Impairment of electroneutral Na⁺ transport and associated downregulation of NHE3 contributes to the development of diarrhea following in vivo challenge with Brachyspira spp.

Abstract

The effect of Brachyspira hyodysenteriae and Brachyspira hampsonii spirochetosis on Na⁺ transport was assessed in the colon to determine its contribution to diarrheal disease in pigs following experimental infection. Electrogenic and electroneutral Na⁺ absorption was assessed in Ussing chambers by radiolabeled ²²Na flux and pharmacological inhibitory studies. Basal radiolabeled ²²Na flux experiments revealed that mucosal-to-serosal flux (J_ms) was significantly impaired in B. hyodysenteriae and B. hampsonii-diseased pigs. Inhibition of epithelial sodium channel via amiloride did not significantly reduce electrogenic short-circuit current (I_sc) in the proximal, apex, and distal colonic segments of diseased pigs over control pigs, suggesting that a loss of electroneutral Na⁺ absorption is responsible for diarrheal development. These findings were further supported by significant downregulation of Na⁺/H⁺ exchanger (NHE1, NHE2, and NHE3) mRNA expression in the proximal, apex, and distal colonic segments paired with decreased protein expression of the critical NHE3 isoform. The decrease in NHE3 mRNA expression appears not to be attributed to the host’s cytokine response as human IL-1α did not modify NHE3 mRNA expression in Caco-2 cells. However, a whole cell B. hampsonii lysate significantly downregulated NHE3 mRNA expression and significantly increased p38 phosphorylation in Caco-2 cells. Together these findings provide a likely mechanism for the spirochete-induced malabsorptive diarrhea, indicated by a decrease in electroneutral Na⁺ absorption in the porcine colon due to Brachyspira’s ability to inhibit NHE3 transcription, resulting in diarrheal disease.

NEW & NOTEWORTHY This research demonstrates that diarrheal disease caused by two infectious spirochete spp. is a result of impaired electroneutral Na⁺ absorption via Na⁺/H⁺ exchanger 3 (NHE3) in the porcine colon. Our findings suggest that the decrease in NHE3 mRNA and protein is not likely a result of the host’s cytokine response. Rather, it appears that these two Brachyspira spp. directly inhibit the transcription and translation of NHE3, resulting in the development of diarrhea.

INTRODUCTION

Electroneutral Na⁺/H⁺ exchange is responsible for developing the osmotic drive for fluid absorption along the gastrointestinal tract of mammals. Inhibition of this exchange process has been shown to contribute to diarrheal disease caused by multiple enteric pathogens. However, the effect of the colonic spirochete Brachyspira on the electroneutral exchange has not been determined. In humans, colonic spirochetosis is caused by Brachyspira pilosicoli and Brachyspira aalborgi (39, 56). In swine, Brachyspira hyodysenteriae and emergent Brachyspira hampsonii cause severe spirochetosis, causing production limiting diarrhea accompanied by varying amounts of fecal blood and mucous (12, 21, 48).

Studies assessing the effect of Brachyspira spp. on solute transport have been limited to the pathogenic swine species B. hyodysenteriae, which has been described as causing malabsorptive diarrhea (4, 51). Ligated colonic loop experiments were utilized to determine the effects on Na⁺ and Cl⁻ transport (4, 51). The authors concluded that B. hyodysenteriae abolished the absorptive capacity of the porcine colon while having no effect on secretion (4, 51). However, these studies were performed using a version of the Berger and Steele equation, and nonlinear changes in net electrolyte movement would lead to misinterpretation of the results (4, 6, 51). Additionally, the transporters that would need to be altered to produce this response were not assessed.

The effect of Brachyspira spp. on the absorptive capacity of the epithelium is greatest in the porcine spiral colon, which can be divided into the proximal, apex, and distal segments. Absorption of Na⁺ and Cl⁻ by the colonic epithelium is primarily carried out by electroneutral Na⁺/H⁺ (NHE) exchangers found in all three segments (32). The remaining Na⁺ absorption is electrogenic and is due to absorption through luminal epithelial sodium channel (ENaC), which is predominantly expressed in the distal colon (44).

Na⁺/H⁺ exchangers (NHE) are electroneutral cation exchangers in which a single extracellular Na⁺ is exchanged for a cytosolic H⁺. NHE isoforms 1–3 are predominantly expressed in intestinal epithelia with NHE2 and NHE3 expressed on the apical surface of epithelial cells in the ileum and colon (1, 8, 26). NHE1 is localized to the basolateral membrane of epithelial cells and plays an important role in the regulation of cellular pH (8). NHE3 is essential for Na⁺ and water absorption in the gastrointestinal tract as NHE3 knockout mice suffered from chronic diarrhea (53), whereas NHE2 knockouts did not, suggesting that NHE3 is responsible for the majority of Na⁺ absorption and water homeostasis in the gastrointestinal tract (52). Furthermore, electroneutral Na⁺ absorption is inhibited by bacterial pathogens such as Salmonella typhimurium, Vibrio cholerae, and Campylobacter jejuni that cause an increase in intracellular Ca²⁺, cAMP, and/or cGMP, inhibiting NHE2 and NHE3 (20, 27, 30, 55). Enteropathogenic Escherichia coli infections significantly downregulate NHE3 activity while NHE1 and NHE2 are stimulated in response to the infection in vitro; however, diarrhea ensues (23).

ENaC is a luminal Na⁺ channel responsible for the electrogenic absorption of Na⁺ in the distal colon (18, 49). The channel is composed of three homologous subunits α, β, and γ (11). Its function is regulated by mineralocorticoids and glucocorticoids such as aldosterone, which have been shown to upregulate mRNA of ENaC’s β- and γ-subunits.

Here we have characterized the pathophysiological effects of B. hyodysenteriae and B. hampsonii on Na⁺ absorption in the porcine colon. Electrogenic and electroneutral Na⁺ absorption via ENaC and NHE isoforms was assessed in Ussing chambers. A decrease in electrogenic Na⁺ absorption was not observed. However, electroneutral radiolabeled ²²Na flux revealed that mucosal-to-serosal flux (J_ms) Na⁺ transport was significantly reduced. These findings along with a decrease in NHE isoforms 1–3 mRNA and reduced NHE3 protein expression suggest a reduction in the ability of the porcine colon to absorb Na⁺ during Brachyspira diarrhea. Loss of NHE3 gene transcripts was not attributed to the host’s cytokine response as IL-1α (16, 57) did not reduce NHE3 mRNA expression in Caco-2 cells. However, a whole cell Brachyspira lysate significantly reduced NHE3 mRNA expression in Caco-2 cells after 48 h of exposure. These findings suggest that Brachyspira directly inhibit ion channel transcriptional processes resulting in diarrheal disease.

