The Combined Escherichia coli Nissle 1917 and Tryptophan Treatment Modulates Immune and Metabolome Responses to Human Rotavirus Infection in a Human Infant Fecal Microbiota-Transplanted Malnourished Gnotobiotic Pig Model

Abstract

Human rotavirus (HRV) is a major cause of childhood diarrhea in developing countries where widespread malnutrition contributes to the decreased oral vaccine efficacy and increased prevalence of other enteric infections, which are major concerns for global health. Neonatal gnotobiotic (Gn) piglets closely resemble human infants in their anatomy, physiology, and outbred status, providing a unique model to investigate malnutrition, supplementations, and HRV infection. To understand the molecular signatures associated with immune enhancement and reduced diarrheal severity by Escherichia coli Nissle 1917 (EcN) and tryptophan (TRP), immunological responses and global nontargeted metabolomics and lipidomics approaches were investigated on the plasma and fecal contents of malnourished pigs transplanted with human infant fecal microbiota and infected with virulent (Vir) HRV. Overall, EcN + TRP combined (rather than individual supplement action) promoted greater and balanced immunoregulatory/immunostimulatory responses associated with greater protection against HRV infection and disease in malnourished humanized piglets. Moreover, EcN + TRP treatment upregulated the production of several metabolites with immunoregulatory/immunostimulatory properties: amino acids (N-acetylserotonin, methylacetoacetyl-CoA), lipids (gamma-butyrobetaine, eicosanoids, cholesterol-sulfate, sphinganine/phytosphingosine, leukotriene), organic compound (biliverdin), benzenoids (gentisic acid, aminobenzoic acid), and nucleotides (hypoxathine/inosine/xanthine, cytidine-5'-monophosphate). Additionally, the levels of several proinflammatory metabolites of organic compounds (adenosylhomocysteine, phenylacetylglycine, urobilinogen/coproporphyrinogen) and amino acid (phenylalanine) were reduced following EcN + TRP treatment. These results suggest that the EcN + TRP effects on reducing HRV diarrhea in neonatal Gn pigs were at least in part due to altered metabolites, those involved in lipid, amino acid, benzenoids, organic compounds, and nucleotide metabolism. Identification of these important mechanisms of EcN/TRP prevention of HRV diarrhea provides novel targets for therapeutics development. IMPORTANCE Human rotavirus (HRV) is the most common cause of viral gastroenteritis in children, especially in developing countries, where the efficacy of oral HRV vaccines is reduced. Escherichia coli Nissle 1917 (EcN) is used to treat enteric infections and ulcerative colitis while tryptophan (TRP) is a biomarker of malnutrition, and its supplementation can alleviate intestinal inflammation and normalize intestinal microbiota in malnourished hosts. Supplementation of EcN + TRP to malnourished humanized gnotobiotic piglets enhanced immune responses and resulted in greater protection against HRV infection and diarrhea. Moreover, EcN + TRP supplementation increased the levels of immunoregulatory/immunostimulatory metabolites while decreasing the production of proinflammatory metabolites in plasma and fecal samples. Profiling of immunoregulatory and proinflammatory biomarkers associated with HRV perturbations will aid in the identification of treatments against HRV and other enteric diseases in malnourished children.

Keywords: Escherichia coli Nissle 1917; human rotavirus infection; lipidomics; malnutrition; metabolomics; neonatal gnotobiotic pigs; tryptophan.

FIG 1

EcN + TRP treatment enhanced B cell immune responses post-VirHRV challenge. Survival rate (A) and normalized weight gain of pigs (B) after HIFM transplantation, supplementation, and VirHRV challenge were monitored for 2 weeks. (C) Virus shedding was determined by cell culture immunofluorescence assay and expressed as log₁₀ FFU/ml. (D) Principal-component analysis (PCA) of immunological parameters. (E) Mean HRV-specific IgA antibody-secreting cells (ASCs) in splenic and duodenal cells. (F to H) Geometric mean titers (log₁₀) HRV-specific IgA antibody in serum, small intestinal contents (SIC), and large intestinal content (LIC). (I) Mean frequencies of CD79β⁺IgA⁺ B cells in splenic and duodenal cells. (J) Mean frequencies of activated antibody-forming (CD79β⁺CD21⁺CD2⁻) B cells in systemic and ileal cells. Data are shown as means ± SEM, and data are compared with untreated group pigs. Gnotobiotic (Gn) pigs were cesarian derived and transplanted with HIFM at 4 days of age, post-HIFM transplantation day (PBTD) 0. Pigs were fed a deficient (Def) and/or sufficient (Suf) diet, supplemented orally with Escherichia Coli Nissle 1917 (EcN) and/or tryptophan (TRP) at PBTD7, subsequently challenged with virulent (Vir) HRV at PBTD9-10/postchallenge day (PCD) 0, and euthanized on PBTD24/PCD14.

