Gut microbial dysbiosis exacerbates long-term cognitive impairments by promoting intestinal dysfunction and neuroinflammation following neonatal hypoxia-ischemia

Abstract

Neonatal hypoxic-ischemic brain damage (HIBD) is considered as a major cause of long-term cognitive impairments in newborns. It has been demonstrated that gut microbiota is closely associated with the prognosis of various neurological disorders. However, the role of microbiota-gut-brain axis on cognitive function following neonatal HIBD remains elusive. In this experiment, the correlation analysis supported the involvement of gut microbial changes following hypoxic-ischemic (HI) insult in the development of long-term cognitive impairments. Subsequent experiment revealed the involvement of the intestinal dysfunction in the hippocampal neuroinflammation and synaptic injury. In causal relationship validation experiments, fecal microbiota transplantation (FMT) from cognitively normal rats could restore gut microbial composition, improve intestinal dysfunction, reduce the serum levels of lipopolysaccharides (LPS) and inflammatory mediators, and alleviate neuroinflammation, synaptic damage and cognitive impairments in neonatal HIBD recipient rats. Conversely, the FMT from neonatal HIBD rats could induce above adverse pathological changes in the normal recipient rats. Moreover, oral administration of anti-inflammatory agent dexamethasone (DEX) exhibited the potential to alleviate these detrimental effects in neonatal HIBD rats, with the efficacy being partly reliant on gut microbiota. Further experiment on the potential molecular mechanisms using RNA sequencing indicated a significant increase in the toll-like receptor 4 (TLR4) gene in the intestinal tissues of neonatal HIBD rats. Additionally, the interventions such as TLR4 inhibitor TLR4-IN-C34 administration, FMT, and oral DEX were demonstrated to modulate intestinal function by inhibiting the LPS/TLR4 signaling pathway, thereby exerting neuroprotective effects. Collectively, these findings underscore the contribution of gut microbial dysbiosis post HI insult in activating the LPS/TLR4 signaling pathway, triggering intestinal inflammation and dysfunction, exacerbating systemic inflammation, and consequently worsening synaptic and cognitive impairments in neonatal HIBD rats. Hence, rectifying gut microbial dysbiosis or regulating intestinal function may represent a promising strategy for alleviating long-term cognitive impairments in neonates affected by HIBD.

Keywords: Neonatal hypoxic-ischemic brain damage; cognitive impairments; fecal microbiota transplantation; microbiota-gut-brain axis; neuroinflammation.

Graphical abstract

Figure 1.

HI insult induced severe long-term cognitive impairments in the neonatal rats. a The experimental flow chart. b Schematic diagram of the MWM test. c, d the swimming speed and the escape latency of rats in the MWM test from 28 to 32 days after HI insult. e, f The number of crossing the platform times and the time spent in the target quadrant of rats in the MWM test on the 33 days after HI insult. g Representative swim tracks of rats from the place navigation trial on the 32 days after HI insult and the probe test on the 33 days after HI insult. h Schematic diagram of the nor test. i The recognition index in nor test of rats on the 36 days after HI insult. j Schematic diagram of the Y maze test. k The rate of Y maze alternation of rats on the 37 days after HI insult. HI, hypoxic-ischemic; MWM, Morris water maze; NOR, novel object recognition; n = 8, per group; *p < 0.05, ns means no significant.

Figure 2.

HI insult induced gut microbial dysbiosis in neonatal rats. a-c Analysis of the α-diversity of gut microbiota in the feces of two neonatal rat groups by Chao 1, Shannon, Simpson analysis. d Principal coordinate analysis (PCoA) was employed to calculate the Bray-Curtis distance matrix to analyze β-diversity of gut microbiota in the feces of two neonatal rat groups. e Analysis of similarities (ANOSIM) revealed significant differences in gut microbiota composition between the fecal samples of two neonatal rat groups. f Average relative abundances of gut microbiota at the phylum levels. g the abundance values of p_Proteobacteria. h Average relative abundances of gut microbiota at the family levels. i Linear discriminant analysis effect size (LEfSe) was used to identified bacterial species exhibiting significant differences at all taxonomic levels of the gut microbiota of two group’s fecal samples. j Indicator species analysis was used to identified bacterial species exhibiting significant differences at family levels of the gut microbiota of two group’s fecal samples. k-o the abundance values of f_Fusobacteriaceae, f_Enterobacteriaceae, f_Prevotellaceae, f_Akkermansiaceae, and f_Enterococcaceae. HI, hypoxic-ischemic; n = 8, per group; *p < 0.05.

