Secondary Logo

Journal Logo

Moderate Traumatic Brain Injury Alters the Gastrointestinal Microbiome in a Time-Dependent Manner

Nicholson, Susannah E.*; Watts, Lora T.; Burmeister, David M.; Merrill, Daniel*; Scroggins, Shannon*; Zou, Yi§; Lai, Zhao§; Grandhi, Ramesh*; Lewis, Aaron M.*; Newton, Larry M.*; Eastridge, Brian J.*; Schwacha, Martin G.*,‡

doi: 10.1097/SHK.0000000000001211
Basic Science Aspects
Free
Editor's Choice

ABSTRACT The microbiome is defined as the collective genomes of the microbes (composed of bacteria, bacteriophage, fungi, protozoa, and viruses) that colonize the human body, and alterations have been associated with a number of disease states. Changes in gut commensals can influence the neurologic system via the brain-gut axis, and systemic insults such as trauma or traumatic brain injury (TBI) may alter the gut microbiome. The objective of this study was to evaluate the gut microbiome in a preclinical TBI cortical impact model. Male rats underwent craniotomy and randomized to a sham group (n = 4), or a moderate TBI (n = 10) using a pneumatic impactor. MRI and behavioral assessments were performed pre-TBI and on days 2, 7, and 14 days thereafter. Microbiome composition was determined with 16s rRNA sequencing from fecal sample DNA pre-TBI and 2 hrs, 1, 3, and 7 days afterward. Alpha- and β-bacterial diversity, as well as organizational taxonomic units (OTUs), were determined. Significant changes in the gut microbiome were evident as early as 2 h after TBI as compared with pre-injured samples and sham rats. While there were varying trends among the phylogenetic families across time, some changes persisted through 7 days in the absence of therapeutic intervention. While large structural lesions and behavioral deficits were apparent post-TBI, there were modest but significant decreases in α-diversity. Moreover, both changes in representative phyla and α-diversity measures were significantly correlated with MRI-determined lesion volume. These results suggest that changes in the microbiome may represent a novel biomarker to stage TBI severity and predict functional outcome.

*Department of Surgery, The University of Texas Health Science Center at San Antonio, San Antonio, Texas

Department of Cellular and Structural Biology, The University of Texas Health Science Center at San Antonio, San Antonio, Texas

The US Army Institute of Surgical Research, Fort Sam Houston, Texas

§Greehey Children's Cancer Research Institute, The University of Texas Health Science Center at San Antonio, San Antonio, Texas

Address reprint requests to Susannah E. Nicholson, MD, Department of Surgery, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dr., San Antonio, TX 78229. E-mail: Nicholson@uthscsa.edu

Received 21 November, 2017

Revised 7 December, 2017

Accepted 19 June, 2018

The project described was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant KL2 TR001118. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Department of the Army and the Department of Defense. Support was also received by the University of Texas Health Science Center at San Antonio Military Health Institute, the Bob Kelso Endowment awarded to the UTHSCSA Department of Surgery and The Greehey Children's Cancer Research Institute Genome Sequencing Facility's Illumina MiSeq Pilot Grant.

The authors report no conflicts of interest.

Back to Top | Article Outline

INTRODUCTION

Traumatic brain injury (TBI) is a leading cause of death in both children and young adults with approximately 2.5 million TBI cases occurring annually in the United States (1). TBI is a complex injury that induces progressive deterioration of brain physiology resulting in varied functional outcomes. The initial physical impact causes direct damage to the brain and is referred to as the primary injury phase. During the secondary injury phase in the subsequent hours and days following a TBI, a myriad of molecular changes lead to mitochondrial dysfunction, neuronal degeneration, cerebral blood flow disruption, inflammation, blood–brain barrier dysfunction, and edema formation (2). These changes contribute to permanent neuronal loss and are responsible for disability, altered quality of life, and mortality (2). Management of patients with TBI focuses on prevention of secondary injury, but TBI progression can be difficult to follow clinically. More reliable methods are needed to better identify patients at risk for secondary injury. Currently, the Glasgow Coma Scale (GCS), loss of consciousness, pupil reactivity, and head computed tomography (CT) are the primary clinical tools used for the early detection of brain injury; however, they have limited sensitivity to predict adverse secondary events or detect subtle damage.

The microbiome is defined as the collective genomes of the microbes (composed of bacteria, bacteriophage, fungi, protozoa, and viruses) that form an ecological community inside and on the human body. The gut microbiota contains tens of trillions of microorganisms, including at least 1,000 different species of known bacteria that outnumber somatic and germs cells (3). Microbiotas have been shown to influence human physiology, metabolism, nutrition, and immune function. The composition and diversity of the GI microbiome play an integral role in intestinal homeostasis, metabolism of nutrients, synthesis of vitamin K, development of organ structures, development and maintenance of the intestinal barrier, and the induction of immunity (4, 5). Recently, disruption of the gut microbiome has been linked to cardiovascular disease, inflammatory bowel disease, obesity, Parkinson disease, and autism (6).