METHODS

Animals.

In 2 separate studies, 30 (²²Na flux study) and 54 (Electrogenic Ussing chamber studies) 6- to 8-wk-old purebred Yorkshire barrows were housed in pairs with 10 pigs per room and provided an antibiotic-free diet with water ad libitum. Treatment groups were each housed in separate BSL2 animal care rooms. Pigs acclimated to their new environment for 7 days before they were inoculated with either B. hyodysenteriae strain G44 (kindly provided by Boehringer-Ingelheim Vet Medica, St. Joseph, MO; n = 12, n = 18), B. hampsonii strain 30446 (kindly provided by Dr. Janet Hill, University of Saskatchewan; n = 12, n = 18), or a mock inoculum of sterile culture media (n = 12, n = 18) as previously described in detail (48). Feeders were removed from the pens 12 h before inoculation, but water was not. Inoculation was conducted by passing a stomach tube and flushing the inoculum into the stomach of a pig, followed by sterile PBS. Pigs were assessed daily for clinical signs of diarrhea, and fecal consistency scores were recorded twice a day to accurately determine the onset of diarrhea. Fecal consistency scores were graded on a 0- to 4-point scale (0 = formed, normal; 1 = soft, wet cement like consistency; 2 = water or runny; 3 = mucoid diarrhea; or 4 = mucohemorrhagic diarrhea) as previously described (48). Brachyspira challenged pigs developed diarrhea within 3–7 days postinoculation and were euthanized 24 h after the onset of mucohemorrhagic diarrhea. Control pigs remained healthy (nondiarrheic) and were euthanized on a matched basis with Brachyspira challenged pigs euthanized on the same day. This research was designed and conducted in accordance with the Canadian Council for Animal Care and approved by the University of Saskatchewan Committee on Animal Care and Supply (Protocol No. 20130034).

Electrogenic Ussing chamber studies.

After euthanasia, ~18-cm segments of proximal (2.5 cm distal of the cecum), apex of the spiral colon (midpoint between cecum and sigmoid colon), and distal (sigmoid colon) were collected and washed with Krebs buffer (pH 7.4) containing the following (in mM): 113 NaCl, 5 KCl, 1.6 Na₂HPO₄, 0.3 NaH₂PO₄·H₂O, 25 NaHCO₃, 1.1 MgCl₂·6H₂O, 2.2 CaCl₂·2H₂O, and 10 glucose and chilled to 4°C. Samples were immediately transported back to the laboratory in Krebs buffer gassed with 95% O₂-5% CO₂. The colonic segments were then stripped with forceps by removal the serosa (visceral peritoneum) and the longitudinal/circular muscle layers of the intestinal wall, leaving only the underlying submucosal elements, remnants of muscle, and the epithelium. The stripped tissues (4 tissue replicates of each segment per pig) were then placed on 1-cm² tissue Ussing chamber inserts and inserted into the Ussing chamber (Physiologic Instruments, San Diego, CA). Each reservoir was independently gassed with 95% O₂-5% CO_2. A heated circulating water bath warmed and maintained the buffer temperature within the Ussing chamber to 37°C. Transepithelial potential differences were short circuited to 0 mV with a voltage clamp on the apical and basolateral chambers that corrected for bath resistance using Ag-AgCl electrodes and 3 M KCl agar bridges.

Tissues were allowed to equilibrate for 20 min before the addition of amiloride. Tissues were pulsed every 30 s with a 1-mV pulse, and the resulting current was used to determine tissue resistance. After the equilibration period, amiloride (A7410; Sigma Aldrich) was added to the apical side of the Ussing chamber at a final concentration of 0.1 mM, selectively inhibiting ENaC.

Characterization of electroneutral absorptive response in healthy and diseased porcine colon ²²Na flux study.

The apex of the spiral colon was collected from control and diseased pigs following euthanasia, and samples were prepared according to the protocol described above. Once the colon tissues (12 tissue replicates per pig) were inserted into the Ussing chambers, short-circuit current was recorded. Tissues were recorded for 20 min to equilibrate to a steady-state current. Tissues were paired with one another based on the resistance of each tissue. Tissues that had resistances that differed <15% of each other were paired. One microcurie of ²²Na was then added to the apical side of chamber 1, and 1 μCi was added to the basolateral side of chamber 2. The side of the chamber that ²²Na is added to is referred to as the “hot side,” and the opposing side is referred to as the “cold side.” Following the addition of ²²Na, 100-μL samples were removed from the hot side of both chambers immediately and placed in glass vials. The 100 μL removed from the Ussing chamber were immediately replaced with 100 μL of fresh Krebs buffer. Five-hundred microliters samples were removed from the cold side of both chambers and placed in glass vials and once again replaced with 500 μL of fresh Krebs buffer. Five-hundred microliter samples were removed from the cold side every 10-min for the first 80 min for steady-state flux to be achieved. All samples collected were counted on a Titertek Plus Series gamma counter.

Cold samples were used to calculate unidirectional mucosal-to-serosal (J_ms), serosal-to-mucosal (J_sm), and net flux (J_net = J_ms – J_sm) for ²²Na as previously described (54). Positive net flux values indicate net absorption while negative values indicate net secretion of the isotope.

Quantitative RT-PCR analysis of ion channel and transporter mRNA expression.

Mucosal samples taken at the time of necropsy were stored in RNAlater (AM7021; Ambion), homogenized in 1 mL of TRIzol reagent (15596018; Life Technologies), and RNA was extracted according to the manufacturer’s protocol. A standard of <500 ng/μL was used as exclusion criteria for RNA samples.

cDNA was created from mRNA using the Go Script Reverse Transcription system (A5001; Promega). cDNA was diluted in RNase-free water and frozen at −80°C. Gene expression was assessed by quantitative RT-PCR (RT-qPCR) using GoTaq qPCR Master Mix (A6002; Promega) and Stratagene Mx5000P real-time qPCR machines. The average C_T (cycle threshold) value was used to calculate the fold difference of each gene using the ΔΔC_T calculation method. Porcine primers were designed for GAPDH, NHE1, NHE2, NHE3, ENaC-α, ENaC-β, and Na-K-ATPase α₁-subunit (ATP1A1) as previously described (16) (Table 1). Porcine GAPDH was used as the reference gene for the analysis. Human primers were designed for GAPDH, NHE2, NHE3, and prostaglandin-endoperoxide synthase 2 (PTGS2) (Table 2). Human GAPDH was used as the reference gene for the analysis.

Table 1.

