Gut Microbiome in Sepsis

Abstract

The gut has been hypothesized to be the “motor” of multiple organ dysfunction in sepsis. Although there are multiple ways in which the gut can drive systemic inflammation, increasing evidence suggests that the intestinal microbiome plays a more substantial role than previously appreciated. An English language literature review was performed to summarize the current knowledge of sepsis-induced gut microbiome dysbiosis. Conversion of a normal microbiome to a pathobiome in the setting of sepsis is associated with worsened mortality. Changes in microbiome composition and diversity signal the intestinal epithelium and immune system resulting in increased intestinal permeability and a dysregulated immune response to sepsis. Clinical approaches to return to microbiome homeostasis may be theoretically possible through a variety of methods including probiotics, prebiotics, fecal microbial transplant, and selective decontamination of the digestive tract. However, more research is required to determine the efficacy (if any) of targeting the microbiome for therapeutic gain. The gut microbiome rapidly loses diversity with emergence of virulent bacteria in sepsis. Restoring normal commensal bacterial diversity through various therapies may be an avenue to improve sepsis mortality.

Sepsis is a dysregulated host response to infection resulting in life-threatening organ dysfunction.¹ Sepsis results in a significant burden to the healthcare system with high mortality and a cost of more than $60 billion annually in the United States alone.² Unfortunately, antibiotic agents represent the only targeted therapy for sepsis that is otherwise managed by supportive therapy.³ The gut has long been hypothesized to be the “motor” of critical illness.^4,5 The gut consists of three compartments: the microbiome, the epithelium, and the immune system. Each of these components is separately altered in sepsis, but crosstalk occurs between the three components.^6–8 Accordingly, we review sepsis-induced changes of the gut microbiome as well as its interactions with intestinal epithelium and immune system. Additionally, we discuss data on targeting the microbiome as a therapeutic strategy in sepsis.

Gut Microbiome

The intestinal microbiome has the same number of cells as the human body (40 trillion) and at least two orders of magnitude more functional genes.^9,10 It is shaped from the earliest time points in life and changes repeatedly based on environmental interactions.¹¹ As the microbiome develops, influences ranging from diet to siblings to disease shape the predominant bacteria^12–14 creating a highly diverse and unique microbial community.¹¹ The predominant bacteria in the gut microbiome are within the phyla of Bacteriodetes and Firmicutes and are commonly referred to as commensal organisms reflecting their ability to support a wide variety of normal functions to maintain homeostasis.^15–18

The developing and mature microbiome also plays a role in shaping local immune responses. Sampling of the intestinal lumen is constant, allowing for normal development of the immune system, as well as providing an opportunity for the mature immune system to detect new antigens. This is highlighted in germ-free animal models. These animals do not acquire an endogenous microbiome and as a result, have less adaptive immune development and function.^19–21

Bacterial products also play a role in health and disease. Short-chain fatty acids (SCFA) are the primary end-products of fermentation of non-digestible carbohydrates available to the gut microbiota.^17,22 Short-chain fatty acids such as acetate, propionate, and butyrate have been shown to play a role in restoring intestinal epithelial gut barrier dysfunction,^23,24 as well as signaling between bacteria and the immune system.²⁵

Sepsis results in a reduction of diversity and collapse of the normal intestinal microbiome. These changes have been referred to as a conversion from a microbiome to a pathobiome and occur almost immediately after a septic insult²⁶ (Fig. 1). The pathobiome can demonstrate a striking lack of diversity with some patients having an ultra-low diversity group of pathogens that number as few as four taxa. These changes are associated with higher morbidity and mortality in intensive care unit (ICU) patients.^27,28 Non-survivors of critical illness have been shown to have a higher abundance of Staphylococcus haemolyticus, Clostridiales sp., Campylobacter ureolyticus, Akkermansia sp., Malassezia sympodialis, Malassezia dermatis, and Saccharomyces cerevisiae, whereas survivors have a greater abundance of Collinsella aerofaciens, Blautia sp., Streptococcus sp., Faecalibacterium prausnitzii, and Bifidobacterium sp.²⁹ The causes of the transition from a microbiome to a pathobiome are multifactorial. Critical illness in and of itself alters gut microbial composition. In addition, unintended effects of several treatments aimed at improving outcomes in critically ill patients—including antibiotics, opiates, proton pump inhibitors, and nutrition support route—also play a factor in this transition.^30–32

FIG. 1.

