Maternal probiotic exposure enhances CD8 T cell protective neonatal immunity and modulates offspring metabolome to control influenza virus infection

Abstract

Maternal gut microbiota composition contributes to the status of the neonatal immune system and could influence the early life higher susceptibility to viral respiratory infections. Using a novel protocol of murine maternal probiotic supplementation, we report that perinatal exposure to Lacticaseibacillus rhamnosus (L.rh) or Bifidobacterium animalis subsp. lactis (B.lac) increases the influenza A/PR8 virus (IAV) clearance in neonates. Following either supplementation, type 1 conventional dendritic cells (cDC1) were amplified in the lymph nodes leading to an enhanced IAV antigen-experienced IFN-γ producing effector CD8 T cells in neonates and IAV-specific resident memory CD8 T cells in adulthood. This was compatible with a higher protection of the offspring upon a secondary infection. Interestingly, only mice born to L.rh supplemented mothers further displayed an increased activation of IFN-γ producing virtual memory CD8 T cells and a production of IL-10 by CD4 and CD8 T cells that could explain a better control of the lung damages upon infection. In the offspring and the mothers, no disturbance of the gut microbiota was observed but, as analyzed through an untargeted metabolomic approach, both exposures modified neonatal plasma metabolites. Among them, we further demonstrated that genistein and 3-(3-hydroxyphenyl)propionic acid recapitulate viral clearance or cDC1 activation in neonates exposed to IAV. We conclude that maternal L.rh or B.lac supplementation confers the neonates specific metabolomic modulations with a better CD8 T cell-mediated immune protection against IAV infection.

Keywords: Bifidobacterium; CD8 T cells; IAV; Lacticaseibacillus rhamnosus; early life; memory response.

Figure 1.

Maternal supplementation with L.rh or B.lac confers a better global protection against IAV infection, limits the viral load and the virus-induced damages in the lungs of neonates. (a) protocol used to study the neonatal immune responses against IAV (innate and adaptative). (b) Survival was observed for 20 dpi and plotted as percentages over time (mantel-cox test) (n = 21–25). (c) The viral load in the neonatal lungs was followed by q-PCR at 3/6/9 dpi (n = 6–15). statistical significance was determined by a two-way ANOVA (Tukey’s multiple comparisons test). (d) lung health was analysed by histology (H&E coloration) at 6 dpi.The percentage of pneumonia area was measured (n = 12–14). Statistical significance was determined by an unpaired t test between the IAV and IAV + probiotics (L.rh or B.lac). *p < 0.05 **p < 0.01 data from three different experiments.

Figure 2.

Maternal supplementation with L.rh or B.lac increases the migration of neonatal cDC1s from the lungs to the mediastinal lymph nodes (MLNs) following PR8 influenza infection. Three-day-old neonates from L.rh-treated, B.lac-treated or untreated (-) mothers were infected with IAV. Uninfected neonates from untreated mothers were used as controls (CTRL). The lungs and the MLNs were collected at 3 dpi. (a) The total cDC1s in the lungs, (b) the CD103-expressing migratory cDC1s, and (c) the CD8.

Figure 3.

Maternal supplementation with L.rh or B.lac increases IAV-specific effector CD8 T cell responses. Three-day-old mice from L.rh-treated, B.lac-treated or untreated (-) mothers were infected with IAV (PR8-strain) and the lungs were harvested at 9 dpi to study the specific IAV CD8 T cells. Uninfected neonates from untreated mothers were used as controls (CTRL) and plotted as a dotted line. (a) The distribution of naïve and effector CD8 T cells and (B) the percentage of Ag-experienced effector cells were measured. (c) The percentage of PA-specific cells among the effector CD8 T cells was also determined. To study their capacity to produce IFN-γ, lung cells were stimulated for 4 hours with the IAV-specific peptide PA(224–233) and analyzed by flow cytometry. (d) The gating strategy and (e) the frequency of IFNγ-producing cells among Ag-experienced effector cells are shown. Statistical significance was determined by an ordinary one-way ANOVA (Dunnett’s comparisons test to untreated (-) group). *p < 0.05 ***p < 0.001. Data from two or three different experiments (n = 8–17/group).

Figure 4.

Maternal supplementation with L.rh increases the recruitment and activation of VM CD8 T cells. three-day-old mice from L.rh-treated or untreated mothers were infected with IAV (PR8-strain) and the lungs were harvested at 6 dpi to study the VM CD8 T cells by flow cytometry. Uninfected neonates from untreated mothers were used as controls (CTRL) and plotted as a dotted line. (a) The relative expression of Ifnγ, Cxcl9, Cxcl10 (e) and Eomesodermin were analyzed by RT-qPCR on the whole lungs. (n = 6–16) (b) the gating strategy used for the identification of IFN-γ–producing VM CD8 T cells by flow cytometry. (c–d) to study IFN-γ production, pulmonary cells were stimulated for 4 hours with r-IL12/r-IL18. (c) The number of VM CD8 T cells and (d) the number of IFN-γ-producing VM CD8 T cells normalized to 100 mg of neonatal lungs were determined. (f) The percentage of Eomes⁺ cells among VM T cells was measured by flow cytometry. Statistical significance was determined by an unpaired t test for the flow cytometry experiments and by a two-way ANOVA (Sidak’s multiple comparisons test) for the RT-qPCR. *p < 0.05 **p < 0.01 ***p < 0.001 ****p < 0.0001. Data from one representative experiment (out of 3).

Figure 5.

