Skip to main content
NCBI home page
American Journal of Respiratory and Critical Care Medicine logoLink to American Journal of Respiratory and Critical Care Medicine
. 2018 Mar 15;197(6):747–756. doi: 10.1164/rccm.201708-1559OC

Bacterial Dissemination to the Brain in Sepsis

Benjamin H Singer 1,2,✉, Robert P Dickson 1,2, Scott J Denstaedt 1, Michael W Newstead 1, Kwi Kim 3, Nicole R Falkowski 1, John R Erb-Downward 1, Thomas M Schmidt 3, Gary B Huffnagle 1,3, Theodore J Standiford 1
PMCID: PMC5855074  PMID: 29232157

Abstract

Rationale: Sepsis causes brain dysfunction and neuroinflammation. It is unknown whether neuroinflammation in sepsis is initiated by dissemination of bacteria to the brain and sustained by persistent infection, or whether neuroinflammation is a sterile process resulting solely from circulating inflammatory mediators.

Objectives: To determine if gut bacteria translocate to the brain during sepsis, and are associated with neuroinflammation.

Methods: Murine sepsis was induced using cecal ligation and puncture, and sepsis survivor mice were compared with sham and unoperated control animals. Brain tissue of patients who died of sepsis was compared with patients who died of noninfectious causes. Bacterial taxa were characterized by 16S ribosomal RNA gene sequencing in both murine and human brain specimens; compared among sepsis and nonsepsis groups; and correlated with levels of S100A8, a marker of neuroinflammation using permutational multivariate ANOVA.

Measurements and Main Results: Viable gut-associated bacteria were enriched in the brains of mice 5 days after surviving abdominal sepsis (P < 0.01), and undetectable by 14 days. The community structure of brain-associated bacteria correlated with severity of neuroinflammation (P < 0.001). Furthermore, bacterial taxa detected in brains of humans who die of sepsis were distinct from those who died of noninfectious causes (P < 0.001) and correlated with S100A8/A9 expression (P < 0.05).

Conclusions: Although bacterial translocation is associated with acute neuroinflammation in murine sepsis, bacterial translocation did not result in chronic cerebral infection. Postmortem analysis of patients who die of sepsis suggests a role for bacteria in acute brain dysfunction in sepsis. Further work is needed to determine if modifying gut-associated bacterial communities modulates brain dysfunction after sepsis.

Keywords: inflammation, microbiome, S100A8, sepsis, cecal ligation and puncture

At a Glance Commentary

Scientific Knowledge on the Subject

Sepsis results in long-term brain dysfunction, and animal models indicate that neuroinflammation plays a role in brain injury during sepsis. Whether brain injury is related to direct translocation of bacteria to the brain and cerebral microvasculature, or is a result of sterile inflammatory mediators is unknown. Evidence from patients with sepsis and animal models of sepsis and pneumonia suggest that translocation of gut bacteria may potentiate lung inflammation during sepsis, but the role of bacterial translocation in other organs is unstudied.

What This Study Adds to the Field

In experimental sepsis, early polymicrobial dissemination of bacteria to the brain is followed by persistence of a limited number of aerobic and anaerobic bacterial taxa that are eventually cleared, even in the absence of bacteremia. In postmortem brain specimens, the bacterial taxa identified from patients who die of sepsis are distinct from those of patients who died of noninfectious causes. The presence of bacteria is associated with an elevated neuroinflammatory marker in both mice and patients. We conclude that sepsis-related brain dysfunction is not a state of chronic bacterial encephalitis but that bacteria are associated with the early brain response to sepsis.

Sepsis, a syndrome of life-threatening organ dysfunction associated with the host response to infection, is frequently associated with acute and chronic brain dysfunction (1–5). Patients who die of sepsis experience cerebral coagulopathy, ischemia, and inflammation, and in animal models sepsis results in persistent neuroinflammation. Although multiple mechanisms related to the host response to infection provoke end-organ injury in sepsis, it is unclear whether the systemic dissemination of gut microbiota is an important part of the development of organ dysfunction, or whether organ dysfunction is solely caused by a dysregulated, but mostly sterile, host response (6–9). We have previously found that sepsis and acute respiratory distress syndrome are associated with enrichment of gut-associated bacteria in the lungs (10). Dissemination of enteric bacteria is not unique to models of abdominal sepsis; Pseudomonas pneumonia results in dissemination of enteric bacteria (11). It is therefore plausible that viable bacteria, and not only circulating cytokines, stimulate inflammation in organs remote from the initial site of infection.

Systemic administration of LPS and other purified bacterial products is often used as a model of sepsis and activation of innate immunity (12). Direct and overwhelming agonism of the signaling pathways downstream of LPS does produce significant neuroinflammation, including morphologic changes and cytokine expression in microglia, the primary innate immune cell of the brain (13). However, growing evidence indicates that even in sterile inflammation, the dose and timing of exposure to inflammatory mediators can lead either to priming or suppression of innate immunity (14). The presence over time of whole bacteria and multiple bacterial antigens within the microcirculation or the brain parenchyma, then, may provoke a significantly different immune response than a single large exposure to specific pathogen-associated molecules.

We hypothesize that neuroinflammation in sepsis and in sepsis survivors is not only provoked by sterile inflammatory mediators, but is related to the presence of bacteria in the brain. We further hypothesize that bacterial translocation is transient, and that the prolonged neuroinflammation observed in animal models of sepsis survival is not simply related to unresolved infection (9, 15). Here, we have used a naturalistic murine model of sepsis to show that viable gut-associated bacteria disseminate to the brain and are found for days after infection even in sepsis-survivor mice. We bolster the plausibility of bacterial dissemination as a driver of the brain response to infection with evidence of bacterial dissemination to the brain associated with neuroinflammation in humans who died of sepsis, even in the absence of clinically notable bacteremia. Some of the results of these studies have been previously reported in abstract form (16).

Methods

Mice

Wild-type mice used in these experiments were male C57BL/6 mice obtained from Jackson Laboratory and used at 8–10 weeks of age. CX3CR1-GFP::CCR2-RFP reporter mice were used at 8–12 weeks of age (17).

Sepsis Model

Cecal ligation and puncture (CLP) was performed as previously described using a single puncture with a 19-gauge needle (18). Imipenem/cilastatin (Merck, 0.5 mg/mouse) and normal saline (1 ml) were administered subcutaneously at the time of surgery. Sham operated animals underwent laparotomy without ligation or puncture of the cecum, and received antibiotics and saline.