Communication between the central nervous system and the gastrointestinal tract occurs via the brain-gut axis, which is composed of a complex network involving neuroendocrine and immunological signaling pathways and direct neural mechanisms. Preclinical models have demonstrated that the GI microbiota may influence the brain-gut axis through all of these pathways (7). In mice, the gut microbiota has been shown to influence permeability of the blood–brain barrier (8). Several preclinical studies have found that stroke injury alters microbial composition and decreases diversity in the gut (9, 10). Alterations in gut microbial populations have also been demonstrated in mice receiving a closed-head mild TBI (10).

In response to an injury, the body's immune system mounts an inflammatory response, which causes a dysbiosis in the microbiome that further amplifies the inflammatory response triggering an even greater imbalance in the microbiome (10, 11). These changes in the gut commensals can then, in turn, influence the neurologic system via feedback through the brain-gut axis and may further affect outcome following TBI. Recent advances in metagenomics sequencing have allowed characterization of the human microbiome beyond the capacity of traditional culture techniques, and have identified the role of microbiota in normal physiologic, immunologic, and metabolic functions. Effort has also focused on the role of commensals in human health and disease; yet data is sparse concerning the microbiome in the context of injury, especially TBI. Given the interaction between the brain and the GI tract through the various components of the brain-gut axis, we sought to determine whether the GI microbiome is affected by moderate TBI. We hypothesized that TBI alters the GI microbiome in a time-dependent manner, which correlates with structural and functional characteristics of the brain injury.

Back to Top | Article Outline

MATERIALS AND METHODS

Experimental model of TBI

All procedures were approved by the Institutional Animal Care and Use Committee of The University of Texas Health Sciences at San Antonio. The open skull cortical impact model was performed as previously described in Watts et al. (12). Briefly, Sprague Dawley rats (225–250 g; n = 10) were anesthetized with 1.1% isoflurane, placed in a stereotaxic frame and underwent a 5 mm craniotomy over the S1 region exposing the dura matter. The dura was directly impacted with a pneumatic impactor (5.0 m/s; 250 μs dwell time and 2 mm depth) to induce a moderate TBI. Sham rats (n = 4) underwent craniotomy without TBI. Following the TBI procedure, rats were moved to the MRI scanner. MRI was used to measure changes in cerebral blood flow, vascular reactivity, diffusion-tensor, and T2 MRI on the day of the TBI procedure and on days 2, 7, and 14 days after onset of TBI to longitudinally track injury progression. Heart rate, respiratory rate, rectal temperature, pulse distention, breath distention, and O2 saturation were continuously monitored and maintained within normal physiological range. MRI was performed on a rodent Bruker Pharmascan 7-T/16-cm under 1.1% isoflurane. Continuous arterial spin labeling technique with a separate neck-labeling coil was used. T2-weighted images were acquired using fast spin-echo sequence, with maps calculated as previously described (13, 14). The image maps of individual animals were coregistered across time points using a transformation matrix generated by QuickVol and MRIAnalysisPak software (15). The regions of interests were defined by pixels that had T2 values higher than the mean plus two standard deviations of the value in the homologous contralesional region. Comparison of MRI scans was made with the evolution of lesion volume, behavioral analysis, and gut microbiome measures.

Back to Top | Article Outline

Behavioral assessments

Behavioral assessments were used to detect functional deficits at corresponding MRI time points (1–3 days prior to TBI and again 1, 2, 7, and 14 days post TBI), as previously described in Watts et al. (12). Behavioral analysis included: Asymmetry (cylinder) test to measure asymmetrical use of the forelimbs for postural support during spontaneous exploration and Foot fault test to assess contralateral and ipsilateral motor impairments of limb functioning and placing deficits during locomotion (12).

The forelimb asymmetry placement test was performed with videotaping. Briefly, the rat was placed in a transparent cylinder (20 cm diameter, 30 cm height) for 5 min or until 30 placements were made. The behavior was scored by counting the number of left or right individual forelimb placements, and the number of simultaneous right and left (both) forelimb placements onto the wall of the cylinder during rearing. The forelimb asymmetry index was calculated as (the number of forelimb placements for each individual limb) + ½ (number of both placements) divided by the total number of placements.

The foot fault test was performed by placing the rat on an elevated grid floor (e.g., size 18 in × 11 in with grid openings of ∼1.56 in2 and 1 in2) for 5 min or until 50 steps were taken with one (right hind) limb. The rat was allowed to move freely on the grid and the total number of steps and the number of times each limb fell below the grid opening were counted. The number of foot faults for each limb was calculated by counting the number of right or left forelimb or hind-limb foot faults.

Back to Top | Article Outline

Microbiome analysis

All rats were fed the same diet throughout the study. Fecal samples (i.e., excreted fecal pellets) were collected prior to the initial TBI and then at 2 h, 1, 3, and 7 days following TBI. Fecal samples were also collected from all sham rats at corresponding time points. DNA was purified from all fecal samples using the Qiagen DNA Stool Mini Kit.

