Enterococcus faecium and Pediococcus acidilactici deteriorate Enterobacteriaceae-induced depression and colitis in mice

Abstract

Gut dysbiosis is closely associated with the outbreak of inflammatory bowel disease (IBD) and psychiatric disorder. The Enterobacteriaceae population was higher in the feces of patients with inflammatory bowel disease (IBD-F) than in those of healthy control volunteers (HC-F). The Enterococcaceae and Lactobacillaceae populations were higher in the feces of IBD patients with depression (IBD/D⁺-F) vs. the feces of IBD patients without depression (IBD/D^--F). Therefore, we examined the effects of Klebsiella oxytoca, Escherichia coli, Cronobacter sakazakii, Enterococcus faecium, and Pediococcus acidolactici overpopulated in IBD/D⁺-F and their byproducts LPS and exopolysaccharide (EPS) on the occurrence of depression and colitis in mice. Oral gavages of Klebsiella oxytoca, Escherichia coli, and Cronobacter sakazakii belonging to Enterobacteriaceae, singly or together, caused dose-dependently colitis and depression-like behaviors in germ-free and specific-pathogen-free mice. Although Enterococcus faecium and Pediococcus acidolactici did not significantly cause colitis and depression-like behaviors, they significantly deteriorated Klebsiella oxytoca- or Escherichia coli-induced colitis, neuroinflammation, and anxiety/depression-like behaviors and increased blood LPS, corticosterone, and IL-6 levels. The EPSs from Enterococcus faecium and Pediococcus acidolactici also worsened Klebsiella oxytoca LPS-induced colitis, neuroinflammation, and depression-like behaviors in mice and increased the translocation of fluorescein isothiocyanate-conjugated LPS into the hippocampus. However, Bifidobacterium longum, which was lower in IBD/D⁺-F vs. IBD/D^--F, or its EPS suppressed them. In conclusion, Enterococcus faecium and Pediococcus acidolactici, known as a probiotic strain, and their EPSs may be a risk factor for the outbreak of depression and IBD.

Figure 1

The fecal microbiota composition of patients with IBD/D⁻ or IBD/D⁺ and healthy individuals (HCs). (A) Effects on the levels of gut bacteria Klebsiella sp., Cronobacter sp., and Escherichia sp. grown in the DHL agar plate, Bifidobacterium sp. in the BL agar plate, and Pediococcus sp. and Enterococcus sp. in the mEn agar plate. (B) Effects on the levels of gut bacteria Klebsiella sp. (a), Klebsiella oxytoca (b), Escherichia coli (c), Pediococcus acidilactici (d), Enterococcus faecium (e), Bifidobacterium longum (f) and Cronobacter sakazekii (g), assessed by qPCR. Data are shown as box plots (HC n = 6; IBD/D⁻ n = 8; IBD/D⁺, n = 7). Means with same letters are not significantly different (p < 0.05). (b,c), Kruskal–Wallis test (nonparametric test).

Figure 2

Effect of gut bacteria on the occurrence of anxiety/depression and colitis in mice. Effects on the time spent in open arms (OT) in the EPMT (a), marble-buried number in MBT (b), immobility time in the TST (c), and forced FST (d), and effects on the BDNF⁺/NeuN⁺ and NF-κB⁺/Iba1⁺ cell population in the hippocampus (e). Effects on the colon length (f) and macroscopic score (g). (h) Effects on the IL-1β, IL-6, LPS, myeloperoxidase (MPO), and corticosterone levels, indicated as compared to control group (CON: 1). Heatmap was generated using Plotly (https://chart-studio.plotly.com). Klebsiella oxytoca (Ko), Escherichia coli (Ec), Cronobacter sakazakii (Cs), Enterococcus faecium (Ef), Pediococcus acidilactici (Pa), or Bifidobacterium longum (Bl) at a dose of 1 × 10⁹ CFU/mouse/day was orally gavaged in mice once a day for 5 days. Control mice (Con) were treated with vehicle (saline) instead of the gut bacterial suspension. Data are shown as box plots (n = 6). *p < 0.05 vs Con. Means with same letters are not significantly different (p < 0.05). All were analyzed using unpaired t test.

Figure 3

Effect of Klebsiella oxytoca on the occurrence of depression and colitis in germ-free mice. Effect on the occurrence of depression-like behaviors (a) and hippocampal IL-1β level (b), BDNF⁺/NeuN⁺ (c), NF-κB⁺/Iba1⁺ (d), LPS⁺/Iba1⁺ (e), and IL-1R⁺ cell populations (f) in germ-free mice. Klebsiella oxytoca (KO, 1 × 10⁷ CFU/mouse/day) were orally gavaged for 5 days in mice (n = 6, in specific-germ-free mice; n = 4, in germ-free mice). Control mice (CON) were treated with vehicle (saline) instead of the bacterial suspension. Data are shown as box plots. Means with same letters are not significantly different (p < 0.05). All were analyzed using unpaired t-test.

Figure 4

Combined effects of two bacteria belonging to Enterobacteriaceae on the occurrence of anxiety/depression and colitis in mice. Effects on the occurrence of anxiety/depression in the EPMT (a) and TST (b). Effects on the BDNF⁺/NeuN⁺ and NF-κB⁺/Iba1⁺ cell population in the hippocampus (c). Effects on the colon length (d) and macroscopic score (e). (f) Effects on the IL-1β, IL-6, LPS, MPO, and corticosterone levels, indicated as compared to control group (CON: 1). Heatmap was generated using Plotly (https://chart-studio.plotly.com). Among Klebsiella oxytoca (Ko), Escherichia coli (Ec), and Cronobacter sakazakii (Cs), two bacterial (KoEc, 1:1 of KO and Ec; KoCs, 1:1 of Ko and Cs; EcCS, 1:1 of Ec and Cs) combinations at a dose of 1 × 10⁸ CFU/mouse/day were orally gavaged once a day for 5 days in mice. Control mice were treated with vehicle (saline) instead of gut bacterial suspension. Data are shown as box plots (n = 6). *p < 0.05 vs Con. All were analyzed using unpaired t test.