Colony-Forming Unit Assay

Aliquots of brain homogenates, blood, or peritoneal lavage fluid were plated onto LB media in 10-fold dilutions and incubated aerobically at 37°C. Colonies were counted after 24 hours of growth and identified through a colony PCR method (10, 19).

Identification and Isolation of Patient-derived Brain Tissue

Human brain specimens were identified from the University of Michigan Brain Bank. Patients were identified as having sepsis if they died with a diagnosis of sepsis or had evidence of infection and organ injury. Patients were identified as dying of noninfectious causes if they died without evidence of infection and with an identifiable cause of death. These cases have also been examined in a separate publication examining the expression patterns of S100A8 in sepsis.

Bacterial DNA Isolation and 16S DNA Sequencing

Genomic DNA was extracted from mouse or human brain tissue. The V4 region of the 16S ribosomal RNA (rRNA) gene was amplified using published primers (20) and the dual-indexing sequencing strategy developed by the laboratory of Patrick D. Schloss and as previously described (10, 21).

Viability Assay of OTU008

Sequences matching the Bacteroidales OTU (OTU008, 100% coverage and identity) were identified using NCBI BLAST and have been previously identified in mouse feces and colonic mucosa (22–24). To determine whether OTU008 represented a viable bacterium or residual nucleic acids from killed organisms, we cultured brain homogenates under anaerobic conditions. PCR using OTU008-specific primers was performed to verify the presence of reproducing bacteria in these mixed cultures.

Statistical Analyses

Sequence data were processed and analyzed using the software mothur version 1.27.0, as previously described (25). Comparisons among means or medians were performed as noted in the text. Detailed methods are presented in the online supplement.

Results

Sepsis Induces Transient Dissemination of Bacteria to the Brain

To study whether sepsis results in dissemination of bacteria to the brain, we used CLP, a well-characterized murine model of polymicrobial sepsis and survivorship (9, 18). We have previously shown that CLP results in leukocyte infiltration and inflammatory cytokine production in the brain for at least 2 weeks after sepsis, with recovery of leukocyte infiltration but sustained cytokine expression over the following 50 days (9). We therefore studied microbiota isolated from the brain and blood via both culture and sequencing of 16S ribosomal RNA-encoding genes with the primary goal of determining whether persistent neuroinflammation after sepsis is simply a result of low-grade chronic bacterial encephalitis.

Given that CLP has the potential to disseminate aerobic and anaerobic bacteria, we performed an unbiased analysis of bacterial taxa present in the brain by 16S rRNA gene sequencing. One day after the induction of sepsis, bacterial sequences detected in the brains of septic mice were distinct from sham-operated and unoperated control animals (Figure 1A) (permutational multivariate ANOVA, P = 0.008). Unoperated control animals did not receive antibiotics. The differences among treatment groups reflects a wide array of bacterial taxa that differ from the background signal detected in unoperated control animals and reflect expected members of the murine gut microbiota, without clear predominance of a single taxon (Figure 1B). No bacteria grew in aerobic culture of the blood 24 hours after CLP, sham operation, or in unoperated control animals (n = 4 per group, not shown).

Figure 1.

Figure 1.

Open in a new tab

Multiple bacterial taxa are detectable in the brain during early sepsis. Abdominal sepsis was induced in mice by cecal ligation and puncture, and bacterial DNA detected in homogenized brain tissue was studied after 24 hours. (A) Brain bacterial communities 24 hours after operation are distinct from unoperated control subjects in both sham-operated and septic animals (permutational multivariate ANOVA, P = 0.03), and all three groups are distinct by principal component analysis (P < 0.001 for all pairwise comparisons). (B) Multiple bacterial taxa are detected in brain homogenates 24 hours after induction of sepsis, without clear predominance of a single taxon. These taxa are generally not represented in the sequences recovered from either sham-operated or unoperated animals. Gray bars indicate background signal detected in brains of unoperated mice. Data presented are means ± SEM. CLP = cecal ligation and puncture; Otu = operational taxanomic unit; PC = principal component.

To determine whether CLP results in prolonged exposure of the brain to bacteria, we examined brain microbiota in sepsis-survivor mice. Five days after intervention, bacterial sequences detected in brains of sepsis-survivor mice were distinct from those derived from sham-operated, antibiotic-treated, and untreated animals (P = 0.002) (Figure 2A). By 14 and 50 days after CLP bacterial sequences were not distinct among groups (P = 0.498 and 0.532) (see Figure E1C in the online supplement). In contrast to acutely ill animals, differences in bacterial community structure at 5 days after sepsis were primarily driven by a single Bacteroidales-classified taxonomic group (OTU), OTU008 followed by an Enterococcus classified OTU (Figure 2B). The relative abundance of this Bacteroidales OTU was significantly increased in sepsis-survivor animals compared with all control mice at 5 days after sepsis (P ≤ 0.01), but returned to background levels by 14 days (Figure 1C).

Figure 2.

Figure 2.

Open in a new tab

Brain microbial communities evolve after cecal ligation and puncture. Sepsis was induced by cecal ligation and puncture, and bacterial DNA detected in homogenized brain tissue was studied in surviving mice and compared with sham-operated, antibiotic-treated, and untreated control animals after 5 days. (A) Brain bacterial communities 5 days after induction of sepsis are distinct from those isolated from all control groups (permutational multivariate ANOVA, P < 0.002). (B) By 5 days post–cecal ligation and puncture, the detected microbial taxa are dominated by Bacteroidales OTU008, which is not detected in control groups. (C) Bacteroidales OTU008 dominates the bacterial community 5 days after sepsis (**P < 0.01) and returns to background levels by 14 days. Gray bars indicate background signal detected in brains of unoperated mice. Values presented as means ± SEM. Statistical significance was determined by Kruskall-Wallis one-way ANOVA with Dunn multiple comparison test (C). OTU = operational taxanomic unit; PC = principal component.

