were deparaffinized and rehydrated using Histo-clear and a graded alcohol series. Sections were then submerged in a solution containing 10 mM sodium citrate and 0.1 M citric acid, pH 6, and heated in a specialized chamber (BioCare Medical, Pacheco, CA) for antigen retrieval. Slides were brought to room temperature and rinsed in ultrapure water before following designated staining procedures (New & Improved Super Sensitive 1-STEP Polymer HRP Kit, catalog number QD600-GPEN, BioGenex, Fremont, CA) with slight modifications. Briefly, after a peroxide block, the incubation period for the Power Block solution was extended to 30 min. After this incubation, the primary MPO Rabbit Polyclonal Antibody (catalog number RB373R7, Thermo Fisher Scientific) was added to sections, which were then incubated overnight at room temperature. The secondary poly-HRP antibody and subsequent diaminobenzidine working solution incubation periods were extended to 1 h and 20 min, respectively. Sections were counterstained, dehydrated, and coverslipped using Permount Mounting Medium. After staining, digital images of slides were captured using the Nanozoomer 2.0-HT Slide Scanner (Hamamatsu, Shizuoka, Japan) at 20× magnification. The built-in visual stain editor on QuPath v0.3.2 software was used to preprocess the images, estimate intensity, and adjust the stain vectors as necessary (Bankhead et al., 2017). One entire tissue selection was selected as the field of view and was analyzed by the “positive cell detection” feature. The detection image for this step was set as “optical density sum,” which detected nuclei based on brightness information, regardless of stain color, and improved the accuracy of overall cell detection. After the automated positive cell detection was completed, any misidentified cells were manually annotated as either positive or negative. The percentage of total cells in an ileal tissue cross section that stained positive for myeloperoxidase was recorded. The same process was repeated for the second section on the slide and the average of the two sections was reported.

Intestinal samples to assess local gene expression

An additional ileal tissue sample from the three selected pigs was flushed of luminal contents, and tissue and mucosal scrapings were collected and snap-frozen in liquid nitrogen. For gene expression analyses, 50 to 100 mg of tissue were homogenized in 1 mL of Trizol (Invitrogen, Carlsbad, CA) through tissue disruption for 2 min at 30 Hz (TissueLyser II, Qiagen, Valencia, CA). Manufacturer instructions were followed to extract RNA from the samples, which was then quantified and checked for purity using a spectrophotometer (NanoDrop ND-1000, Nano-Drop Technologies, Wilmington, DE). Complementary DNA (cDNA) was reverse-transcribed from extracted RNA using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific Inc.). Samples were placed in a thermocycler (Bio-Rad, Hercules, CA) for 10 min at 25 °C, 120 min at 27 °C, and 5 min at 85 °C before being cooled at 4 °C. After this process, samples were stored at −20 °C until plating. The TaqMan Gene Expression Assay (Thermo Fisher Scientific Inc.) was used to perform qPCR for target genes interleukin 6 (IL-6; NM_214399.1), interleukin 8 (IL-8; NM_213867.1), tumor necrosis factor α (TNF-α; NM_214022.1), granulocyte-macrophage colony-stimulating factor 2 (CSF-2; NM_214188.2), vascular endothelial growth factor A (VEGF-A; NM_214084.1), claudin 1 (CLDN-1; NM_001244539.1), and occludin (OCLN; NM_001163647.2; Thermo Fisher Scientific). The swine ribosomal protein L19 (RPL-19; XM_003131509.5; Thermo Fisher Scientific) served as the reference gene. The cDNA for each sample was amplified using TaqMan (Thermo Fisher Scientific Inc.) oligonucleotide probes containing 5ˊ fluorescent reporter dye and 3ˊ nonfluorescent quencher dye. Fluorescence was measured using the QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems, Foster City, CA) set to a maximum of 40 cycles. The expression of target genes was normalized through parallel amplification of the RPL-19 reference gene for each sample. The comparative Ct method described by Schmittigen and Livak (2008) was used to calculate relative gene expression. All results are expressed as fold-changes relative to the UC group.

