Pattern recognition receptors as therapeutic targets for bacterial, viral and fungal sepsis

Abstract

Sepsis remains to be a significant health care problem associated with high morbidities and mortalities. Recognizing its heterogeneity, it is critical to understand our host immunological responses to develop appropriate therapeutic approaches according to the type of sepsis. Because pattern recognition receptors are largely responsible for the recognition of microbes, we reviewed their role in immunological responses in the setting of bacterial, fungal and viral sepsis. We also considered their therapeutic potentials in sepsis.

Background

Sepsis is a very complex, heterogeneous and pathological syndrome resulting from microbial infection. Sepsis remains to be associated with high morbidities and mortalities but there are no specific therapies available other than supportive management ¹. Supportive management largely consists of antibiotics therapy, fluid resuscitation and respiratory support as needed ². As sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to any infection in the third International Consensus Definitions Task Force (Sepsis-3) ³, responsible pathogens range widely from bacteria, virus to fugal species. Causative microbes are identified in only 60-70% of cases ⁴ and bacteria account for most of the identified organisms, followed by fungus. The reported percentage of viral sepsis is only in a range of 1% ^5,6. Viral panel is not necessarily included in a routine sepsis workup algorithm, however, so that viral sepsis may be under-represented ⁷ The heterogeneity of sepsis has been well recognized and could at large explain the fact that the clinical trials of various anti-cytokine antibodies universally failed. In addition to current supportive treatment, a new therapy is urgently warranted.

Our immune system is mounted with a number of pattern recognition receptors (PRRs) to respond to a wide variety of microbes in the environment. PRRs include Toll-like receptors (TLRs), nucleotide binding oligomerization domain (NOD)-like receptors (NLRs), retinoic acid-inducible gene (RIG)-I-like receptors (RLRs), C-type lectin receptors (CLRs) and DNA-sensing molecules ⁸. Understanding how our host immunity responds to microbial components by using these PRRs is critical to decipher sepsis pathophysiology. With PRRs as new targets for sepsis treatment, we reviewed host immunological responses to bacterial, fungal and viral sepsis for their similarities and differences in respect to PRR responses.

Bacterial sepsis

Bacterial sepsis is the most common type of sepsis. Bacterial species are largely divided into Gram-positive and Gram-negative bacteria. According to the Extended Prevalence of Infection in Intensive Care (EPIC) II study, 62.2% of patients had positive blood cultures harboring Gram-negative bacteria and 46.8% were infected with Gram-positive bacteria ^9,10. A group of patients experienced polymicrobial sepsis with both Gram-positive and Gram-negative bacteria. While Escherichia coli (E. coli) can be found in approximately 1 in 6 culture-positive septic patients, and other predominant Gram-negative bacteria species in sepsis include Pseudomonas, Klebsiella and Enterobacter species. Gram-positive bacterial species such as methicillin-resistant Staphylococcus aureus (MRSA) have made up an increasing percentage of sepsis with the advent of excessive antibiotic treatment.

Immunological responses in bacterial sepsis

The components of bacterial cell walls are main targets that host immune cells use as to recognize bacteria. Among PRRs, TLR and NLR members play a pivotal role in sensing bacteria. The cell walls of Gram-positive bacteria consist mainly of peptidoglycan, glycolipid lipoteichoic acid, and lipoproteins. Peptidoglycan is a unique component of the cell wall of virtually all the bacteria and recognized by TLR2, and NOD-containing protein-1 and -2 (NOD-1, NOD-2). Lipoteichoic acid and lipoproteins are recognized by TLR2 ⁸. In addition to these cell wall components, TLR9 recognizes unmethylated CpG motifs present in bacterial DNA ¹¹. The cell walls of Gram-negative bacteria contain peptidoglycan, lipopolysaccharide (LPS), phospholipids and proteins. Lipid A portion of LPS is characteristic of Gram-negative bacteria and consists of a mono- or biphosphorylated disaccharide backbone acetylated with fatty acids. The level of both phosphorylation and acylation determines the immunostimulatory potency of lipid A/LPS. LPS is recognized by myeloid differentiation factor 2 (MD-2)/TLR4 complex. Flagellin is a major component of flagella, mainly found in Gram negative bacteria. Flagellin is recognized by TLR5.