To determine whether the sequences detected by 16S rRNA gene analysis represented viable bacteria or residual nucleic acids from killed organisms, we cultured brain homogenates from sepsis-survivor mice. Enterococcus faecalis was isolated from aerobic culture of brains of sepsis-survivor mice 5 days after CLP (Figure 3A), but was not cultured from the blood (see Figure E1A). E. faecalis was not cultured from the brain 14 days or 50 days after CLP, or from sham-operated animals or antibiotic-treated unoperated control animals (Figure 1A; see Figure E1B). To examine the viability of OTU008, we cultured brain homogenates 5 days after CLP and from unoperated control mice under anaerobic conditions. OTU008 was undetectable by PCR of raw brain lysates of sepsis-survivor or control mice, but was readily amplified from anaerobic cultures of brain tissue, indicating that viable bacteria grew in vitro and are thus present in the brains of sepsis-survivor mice (Figure 3B).

Figure 3.

Figure 3.

Open in a new tab

Experimental sepsis results in translocation of live aerobic and anaerobic bacteria to the brain. (A) Five days after cecal ligation and puncture, Enterococcus faecalis was isolated from the brains of five of six postsepsis mice but was undetectable in the brains of mice in all control groups (P = 0.002). (B) OTU008 corresponds to a live bacterium, not carryover of nucleic acids. PCR specific to OTU008 shows the presence of bacterial growth in brain homogenates in anoxic conditions after sepsis but fails to detect OTU008 signal in raw brain homogenates or in unoperated control animals. Values presented as means ± SEM. Statistical significance was determined by Kruskall-Wallis one-way ANOVA with Dunn multiple comparison test (A). OTU = operational taxanomic unit.

Neuroinflammation Is Associated with Altered Brain Microbial Communities in Patients with Sepsis and in Experimental Sepsis

Given this evidence of significant bacterial exposure in the brain in our experimental model of sepsis, we questioned whether brain microbial communities in patients with sepsis differ from those in patients who died of nonsepsis causes. We reviewed records of subjects in the University of Michigan Brain Bank who had died either of sepsis (see Table E1) or of a clear noninfectious cause (see Table E2).

We found marked variability in the microbiota detected in the brains of patients with sepsis, including high abundance of taxa associated with both the upper respiratory and lower gastrointestinal tracts (Figure 4A). The separation between sepsis and nonsepsis brain specimens as a group was driven primarily by differences in relative abundance of Haemophilus, Neisseria, and Moraxella species, with brain specimens of individual patients with sepsis dominated by gut-associated taxa, such as Enterobacteriaceae sp. and Bacteroides sp. None of the sepsis-associated microbial taxa were enriched in the negative sequencing control specimens. Overall, bacterial taxa identified by 16S rRNA gene sequencing in cerebral cortex tissue were distinct in patients who died of sepsis compared with nonsepsis control subjects (PERMANOVA, P = 0.0008).

Figure 4.

Figure 4.

Open in a new tab

Neuroinflammation is associated with altered brain microbiota in patients with sepsis. Bacterial DNA was isolated and sequenced from postmortem brain specimens in patients who died of sepsis or a noninfectious illness. (A) Heterogeneous bacterial taxa are detected in brains of patients with and without sepsis. Model-based analysis of multivariate abundance identified Haemophilus, Neisseria, and Moraxella spp. as differentially abundant in sepsis and nonsepsis brain communities. A single brain specimen was enriched with Enterobactericiae sp. (OTU0017) and two were enriched with Bacteroides sp. (OTU0014). Pink boxes indicate upper airway–associated taxa, and brown boxes indicate gut-associated taxa. (B) Bacterial communities detected in cerebral cortex specimens from patients with sepsis were distinct from those of patients who died of noninfectious illness (P < 0.001). S100A8/A9 (S100 calcium binding protein A8 and A9 heterodimer) concentration is correlated with differences in bacterial taxa (P < 0.05). The area of each circle is proportional to the level of S100A8/A9 protein. Statistical significance was determined by Mann-Whitney U test (A) and permutational multivariate ANOVA (adonis) (B). OTU = operational taxanomic unit; PC = principal component.

Case reports have suggested that patients who die of sepsis demonstrate increased brain cytokine expression associated with neuroinflammation (26). We therefore examined expression of IL-1α, IL-1β, IL-1RA, IL-4, IL-6, IL-8, CCL2, CCL3, CCL4, and tumor necrosis factor-α in brain lysates by multiplex assay (Luminex, R&D Systems). Once adjusted for multiple comparisons, there was no significant difference in the expression of any of these inflammatory mediators in the brains of patients who died of sepsis and noninfectious causes (see Figure E2). Based on prior animal studies, we hypothesized that expression of the endogenous inflammatory mediator S100A8/A9 would be elevated in patients with sepsis (9).

Using S100A8/A9 protein concentration as a covariate in analysis of 16S sequences identified from human brain specimens, we found that neuroinflammation as measured by S100A8/A9 protein concentration was significantly correlated with the community structure of bacterial taxa (PERMANOVA, P = 0.034) (Figure 4B), indicating that human brain specimens with similar concentrations of S100A8/A9 had similar bacterial community structure as measured by 16S rRNA gene sequencing.

S100A8 expression is associated with brain microbiota even in early survivors of murine experimental sepsis. We measured brain expression of S100A8 in cerebral cortex tissue from the same animals used to study changes in microbiota. S100A8 gene expression was markedly increased in the cerebral cortex 5 days after sepsis compared with control conditions (P ≤ 0.001) (Figure 5A). S100A8 expression was significantly associated with the community structure of bacterial taxa in the brains of mice, driven by differences in the relative abundance of the previously described Bacteroidales OTU008 (P ≤ 0.001) (Figure 5B). Sepsis survivor mice also display evidence of altered microglial responses 5 days after CLP. Immunoreactivity for the microglial marker Iba1 was increased in the hippocampus and motor cortex compared with sham-operated animals (Mann-Whitney U test; P = 0.035) (Figures 5C and 5D). Isolated microglia and tissue macrophages, identified using a GFP reporter of CX3CR1 expression (Figure 5E), demonstrate priming of tumor necrosis factor-α production on ex vivo stimulation with LPS (Figure 5F). Neuroinflammation in sepsis survivor mice, which we have previously characterized up to 2 months following CLP (9), could simply be driven by chronic intraperitoneal infection and systemic inflammation. However, peritoneal infection (see Figure E3A) and neutrophil exudation (see Figure E3B) both improve significantly over the first week following CLP.

Figure 5.

Figure 5.

