Introduction

Sepsis is not only a process of systemic inflammatory response or immune disorder, it rather involves changes in the function of multiple organs in the body. As depicted in Figure 1, on the cellular and molecular levels, the pathogenesis of sepsis is extremely complex, including imbalance in inflammatory response, immune dysfunction, mitochondrial damage, coagulopathy, neuroendocrine immune network abnormalities, endoplasmic reticulum stress, autophagy, and other pathophysiological processes, and ultimately leads to organ dysfunction, which will be dissected in the following.

Figure 1. The complex pathogenesis of sepsis.

Inflammation Imbalance

Inflammatory imbalance represents the most critical basis of sepsis pathogenesis and occurs throughout the whole process of sepsis, and the pathogens eliciting the response include organisms such as bacteria, fungi, parasites, and viruses. The host’s initial acute response to invasive pathogens typically causes macrophages to engulf the pathogens and produce a range of pro-inflammatory cytokines, and this can trigger cytokine storms and activate the innate immune system [20,21]. Obviously, the activation of the innate immune system is mediated by pattern-recognition receptors (PRRs), which initiate a series of activation in immune cells by detecting damage-associated molecular patterns (DAMPs) or pathogen-associated molecular patterns (PAMPs), and thus upregulate the expression of inflammation-related genes [22]. In the immune response to sepsis, both exogenous factors derived from the pathogen (e.g., lipopolysaccharide (LPS)) and endogenous factors released by injured cells (e.g., high-mobility group box-1 (HMGB-1) protein) can interact with various PRRs, such as Toll-like receptors (TLRs), C-type lectin receptors (CLRs), RIG-I like receptors (RLRs), and NOD-like receptors (NLRs) [22,23]. Among these receptors, TLRs have been most widely studied. The interaction between TLRs and their ligands is induced by their TIR domains, which leads to the activation of c-Jun N-terminal kinase (JNK), extracellular signal-regulated kinase 1/2 (ERK1/2), p38 mitogen-activated protein kinase (MAPK), and nuclear factor-κB (NF-κB) signaling pathways through the myeloid differentiation factor 88-dependent pathway. These events are followed by the production of inflammatory cytokines such as interleukin (IL)-1, IL-6, tumor necrosis factor-α (TNF-α), interferon (IFN) regulatory factor 7 (IRF7), and adaptor protein 1 (AP-1) [24]. These cascades of events of multiple signaling pathways are tightly controlled processes that involve many cytoplasmic and membrane-bound proteins, such as interleukin-1 receptor-associated kinase-M (IRAK-M), toll interacting protein (TOLLIP), suppressor of cytokine signaling 1 (SOCS1), single Ig IL-1R-related molecule (SIGIRR), growth Stimulation expressed gene 2 (ST2), and so on [25]. Moreover, TLR signaling is also regulated by a tight control of TLR expression on the cell membrane. Thereby, TLR4 and TLR2 mRNAs are highly expressed in patients with sepsis [26].

In addition, soluble cytosolic PRRs such as NLRs are also involved in sepsis-induced immune imbalance. NLRs contain the nucleotide-binding oligomeric domain (NOD) and LRR domain (similar to TLRs) [27]. The activation of some NLRs is regulated by the adaptor protein receptor-interacting protein kinase 2 (RIP2) (also known as RICK)  leading to the activation of NF-κB and AP-1, while some other NLRs (such as NLRP and NLRC4) participate in the formation of protein complexes referring to the inflammasome [28]. The inflammasome cleaves the caspase-1 precursor into active caspase-1, and activated caspase-1 cleaves IL-1β and IL-18 precursors to release the cytokines IL-1β and IL-18 [29].

Thirdly, the CLR family includes dectins, DC-SIGN, and mannose-binding lectins. Dectins induce reactive oxygen species (ROS) production and activates inflammatory responses through Src and Syk kinases, whereas DC-SIGN participates in the recognition of Leishmania, viruses, and fungi, which acts through Raf-1-mediated signaling pathway; however, the negative regulation of CLR responses require further study [30].

Interestingly, in addition to the above-mentioned PPRs, the receptors for double-stranded RNA including RIG-I, MDA5, and LGP2 have been also found to be involved in sepsis-induced immune dysfunction [31–35].