A number of PRRs used for bacterial recognition are distributed on different cell types and activate overlapping but redundant signaling pathways. TLRs are expressed on the cellular membrane or the endocytic vesicles, with characteristic horseshoe conformation consisting of leucine-rich repeats (LRRs) that serve as ligand binding sites. TLR2 is expressed on myeloid cells, a subset of lymphoid cells, epithelial cells and endothelial cells ^12,13. TLR4 is also expressed on myeloid cells, a subset of lymphoid cells and endothelial cells ¹⁴. TLR5 is expressed on myeloid cells and epithelial cells ^15,16. TLR9 is expressed on myeloid cells and B cells ¹⁷. TLR2, TLR4, TLR5 and TLR9 activate MyD88 pathway, thereby producing nuclear factor kappa B (NFκB) mediated proinflammatory cytokines (Figure 1). TLR4 also activates TRIF pathway, which induces Type I interferon. NOD-1 and NOD-2 are among NLR family and structurally related cytosolic proteins with LRR domains that are similar to those found in TLRs ¹⁸. NOD-1 and NOD-2 are expressed by epithelial cells and antigen-presenting cells ¹⁹ . NOD-1 senses a peptidoglycan-derived peptide D-gamma-Glu-mDAP (iEDAP) from all Gram-negative bacteria and a subset of Gram-positive bacteria. NOD-2 senses muramyldipeptide (MDP), a highly conserved peptidoglycan motif present in almost all bacteria. The binding of ligands to NOD-1 and NOD-2 induces their oligomerization and subsequent activation of NFkB and other effector pathways ²⁰.

Figure 1.

Pattern recognition receptors involved in bacterial, viral and fungal sepsis

Excessive inflammation and immunosuppression are a characteristic, immunological progression pattern in sepsis ²¹. Overall, PRRs interact with a number of bacterial factors and contribute to the immunological responses. However, we need to understand immunological response profiles driven by an individual PRR role to consider the development of PRR targeted therapeutic approaches. The experimental polymicrobial abdominal sepsis model induced by cecal ligation and puncture (CLP) surgery is the most commonly used preclinical model to study bacterial sepsis pathophysiology because it recapitulates human sepsis of warm and cold shock phases ²². One of the advantages in this model is that the severity of CLP sepsis can be easily adjusted by the length of ligated cecum and the number and size of puncture hole in the cecum. However, it is important to note that CLP sepsis with different severity may represent different pathophysiology ²². Thus, the severity of CLP sepsis needs to be paid attention.

TLR2 forms a heterodimer either with TLR1 or TLR6 to manifest its function. TLR2 KO mice showed a better survival in CLP sepsis, associated with the attenuation of CXCR2 downregulation and adequate neutrophil migration to the peritoneal cavity compared to WT mice ²³. The downregulation of CXCR2, a neutrophil chemoattractant receptor for CXCL2 presumably results from TLR2 stimulation. The downregulation of CXCR2 on neutrophils at an environment rich with TLR2 ligands will minimize their migration to other locations. High levels of TLR2 ligands in circulating blood during sepsis, if exists, however, may hinder neutrophil recruitment to the infected tissues/ organs. A chimeric mouse experiment using WT and TLR2 KO mice demonstrated that altered neutrophil functions in TLR2 KO mice were derived from TLR2 KO non-hematopoietic cells ²⁴. However, the mechanism of how the activation of TLR2 expressing non-hematopoietic cells leads to CXCR2 downregulation on neutrophils remains to be clarified. TLR2 also affects cardiac function during sepsis as TLR2 KO mice showed a better ventricular function in CLP sepsis ²⁵. Non-hematopoietic cells also affected TLR2-mediated cardiac dysfunction during sepsis as they did for neutrophils ²⁴.