Open in a new tab

Neuroinflammation is enhanced in postsepsis brains and correlated with changes in brain microbiota in experimental sepsis. S100A8 (S100 calcium binding protein A8) expression was correlated both with OTU008 abundance (A, permutational multivariate ANOVA [adonis], P < 0.001) and overall microbial community structure 5 days after operations (B, permutational multivariate ANOVA [adonis], P ≤ 0.001). (B) Relative expression of S100A8 is proportional to marker size. In addition to increased tissue damage–associated molecular pattern expression, sepsis survivor mice exhibit increased reactivity by microglia and brain tissue macrophages. (C) Immunfluorescence of motor cortex show markedly increased size and density of Iba1+ microglial cell processes. (D) Immunoreactivity for Iba1 is increased in cecal ligation and puncture survivors on Day 5 post–cecal ligation and puncture in both area CA1 and overlying motor cortex (Mann-Whitney U test, P = 0.035 in both areas). *P < 0.05. (E) Microglia/macrophages were isolated by flow cytometry as a CX3CR1+CCR2− cell population. (F) When stimulated ex vivo with LPS, CX3CR1+CCR2− cells isolated from sepsis survivor mice 5 days post–cecal ligation and puncture produce increased tumor necrosis factor-α compared with sham-operated (ANOVA, P < 0.0001; Sidak post hoc, P = 0.0002) or antibiotic-treated mice (Sidak post hoc, P < 0.0001). ***P < 0.001. Values are presented as means ± SEM. A.U. = arbitrary units; CCR2 = C-C motif chemokine receptor 2; CLP = cecal ligation and puncture; CX3CR1 = C-X3-C motif chemokine receptor 1; GFP = green fluorescent protein; Iba1 = ionized calcium binding adapter molecule 1; OTU = operational taxanomic unit; PC = principal component; RFP = red fluorescent protein; S1008A8/A9 = S100 calcium binding protein A8 and A9 heterodimer; TNF = tumor necrosis factor.

Discussion

The central finding of our study is that sepsis results in dissemination of bacteria to the brain for several days, which ultimately resolves over the course of weeks after the onset of illness. The presence of distinct bacterial taxa was associated with S100A8 expression, a marker of neuroinflammation, both in animals and humans who died of sepsis. Our results suggest that viable bacteria, and not only sterile circulating inflammatory mediators, play a role in the pathogenesis of brain dysfunction after sepsis.

The brain is vulnerable to numerous insults during sepsis. Cerebral infarcts, hemorrhage, and microabscesses and macroabscesses have been demonstrated in postmortem studies of patients who die of sepsis (27–29). One could assume that any patient with bacteremia is at risk for developing microabscesses in multiple organ systems, and that this process underlies organ dysfunction in sepsis. Some common bacteria, such as Staphylococcus aureus, do form disseminated abscesses in the case of fulminant infection, including in the brain (30). Likewise, direct dissemination of Streptococcus to the myocardium contributes to cardiac dysfunction in models of pneumonia (31). Although the prevalence of focal infection at autopsy is common in sepsis, disseminated microabscesses are rare, and encephalitis or brain abscesses were nearly absent in several case series (32–34). Similarly, in the present study only a single patient had gross or histopathologic evidence of overt brain infection at autopsy (see Table E1).

If histologic evidence of bacterial dissemination is absent in many cases of sepsis, then is brain dysfunction in sepsis related to the presence of bacteria at all? Bacteremia is detected in only about a third of patients with sepsis, leading to the hypothesis that sterile mediators, such as LPS, or a dysregulated host response to infection are primary drivers of organ dysfunction (1, 35–38). Significant controversy surrounds the fidelity of sterile endotoxemia as a model of human sepsis, however (12, 39, 40). Our data suggest that viable bacteria do translocate to the brain during sepsis, even in the absence of overt infection. In murine sepsis survivors, cultivable aerobic bacteria are detected in the brain without bacteremia, suggesting that translocation either occurs very briefly or that the biomass of bacteria in blood is low, and bacteria are more readily detected once captured in the brain microvasculature. The identification of uncultured anaerobes in the brain, however, suggests that many disseminated bacteria may simply be undetectable with culture-dependent methods.

Given that the immune response to pathogen-associated molecules varies immensely with the chronicity, multiplicity, and intensity of exposure, understanding dissemination of even small numbers of bacteria is of critical importance to predicting and understanding the pathogenesis of organ dysfunction in this syndrome (14). Although we found that alterations in brain microbiota are associated with neuroinflammation, we do not claim that exposure to microbial pathogens is the sole pathway to developing acute or chronic brain dysfunction in critical illness. Delirium is associated with multiple causes of critical illness, and survivors of respiratory failure and shock suffer from chronic brain dysfunction (5, 41, 42). Microglia are activated in the brains of patients with delirium, even in the absence of sepsis (43). Multiple mechanisms, including ischemia, neurotransmitter imbalance, and vascular dysfunction likely contribute to brain dysfunction in critical illness (7). Nevertheless, there is evidence that amyloid fibrils, a pathologic hallmark of Alzheimer’s disease, play a role in host defense against bacteria, suggesting immune mechanisms have evolved in response to direct exposure of the brain to bacteria (44). Even in primarily noninfectious states of critical illness, such as stroke, increases in gut permeability may increase risk of bacterial translocation, and modulation of the gut microbiota may be a target for treatment and reduction of complications (45).

The Bacteriodales OTU008 identified in our mouse studies is most closely homologous to the candidate family Homeothermaceae. Although the potential pathogenicity of these bacteria is unknown, they do carry genes for tolerance of oxidative stress and assembly of fimbriae, similar to pathogens in the genus Porphyromonas (46). A member of this family has recently been cultured for the first time, opening the possibility for cultivating other members of the Homeothermaceae and directly investigating their pathogenicity (47).

Although our results support the hypothesis that bacterial translocation is involved in the initiation of neuroinflammation, we did not find evidence of chronic bacterial infection of the brain in this model. In some forms of severe CLP, chronic intraabdominal infection is established (48, 49), but in our model peritoneal infection and inflammation were controlled. We did not find evidence of bacteria in the brain 2 weeks after sepsis, and thus persistent neuroinflammation after sepsis is likely caused by reprogramming of the host immune response and is not caused by persistent or chronic infection (9, 50, 51).