Genomic DNA was then quantified and the V1–V3 variable region of the 16S rRNA genes amplified with custom-designed primers (F27/R534). The Forward Bosshard sequence was AGAGTTTGATCMTGGCTCAG (27F) and the Reverse Bosshard sequence was GTATTACCGCTGCTG (534R) with the amplicon size of V1–V3 about 510 bp (534-27) (9). Libraries for all samples were prepared and sequenced by Illumina 600 cycle V3 kit Paired-end sequencing (2 × 300 bp) using the Illumina MiSeq platform. The average of 392,577 raw reads per sample was generated with read length of 301bps. Raw sequences were quality trimmed (Q20) by sickle, and reads shorter than 200 bases were removed. Therefore, only the forward read was used for this analysis. The trimmed sequences were exported as FASTA files. Subsequently, FASTA files were processed through the software package Quantitative Insights Into Microbial Ecology (QIIME). The operational taxonomic units (OTU) were clustered based on at 97% similarity and aligned to the Greengenes database according to standard phylogenetic methods (16). Alpha rarefaction was performed using the observed species metrics and α-diversity was evaluated using the following indices: PD Whole Tree, Chao1, observed species number, and Shannon Index. PD Whole Tree measures phylogenetic diversity, or the minimum number of phylogenetic tree branches required to span a given set of taxa (17). Chao1 and observed species measure the total number of species, with the Chao index being particularly useful for data with low-abundance organisms. Shannon Index is a measure of species richness. Beta diversity was measured with the Greengenes-aligned OTUs estimated by computing weighted and unweighted UniFrac distances to calculate species composition similarities and differences. STAMP and GraphPad Prism were used for the visualization and the statistics of the comparative metagenomic data sets.

Back to Top | Article Outline

Statistics

One-way analysis of variance (ANOVA) with Sidak multiple comparisons test was calculated at the phylum and family level to compare changes at all time points, as well as the volume of MRI lesions. Two-way ANOVA was also used to evaluate the α-diversity indices over time as well as behavioral functional outcomes. A permutational multivariate ANOVA (PERMANOVA) was performed to determine differences in β-diversity (i.e., PCA plot) over time. Linear regression was performed to identify relationships between the microbiome, MRI lesions volume and behavioral data. All data is represented as Mean ± SD, and P < 0.05 was used as the alpha-value.

Back to Top | Article Outline

RESULTS

MRI determined lesion volumes

The images in Figure 1A show representative multislice T2 maps from a TBI rats at 3 h, 1, 2, 7, and 14 days post-TBI. The T2 MRI at 3 h post-TBI showed changes within the area underlying the somatosensory cortex (impacted area). The T2 determined lesion volumes increased significantly from 5.2 ± 0.43 mm3 3 h postinjury to 8.6 ± 1.1 mm3 on day 2 postinjury (Fig. 1B). The lesion volume slightly declined by day 7 to 6.6 ± 1.0 mm3 (Fig. 1B). The T2 map below the impacted area showed heterogeneous contrast at the various time points evaluated, with hyperintensity indicating vasogenic edema and hypointensity indicating possible hemorrhage. The average ipsilesional T2-weighted intensity measures demonstrated a significant increase in intensity measurements from sham (63.7 ± 0.6) to TBI (93.2 ± 1.7) (Fig. 1C, P < 0.0001). The data suggest an increase in vasogenic edema formation in the TBI animals in the area underlying the initial impact area. The average contralesional T2-weighted intensity measures were non-significant with sham animals measuring 66.1 ± 0.1 and TBI animals measuring 65.2 ± 1.3 (Fig. 1C). This heterogeneity was evident within 1 h after TBI induction and continued to increase through day 2 before returning toward baseline levels by day 14.

Fig. 1

Fig. 1

Back to Top | Article Outline

Behavioral assessments

Sensorimotor function was assessed using the asymmetry forelimb placement (cylinder) test and foot-fault test verify functional deficits following TBI. The mean scores for the behavioral tests (foot fault and forelimb asymmetry) are shown in Figure 2. The average number of pre-TBI foot fault scores was 1 ± 0.25 and 2 ± 0.48 for sham and TBI animals, respectively. This difference in foot fault scores was not significant and is a normal variation seen among animals. Following TBI, foot fault scores markedly worsened in the affected forelimb on day 2 (6.0 ± 0.9, P < 0.05 compared with shams). These scores improved slightly by 7 and 14 days post-TBI (Day 7; 5.0 ± 1.0 and Day 14; 5 ± −0.8; P < 0.05), but remained significantly higher than sham animals (Fig. 2A). Pre-TBI forelimb asymmetry was 50% ± 0.3% for shams and 49% ± 1.0% for TBI animals, indicating symmetrical use of both forelimbs prior to injury (Fig. 2B). The asymmetry score was the worst on day 2 post-TBI (50% ± 0.4% for Sham animals and 63% ± 6.5% for TBI animals, respectively). This indicates increased utilization of the unaffected forelimb (left forelimb); however, these values were not significantly different from the sham animals.