PRRs can be activated by exogenous PAMPs and endogenous DAMPs. In the case of endogenous sepsis, liver cells are reported to release a large amount of HMGB-1, which binds to bacterial endotoxin (LPS); the bacterial endotoxin is transported to the cytoplasm through RAGE receptors expressed on vascular endothelial cells and macrophages, and this leads to cysteinase caspase-11-mediated cell death (pyroptosis) and results in shock, multiple organ failure, and death [36,37].

Immune Dysfunction

The pathogenesis of sepsis includes a decrease in HLA-DR, lymphocyte replication, programmed cell death/apoptosis induction, anti-inflammatory molecules expression increasing, and cell-associated co-suppressor receptors and ligands upregulation [38,39].

When inflammation occurs during sepsis, neutrophils interact with endothelial cells and migrate, driven by chemokines, to the inflammation site, where they recognize and phagocytose pathogens, release various active factors and proteolytic enzymes, and eliminate pathogens [40]. Mononuclear/macrophage cells are activated when stimulated by cytokines (e.g., granulocyte-macrophage colony-stimulating factor (GM-CSF), TNF-α, INF-γ) or by pathogenic microorganisms, chemical mediators, immune complexes, etc., and the activated cells phagocytose and kill multiple pathogens and present antigens³⁰. Differentiated effector T cells further promote the activation of macrophages and secrete a large amount of active medium to cause damage and fibrosis of the tissues [41]. During sepsis, the maturation process of dendritic cells (DCs) of the spleen and lymph nodes has been impeded during sepsis [42]. During sepsis, DC activation also causes a rapid accumulation of innate immune cells (including monocytes, natural killer (NK) cells, and granulocytes). Monocytes play an important role in the pathophysiology of sepsis. In the sepsis patients, the defects of monocyte metabolism are an expression of immunosuppression, which is characterized by an extensive inhibition of metabolic processes such as glycolysis, fatty acid oxidation, and oxidative phosphorylation [43]. NK cells can accumulate to produce elevated levels of INF-γ, but lost the ability to support the Th1 immune response required for the clearance of bacterial infection [44]. Although a large portion of sepsis patients might die during the initial cytokine storm, patients who survive this stage could develop immunosuppression, which includes failure to clear primary infections, the development of secondary opportunistic infections, and the reactivation of potential viruses. Sepsis-induced immunosuppression involves both innate and adaptive immunity. Immunosuppression after sepsis has been described as compensatory anti-inflammatory response syndrome²⁷, and is regulated by co-stimulatory molecules such as CD80/B7-1, which are produced by activation of the TLR signaling pathway, and the naive T cells transformed into regulatory T cells induced by cytokines, and this results in a reduced expression of antigen presentation-related transcription factors (e.g., IRF4, MUM1) [45–47].

2.3. Mitochondrial Damage

Mitochondria are the major micro-organelles involved in energy production, protein synthesis, and catabolism. However, sepsis-induced mitochondrial damage or dysfunction can result in cellular metabolic disorders, insufficient energy production, and oxidative stress, which give rise to the apoptosis of organ cells and immune cells, thus ultimately generate immune disorders, multiple organ failure, and even death.

As depicted in Figure 2, during sepsis, on account of a limited oxygen supply and incomplete oxidative reaction, as well as hypoxia, the free radical production increases dramatically while the machinery of the antioxidant system becomes damaged [48]. When exposed to DAMPs or PAMPs, activated leukocytes release inflammatory cytokines, which trigger the expression of NADPH oxidase [49]. As shown in the Figure 2, cytokines cause an overproduction of reactive nitrogen species (RNS) and NO by promoting inducible nitric oxide synthase (iNOS) activity. NO can bind to ROS peroxides to form RNS, and this leads to an irreversible inhibition of electron transfer chain (ETC) activity. Given this dysfunction of the ETC, mitochondria themselves represent a source of additional ROS production during sepsis, which unfortunately brings about further damage to mitochondria, including intimal damage, an inhibition of ETC activity, and mitochondrial DNA damage [50]. Ultimately, the mitochondrial matrix swells, the mitochondrial membrane ruptures, and apoptosis is initiated. A high rate of apoptosis takes place among splenic lymphocytes and among cells of other organs during sepsis, and an inhibition of apoptosis by using caspase inhibitors has been found to increase the survival rate in sepsis [51]. During LPS-induced sepsis, the nuclear respiratory factor-1 (NRF-1), which is a transcriptional activator of mitochondrial transcription factor A (TFAM), is upregulated in hepatocytes [52]. Autophagy is activated in response to the clearance of irreversibly damaged mitochondria. Mitochondrial biogenesis is regulated by the AMPK/PGC-1α/NRF-1/2 signaling pathway; nevertheless, the disproportionality of the ATP/ADP ratio as a result of insufficient ATP production disrupts the activation of AMPK and the subsequent PGC-1α/NRF-1/2 pathway and thereby contributes to TFAM expression [53,54]. As a transcription promoter, TFAM translocates into the mitochondrial matrix and causes mitochondrial DNA expression after mitochondrial biogenesis. It appears that the mitochondrial density continues to decline after a severe sepsis episode [55]. More experimental investigations remain to be launched for the role of the mitochondrial control mechanisms especially in inducing the multiple organ failure that is characteristic of sepsis and as potential therapeutic targets [56–58].

