Repetitive, mild traumatic brain injury results in a progressive white matter pathology, cognitive deterioration, and a transient gut microbiota dysbiosis

Abstract

Traumatic brain injury (TBI) is often accompanied by gastrointestinal and metabolic disruptions. These systemic manifestations suggest possible involvement of the gut microbiota in head injury outcomes. Although gut dysbiosis after single, severe TBI has been documented, the majority of head injuries are mild, such as those that occur in athletes and military personnel exposed to repetitive head impacts. Therefore, it is important to determine if repetitive, mild TBI (rmTBI) will also disrupt the gut microbiota. Male mice were exposed to mild head impacts daily for 20 days and assessed for cognitive behavior, neuropathology and disruptions in the gut microbiota at 0, 45 or 90 days after injury. Deficits in recognition memory were evident at the late post-injury points. Brains show an early increase in microglial activation at the 0-day time point that persisted until 90 days post-injury. This was compounded by substantial increases in astrocyte reactivity and phosphorylated tau at the 90-day time point. In contrast, changes in the microbial community were minor and transient, and very few differences were observed in mice exposed to rmTBI compared to controls. While the progressive emergence of white matter damage and cognitive alterations after rmTBI resembles the alterations observed in athletes and military personnel exposed to rmTBI, these changes could not be linked to systematic modifications in the gut microbiota.

Figure 1

Effect of rmTBI on ROR. Mice were subjected to 20 head impacts and ROR was recorded after each impact daily. Controls were anesthetized but were not subjected to head impacts. The rmTBI group contained 18 mice and the control group contained 17 mice. The symbols indicate ****p < 0.0001; ***p < 0.001 and **p < 0.01.

Figure 2

Effects of rmTBI on recognition memory (NOR test) at 0, 45, and 90 days post-injury. The percentage of time spent exploring each object (novel vs familiar) was plotted using the total time investigating to normalize measures among individuals. Mice were subjected to 20 head impacts and tested in the NOR paradigm prior to sacrifice at each time point. Each group contained 5–6 subjects and bars represent mean ± SEM values. The symbols indicate ****p < 0.0001; ***p < 0.001 and *p < 0.05.

Figure 3

Effects of rmTBI on astrocyte (a), microglial (c) and p-tau (e) immunoreactivity in the optic tract. Levels of GFAP (b), Iba-1 (d) and p-tau (f) among the treatment groups quantified by pixel density from immunohistochemical analyses. Mice were subjected to 20 head impacts and brains were harvested at the indicated times and subjected to immunohistochemical analysis using antibodies against astrocytes (GFAP), microglia (Iba-1) or p-tau (AT8). Results are presented as means ± SEM. Groups are indicated as follows: Con-0 (control-0 days), TBI-0 (rmTBI-0 days), Con-45 (control-45 day time point), TBI-45 (rmTBI-45 day time point), Con-90 (control-90 day time point) and TBI-90 (rmTBI-90 day time point). The symbols indicate the levels of significance as follows for the indicated comparisons: ****p < 0.0001, ***p < 0.001, **p < 0.01 and *p < 0.05.

Figure 4

Effects of rmTBI on α- and β-diversity. Data for α-diversity are presented as the indices of Chao-1 (a), Simpson 1-D (b) and Shannon (c). Data are plotted as mean ± SEM. Each group contained 5–6 mice. The symbol * in panel C represents the significance level for the indicated post hoc comparisons of p < 0.05. Data for β-diversity (d) is presented as a PCoA and shows differences in the similarities of the gut microbiota clustering profiles for rmTBI and control groups for the 0 day, 45 day and 90 day time points using the Bray Curtis index for community structure.

Figure 5

Bacterial taxa that were differentially abundant across treatments. LEfSe was carried out using the Galaxy Project and the results are displayed as bars the length of which is indicative the linear discriminant analysis score for each ASV. The taxonomic identity of each ASV is indicated to the left of each bar. All groups are statistically significant compared to each other (LDA > 3.0).

Figure 6

Percent relative abundances of phyla in the treatment groups. Stacked columns for the 10 most prominent bacterial phyla are included (a). The relative abundance of taxa below the level of phylum (b) are presented as mean % ± SEM. The symbols indicate the levels of significance as follows for the indicated comparisons: ****p < 0.0001 and *p < 0.05.