Acute infection with Brachyspira hyodysenteriae affects mucin expression, glycosylation, and fecal MUC5AC

Abstract

Introduction: Infection with strongly β-hemolytic strains of Brachyspira hyodysenteriae leads to swine dysentery (SD), a production-limiting disease that causes mucohemorrhagic diarrhea and typhlocolitis in pigs. This pathogen has strong chemotactic activity toward mucin, and infected pigs often have a disorganized mucus layer and marked de novo expression of MUC5AC, which is not constitutively expressed in the colon. It has been shown that fucose is chemoattractant for B. hyodysenteriae, and a highly fermentable fiber diet can mitigate and delay the onset of SD.

Methods: We used lectins targeting sialic acids in α-2,6 or α-2,3 linkages, N-acetylglucosamine (GlcNAc), α-linked L-fucose, and an immunohistochemical stain targeting N-glycolylneuraminic acid (NeuGc) to investigate the local expression of these mucin glycans in colonic tissues of pigs with acute SD. We used a commercial enzyme-linked immunosorbent assay (ELISA) to quantify fecal MUC5AC in infected pigs and assess its potential as a diagnostic monitoring tool and RNA in situ hybridization to detect IL-17A in the colonic mucosa.

Results: Colonic mucin glycosylation during SD has an overall increase in fucose, a spatially different distribution of GlcNAc with more expression within the crypt lumens of the upper colonic mucosa, and decreased expression or a decreased trend of sialic acids in α-2,6 or α-2,3 linkages, and NeuGc compared to the controls. The degree of increased fucosylation was less in the colonic mucosa of pigs with SD and fed the highly fermentable fiber diet. There was a significant increase in MUC5AC in fecal and colonic samples of pigs with SD at the endpoint compared to the controls, but the predictive value for disease progression was limited.

Discussion: Fucosylation and the impact of dietary fiber may play important roles in the pathogenesis of SD. The lack of predictive value for fecal MUC5AC quantification by ELISA is possibly due to the presence of other non-colonic sources of MUC5AC in the feces. The moderate correlation between IL-17A, neutrophils and MUC5AC confirms its immunoregulatory and mucin stimulatory role. Our study characterizes local alteration of mucin glycosylation in the colonic mucosa of pigs with SD after B. hyodysenteriae infection and may provide insight into host-pathogen interaction.

Keywords: Brachyspira; GlcNAc; MUC2; MUC5AC; fucose; glycosylation; sialic acid; swine dysentery.

Figure 1

Study design. ELISA to detect MUC5AC detection in feces was performed on samples obtained from DPI-7, and on samples obtained every other day from DPI 4 to the day of euthanasia or at the end of the study on DPI 16. Formalin-fixed, paraffin-embedded tissue samples from the apex of the spiral colon were used for IHC, lectin stain, and RNA-ISH.

Figure 2

Staining intensity of LTL targeting α-linked L-fucose was significantly different between the three groups (A, B) and was low in the control (C). Pigs from the Bhyo group (D, P < 0.05) had higher intensity of LTL throughout the full thickness of the mucosa compared to the controls (C), and the Bhyo-RS group had a higher staining intensity tendency in the lower mucosa (A, P= 0.065) and significantly higher staining intensity in the upper colonic mucosa compared to the controls (B, E P < 0.05). Bar = 200 µm. **, P ≤ 0.01; ****, P ≤ 0.0001; ns, not significant.