Microbiome to pathobiome transformation. This figure demonstrates how stress events trigger a transformation from a healthy microbiome to a pathobiome.

The microbiome composition prior to the onset of sepsis may also impact outcomes after the onset of critical illness. In a recent pre-clinical study, genetically identical animals with different baseline microbiome diversity based on vendor of origin had divergent mortality after polymicrobial intra-abdominal sepsis. Animals with greater baseline diversity had improved survival and an altered immune response. This survival advantage was mechanistic because of the microbiome as co-housing the animals, resulting in animals having a similar microbiome regardless of their vendor of origin, led to abolition of mortality and immune differences.³³ Co-housing represents a unique approach to adoptive immunity that may be readily explored using animal models and begins to provide a mechanistic understanding for microbiome transfer therapies such as fecal microbiota transplantation.

As bacterial composition changes and diversity decreases in the setting of sepsis, SCFA production also changes. Stool samples in critically ill ICU patients demonstrate decreased SCFA levels with lower butyric and isobutyric acid production; these findings are associated with higher rates of infection.³⁴ Although convincing clinical data do not exist for SCFA administration in patients, pre-clinical models show promise with rescue of sepsis-induced mortality from pneumonia with oral SCFA administration after intestinal microbiome depletion after oral antibiotic administration.³⁵

Intestinal Epithelium

The intestinal epithelium is a single-cell–layer inner lining that separates the microbiome residing in the gut lumen from the sterile host.⁹ The epithelial barrier is semipermeable, controlled by intercellular tight junction proteins that allow passage of water, nutrients, and antigens to sustain the host and enable appropriate immune system function.^36–38 During sepsis, intestinal hyperpermeability and enterocyte apoptosis occur and result in barrier dysfunction that increases the antigens and bacterial translocation.^39,40 In pre-clinical models, both reducing apoptosis and barrier permeability improve survival after sepsis.^41,42 Furthermore, decreasing enterocyte apoptosis improves intestinal hyperpermeability, and suggests a link between enterocyte survival and gut barrier function competence.⁴³

Evidence also links the microbiome to the intestinal epithelium. Mice lacking CYP1A1, a cytochrome produced by intestinal epithelial cells, have an altered gut microbiome, but improved survival after intra-peritoneal methicillin-resistant Staphylococcus aureus (MRSA) injection.⁴⁴ This altered microbiome results in decreased cadaverine production, a compound that worsens intestinal permeability by activating the histamine H4 receptor/nfκB/MLCK signaling pathway. Mechanistically, the altered microbiome identified in septic knockout mice demonstrates low abundance of Enterococcus faecalis, which is a primary species responsible for cadaverine conversion from lysine. Importantly, re-introduction of that bacteria worsened intestinal permeability and abrogated the survival benefit seen in knockout mice. Notably, increased stool and blood cadaverine levels were also detected in critically ill patients infected with MRSA compared with healthy controls.

Short-chain fatty acids have also been shown to act as energy substances to protect intestinal barrier integrity.⁴⁵ Acetate, propionate, and butyrate alone or in combination significantly increase transepithelial resistance of CACO-2 cells in vitro, and stimulate tight junction formation. Further, SCFAs decrease lipopolysaccharide (LPS)-induced hyperpermeability and prevent morphological disruption of the tight junction proteins ZO-1 and occludin while simultaneously inhibiting LPS-induced activation of NLRP3 inflammasome and autophagy. Clearly, SCFAs appear beneficial with regard to intestinal barrier function.

Mucosal Immune System

The mucosal immune system provides protection to the host in multiple ways.⁴⁶ This protection is conferred, in part, through constant surveillance of the intestinal lumen and effector action to maintain luminal homeostasis or act if there is barrier breach.^47–49 Furthermore, there is a symbiotic relation between the gut microbiome with the immune system across the intestinal epithelium whereby gut microbes influence and shape the immune system milieu and vice versa.^50,51

The first line of immunologic defense is the innate immune system. Resident microbes interact with the innate immune system in an antigen non-specific manner through pathogen recognition receptors.⁵¹ Other innate system cells are then activated, releasing cytokines, and producing antimicrobial molecules. For the adaptive immune system, antigen presenting cells, such as dendritic cells, sample contents from the intestinal lumen and present them to lymphocytes in the mesenteric lymph nodes to stimulate differentiated T cells and B cells.⁵²