TLR4 has been a main target of investigation in Gram-negative sepsis ^26,27 The importance of TLR4 in infection is well demonstrated in patients with TLR4 Asp-299-Gly mutation, which depresses TLR4 signaling and is associated with an increased susceptibility to Gram-negative infection ²⁸. Although primary Gram-negative bacterial infection certainly uses TLR4-mediated immunological responses, the release of LPS from gut microbiota as a result of intestinal hypoperfusion caused by Gram-positive bacterial sepsis and fungal sepsis is also indicated, suggesting that TLR4 signals are involved in a wide variety of microbial sepsis ^29,30. Like TLR2, TLR4 is expressed on immune cells as well as non-immune cells. In myeloid cells, TLR4 is required for phagocytosis and bacterial clearance along with cytokine production ³¹. TLR4 in hepatocytes is in charge of LPS clearance. TLR4 in endothelial cells contributes to the sequestration of leukocytes to the tissues ³². Consistent with these TLR4 roles, TLR4 KO mice showed a worse survival in less severe CLP due to poor bacterial clearance ³¹. However, TLR4 KO mice showed a better outcome in severe CLP sepsis presumably due to a significantly less production of proinflammatory cytokines ³³.

Flagellin administration was effective to reduce CLP sepsis mortality by promoting macrophage phagocytosis and reactive oxygen species (ROS) formation via its interaction with TLR5 ³⁴. In fact, neutralizing TLR5 antibody administration significantly attenuated beneficial effect of flagellin administration on CLP sepsis outcome. However, higher dose of flagellin was associated with significant proinflammatory responses and organ injury instead ³⁵.

TLR9 KO showed a better survival with a greater number of peritoneal dendritic cells (DCs) and neutrophils along with less proinflammatory cytokine levels than WT mice ³⁶. TLR9 KO DCs increased neutrophil recruitment presumably by producing neutrophil chemoattractants, which led to survival benefit. Does this mean that TLR9 is not necessary? In pneumococcal infections, TLR9 KO mice were much more susceptible to the infection and showed a worse survival than WT, TLR1 KO, TLR2 KO, TLR4 KO and TLR6 KO mice ³⁷. This was due to impaired uptake of Streptococcus pneumoniae (S. pneumoniae) by TLR9 KO macrophages. Thus, TLR9 is also associated with effector functions of myeloid cells aside from proinflammatory cytokine production, similar to other TLRs.

Overall TLR studies using CLP sepsis indicated that the elimination of the TLR stimulation would inhibit the hyper-inflammatory systemic responses that could lead into death in severe sepsis, thereby posing a survival benefit. However, in less severe sepsis, the presence of TLR signaling has been shown protective ³⁸. TLR2, TLR4, TLR5 and TLR9 share MyD88 signaling pathway. A study by Feng et al. showed that MyD88 KO mice demonstrated a significantly better survival along with lower proinflammatory cytokine levels and improved cardiac function following CLP surgery ³⁸. MyD88 KO mice also showed less recruited neutrophils to the peritoneal cavity with slightly higher mean bacterial loads (not statistically significant) compared to WT mice, suggesting that their outcome advantage was likely derived from the less profound proinflammatory cytokine profiles. In contrast, a study by Sonego et al. using less severe CLP sepsis showed that MyD88 KO mice had worse survival with less neutrophil recruitment to the peritoneal cavity and more bacterial loads than WT mice ³⁹, which is in line with the aforementioned discussion about the importance of maintaining effector functions in sepsis. The study of S. pneumoniae by Albiger et al. also demonstrated an impaired uptake of bacteria by MyD88 KO macrophages ³⁷, demonstrating the involvement of MyD88 in effector functions other than proinflammatory cytokine production.

TLR4 also activates TRIF pathway. In contrast to MyD88 KO mice, TRIF KO mice did not show any difference in the CLP outcome ³⁸. No significant difference in proinflammatory cytokine levels and neutrophil phagocytic activity between WT and TRIF KO mice was observed as well. The absence of NOD-1 or NOD-2 did not affect the CLP sepsis outcome ³⁹. This raises a question whether TLR and NOD signaling are both critical for innate immunity against bacterial infection and what the rank order is regarding the importance of TLRs and NLRs. It was generally presumed that the membrane bound TLRs survey the extracellular bacteria, while the cytosolic NOD-1 and NOD-2 detect the intracellular bacteria ⁴⁰. However, it was shown that S. aureus, for example, could be detected by NOD-2 ⁴¹. Zhou et al. also showed that TLR and NOD signaling are both required to induce a strong inflammatory response and phagosome maturation via crosstalk between them ²⁰, suggesting that both are required. The rank order of TLR an NOD signaling use in sepsis caused by different bacteria species remains to be determined in the future.