FIG 2

EcN + TRP treatment enhanced T cell immune responses post-VirHRV challenge. (A) Total numbers of mononuclear cells (MNCs) in the ileum. (B) Mean frequencies of lymphocytes among total MNCs in systemic tissues. (C) Mean frequencies of T-helper (CD3+CD4⁺) cells in the blood and ileal tissues. (D and E) Mean frequency of cytotoxic (CD3⁺CD8⁺) T and activated (CD4⁺CD25⁺FOXP3⁻) T-regulatory cells in blood cells. (F) Mean frequency of natural (CD8⁺CD25⁺FOXP3⁺) T-regulatory cells in duodenal tissues. (G and H) Mean frequency of CD4/CD8⁺TGF-β + T cells in splenic cells. (I) Cytokines profile was determined in serum samples collected at the terminal time point (PCD7). Data are shown as means ± SEM. Gnotobiotic (Gn) pigs were cesarian derived and transplanted with human infant fecal microbiota (HIFM) at 4 days of age, post-HIFM transplantation day (PBTD) 0. Pigs were fed a deficient (Def) and/or sufficient (Suf) diet, supplemented orally with Escherichia Coli Nissle 1917 (EcN) and/or tryptophan (TRP) at PBTD7, subsequently challenged with virulent (Vir) HRV at PBTD9-10/postchallenge day (PCD) 0, and euthanized on PBTD24/PCD14. *, P < 0.05; **, P < 0.01, and ***, P < 0.001.

FIG 3

EcN + TRP treatment enhanced innate immune responses post-VirHRV challenge. (A) Mean frequency of CD103⁺ pDCs among total mononuclear cells (MNCs) in the ileum. (B) Mean frequencies of MNCs expressing TLR4 in systemic and intestinal tissues, (C) TLR3 in splenic and intestinal tissues, and (D) TLR9 in systemic tissues. (E) Mean frequency of NK (SWC3a⁺CD16⁺) and (F) NK cell function in systemic and blood MNCs. Blood MNCs and carboxyfluorescein diacetate succinimidyl ester (CFSE) stained K562 tumor cells were used as effector and target cells, respectively, and cocultured at set ratios to assess the NK cytotoxic function. For the effector, target cell cocultures were stained with 7-aminoactinomycin D (7AAD) after 12 h of incubation at 37°C, and the frequencies of CFSE-7AAD double-positive cells (lysed K562 target cells) were assessed by flow cytometry. (G) Mean frequencies of apoptotic MNCs (PI-/Annexin-APC-) among total blood MNCs. Data are shown as means ± SEM. Gnotobiotic (Gn) pigs were cesarian derived and transplanted with human infant fecal microbiota (HIFM) at 4 days of age, post-HIFM transplantation day (PBTD) 0. Pigs were fed a deficient (Def) and/or sufficient (Suf) diet, supplemented orally with Escherichia coli Nissle 1917 (EcN) and/or tryptophan (TRP) at PBTD7, subsequently challenged with VirHRV at PBTD9-10/postchallenge day (PCD) 0, and euthanized on PBTD24/PCD14. *, P < 0.05.

FIG 4

Principal-component analysis (PCA) of plasma/fecal metabolomics and lipidomics. Gnotobiotic (Gn) pigs were cesarian-derived and transplanted with human infant fecal microbiota (HIFM) at 4 days of age, post-HIFM transplantation day (PBTD) 0. Pigs were fed a deficient (Def) and/or sufficient (Suf) diet, supplemented orally with Escherichia coli Nissle 1917 (EcN) and/or tryptophan (TRP) at PBTD7, subsequently challenged with VirHRV at PBTD9-10/postchallenge day (PCD) 0, and euthanized on PBTD24/PCD14. PM, plasma metabolomics; PL, plasma lipidomics; FM, fecal metabolomics; FL, fecal lipidomics.

FIG 5

Heatmap of significant metabolites of plasma/fecal metabolomics and lipidomics. Heatmap visualizes the intensity/view of metabolites. More intense color represents the higher magnitude while the less intense color represents the lower magnitude of metabolite. In addition, the magnitude of metabolite is arranged to show their relatedness in the cluster heatmap. Gnotobiotic (Gn) pigs were cesarian derived and transplanted with human infant fecal microbiota (HIFM) at 4 days of age, post-HIFM transplantation day (PBTD) 0. Pigs were fed a deficient (Def) and/or sufficient (Suf) diet, supplemented orally with Escherichia coli Nissle 1917 (EcN) and/or tryptophan (TRP) at PBTD7, and, subsequently challenged with VirHRV at PBTD9-10/postchallenge day (PCD) 0, and euthanized on PBTD24/PCD14. PM, plasma metabolomics; PL, plasma lipidomics; FM, fecal metabolomics; FL, fecal lipidomics.

FIG 6

Experimental design showing time points for hysterectomy, human infant fecal microbiota (HIFM) transplantation, E. coli Nissle (EcN) 1917 probiotic/tryptophan (TRP) treatment, virulent human rotavirus (VirHRV) challenge, and euthanasia. PBTD, postbacterial transplantation day; PCD, PBTD; postchallenge.