Gut Microbiota of Obese Children Influences Inflammatory Mucosal Immune Pathways in the Respiratory Tract to Influenza Virus Infection: Optimization of an Ideal Duration of Microbial Colonization in a Gnotobiotic Pig Model

Abstract

The impact of obesity on the human microbiota, immune maturation, and influenza virus infection has not been yet established in natural host animal models of influenza. In this study, gnotobiotic (Gn) pigs were colonized with human fecal microbiota (HFM) of obese (oHFM) or healthy lean (hHFM) children and infected at different periods (2-, 3-, and 5-weeks post-transplantation) using a zoonotic influenza virus strain. The infected oHFM pigs were characterized by lower levels of Firmicutes (Lactococcus, Lactobacillus, Turicibacter, and Streptococcus) and Actinobacteria (Bifidobacterium), which was associated with higher levels of Proteobacteria (Klebsiella), Bacteroidetes, and Verrucomicrobia (Akkermansia) compared with the infected hHFM group (P < 0.01). Furthermore, these genera significantly correlated with the expression of immune effectors, immune regulators, and inflammatory mediators, and displayed opposite trends between oHFM and hHFM groups (P < 0.01). The lymphoid and myeloid immune cell frequencies were differently modulated by the oHFM and hHFM colonization, especially apparent in the 5-weeks HFM colonized piglets. In addition, oHFM group had higher pro-inflammatory cytokines (IL-6, IL-12, TNF-α, and IFNγ) gene expression in the respiratory tract compared with the hHFM colonized pigs was detected. In conclusion, pigs colonized for longer duration, established oHFM increased the immune maturation favoring the activation of inflammatory mediators, however, the influenza virus load remained comparable with the hHFM group. Further, a longer duration of microbial colonization (5 weeks) may be required to reveal the impact of microbiome on the host immune maturation and susceptibility to influenza virus infection in the humanized Gn pig model. IMPORTANCE The diversity of gut microbiome of obese people differs markedly from that of lean healthy individuals which, in turn, influences the severity of inflammatory diseases because of differential maturation of immune system. The mouse model provides crucial insights into the mechanism(s) regulating the immune systems mediated by the gut microbiota but its applicability to humans is questionable because immune cells in mice are poorly activated in microbiota humanized mice. Several important strains of Bifidobacterium, Lactobacillus, and Clostridium fails to colonize the murine gut. Thus, understanding the role of certain important commensal gut bacterial species influences upon health and disease, a suitable large animal model like pig that supports the growth and colonization of most of the important human gut bacteria and possess comparable immunology and physiology to humans is beneficial to improve health.

Keywords: gnotobiotic pigs; healthy; human fecal microbiota; immune maturation; inflammation; influenza virus; obese.

FIG 1

Schematic illustration of the experimental overview, longitudinal sampling time points, and details of the analysis. Initially, Gn pigs were transplanted with fecal material of obese and healthy lean children and colonized human microbiota in pig intestines (outgrowth) was collected and transplanted to all the experimental Gn pigs.

FIG 2

Influence of diverse gut microbiota on influenza virus infection. Gnotobiotic piglets were inoculated with healthy lean and obese outgrowth HFM and at 2-, 3-, and 5-weeks post transplantation challenged with SwIAV H1N1-OH7. Samples collected at different postchallenge day (PCD) were analyzed for the challenge virus load in (A) nasal swab at PCD 2; (B) nasal swab at PCD 4; and (C) BAL fluid at PCD 7. Data are presented as the mean of 4 to 6 pigs ± SEM analyzed using unpaired t test. Asterisks denote significant difference (*, P < 0.05).

FIG 3

Activation of pro-inflammatory cytokines expression in the mucosal lymph nodes of pigs mediated by gut microbiota and influenza virus infection. Gnotobiotic piglets were inoculated with healthy lean and obese outgrowth HFM and at 2-, 3-, and 5-weeks post transplantation challenged with SwIAV H1N1-OH7. At PCD 7 animals were euthanized and RNA extracted from tissues were analyzed for mRNA expression by qRT-PCR in TBLN (A) TNF-α and (B) IL-12; and in MLN (C) TNF-α; (D) IFNγ; and (E) IL-6. Data are presented as the mean of 4 to 6 pigs ± SEM analyzed using unpaired t test. Asterisks denote significant difference (*, P < 0.05).