Therapeutic consideration

A large number of antibiotics have been developed so far to eradicate bacteria. There is no doubt that early antimicrobial therapy is critical. However, there is no direct therapy against organ injury in sepsis. Attenuating organ injury while maintaining the ability to kill bacteria by host immunity as well as anti-microbial drugs would be ideal.

Among major PRRs in bacterial sepsis, TLR4 attracted a significant attention as a potential target. LPS signal is initiated by its binding to MD-2 followed by the binding of LPS-MD-2 complex to TLR4. Eritoran (E5564) is a synthetic analog of lipid A and inhibits lipid A binding to MD-2, thereby terminating TLR4-mediated signaling (Table 1). In Phase 2 trial, 300 adult patients with severe sepsis were either given placebo or intravenous Eritoran every 12 hours for 6 days (either 45 mg total or 105 mg total) initiated within 12 hours from the first onset of organ injury. The 28-day mortality rate of Eroitan-105 mg group, Eroitan-45 mg group and placebo was 26.6%, 32.0% and 33.3%, respectively. Subjects at highest risk of mortality by APACHE II score showed that Eritoran-105 mg group had mortality rate of 33.3%, while placebo had 56.3% ⁴².

Table 1.

Pattern recognition receptors involved in sepsis, their functions, and known agonists and antagonists

