Clinical and postmortem diagnosis of CCD
After inspection of the randomly chosen pens (n = 10), 20 pigs were assessed as ‘diarrheic’ and 10 pigs as ‘clinically healthy controls’. Diarrheic pigs had watery and lose diarrhea with shiny mucus on the stool and they were pronounced diarrheic if fecal DM was < 18% (Fig. 1A). The NoDiar group had significantly higher dry matter content of stool compared to diarrheal groups (24.1 vs. 12.5%). Histological examination failed for 9 pigs; hence all the results are from 21 animals since we used histology as the benchmark of our diagnosis. Based on our histological results, pigs were classified as healthy controls (Fig. 1B) without clinical and postmortem signs of diarrhea and inflammation (NoDiar; n = 5), diarrheal without colonic inflammation (DiarNoInfl; n = 4), and diarrheal with inflammation (DiarInfl, Fig. 1C; n = 12). None of the tested pigs showed shedding of specific pathogens, e.g. B. hyodysenteria, B. pilosicoli and L. intracellularis, in the stool.
Alpha Diversity
Figure 2A-C show alpha diversity metrics for different samples from digesta vs. mucus (Fig. 2A) and in different diarrheal groups for digesta (Fig. 2B) and mucus (Fig. 2C). Except for FaithPD, all alpha diversity metrics showed to be different between digesta and mucus; higher in digesta vs. mucus. In digesta, DiarNoInfl showed the lowest values (P < 0.05) for Chao1, Shannon and FaithPD. The same pattern was observed for mucosal samples except for Shannon, which was constant for all groups. Gender and segment had no effect on alpha diversity indices in digesta and mucus; therefore, these samples obtained from Co2 and Co3 were considered similar.
Beta Diversity
Differences in bacterial composition between digesta and mucus, and between different groups are shown in Fig. 2. Regardless of diarrheal status, beta diversity based on Bray-Curtis dissimilarity, derived from a Principal Coordinate Analysis (PCoA) showed to be different between digesta and mucus (Fig. 3A), as confirmed by a graph-based analysis (Fig. 3B). The graph shows that samples from digesta formed solid edges together and mixed with mucosal samples, while mucosal samples formed solid edges only together, indicating differences in distribution of data originating from the two sample types (P < 0.01). The results of dbRDA showed that in digesta, there was a significant difference (R2 = 0.15, P < 0.01) between groups based on Bray-Curtis dissimilarity (Fig. 3C) and that the three groups formed separate clusters on ordination plots. In mucus, diarrheal status also had significant influence on beta diversity (Fig. 3D; R2 = 0.10, P = 0.02). There was no significant difference for beta diversity between Co2 and Co3 in both digesta and mucosal samples. Gender showed no effect on beta diversity for digesta samples; however, gender did influence beta diversity in mucosal samples. To verify the validity of dbRDA model, a test was performed on dispersion of variance around the centroids for sample type and for groups in digesta and mucus separately (Fig. S1A-C). Samples from mucus and digesta were different within group variances (Fig. S1A; P < 0.01); therefore, the dataset was split into digesta and mucus, which showed to be variance homoscedastic (P > 0.05) according to the group (Fig. S1B-C).
Bacterial Composition And Differential Abundance
The relative abundances of different phyla were different for digesta compared to mucus and there were also differences between groups (Fig. 4A-B). Numeric relative abundance of Proteobacteria was higher for DiarNoInfl and DiarInfl in both digesta and mucus, when compared to the NoDiar group. Regardless of diarrheal status, the relative abundance of Actinobacteriota, Planctomycetota, Patescibacteria, Firmicutes, and Bacteroidota was higher in digesta compared to mucus, while it was higher for Verrucumicrobiota, Campilobacterota, Deferribacterota, and Spirochaetota in mucosal samples vs. digesta samples (Fig. 4C).
At genus level, six genera, four belonging to Firmicutes, decreased in abundance when moving from digesta to mucus, while abundance of 24 genera increased in mucus compared to digesta, with the magnitude of this increase being observed for genera Acetivibrio ethanolgignens group and Brachyspira, 10.7 and 8.57 LFC, respectively (Fig. 4D).