Both our animal and human studies demonstrate only an association, and not a causative relationship between brain microbial exposure and neuroinflammation. It is possible that differences in microbiota and neuroinflammation observed in the human and animal studies here are unrelated results of systemic inflammation in sepsis. A limitation of using CLP as a model of sepsis is that the gut serves as the primary source of infection. Antibiotic treatment to eliminate gut microbiota would alter the fundamental severity of illness in CLP, and therefore not illuminate the independent role of bacterial dissemination in neuroinflammation. We are developing other models of sepsis survival to address the independent contribution of the gut microbiota to long-term organ dysfunction after sepsis. Furthermore, although the use of the animal model allows us to study variation in bacterial taxa in the brain over time, the study of bacteria in the brains of patients with sepsis is necessarily limited by the requirement that patients die of their acute illness, which introduces several sources of confounding. The small number of patients in this study preclude adjusting for these factors, but they provide useful guidance for the design of future studies.

Several important factors other than brain microbiota differ among the sepsis and nonsepsis groups in our study. Patients in the sepsis group received more antibiotics than patients who died of nonsepsis causes. Although we postulate that the brain parenchyma is sterile in health, patient-derived samples are subject to translocation and outgrowth of bacteria during the postmortem interval (52–54). Administration of antibiotics during life could thus alter the 16S sequences observed in our study. The gut microbiome contributes to maintenance of the blood-brain barrier, cerebral leukocyte trafficking, and neurogenesis via circulating bacterial metabolites (55–57). Thus, antibiotic-induced dysbiosis could contribute to brain dysfunction even in the absence of sepsis.

Neurodegenerative disease was also more prevalent in the sepsis group than nonsepsis control subjects. Not all patients with elevated brain levels of S100A8/A9 had an underlying diagnosis of Alzheimer’s disease. Alzheimer’s disease is associated with increased levels of S100A8 and S100A9, although whether the fibrillar forms of these proteins would react to the heterodimer-specific assay used here is unknown (58–60). As damage-associated molecular pattern, S100A8/A9 is also elevated by ischemia and hypoxia, not only in sepsis or infection (61–63). Patients in the sepsis group did have a prolonged agonal course compared with patients without sepsis, and signs of hypoxia have been identified previously in postmortem brain specimens from patients with sepsis and delirium (34). We unsuccessfully attempted to find alternative protein markers of brain inflammation in these tissues. The utility of S100A8/A9 as a marker of inflammation in this setting may be caused by high initial expression, or the robustness of this marker to the storage conditions in banked samples.

Despite these confounding factors, our human and murine data illuminate promises and pitfalls of using 16S sequencing to analyze brain microbiota. Although a previous study examined human brain microbiota using culture-independent techniques, it included patients with a heterogeneous group of infections and acute brain pathologies (64). Sequencing of 16S rRNA gene sequences in perfused brain tissue obtained under controlled conditions from antibiotic-treated control mice produces signals associated with multiple bacteria, including enteric pathogens, which are not present in reagent-only control animals. Several factors account for this discrepancy. First, microbial DNA may be present in the circulation or within circulating leukocytes in the absence of viable organisms. Nevertheless, we showed that the most abundant Bacteroidales 16S DNA detected in this study was derived from a live organism. Second, errors in both PCR amplification and gene sequencing are likely to be more frequent when abundant nucleic acids are present in a biologic sample than in reagent control subjects (65). It is unlikely that viable aerobic bacteria are consistently present within the healthy mouse brain with an abundance below the limit of detection by culture-based methods.

We have shown that sepsis results in altered brain bacterial taxa, which are associated with neuroinflammation in both a murine model of sepsis survival and in patients who die of sepsis, suggesting that direct translocation of bacteria to the brain plays a role in brain dysfunction in sepsis. Currently, antimicrobial therapy in sepsis is often directed at the presumed source of infection, and frequently does not include empiric treatment of gut-associated anaerobic bacteria. Clinical trials have demonstrated that selective decontamination of gut bacteria reduces organ dysfunction and improves outcomes in sepsis, although the practice has not been adopted worldwide (66). This study and others suggest that renewed attention should be paid to the effects of uncultured bacteria on specific end-organ dysfunctions in sepsis, and whether early antibiotic treatment of central nervous system infection even in the absence of encephalitis will modify the incidence of brain dysfunction after sepsis. Further study is needed, however, to delineate whether specific bacteria play a causative role in neuroinflammation.

Acknowledgments

Acknowledgment

The authors acknowledge the University of Michigan Multidisciplinary Intensive Care Research Workgroup for discussions, and Drs. Dolores Hambardzumyan and Richard Ransohoff for the gift of CX3CR1-GFP::CCR2-RFP mice. Sequences are available from the NCBI Sequence Read Archive (accession number SRP067513).

Footnotes

Supported by NIH grants T32HL00774921 (B.H.S.), K08NS101054 (B.H.S.), UL1TR000433 (R.P.D.), K23HL130641 (R.P.D.), R01HL123515 (T.J.S.), U01HL098961 (G.B.H.), and R01HL114447 (G.B.H.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Support was provided by the Michigan Institute for Clinical and Health Research (R.P.D.), the Host Microbiome Initiative of the University of Michigan (R.P.D. and B.H.S.), the University of Michigan Center for Integrative Research in Critical Care (R.P.D. and B.H.S.), and the University of Michigan Depression Center (B.H.S.). The University of Michigan Brain Bank is supported by National Institute of Aging P30 AG 053760 and the University of Michigan Protein Folding Disorders Initiative.

Author Contributions: Designed the study, B.H.S., R.P.D., T.M.S., G.B.H., and T.J.S. Performed experiments and analysis of data, B.H.S., R.P.D., S.J.D., M.W.N., K.K., N.R.F., and J.R.E.-D. Drafted the manuscript, B.H.S. and R.P.D. Approved final version of the manuscript, B.H.S., R.P.D., S.J.D., M.W.N., K.K., N.R.F., J.R.E.-D., T.M.S., G.B.H., and T.J.S.

This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org.

Originally Published in Press as DOI: 10.1164/rccm.201708-1559OC on December 12, 2017

Author disclosures are available with the text of this article at www.atsjournals.org.