Porcine primer sequences for quantitative RT-PCR

Gene Name	Forward (5′-3′)	Reverse (5′-3′)
GAPDH	ACATCAAGAAGGTGGTGAAGCAGG	TGAGCTTGACGAAGTGGTCGTTGA
NHE1	CAGAGGACTGCTTCCACAAA	CAAAGTGAGACCTGGGACATAG
NHE2	GGCAGAGACTGGGATGATAAG	TCGCTGACGGATTTGATAGAG
NHE3	GGTGCTCTTCATCATCGTCTTC	CCAGCGTCACGAAAGATTCA
ENaC-α	TGCACTGGGCAAATTCATCTTCGC	ACATCCAGAGGTTGGAGCTGTTCT
ENaC-β	TCTTCCACCCGGATTATGGCAACT	ATGTCCAGGATCAACTTCAGGCC
ATP1A1	TGTGAAGAACTTGGAGGCTGTGGA	TGGCCGAAGTCTTGTCGAATGAGA

NHE, Na⁺/H⁺ exchanger; ENaC, epithelial sodium channel; ATP1A1, Na-K-ATPase α₁-subunit.

Table 2.

Human primer sequences for quantitative RT-PCR

Gene Name	Forward (5′-3′)	Reverse (5′-3′)
GAPDH	CAAGGTCATCCATGACAACTTTG	GGGCCATCCACAGTCTTCTG
NHE2	CCGATTTGGGAGTGAGAAG	CGCCCTCCAGAAGTATTTAG
NHE3	GAGACAAGGTCAAGGAGAAG	GAGAGGATGTGGTCGAAAG
PTGS2	GAAGCCTTCTCTAACCTCTC	GGATCAGGGATGAACTTTCT

NHE, Na⁺/H⁺ exchanger; ENaC, epithelial sodium channel; ATP1A1, Na-K-ATPase α₁-subunit; PTGS2, prostaglandin-endoperoxide synthase.

Western blot analysis of NHE3.

Protein was extracted from control and diarrheic pigs from the apex segment of the colon using the ProteoExtract transmembrane protein kit (71772-3; Novagen). Samples were boiled in 2× denaturing buffer (20% glycerol, 4% SDS, 125 mM Tris pH 6.8, and 0.3 mM bromophenol blue) containing 10% β-mercaptoethanol (M6250; Sigma-Aldrich) for 5 min and analyzed by 10% SDS-PAGE.

For Western blot analysis, proteins were electroblotted onto PVDF membrane (RPN303LFP; GE Healthcare Life Sciences) at 0.2 mA for 4 h at 4°C onto a PVDF membrane with transfer buffer (25 mM Tris, 192 mM glycine, and 20% methanol). Membranes were blocked for 1 h at room temperature with a 10% RapidBlock (M325-AMRESCO) blocking solution and subsequently probed overnight at 4°C with primary antibodies anti-NHE3 (ARP43870_P050; AVIVA Biosciences), verified by supplier to react with pig and anti-β-actin (C-4; sc-47778; Santa Cruz Biotechnology) in a 10% RapidBlock solution. The membranes then incubated for 1 h at room temperature with secondary antibodies Alexa Fluor 488 Goat anti-Rabbit IgG antibody (A-11008) and ECL Plex Goat anti-Mouse IgG-Cy5 antibody (PA45009; Amersham Biosciences) in 10% RapidBlock solution. Proteins were detected and analyzed using Typhoon Trio and ImageQuant TL System (63005583; GE Healthcare Life Sciences). The sizes of detected proteins were compared with a prestained rec protein ladder loaded on each blot (BP3603500; Fisher BioReagents).

Cell culture.

The Caco-2 cell line derived from human colorectal adenocarcinoma (HTB-37; ATCC, Manassas, VA) was cultured in DMEM (10-0130CM; Corning, Manassas, VA) containing 10% heat-inactivated fetal bovine serum (Gibco, Burlington, ON, Canada), 1% penicillin-streptomycin (15140-122; Life Technologies), and 1% MEM nonessential amino acids (Gibco, Grand Island, NY) at 37°C in a humidified atmosphere with 5% CO₂. Cells were plated on polyester Transwell permeable supports (0.4-µm pores, 24-mm diameter; Corning) and cultured under standard conditions until confluency. Cells were maintained for 10 days after confluency was achieved before being used in downstream experiments.

Exposure of Caco-2 monolayers to recombinant human IL-1α.

To determine if the single proinflammatory cytokine that was significantly upregulated throughout the colon of pigs infected with B. hyodysenteriae and B. hampsonii (16) was responsible for the decrease in NHE3 mRNA expression in vivo, polarized Caco-2 monolayers were exposed to human recombinant IL-1α (I2778; Sigma-Aldrich) and subsequently probed for changes in NHE2 and NHE3 mRNA expression by RT-qPCR (Tables 1 and 2). Both apical and basolateral surfaces were exposed to IL-1α at concentrations of 10, 100, and 500 ng/mL for 24 h. The Caco-2 cell line was chosen as no porcine colon-derived cell lines are commercially available. To ensure that IL-1α had a biological effect on Caco-2 monolayers, qPCR primers were developed for PTGS2, which has been previously shown to become upregulated in multiple cell types after IL-1α exposure (14, 25, 46).

Caco-2 monolayer exposure to B. hampsonii lysate.

To determine if B. hampsonii has a direct effect on modulating ion channel expression, a whole cell lysate was used to determine if the bacterium was capable of downregulating cystic fibrosis transmembrane conductance regulator (CFTR) mRNA expression in Caco-2 cells. The B. hampsonii lysate was prepared by centrifuging 50 mL of culture broth containing actively motile spirochete bacteria (10⁸/mL) at 10,000 g for 40 min, after which the supernatant was poured off (41). Bacterial pellets were resuspended in phosphate-buffered saline (10010-023, Life Technologies, Gibco) and vortexed until the pellet completely dissolved. This material was subsequently disrupted by sonication (Sonics & Materials, Danbury, CT) at 50% duty for 120 s at 4°C (41). Lysate total protein concentrations were determined using BCA protein assay (23225; ThermoFisher) using bovine serum albumin (BSA) as a standard. Polarized Caco-2 monolayers were exposed apically to the whole cell B. hampsonii lysate (0.005–50 µg/mL) for 48 h. Fifty micrograms per milliliters of total bacterial protein were determined to be ~10⁸ genomic equivalents/ml and considered a physiological concentration.

Western blot analysis of p38, ERK, JNK, and NF-κB signaling pathways.

