During septic ALI/ARDS, increased cytokine levels may cause impairment of the intestinal barrier. This selective barrier is composed of intestinal mucosal epithelial cells and inter-cell connections. When the selective barrier functions normally, it allows the cell-side movement of water, solutes, and immunomodulatory factors but prevents the movement of macromolecules and microorganisms (Weström et al., 2020). Tight junctions between cells play a critical role in the gut barrier function, and the cells are connected to the intracellular cytoskeleton by tight junction proteins [e.g., occludin, claudin family, junctional adhesion molecules (JAM), myosin light chain, zonula occludens (ZO); Krug and Fromm, 2020]. The increase of inflammatory cytokines during sepsis results in intestinal hyperpermeability through the up-regulation of claudin 2 and JAM and down-regulation of claudin 5 and ZO-1 (Yoseph et al., 2016). Alternatively, sepsis can also redistribute claudins 1, 3, 4, 5, and 8, resulting in intestinal barrier dysfunction (Li et al., 2009). Claudins are a protein family of up to 27 members in mammals. Expression of claudins 1–19 has been examined throughout the rat and mouse intestine and that of claudins 20–24 in the mouse upper small intestine. Members of the claudin family are major driving forces in the formation of paracellular barriers (Günzel and Yu, 2013). Claudins can be roughly divided into sealing or pore-forming claudins. Alteration of claudin expression may result in a decreased or increased paracellular transport of solutes as well as an increased permeability to macromolecules (Günzel, 2017). To date, claudin-2, -10b, and -15 qualify as cation pores and claudin-10a and -17 as anion pores, which are both acknowledged as pore-forming claudins (Odenwald and Turner, 2013). Therefore, an increase in a pore-forming protein, such as claudin-2, would directly cause hyperpermeability and deterioration of the barrier function, while a decrease in a sealing protein, such as claudin-5, would lead to the same result in a mechanistically complementary manner. Beyond that, the activation of myosin light chain kinase (MLCK) by inflammatory cytokines (TNF-α, IL-6, and IL-1β) is also associated with paracellular hyperpermeability. Cytokines further activated MLCK in a feed-forward mechanism, partly via altering ZO-1 and claudin 15 (Lorentz et al., 2017). MLCK phosphorylates the myosin regulatory light chain, which leads to contraction of the actin-myosin ring, increasing intestinal permeability (Su et al., 2020). Indeed, the significant increase of the intestinal permeability is closely interrelated to the occurrence and development of sepsis (Assimakopoulos et al., 2018), such as sepsis secondary intratracheal instillation of Pseudomonas aeruginosa due to gut mucosal and microvascular injury and gut barrier dysfunction (Yu and Martin, 2000). Recent studies have reported that gut microbes, represented by Bacteroidetes and Enterobacteriaceae, translocate across the intestinal mucosa and even enter the lung in sepsis and ARDS patients (Dickson et al., 2016; Panzer et al., 2018). However, there is no direct evidence that gut-derived bacteria or bacterial products, such as endotoxin, are present in the blood of septic patients (Haussner et al., 2019). Studies have suggested that ligation of mesenteric lymphatic vessels attenuates lung injury and neutrophil activation and improves survival in mice with endotoxemia (Badami et al., 2008). Moreover, mesenteric lymphatic vessels collected from critically ill mice induces lung injury upon intravenous injection in healthy mice (Senthil et al., 2007). Thus, translocation via intestinal lymphatic vessels is regarded (Haussner et al., 2019), which is called gut–lymph hypothesis. In detail, bacterial translocation gives rise to a local activation of the mucosal immune system (MIS). MIS stimulates the production of inflammatory substances that enter the lung and systemic circulation through the mesenteric lymphatics. These danger-associated molecular patterns (DAMPs) are recognized by innate immune cells to promote pro-inflammatory pathways. A massive release of DAMPs accelerates the progression of organ damage and multiple organ dysfunction syndrome (MODS), further aggravating gut injury, forming a vicious cycle (Deitch, 2012; Assimakopoulos et al., 2018; Haussner et al., 2019). Therefore, the destruction of the gut barrier contributes to the translocation of gut microbes and/or their products to the systemic circulation via the mesenteric lymphatics, exacerbating significant host inflammation and acute pulmonary edema with a positive feedback (Figure 2).