References

  • 1.Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) JAMA. 2016;315:801–810. doi: 10.1001/jama.2016.0287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Widmann CN, Heneka MT. Long-term cerebral consequences of sepsis. Lancet Neurol. 2014;13:630–636. doi: 10.1016/S1474-4422(14)70017-1. [DOI] [PubMed] [Google Scholar]
  • 3.Ely EW, Shintani A, Truman B, Speroff T, Gordon SM, Harrell FE, Jr, et al. Delirium as a predictor of mortality in mechanically ventilated patients in the intensive care unit. JAMA. 2004;291:1753–1762. doi: 10.1001/jama.291.14.1753. [DOI] [PubMed] [Google Scholar]
  • 4.Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA. 2010;304:1787–1794. doi: 10.1001/jama.2010.1553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Salluh JI, Soares M, Teles JM, Ceraso D, Raimondi N, Nava VS, et al. Delirium Epidemiology in Critical Care Study Group. Delirium epidemiology in critical care (DECCA): an international study. Crit Care. 2010;14:R210. doi: 10.1186/cc9333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Sonneville R, Verdonk F, Rauturier C, Klein IF, Wolff M, Annane D, et al. Understanding brain dysfunction in sepsis. Ann Intensive Care. 2013;3:15. doi: 10.1186/2110-5820-3-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sharshar T, Bozza F, Chrétien F. Neuropathological processes in sepsis. Lancet Neurol. 2014;13:534–536. doi: 10.1016/S1474-4422(14)70064-X. [DOI] [PubMed] [Google Scholar]
  • 8.Weberpals M, Hermes M, Hermann S, Kummer MP, Terwel D, Semmler A, et al. NOS2 gene deficiency protects from sepsis-induced long-term cognitive deficits. J Neurosci. 2009;29:14177–14184. doi: 10.1523/JNEUROSCI.3238-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Singer BH, Newstead MW, Zeng X, Cooke CL, Thompson RC, Singer K, et al. Cecal ligation and puncture results in long-term central nervous system myeloid inflammation. PLoS One. 2016;11:e0149136. doi: 10.1371/journal.pone.0149136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Dickson RP, Singer BH, Newstead MW, Falkowski NR, Erb-Downward JR, Standiford TJ, et al. Enrichment of the lung microbiome with gut bacteria in sepsis and the acute respiratory distress syndrome. Nat Microbiol. 2016;1:16113. doi: 10.1038/nmicrobiol.2016.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hunninghake GW, Doerschug KC, Nymon AB, Schmidt GA, Meyerholz DK, Ashare A. Insulin-like growth factor-1 levels contribute to the development of bacterial translocation in sepsis. Am J Respir Crit Care Med. 2010;182:517–525. doi: 10.1164/rccm.200911-1757OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Nemzek JA, Hugunin KMS, Opp MR. Modeling sepsis in the laboratory: merging sound science with animal well-being. Comp Med. 2008;58:120–128. [PMC free article] [PubMed] [Google Scholar]
  • 13.Hoogland ICM, Houbolt C, van Westerloo DJ, van Gool WA, van de Beek D. Systemic inflammation and microglial activation: systematic review of animal experiments. J Neuroinflammation. 2015;12:114. doi: 10.1186/s12974-015-0332-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Morris MC, Gilliam EA, Li L. Innate immune programing by endotoxin and its pathological consequences. Front Immunol. 2015;5:680. doi: 10.3389/fimmu.2014.00680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Barichello T, Martins MR, Reinke A, Feier G, Ritter C, Quevedo J, et al. Long-term cognitive impairment in sepsis survivors. Crit Care Med. 2005;33:1671. doi: 10.1097/01.ccm.0000170192.54682.c1. [DOI] [PubMed] [Google Scholar]
  • 16.Singer BH, Dickson RP, Newstead MW, Falkowski NR, Kim K, Seiber JR, et al. Bacterial dissemination to the brain in a murine model of treated sepsis [abstract] Am J Respir Crit Care Med. 2016;193:A6301. doi: 10.1164/rccm.201708-1559OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Yamasaki R, Lu H, Butovsky O, Ohno N, Rietsch AM, Cialic R, et al. Differential roles of microglia and monocytes in the inflamed central nervous system. J Exp Med. 2014;211:1533–1549. doi: 10.1084/jem.20132477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Walley KR, Lukacs NW, Standiford TJ, Strieter RM, Kunkel SL. Balance of inflammatory cytokines related to severity and mortality of murine sepsis. Infect Immun. 1996;64:4733–4738. doi: 10.1128/iai.64.11.4733-4738.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Paster BJ, Boches SK, Galvin JL, Ericson RE, Lau CN, Levanos VA, et al. Bacterial diversity in human subgingival plaque. J Bacteriol. 2001;183:3770–3783. doi: 10.1128/JB.183.12.3770-3783.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Lozupone CA, Turnbaugh PJ, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proc Natl Acad Sci USA. 2011;108:4516–4522. doi: 10.1073/pnas.1000080107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microbiol. 2013;79:5112–5120. doi: 10.1128/AEM.01043-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Devkota S, Wang Y, Musch MW, Leone V, Fehlner-Peach H, Nadimpalli A, et al. Dietary-fat-induced taurocholic acid promotes pathobiont expansion and colitis in Il10-/- mice. Nature. 2012;487:104–108. doi: 10.1038/nature11225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Evans CC, LePard KJ, Kwak JW, Stancukas MC, Laskowski S, Dougherty J, et al. Exercise prevents weight gain and alters the gut microbiota in a mouse model of high fat diet-induced obesity. PLoS One. 2014;9:e92193. doi: 10.1371/journal.pone.0092193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Musch MW, Wang Y, Claud EC, Chang EB. Lubiprostone decreases mouse colonic inner mucus layer thickness and alters intestinal microbiota. Dig Dis Sci. 2013;58:668–677. doi: 10.1007/s10620-012-2509-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ballinger MN, Newstead MW, Zeng X, Bhan U, Horowitz JC, Moore BB, et al. TLR signaling prevents hyperoxia-induced lung injury by protecting the alveolar epithelium from oxidant-mediated death. J Immunol. 2012;189:356–364. doi: 10.4049/jimmunol.1103124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Warford J, Lamport A-C, Kennedy B, Easton AS. Human brain chemokine and cytokine expression in sepsis: a report of three cases. Can J Neurol Sci. 2017;44:96–104. doi: 10.1017/cjn.2016.310. [DOI] [PubMed] [Google Scholar]