In digesta from DiarNoInfl group, the abundance of Fibrobacterota and Cyanobacteria phyla decreased, while it increased for Proteobacteria, when compared to the NoDiar group (Fig. 5A). The DiarInfl, compared to NoDiar group, showed increased abundance of Proteobacteria as well as Spirochaetota, while it had reduced LFC for Actinobacteriota, Cyanobacteria and Firmicutes (Fig. 5B). Comparing DiarInfl with DiarNoInfl group revealed that the former had higher abundance of Spirochaetota (2.0 LFC) and Fibrobacterota and lower in Proteobacteria and Verrucumicrobiota (Fig. S2A). Digesta from the DiarNoInfl compared to the NoDiar group, showed to have 18 genera reduced in abundance (mainly belong to Firmicutes), such as F082 group, Fibrobacter, and Mailhella; and three increased in abundance including Bifidobacterium, T34, and Turicibacter (Fig. 5C). As for DiarInfl, four Firmicutes genera were reduced, e.g. Syntrophococcus and Shuttleworthia, and 11 genera increased in abundance, including Tyzzerella, Bifidiobacterium, Escherichia-Shigella and Helicobacter, when compared to the NoDiar group (Fig. 5D). Comparison of digesta between the two diarrheal groups showed that DiarInfl increased the abundance of 27 genera (chiefly from Firmicutes and Spirochaetota), compared with DiarNoInlf, and it decreased the abundance of six genera belonging to Firmicutes (Fig. S2C).
In mucus of DiarNoInfl pigs, Fibrobacterota (LFC = -7.0) and Cyanobacteria (LFC = -1.50) phyla were reduced and Proteobacteria (LFC = 2.90) increased in abundance (FDR < 0.05) compared with the NoDiar group (Fig. 6A). DiarInfl vs. NoDiar only resulted in increased abundance of Proteobacteria with LFC = 1.80 (Fig. 6B). At the genus level, DiarNoInfl vs. NoDiar showed reduced (FDR < 0.05) abundance of 20 genera, mainly belonging to Firmicutes, Spirochaetota, and Fibrobacterota and increased abundance of four genera in mucus, such as T34 (Fig. 6C). The DiarInfl group showed to have lower abundance of five genera (e.g., Lawsonia, Syntrophococcus and Shuttleworthia) and higher abundance of 10 mucosal genera, compared with NoDiar (Fig. 6D). In mucosal samples, comparison between DiarInfl and DiarNoInfl showed that DiarInf had lower abundance of Lawsonia (from Desulfobacterota phylum; Fig. 2SB) and higher abundance of Sphaerochaeta (belonging to Spirochaetota; Fig. S2D).
Microbial Fermentation Products
Table 1 shows the pH and concentration (mmol/kg digesta) of SCFA in digesta recovered from Co2 and Co3 of the different groups. Among all SCFAs, only butyrate was affected by gender, it was lower in females than in males; 12.4 vs. 8.80 mmol/kg digesta, respectively. Sample type had no effect on pH and SCFA concentration; however, since the estimate for sample type was not ignorable relative to other factors, the results are presented for both Co2 and Co3. Concentration of total SCFA and pH of digesta were not different among groups, while the concentration of individual SCFA such as butyrate, valerate and iso-acids differed between groups. Compared to NoDiar, DiarInfl showed on average 36.4% less butyrate concentration in Co2; and in Co3, DiarNoInfl and DiarInfl had on average 41.2% less butyrate compared with NoDiar. Valerate, and iso-acids were also lower in two groups with diarrhea, compared with NoDiar group.
Table 1
Digesta pH and concentration of SCFA (mmol/kg wet sample) in Co2 and Co3.
|
Groups1
|
|
NoDiar
|
DiarNoInfl
|
DiarInfl
|
Co2
|
|
|
|
pH
|
6.0 (5.75–6.36)
|
6.0 (5.64–6.31)
|
6.20 (6.02–6.48)
|
SCFA2
|
118 (96.9–145)
|
112 (91.1–137)
|
111 (93.9–131)
|
Butyrate
|
15.1 (10.0-22.7) b
|
10.0 (6.55–15.3) ab
|
9.60 (6.86–13.5) a
|
Propionate
|
29.2 (22.6–37.8)
|
28.2 (21.7–36.7)
|
28.0 (22.5–34.9)
|
Acetate
|
69.1 (58.5–81.7)
|
69.1 (58.4–81.8)
|
70.9 (63.0-80.3)
|
Valerate
|
3.50 (1.80–6.81)
|
3.10 (1.53–6.18)
|
2.1 (1.17–3.79)
|
Iso-acids
|
1.70 (1-2.8) b
|
0.70 (0.42–1.22) a
|
1.0 (0.63–1.58) a
|
Co3
|
|
|
|
pH
|
6.40 (6.04–6.69)
|
6.30 (5.90–6.74)
|
6.40 (6.18–6.61)
|
SCFA
|
108 (88.0-131.5)
|
96.3 (76.6–121)
|
101 (85.6–119)
|
Butyrate
|
13.6 (9.03–20.6) b
|
7.40 (4.60–11.9) a
|
8.60 (6.19-12.0) a
|
Propionate
|
23.6 (18.2–30.5)
|
22.0 (16.4–29.5)
|
23.8 (19.2–29.5)
|
Acetate
|
65.0 (54.9–77.0)
|
63.6 (52.1–77.6)
|
66.2 (58.6–74.8)
|
Valerate
|
3.20 (1.66–6.32) b
|
1.70 (0.79–3.70) ab
|
1.80 (1.02–3.27) a
|
Iso-acids
|
2.30 (1.39–3.96) c
|
0.70 (0.39–1.30) a
|
1.20 (0.77–1.90) b
|
1 Diarrheal groups: no diarrheal control (NoDiar; n = 10), diarrheal without inflammation in colon (DiarNoInfl; n = 6), and diarrheal with inflammation in colon (DiarInfl; n = 22).