Fig. 2

Fig. 2

Back to Top | Article Outline

Sample sequencing and OTU analysis

The NGS generated a maximum of 918,424 raw sequences (mean of 392,577 sequences per sample) with a read length of 301 base pairs. After 3’-end trimming by sickle (Q ≥ 20 and length ≥ 200), 699,097 high quality reads remained (mean of 258,441 sequences per sample). The sequence reads and OTUs by time point are listed in Table 1. OTUs clustered based on 97% similarity demonstrated that the number of OTUs on day 3 were significantly less than pre-TBI levels, indicating that the total number of taxonomic units represented microbial species changed after 3 days postinjury.

Table 1

Table 1

Back to Top | Article Outline

α and β diversity analyses

The α-diversity, as measured by the Chao1 index and total species number, was significantly lower on day 3 compared with the pre-TBI rats demonstrating that the microbial species diversity was decreased by TBI. The PD Whole Tree Index and the Shannon Index demonstrated a trend toward lower number of species by day 3, but this measure of α-diversity did not reach significance (Fig. 3). Conversely, differences in β-diversity and microbial profile, as depicted in the principle components analysis (PCA) plot, were observed over time following TBI, as shown in Figure 4. Each point on the plot in Figure 4 represents the gut microbial composition of an individual rat, with red points representing the animals prior to TBI. Prior to TBI, variance in the microbial composition between animals is clearly evident, with not much change at 2 h post-TBI (green points). By 1 day after TBI (blue points), a shift in microbial composition occurs, primarily along PC1, which accounted for 38% of the inter-sample variation. This pattern is maintained by 3 days postinjury (purple points) with a return to a more normal (i.e., pre-TBI) composition by day 7 (yellow points). While PERMANOVA analysis confirmed a significant effect within each animal (P < 0.05), the effect of time did not quite reach statistical significance. Nevertheless, these PCA results suggest that TBI disrupts the GI microbiome.

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Back to Top | Article Outline

Organism classifications

The representative and most abundant phyla included Firmicutes, Bacteroidetes, Verrucomicrobia, Proteobacteria, Cyanobacteria, Tenericutes, and Deferribacteres in all samples, irrespective of the time after injury (Fig. 5A). The dominant phyla in both the pre-TBI and the post-TBI groups were Firmicutes and Bacteroidetes. These two phyla comprised 89.8% of the bacteria in the pre-TBI group, and total levels of these two phyla decreased by about 10% by day 3 post-TBI, even though there was an overall increase in Bacteroidetes. Conversely, the phylum Proteobacteria, containing many pathogenic bacteria, more than doubled during the same time post-TBI. The ratio of Firmicutes to Bacteroidetes decreased significantly with the nadir occurring on day 1 post-TBI (2.4 for pre-TBI vs 1.1 for 3 days post-TBI; P = 0.012). Similar data for sham animals are shown in Figure 5B. There was no difference in the GI microbiome between sham and TBI animals prior to craniotomy/TBI. Additionally, in sham animals, the microbiome phyla remained relatively unchanged over time, with the lone exception of a small decrease in the phylum Bacteroidetes on day 3 post-craniotomy compared to pre-craniotomy levels. This finding is the opposite of what is observed in TBI animals, and otherwise there was no change in the microbiome in sham animals (Fig. 5B).

Fig. 5

Fig. 5

The data in Table 2 shows specific changes in the families of each phylum after TBI. There were significant decreases in relative abundance at the phylum level seen in Firmicutes at days 1 and 3 and Deferribacteres at day 1 following TBI (P < 0.05; Table 2 and Fig. 5A). Conversely, an increased relative abundance in the phyla Bacteroidetes and Deferribacteres was observed at day 1 following TBI (P < 0.05; Table 2 and Fig. 5A). Within the phylum Firmicutes, there was an initial decrease of ∼50% in proportion of sequences in the family Lachnospiraceae at 2 h after TBI, which remains suppressed through 3 days and returned to baseline by 7 days. The Mogibacteriaceae family was also significantly decreased (P < 0.05) at 1 and 3 days, and returned to baseline levels by 7 days. Ruminococcaceae levels decreased at 1 day after TBI only. Lactobacilliaceae levels were not significantly different after TBI. Levels of bacteria from the family Anaeroplasmataceae, in the phylum Tenericutes decreased by ∼60% at 2 h after TBI and remained markedly decreased over time. Deferribacteraceae (a member of the Deferribacteria phylum) levels), were reduced ∼75% at 1 and 3 days and approached baseline by 7 days after TBI). The Bacteroidete family Bacteroidaceae demonstrated a 50% increase in the proportion of sequences by 1 day after TBI with a return to baseline by 7 days). Verrucomicrobia, from the phylum Verrucomicrobia, demonstrated a similar trend with a decrease seen at 1 day and return to baseline levels by 7 days). Enterobacteriaceae and Pseudomonadaceae, members of the Proteobacteria phylum, all demonstrated markedly increased levels by 3 days post-TBI with all levels in these families returning to baseline by 7 days (P < 0.05).