Minimum inhibitory concentration to assess the in vitro repressive effect

The repressive effect of the monoglyceride combination was evaluated in vitro by counting the number of bacterial colonies growing on Mueller–Hinton agar plates which contained concentrations of the monoglycerides ranging from 0.3125 to 40 mg/mL. In brief, a density of 0.5 Mcfarland of bacterial suspension, containing approximately 1 × 10⁸ CFU/mL of F18 ETEC, was prepared using Sensititre nephelometer (Thermo Fisher Scientific). This suspension was further diluted in the Mueller–Hinton broth (Becton, Dickinson and Company, Franklin Lakes, NJ) to obtain the final inoculum containing 1 × 10³ CFU/mL of E. coli. A total of 100 µL of the above solution was plated onto a Mueller–Hinton agar plate with a known concentration of monoglycerides. The plate was incubated at 37 °C for 24 h, and the bacterial number was accessed by counting the colonies growing on the plate. The experiment was performed in triplicate and the reported colony counts were the average of the three plates. The lowest concentration plates without growth were reported as the minimum inhibitory concentration (MIC).

Statistical analyses

Aside from the UC group, which was housed separately and only had four replicate pens, the other two treatments had six replicate pens that were arranged in a completely randomized design. As mentioned before, for all fecal and intestinal samples, 3 pigs from every pen (48 pigs total) were repeatedly used for the 5 and 10 DPI collections, as well as the intestinal integrity outcomes collected on 4 and 9 DPI. The individual pig served as the experimental unit for all outcomes aside from growth performance, where the pen was the experimental unit. Data were corrected for mortality and subjected to an analysis of variance (ANOVA) using the MIXED procedure of SAS (version 9.4; SAS Institute, Cary, NC). A one-way ANOVA was used to determine whether the model was significant, and, if significant, means separation was conducted assuming an absolute alpha level of 0.05. Values were considered outliers and removed from analysis if their absolute Studentized residual values were greater than 3. Estimates of the least square means and SEM were derived from this one-way ANOVA and significance was accepted if the P-value was less than 0.05.

The data obtained from bacterial enumeration of F18 ETEC were additionally subject to the PLM procedure of SAS, specifically a Tukey–Kramer test. Before this statistical analysis, data were converted into discrete values of either 0, 1, 2, or 3, depending on the number of pigs that tested positive (Ct ≤ 37) out of the three animals sampled per pen. The mean number reported was calculated by taking the weighted average of the cumulative probabilities and indicated the predicted number of pigs, out of the three tested in each pen, that would be positive for F18 ETEC based on their treatment group.

Results

Animal attrition and removal

Upon arrival at the facility, animals appeared healthy and in good condition; however, 3 d after arrival (DPI −4) and before ETEC inoculation, a high incidence of watery diarrhea was observed. Fecal samples were collected from all 16 pens, and adjacent pens were pooled such that eight pooled samples were submitted to the University of Illinois Veterinary Diagnostic Laboratory to test for PEDV, PDCoV, TGEV, and rotavirus. It was confirmed the following day (DPI −3) that this diarrhea outbreak was the result of a widespread rotavirus infection likely acquired from the source farm as all but one pooled sample tested positive for rotavirus. There was no detection of PEDV, PDCoV, or TGEV in any of the pools. This infection, combined with the other stressors associated with weaning, likely contributed to the loss of four pigs before the ETEC challenge. To help combat this unanticipated health challenge, all pigs were provided electrolytes via the water starting on −2 DPI and ending the morning of the first inoculation on 0 DPI. During the remainder of the study, 12 additional pigs died, with 10 of these being in the ETEC-inoculated groups.

Growth performance

Growth performance results can be found in Table 3. The only significant difference observed during the peak-infection period (0 to 5 DPI), was within the UC group which had a greater (P < 0.05) G:F than the other two treatments. No significant differences were detectable for ADG or ADFI between the three treatments during the preinoculation, postinoculation, or overall phases.

Table 3.

Effects of monoglyceride supplementation on the growth performance of weanling pigs infected with ETEC¹

Item	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
BW, kg
−7 DPI	6.44	6.39	6.46	0.167	0.940
0 DPI	6.36	6.61	6.51	0.201	0.644
5 DPI	7.67	7.69	7.61	0.272	0.965
10 DPI	9.38	9.14	9.38	0.457	0.876
−7 to 0 DPI
ADG, g/d	−8.49	22.5	12.5	15.659	0.336
ADFI, g/d	99.0	107	111	12.704	0.767
G:F	−0.123	0.183	0.065	0.150	0.321
0 to 5 DPI
ADG, g/d	257	202	207	28.861	0.313
ADFI, g/d	286	306	299	28.583	0.865
G:F	0.893a	0.655b	0.685b	0.057	0.016
5 to 10 DPI
ADG, g/d	323	280	335	60.813	0.721
ADFI, g/d	523	432	462	47.632	0.365
G:F	0.633	0.623	0.723	0.106	0.686
−7 to 10 DPI
ADG, g/d	173	151	162	27.895	0.833
ADFI, g/d	237	236	240	21.261	0.990
G:F	0.720	0.620	0.662	0.070	0.560

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, each initially housing 10 pigs. Pigs were removed from analysis for ADG, ADFI, and G:F if their BW measurement was deemed a statistical outlier. All pigs received allotted treatment diet starting at −7 DPI.