Notably, toll-like receptor 4 (TLR4) is widely expressed in intestinal stem cells and regulates their proliferation or apoptosis (Chen et al., 2018). During sepsis, increased cytokine levels inhibit intestinal cell regeneration and promote apoptosis in a TLR4-dependent manner (Mazmanian et al., 2008). At the cellular level, crypt cell proliferation is markedly decreased (Coopersmith et al., 2003), with increased intestinal epithelial cells (IECs; Coopersmith et al., 2002) and crypt cell apoptosis (Perrone et al., 2012) following sepsis. Although IECs migrate in a TLR-4-dependent manner (Neal et al., 2013), changes in IEC proliferation and apoptosis far exceed this slow course, leading to a shorter villi length (Dominguez et al., 2011). Moreover, cytokines cause an abnormal intestinal and barrier function of the mucous layer, characterized by reduced thickness, diminished lumen coverage, and poor adhesion (Chang et al., 2012). All of these factors further decrease the effectiveness of the gut barrier.

All in all, an increase of inflammatory cytokines induces gut barrier dysfunction, intestinal hyperpermeability, bacterial translocation, and amplification of the inflammatory responses. Subsequently, this expansion of the systemic inflammatory response contributes to lung injury.

Sepsis-associated ALI/ARDS results in the alteration of lung microbiota both in the mice model and in patients. The etiology of pulmonary dysbiosis in sepsis patients is complex and includes endogenous (e.g., hypoxia and ischemia-reperfusion injury) and external factors (e.g., endotracheal intubation, mechanical lung ventilation, inhaled oxygen, and antibiotics). Poroyko et al. (2015) showed in an ALI mouse model induced by the intratracheal instillation of lipopolysaccharide (LPS) that the bacterial DNA burden in bronchoalveolar lavage (BAL) was increased fivefold, whereas the community complexity measured by the Shannon diversity index was significantly decreased. The major trend in the microbial community reaction to LPS-induced ALI was the loss of Firmicutes, represented by Alicyclobacillaceae, and the bloom of Proteobacteria, represented by Brucellaceae and Xanthomonadaceae. Dickson et al. (2016) also found that the community richness was also higher in a mouse model of sepsis induced by caecal ligation and puncture (CLP). The lower respiratory tract was rapidly enriched with bacteria in the gastrointestinal tract, including members of the Bacteroidales order, Enterococcus species, and Lachnospiraceae species, and remained in this status for 5 days. Furthermore, they sequenced lung microbiota of mice exposed to intratracheal LPS to model direct lung injury and observed enrichment of Enterobacteriaceae-classified OTU in post-sepsis lungs. Based on the above experimental evidence, they further analyzed bacterial communities from the BAL fluid of 68 patients with ARDS. The gut-associated Bacteroides OTU, the most abundant genus in the human gut microbiome, were common in the lungs of ARDS patients (41%) but absent in the lungs of the healthy controls. Furthermore, the gut-associated Bacteroides OTU are most strongly correlated with the severity of systemic and alveolar inflammation. Another study conducted by Dickson reconfirmed prior findings (Dickson et al., 2020). They found the bacterial load was greater in patients with ARDS. In addition, they compared the lung bacterial community composition in BAL specimens from ARDS and non-ARDS patients. The predominant clusters in non-ARDS patients were Streptococcaceae, Veillonellaceae, Prevotellaceae, Verrucomicrobiaceae, and Flavobacteriaceae, which are common in healthy lungs and negative sequence control samples. However, in ARDS patients, the bacterial community in the lungs was characterized by Pasteurellaceae and Enterobacteriaceae. Similar to their prior results, gut-derived Enterobacteriaceae are also correlated with the ARDS status. Notably, the Enterobacteriaceae taxonomy was highly aligned with the ARDS-associated bacterial taxon, as identified by Panzer et al. (2018). They also proved that the microbiome is associated with ARDS development. A recent clinical study (Schmitt et al., 2020) compared the pulmonary microbiota in 15 patients with sepsis-induced ARDS undergoing abdominal surgery and 15 patients post esophagectomy. In the ARDS group, the α-diversity index of the pulmonary microbiome was significantly decreased, which was related to the length of the ICU stay and the need for ventilator use, suggesting that alteration of the lung microbiome may represent a mechanism of pathogenesis in septic ALI/ARDS.