Table 2

Table 2

Back to Top | Article Outline

Correlative analysis of MRI lesion volume, behavioral assessments, and the GI microbiome

To directly compare the magnitude of functional and structural changes to the microbiome alterations, linear regression was performed. Linear regression revealed that changes in both Firmicutes (y = −7.46x+110.7, r2 = 0.914, P < 0.0001) and Proteobacteria (y = 3.29x−15.61, r2 = 0.805, P = 0.001; Fig. 6A) were strongly correlated to MRI lesion volume, with a weaker but still significant relationship to the phylum Verrucomicrobia (y = 0.96x+0.88, r2 = 0.55, P = 0.02). Moreover, regression analysis also revealed that larger brain lesion volume resulted in a significantly greater reduction in α-diversity by all indices measured (Fig. 6B). Although not quite as strongly, changes in gut microbial phyla and diversity were also associated with behavioral/functional outcomes as measured by the foot fault test. Specifically, the foot fault scores also tended to positively correlate to Proteobacteria, although this was not quite significant (P = 0.081). However, there was a significant correlation to the phylum Verrucomicrobia (y = 1.25x−0.615, r2 = 0.773, P = 0.0018, Fig. 6C), as well as significant negative correlations to α-diversity measures (Fig. 6D).

Fig. 6

Fig. 6

Back to Top | Article Outline

DISCUSSION

The current study was designed to detect changes in the GI microbiome due to a moderate-TBI injury over time. Significant changes in the GI microbiome were evident as early as 2 h after TBI as compared with pre-injured samples and sham rats, with varying trends among the phylogenetic families. This indicates that TBI alters the gastrointestinal microbiome and creates a dysbiosis in the absence of other injury patterns, resuscitation, antibiotics, and analgesics. Peak MRI lesion volume, functional deficits, microbial composition alterations, and the greatest reduction in α-diversity occurred in a similar timeline (i.e., between 2 and 3 days). Perhaps the most salient finding is that the evolution of the lesion directly correlated with changes in the GI microbiome. Specifically, a larger brain lesion was associated with greater decreases in levels of Firmicutes (traditionally beneficial bacteria) and an exacerbated increase in Proteobacteria (with many pathogenic bacteria). Furthermore, a larger brain lesion size was associated with a significant reduction in α-diversity. To our knowledge, this is the first such finding directly correlating GI microbiome changes to lesion size due to TBI.

A decrease in traditionally beneficial bacteria at the phylum and family levels was observed after TBI with parallel increases in families that contain opportunistic pathogens including Bacteroidaceae, Enterobacteriaceae and Pseudomonadaceae. These findings herein are consistent with previous research in human burn victims, which showed a decreased overall diversity of the microbiome and an associated increase in gut permeability (11). This group also found an increase in gram-negative bacteria, particularly in the family Enterobacteraceae. Krezalek et al. found that the intestinal microbiota was disrupted within hours of injury, and that these disturbances were predictive of sepsis-associated mortality in this patient population. Moreover, in patients who developed sepsis, the composition of the gut microbiome dramatically changed such that pathogen communities of extremely low diversity emerged and triggered further virulence in the host leading to the development of a pathobiome (18). In a model of middle cerebral artery stroke, Singh et al. (9) similarly observed reduced diversity of the gut microbiome, with particular decreases in the phylum Bacteroidetes.

The normal microbiome is dominated by the Bacteroidetes and Firmicutes phyla, with more than 90% of species falling within these groups (3, 4, 6). Other representative phyla in the gut microbiome include Actinobacteria, Fusobacteria, Proteobacteria, Verrucomicrobia, and Cyanobacteria(6). Native gut bacteria provide many benefits in that they metabolize indigestible polysaccharides, produce essential vitamins, maintain tissue homeostasis, and protect against invasion of pathogens (19). In Clostridium difficile colitis, there is an increase in Proteobacteria and decreases in the phyla Bacteroidetes and Firmicutes(20). Moreover, an increased ratio of Firmicutes to Bacteroidetes is observed in obesity that is thought to be related to the promotion of Firmicutes from a high fat diet leading to increased caloric intake (21). Alternatively, we found a reduced ratio of Firmicutes to Bacteroidetes, which may be a result of the stress response following TBI.

Reductions in beneficial commensal bacterial populations have been associated with acute and chronic disease states (3). Similarly, we observed decreases in families within the Firmicutes phylum including Lachnospiraceae, Mogibacteriaceae, and Ruminococcaceae, traditionally believed beneficial families of bacteria. Bacteria in the gram-positive family Lachnospiraceae, which includes Clostridia species, generate butyrate by fermenting carbohydrates. Lachnospiraceae have been shown to prevent inflammation in colitis models, and decreased levels are found in inflammatory bowel disease (5, 22). Other beneficial families including Anaeroplasmataceae and Verrucomicrobiaceae were also reduced in animals sustaining TBI in our study. Recent research has shown a correlation between changes in Verrucomicrobia and glucose tolerance, obesity, and diabetes (23). Our data suggest that similar changes observed in our model may potentially contribute to secondary brain injury following TBI.