Sepsis alters the immune system, and in turn, its ability to respond to gut dysbiosis. Neutrophils are a key effector of the innate immune system and act as a first line of defense against infection through an arsenal of phagocytosis, release of chemoattractants, and creation of neutrophil extracellular traps. When neutrophils have early access to the site of infection, sepsis mortality is reduced, at least in pre-clinical studies.⁵³ Another population of innate immune cells that have been recently described are innate lymphoid cells (ILC). These cells are predominantly found at mucosal sites and have the ability to respond to bacterial exposure. Innate lymphoid cells are technically lymphocytes because they have a common lymphoid progenitor within the bone marrow.⁵⁴ Furthermore, ILCs have been shown to be induced by the microbiome,⁵⁵ and depletion of the microbiome in pre-clinical studies results in a loss of intestinal ILCs with subsequent disappearance of neutrophils and an increased susceptibility to neonatal sepsis.^56,57 Additionally, when ILCs are lost, systemic dissemination of commensal bacteria results.⁵⁸

Sepsis also induces adaptive immune dysfunction with increased sepsis-induced T cell apoptosis playing a crucial role in mediating mortality.^58,59 The microbiome's relation with the adaptive immune system may also play a role in regulating intestinal epithelial cell apoptosis and gut barrier dysfunction. Septic RAG^-/- animals that lack mature T lymphocytes have worsened enterocyte apoptosis, but the absence of a gut microbiome reverses the apoptosis back to the level noted in wild-type controls.⁶⁰ B cells are the other component of the adaptive immune system and have a relation with the microbiome through immunoglobulin A (IgA) induction. Through sampling of the luminal contents, the microbiota stimulate IgA production that, in turn, feeds back and prevents pathogenic bacteria from overgrowing while supporting commensal microbiota proliferation.^61,62 In the setting of critical illness, IgA also plays a protective role. Mice exposed to bacteria in the phylum Proteobacteria produce more IgA specific to these pathogenic bacteria and are protected against intra-abdominal sepsis.⁶³

Recent pre-clinical mechanistic experiments demonstrate crosstalk connecting the gut barrier with the immune system, and the gut microbiome.⁶⁴ Junctional adhesion molecule A (JAM-A) is a tight junction protein upregulated during sepsis.³⁹ When JAM-A is constitutively deleted from mice, intestinal permeability is increased.⁶⁵ When sepsis is induced in knockout animals, intestinal permeability worsens, whereas survival paradoxically improves as a reflection of reduced systemic inflammation and bacteremia. Additionally, the microbiome of JAM-A^-/- animal's reveals increases in the phylum Proteobacteria. To determine potential mechanisms underlying the improved survival in knockout mice, a survey of the immune system disclosed increased T cells with decreased cytokine production, as well as higher circulating and gut IgA. However, JAM-A^-/- x RAG^-/- mice have improved survival compared with RAG^-/- mice suggesting that additional mechanisms beyond the adaptive immune response are responsible for the survival benefit. Examination of the innate immune system demonstrates that neutrophil phagocytosis is increased in JAM-A^-/- animals. Importantly the overall survival benefit is lost after neutrophil depletion. Thus, increased neutrophil phagocytosis of opsonized pathogenic microbiome bacteria in the setting of baseline intestinal hyperpermeability represents an integrated mechanism through which multiple components of the gut potentially drive the host response to sepsis.

Targeting the Microbiome in the ICU

In light of evidence demonstrating an association between the pathobiome and poor outcomes in critical illness, multiple approaches have taken aim at targeting the intestinal microbiome as a potential therapeutic intervention. Current methods include probiotics, prebiotics, fecal microbial transplantation (FMT), and selective decontamination of the digestive tract (SDD). Although evidence supporting each of these approaches varies, data are not compelling enough to recommend any of them as standard of care for microbiome manipulation to improve outcomes from critical illness.