Changes in the lung flora may affect the composition of the gut flora in sepsis-induced ALI/ARDS. One previous study manifested that lung microbiota imbalance in a sepsis-related ALI murine model increased the total bacterial count in the cecum (Sze et al., 2014). In clinical cases, the most common source of sepsis is the lung (Wang et al., 2013). Rosa et al. (2020) found that Vancomycin treatment of acute Pseudomonas aeruginosa pneumonia in mice can induce intestinal dysbacteriosis, resulting in an increase in the number of Proteus, a decrease in the number of bacteroides, and inflammatory changes in the intestinal tract. After fecal microbiota transplantation, the susceptible phenotype and tissue injury phenotype were reversed in mice. Moreover, the pulmonary allergic response also influences the composition of the intestinal microbiota (Vital et al., 2015). In the context of experimental influenza infection, it has been reported that the IFN-γ produced by lung-derived CCR9 + CD4 + T cells changed the composition of the gut microbiota, and caused intestinal immune injury (Wang et al., 2014). Moreover, the pulmonary production of IFN-Is promotes the consumption of obligate anaerobic bacteria and the enrichment of proteobacteria in the gut, leading to significant intestinal dysregulation (Deriu et al., 2016). Thus, lung inflammation directly changes the intestinal bacterial community structure and further worsens lung inflammation (Vital et al., 2015). However, only very limited data have been reported on how long dysbiosis in sepsis-induced ALI/ARDS causes gut dysbiosis. The alveolo-capillary membrane becomes increasingly permeable in sepsis-induced ALI/ARDS as a result of a direct (primarily epithelial) or indirect (primarily endothelial) injury. It is reasonable to posit that alveolo-capillary permeability might be at the highest risk of gut–lung bacterial translocation. Considerable efforts are needed to increase our knowledge about the influence of post-sepsis lung injury on gut microbiota.

The human gastrointestinal tract is the harbor of 100 trillion bacteria, which are ten times more abundant than somatic and germ line cells of the human body (Morais et al., 2021). Regarding human health, the gut microbiota contributes to prevent infections caused by pathogens, provide nutrients, participate in metabolism, shape the immune system, and serve as a biological barrier (Haak et al., 2018). In turn, the immune system will affect the microbiota composition (Honda and Littman, 2016). In various disease states, the loss of “health-promoting” bacteria and overgrowth of “disease-promoting” pathogenic bacteria make patients more succumb to sepsis and MODS (Alverdy and Krezalek, 2017). A multi-center study has shown that patients with sepsis have an increased abundance of intestinal microbiota, which is closely related to inflammation caused by Parabacteroides, Clostridium, Bilophila, and other species. Concomitantly, researchers have detected an increased abundance of Enterococcus and other pathogenic bacteria in sepsis patients who died, suggesting that these bacteria may be potential biomarkers for ICU care (Agudelo-Ochoa et al., 2020). It has also been shown that septic patients have a decreased abundance of Faecalibacterium associated with reduced intestinal inflammation (Lankelma et al., 2017). In a single-center case control study of children, the abundance of the following 13 bacteria in septic children were significantly higher than that in the healthy control group: Proteobacteria, Bacilli, Gammaproteobacteria, Enterobacteriales, Pseudomonadales, Lactobacillales, Enterococcaceae, Enterobacteriaceae, Moraxellaceae, Enterococcus, Clostridium innocuum group, Acinetobacter, and Enterococcus durans. Among these bacteria, Enterococcaceae, Enterococcus, and Enterococcus durans showed an increase in their abundance that is positively correlated with the inflammatory indicators CRP and WBC. Furthermore, the abundance of the following seven bacteria was significantly decreased in the guts of septic children: Bifidobacteria, Selenomonas, Aminococcus acidaceae, Daentosaceae