The Bacteroidaceae family typically acts as one of the dominant families endemic to the human microbiome but also includes the pathogenic bacteria Bacteroides fragilis, whose enterotoxigenic strain is commonly associated with diarrhea in inflammatory bowel disease and colorectal cancer (24). Members of the phylum Proteobacteria, particularly those within the family Enterobacteriaceae, contain many opportunistic pathogenic bacteria, including those from the genera Escherichia, Klebsiella, Proteus, and Citrobacter, which are commonly seen in sepsis (24). Proteus mirabilis and Klebsiella pneumonia, bacteria in the family Enterobacteriaceae, are associated with colitis and can elicit inflammation and spontaneous colitis when transferred to wild-type mice (25). In burn patients, there is an overgrowth of Enterobacteriaceae and Bacteroidaceae(11). Increased abundance of traditionally opportunistic bacteria was also observed in this study, with increases of Bacteroidaceae, Enterobacteriaceae, and Pseudomonadaceae. Moreover, the extent of the brain lesion in our model positively correlated with the increase in Proteobacteria, which includes two of those families suggesting that increased pathogenic flora may contribute to inflammatory and infectious complications. A feedback loop involving the virulent bacterial species within the GI microbiome and the brain-gut axis may also potentiate a neuroinflammatory cascade, leading to secondary brain injury and thus influencing functional outcome.

Specific to the brain-gut axis, increases of Enterobacteriaceae in cirrhotic patients are associated with hyperammonemia-astrocytic changes, hepatic encephalopathy, higher model end stage liver disease (MELD) scores, increased brain magnetic resonance spectroscopy manifestations, and increased systemic and local inflammatory effects (26). The ability of these bacteria to produce toxins that increase levels of gut ammonia and potentiate other metabolic and systemic effects may also play a key role in negative sequelae in patients with TBI via the brain-gut axis. Furthermore, in patients with Parkinson the abundance of Enterobacteriaceae is positively associated with postural instability and gait disturbance severity (27). Opportunistic bacteria such as Pseudomonas aeruginosa play a key role in a number of nosocomial infectious complications such as urinary tract infections and surgical site infections (28). We also found increases in the opportunistic families Pseudomonadaceae and Enterobacteriaceae suggesting that they may also contribute to negative secondary sequelae and outcomes in TBI patients.

Preclinical models have demonstrated that the gastrointestinal microbiota may influence the brain-gut axis via immunological, neuroendocrine, and direct neural mechanisms (7, 29). Afferent and efferent connections from vagal nuclei compose the enteric nervous system and directly connect the brain to the gut (29). Furthermore, neuroendocrine and immunologic communication from areas of the brain such as the hypothalamus, insular cortex, and the cingulate is integrated with visceral signaling via the vagus complex and the pontine nucleus tractus solitaries (29). TBI can disrupt the corticopontine communication between the intestine and the vagal complex causing dysautonomia, which is characterized by tachycardia, hyperthermia, hypertension, increased sweating, muscle tone, and posturing (30). The release of stress hormones in the hypothalamic–pituitary–adrenal axis following injury can also result in intestinal changes in permeability, motility, and secretion (31). Future investigation will help determine whether these phenomena are involved in our model.

Improvements in emergency response times and care have increased survivability and have brought to light an increasing need for developing reliable methods to identify patients at risk of developing secondary complication pathologies. Objective diagnosis and classification of TBI can be challenging with the current clinical modalities. While head computed tomography (CT) and MRI are excellent at determining the presence and extent of hemorrhage, edema, and other physical changes in the brain, they have limited ability to predict adverse secondary events or to detect subtle damage. Moreover, current methods of classification for TBI include the Glasgow Coma Scale, Abbreviated Injury Scale, and the Abbreviated Trauma Scale, all of which struggle to clearly define the severity of TBI due to the heterogeneous presentations. Recent studies have identified several protein molecules that may serve as candidate biomarkers in TBI (32, 33). Recent publications from the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) Study have demonstrated the utility of sampling plasma for glial, axonal, and neuronal breakdown molecules following TBI as biomarkers (34, 35). To this end, changes in the microbiome may represent a novel biomarker to stage TBI severity and predict functional outcome, especially when viewed in the context of the brain-gut axis. By characterizing the changes in bacterial families and degree of loss of diversity, it may be possible to correlate TBI severity to the specific changes in the microbiome. Therefore, future studies will focus on determining the impact of varying the severity of TBI on changes in the microbiome.