^abMeans lacking a common superscript letter differ (P < 0.05).

Abbreviations: ADFI, average daily feed intake; ADG, average daily gain; BW, body weight; DPI, days postinoculation; ETEC, enterotoxigenic E. coli; G:F, gain:feed/feed efficiency.

Rectal temperatures and fecal characteristics

The results for the rectal temperatures and DM analysis of fecal samples can be found in Table 4. At 5 DPI, pigs in the UC group had a lower (P < 0.05) rectal temperature of 39.12 °C compared with the IC and MG groups of 39.62 °C and 39.56 °C, respectively. The UC additionally had a greater (P < 0.05) fecal DM percentage of 46.73% compared with 23.04% in the IC group and 21.44% in the MG group. Rectal temperatures and fecal DM did not differ between the three treatments at 10 DPI.

Table 4.

Effects of monoglyceride supplementation on fecal characteristics and rectal temperature of weanling pigs infected with ETEC¹

Item	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
Fecal DM, %
0 DPI2	20.56	19.39	18.45	1.251	0.432
5 DPI	46.73a	23.04b	21.44b	3.659	<0.001
10 DPI	26.51	24.68	25.11	2.554	0.863
Rectal temperature, °C
5 DPI	39.12a	39.62b	39.56b	0.106	0.002
10 DPI	39.77	40.22	40.26	0.161	0.050

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, sampling three pigs per pen. All pigs received allotted treatment diet starting at −7 DPI.

²On −3 DPI it was confirmed by the University of Illinois College of Veterinary Medicine Veterinary Diagnostic lab that fecal samples submitted the previous day tested positive for rotavirus, explaining the low % DM for samples on 0 DPI. All pigs were provided electrolytes via the water starting on −2 DPI and ending the morning of the first inoculation on 0 DPI.

^abMeans lacking a common superscript letter differ (P < 0.05).

Abbreviations: DM, dry matter; DPI, days postinoculation; ETEC, enterotoxigenic E. coli.

Data for fecal bacterial enumeration are summarized in two different ways. First, Ct values for total bacteria, total E. coli, and total F18 ETEC are reported in Table 5. Any samples that were nondetectable were assigned a Ct of 40.0, which was the total number of cycles used for qPCR determination. A larger Ct value represents a greater number of PCR cycles required for detection and indicates a lower number of bacteria present in the sample. Total bacteria, total E. coli, and total F18 ETEC were all lower (P < 0.05) on 5 DPI in the feces of the UC group compared with the IC and MG treatment groups, as demonstrated by the greater mean Ct value. There were no statistically significant differences observed amongst treatment groups for any of the fecal bacterial profiles at 10 DPI.

Table 5.

Effects of monoglyceride supplementation on fecal bacterial profiles (Ct values) in weanling pigs infected with ETEC¹

Item	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
Total bacteria (rrs)
5 DPI	24.94a	21.46b	20.76b	1.074	0.012
10 DPI	25.08	24.89	23.42	1.011	0.362
Total E. coli (gadAB)
5 DPI	31.50a	25.14b	24.48b	1.480	0.001
10 DPI	30.39	29.06	27.61	1.452	0.346
F18 ETEC (fedA)2
5 DPI
Ct	ND	26.65	23.43	1.674	0.157
Detectable	0/12	15/18	11/18	—	—
10 DPI
Ct	ND	30.41	32.64	1.528	0.301
Detectable	0/12	7/17	9/16	—	—

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, sampling three pigs per pen. All pigs received allotted treatment diet starting at −7 DPI. For all values, a larger Ct value represents more cycles required for detection and thus a lower number of bacteria and vice versa.

²All uninfected pigs were undetectable (Ct > 37) at both time points, so this treatment was not included for statistical analysis and SEM and P-values are representative of only the infected control and infected supplemented groups. The number of pigs detectable for F18 ETEC at each time point out of 18 total per treatment group are listed as well. Totals less than 18 on 10 DPI are due to pigs that were removed as outliers or from which no sample was obtained.