Caco-2 monolayers were exposed to 50 µg/mL of whole cell B. hampsonii lysate (n = 4) or PBS (n = 4) for 48 h as described previously, after which point total soluble protein was isolated using M-PER mammalian protein extraction reagent (78505; ThermoFisher) containing 1× halt protease and phosphatase inhibitor cocktail (78446; Thermo Fisher). Lysate total protein concentrations were determined using the BCA protein assay (23225; Thermo Fisher) using BSA as a standard. The protein concentrations of all samples were normalized to 1,500 µg/mL, after which point the samples were separated by 10% SDS-PAGE and transferred onto a PVDF membrane as described previously. Membranes were blocked for 1 h with 10% RapidBlock (M325-AMRESCO) solution and probed with rabbit p-p38 T180/Y182 (4511S; Cell Signaling Technology), rabbit p-JNK T182/Y185 (9251S; Cell Signaling Technology), rabbit p-44/42 MAPK (p-ERK1/2) T202/Y254 (4377S; Cell Signaling Technology), or p-IκBα (2859L; Cell Signaling Technology) in 10% RapidBlock solution at 4°C overnight, washed three times with PBS + 0.1% Tween 20, probed with Alexa Fluor 488 Goat anti-Rabbit IgG antibody (A-11008; Thermo Fisher Scientific), and visualized using Typhoon Trio and ImageQuant TL System (63005583; GE Healthcare Life Sciences). The PVDF membranes were then stripped with 100 mM β-mercaptoethanol, 2% SDS, and 62.5 mM Tris·HCl pH 6.7 at 50°C for 30 min, washed two times with PBS + 0.1% Tween 20, blocked for 1 h in 10% RapidBlock solution with 10% RapidBlock (M325-AMRESCO) solution and probed with rabbit p38 (9212S; Cell Signaling Technology), rabbit JNK (9252S; Cell Signaling Technology), rabbit 44/42 MAPK (ERK1/2) (9102S; Cell Signaling Technology), or rabbit IκBα (4812S; Cell Signaling Technology) in 10% RapidBlock solution at 4°C overnight, washed three times with PBS + 0.1% Tween 20, probed with Alexa Fluor 488 Goat anti-Rabbit IgG antibody (A-11008; Thermo Fisher Scientific), and visualized using Typhoon Trio and ImageQuant TL System (63005583; GE Healthcare Life Sciences). The sizes of detected proteins were compared with a prestained rec protein ladder loaded on each blot (BP3603500; Fisher BioReagents).

Statistical analysis.

The electrogenic Ussing chamber and ²²Na Flux studies data were not normally distributed (Shapiro-Wilk test; P < 0.05) and are expressed as median ± interquartile range. Data obtained by RT-qPCR and Western blot were normally distributed and are expressed as means ± SE (Shapiro-Wilk test; P > 0.05). Change in I_sc between diseased and control colon segments was analyzed by Kruskal-Wallis one-way ANOVA. Changes in rate of ²²Na transport was analyzed by Kruskal-Wallis one-way ANOVA and Dunn’s test. Fold changes in mRNA expression of diseased colonic segments compared with control obtained by RT-qPCR were log₁₀ transformed and analyzed by one-way ANOVA and Tukey post hoc. Western blot analysis of NHE3 protein and cell signaling pathways were analyzed by Student’s t test (control vs. Brachyspira species in 2 separate analyses for NHE3 protein, control vs Brachyspira lysate treated for cell signaling pathways). Fold changes obtained by RT-qPCR for in vitro data were analyzed by Student’s t test. Significance was set a priori at P < 0.05.

RESULTS

B. hyodysenteriae and B. hampsonii Reduce Na⁺ absorption in the apex of the porcine spiral colon during ²²Na equilibration.

During the 80-min ²²Na equilibration, diseased colon tissues had a significantly decreased unidirectional J_ms flux both immediately after clamping the tissues (basal) and after 80 min (P < 0.05) (Table 3). Interestingly, at no point was there a significant difference in J_ms between bacterial strains. Additionally, J_sm ²²Na flux in B. hyodysenteriae-infected tissues was not significantly different than control either immediately after clamping or after 80 min (Table 3). However, a significant reduction in J_sm flux was observed in B. hampsonii-infected tissues when compared with control (P < 0.05) after 80-min of fluxing. J_net revealed that control colonic tissues exhibited net Na⁺ absorption during the equilibration period (Table 3). B. hyodysenteriae-diseased tissues had a net secretory response throughout the equilibration and was significantly different (P < 0.05) than control 80 min after the beginning of the equilibration period. B. hampsonii exhibited a different trend with a small rate of net Na⁺ absorption. However, this rate was significantly less (P < 0.05) than the net absorption observed in control tissues and was not significantly different than the net secretion noted in tissues infected with B. hyodysenteriae (Table 3).

Table 3.

Brachyspira hyodysenteriae and Brachyspira hampsonii decrease basal J_ms and net Na⁺ absorption during ²²Na equilibration in the apex of the porcine spiral colon

	JNa+, µeq·cm2·h
	M-S		S-M		Net		TEER, ohm·cm2		Potential Difference, mV
	Basal	80 Min	Basal	80 Min	Basal	80 Minutes	Basal	80 Min	Basal	80 Min
Control	17.38 ± 3.36a	59.32 ± 18.10a	14.83± 3.05a	40.94 ± 7.70a	2.59 ± 4.62a	17.01 ± 14.61a	60.62 ± 32.94a	58.82 ± 28.39a	0.1 ± 1.4a	0.0 ± 1.5a
B. hyo	10.14 ± 11.12b	37.15 ± 22.53b	10.23 ± 9.79a	40.62 ± 15.11a,b	−0.78 ± 4.85b	−0.78 ± 11.14b	58.16 ± 17.46a	52.38 ± 13.23b	1.0 ± 1.2b	0.8 ± 1.0b
B. hamp	11.50 ± 15.78b	31.95 ± 19.53b	8.91 ± 16.80a	30.86 ± 16.43c	1.15 ± 4.69a	0.65 ± 12.29b	55.40 ± 26.53a	49.56 ± 19.23b	1.0 ± 1.0b	0.7 ± 1.0b

Data are median ± interquartile range, analyzed using Kruskal-Wallis one-way ANOVA and Dunn’s post hoc; n = 12 ctrl, n = 12 B. hyodysenteriae, and n = 12 B. hampsonii. Basal mucosal-to-serosal flux (J_ms), serosal-to-mucosal flux (J_sm), and net flux rates during ²²Na equilibration across apex colonic segments from control, B. hyodysenteriae, and B. hampsonii-diseased pigs were measured immediately following mounting and clamping of the tissue (basal) and at the end of the 80-min equilibration period. M-S, unidirectional mucosal-to-serosal ²²Na flux; S-M, unidirectional serosal-to-mucosal ²²Na flux; Net, net ²²Na flux; TEER, transepithelial electric resistance. Significant differences within tissue groups across the basal and 80-min time points are in underlined. Significantly different values within each group are marked in bold. Each value marked with a different superscript letter differs by at least P < 0.05 compared with other groups at the same time points.