2 SCFA: Short-chain fatty acids (mg/kg wet sample). Estimated Marginal Means are reported with their 95% confidence intervals and rows with different superscript letters indicate different EMMs (P < 0.05) with pairwise comparison adjusted by BH.
Concentrations of biogenic amines were not significantly different between two segments of colon (Table 2). The NoDiar group had significantly higher concentrations of total biogenic amines, when compared to DiarInfl in Co2 (P = 0.01) and in Co3 (P < 0.05). L-lysine was lowest in the DiarNoInfl group, but putrescine concentration was highest in the DiarNoInfl group. Gender had no effects on total concentration of biogenic amines, while in individual biogenic amines, males showed to have significantly higher levels of putrescine and cadaverine compared to females. Concentrations of L-threonine, agmatine, L-valine and L-lysine were higher in digesta from females compared to males (data not shown).
Table 2
Concentration of biogenic amines in digesta (mmol/kg wet sample) from Co2 and Co3.
|
Groups1
|
|
NoDiar
|
DiarNoInfl
|
DiarInfl
|
Co2
|
|
|
|
Biogenic Amines
|
688 (488–970) b
|
585 (386–885) ab
|
416 (316–548) a
|
L-threonine
|
43.7 (22.3–85.6)
|
27.0 (11.3–64.7)
|
36.6 (22.4–59.8)
|
Agmatine
|
43.1 (26.5–70.0)
|
43.9 (23.4–82.3)
|
40.6 (28.7–57.5)
|
DL-methionine
|
14.7 (2.99–72.7)
|
12.8 (1.64–100)
|
14.1 (4.43–45.1)
|
L-valine
|
64.1 (38.1–108)
|
31.1 (15.9–60.8)
|
53.8 (36.4–79.7)
|
L-lysine
|
188 (135–261) b
|
99.3 (64.8–152) a
|
146 (115–186) ab
|
Putrescine
|
74.4 (48.3–115) b
|
85.7 (47.7–154) b
|
27.8 (20.3–38.0) a
|
Cadaverine
|
232 (93.7–573) b
|
211 (70.8–629) ab
|
86.0 (41.8–177) a
|
Co3
|
|
|
|
Biogenic Amines
|
631 (443–899) b
|
689 (452–1050) b
|
450 (342–594) a
|
L-threonine
|
32.4 (16.3–64.2)
|
22.6 (8.9–57.5)
|
37.7 (22.9–61.9)
|
Agmatine
|
45.1 (27.8–73.0)
|
49.5 (25.8–95.2)
|
52.2 (36.4–74.7)
|
DL-methionine
|
12.6 (2.48–64.2)
|
11.1 (1.16–105.8)
|
14.9 (4.55–48.6)
|
L-valine
|
53.8 (30.3–95.7)
|
30.8 (14.6–64.6)
|
59.1 (39.0-89.7)
|
L-lysine
|
175 (125–245) b
|
94.5 (59.7–150) a
|
165 (130–211) b
|
Putrescine
|
44.7 (29.2–68.3) a
|
89.4 (51.5–155) b
|
30.7 (22.5–42.0) a
|
Cadaverine
|
219 (89.1–538) b
|
234 (76.2–717) b
|
78.0 (38.1–159) a
|
1 Diarrheal groups: no diarrheal control (NoDiar; n = 10), diarrheal without inflammation in colon (DiarNoInfl; n = 6), and diarrheal with inflammation in colon (DiarInfl; n = 22). Estimated Marginal Means are reported with their 95% confidence intervals and rows with different superscript letters indicate different EMMs (P < 0.05) with pairwise comparison adjusted by BH.