^abMeans lacking a common superscript letter differ (P < 0.05).

Abbreviations: Ct, cycle threshold; DPI, days postinoculation; ETEC, enterotoxigenic E. coli; fedA, F18 fimbrial subunit protein; gadAB, glutamate decarboxylase A and B genes; ND, nondetectable; rrs, ribosomal RNA gene.

Second, due to many samples having nondetectable levels of F18, we chose to also convert the data into discrete values of either 0, 1, 2, or 3, with each number corresponding to the number of pigs that tested positive for F18 ETEC out of the three pigs sampled per pen (Table 6). Samples with a Ct ≤ 37 were considered positive and any sample with a Ct > 37 was considered negative. While significance could not be assigned to these values, there was an 83% cumulative probability that all three pigs sampled in the IC group on 5 DPI would test positive for F18 ETEC, whereas this probability was only 16% for the MG treatment. By 10 DPI, the cumulative probability of all three pigs testing positive in the IC and MG groups was 5% and 12%, respectively.

Table 6.

Effects of monoglyceride supplementation on the number of pigs positive for F18 ETEC per pen¹

Item	Number of pigs positive for F18 per pen2				Mean number
	0	1	2	3
5 DPI
Uninfected control	1.00	0.00	0.00	0.00	0.00
Infected control	0.00	0.01	0.16	0.83	2.82
Infected supplemented	0.00	0.17	0.67	0.16	1.99
10 DPI
Uninfected control	1.00	0.00	0.00	0.00	0.00
Infected control	0.07	0.51	0.37	0.05	1.40
Infected supplemented	0.03	0.31	0.54	0.12	1.75

¹Values represent the cumulative probability of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups that 0, 1, 2, or 3 pigs, out of the three tested per pen, test positive for F18 ETEC, with a Ct value ≤ 37 considered to be positive.

²A maximum value of 1.00 represents 100%. The mean number was calculated by taking the weighted average of the cumulative probabilities and indicates the predicted number of pigs, out of the three tested in each pen, that would test positive for F18 ETEC based on their treatment group.

Abbreviations: DPI, days postinoculation; ETEC, enterotoxigenic E. coli; F18, fimbrial subunit protein.

Blood samples to assess intestinal integrity

Plasma concentrations of lactulose and mannitol, as well as their ratio, are summarized in Table 7. The MG group had a greater (P < 0.05) concentration of plasma mannitol on 9 DPI compared with the IC group. No significant differences between dietary treatment groups were observed for the L:M ratio.

Table 7.

Effects of monoglyceride supplementation on lactulose and mannitol concentrations in blood plasma of weanling pigs infected with ETEC¹

Item	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
Lactulose, µmol/L
4 DPI	15.15	15.83	11.85	2.201	0.327
9 DPI	10.36	12.60	11.04	2.758	0.800
Mannitol, µmol/L
4 DPI	24.77	20.39	19.07	3.979	0.550
9 DPI	33.12ab	20.59a	39.19b	5.088	0.010
L:M
4 DPI	0.778	1.024	0.821	0.155	0.438
9 DPI	0.426	0.809	0.421	0.162	0.083

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, sampling three pigs per pen. All pigs received allotted treatment diet starting at −7 DPI.

^abMeans lacking a common superscript letter differ (P < 0.05).

Abbreviations: DPI, days postinoculation; ETEC, enterotoxigenic E. coli; L:M, lactulose:mannitol ratio.

Intestinal antibody expression and neutrophil infiltration

Mucosal scrapings and cross sections of ileal tissue were used to quantify ileal sIgA content and MPO expression, respectively, on both 5 and 10 DPI. However, no significant differences were observed amongst the three treatments for these outcomes as shown in Table 8.

Table 8.

Effects of monoglyceride supplementation on sIgA and MPO levels in the ileum of weanling pigs infected with ETEC¹

Item	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
sIgA, µg/mg protein2
5 DPI	1.281	1.169	1.407	0.153	0.423
10 DPI	2.160	2.299	1.673	0.345	0.287
MPO, %3
5 DPI	0.4608	0.6063	0.5795	0.089	0.431
10 DPI	0.4117	0.4034	0.3385	0.079	0.701

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, sampling three pigs per pen. All pigs received allotted treatment diet starting at −7 DPI.

²Ileal mucosal scrapings used to determine total sIgA and total protein.