PRRs	Functions	Agonist/antagonist
TLR2	Activates MyD88 pathway, thereby producing nuclear factor kappa B (NFκB) mediated proinflammatory cytokines. Downregulates CXCR2, hindering neutrophil migration to other locations including to the infected tisues/organs during bacterial sepsis. Increases the levels of IL-10, CD4+CD25+ Treg population and MIP-2 resulting to neutrophil recruitment in fungal sepsis. TLR2 is also involved in HSV genomic recognition in viral sepsis	CBLB612 (TLR2 agonist/drug) SV-283 (NY-ESO-1) (TLR2 agonist/adjuvant) IS A-201 (TLR2/1 agonist/adjuvant) OPN-305-110 (TLR2 antagonist/drug) OPN-305 (TLR2 antagonist/drug)
TLR3	Detects extracellular viral RNA that enter the cell from the endocytic pathway. Upon dsRNA recognition, TLR3 induces IFN-β and proinflammatory cytokine production from host cells	Poly(I:C) (TLR3 agonist) Poly(I:C12U) or Ampligen® (TLR3 agonist) Poly-ICLC or Hiltonol (TLR3 agonist) polyA:U (TLR3 agonist) Riboxxol (also known as RGIC®50) (TLR3 Agonist) ARNAX (TLR3 Agonist)
TLR4	Activates MyD88 pathway, thereby producing nuclear factor kappa B (NFκB) mediated proinflammatory cytokines. TLR4 also activates TRIF pathway, which induces Type I interferon. TLR4 is required for phagocytosis and bacterial clearance along with cytokine production in bacterial sepsis. TLR4 in hepatocytes is in charge of LPS clearance. TLR4 in endothelial cells contributes to the sequestration of leukocytes to the tissues. Is associated with increased chemokine (KC, MIP-2) expression and neutrophil recruitment in fungal sepsis.	Monophosphoryl lipid A (TLR4 agonist) Eritoran tetrasodium (E5564) (TLR4 antagonist) TAK-242 (Res atorvid) (TLR4 antagonist) CRX-526 (TLR4 antagonist)
TLR5	Activates MyD88 pathway, thereby producing nuclear factor kappa B (NFκB) mediated proinflammatory cytokines. Promotes macrophage phagocytosis and reactive oxygen species (ROS) formation Ada its interaction with flagelin in CLP sepsis.	Entolimod (CBLB502) (TLR5 agonist)
TLR7	TLR7 detects extracellular viral RNA (ssRNA) on pDCs and mediates protection against lethal viral dose.	Imiquimod or Aldara™ (TLR7 agonist) PF-04878691 or 852A (TLR7 agonist) ANA975 (TLR7 agonist) ANA773 (TLR7 agonist) GS9620 (TLR7 agonist) 2-O-methyl-modified RNA (TLR7 antagonist)
TLR9	Activates MyD88 pathway, thereby producing nuclear factor kappa B (NFκB) mediated proinflammatory cytokines. Suppresses Dcs and neutrophil recruitement and promotes macrophage uptake of pathogens in bacterial sepsis. Is also associated with effector functions of myeloid cells and with viral loads in viral sepsis.	CPG10101 or Actilon™ (TLR9 agonist) IMO-2125 (TLR9 agonist) SD-101 (TLR9 agonist)
NOD-1	NOD-1 senses a peptidoglycan-derived peptide D-gamma-Glu-mDAP (iEDAP) from all Gram-negative bacteria and a subset of Gram-positive bacteria, inducing its oligomerization and subsequent activation of NFkB and other effector pathways. NOD-1 activation leads to proinflammatory mediator production in collaboration with TLRs. NOD-1 is also a key innate immune trigger to drive the development of immunity.	KF1B (NOD-1 agonist) FK565 (NOD-1 Agonist)
NOD-2	NOD-2 is expressed by epithelial cells and antigen-presenting cells and senses muramyldipeptide (MDP), a highly conserved peptidoglycan motif present in almost all bacteria, inducing its oligomerization and subsequent activation of NFkB and other effector pathways.	CL429 (NOD-2 agonist)
Dectin-1	Dectin-1 recognizes β-(1,3)- glucan and is important in fungal sepsis. Dectin-1 signaling induces TLR-dependent and –independent signaling activation leading into phagocytosis, respiratory burst, and the production of various inflammatory mediators including IL-1β, IL-6 and IL-23 [64]. TLR-independent signal includes the activation of integrin Mac-1 (CD11b/CD18) by Dectin-1. The proinflammatory cytokines are important to prime CD4+ T cells to differentiate into Th1/Th17 cells.	β-glucan peptide (Dectin-1 agonist)
Dectin-2	Dectin-2 recognizes α-mannose and is important for fungal sepsis. Dectin-2 forms a heterodimer with Dectin-3, important for Th17 immune responses. Dectin-2 KO mice are more susceptible to Candida albicans with less neutrophil recruitment and macrophage phagocytosis.	Furfurman (Dectin-2 agonist)
RIG-I	RIG-I senses viral replication intermediates in infected cells. RIG-I signaling mediates protection against lethal viral dose.	ppp-RNA (M8. CBS-13-BPS, SLR10. SLR14) (RIG-I agonists) KIN131A (RIG-I agonist)
MDA-5	Senses viral dsRNA	Poly (IC:LC) (MD5 and TLR3 agonist) rb-dsRNA (MD5 and TLR3 agonist) NAB2 (MD5 and TLR3 agonist)

In Phase 3 trial, 1304 patients with severe sepsis received Eritoran-105 mg and 657 patients received placebo. In contrast to the Phase 2 trial, the 28-day mortality of Eritoran-105 mg group and placebo was 28.1% and 26.9%, respectively ⁴³. Rather lower mortality in the placebo group than expected (i.e. less severe sepsis) and low circulating LPS levels in this cohort compared to the patients in Phase 2 trial were pointed out as potential explanation for the different results between Phase 2 and Phase 3 trial. The result from Phase 2 and Phase 3 trial is in line with the rodent studies mentioned above, where TLR4 deficiency offered a survival benefit by attenuating proinflammatory responses in severe sepsis, but worsened the outcome in less severe sepsis by impairing effector cell functions of myeloid cells. TLR4-MyD88 pathway controls proinflammatory cytokine production via NFκB signaling pathway, and effector functions such as phagocytosis via p38 signaling pathway ⁴⁴ Thus, it may be important to block NFκB signaling via TLR4 specifically. This discussion may be applicable for TLR9 as well.

