Lipopolysaccharide

Much of our understanding of the innate immune system and the pathophysiology of inflammation has come from experimental studies with a compound called lipopolysaccharide (LPS) or endotoxin, which is a proinflammatory component of the cell wall of gram-negative bacteria. When experimental animals are injected with purified LPS, they manifest clinical and biochemical findings reminiscent of those observed in patients with severe sepsis or septic shock. Depending on myriad factors (e.g., the animal species being studied, the dose of LPS, its route of administration), the features of acute endotoxemia can include fever (or hypothermia), systemic arterial hypotension, leukocytosis or leukopenia, renal dysfunction, pulmonary dysfunction, hepatocellular damage, and metabolic acidosis.

LPS is a complex glycolipid composed of a polysaccharide tail attached to a lipophilic domain called lipid A. The polysaccharide portion of the molecule tends to be structurally different in different species and strains of gram-negative bacteria, whereas the structure of lipid A (as well as a few neighboring sugar residues) is highly conserved across different species and strains of gram-negative microorganisms. A complex of LPS and a serum protein, LPS-binding protein (LBP), initiates the activation of monocytes and macrophages by binding to a surface protein, CD14. Because it is a glycophosphatidylinositol-anchored membrane protein, CD14 lacks a cytosolic domain and is unable to initiate intracellular signaling directly. Accordingly, investigators sought to identify another protein that presumably participates with CD14 to initiate the cellular response to LPS. The putative LPS coreceptor was ultimately identified as a Toll-like receptor (TLR).⁴

Toll-Like Receptors

TLR4, as well as other members of the TLR family of PRRs, is a homologue of a protein, Toll, which plays roles in embryogenesis as well as antifungal immunity in fruit flies. TLR4 was originally identified by studying an inbred strain of mice, C3H/HeJ, that is congenitally hyporesponsive to endotoxin. Subsequently, TLR4 knockout mice were generated and shown to be as hyporesponsive to LPS as C3H/HeJ mice, thus confirming the concept that expression of functional TLR4 is necessary for the activation of macrophages and monocytes by endotoxin. TLR4 mutations are also associated with endotoxin hyporesponsiveness in humans. MD-2, another protein associated with the extracellular domain of TLR4, is required for LPS responsiveness.

In addition to LPS, other PAMPs and alarmins are recognized by various TLRs (Table 4-1). For example, TLR2 recognizes various bacterial lipoproteins, as well as peptidoglycan derived from gram-positive bacteria. TLR5 recognizes flagellin, a 55-kDa protein found in the flagella of certain bacteria. TLR9 recognizes certain oligonucleotides containing unmethylated CpG motifs that are more common in bacterial DNA than in mammalian DNA.

Table 4-1 Recognition of Pathogen-Associated Molecular Patterns and Alarmins by Pattern Recognition Receptors

Among the TLRs, TLR4 seems to be particularly important, because this receptor recognizes not only the PAMP, LPS, but several endogenous danger signals as well. These endogenous ligands for TLR4 include the following: heat shock protein (HSP) 70, an inducible cytosolic protein, which is important for the proper folding of nascent proteins; high-mobility group box-1(HMGB1), an abundant DNA-binding protein, which is important for transcription and repair of DNA; extra domain A of fibronectin, an abundant protein in the extracellular matrix; and fragments of hyaluronan, a glycosaminoglycan, which is one of the chief components of the extracellular matrix. Some of these alarmins, such as HMGB1, are actively secreted by immunostimulated macrophages or enterocytes, whereas others, such as hyaluronan fragments, are probably generated as a consequence of trauma to tissues. Accumulating evidence obtained by the Billiar group at the University of Pittsburgh has suggested that many of the deleterious host responses to severe trauma and/or hemorrhagic shock are mediated by the interaction of endogenous alarmins with TLR4.⁴

TLRs are glycoproteins. Their structure includes a ligand-binding domain, containing leucine-rich repeat (LRR) motifs, and a signaling domain, which is homologous to the signaling domain for the receptor for the cytokine IL-1 (see later). To date, 10 TLRs have been identified in humans, and these receptors can be divided into subfamilies based on the ligands they recognize. The receptors TLR3, TLR7, TLR8, and TLR9, are located intracellularly on membrane-bound endosomes, whereas the remaining members of the TLR family of receptors are situated so that they span the cytosolic membrane on the surface of cells.