³Represented as the percentage of total cells in an ileal tissue cross section that stained positive for myeloperoxidase.

Abbreviations: DPI, days postinoculation; ETEC, enterotoxigenic E. coli; sIgA, secretory immunoglobulin A; MPO, myeloperoxidase.

Intestinal gene expression

Gene expression of inflammatory cytokines, angiogenic growth factors, and tight junction proteins was measured in ileal tissue, as displayed in Table 9. The expression of TNF-α, VEGF-A, and OCLN was lower (P < 0.05) in both the IC and MG treatment groups compared with the UC group at 5 DPI. By 10 DPI, there were no longer any differences between treatment groups in the expression of the genes measured.

Table 9.

Effects of monoglyceride supplementation on gene expression in ileal tissue of weanling pigs infected with ETEC¹

Item2	Treatment			SEM	P-value
	Uninfected control	Infected control	Infected supplemented
IL-6
5 DPI	1.000	1.154	0.651	0.278	0.290
10 DPI	1.000	1.022	0.761	0.191	0.436
IL-8
5 DPI	1.000	0.597	0.798	0.150	0.126
10 DPI	1.000	1.324	1.363	0.351	0.753
TNF-α
5 DPI	1.000a	0.538b	0.433b	0.166	0.031
10 DPI	1.000	0.860	0.644	0.139	0.117
CSF-2
5 DPI	1.000	0.621	0.629	0.142	0.079
10 DPI	1.000	0.777	1.059	0.265	0.625
VEGF-A
5 DPI	1.000a	0.443b	0.489b	0.091	< 0.001
10 DPI	1.000	0.884	0.932	0.208	0.908
CLDN-1
5 DPI	1.000	1.198	1.031	0.389	0.907
10 DPI	1.000	1.304	0.933	0.345	0.616
OCLN
5 DPI	1.000a	0.463b	0.554b	0.134	0.008
10 DPI	1.000	0.756	0.728	0.194	0.506

¹Values represent least square means of four pens for the uninfected control group and six pens for the infected control and infected supplemented groups, sampling three pigs per pen. All pigs received allotted treatment diet starting at −7 DPI.

²Gene expression levels are all displayed relative to the uninfected control treatment group within a given time point (i.e., row).

^abMeans lacking a common superscript letter differ (P < 0.05).

Abbreviations: CLDN-1, claudin 1; CSF-2, colony-stimulating factor 2; DPI, days postinoculation; ETEC, enterotoxigenic E. coli; IL-6, interleukin 6; IL-8, interleukin 8; OCLN, occludin; TNF-α, tumor necrosis factor alpha; VEGF-A, vascular endothelial growth factor A.

MIC assay

The MIC for the monoglyceride combination was 10 mg/mL, with the repressive effect on F18 ETEC being dose dependent. Colony counts increased from 2.10 × 10⁷ to 7.57 × 10⁷ to 8.50 × 10⁷ CFU/mL as the concentration of monoglycerides decreased from 5 to 2.5 to 1.25 mg/mL, respectively (Table 10).

Table 10.

Effects of monoglyceride supplementation on average F18 ETEC colony counts in vitro¹

MG concentration, mg/mL	Mean F18 ETEC colony count, × 107 CFU/mL	SEM
40.0000	0.00	0.000
20.0000	0.00	0.000
10.0000	0.00	0.000
5.0000	2.10	4.163
2.5000	7.57	2.848
1.2500	8.50	3.215
0.6250	8.83	2.186
0.3125	8.63	4.410
0.0000	8.27	6.960

¹Values represent the average reported colony count from three plates containing the same concentration of the monoglyceride supplement.

Abbreviations: CFU, colony-forming units; ETEC, enterotoxigenic Escherichia coli; MG, monoglyceride.

Discussion

The plethora of stressors associated with the weaning period in swine production often results in reduced growth performance and leaves the young pig vulnerable to diseases. Postweaning diarrhea, frequently caused by infection with F18 ETEC, is one such disease that further decreases performance and increases mortality. To diminish the sizable economic loss associated with diseases, there is much investigation into interventions that may confer growth and health benefits. A combination of the monoglycerides of butyric, caprylic, and capric fatty acids is one potential nutraceutical evaluated in this study for its effects on performance, inflammation, and intestinal integrity in weanling pigs challenged with F18 ETEC.