Overall, the unidirectional J_ms flux suggests that diseased tissues have a reduced absorptive capacity when compared with control tissues. J_sm flux did not account for the decrease in net Na⁺ absorption in diseased samples. Additionally, the decreased absorptive capacity did not appear to be due to loss of barrier function in Brachyspira-infected tissue. While the transepithelial resistances of B. hyodysenteriae and B. hampsonii-treated tissues were significantly lowered after 80 min of equilibration, no differences were found between different groups at the basal time point or within similar groups at the 80-min time point (Table 3). Additionally, the lower trans-epithelial resistances of B. hyodysenteriae and B. hampsonii-treated tissues was not associated with an increase in paracellular transport, as Na⁺ J_sm was either unchanged (B. hyodysenteriae) or lowered (B. hampsonii) in Brachyspira-infected tissues. Interestingly, both B. hyodysenteriae- and B. hampsonii-treated tissues exhibited an increased potential difference across the membrane when compared with control at both the basal and 80-min time points (Table 3), with the B. hyodysenteriae- and B. hampsonii-treated tissue potential differences significantly decreasing after 80 min. Thus Na⁺ absorption in the colon appears to be primarily due to electroneutral movement of Na⁺ and H⁺ through NHEs and electrogenic absorption via ENaC. The unidirectional J_ms suggests that either one or both of these routes of absorption have been negatively altered contributing to malabsorptive diarrhea in experimentally challenged pigs.

B. hyodysenteriae and B. hampsonii infections do not reduce electrogenic Na⁺ absorption in the porcine spiral colon.

The electrogenic Na⁺ absorptive capacity of the porcine spiral colon was measured in diseased and healthy colon segments in Ussing chambers to determine its contribution to the reduction in ²²Na J_ms flux. The addition of amiloride to the mucosa of proximal, apex, and distal colon segments revealed a numerically decreased trend (P > 0.05) in inhibitable I_sc throughout all segments of diarrheic pigs; however, no significant difference was noted when compared with control (Fig. 1). It was noted, however, that electrogenic inhibitable I_sc via amiloride was greater in the distal colonic segment compared with the proximal. This suggests that electrogenic Na⁺ absorption throughout the colon is not affected significantly in diseased animals, leading to the conclusion that the significant decrease in J_ms is due to loss of electroneutral Na⁺ transport in the spiral colon of diarrheic pigs.

Fig. 1.

Brachyspira hyodysenteriae and Brachyspira hampsonii do not affect electrogenic Na⁺ absorption through epithelial sodium channel (ENaC). Change in short-circuit current (I_sc) in response to the addition of ENaC channel inhibitor amiloride in the proximal, apex, and distal colonic segments of control and diseased pigs. Data are median ± interquartile range, analyzed using Kruskal-Wallis one-way ANOVA; n = 18 ctrl, n = 17 B. hyodysenteriae, and n = 16 B. hampsonii.

RT-qPCR analyses of Na⁺ channels and transporters.

The ²²Na flux experiments indicated that the unidirectional J_ms was significantly impaired in the apex of diarrheic pigs. To determine if these differences were attributed to altered Na⁺ channel and transporter mRNA expression, colonic mucosal samples were analyzed by RT-qPCR. In diseased pigs, downregulation of NHE isoforms 1–3 in the proximal, apex, and distal colonic segments of diarrheic pigs experimentally challenged with B. hyodysenteriae and B. hampsonii compared with controls further supports the decreased J_ms. NHE2 and NHE3 transcripts were significantly reduced compared with controls in all three segments of the colon (Table 4). These two transporters have been shown to drive electroneutral Na⁺ absorption and are involved in diarrheal disease (52).

Table 4.

Fold changes in Na⁺ channel and transporter mRNA expression in the porcine colon

	Segment of Porcine Colon
	Proximal			Apex			Distal
Ion Channel	Control	B. hyo	B. hamp	Control	B. hyo	B. hamp	Control	B. hyo	B. hamp
NHE1	2.15 ± 1.42a	0.27 ± 0.10b	0.33 ± 0.08a	1.02 ± 0.11a	0.44 ± 0.11b	0.47 ± 0.10b	1.76 ± 0.49a	0.51 ± 0.24b	0.38 ± 0.08b
NHE2	1.30 ± 0.47a	0.38 ± 0.12b	0.31 ± 0.06b	1.04 ± 0.13a	0.08 ± 0.01b	0.18 ± 0.09b	1.42 ± 0.38a	0.26 ± 0.03b	0.32 ± 0.11b
NHE3	1.16 ± 0.30a	0.11 ± 0.02b	0.24 ± 0.04b	1.17 ± 0.36a	0.36 ± 0.11b	0.28 ± 0.11b	1.33 ± 0.44a	0.29 ± 0.10b	0.14 ± 0.04b
ENaC-α	1.49 ± 0.61a	0.76 ± 0.29a	0.38 ± 0.13a	1.05 ± 0.17a	0.09 ± 0.04b	0.03 ± 0.01b	1.59 ± 0.67a	0.44 ± 0.10a	0.17 ± 0.05b
ENaC-β	1.45 ± 0.76a	0.26 ± 0.06a	2.32 ± 0.98a	1.21 ± 0.31a	0.47 ± 0.13a	1.42 ± 0.28a	1.61 ± 0.65a	1.44 ± 0.18a	2.66 ± 0.88a
ATP1A1	1.36 ± 0.48a	0.21 ± 0.03b	0.31 ± 0.06b	1.27 ± 0.45a	0.57 ± 0.16a	0.47 ± 0.09a	1.48 ± 0.44a	0.85 ± 0.12a	0.60 ± 0.10a

Data are means ± SE of fold difference in gene expression measured by quantitative RT-PCR; n = 6 ctrl, n = 6 B. hyodysenteriae, and n = 6 B. hampsonii. Fold changes in Na⁺ channel and transporter mRNA expression in 3 segments (proximal, apex, and distal) of the colon in B. hyodysenteriae, and B. hampsonii-diseased pigs compared with control are shown. NHE, Na⁺/H⁺ exchanger; ENaC, epithelial sodium channel; ATP1A1, Na-K-ATPase α₁-subunit. Significantly different values are marked in bold.

^a,bEach gene marked with a different superscript letter differs by at least P < 0.05 compared with control in each segment.

In contrast, NHE1 primarily localizes to the basolateral membrane of epithelial cells and plays a key role in regulation of cellular pH (8). NHE1 was downregulated (P < 0.05) in the proximal, apex and distal segments of the colon in diarrheic pigs infected with B. hyodysenteriae and in the apex and distal segments of pigs infected with B. hampsonii (Table 4). Loss of NHE1 mRNA and protein has been observed in colonic biopsies from patients with inflammatory bowel disease that has been suggested to be a contributing factor to the pathogenesis of the disease (29).