  • 27.Sharshar T, Annane D, de la Grandmaison GL, Brouland JP, Hopkinson NS, Françoise G. The neuropathology of septic shock. Brain Pathol. 2004;14:21–33. doi: 10.1111/j.1750-3639.2004.tb00494.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Pendlebury WW, Perl DP, Munoz DG. Multiple microabscesses in the central nervous system: a clinicopathologic study. J Neuropathol Exp Neurol. 1989;48:290–300. doi: 10.1097/00005072-198905000-00006. [DOI] [PubMed] [Google Scholar]
  • 29.Jackson AC, Gilbert JJ, Young GB, Bolton CF. The encephalopathy of sepsis. Can J Neurol Sci. 1985;12:303–307. doi: 10.1017/s0317167100035381. [DOI] [PubMed] [Google Scholar]
  • 30.Astrup LB, Nielsen MV, Iburg TM, Leifsson PS, Jensen HE, Nielsen OL, et al. Brain microabscesses in a porcine model of Staphylococcus aureus sepsis. Acta Vet Scand. 2013;55:76. doi: 10.1186/1751-0147-55-76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Reyes LF, Restrepo MI, Hinojosa CA, Soni NJ, Anzueto A, Babu BL, et al. Severe pneumococcal pneumonia causes acute cardiac toxicity and subsequent cardiac remodeling. Am J Respir Crit Care Med. 2017;196:609–620. doi: 10.1164/rccm.201701-0104OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Bode-Jänisch S, Schütz S, Schmidt A, Tschernig T, Debertin AS, Fieguth A, et al. Serum procalcitonin levels in the postmortem diagnosis of sepsis. Forensic Sci Int. 2013;226:266–272. doi: 10.1016/j.forsciint.2013.01.041. [DOI] [PubMed] [Google Scholar]
  • 33.Torgersen C, Moser P, Luckner G, Mayr V, Jochberger S, Hasibeder WR, et al. Macroscopic postmortem findings in 235 surgical intensive care patients with sepsis. Anesth Analg. 2009;108:1841–1847. doi: 10.1213/ane.0b013e318195e11d. [DOI] [PubMed] [Google Scholar]
  • 34.Janz DR, Abel TW, Jackson JC, Gunther ML, Heckers S, Ely EW. Brain autopsy findings in intensive care unit patients previously suffering from delirium: a pilot study. J Crit Care. 2010;25:538.e7–12. doi: 10.1016/j.jcrc.2010.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Phua J, Ngerng W, See K, Tay C, Kiong T, Lim H, et al. Characteristics and outcomes of culture-negative versus culture-positive severe sepsis. Crit Care. 2013;17:R202. doi: 10.1186/cc12896. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Guidet B, Barakett V, Vassal T, Petit JC, Offenstadt G. Endotoxemia and bacteremia in patients with sepsis syndrome in the intensive care unit. Chest. 1994;106:1194–1201. doi: 10.1378/chest.106.4.1194. [DOI] [PubMed] [Google Scholar]
  • 37.Opal SM, Garber GE, LaRosa SP, Maki DG, Freebairn RC, Kinasewitz GT, et al. Systemic host responses in severe sepsis analyzed by causative microorganism and treatment effects of drotrecogin alfa (activated) Clin Infect Dis. 2003;37:50–58. doi: 10.1086/375593. [DOI] [PubMed] [Google Scholar]
  • 38.Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J Med. 2013;369:840–851. doi: 10.1056/NEJMra1208623. [DOI] [PubMed] [Google Scholar]
  • 39.Takao K, Miyakawa T. Genomic responses in mouse models greatly mimic human inflammatory diseases. Proc Natl Acad Sci USA. 2015;112:1167–1172. doi: 10.1073/pnas.1401965111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Seok J, Warren HS, Cuenca AG, Mindrinos MN, Baker HV, Xu W, et al. Inflammation and Host Response to Injury, Large Scale Collaborative Research Program. Genomic responses in mouse models poorly mimic human inflammatory diseases. Proc Natl Acad Sci USA. 2013;110:3507–3512. doi: 10.1073/pnas.1222878110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Adhikari NKJ, McAndrews MP, Tansey CM, Matté A, Pinto R, Cheung AM, et al. Self-reported symptoms of depression and memory dysfunction in survivors of ARDS. Chest. 2009;135:678–687. doi: 10.1378/chest.08-0974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Pandharipande PP, Girard TD, Jackson JC, Morandi A, Thompson JL, Pun BT, et al. BRAIN-ICU Study Investigators. Long-term cognitive impairment after critical illness. N Engl J Med. 2013;369:1306–1316. doi: 10.1056/NEJMoa1301372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Munster BC, Aronica E, Zwinderman AH, Eikelenboom P, Cunningham C, Rooij SE. Neuroinflammation in delirium: a postmortem case-control study. Rejuvenation Res. 2011;14:615–622. doi: 10.1089/rej.2011.1185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Kumar DKV, Choi SH, Washicosky KJ, Eimer WA, Tucker S, Ghofrani J, et al. Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer’s disease. Sci Transl Med. 2016;8:340ra72. doi: 10.1126/scitranslmed.aaf1059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Stanley D, Mason LJ, Mackin KE, Srikhanta YN, Lyras D, Prakash MD, et al. Translocation and dissemination of commensal bacteria in post-stroke infection. Nat Med. 2016;22:1277–1284. doi: 10.1038/nm.4194. [DOI] [PubMed] [Google Scholar]