Figure 3.

The correlation analyses supported the involvement of the gut microbiota in the development of long-term cognitive impairments in the neonatal HIBD rats. a-y the correlation analysis between the relative abundance of five significantly altered microbial groups (including f_Fusobacteriaceae, f_Enterobacteriaceae, f_Prevotellaceae, f_Akkermansiaceae, and f_Enterococcaceae) and relevant index data from cognitive behavior tests (including escape latency of D32, the time of crossing, the time spent in target quadrant, recognition index, and Y maze alternation). HIBD, hypoxic-ischemic brain damage; HI, hypoxic-ischemic; D32, the 32 days after HI insult; n = 16 in the correlation analysis.

Figure 4.

HI insult induced intestinal dysfunction, and increased the serum levels of LPS and inflammatory mediators of neonatal rats. a The experimental flow chart. b, c HE staining of the colons and the pathology scores. d-g the expression levels of the tight junction protein occludin (red) and ZO-1 (red) by the if staining in the colons and mean fluorescence intensity of occludin and ZO-1. h-j The expression levels of the IL-17a and IL-22 in the colons with the IHC staining and the mean IOD/area of these two cytokines. k-n the serum concentrations of LPS, tnf-α, IL-6, and IL-1β. o-r the correlation analysis between the concentrations of pro-inflammatory mediators (including LPS and IL-1β) in the serum and the intestinal function indicators (including the mean fluorescence intensity of ZO-1 and the mean IOD/area of IL-17a in the colon. HI, hypoxic-ischemic; LPS, lipopolysaccharide; HE staining, hematoxylin-eosin staining; if staining, immunofluorescence staining; IHC staining, immunohistochemical staining; IOD, integrated optical density; n = 5, per group; n = 10 in the correlation analysis; *p < 0.05.

Figure 5.

HI insult induced hippocampal neuroinflammation, synaptic injury and neuronal damage in neonatal rats. a, b The expression levels of the microglial marker IBA-1 (green) and astrocyte marker GFAP (green) with the if staining in the hippocampal CA1 region. c Confocal images showed an enlarged view of iba-positive microglia in the hippocampal CA1 region and sholl analyses of microglia imaging were used. d-f the calculated mean intensities of IBA-1 and GFAP in per view, and the number of branches of microglia in the hippocampal CA1 region. g The synaptic ultrastructure changes in the hippocampal CA1 region by TEM. h, i the calculated PSD thickness and synaptic cleft width. j the morphology of pyramidal neurons and synaptic structures in the CA1 region of the hippocampus by Golgi staining. k, l the calculated number of dendritic branches and the dendritic spines density. m the neuropathological alterations in the hippocampal CA1 region by Nissl staining. n, o The calculated average number per view of CA1 neurons and the thickness of CA1 were calculated. HI, hypoxic-ischemic; if staining, immunofluorescence staining; TEM, transmission electron microscope; PSD, postsynaptic density; n = 5, per group; *p < 0.05.

Figure 6.

The correlations analyses supported the involvement of the intestinal dysfunction in the hippocampal pathological changes in the neonatal HIBD rats. a-l the correlation analysis between the intestinal function indicators (including the mean fluorescence intensity of ZO-1 and the mean IOD/Area of IL-17a in the colon) and the representative indicators of hippocampal pathological changes (mean intensities of IBA-1 and GFAP per view, endpoints/cell of microglia, thickness of PSD, dendritic spines density, and average number of CA1 neurons per view). Aimed to explore the relation between intestinal function and hippocampal pathological changes induced by HI insult. HIBD, hypoxic-ischemic brain damage; HI, hypoxic-ischemic; PSD, postsynaptic density; IOD, integrated optical density; n = 10 in the correlation analysis.