Figure 5

Combined effects of Enterococcus faecium (Ef), Pediococcus acidilactici (Pa), or Bifidobacterium longum (Bl) with Enterobacteriaceae on the occurrence of anxiety/depression and colitis in mice. Effects on the occurrence of anxiety/depression in the EPMT (a) and TST (b). Effects on the BDNF⁺/NeuN⁺ and NF-κB⁺/Iba1⁺ cell population in the hippocampus (c). Effects on the colon length (d) and macroscopic score (e). (f) Effects on the IL-1β, IL-6, LPS, MPO, and corticosterone levels, indicated as compared to control group (CON: 1). Heatmap was generated using Plotly (https://chart-studio.plotly.com). Each two bacteria (KoEf, 1:1 of Ko and Ef; KoPa, 1:1 of Ko and Pa; KoBl, 1:1 of Ko and Bl; EcEf, 1:1 of Ec and Ef; EcPa, 1:1 of Ec and Pa; EcBl, 1:1 of Ec and Bl) combination at a dose of 1 × 10⁷ (KoEf), 1 × 10⁸ (KoPa), or 1 × 10⁹ CFU/mouse/day (EcEf, EcPa, EcBl, Ef, Pa, and Bl) was orally gavaged once a day for 5 days in mice. Control mice were treated with vehicle (saline) instead of gut bacterial suspension. Data are shown as box plots (n = 8). *p < 0.05 vs Con. All were analyzed using unpaired t test.

Figure 6

Effects of gut bacteria EPS and LPS on the occurrence of anxiety/depression and colitis in mice. (A) Effects of orally gavaged EPS and/or LPS. Effects on the occurrence of anxiety/depression in the EPMT (a) and TST (b). Effects on the BDNF⁺/NeuN⁺ and NF-κB⁺/Iba1⁺ cell population in the hippocampus (c). Effects on colon length (d) and macroscopic score (e). (f) Effects on the IL-1β, IL-6, LPS, MPO, and corticosterone levels, indicated as compared to control group (CON: 1). Heatmap was generated using Plotly (https://chart-studio.plotly.com). (B) Effects of intraperitoneally injected EPS and/or LPS on the occurrence of anxiety/depression in the EPMT (a) and TST (b) and expression of IL-1β in the hippocampus (c). (C) Effects of EPSs on the translocation of FITC-conjugated LPS into the brain. Test agents (CON, vehicle [saline]; EE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Enterococcus faecium exopolysaccharide; PE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Pediococcus acidilactici exopolysaccharide; BE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Bifidobacterium longum exopolysaccharide; KL, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Klebsiella oxytoca lipopolysaccharide; KLEE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Ko and Ef (1:1); KLPE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Ko and Pa (1:1); KLBE, 20 μg/kg (i.p.) or 20 mg/kg (p.o.) of Ko and Bl (1:1) were treated in mice once a day for 5 days. FITC-conjugated LPS was orally gavagaed daily for 2 days from 24 h after EPS or LPS conjugated without FITC was gavaged for 3 days. Control mice were treated with vehicle (saline) instead of gut bacterial suspension. Data are shown as box plots (n = 8). *p < 0.05 vs Con. All were analyzed using unpaired t test.

Figure 7

Effects of Enterococcus faecium, Pediococcus acidilactici, and their exopolysaccharides on Klebsiella oxytoca- or its lipopolysaccharide-induced IL-1β and IL-6 expression and NF-κB activation in macrophages. (A) Effects of gut bacteria on IL-1β (a) and IL-6 expression (b) in macrophages. (B) Effects of gut bacterial LPS and EPS on IL-1β (a) and IL-6 expression (b) in macrophages. Test agents (CON, vehicle [saline]; Ef, 1 × 10⁵ CFU/mL of Enterococcus faecium (Ef); Pa, 1 × 10⁵ CFU/mL of Pediococcus acidilactici; Bl, 1 × 10⁵ CFU/mL of Bifidobacterium longum; Ko, 1 × 10⁵ CFU/mL of Klebsiella oxytoca; KoEf, 1 × 105 CFU/mL of Ko and Ef (1:1); KoPa, 1 × 10⁵ CFU/mL of Ko and Pa (1:1); KoBl, 1 × 10⁵ CFU/mL of Ko and Bl (1:1). KL, 100 ng/mL of Ko lipopolysaccharide (LPS); EE, 100 ng/mL of Ef expolysaccharide (EPS); PE, 100 ng/mL of Pa; BE, 100 ng/mL of Bl; KLEE, 100 ng/mL of Ef EPS with 100 ng/mL of Ko LPS; KLPE, 100 ng/mL of Pa EPS with 100 ng/mL of Ko LPS; KLBE, 100 ng/mL of Bl EPS with 100 ng/mL of Ko LPS) was treated in macrophages. The macrophages were treated with LPS (80 ng/mL) in the presence or absence of Ef, Pa, their EPSs, Ko, and/or its LPS for 20 h. Cytokine levels were assayed using ELISA kits. Data are shown as box plots (n = 6). *p < 0.05 vs Con. All were analyzed using unpaired t test.