Although previous research using the F18-challenge model has frequently concluded that ETEC infection significantly decreased ADG and ADFI (Song et al., 2012; Li et al., 2019; Becker et al., 2020), this study found no significant effects due to induction of this infection. However, these results are consistent with those reported by Sun et al. (2021), and a previous trial conducted within our lab (Wojnicki et al., 2019). When faced with an enteric pathogen, sick animals are less efficient, as partially explained by the diversion of dietary energy and nutrients away from tissue accretion and towards immune response activation (Patience et al., 2015). Therefore, a significant decrease in G:F in the infected groups during the peak-infection period was expected. A total of 16 out of the 160 total pigs died over the course of this experiment, which is consistent with the level of mortality observed in previous studies using the same strain of ETEC (Song et al., 2012; Wojnicki et al., 2019). A spike in mortalities near the end of the experiment was likely the result of edema disease. Necropsy confirmed edema disease in two pigs, and the others displayed clinical signs characteristic of this illness including facial swelling and a staggered gait. The F18 ETEC strain used in this experiment does produce the Shiga-toxin associated with the development of edema disease, and this complication was similarly observed by Song et al. (2012).

As mentioned previously, a widespread rotavirus outbreak was confirmed on −3 DPI, before ETEC inoculation. This infection resulted in lower-than-expected values across all treatments for fecal DM on 0 DPI and growth and intake during the acclimation period (−7 to 0 DPI). Due to this decreased feed intake, it is possible that pigs may not have ingested enough of the monoglyceride treatment to influence the different measures that were analyzed. Additionally, villous atrophy, due to the rotavirus infection, may have diminished the effects of ETEC challenge due to the loss of enterocytes for ETEC adherence and toxin entry. While co-infection with rotavirus and E. coli is possible in commercial settings, the results of this experiment are most often compared with previous research that had only one health challenge intervention. Therefore, even with careful control over the experimental ETEC status of these pigs, it is difficult to determine exactly how these two enteric infections interacted with one another and led to other unexpected results.

Due to the adhesins and toxins of F18 ETEC, diarrhea is one of the most prominent clinical manifestations of ETEC infection. Therefore, a significant difference in fecal DM percentages between treatment groups at 5 DPI was expected. At this time point, ETEC-inoculated pigs in the IC and MG groups had more watery feces, as indicated by lower DM percentages. In addition to DM analysis, fecal samples were used for bacterial enumeration. Previous research indicated that butyrate supplementation to pigs reduced fecal E. coli levels, where the purported mechanism involved butyrate functioning as a histone deacetylase inhibitor to upregulate the gene expression of certain endogenous host defense peptides (Xiong et al., 2016). Along with butyrate, both Marounek et al. (2003) and Skřivanová et al. (2006) reported that caprylic and capric fatty acids directly inhibited the growth of E. coli in vitro, which could further decrease the prevalence of E. coli in the feces. Finally, in our own in vitro research, we discovered that the butyric, caprylic, and capric fatty acid glyceride combination inhibited F18 ETEC in a dose-dependent manner, with a MIC of 10 mg/mL. However, in vivo, we observed no difference between the IC and MG groups in terms of total bacteria, total E. coli, or total F18 ETEC levels. In general, it is difficult to translate in vitro results to in vivo experiments, but the reduced feed intake at the beginning of the experiment, possibly leading to less ingestion of the MG treatment, may have contributed to the discrepancy between the expected and observed bacterial counts. An additional limitation is that bacterial expression could not be quantified on a DM basis, so the variable water content of fecal samples could have affected this outcome. While bacterial levels did not differ between treatments, not all pigs in the two infected groups had detectable levels of F18 ETEC. Whether the numerical difference in probabilities between the MG and IC groups on 5 DPI can be attributed to treatment benefits or variation in genetic susceptibility to F18 warrants further research.

The normal rectal temperature of a young pig ranges from 38.6 to 39.2 °C, although in weanling pigs the upper end of this range may be closer to 39.6 °C (Sipos et al., 2013; Becker et al., 2020). Therefore, whereas the ETEC-infected groups had significantly greater rectal temperatures than the UC during the peak-infection period, these temperatures were still within the typical range. A high fever is not a common clinical sign of ETEC, and so the lack of a significant febrile response was not surprising (Zhang et al., 2022). Furthermore, while our collection time points were based upon the estimated timeline of ETEC infection, it is possible that the peak response occurred before our 5 DPI collection, leading to a lack of significant differences when it came to the febrile and inflammatory response outcomes. Elevated, yet not significantly different, rectal temperatures on 10 DPI may be explained by higher environmental temperatures on this day of sampling.