FIG 4

Modulation in the population of lymphoid and myeloid immune cells in the mucosal and systemic sites of pigs influenced by colonized HFM and influenza virus. Gnotobiotic piglets were inoculated with healthy lean and obese outgrowth HFM and at 2-, 3-, and 5-weeks post transplantation challenged with SwIAV H1N1-OH7. At PCD 7 animals were euthanized and MNCs isolated from TBLN, MLN, and blood (PBMCs) were immunostained and analyzed for the frequency of different lymphoid and myeloid cells by flow cytometry. (A, D, G) T helper/memory cells; (B, E, H) Cytotoxic T lymphocytes; and (C, F, I) CD172⁺ myeloid cells. Data are presented as the mean of 4 to 6 pigs ± SEM analyzed using unpaired t test. Asterisks denote significant difference (*, P < 0.05, **, P < 0.01, and ***, P < 0.001).

FIG 5

Microbiota composition at the phylum level in fecal, intestinal, and systemic tissues across the three challenging experiments. Relative abundance in rectal (A to C) and nasal (D to F) swabs at postchallenging day 0 (A0), 2 (A2), and 7 (A7). Relative abundance in intestinal (ileum and colon; G to I), respiratory (lung and BAL; J to L) and systemic tissues (MLN; M to O) at postchallenging day 7 (PCD 7). Obese and healthy: pigs transplanted with an obese or healthy human fecal microbiota (HFM), respectively. 2-, 3-, and 5-weeks post-transplantation: maturation time between HFM transplantation and influenza challenge. NS, nasal swab; RS, rectal swab; BAL, bronchoalveolar lavage fluid; MLN, mesenteric lymph node. Details about the significant differences in relative abundance (from phylum to species levels) detected between hHFM and oHFM for a designated incubation time post-transplantation, time point post influenza challenge and sample type are displayed in Table S3.

FIG 6

Differences in microbiota composition at the genus level between healthy and obese HFM pig samples. Microbiome differences observed in the 2-weeks (A) and 3-weeks (B) post-transplantation experiments. Data presented in the principal-component analysis plot were generated based on the associated heatmap data presented for each time point (0, 2, and 7 days PCD). The PCA displays the significant differences detected between oHFM and hHFM, across all time points and sample types. Only predominant OTUs (relative abundance > 1%) showing significant differences (P < 0.01) between the healthy and obese groups are displayed in these figures. O and H: pigs transplanted with an obese or healthy human fecal microbiota (HFM), respectively. The color of the cell is associated with the relative abundance of the OTU in the designated tissue and time point. NS, nasal swab; RS, rectal swab; BAL, bronchoalveolar lavage fluid; MLN, mesenteric lymph node. Details about the significant differences in relative abundance (from phylum to species levels) detected between hHFM and oHFM for a designated incubation time post transplantation, time point post influenza challenge and sample type are displayed in Table S3.

FIG 7

Distribution of the microbiota collected from intestinal (rectal swab, colon, and ileum) and respiratory (nasal swab, lung, and BAL fluid) tissues of the healthy human fecal microbiota (hHFM) pig group at 7 days postchallenge (PCD 7), across the three experiments (2-, 3-, and 5-weeks post-transplantation) based on its correlations between the immune responses (immune regulators, immune effectors, and inflammatory responses). Clusters A (n = 10) and C (n = 15) indicates OTUs frequently negatively or positively correlated with the immune parameters, respectively. Cluster B (n = 16) indicates the OTUs frequently negatively and positively correlated with the immune parameters. Clusters D (n = 40) indicates the OTUs not frequently correlated with the immune parameters. Inflammatory responses (TNF-α, IL-6, and IFNγ), immune effectors (CD3+CD4+CD8α + β- and CD3+CD4-CD8α + β+T cells), and immune regulators (IL-12 and CD3-CD172⁺).

FIG 8

Microbiome-immune response correlations in intestinal and systemic tissues across the three experiments. The two-way clustering plots were generated using only significant correlations (P > 0.01 and r² lower than −0.25 or higher than 0.25) at PCD 7. The color of the cells is associated with the type of correlations between the designated OTU and immune parameter. Blue and red cells: negative or positive correlation, respectively, between the designated OTU and immune parameter in the healthy HFM group. The opposite trend was observed in the obese HFM group. Obese and healthy group: pigs transplanted with an obese or healthy human fecal microbiota (HFM), respectively. I, inflammatory responses (TNF-α, IL-6, and IFNγ); E, immune effectors (CD3+CD4+CD8α + β- and CD3+CD4-CD8α + β+T cells); and R, immune regulators (IL-12 and CD3-CD172⁺).