Probiotics are enteral supplements containing live micro-organisms with the goal of diversifying the gut microbiome to improve health or outcome from disease. Probiotics have been explored extensively as part of the care approach for critically ill patients. The largest study in adult patients was a randomized placebo-controlled trial across 44 ICUs with more than 2,500 patients examining the effects of Lactobacillus rhamnosus GG on patients expected to require mechanical ventilation for greater than 72 hours.⁶⁶ No difference was noted in the primary outcome of ventilator-associated pneumonia incidence. Additionally, none of 20 pre-specified secondary outcomes demonstrated any differences after probiotic supplementation. Notably, 15 patients had the adverse effect of the administered probiotic being recovered from a sterile site or as the sole bacteria identified in a non-sterile site. These data suggest that enterally supplemented bacteria may escape the gastrointestinal tract lumen during critical illness as a manifestation of increased intestinal permeability, reduced barrier function, and impeded immune competence.

These results are somewhat disparate from a number of smaller studies that showed that probiotics decrease ventilator-associated pneumonia in patients receiving invasive mechanical ventilation with variable impact on secondary outcomes. In light of the conflicting data often derived from low-quality studies, multiple meta-analyses have been recently published on the impact or probiotics.^67–70 The largest of these includes 65 randomized controlled trials of 8,483 patients.⁶⁷ This meta-analysis shows that probiotics may reduce ventilator-associated pneumonia, healthcare-associated pneumonia, ICU length of stay, hospital length of stay, and duration of mechanical ventilation without altering mortality. Notably, the benefits identified with probiotics were negated in post hoc sensitivity analyses when excluding studies with a high risk of bias. It is important to note that if there is a benefit of probiotics, numerous uncertainties remain including optimal strain(s) of bacteria, concentration, timing of initiation, duration of administration, and patient population who might benefit. Additionally, patient populations are also subject to substantial heterogeneity because baseline microbiome characteristics that are individually unique prior to probiotic administration may influence study outcomes.⁷¹ It should also be noted that several case reports and case series have reported evidence of bacteremia with organisms administered as probiotics.^72,73 Therefore, even therapeutics that appear relatively benign may have deleterious impact on those with altered gut and immune function. Nonetheless, a closely related therapeutic, prebiotics have been well explored as a perhaps less risky microbiome support approach.

Prebiotics are indigestible materials that promote commensal bacterial growth. Prebiotics differ from probiotics in that they do not contain living organisms. Fiber is the most common prebiotic investigated. Fiber acts as a microbial food source for SCFA-producing organisms. Pre-clinical sepsis models have shown promise as they demonstrate decreased mortality in mice treated with a high-fiber diet for two weeks prior to sepsis onset compared with mice receiving low to normal fiber intake.⁷⁴ Although the clinical relevance of a pre-treatment model is limited, this is potentially mechanistically important as the survival benefit induced by fiber was both microbiome-mediated and negated by antibiotic administration. A meta-analysis of 19 studies of fiber use in critically ill patients showed that enteral formulas with fiber may help reduce diarrhea incidence and severity as well as overall gastrointestinal complications without impacting mortality, ICU length of stay, or hospital length of stay.⁷⁵ The certainty of these conclusions is low due to lower quality of many of the included studies. Additionally, a pilot study including 20 critically ill patients on broad-spectrum antibiotic agents given enteral fiber compared with no fiber demonstrated a trend toward increased relative abundance of SCFA-producing bacteria and increased SCFA levels in fiber receiving patients.⁷⁶ Butyrate also rescues LPS-induced decreases in mitochondrial respiration in vitro, suggesting a potential additional mechanism through which SCFAs may help to mitigate the immunoinflammatory response in sepsis.⁷⁷ Prebiotics may be most effective when a relatively normal microbiome remains resident and functional within a patient's gastrointestinal tract. For those with a substantially deranged microbiome, one well documented consequence is acute illness from Clostridium difficile overgrowth.

Fecal microbiota transplantation is a technique by which the entire native intestinal microbiome is replaced with a donor stool sample. This procedure is theoretically beneficial when there has been substantial alteration of the normal host diversity with an increase in normally low concentration pathogens. The ultimate goal is to restore normal microbial diversity by providing high volumes of commensals and preventing re-colonization of pathogenic bacteria. This dual approach theoretically allows for restoration of normal adaptive host metabolism and immune function.