The recognition of lipopolysaccharide, found in the outer membrane of Gram-negative bacteria like ETEC, plays a key role in the initiation of the innate immune response and production of proinflammatory cytokines (Webel et al., 1997; Vaure and Liu, 2014). Although they have many effects on the body, these proinflammatory cytokines are also mediators of the febrile response (Marshall et al., 2018). Therefore, while we hypothesized that expression of TNF-α and IL-6 in localized ileal tissue would be increased in ETEC-inoculated pigs but possibly attenuated by monoglyceride supplementation, the minimally elevated rectal temperatures indicated that there was likely not a significant amount of inflammation occurring at the local intestinal level. While there were no significant differences in terms of IL-6 or IL-8 expression, the decreased expression of TNF-α in the two infected groups on 5 DPI was contrary to both our hypothesis and previous literature (Webel et al., 1997; Liu et al., 2014; Vaure and Liu, 2014). One possible explanation is that weaning stress and rotavirus infection, two events that occurred before the ETEC challenge in our study, resulted in increased TNF-α expression in ileal tissue and circulation of pigs. Previous studies have demonstrated such increases and additionally discovered that TNF-α remained elevated in the distal small intestine of weaned pigs and in the serum of rotavirus-infected pigs (Pié et al., 2004; Azevedo et al., 2006). Due to the inflammatory nature of TNF-α, there are many regulatory mechanisms in place to prevent the damage associated with prolonged inflammation. These complex processes involve regulating gene transcription and translation, receptor expression and activation, and intracellular signaling (Moelants et al., 2013). Inflammation from weaning and rotavirus infection, exacerbated by 3 d of ETEC inoculation, may have led to the induction of one or many of these negative feedback mechanisms and caused the reduction in TNF-α expression on 5 DPI; further research is needed to corroborate this hypothesis.

As sIgA is the primary antibody of the small intestinal mucosa, we expected to see an increase in sIgA concentrations within the ileal mucosal scrapings as the immune response of ETEC-inoculated pigs was activated (Bianchi et al., 1999). Similarly, we expected to see an increase in MPO in the ileal cross sections as it is an enzyme released by neutrophils during an immune response, and therefore serves as a marker for neutrophil infiltration and the overall severity of intestinal inflammation (Klebanoff, 2005). This was not the case as there were no significant differences for either outcome. This further indicated that, similar to the rectal temperatures and gene expression, infection and its associated inflammation was either not severe enough to cause dramatic changes in sIgA or MPO production, or this response occurred before the 5 DPI collection. Alternatively, the change in F18-specific sIgA activity may have been masked by nonspecific IgA production; a phenomenon observed by Trevisi et al. (2015). They discovered that total IgA activity did not follow the same trends as K88 ETEC-specific IgA, and significant differences were only seen when K88-specific IgA was measured. As we quantified total, not pathogen-specific, sIgA in our analysis, this may help to explain the lack of differences between treatments.

Angiogenesis is the process by which new blood vessels are formed that occurs throughout life, regardless of health status (Adair and Montani, 2010). One of the most well-characterized mediators of this process is VEGF-A, which plays a role in promoting both physiological and pathological angiogenesis (Scaldaferri et al., 2009). The expression of VEGF, and thus angiogenesis, is regulated by CSF-2 (Cianfarani et al., 2006; Zhao et al., 2014). The production of this growth factor may be induced by bacterial endotoxins or inflammatory cytokines, indicating that expression levels of these two genes might be elevated in our infected treatment groups (Shi et al., 2006). Additionally, natural and synthetic sources of monobutyrin have been implicated as stimulators of angiogenesis, leading us to hypothesize that expression of VEGF-A and CSF-2 would be elevated in our treatment group fed the monoglyceride supplement (Dobson et al., 1990). However, we once again observed the exact opposite, as VEGF-A expression was significantly lower and CSF-2 expression was numerically lower on 5 DPI. While there is limited information regarding where these angiogenic factors are expressed in the small intestine, Holmes et al. (2008) discovered that VEGF-A expression was the strongest in the epithelial cells and lamina propria of sheep villi. Similarly, CSF-2 appears to be broadly expressed by the small intestine epithelial cells of mice (Egea et al., 2013). Both rotavirus and ETEC infection decrease intestinal villus height and, because all genes were measured locally in ileal tissue, it may be possible that the cells producing these growth factors were sloughed during the peak of infection at 5 DPI (DebRoy and Maddox, 2001; Becker et al., 2020; Sun et al., 2021). This hypothesis would also align with the fact that VEGF-A and CSF-2 expression was like that of the UC group by 10 DPI, as recovery should have begun by this time point.