Flagellin administration, if high dose was not used, was beneficial to enhance CLP sepsis survival by augmenting macrophage effector functions. At high dose, however, significant proinflammatory responses caused organ injury. Thus, a case of TLR5 should be similar; avoiding significant activation of NFκB pathway is a key. Thus, approaches to directly target NF NFκB pathway via TLRs would be a potential approach. TRAF6 (TNF receptor associated factor 6)-TAK1(transforming growth factor-β-activated kinase-1)-IKK(IκB kinase)α/β are upstream signaling molecules for NFkB activation via TLRs ⁴⁵. So far targeting these molecules have not been described yet.

In addition to TLR4, there is an increasing interest to target NOD-1 for infectious and non-infectious related multi-organ failure given that NOD-1 activation leads to proinflammatory mediator production in collaboration with TLRs ⁴⁶. Given that NOD-1 is a key innate immune trigger to drive the development of adaptive immunity ⁴⁷, however, a caution should be made to simply target NOD-1 as in the case of TLR4.

Fungal sepsis

Fungi are usually commensal. Although an estimated 5,000,000 fungal species exist in our environment, only 100-300 of them regularly infect humans ^48,49. Fungi exhibit various morphologic states including spherical yeasts, molds with branching hyphae and pseudo-hyphae Since 1980, the prevalence of opportunistic fungal diseases has steadily increased in parallel to an increase in the number of individuals with acquired immune deficiencies or those receiving immune suppressive or myeloablative therapies ⁵⁰. Among fungus, Candida species are by far the predominant cause of fungal sepsis accounting for 10% to 15% of health-care associated infections and 5% of all cases of sepsis ⁵¹. Invasive candidiasis has become the fourth leading cause of bloodstream infection, causing up to 40% mortality ⁵². Therefore, understanding our host immunological responses via PRRs will be important for the development of potential therapeutics.

Immunological responses in fungal sepsis

The cell wall of fungi predominantly comprises rigid polysaccharide layers. An inner layer consists of a chitin layer and an adjacent layer of β-(1,3) and β-(l,6)-glucans. An outer layer consists of N- and O-linked mannoses, β-glucans occupy 50% of the dry weight of the fungal cell wall ⁵³. Innate responses to fungal pathogens are initiated by fungal component recognition via an array of PRRs including C-type lectin receptors (CLRs), TLRs and complement receptors By recognizing the fugal cell wall components, TLRs and CLRs play important roles in the initiation of innate immune response for the immediate control of fungal propagation, and the differentiation of CD4⁺ Th1 and Th17 cells for later control and long-term memory of fungal infection ⁴⁸. TLR4 recognizes O-linked mannose ⁵⁴, while TLR2 primarily recognizes cellular wall glycolipid component phospholipomannan ⁵⁵. The major CLRs involved in antifungal recognition include Dectin-1, Dectin-2 and Dectin-3. Dectin-1 recognizes β-(1,3)- glucan, and Dectin-2 and Dectin-3 recognize α-mannose. Recruited neutrophils to the fungus-infected tissue perform reactive oxygen species (ROS) production and neutrophil extracellular traps (NETs) formation for pathogen clearance ⁵⁶. DCs and monocytes can produce IL-23, which helps to promote T cell differentiation toward Th17. Th17 produces IL-17A and IL-17F, which interact with epithelial cells to produce β-defensins to kill fungus. Antibodies are also produced through fungal infection. However, antibodies that react with fungal antigens can be protective, neutral or disease-enhancing so that it is difficult to decipher their roles ⁵⁷.