Fecal microbiota transplantation has been studied most extensively in patients with recurrent as opposed to initial Clostridium difficile infection. Whereas oral antimicrobials remain the first-line treatment of Clostridium difficile, treatment failure occurs in a subset of patients.⁷⁸ Based on robust efficacy data, FMT has become part of medical professional organization recommendations for recurrent disease management.^79,80 With an increased understanding of sepsis-induced dysbiosis, there has been increased interest in the use of FMT in the ICU. Mechanistically, recent pre-clinical work suggests that FMT restores SCFA-producing microbial communities that, in turn, stimulate immune function to clear overgrown pathogens. Mortality is reduced after FMT in mice in an interferon regulatory factor 3-dependent manner with an associated increase in SCFA-producing Bacteriodetes.⁸¹ Additionally, FMT after murine sepsis improves gut barrier function by reducing epithelial cell apoptosis, improving the composition of the mucus layer, reducing intestinal permeability with upregulation of tight junction proteins, and improving the host inflammatory response.⁸²

Despite the conceptual appeal of FMT and pre-clinical data supporting its use, numerous barriers currently exist to use FMT routinely in critically ill patients. A major concern is that most critically ill patients receive antibiotic agents during the ICU stay, a therapy that would be expected to maladaptively alter the transplanted flora after FMT. Furthermore, the implications of transplanting live microbes into a host often encumbered by sepsis-induced immunosuppression is unknown. Additionally, the lack of standardization of dose, frequency, and bacterial composition make therapy generalization quite difficult. Finally, there have been case reports of multi-drug–resistant organism bacteremia in patients receiving FMT including one attributable death.⁸³ Although such a dire complication is extremely rare, it reinforces the need to focus on safety if this approach is to be used for critically ill patients who demonstrate compromised immune competence that is related, at least in part, to the gastrointestinal tract microbiome and gut barrier function. Accordingly, despite promising case reports documenting benefit, FMT as a routine therapy for critically ill patients remains experimental.⁸⁴ Instead, a durably investigated approach leverages a common therapeutic, namely oral antibiotic therapy to alter gastrointestinal tract bacterial bioburden.

Loosely related to colonic preparation in advance of resectional surgery with a planned anastomosis, SDD eliminates bacteria from the intestinal tract using oral antibiotic therapy. This strategy relies on removing pathogenic bacteria that would sense a host stressor and become more virulent. Unfortunately, the process is non-specific and also removes commensal organisms. The most common application for this technique is for patients undergoing elective bowel surgery because pre-operative bowel preparation that includes antibiotic agents has been shown to reduce anastomotic leaks and surgical site infections.⁸⁵ In critically ill patients, there is robust evidence regarding SDD. A meta-analysis of 41 trials (most with poor reporting quality) and more than 11,000 patients receiving invasive mechanical ventilation for at least 48 hours showed that SDD (using both oral and system antibiotic agents) decreased mortality and respiratory infections.⁸⁶

Furthermore, oral antibiotic agents alone may decrease mortality and respiratory infection. It is important to note, however, that these studies were performed mostly in a single country characterized by a low baseline antibiotic resistance profile and may not be directly applicable to developed nations with substantial resistance profiles. In contrast, a randomized controlled trial of more than 8,600 patients in ICUs with moderate to high antimicrobial resistance failed to find any SDD benefit in decreasing infections or mortality.⁸⁷ Given the conflicting data on SDD and the concern for amplifying antibiotic resistance as a result of using SDD, this therapy has not become routinely used throughout most of the world. Nonetheless, it should be noted that a recent study of more than 6,000 patients receiving either oral antibiotic agents or no therapy over a five-year period revealed no differences in the incidence of multi-drug–resistant bacteria, except for a lower rate of ESBL-producing Klebsiella pneumoniae and a higher rate of vancomycin-resistant Enterococcus faecium in the treatment group.⁸⁸ Therefore, SDD remains a tantalizing but untargeted therapeutic that may benefit from specific refinement in a precision fashion that targets uniquely identified pathogenic bacteria present in a specific patient. The ability to provide such precise targeting remains elusive at present.

Conclusions

Our understanding of the interface of the gut microbiome with the immune system during health and disease continues to expand. Communication between the intestinal microbiome, the intestinal epithelium, and the immune system prior to and after sepsis impact intestinal barrier function and lead to local and systemic changes that influence outcome. Although targeting the pathobiome has conceptual appeal, more research needs to be done to determine the best approach (if any) toward manipulating the microbiome in critically ill patients for therapeutic gain.

Funding Information

This work was supported by funding from the National Institutes of Health (GM072808, GM104323, AA027396)

Author Disclosure Statement

The authors have no conflicts of interest to disclose.