Table 3 shows the concentration of NH4+ and indoles in two segments of colon, Co2 and Co3. Concentration of NH4+, total indoles and indole-3-methylindole was lowest in the DiarInfl, compared with the NoDiar and DiarNoInfl groups in both Co2 and Co3. Indole-3-acetate was remarkably high in DiarNoInfl for both Co2 and Co3 digesta with 14.1 and 10.7 mmol/kg wet sample, respectively, compared to NoDiar and DiarInfl. Gender and segment had no effects on NH4+ and indoles (data not shown).
Table 3
Concentration of indoles (µg/kg wet sample) and NH4+ (mmol/kg wet sample) in digesta from Co2 and Co3.
|
Groups1
|
|
NoDiar
|
DiarNoInfl
|
DiarInfl
|
Co2
|
|
|
|
NH4+
|
8.45 (5.29–13.5) b
|
7.82 (4.64–13.2) b
|
4.76 (3.13–7.25) a
|
Indoles
|
41.7 (25.9–67.0) b
|
37.7 (20.1–70.6) b
|
17.3 (12.2–24.6) a
|
Indole-3-acetate
|
1.31 (0.74–2.33) a
|
14.1 (6.65–29.8) b
|
2.27 (1.51–3.4) a
|
Indole-3-propionate
|
1.54 (0.93–2.54)
|
1.11 (0.62–1.99)
|
1.55 (1.01–2.39)
|
Indol-1-benzopyrrol
|
2.98 (1.66–5.35)
|
2.3 (1.06–5.03)
|
1.47 (0.95–2.28)
|
Indole-3-methylindole
|
22.4 (10.0-50.2) b
|
9.67 (3.32–28.1) ab
|
6.78 (3.73–12.3) a
|
L-Tryptophan
|
12.5 (4.36–36.1)
|
10.6 (2.64–42.5)
|
4.85 (2.23–10.5)
|
Co3
|
|
|
|
NH4+
|
11.5 (7.15–18.4) b
|
7.51 (4.43–12.7) ab
|
5.82 (3.82–8.86) a
|
Indoles
|
42.1 (26.1–67.9) b
|
35.4 (18.8–66.4) b
|
16.7 (11.9–23.7) a
|
Indole-3-acetate
|
1.15 (0.66-2.0) a
|
10.7 (5.12–22.3) b
|
1.81 (1.2–2.74) a
|
Indole-3-propionate
|
1.35 (0.82–2.23)
|
0.84 (0.46–1.54)
|
1.29 (0.84–1.98)
|
Indol-1-benzopyrrol
|
3.7 (2.04–6.71) b
|
1.3 (0.6.0-2.83) a
|
1.83 (1.2–2.79) a
|
Indole-3-methylindole
|
22.3 (9.83–50.6) b
|
9.0 (3.07–26.4) ab
|
6.46 (3.59–11.6) a
|
L-Tryptophan
|
12.8 (4.4–37.1)
|
13.1 (3.2–53.8)
|
5.12 (2.37–11.1)
|
1 Diarrheal groups: no diarrheal control (NoDiar; n = 10), diarrheal without inflammation in colon (DiarNoInfl; n = 6), and diarrheal with inflammation in colon (DiarInfl; n = 22). Estimated Marginal Means are reported with their 95% confidence intervals and rows with different superscript letters indicate different EMMs (P < 0.05) with pairwise comparison adjusted by BH.
Figure 7 represents the association of top 100 genera with different microbial fermentation products in digesta collected from Co2 and Co3 of the three groups. In total, 30, 76, and 74 genera showed significant association with the production of SCFA, biogenic amines, and indoles, respectively. The genera Shuttleworthia (r = 0.72 for butyrate; r = 0.52 for iso-acid; r = 0.72 for valerate), Syntrophococcus (r = 0.68 for butyrate; r = 48 for iso-acids; r = 0.59 for valerate), Acidaminococcus (r = 0.71 for butyrate; r = 0.60 valerate), Turicibacter (r = -0.47 for iso-acid), and Helicobacter (r = -0.62 for butyrate; r = -0.55 for valerate) were significantly associated with different SCFA production and they were changed in digesta of DiarNoInfl and DiarInfl vs. NoDiar. In both DiarNoInfl and DiarInfl, Shuttleworthia was reduced in abundance compared with NoDiar and this genus was positively associated with butyrate production. In addition, Syntrophococcus and Acidaminococcu were positively associated with butyrate concentration in digesta, which was reduced in abundance for DiarInfl vs. NoDiar, while Helicobacter was increased in abundance, and it showed negative association with butyrate concentration of digesta. Turicibacter with negative association with butyrate concentration was increased in abundance in DiarNoInfl vs. NoDiar.
The DiarInfl group showed lower abundance of Syntrophococcus, Acidaminococc, and Shuttleworthia compared with NoDiar and these genera were negatively associated with the concentration of total ammonia, indoles, and indole-3-methylindole.