While the findings of our study support a relationship between TBI and the gut microbiome, there are limitations. This is a preclinical model of TBI in rodents, so conclusions drawn may not translate to patients with TBI. Further clinical studies are needed to better assess the effects on TBI and the microbiome. In addition, gnotobiotic rodents were not utilized in these experiments, but all comparisons were made between rats with TBI and shams with all reported changes reaching significance. As mentioned, PERMANOVA demonstrated that β-diversity was significantly altered within individual animals over time, but failed to reach significance when time points were compared. Given the trend, significance would likely be reached at days 1 and 3 with a larger sample size. In addition, this model utilized a moderate TBI, and given the important relationships with lesion volume, a staged model incorporating mild and severe TBI would likely prove revealing. While the effects of the craniotomy cannot be entirely excluded, only decreased levels of Bacteroidetes were noted in the sham animals, which was the opposite of findings in TBI animals. This was the only change seen in any of the phyla or families in the entire GI microbiota and may be accounted for by only having four sham animals. Finally, since the fecal samples were from excreted pellets, they only represent the microbiota of the large intestine and changes in the small intestinal microbiome cannot be inferred. Similarly, anatomic variation of luminal versus mucosal-associated microbes was not explored.

In summary, we observed changes in the gastrointestinal microbiome as early as 2 h post-TBI, which, in some families, persisted through 7 days in the absence of therapeutic intervention. Many of the alterations at both the phylum and family levels were noted by 3 days and correlated with peak lesion volume on MRI and loss of behavioral function. Many of these changes were directly related to the size of the lesion volumes measured via MRI scanning. The physiologic and clinical implications of these changes to the microbiota in the setting of TBI remain unseen. Further evaluation of the gut microbiome following TBI has the potential to improve clinical detection of TBI and outcome, serve as a potential therapeutic target, and enhance quality-of-life for patients with TBI.

Back to Top | Article Outline

Acknowledgments

The authors thank the following individuals for their support: Basil A. Pruitt, Jr., Dawn Garcia for 16S sequencing sample processing, Yidong Chen, PhD for bioinformatics support.