UCG-003, Dialister, Dorea longicatena, and Ruminococcus sp.5_1_39bFAA. The decrease of intestinal bifidobacteria abundance is negatively correlated to WBC (Liu et al., 2021). The mechanisms behind these shifts in the microbiota composition are unclear (Haak and Wiersinga, 2017). One of the reasons for this phenomenon may be clinical intervention, such as enteral/parenteral feeding, selective oral decontamination/selective decontamination of the digestive tract, as well as the administration of antibiotics, proton pump inhibitors, opioids, catecholamines, and sedatives (Haak and Wiersinga, 2017). In addition, due to the acute phase of inflammatory response and intestinal dysfunction, critically ill patients with sepsis have a greater risk of malnutrition (Liu et al., 2014). Nutritional deficiency is associated with intestinal dysbiosis, characterized as an increase in proteus numbers and a decrease in α-diversity, and it makes epithelial barrier function weaken, which predisposes to bacterial translocation (Ralls et al., 2015; Levesque et al., 2017). And decreased gastrointestinal motility and loss of intestinal epithelial integrity in sepsis patients cause a decrease in anaerobic bacteria (e.g., Lachnospiraceae and Ruminococcaceae), which further impairs the intestinal epithelial function and allows the proliferation of opportunistic pathogenic bacteria (Haak et al., 2018). Fay et al. (2019) examined genetically identical septic C57BL/6 mice from two vendors with different microbiome compositions. Following CLP, significant differences were noted in the mortality and immunophenotype (especially splenic IFN-γ + CD4 + T cells, effector memory CD4 + T cells, central memory CD4 + T cells, and Peyer’s patch effector memory CD4 + T cells). In addition, CLP was performed of naive mice from different suppliers after 3 weeks of cohousing, and the differences in the mortality and immunophenotype disappeared. These findings suggest that the microbiome plays a critical role in the survival of sepsis and the host immune response. This conclusion has also been confirmed by other studies. Depletion of intestinal microbiota with broad-spectrum antibiotics could exacerbate lung and intestinal damage and increase the mortality in sepsis (Xu et al., 2021). Notably, the differential immune responses to sepsis are determined by the microbiome (Cho and Blaser, 2012). A recent study has shown that the lung microbiome is enriched with gut-derived bacteria in murine sepsis and in human patients with ARDS. The presence of gut-specific communities (Bacteroides) is correlated with the intensity of systemic inflammation (Dickson et al., 2016). When the bacterial burden of the gut is minimized, the inflammation and injury sustained with distal organs are lessened during sepsis (Cuevas et al., 1972). Imbalance of the intestinal microbiome regulates the Toll-like receptor 4 (TLR4)/nuclear factor-κB (NF-κB) signaling pathway in the lung immune system, which activates pulmonary oxidative stress and mediates lung injury (Tang et al., 2021). Enhancing the α-diversity of the intestinal flora in mice changes the immune response to sepsis and improves the survival rate of sepsis, which is mediated by the powerful CD4 + T cell response (Fay et al., 2019). The intestinal flora affects not only the cellular immunity but also the humoral immunity. Commensal bacteria directly produce IgA, which has a protective activity against mucosal barrier disruption and the resulting sepsis (Wilmore et al., 2018). These studies indicate that the gut is the predominant motor of sepsis (Klingensmith and Coopersmith, 2016).

In recent years, with the developed in-depth understanding of gut microecology, studies have shown that not only gut microbiota itself but also the metabolites of gut microbiota are involved in various functions of the intestinal microenvironment. Bacteria can produce a variety of metabolites, including: SCFAs, vitamins, bile acids, choline metabolites, aromatic compounds, amines, etc., (Van Treuren and Dodd, 2020). In the following sections, we summarized several metabolites affecting the occurrence and development of sepsis.