Other Families of Pattern Recognition Receptors

In addition to members of the TLR family, there are two other families of PRRs that are important for recognizing DAMPs and initiating innate immune responses. These two families are the retinoid acid-inducible gene I (RIG-I)–like receptors (RLRs) and the nucleotide-binding oligomerization domain (NOD)–like receptors (NLRs).⁵ The two RIG-I–like receptors, RIG-I and melanoma differentiation-associated gene (MDA) 5, play a pivotal role in sensing the presence of viral double-stranded (ds) RNA in the cytoplasm. The interaction of ds-RNA with the C-terminal domains of RLRs initiates a signaling cascade, leading ultimately to the expression of cytokines important in antiviral immunity.

The two most extensively studied members of the NLR family of receptors are NOD1 and NOD2.⁵ These PRRs sense PAMPs derived from the synthesis and degradation of bacterial peptidoglycan. NOD1 is activated by diaminopimelic acid produced by gram-negative bacteria, whereas NOD2 is activated by muramyl dipeptide (MDP), produced by gram-negative and gram-positive bacteria. As will be discussed in greater detail, NLRs are not only important for sensing certain intracellular pathogens, but these receptors also play a key role in the processing for secretion of two important proinflammatory cytokines, IL-1β and IL-18.

The receptor for advanced glycation end products (RAGE) is a receptor that has multiple potential ligands, including HMGB1, amyloid-β peptide, and certain members of the S100-calgranulin family of proteins.⁶ Because RAGE-dependent signaling may be important for transducing some of the proinflammatory effects the alarmin, HMGB1, RAGE can be considered a PRR involved in innate immunity.

High-Mobility Group Box 1

When mice are injected with a lethal bolus dose of LPS, circulating levels of TNF peak approximately 60 to 90 minutes later and are almost undetectable within 4 hours. Although mice show clinical signs of endotoxemia (e.g., decreased activity and ruffled fur) within a few hours after the injection of LPS, mortality typically does not occur until more than 24 hours later, long after circulating levels of the so-called alarm phase cytokines, TNF and IL-1β, have returned to normal. These observations suggested the possibility to Wang and colleagues that LPS-induced lethality might be mediated by a previously unidentified factor that is released much later than TNF or IL-1β.^6a Prompted by this idea, these investigators carried out a prolonged search for the putative late-acting mediator. This research program ultimately resulted in the identification of HMGB1 (formerly called HMG-1) as a novel mediator of LPS-induced lethality.

HMGB1 was originally identified in 1973 as a nonhistone nuclear protein with high electrophoretic mobility. A characteristic feature of the protein is the presence of two folded DNA-binding motifs termed the A domain and the B domain. Both these domains contain a characteristic grouping of aromatic and basic amino acids within a block of 75 residues termed the HMG box. HMGB1 has several functions within the nucleus, including facilitation of DNA repair and support of the transcriptional regulation of genes. When released by cells into the extracellular milieu, HMGB1 can interact with several different receptors, including TLR2, TLR4, and RAGE, on macrophages, endothelial cells, and enterocytes.⁷ Activation of these receptors leads to the release of other proinflammatory mediators, such as TNF and NO·.

Although HMGB1 is normally not secreted by cells and levels of this protein are usually undetectable in plasma or serum, high circulating concentrations of HMGB1 can be detected in mice 16 to 32 hours after the onset of endotoxemia. Immunostimulated macrophages and enterocytes actively secrete HMGB1. Moreover, necrotic but not apoptotic cells release nuclear HMGB1. In this way, unexpected cell death, such as that secondary to trauma or infection, can act as a danger signal and lead to the induction of an inflammatory response.

Delayed passive immunization of mice with antibodies against HMGB1 confers significant protection against LPS-induced mortality. Furthermore, the administration of highly purified recombinant HMGB1 to mice is lethal. Thus, HMGB1 fulfills a modified version of Koch’s criteria for being a mediator of LPS-induced lethality in mice. Direct application of HMGB1 into the airways of mice initiates an acute inflammatory response and lung injury that is reminiscent of ARDS in humans. In addition, HMGB1 (or a truncated form of the protein, including only the B box domain) increases the permeability of human enterocyte-like monolayers in culture and promotes intestinal barrier dysfunction when injected into mice.⁸ Thus, it seems plausible that HMGB1 contributes to the development of organ dysfunction in human sepsis, a notion that is supported by the observation that circulating HMGB1 concentrations are significantly higher in patients with ultimately fatal sepsis than in patients with a less severe form of the syndrome.⁹ Circulating levels of HMGB1 are also increased in victims of trauma¹⁰ or burn injury.¹¹ Administration of a neutralizing anti-HMGB1 antibody improves survival in mice subjected to lethal hemorrhagic shock.¹² Ethyl pyruvate, a compound that blocks the release of HMGB1 from LPS-stimulated murine macrophage-like cells and inhibits release of the mediator in vivo, improves survival in mice with bacterial peritonitis, even when treatment with the compound is delayed for 24 hours after the onset of infection.¹³

Heat Shock Proteins

The heat shock proteins were first identified as a family of molecules that are induced when cells or experimental animals are subjected to sublethal thermal stress. These proteins are also induced by many other stimuli, such as inflammation, oxidative stress, and infection. The primary role of HSPs is to serve as molecular chaperones to facilitate the proper folding of nascent proteins.