Limited research has been conducted using the dual sugar test as a measure of intestinal permeability in young pigs, with a dearth of information on concentrations of lactulose and mannitol in plasma, as opposed to quantification in urine samples. Hu et al. (2023) orally dosed pigs with a lactulose and mannitol mixture and discovered that L:M ratios obtained from plasma were strongly correlated to those measured in urine. As blood collection in animals is often simpler and more practical than urine collection, these results are promising that this may be an alternative way to measure barrier function in pigs. However, we did not observe any differences in the plasma L:M ratio between treatments at 5 or 10 DPI. Wensley et al. (2023) and Wilson et al. (2022) were two of the only other groups to measure L:M ratios in the plasma of nursery pigs, and, similarly, neither found significant treatment differences for this measure of intestinal integrity.

The tight junction proteins OCLN and CLDN-1 are crucial in maintaining the selective passage of molecules across the intestinal epithelium but are often disrupted or downregulated by ETEC infection (Mukiza and Dubreuil, 2013; Li et al., 2019; Becker et al., 2020). Therefore, we anticipated decreased expression of OCLN and CLDN-1 in ETEC-infected pigs but hypothesized that feeding a combination of monoglycerides may alleviate some of the damage as butyrate has the potential to increase tight junction protein expression and assembly (Wang et al., 2012; Yan and Ajuwon, 2017; Feng et al., 2018). We did observe a decrease in OCLN expression at 5 DPI; however, this reduction was not ameliorated with the monoglyceride treatment. Expression of CLDN-1 did not mimic what was seen with OCLN, as there were no significant differences between treatment groups.

In conclusion, these results suggest that supplementation with a combination of butyric, capric, and caprylic fatty acid monoglycerides at 0.3% did not improve growth or health measures in weanling pigs that were infected with F18 ETEC. During the peak-infection period, ETEC-challenged pigs exhibited increased rectal temperatures and total bacteria, total E. coli, and total F18 ETEC in the feces, as indicated by significantly lower Ct values. Challenged pigs additionally had decreased G:F, fecal DM, and TNF-α, VEGF-A, and OCLN gene expression at this time. However, these were the only significant differences observed during this study and they were not impacted by the nutritional intervention. It is possible that diminished feed intake, due to an uncontrolled rotavirus infection at the beginning of the study, may have blunted the effects of monoglyceride supplementation. Alternatively, the monoglyceride inclusion rate necessary to demonstrate effectiveness may be greater than that used for this study. Furthermore, the expression of TNF-α and VEGF-A was analyzed locally in ileal tissue samples that had previously been impacted by rotavirus, which could explain why their decreased expression in infected pigs was contrary to what was initially expected. Overall, more research needs to be performed, and at varying doses, to determine if dietary supplementation with a combination of monoglycerides can adequately mitigate the deleterious effects of an ETEC challenge in weanling pigs.

Glossary

Abbreviations:

ADFI

average daily feed intake

ADG

average daily gain

ANOVA

analysis of variance

BW

body weight

cDNA

complementary DNA

CFU

colony-forming units

CLDN-1

claudin 1

CSF-2

granulocyte-macrophage colony-stimulating factor 2

Ct

cycle threshold

DM

dry matter

DPI

days postinoculation

ETEC

enterotoxigenic Escherichia coli

G:F

gain:feed

HPLC

high-performance liquid chromatography

IHC

immunohistochemistry

IL-6

interleukin 6

IL-8

interleukin 8

L:M

lactulose:mannitol

MIC

minimum inhibitory concentration

MPO

myeloperoxidase

OCLN

occludin

PBS

phosphate-buffered saline

PDCov

porcine deltacoronavirus

PEDV

porcine epidemic diarrhea virus

PWD

postweaning diarrhea

qPCR

quantitative polymerase chain reaction

RPL-19

ribosomal protein L19

sIgA

secretory immunoglobulin A

TGEV

transmissible gastroenteritis virus

TNF-α

tumor necrosis factor α

VEGF-A

vascular endothelial growth factor A

Conflict of interest statement. Adriana Barri and Adebayo Sokale are employees of BASF (Ludwigshafen, Germany). No other authors have conflicts of interest to declare.