An Enteric Bacterial Infection Triggers Neuroinflammation and Neurobehavioral Impairment in 3xTg-AD Transgenic Mice

Abstract

Background: Klebsiella pneumoniae is infamous for hospital-acquired infections and sepsis, which have also been linked to Alzheimer disease (AD)-related neuroinflammatory and neurodegenerative impairment. However, its causative and mechanistic role in AD pathology remains unstudied.

Methods: A preclinical model of K. pneumoniae enteric infection and colonization is developed in an AD model (3xTg-AD mice) to investigate whether and how K. pneumoniae pathogenesis exacerbates neuropathogenesis via the gut-blood-brain axis.

Results: K. pneumoniae, particularly under antibiotic-induced dysbiosis, was able to translocate from the gut to the bloodstream by penetrating the gut epithelial barrier. Subsequently, K. pneumoniae infiltrated the brain by breaching the blood-brain barrier. Significant neuroinflammatory phenotype was observed in mice with K. pneumoniae brain infection. K. pneumoniae-infected mice also exhibited impaired neurobehavioral function and elevated total tau levels in the brain. Metagenomic analyses revealed an inverse correlation of K. pneumoniae with gut biome diversity and commensal bacteria, highlighting how antibiotic-induced dysbiosis triggers an enteroseptic "pathobiome" signature implicated in gut-brain perturbations.

Conclusions: The findings demonstrate how infectious agents following hospital-acquired infections and consequent antibiotic regimen may induce gut dysbiosis and pathobiome and increase the risk of sepsis, thereby increasing the predisposition to neuroinflammatory and neurobehavioral impairments via breaching the gut-blood-brain barrier.

Keywords: Klebsiella pneumoniae; Alzheimer disease; dysbiosis; microbiome; sepsis.

Figure 1.

Klebsiella pneumoniae proliferates during dysbiosis and subsequently translocates to the brain through circulation. A, Schematic study design created with BioRender.com. B, Principal coordinate analysis based on Bray-Curtis dissimilarity was utilized to represent beta-diversity. Significance was calculated using PERMANOVA with 999 random permutations. Alpha-diversity was determined using the Shannon index. Significance between groups was calculated using the nonparametric Kruskal-Wallis test. C, Klebsiella levels determined by CFU counting and their relative abundance measured by sequencing at weeks 1 and 3. The correlation between microbial diversity (Shannon index) and Klebsiella relative abundance was calculated using Spearman rank correlation. The relative abundance of bacterial microbiome composition at the genus level is presented. D, Prevalence and relative abundance of Klebsiella in each part of the intestine (duodenum, jejunum, ileum, and colon), serum, and brain at week 6. Data are presented as mean ± SD. *P < .05, **P < .01, ***P < .001. TgCt, 3xTg-AD mice control group (n = 6); Kpn, 3xTg-AD mice infected with K. pneumoniae (n = 8, BL and W1; n = 7, W3); Kpn + A, 3xTg-AD mice infected with K. pneumoniae and treated with antibiotic (n = 10, BL; n = 8, W1; n = 7, W3). Abbreviations: A, antibiotic; AD, Alzheimer disease; BL, base line; CFU, colony-forming units; PC, principal component; W, week.

Figure 2.

Antibiotic-induced dysbiosis opens favorable niche for Klebsiella pneumoniae colonization. A, The microbial abundance at the phylum, family, and genus levels of each sample is shown for week 1. For the abundance heatmap, z-scores were calculated. Dendrograms were generated using hierarchical clustering results with the average linkage method based on Bray-Curtis dissimilarity (● P < .05). B, Volcano plots display differentially abundant ASVs (log₂ fold change) at week 1. Dots above dotted line (y = 1.301) represent significantly (P < .05) higher abundance in the TgCt, Kpn, or Kpn + A group. Cladograms illustrate differentially abundant taxa analyzed using the LDA effect size algorithm at week 1. Significantly different taxa between groups from phylum to genus level are indicated with alphabet. C, A circular heatmap illustrates the differences in genus at week 1 on Kpn infection and antibiotic treatment, as determined using a 2-way ANOVA model (● P < .05). D, Taxa significantly correlated (P < .05) with the Klebsiella genus at week 1 were determined using Spearman rank correlation. TgCt, 3xTg-AD mice control group (n = 6); Kpn, 3xTg-AD mice infected with K. pneumoniae (n = 8, BL and W1; n = 7, W3); Kpn + A, 3xTg-AD mice infected with K. pneumoniae and treated with antibiotic (n = 10, BL; n = 8, W1; n = 7, W3). Abbreviations: AD, Alzheimer disease; ASV, amplicon sequence variants; BL, base line; LDA, linear discriminant analysis; W, week.

Figure 3.

Klebsiella pneumoniae enteric infection induces neurobehavioral impairment. A, The survival rate and body weight from the baseline to week 6 for each group and photographs displaying a normal brain in the TgCt group of mice and a brain covered by severe abscesses found in the Kpn + A group of mice. B, Gene expression levels of inflammatory markers (IL-1β, IL-6, IL-8, IL-10, and TNF-α) in the hippocampus (n = 5–6 per group), hypothalamus (n = 3–4 per group), and frontal cortex (n = 5–6 per group) are presented. C, Results encompassing the total traversed distance and freezing time, as well as the discrimination index value in the novel object test. The outcomes of the T-maze spontaneous test, including the average time spent making decisions and the alternation score, which tallies alternative decisions compared to the previous trial. The results of the inclined screen test, determined by the maximum angle at which mice maintain balance and exhibit climbing behavior (n = 5–6 per group for all tests). D, The correlation between the levels of inflammatory markers in the hippocampus, hypothalamus, and FC (left), and behavioral assay outcomes (right) with the relative abundance level of the Klebsiella genus in each part of the intestine (duodenum, jejunum, ileum, and colon) is depicted. Data are presented as mean ± SE. Significance was determined by the 1-way ANOVA test with post hoc Dunn test. ♦ TgCt versus Kpn, P < .05; † TgCt versus Kpn + A, P < .05; #P < .1, *P < .05, **P < .01, ***P < .001. TgCt, 3xTg-AD mice control group; Kpn, 3xTg-AD mice infected with K. pneumoniae; Kpn + A, 3xTg-AD mice infected with K. pneumoniae and treated with antibiotic. Abbreviations: AD, Alzheimer disease; DI, discrimination index; FC, frontal cortex; Hippo, hippocampus; Hypo, hypothalamus; IL, interleukin; NO, novel object; TNF-α, tumor necrosis factor-α; W, week.

Figure 4.

Klebsiella pneumoniae enteric infection during dysbiosis elevates total tau level. A, Triton-insoluble and (C) Triton-soluble levels of AD biomarkers (amyloid-β precursor protein (APP) and total/phosphorylated tau) in brain were quantified by western blot (n = 5–6/group). The value normalized by β-actin is denoted under each column. The (B) Triton-insoluble and (D) Triton-soluble protein levels are presented by fold-change calculated compared to TgCt group. Data are presented as mean ± SE. Significance was determined by the Kruskal-Wallis H test with post hoc Dunn test. #P < .1, **P < .01. TgCt, 3xTg-AD mice control group; Kpn, 3xTg-AD mice infected with K. pneumoniae; Kpn + A, 3xTg-AD mice infected with K. pneumoniae and treated with antibiotic. Abbreviations: AD, Alzheimer disease; Nor., normalized; Phospho, phosphorylated.