Back to Top | Article Outline

REFERENCES

1. Centers for Disease Control and Prevention: Report to Congress on Traumatic Brain Injury in the United States: Epidemiology and Rehabilitation. 2015.
2. Adelson PD, Whalen MJ, Kochanek PM, Robichaud P, Carlos TM. Blood brain barrier permeability and acute inflammation in two models of traumatic brain injury in the immature rat: a preliminary report. Acta Neurochir Suppl 71:104–106, 1998.
3. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, Gordon JI. The human microbiome project. Nature 449 7164:804–810, 2007.
4. Lloyd-Price J, Abu-Ali G, Huttenhower C. The healthy human microbiome. Genome Med 8 1:51, 2016.
5. Ivanov II, Honda K. Intestinal commensal microbes as immune modulators. Cell Host Microbe 12 4:496–508, 2012.
6. Cho I, Blaser MJ. The human microbiome: at the interface of health and disease. Nat Rev Genet 13 4:260–270, 2012.
7. Mayer EA, Tillisch K, Gupta A. Gut/brain axis and the microbiota. J Clin Invest 125 3:926–938, 2015.
8. Braniste V, Al-Asmakh M, Kowal C, Anuar F, Abbaspour A, Toth M, Korecka A, Bakocevic N, Ng LG, Kundu P, et al. The gut microbiota influences blood-brain barrier permeability in mice. Sci Transl Med 6 263:263ra158, 2014.
9. Singh V, Roth S, Llovera G, Sadler R, Garzetti D, Stecher B, Dichgans M, Liesz A. Microbiota dysbiosis controls the neuroinflammatory response after stroke. J Neurosci 36 28:7428–7440, 2016.
10. Houlden A, Goldrick M, Brough D, Vizi ES, Lenart N, Martinecz B, Roberts IS, Denes A. Brain injury induces specific changes in the caecal microbiota of mice via altered autonomic activity and mucoprotein production. Brain Behav Immun 57:10–20, 2016.
11. Earley ZM, Akhtar S, Green SJ, Naqib A, Khan O, Cannon AR, Hammer AM, Morris NL, Li X, Eberhardt JM, et al. Burn injury alters the intestinal microbiome and increases gut permeability and bacterial translocation. PLoS One 10 7:e0129996, 2015.
12. Watts LT, Long JA, Chemello J, Van Koughnet S, Fernandez A, Huang S, Shen Q, Duong TQ. Methylene blue is neuroprotective against mild traumatic brain injury. J Neurotrauma 31 11:1063–1071, 2014.
13. Shen Q, Meng X, Fisher M, Sotak CH, Duong TQ. Pixel-by-pixel spatiotemporal progression of focal ischemia derived using quantitative perfusion and diffusion imaging. J Cereb Blood Flow and Metab 23:1479–1488, 2003.
14. Shen Q, Ren H, Cheng H, Fisher M, Duong TQ. Functional, perfusion and diffusion MRI of acute focal ischemic brain injury. J Cereb Blood Flow and Metab 25:1265–1279, 2005.
15. Liu ZM, Schmidt KF, Sicard KM, Duong TQ. Imaging oxygen consumption in forepaw somatosensory stimulation in rats under isoflurane anesthesia. Magn Reson Med 52 2:277–285, 2004.
16. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C, Nielsen T, Pons N, Levenez F, Yamada T, et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 464 7285:59–65, 2010.
17. Faith DP, Baker AM. Phylogenetic diversity (PD) and biodiversity conservation: some bioinformatics challenges. Evol Bioinform Online 2:121–128, 2007.
18. Krezalek MA, DeFazio J, Zaborina O, Zaborin A, Alverdy JC. The shift of an intestinal “microbiome” to a “pathobiome” governs the course and outcome of sepsis following surgical injury. Shock 45 5:475–482, 2016.
19. Caricilli AM, Castoldi A, Camara NO. Intestinal barrier: a gentlemen's agreement between microbiota and immunity. World J Gastrointest Pathophysiol 5 1:18–32, 2014.
20. Carlucci C, Petrof EO, Allen-Vercoe E. Fecal microbiota-based therapeutics for recurrent clostridium difficile infection, ulcerative colitis and obesity. EBioMedicine 13:37–45, 2016.
21. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Microbial ecology: human gut microbes associated with obesity. Nature 444 7122:1022–1023, 2006.
22. Atarashi K, Tanoue T, Oshima K, Suda W, Nagano Y, Nishikawa H, Fukuda S, Saito T, Narushima S, Hase K, et al. Treg induction by a rationally selected mixture of Clostridia strains from the human microbiota. Nature 500 7461:232–236, 2013.
23. Zhang X, Shen D, Fang Z, Jie Z, Qiu X, Zhang C, Chen Y, Ji L. Human gut microbiota changes reveal the progression of glucose intolerance. PLoS One 8 8:e71108, 2013.
24. Keenan JI, Aitchison A, Purcell RV, Greenlees R, Pearson JF, Frizelle FA. Screening for enterotoxigenic Bacteroides fragilis in stool samples. Anaerobe 40:50–53, 2016.
25. Garrett WS, Gallini CA, Yatsunenko T, Michaud M, DuBois A, Delaney ML, Punit S, Karlsson M, Bry L, Glickman JN, et al. Enterobacteriaceae act in concert with the gut microbiota to induce spontaneous and maternally transmitted colitis. Cell Host Microbe 8 3:292–300, 2010.
26. Ahluwalia V, Betrapally NS, Hylemon PB, White MB, Gillevet PM, Unser AB, Fagan A, Daita K, Heuman DM, Zhou H, et al. Impaired gut-liver-brain axis in patients with cirrhosis. Sci Rep 6:26800, 2016.
27. Scheperjans F, Aho V, Pereira PA, Koskinen K, Paulin L, Pekkonen E, Haapaniemi E, Kaakkola S, Eerola-Rautio J, Pohja M, et al. Gut microbiota are related to Parkinson's disease and clinical phenotype. Mov Disord 30 3:350–358, 2015.
28. System N. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2003, issued August 2003. Am J Infect Control 31 8:481–498, 2003.
29. Travagli RA, Hermann GE, Browning KN, Rogers RC. Brainstem circuits regulating gastric function. Annu Rev Physiol 68:279–305, 2006.
30. Baguley IJ, Heriseanu RE, Cameron ID, Nott MT, Slewa-Younan S. A critical review of the pathophysiology of dysautonomia following traumatic brain injury. Neurocrit Care 8 2:293–300, 2008.
31. Grenham S, Clarke G, Cryan JF, Dinan TG. Brain-gut-microbe communication in health and disease. Front Physiol 2:94, 2011.
32. Zetterberg H, Smith DH, Blennow K. Biomarkers of mild traumatic brain injury in cerebrospinal fluid and blood. Nat Rev Neurol 9 4:201–210, 2013.
33. Paziana K, Korley FK. Emerging themes from the literature on circulating biomarkers of traumatic brain injury. Future Neurol 10 3:281–291, 2015.
34. Korley FK, Diaz-Arrastia R, Wu AH, Yue JK, Manley GT, Sair HI, Van Eyk J, Everett AD, investigators T-T, Okonkwo DO, et al. Circulating brain-derived neurotrophic factor has diagnostic and prognostic value in traumatic brain injury. J Neurotrauma 33 2:215–225, 2016.
35. Korley FK, Yue JK, Wilson D, Hrusovsky K, Diaz-Arrasta R, Ferguson AR, Yuh EL, Mukherjee P, Wang KKW, Valadka A, et al. Performance evaluation of a multiplex assay for simultaneous detection of four clinically relevant TBI biomarkers. J Neurotrauma 2018; [Epub ahead of print].
Keywords:

Brain-gut axis; gastrointestinal (GI); gut; commensals; microbiome; traumatic brain injury (TBI)

© 2019 by the Shock Society