Like HMGB1, heat shock proteins are normally found inside cells but, under certain conditions, these proteins can be detected in the extracellular milieu. For example, elevated circulating levels of HSP70 have been found in trauma patients and patients in the immediate period after coronary artery bypass graft surgery. In addition, immunostimulated monocytes appear to be capable of actively secreting HSP70. Extracellular HSP70 (and the related protein, HSP60) can activate innate immune cells via a TLR4-dependent mechanism. Thus, like HMGB1, these proteins may serve as endogenous danger signals and trigger activation of the inflammatory response after damage to tissues.

Interferon-γ and Granulocyte-Macrophage Colony-Stimulating Factor

The interferons, named for their ability to interfere with viral infection, were initially discovered in the 1950s as soluble factors secreted by leukocytes. The type 1 interferons, IFN-α and IFN-β, are primarily involved as mediators of innate (and acquired) immune responses to viral infection. IFN-γ, although also important in the immune response to viral infection, has much broader activity as a proinflammatory mediator.

For the most part, IFN-γ is produced by three types of cells—CD4⁺ Th1 cells, CD8⁺ Th1 cells, and natural killer (NK) cells. IFN-γ, along with IL-12, plays a critical role in promoting the differentiation of CD4⁺ T cells into the Th1 phenotype. Because Th1 cells also produce IFN-γ, the potential exists for a positive feedback loop. IL-12, produced by monocytes and macrophages, stimulates the production of IFN-γ by Th1 and NK cells. In turn, IFN-γ further activates monocytes and macrophages, thereby creating another positive feedback loop.

In addition to promoting the differentiation of uncommitted CD4⁺ T cells into Th1 cells, IFN-γ inhibits the differentiation of lymphocytes into cells with the Th2 phenotype. Because Th2 cells secrete the counterregulatory cytokines IL-4 and IL-10, the effect of IFN-γ to downregulate the production of these cytokines by Th2 cells further promotes the development of an inflammatory response to an invading pathogen. In target cells, such as macrophages or enterocytes, IFN-γ induces the expression or activation of a number of key proteins involved in the innate immune response to microbes. Among these proteins are other cytokines, such as TNF and IL-1, and enzymes, such as iNOS and the reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex. Thus, IFN-γ stimulates the release of a number of other proinflammatory mediators, including cytokines, such as TNF, and small molecules, such as superoxide radical anion (O₂⁻·), an oxidant produced by NADPH oxidase, and NO·, produced by iNOS. Secretion of these inflammatory mediators by activated macrophages and other cell types is inhibited by IL-4 and IL-10. Accordingly, IFN-γ–mediated downregulation of the Th2 phenotype—and thereby production of IL-4 and IL-10—further promotes the development of an inflammatory response.

The crucial role of IFN-γ in the host’s innate immune response to microbial invasion, particularly by intracellular pathogens, has been emphasized by experiments using transgenic mice with targeted disruption of the genes coding for IFN-γ or the ligand-binding subunit of the IFN-γ receptor (IFN-γR). These knockout mice manifest increased susceptibility to infections caused by Listeria monocytogenes, Mycobacterium tuberculosis, or bacille Calmette-Guérin.

When responsive target cells are exposed to IFN-γ, a number of genes are activated within minutes and without the synthesis of new copies of intermediate signaling proteins. IFN-γ–induced signal transduction occurs through the activation of a protein tyrosine phosphorylation cascade known as the JAK-STAT pathway (Fig. 4-2). JAK initially stood for “just another kinase” because the biologic role of these proteins was not established when they were initially discovered. Because these receptor-associated kinases look both outside and inside the cell, JAK has now come to stand for Janus kinases, after the two-faced Roman god. The moniker STAT, an acronym for signal transducers and activators of transcription, was appropriately chosen because, in medical parlance, an action to be carried out immediately is a stat order and signaling involving these proteins similarly occurs without delay. In addition to IFN-γ, a large number of other cytokines, including IL-6 and IL-11 (see later), also use versions of the JAK-STAT signaling mechanism. In mammals, there are seven mammalian STAT proteins (STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6) and four JAK proteins (JAK1, JAK2, JAK3, and TYK2).