The electrogenic Na⁺ transporter ENaC had significantly downregulated mRNA expression of the channels α-subunit (P < 0.05) in the apex of diarrheic pigs infected with B. hyodysenteriae and in the apex and distal segments of pigs infected with B. hampsonii (Table 4). ENaC’s β-subunit was unchanged compared with control in all three segments of the colon (Table 4). Previous findings have suggested that ENaC can form a functional channel capable of transporting Na⁺ without all three subunits (22). Thus downregulation of ENaC’s α-subunit in diseased colonic samples may not result in a decrease in amiloride inhibitable I_sc.

Na⁺-K⁺-ATPase α₁-subunit is responsible for the transport of three Na⁺ out of the cell and two K⁺ in, maintaining the cells osmotic balance (34). Its reduction could impact the ²²Na flux. However, Na⁺-K⁺-ATPase α₁ expression was unchanged in all colonic segments infected with B. hyodysenteriae and B. hampsonii, except the proximal colon, which was downregulated (P < 0.05, Table 4). This finding indicates that the impaired ²²Na flux in the apex of diseased pigs is not due to changes in this transporter. Furthermore, the decrease in Na⁺-K⁺-ATPase α₁ did not significantly affect the electrogenic Na⁺ transport via ENaC, indicating that the decreased expression minimally contributed to the diarrhea.

Western blot analysis of NHE3.

Although all three NHE isoforms were downregulated, previous studies have indicated that the NHE3 isoform is critical for Na⁺ absorption in the gastrointestinal tract (53). Thus, Western blot for NHE3 was performed to support inhibitor and RT-qPCR results for this vital transporter. NHE3 channel protein was significantly decreased in the apex of diarrheic pigs infected with B. hyodysenteriae (P < 0.001) and B. hampsonii (P < 0.001) compared with control (Fig. 2). This decrease in channel protein correlates with the decrease in mRNA expression of NHE3 and impairment of Na⁺ absorption in the colon of Brachyspira-infected pigs.

Fig. 2.

Brachyspira hyodysenteriae and Brachyspira hampsonii significantly reduce Na⁺/H⁺ exchanger 3 (NHE3) protein expression in the apex of the porcine colon. Western blot (A) and densitometry (B) of NHE3 (~93 kDa) compared with β-actin (~43 kDa) reference for B. hyodysenteriae and B. hampsonii-infected apex porcine colon compared with control. Sizes of molecular weight markers are indicated. Control and B. hyodysenteriae/control and B. hampsonii bands were taken from consecutive lanes on the same blot as indicated. Data are means ± SE, analyzed using Student’s t test. *P < 0.05; n = 3 ctrl, n = 3 B. hyodysenteriae, and n = 3 B. hampsonii.

IL-1α does not decrease NHE3 mRNA expression in polarized Caco-2 monolayers.

To investigate whether the upregulation of IL-1α mRNA expression throughout the colon of diarrheic pigs (16) was responsible for the decreased expression of NHE3 mRNA, polarized Caco-2 monolayers were exposed to varying concentrations (10, 100, and 500 ng/ml) of human recombinant IL-1α for 24 h. RT-qPCR revealed that IL-1α had no effect on modulating NHE3 gene expression at any concentration when compared with control (Table 5). To ensure that IL-1α was biologically active, PTGS2 mRNA expression was assessed. IL-1α exposure has been shown to result in elevated PTGS2 mRNA expression in several cell types (14, 25, 46). We found that all monolayers independent of IL-1α concentration had a significant twofold increase in PTGS2 mRNA expression (P < 0.05, Table 5).

Table 5.

Fold changes in NHE3 and PTGS2 mRNA expression in polarized Caco-2 cell monolayers after 24-h exposure to IL-1α

		Concentration of Recombinant Human IL-1α
Gene Name	Control	10 ng/ml	100 ng/ml	500 ng/ml
NHE3	1.07 ± 0.18a	1.48 ± 0.32a	1.60 ± 0.49a	1.28 ± 0.29a
PTGS2	1.00 ± 0.05a	2.00 ± 0.24b	2.38 ± 0.34b	2.02 ± 0.20b

Data are means ± SE of fold differences in gene expression measured by quantitative RT-PCR; n = 6. NHE3, Na⁺/H⁺ exchanger; PTGS2, prostaglandin-endoperoxide synthase 2. Significantly different values are marked in bold.

^a,bEach gene marked with a different superscript differs by at least P < 0.05 compared with control.

B. hampsonii lysate decreases NHE3 mRNA expression and induces p38 phosphorylation in polarized Caco-2 monolayers.

As IL-1α had no effect on NHE3 mRNA expression, we determined if a Brachyspira lysate was capable of directly altering NHE3 mRNA expression. Polarized Caco-2 monolayers were exposed to varying concentrations of lysate (0.005–50 µg/ml) for 48 h. Our results revealed that a lysate concentration of 50 µg/ml was capable of significantly downregulating NHE3 expression (P = 0.009) compared with control (Fig. 3). Our previous findings revealed that this Brachyspira lysate significantly upregulated IL-1α in exposed monolayers, recapitulating other cellular events observed in vivo (16). To further examine the mechanism by which this B. hampsonii lysate inhibited NHE3 transcription polarized Caco-2 monolayers were then exposed to 50 µg/mL lysate for 48 h, after which point total soluble protein was extracted and the activation of MAPK and NF-κB signaling pathways was assessed via Western blotting. These data showed both a significant increase in the level of p38 phosphorylation upon treatment with Brachyspira lysate (P = 0.036, Fig. 4A). Additionally, high amounts of phosphorylated ERK1/2 were detected in both Brachyspira lysate-treated and control groups; however, no significant differences in phosphorylation were observed between groups (P = 0.461, Fig. 4B). While endogenous JNK and IκBα were detected in Caco-2 monolayers, we were unable to detect the phosphorylated forms of these proteins (data not shown). Together these findings provide evidence that loss of NHE3 mRNA transcripts is likely due to a direct effect of the bacteria and suggests a potential role of p38 MAPK activation in this transcriptional repression.

Fig. 3.

Brachyspira hampsonii lysate downregulates Na⁺/H⁺ exchanger 3 (NHE3) mRNA expression in Caco-2 monolayers. Fold change in NHE3 mRNA expression compared with control as measured by quantitative RT-PCR after exposure to a Brachyspira hampsonii whole cell lysate at 5 concentrations (0.005–50 μg/ml) for a period of 48 h. Data are means ± SE, analyzed using one-way ANOVA. *P < 0.05; n = 6.