Figure 7.

The effects of FMT on the intestinal function and serum inflammation levels in the neonatal rats. a The experimental flow chart. b, c Quantification of f_Enterobacteriaceae and f_Akkermansiaceae in fecal samples. d-f the expression levels of IL-17a and IL-22 in the colons with the IHC staining and the mean IOD/area of these two cytokines. g-i The expression levels of the tight junction protein occludin (red) and ZO-1 (red) by the if staining in the colons and mean fluorescence intensity of occludin and ZO-1. j-m the serum concentrations of LPS, tnf-α, IL-6, and IL-1β. FMT, fecal microbiota transplantation; HIBD, hypoxic-ischemic brain damage; HI, hypoxic-ischemic; IHC staining, immunohistochemical staining; IOD, integrated optical density; if staining, immunofluorescence staining; LPS, lipopolysaccharide; n = 5 or 8, per group; *p < 0.05.

Figure 8.

The effects of FMT on the hippocampal neuroinflammation, synaptic structure and long-term cognitive function in the neonatal rats. a The expression levels of microglial marker IBA-1 (green) and astrocyte marker GFAP (green) with the if staining in the hippocampal CA1 region. b, c The mean intensity in per view of IBA-1 and GFAP in the hippocampal CA1 region. d-f The morphology of pyramidal neurons and synaptic structures in the hippocampal CA1 region with the golgi staining and calculated number of dendritic branches and the dendritic spines density. g, h The swimming speed and escape latency of rats in the MWM test from 28 to 32 days after HI insult. i, j The number of crossing the platform and the time spent in the target quadrant of rats in the MWM test on the 33 days after HI insult. k The recognition index in the nor test of rats on the 36 days after HI insult. l The rate of Y maze alternation of rats on the 37 days after HI insult. FMT, fecal microbiota transplantation; HI, hypoxic-ischemic; if staining, immunofluorescence staining; MWM, Morris water maze; NOR, novel object recognition; n = 5 or 8, per group; *p < 0.05, ns means no significant.

Figure 9.

Oral anti-inflammatory agent DEX treatment improved intestinal dysfunction, reduced systemic inflammation, and alleviated synaptic and cognitive impairments in neonatal HIBD rats. a the experimental flow chart. b-d the expression levels of the IL-17a and IL-22 in the colons with the IHC staining and the mean IOD/area of these two cytokines. e-g The expression levels of the tight junction protein Occludin (red) and ZO-1 (red) by the if staining in the colons and mean fluorescence intensity of Occludin and ZO-1. h-k The serum concentrations of LPS, tnf-α, IL-6, and IL-1β. l, m the expression levels of the microglial marker IBA-1 (green) and astrocyte marker GFAP (green) in the hippocampal CA1 region with the if staining. n, o The mean intensities of IBA-1 and GFAP per view in the CA1 region of hippocampus. p-r the morphology of pyramidal neurons and synaptic structures in the CA1 region of hippocampus with the Golgi staining and calculated number of dendritic branches and the dendritic spines density. s the experimental flow chart. t, u The swimming speed and the escape latency of rats in the MWM test from 28 to 32 days after HI insult. v, w the number of crossing the platform and the time spent in the target quadrant of rats in the MWM test on the 33 days after HI insult. x the recognition index in nor test of rats on the 36 days after HI insult. y the rate of Y maze alternation of rats on the 37 days after HI insult. DEX, dexamethasone; HIBD, hypoxic-ischemic brain damage; HI, hypoxic-ischemic; IHC staining, immunohistochemical staining; IOD, integrated optical density; if staining, immunofluorescence staining; LPS, lipopolysaccharide; MWM, Morris water maze; NOR, novel object recognition; n = 5 or 8, per group; *p < 0.05, ns means no significant.

Figure 10.