Because Candida species are major culprit for fungal infection, we will review PRR responses to Candida albicans here, primarily focusing on TLRs and CLRs. Netea et al. showed that C3H/HeJ mice, possessing a defective TLR4 due to the point mutation ⁵⁸, were significantly more susceptible to Candida albicans compared to C3H/HeN control mice, which possess intact TLR4, associated with impaired chemokine (KC, MIP-2) expression and neutrophil recruitment ⁵⁹. Interestingly, there was no significant change in proinflammatory cytokine levels between C3H/HeJ and C3H/HeN mice. There was also no difference in phagocytic ability between C3H/HeJ and C3H/HeN mice. TLR2 antibody administration did not affect KC and MIP-2 production, but attenuated the production of the proinflammatory cytokines instead. In contrast to the TLR4 experiment, the study by the same group showed that TLR2 KO mice were more resistant to Candida albicans, associated with less IL-10 level and less CD4⁺CD25⁺ Treg population ⁶⁰. Interestingly, TNF-α and IL-1 production did not differ between WT and TLR2 KO mice. Detection of fungal ligands via TLR2 might have been compensated by other PRRs, while immunosuppressive IL-10 signal seemed to derive from TLR2. However, a study by Viliam on et al. showed that TLR2 KO mice were susceptible to Candida albicans infection ⁶¹. In this study, TLR2 KO mice were associated with less MIP-2 production and neutrophil recruitment. A study by Bellocchio et al. showed that fungal morphotypes (yeast vs hyphae) affected TLR2/4 signaling ⁶², which may potentially explain the difference in the results between these studies. MyD88 was found to be critical in fungal defense, as MyD88 KO mice were extremely sensitive to Candida infection ⁶². Specifically, MyD88 KO neutrophils showed impairment in killing ability and MyD88 KO DCs were ineffective to prime antifungal Th1 responses.

CLRs contain at least one C-type lectin-like domain and play a major role in fungal recognition. They are preferentially expressed on myeloid cells ^63,64. Many studies have demonstrated that mutations or deletions in CLRs or downstream molecules lead to enhanced susceptibility to fungal infections, demonstrating their essentiality ⁴⁸. Dectin-1 signaling induces TLR-dependent and –independent signaling activation leading into phagocytosis, respiratory burst, and the production of various inflammatory mediators including IL-1β, IL-6 and IL-23 ⁶⁵. TLR-independent signal includes the activation of integrin Mac-1 (CD11b/CD18) by Dectin-1 ⁶⁶. The proinflammatory cytokines are important to prime CD4⁺ T cells to differentiate into Th1/Th17 cells. Dectin-1 KO mice demonstrated significant susceptibility to Candida albicans, associated with defective activation of tissue macrophages and subsequent innate immune responses ⁶⁷ Dectin-2 forms a heterodimer with Dectin-3, important for Th17 immune responses ⁶⁸. Dectin-2 KO mice were more susceptible to Candida albicans with less neutrophil recruitment and macrophage phagocytosis ⁶⁹. In line with the result of Dectin-2 KO mice, Dectin-3 KO mice were also susceptible to Candida ⁷⁰.

Therapeutic consideration

While in vivo results from TLRs were rather inconsistent, Dectin receptors have received a major interest regarding the development of therapeutics by targeting PRRs for fungal sepsis. For example, Dectin-1 Y298X is associated with susceptibility to Candida infection ^71,72. Dectin-1 and-2 are essential for fungal defense and negatively regulated by E3 ubiquitin ligase Casitas B lymphoma-b (CBLB). In addition, kinase SYK, which is located at the downstream signaling of Dectin receptors, is negatively regulated by CBLB as well. Following fungal recognition, CBLB ubiquitinates SYK, Dectin-1 and Dectin-2 ⁷³. CBLB KO mice showed resistance to Candida infection with enhanced inflammasome activation, ROS production and increased fungal killing ^73,74. Furthermore, the administration of cblb siRNA or a peptide inhibiting CBLB activity improved the outcome of Candida infection in mice, indicating that CBLB is a promising target ^73,74. Other approaches to optimize Dectin receptor function should be also considered in the future.

Viral sepsis

Viral sepsis is likely under-represented among all sepsis. For example, a third of adult patients requiring ICU management for severe pneumonia had viral infections by PCR analysis ⁷⁵. Patients with culture-negative sepsis had significantly lower levels of procalcitonin than those with documented sepsis ⁷⁶. Furthermore, recent COVID-19 outbreak clearly demonstrated the seriousness of viral sepsis ^77,78. Viruses can be divided into single-stranded RNA (ssRNA), double-stranded RNA (dsRNA), or DNA virus ⁷⁹. Over 200 viruses are known to infect humans. Herpes simplex virus (HSV) and enterovirus are the most common viral causes of neonatal sepsis ⁸⁰, while enterovirus is the most common cause of viral sepsis in young children. Influenza virus also causes a major morbidity and mortality in various age groups ⁴ HSV is DNA virus, and enterovirus and influenza virus are ssRNA virus.