Fig. 4.

Brachyspira Hampsonii lysate increases p38 MAPK phosphorylation and has no effect on ERK1/2 activation in Caco-2 monolayers. Western blot (A) and densitometry (B) of p-p38 (~38 kDa) compared with total p38 (~38 kDa) reference for Brachyspira hampsonii lysate-treated Caco-2 monolayers compared with control. Sizes of molecular mass markers are indicated. Lysate and control bands were taken from 2 nonconsecutive lanes on the same blot. Western blot (C) and densitometry (D) of p-ERK1/2 (~44/42 kDa) compared with total ERK1/2 (~44/42 kDa) reference for Brachyspira hampsonii lysate-treated Caco-2 monolayers compared with control. Sizes of molecular mass markers are indicated. Lysate and control bands were taken from two consecutive lanes on the same blot. Data are means ± SE, analyzed using Student’s t test. *P < 0.05; n = 4 control for all proteins tested; n = 4 Brachyspira lysate-treated for all proteins tested.

DISCUSSION

Here we identify colonic electroneutral Na⁺ absorption and associated genes as the cause of spirochete-induced diarrheal disease caused by B. hyodysenteriae and B. hampsonii. Diarrhea in pigs caused by B. hyodysenteriae has been previously described as malabsorptive in which Na⁺ and Cl⁻ absorption is abolished in diarrheic animals (4, 51). An increase in luminal anion secretion does not contribute to the development of diarrheal disease (4, 16, 51). However, a decrease in agonist induced anion secretion in diarrheic pigs experimentally challenged with B. hyodysenteriae and B. hampsonii throughout the colon, due to downregulation of anion channel mRNA and protein, is thought to modify mucin properties and aid in the colonization and development of spirochetosis (16). Furthermore, colonic permeability to mannitol and PEG-400 in the porcine colon remained unchanged, and transepithelial electrical resistances were not different from control, suggesting that tight junction integrity is not compromised (4, 16). The pathophysiological mechanism of transporter and transporter regulation during spirochetosis causing malabsorptive diarrhea has not been characterized in any Brachyspira spp.

Impairment of electroneutral Na⁺ absorption contributes to diarrheal development.

The current study provides strong evidence supporting the loss of the Na⁺ absorptive capacity of the colon in diarrheic pigs infected with B. hyodysenteriae and B. hampsonii. The ²²Na flux equilibration period further supports previous findings, by which J_ms is severely impaired in the colon of diseased pigs (4, 51). Inhibition of electrogenic Na⁺ transport via ENaC in Ussing chambers revealed that there was no significant difference in inhibitable current between control and diseased tissues. Therefore, these findings point to impairment of electroneutral Na⁺ transport as the driving force responsible for the production of diarrhea in Brachyspira-diseased pigs.

Gene transcripts of colonic samples taken from diarrheic pigs revealed that electroneutral Na⁺ transporters NHE isoforms 1–3 were significantly downregulated compared with control in all three segments of the colon. Loss of NHE3 in a complete Slc9a3^−/− mouse model resulted in mild diarrhea, enlargement of the gastrointestinal tract, increase in GI-contents, and decrease in pH in the colon relative to the cecum (52). Interestingly, NHE2 knockout (Slc9a2^−/−) mice showed no changes in acid-base homeostasis or disruptions in electrolyte absorption suggesting that the NHE3 isoform is of greater importance in the gastrointestinal tract (52). Significant downregulation of NHE3 protein expression in the apex of diarrheic pigs further supports our conclusion that impairment of electroneutral Na⁺ absorption is responsible for the development of diarrhea.

During diarrheal episodes, chronic stimulation by mineralocorticoids elicits the expression of ENaC throughout the colon and into the ileum to help mitigate fluid and electrolyte loss (52, 58). This compensatory mechanism appears not to be present in diarrheic pigs suffering from Brachyspira colitis. Interestingly, the α-subunit mRNA of ENaC was significantly downregulated in some segments of the colon in diseased pigs while the β-subunit was unchanged from control. Previous studies have suggested that not all subunits are required to form a functional channel supporting the transport of Na⁺ (22). The current electrogenic Ussing chamber studies lead to the conclusion that Na⁺ absorption is not significantly affected in the colon of diarrheic pigs. It is possible, however, that Na⁺ absorption through ENaC accounts for the majority of Na⁺ absorption in diseased pigs when electroneutral Na⁺ absorption is impaired.

Na⁺-K⁺-ATPase’s α-subunit was downregulated in the proximal segment of the colon in B. hyodysenteriae- and B. hampsonii-diseased pigs. Na⁺-K⁺-ATPase is essential for maintaining the electrolyte gradient across the cellular membrane, regulating osmotic balance and cell volume (7, 28). Malabsorption of bile acids in the colon associated with chronic diarrhea in human patients has been shown to cause downregulation of Na⁺-K⁺-ATPase’s subunits (9). This was further supported with T84 cells exposed to 150 µM deoxycholic acid for 48 h, which downregulated both ATP1A1 and ATP1B1 at both the mRNA and protein level in vitro (9). Since Na⁺-K⁺-ATPase is essential for driving ion transport within enterocytes, loss of its function would be a main contributor to the production of diarrhea observed in animals infected with Brachyspira species. However, it was not downregulated in the apex of the colon and is, therefore, unlikely to be responsible for the decrease in Na⁺ absorption observed herein. Additionally, the decrease in mRNA expression in the proximal colon did not have a significant effect on electrogenic Na⁺ absorption in this segment. Together these findings suggest that downregulation of ATP1A1 in the proximal segment is of minimal importance during spirochetosis.

IL-1α does not decrease NHE3 mRNA expression in polarized Caco-2 monolayers.

Cytokines have been previously shown to have strong negative regulation of Na⁺ channel and transporters (2, 47, 60). Kruse et al. (31) reported that pigs experimentally challenged with B. hyodysenteriae had elevated levels of IL-1β and TNF-α in their blood during peak dysentery while IFN-γ was not detected. Furthermore, colonic explants exposed to live B. hyodysenteriae had elevated IL-1α expression after 8 h of exposure (57). Our previous findings have shown that diarrheic pigs challenged with B. hyodysenteriae or B. hampsonii had significantly elevated mRNA expression of proinflammatory cytokine IL-1α throughout their colons while TNF-α and IFN-γ were downregulated or unchanged compared with control (16).