Figure 2. The regulation mechanisms of mitochondrial damage during sepsis.

Coagulation Disorders

The interaction between inflammation and coagulation is widely considered to be a key point in the pathogenesis of sepsis. Inflammation can induce a coagulation reaction in sepsis, and activation of the coagulation reaction promotes the inflammatory response [59].

Under normal conditions, the activation of coagulation is regulated by three critical physiological anticoagulant pathway systems, including the tissue factor pathway inhibitor system, the activated protein C (APC) system, and the antithrombotic system, which regulated the activation of coagulation [60]. During sepsis, all three pathways exhibit a certain degree of disorder. Due to impaired protein synthesis, the levels of sustained consumption and protein degradation in the three coagulation-inhibitor pathways are low. Thrombomodulin (TM) and endothelial protein C receptor expression is downregulated due to the conversion of protein C to APC under inflammatory conditions [61]. Furthermore, during maximal activation of coagulation, the endogenous fibrinolysis diminishes substantially in sepsis; when plasminogen activator (i.e., tissue plasminogen activator (t-PA)) and urokinase-type plasminogen activator (u-PA) are released from vascular endothelial cell storage sites, plasminogen activator stimulation and sub-quantitative plasmin production increase, whereas a continued increase in plasminogen activator inhibitor-1 (PAI-1) makes this effect disappear [62]. The PAI-1 polymorphism has been shown to increase the risk of septic shock caused by meningococcal infection. Patients featured with the 4G/4G genotype present highly elevated concentrations of PAI-1 and increased mortality related to clinical outcome in Gram-negative sepsis [63].

Neuroendocrine–Immune Network Abnormalities

The homeostasis that depends on the interaction of the neuroendocrine–immune system is also considered to be a crucial part of the host response during septic shock [64,65]. In the case of threats, the central nervous system responds to sepsis through three main mechanisms: (1) the autonomic nervous system, in which primary afferent nerves (vagus and trigeminal nerves) and sensory nerves are associated with PAMPs and lead to inflammatory cytokine activation; (2) circulatory inflammatory mediators, via the choroid plexus and the ventricle organs connected to the central nervous system; and (3) by means of activation of endothelial cells through the blood–brain barrier, causing the release inflammatory mediators (NOS metabolites) [66].

The immune system can be regarded as a “diffuse sensory organ” that signals to the brain through distinct pathways, such as through the vagus nerve and endothelial activation/dysfunction, and this results in the liberation of cytokines and neurotoxic mediators. These afferent signals trigger the efferent response of the central nervous system and thereby activate the autonomic nervous system (including the sympathetic nervous systems and parasympathetic nervous systems), associated with the hypothalamic–pituitary–adrenal (HPA) axis activation, which is the most crucial part. Under normal conditions, the hypothalamic paraventricular nucleus and the supraoptic nucleus release corticotropin (CRH) and arginine vasopressin (AVP). In the pituitary gland, AVP can enhance CRH release. Subsequently, AVP and CRH stimulate the release of adrenocorticotrophic hormone (ACTH), which, in turn, is responsible for the secretion of adrenal cortisol, which counteracts the inflammatory process and restores cardiovascular homeostasis. HPA axis dysfunction during sepsis reduces serum levels of CRH, ACTH, and adrenal cortisol, leading to adrenal insufficiency syndrome [67]. Moreover, a branch of the autonomic nervous system, mainly the sympathetic branch, regulates cytokine production [68]. Evidence indicates that noradrenaline (NA) can respond to LPS; NA is released from the non-synaptic ends of sympathetic axons, is detected in increased concentrations near immune cells, and can inhibit the expression of the pro-inflammatory factors TNF-α and IL-12 and promote expression of the anti-inflammatory cytokine IL-10 [69,70]. These results support the idea that this effect is mediated by β2-adrenergic receptors, which are expressed on immune cells and coupled to cAMP [71].