The FMT from the HI+DEX group donors alleviated hippocampal neuroinflammation, synaptic injury, and cognitive impairments in the neonatal HIBD rats. a The experimental flow chart. b, c Quantification of f_Enterobacteriaceae and f_Akkermansiaceae in fecal samples. d, e The expression levels of the microglial marker IBA-1 (green) and astrocyte marker GFAP (green) in the hippocampal CA1 region with the if staining. f, g The mean intensities of IBA-1 and GFAP per view in the CA1 region of hippocampus. h-j The morphology of pyramidal neurons and synaptic structures in the hippocampal CA1 region with the Golgi staining and calculated number of dendritic branches and the dendritic spines density. k, l The swimming speed and the escape latency of rats in the MWM test from 28 to 32 days after HI insult. m, n the number of crossing the platform and the time spent in the target quadrant of rats in the MWM test on the 33 days after HI insult. o the recognition index in nor test of rats on the 36 days after HI insult. p the rate of Y maze alternation of rats on the 37 days after HI insult. FMT, fecal microbiota transplantation; HI, hypoxic-ischemic; DEX, dexamethasone; if staining, immunofluorescence staining; MWM, Morris water maze; NOR, novel object recognition; n = 5 or 8, per group; *p < 0.05, ns means no significant.

Figure 11.

Oral TLR4-IN-C34 treatment improved intestinal dysfunction and alleviated synaptic and cognitive impairments in the neonatal HIBD rats. a the experimental flow chart. b the cluster heat map showing differentially expressed RNA between the sham and HI groups (n = 4). c the volcano plots displayed the DEGs as red dots (upregulated) and blue dots (downregulated). d significantly upregulating DEGs were categorized into classes based on GO enrichment terms. e-h the expression levels of TLR4 of colon by WB and IF analyses. i-k the expression levels of the IL-17a and IL-22 in the colons with the IHC staining and the mean IOD/area of these two cytokines. l-n the expression levels of the tight junction protein Occludin (red) and ZO-1 (red) by the if staining in the colons and mean fluorescence intensity of Occludin and ZO-1. o-q the morphology of pyramidal neurons and synaptic structures in the hippocampal CA1 region with the Golgi staining and calculated number of dendritic branches and the dendritic spines density. r, s The swimming speed and the escape latency of rats in the MWM test from 28 to 32 days after HI insult. t, u The number of crossing the platform and the time spent in the target quadrant of rats in the MWM test on the 33 days after HI insult. v the recognition index in nor test of rats on the 36 days after HI insult. w the rate of Y maze alternation of rats on the 37 days after HI insult. TLR4, toll-like receptor 4; HIBD, hypoxic-ischemic brain damage; GO, gene ontology; DEGs, differentially expressed genes; HI, hypoxic-ischemic; WB, western blot; IF, immunofluorescence; IHC staining, immunohistochemical staining; IOD, integrated optical density; if staining, immunofluorescence staining; MWM, Morris water maze; NOR, novel object recognition; n = 4, 5, 6 or 8, per group; *p < 0.05, ns means no significant.

Figure 12.

Schematic diagram of the findings from this study. The gut microbial dysbiosis following the HI insult activates LPS/TLR4 signaling pathway, leading to intestinal inflammation and intestinal barrier damage. This intestinal dysfunction can elevate serum levels of pro-inflammatory mediators, including LPS from gut microbiota and tnf-α, IL-6, and IL-1β secreted by intestinal epithelial cells, thereby activating microglia and astrocytes in the hippocampus via the microbiota-gut-brain axis. Consequently, this process can exacerbate hippocampal neuroinflammation, synaptic damage, and long-term cognitive impairments induced by HIBD. Rectifying the gut microbial dysbiosis by the FMT or oral DEX can counteract these detrimental effects and alleviate long-term cognitive impairments caused by neonatal HIBD. Moreover, oral TLR-IN-C34 treatment also can exert neuroprotective effect by inhibiting the intestinal LPS/TLR4 signaling pathway. HI, hypoxic-ischemic insult; LPS, lipopolysaccharide; TLR4, toll-like receptor 4; HIBD, hypoxic-ischemic brain damage; FMT, fecal microbiota transplantation; DEX, dexamethasone; by Figdraw.