The majority of previous research on cytokine modification of electroneutral Na⁺ transporters has focused on proinflammatory cytokines IFN-γ, TNF-α, and IL-1β. This is likely due to the elevated expression of these proinflammatory cytokines in patients suffering from inflammatory bowel disease (5, 17, 37, 40). IFN-γ has been shown to decrease NHE2 and NHE3 mRNA and protein expression in Caco-2 BBE cells and rat ileum and colon in a dose- and time-dependent manner (2, 47). Additionally, TNF-α also decreased NHE3 gene expression in C2BBe1 cells (2). Both IFN-γ and TNF-α repress the NHE3 gene by phosphorylation of Sp1 and Sp3 transcription factors by a cAMP-dependent protein kinase (2).

A recent study has shown that IL-1β strongly reduces mRNA and protein expression of PDZK₁ (NHERF3) in Caco-2BBE cells (35). Reduced expression of PDZK₁ has been identified in the inflamed intestine of both ulcerative colitis patients and murine colitis models (33, 59). Absence of PDZK₁ leads to dysfunction of NHE3; however, it is not a result of decreased mRNA or protein expression but rather an increase in protein turnover due to reduced membrane retention time (15, 24). Other studies have shown that NHERF1- and NHERF2-deficient mice have reduced NHE3 activity and transporter abundance on the brush border membrane of jejunum and colon while mRNA levels remain unchanged (10, 13).

These three cytokines do not account for the decrease in NHE2 and NHE3 mRNA and protein expression in the colon of Brachyspira-diseased pigs. As IL-1α was the only cytokine upregulated throughout the colon of diarrheic pigs we tested whether IL-1α was capable of downregulating NHE3 mRNA expression in polarized Caco-2 monolayers (16). We found that recombinant IL-1α did not downregulate NHE3 in Caco-2 cells at any concentration (10, 100, and 500 ng/ml) after 24 h of exposure. However, the elevated mRNA expression of PTGS2 confirmed that IL-1α had a biological effect (14, 25, 46).

Previous studies have found that basolateral stimulation of rabbit ileum with human IL-1α decreases both Na⁺ and Cl⁻ absorption while stimulating Cl⁻ secretion (14). These effects were mirrored closely when ileal samples were stimulated with PGE₁; however, when ileal samples were first treated with IL-1α and subsequently exposed to PGE₁, the magnitude of change in net flux and electrical properties was reduced (14). These findings suggested that the effects of IL-1α and PGE₁ are not additive in nature (14). Additionally, in another study assessing enterocyte-subepithelial myofibroblast interaction exposure, 18Co or P2JF cells preincubated with IL-1α and grown acutely juxtaposed to T84 cells caused significantly elevated basal I_sc (25). This effect was attributed to upregulation of PTGS1 and PTGS2 expression in 18Co and P2JF cell lines and production of PGE₂ (25). Additional studies have identified that prostanoids stimulate Cl⁻ secretion and inhibit electroneutral Na⁺ absorption by elevating cAMP (3, 42, 61). However, in the colon of diarrheic pigs infected with Brachyspira, anion secretion is impaired, and there is a reduction in NHE3 mRNA and protein expression. The effects of prostanoids are receptor mediated and do not directly affect ion channel mRNA and protein expression (38), suggesting that prostanoid induction by IL-1α is not responsible for the observed decrease in Na⁺ absorption. Thus the reduction in Na⁺ channel mRNA and protein in vivo, the inability of IL-1α to downregulate NHE3 in Caco-2 cells, and previous literature suggests that Brachyspira are capable of modulating Na⁺ transporter transcriptional processes.

B. hampsonii lysate downregulates NHE3 mRNA expression and induces p38 phosphorylation in Caco-2 cells.

As IL-1α was not capable of downregulating NHE3 mRNA in Caco-2 cells, a whole cell B. hampsonii lysate was used to determine its effect on NHE3 gene expression. Exposure of polarized Caco-2 monolayers to a lysate concentration of 50 µg/ml caused a significant decrease in NHE3 mRNA. This suggests that a component of the bacteria is responsible for the decrease in transporter mRNA expression. This is not the first time that destruction of Brachyspira results in modification of cellular processes. Live B. hyodysenteriae treated with neutrophil elastase, a serine protease, caused elevated mucin production and transport rates in HT29 MTX-E12 cells (45). Furthermore, we have previously shown that a B. hampsonii lysate decreased CFTR mRNA expression in polarized Caco-2 monolayers (16). Additionally, Western blot analyses of Caco-2 monolayers treated with B. hampsonii lysate indicated that activation of the p38 MAPK signaling pathway accompanied the aforementioned decrease in NHE3 transcription. Previous studies have indicated that IFN-γ and TNF-α are capable of inhibiting NHE3 transcription in intestinal cells (2, 47), with this inhibition being mediated through phosphorylation of the transcription factors SP1 and SP3 by PKA (2). Given that p38 MAPK is also known to phosphorylate SP1 (62), these data therefore suggest a potential signaling mechanism by which the Brachyspira lysate decreases NHE3 mRNA levels. Thus we provide strong evidence supporting the direct effects of Brachyspira spp. that contribute to diarrheal development in the colon of infected pigs.

Conclusion.

The current study provides new insight elucidating the pathophysiological mechanisms of B. hyodysenteriae and B. hampsonii infections and how they contribute to the development of diarrheal disease. Impairment of the electroneutral Na⁺ absorptive capacity of the porcine colon is a major contributor to the development of diarrhea. This decrease in Na⁺ absorption is supported by the downregulation of NHE isoforms 1–3 mRNA and reduced protein expression of the critical NHE3 isoform in the colon of diseased pigs. The host’s cytokine response was determined not to be responsible for the decrease in NHE3 mRNA expression as human recombinant IL-1α had no effect on modulating NHE3 expression in Caco-2 monolayers. A B. hampsonii lysate, however, decreased NHE3 mRNA expression and induced p38 MAPK phosphorylation in Caco-2 monolayers providing further evidence that these Brachyspira spp. are able to impair ion channel and transporter transcriptional processes. Together, these findings provide new evidence supporting a decrease in electroneutral Na⁺ absorption contributing to diarrheal development during episodes of spirochetosis.

GRANTS

This research was supported by Alberta Livestock Meat Association (ALMA 2013R054R) and Natural Sciences and Engineering Research Council of Canada Discovery Grant 371364-2010 (to M. E. Loewen).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.B.E., B.A.K., J.C.H., and M.E.L. conceived and designed research; C.B.E., B.A.K., N.C., J.C.H., and M.E.L. performed experiments; C.B.E., B.A.K., and M.E.L. analyzed data; C.B.E., B.A.K., and M.E.L. interpreted results of experiments; C.B.E. and B.A.K. prepared figures; C.B.E. and B.A.K. drafted manuscript; C.B.E., B.A.K., J.C.H., and M.E.L. edited and revised manuscript; C.B.E., B.A.K., N.C., J.C.H., and M.E.L. approved final version of manuscript.