  • 46.Ormerod KL, Wood DLA, Lachner N, Gellatly SL, Daly JN, Parsons JD, et al. Genomic characterization of the uncultured Bacteroidales family S24-7 inhabiting the guts of homeothermic animals. Microbiome. 2016;4:36. doi: 10.1186/s40168-016-0181-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Lagkouvardos I, Pukall R, Abt B, Foesel BU, Meier-Kolthoff JP, Kumar N, et al. The Mouse Intestinal Bacterial Collection (miBC) provides host-specific insight into cultured diversity and functional potential of the gut microbiota. Nat Microbiol. 2016;1:16131. doi: 10.1038/nmicrobiol.2016.131. [DOI] [PubMed] [Google Scholar]
  • 48.Doerschug KC, Powers LS, Monick MM, Thorne PS, Hunninghake GW. Antibiotics delay but do not prevent bacteremia and lung injury in murine sepsis. Crit Care Med. 2004;32:489–494. doi: 10.1097/01.CCM.0000109450.66450.23. [DOI] [PubMed] [Google Scholar]
  • 49.Iskander KN, Vaickus M, Duffy ER, Remick DG, Moldawer L, Wollenberg G. Shorter duration of post-operative antibiotics for cecal ligation and puncture does not increase inflammation or mortality. PLoS One. 2016;11:e0163005. doi: 10.1371/journal.pone.0163005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Chavan SS, Huerta PT, Robbiati S, Valdes-Ferrer SI, Ochani M, Dancho M, et al. HMGB1 mediates cognitive impairment in sepsis survivors. Mol Med. 2012;18:930–937. doi: 10.2119/molmed.2012.00195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Barichello T, Martins MR, Reinke A, Feier G, Ritter C, Quevedo J, et al. Cognitive impairment in sepsis survivors from cecal ligation and perforation. Crit Care Med. 2005;33:221–223, discussion 262–263. doi: 10.1097/01.ccm.0000150741.12906.bd. [DOI] [PubMed] [Google Scholar]
  • 52.Javan GT, Finley SJ, Can I, Wilkinson JE, Hanson JD, Tarone AM. Human thanatomicrobiome succession and time since death. Sci Rep. 2016;6:29598. doi: 10.1038/srep29598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Burcham ZM, Hood JA, Pechal JL, Krausz KL, Bose JL, Schmidt CJ, et al. Fluorescently labeled bacteria provide insight on post-mortem microbial transmigration. Forensic Sci Int. 2016;264:63–69. doi: 10.1016/j.forsciint.2016.03.019. [DOI] [PubMed] [Google Scholar]
  • 54.Can I, Javan GT, Pozhitkov AE, Noble PA. Distinctive thanatomicrobiome signatures found in the blood and internal organs of humans. J Microbiol Methods. 2014;106:1–7. doi: 10.1016/j.mimet.2014.07.026. [DOI] [PubMed] [Google Scholar]
  • 55.Braniste V, Al-Asmakh M, Kowal C, Anuar F, Abbaspour A, Toth M, et al. The gut microbiota influences blood-brain barrier permeability in mice. Sci Transl Med. 2014;6:263ra158. doi: 10.1126/scitranslmed.3009759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Campos AC, Rocha NP, Nicoli JR, Vieira LQ, Teixeira MM, Teixeira AL. Absence of gut microbiota influences lipopolysaccharide-induced behavioral changes in mice. Behav Brain Res. 2016;312:186–194. doi: 10.1016/j.bbr.2016.06.027. [DOI] [PubMed] [Google Scholar]
  • 57.Möhle L, Mattei D, Heimesaat MM, Bereswill S, Fischer A, Alutis M, et al. Ly6C(hi) monocytes provide a link between antibiotic-induced changes in gut microbiota and adult hippocampal neurogenesis. Cell Reports. 2016;15:1945–1956. doi: 10.1016/j.celrep.2016.04.074. [DOI] [PubMed] [Google Scholar]
  • 58.Shepherd CE, Goyette J, Utter V, Rahimi F, Yang Z, Geczy CL, et al. Inflammatory S100A9 and S100A12 proteins in Alzheimer’s disease. Neurobiol Aging. 2006;27:1554–1563. doi: 10.1016/j.neurobiolaging.2005.09.033. [DOI] [PubMed] [Google Scholar]
  • 59.Wang C, Klechikov AG, Gharibyan AL, Wärmländer SKTS, Jarvet J, Zhao L, et al. The role of pro-inflammatory S100A9 in Alzheimer’s disease amyloid-neuroinflammatory cascade. Acta Neuropathol. 2014;127:507–522. doi: 10.1007/s00401-013-1208-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Lodeiro M, Puerta E, Ismail MA, Rodriguez-Rodriguez P, Rönnbäck A, Codita A, et al. Aggregation of the inflammatory S100A8 precedes Aβ plaque formation in transgenic APP mice: positive feedback for S100A8 and Aβ productions. J Gerontol A Biol Sci Med Sci. 2017;72:319–328. doi: 10.1093/gerona/glw073. [DOI] [PubMed] [Google Scholar]
  • 61.Biswal S, Sharma D, Kumar K, Nag TC, Barhwal K, Hota SK, et al. Global hypoxia induced impairment in learning and spatial memory is associated with precocious hippocampal aging. Neurobiol Learn Mem. 2016;133:157–170. doi: 10.1016/j.nlm.2016.05.011. [DOI] [PubMed] [Google Scholar]
  • 62.Ziegler G, Prinz V, Albrecht MW, Harhausen D, Khojasteh U, Nacken W, et al. Mrp-8 and -14 mediate CNS injury in focal cerebral ischemia. Biochim Biophys Acta. 2009;1792:1198–204. doi: 10.1016/j.bbadis.2009.10.003. [DOI] [PubMed] [Google Scholar]
  • 63.van Zoelen MAD, Vogl T, Foell D, Van Veen SQ, van Till JWO, Florquin S, et al. Expression and role of myeloid-related protein-14 in clinical and experimental sepsis. Am J Respir Crit Care Med. 2009;180:1098–1106. doi: 10.1164/rccm.200810-1552OC. [DOI] [PubMed] [Google Scholar]
  • 64.Branton WG, Ellestad KK, Maingat F, Wheatley BM, Rud E, Warren RL, et al. Brain microbial populations in HIV/AIDS: alpha-proteobacteria predominate independent of host immune status. PLoS One. 2013;8:e54673. doi: 10.1371/journal.pone.0054673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Schloss PD, Gevers D, Westcott SL. Reducing the effects of PCR amplification and sequencing artifacts on 16S rRNA-based studies. PLoS One. 2011;6:e27310. doi: 10.1371/journal.pone.0027310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Silvestri L, de la Cal MA, van Saene HKF. Selective decontamination of the digestive tract: the mechanism of action is control of gut overgrowth. Intensive Care Med. 2012;38:1738–1750. doi: 10.1007/s00134-012-2690-1. [DOI] [PubMed] [Google Scholar]

Articles from American Journal of Respiratory and Critical Care Medicine are provided here courtesy of American Thoracic Society

RESOURCES