As shown in Figure 3, the neurotransmitter acetylcholine (ACh) signaling plays an important role in modulating inflammatory responses. The concept of the “cholinergic anti-inflammatory pathway (CAP)” has been proposed in a study that examines how the vagus nerve participates in the regulation of sepsis [72], which opens up a new direction for the treatment of sepsis. The CAP activation can lead to inhibition of the synthesis and release of cytokines, and the differentiation and maturation of T cells followed by a substantial reduction of the killing function of monocytes and neutrophils [73]. CAP is currently considered to exert anti-inflammatory effects mainly through the vagus nerve, Ach, and its specific α7 nicotinic ACh receptor (α7 nAChR), and related intracellular signal transduction pathways. Upon interaction with ACh, α7 nAChR causes a decrease in the level of pro-inflammatory factors, reduces the expression of chemokines and adhesion molecules, alters the differentiation and activation of immune cells, and regulates homeostasis [72]; thus, it acts as an “effector” to exert anti-inflammatory effects in CAP. After the vagus nerve is severed, the susceptibility to endotoxic shock is increased [74,75]. Therefore, systemic inflammatory response can be alleviated by stimulating the efferent vagus nerve or activating α7 nAChR. In several in vitro cellular models, the α7 nAChR also plays a key role in anti-inflammatory effects via the activation of diverse signaling pathways. Therefore, α7 nAChR is considered to be a target for regulating the release of inflammatory cytokines and the anti-inflammatory effects of CAP signaling.

Figure 3. The function of the cholinergic anti-inflammatory pathway (CAP) in sepsis.

Endoplasmic Reticulum Stress

The endoplasmic reticulum (ER) is an intracellular organelle that is involved in protein translocation, folding, posttranslational modification, and further transport to the Golgi apparatus. The unfolded or misfolded proteins are accumulated in the ER during sepsis, altering its homeostasis, and leading to oxidative stress and severe calcium disorders that result in ER stress [76]. Under ER stress, unfolded protein response sensors might switch their signals to stimulate the cell death by unique registration signaling mechanisms, which include several steps: the first is transcriptional activation of the CEBP homologous protein (CHOP) gene, mediated by PKR-like endoplasmic reticulum kinase (PERK), inositol-requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6); the activation of the JNK pathway, mediated by IRE, followed by the subsequent activation of TNF receptor-associated factor 2 and apoptotic signal-regulated kinase 1; finally, the activation of caspase-12 associated and the activated caspase-12 migrates from the ER to the cytosol and then cleaves caspase-9, and finally activates caspase-3 [77–79]. In sepsis animal models, markers of increased ER stress (such as glucose-regulated protein 94 (GRP94), CHOP, and caspase-12) are detected in several organs including the heart and liver, as well as these markers are directly connected with the extent of organ dysfunction, which may be a major cause for sepsis-induced multiple organ failure [80]. ER stress brings about abnormal apoptosis in sepsis animals, suggesting that ER stress-mediated apoptosis represents a potential new target for clinical prevention and treatment for sepsis.

Autophagy

Autophagy refers to the natural process by which a cytoplasmic substance or pathogen is engulfed by the autophagosome, which is then fused with a lysosome to be degraded. Autophagy is a critical defense mechanism used by the host to resist external pathogens and dangerous signals, and plays a critical role in the induction and regulation of natural immune-cell inflammatory response, and is a key factor affecting sepsis development [81]. Autophagy exerts a protective effect in sepsis probably through the following mechanisms: pathogen clearance, the neutralization of microbial toxins, the regulation of cytokine release, the reduction of apoptosis oncotarget, and the promotion of antigen expression [82–84]. In a study that genetic ablates the autophagy protein ATG16L1gene, the endotoxin effects were enhanced, and the autophagy-deficient mice became more susceptible to LPS challenge, since the immune response is inhibited due to T-cell autophagy deficiency [85]. In a mouse model of sepsis developed using cecal ligation and puncture (CLP), incomplete autophagy might lead to cardiac dysfunction in sepsis, and autophagy activation by rapamycin restores cardiac function and reduces CLP-induced myocardial damage [19].