Figure 3

Staining intensity of GS-II targeting GlcNAc was significantly different between the three groups, both in the lower and the upper half of the colonic mucosa (A, B, P < 0.05). The positivity of GS- II targeting GlcNAc was more confined to the goblet cells in the base of the colonic mucosa in the controls (C, P < 0.05), while there was abundant GS-II staining within the crypts and in the upper half of the colonic mucosa in the Bhyo and Bhyo-RS (D) groups compared to the controls (P < 0.05). Staining intensity of SNA targeting NeuAc attached to terminal galactose in an α-2,6 linkage was significantly different between the three groups, both in the lower and the upper half of the colonic mucosa (E, F, P < 0.05). There was marked positivity in the controls, which was evenly distributed in the goblet cells throughout the full thickness of the mucosa (G). In the Bhyo and Bhyo-RS (H) group, there was little to no intensity of α-2,6 linked NeuAc in the lower colonic mucosa (E) and little to moderate positivity in the upper half of the mucosa (F, H). Staining intensity of MAL II targeting NeuAc in an α-2,3 linkage was significantly different between the three groups in the lower half (I, P < 0.05) but not in the upper half of the colonic mucosa (J). Positivity was abundant and well confined to the goblet cells in the controls throughout the full thickness of the colonic mucosa (K), while the Bhyo and Bhyo-RS (L) groups had staining intensity mostly in the upper half of the mucosa that was more abundant in the crypt lumens. Bar = 200 µm. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ns, not significant.

Figure 4

No significant difference in the total expression of NeuGc was detected between the three groups (A). Controls slightly tend to show abundant positive staining within goblet cells and epithelial cytoplasm (B), whereas positivity in pigs infected with B.hyodysenteriae was often presented within crypt lumens and less in the cytoplasm (C). MUC5AC expression was significantly different between the three groups (D, P < 0.05). There was a marked de novo expression of MUC5AC in the Bhyo and Bhyo-RS (F) groups compared to the controls (E, P < 0.05). Bar = 200 µm. ***, P ≤ 0.001; ns, not significant.

Figure 5

MUC2 expression was significantly different between the three groups, both in the lower and the upper half of the colonic mucosa (A, B), P < 0.05). There was a moderate correlation between the expression of MUC2 and MAL II (C), lower half, R=0. 5855, P < 0.05, and upper half, (D), R=0. 5469, P < 0.05). There was more MUC2 expression in the lower half of the colonic mucosa in the controls (E), while the Bhyo (F) and Bhyo-RS groups had more expression in the upper halves (P < 0.05). Bar = 200 µm. **, P ≤ 0.01; ***, P ≤ 0.001; ns, not significant.

Figure 6

IL-17A expression was significantly different between the three groups (P < 0.05). The post hoc Dunn’s test showed (A, P=0.057) showed that the Bhyo-RS group (E) had a tendency of higher IL-17A expression compared to the controls (D). A moderate correlation was found between neutrophil count and IL-17A expression (B, R=0. 6841, P < 0.05), and there was a moderate correlation between IL-17A and immunohistochemical MUC5AC expression (C, R=0. 4727, P < 0.05). Bar = 50 µm. ns, not significant.

Figure 7

Significant differences in fecal MUC5AC concentration measured by ELISA was found on DPI 6 (A, P < 0.05, ANOVA) and DPI 10 (A, P < 0.05, Kruskal-Wallis). On DPI 12 (P=0.055, Kruskal-Wallis) and DPI 14 (P = 0.051, Kruskal-Wallis), there was a trend of an overall difference between groups. The two inoculated groups had significantly higher fecal MUC5AC concentrations than the controls on DPI 10 (A, P < 0.05, Dunn’s). When comparing the endpoint fecal MUC5AC and colonic MUC5AC concentrations, the two inoculated groups have significantly increased endpoint fecal MUCAC and colonic MUC5AC concentrations (B, C, P < 0.05, Tukey’s) compared to the controls. Fecal MUC5AC concentrations from the Bhyo-RS group were higher than the Bhyo group (B, P < 0.05, Tukey’s). There was weak to no correlation found between fecal MUC5AC (ELISA), colonic MUC5AC (ELISA) concentrations, and immunohistochemical MUC5AC expression on the colonic tissues. *, P ≤ 0.05; **, P ≤ 0.01; ****, P ≤ 0.0001; ns, not significant.