Maternal supplementation with L.rh or B.lac increases the pool of memory cells into the lungs of young mice 30 days after a neonatal IAV infection. (a) Experimental plan to study the memory response. Three-day-old mice from L.rh-treated, B.lac-treated or untreated (-) mothers were infected with IAV (PR8-strain) and grew until 30 days post-neonatal infection. Uninfected young mice (33-day-old) were used as controls (CTRL) and plotted as a dotted line. The lungs were harvested to study the pool of TRM cells. (b) The number of CD8 T cells and (c) the number of PA⁺ CD8 TRM cells were analyzed. (d) To study their capacity to produce IFN-γ, lung cells were stimulated for 4 hours with the PA(224–233) peptide. The number of IFN-γ⁺ CD8 TRM cells was determined by flow cytometry. Statistical significance was determined by an ordinary one-way ANOVA (Dunnett’s comparisons test to untreated (-) group). *p < 0.05 **p < 0.01.

Figure 6.

Maternal supplementation with L.rh or B.lac induces a protective memory responses against a secondary IAV challenge. Three-day-old mice from L.rh-treated, B.lac-treated or untreated (-) mothers were infected with IAV (PR8-strain) and grew until 30 d post-neonatal infection when a recall infection with a LD50 of the same IAV strain was done to follow (a) the viral load in the lungs at 3/6 d post-recall. (b) The survival and (c) the weight was followed for 2 weeks (Mantel–Cox test) (n = 10–18). (d) Histological staining was done at 5 dp recall and “induced bronchus-associated lymphoid tissues” (iBALT) were analyzed (white circle) (magnification x10). (e) The percentage of iBALT area and (f) the number of iBALT were measured (n = 7–10). Statistical significance was determined by a two-way ANOVA (Turkey’s multiple comparisons test) or a Kruskal–Wallis test (Dunn’s multiple comparisons test). *p < 0.05 **p < 0.01 ****p < 0.0001.

Figure 7.

Evolution of the offspring’s gut microbiota α-diversity and β-diversity during the first days of life. (a) the protocol to study the gut microbiota composition in neonates. (b-c) the neonatal gut microbiota richness and evenness expressed as Shannon index for (b) L.rh and (c) B.Lac groups. Each pup group is depicted with ART ANOVA p-values for each factor (treatment, age) and their interaction (Treatment:age; significant values at p < 0.05 depicted in green text). (d–e) Principal coordinate analysis plots based on weighted (wUnifrac; left) and unweighted (unwUnifrac; right) UniFrac, respectively, are depicted with ADONIS p-values for each factor (age, treatment) and their interaction (Age:treatment) for (d) L.rh and (e) B.lac groups (significant values at p < 0.05 depicted in green, values at p < 0.1 are depicted in yellow. (n = 10–16, each litter contained 6–10 pups (D1), 5–8 pups (D3) or 3–4 pups (D7)).

Figure 8.

Maternal supplementation with L.rh or B.lac induces changes in the neonate plasma metabolome. (a) protocol to study the plasma metabolome of neonates. (b) Assessment of hydrophilic (panel up) and hydrophobic (panel down) plasma metabolites from neonates, by partial least squares discriminant analysis (PLS-DA). (c) One-way ANOVA (threshold FDR p < 0.05) with Tukey’s posttest in the both hydrophilic and hydrophobic matrix to identify metabolites with significant differences in the peak intensity comparing the pups from B.lac or L.rh treated-mothers versus control. (d) Hierarchical cluster analysis and heatmap of all significant metabolites selected by one-way analysis of variance (ANOVA) (n = 102 metabolites); p-value (FDR) <0,05 cutoff; post-hoc analysis by Tukey’s honestly significant difference (HSD) with raw p-value cutoff < 0,05. Each colored cell on the map corresponds to a normalized abundance of peak intensity value, with metabolites in rows and samples in columns. Just metabolites are hierarchical clustered by one minus Pearson correlation with linkage method by average, and grouped rows by cluster (K-means) and class of metabolites. n = 9–10. The figures were drawn via metaboanalyst software v 6.0 (https://www.Metaboanalyst.ca/) and phantasus v 1.21.5 (https://ctlab.Itmo.ru/phantasus).

Figure 9.

Identified xenobiotic metabolites (genistein and 3-(3-hydroxyphenyl)propionic acid) increase migratory cDC1s in MLN and/or viral load following PR8 influenza infection. (a) Rank of 15 most important metabolites from the 102 significant metabolites, identified by the random forest according to the mean decrease accuracy on the x-axis. Colored boxes indicate the relative peak intensity of the corresponding metabolite in each group. (b) Spearman rank correlation between 3-(3-hydroxyphenyl)propionate sulfate (33HPP-S) and metabolites. Group (ctrl, B.lac and L.rh) was selected as covariant of interest. The arrows point to significant metabolites from the xenobiotic class. (n = 9–10) (C) the protocol to study the impact of metabolites given to neonates during the influenza infection. (d) The viral load in the neonatal lungs was monitored by q-pcr at 3/6 dpi (n = 5–14). Statistical significance was determined by a two-way ANOVA (Tukey’s multiple comparisons test). Uninfected neonates from untreated mothers were used as controls (CTRL) and plotted as a dotted line. (e) The CD103-expressing migratory cDC1s were analyzed by flow cytometry in the MLN at 3dpi. Statistical significance was determined by an ordinary one-way ANOVA (Holm-Sidak’s multiple comparisons test). *p < 0.05 **p < 0.01 ***p < 0.001. Data are from two different experiments. (f) Lung health was analyzed by histology (H&E coloration) at 6 dpi and the percentage of pneumonia area was measured (n = 7–8). Statistical significance was determined by an unpaired T test or Mann–Whitney test between the IAV and IAV + metabolites (cin, Gen or 33HPP). *p < 0.05 **p < 0.01.