Immunological responses in viral sepsis

TLR3, TLR7 and TLR9 detect extracellular viral RNAs and DNAs that enter the cell from the endocytic pathway. TLR3, TLR7, or TLR9 serves as a receptor for dsRNA, ssRNA, or DNA, respectively. TLR3 is expressed on macrophages and DCs along with epithelial cells ⁸¹. TLR7 is expressed on DCs ⁸². TLR9 is expressed on macrophages, DCs and B cells as above ¹⁷. In contrast, viral RNAs produced within a cell are sensed by a separate family of proteins called the RIG-I-like receptors including ssRNA sensor RIG-I and dsRNA sensor MDA-5 (melanoma differentiation-associated 5). While eukaryotic RNA undergoes 5’-capping in the nucleus, viral RNA is produced in the cytosol without this capping. Viral DNAs also activate cGAS (cyclic guanosine monophosphate-adenosine monophosphate synthase), then subsequently STING (stimulator of interferon genes). Eukaryotic DNA exists only in the nucleus, while viral DNA is in the cytosol. RIG-I-like receptors and cGAS exist in many cell types. An interaction of viral RNA and DNA with these PRRs will induce type I interferons (IFNs) that play a central role in driving an antiviral state. Type I IFNs consist of IFN-α, IFN-β, IFN-κ, IFN-ε and IFN-ω. IFN-α and IFN-β have been studied most. Plasmocytoid dendritic cells (pDCs) are the most efficient IFNs producing cells whose TLRs can recognize virus and produce IFNs up to 1,000 times more than other cells. Type I IFNs show anti-viral effects by a number of mechanisms; 1) Induction of endoribonuclease to degrade viral RNA, 2) induction of Mx (myxoma resistant) proteins to interfere with viral replication, 3) Increase of MHC I expression and antigen presentation by host cells, 4) Activation of NK cells and antigen presenting cells, and 5) Induction of T cell chemoattractants. Cytotoxic T cells can directly kill virus-infected cells. Helper T cells can help B cells to make antibodies against viruses.

The use of TLR9 in HSV genomic recognition is well characterized ⁸³. Furthermore, TLR2 is also involved in HSV genomic recognition ⁸⁴. Either TLR2 mice or TLR9 KO mice had viral loads comparable to WT mice, while TLR2/TLR9 double KO mice had significantly higher viral loads. Most cases of human Influenza are caused by influenza A and B viruses. Influenza A virus, for example, is recognized by TLR7 in the endosome and RIG-I in the cytosol ⁸⁵. RIG-I replies on sensing of viral replication intermediates in infected cells such as airway epithelial cells, while detection of ssRNA by TLR7 on pDCs occurs without viral replication. Both TLR7 and RIG-I signaling mediates protection against lethal viral dose ⁸⁵.

Therapeutic consideration

Current antiviral interventions focus on the use of direct-acting antivirals, which target the essential components in the life cycle of a virus ⁸⁶. So far animal models demonstrated that adequate TLR and RLR function would be critical in the setting of viral infection. For example, a RIG-I agonist has been shown to work as vaccine adjuvants and antiviral therapy ⁸⁶. TLR agonists also show protective effect against viral infection ⁸⁷. However, it remains to be determined if rodent viral infection models recapitulate human viral sepsis, which is an important consideration when we target PRRs for viral sepsis.

Conclusion

Here we reviewed our host responses associated with PRRs in the setting of bacterial, fungal and viral sepsis. PRRs and downstream signaling pathways can offer overlapping, but some different roles depending on the severity of infection and the type of microbes. Furthermore, the majority of the studies to determine the role of PRRs in sepsis so far have been done using preclinical models. Understanding human immunological profiles in the setting of various sepsis will be critical to translate the findings in the preclinical studies to the development of therapeutic interventions.

Highlight.

• Sepsis is a heterogeneous disease with an array of immunological responses.

• Different combination of pattern recognition receptors is activated via diverse pathogens.

• Pattern recognition receptors can be potential targets for sepsis.