FIGURE 4-2 Simplified representation of intracellular signaling mediated by binding of IFN-γ to its receptor (IFN-γR). A, IFN-γR is a dimer that consists of a ligand-binding α chain and a transmembrane signaling β chain. B, Binding of IFN-γ leads to dimerization of IFN-γR and brings two signaling proteins, JAK1 and JAK2, into association with the receptor complex. C, The association of JAK1 and JAK2 with the receptor leads to mutual tyrosine phosphorylation of these proteins, as well as phosphorylation of tyrosine residues on the ligand-binding chains of IFN-γR and docking of two copies of the preformed transcription factor STAT1α to the receptor complex. After tyrosine phosphorylation, STAT1α forms a homodimer. The homodimer dissociates from the receptor complex and translocates to the nucleus, where binding to the promoter regions of various IFN-γ–responsive genes leads to transcriptional activation.

IFN-γR is a heterodimer that consists of a 90-kDa glycoprotein, the α chain, which is required for binding of the ligand, and a transmembrane protein, the β chain, which is required for signaling. Associated with the receptor are two members of the JAK family of kinases, JAK1 and JAK2. Interaction of IFN-γ with its receptor results in the dimerization of IFN-γR, which brings JAK1 and JAK2 into close association and leads to mutual phosphorylation and activation (see Fig. 4-2). The activated JAK kinases then catalyze the phosphorylation of tyrosine residues on the α chains of IFN-γR, which results in docking to the receptor complex by the transcription factor STAT1. After tyrosine phosphorylation, two copies of STAT1 form a homodimer (IFN-γ activation factor [GAF]) that subsequently dissociates from the receptor complex and translocates to the nucleus, where binding to the regulatory regions of target genes containing the IFN-γ activation site (GAS) nucleotide sequence leads to transcriptional activation.

JAK-STAT–dependent signaling is regulated in cells by a variety of mechanisms. Because STATs are activated by tyrosine phosphorylation, phosphotyrosine phosphatases are implicated in the negative regulation of JAK-STAT signaling pathways. In this regard, the first to be described were Src homology 2 domain (SH2)–containing tyrosine phosphatases such as SHP1 and SHP2. The presence of a characteristic amino acid sequence, the SH2 domain, in these cytoplasmic enzymes promotes the association of these phosphatases with phosphotyrosines present on activated receptors or on signaling molecules, as well as on activated JAKs.¹⁴ The transmembrane tyrosine phosphatase CD45, which is expressed on T and B cells, also downregulates JAK-STAT signaling. Two other important classes of proteins that regulate JAK-STAT signaling are the protein inhibitors of activated STAT (PIAS) and the inducible suppressors of cytokine signaling (SOCS).

The pivotal role played by IFN-γ in the regulation and expression of innate immunity to microbial pathogens led investigators to use this cytokine as a therapeutic agent to increase host resistance to infection, particularly for patients with congenital or acquired immunosuppression. For example, prophylactic treatment with recombinant IFN-γ has been shown to reduce the frequency of infections markedly in patients with chronic granulomatous disease, a life-threatening condition caused by an inherited defect in NADPH oxidase, the enzyme complex responsible for generating ROS in phagocytes. IFN-γ has been approved for this indication by the U.S. Food and Drug Administration (FDA). Severe trauma and burns are associated with defects in host antibacterial and antifungal defense and, in animal models of these conditions, treatment with IFN-γ has been found to increase resistance to infection. Three major clinical trials of prophylactic IFN-γ treatment were conducted in patients with multiple trauma or major thermal injury. Unfortunately, in all three studies, the incidence of infection and mortality was similar in cytokine- and placebo-treated patients.

It is unclear why treatment with IFN-γ failed to improve outcomes in these trials. However, treatment with IFN-γ was not individualized according to immunologic phenotype, and thus some of the deleterious effects of inflammation might have been fostered in certain subjects by administration of this potent proinflammatory cytokine. This concept is supported by results from an uncontrolled trial in which patients with sepsis and laboratory findings indicative of excessive immunosuppression (downregulation of human leukocyte antigen [HLA]-DR expression on circulating monocytes) were treated with IFN-γ. In this small study, administration of IFN-γ resulted in the resolution of sepsis in eight of nine patients. A small pilot study evaluated the use of prophylactic perioperative IFN-γ therapy to decrease the risk for infection in anergic high-risk patients undergoing major operations. Results from this study were inconclusive.

Another approach may be to substitute granulocyte-macrophage colony-stimulating factor (GM-CSF) for IFN-γ. GM-CSF is a hematopoietic growth factor and proinflammatory cytokine produced by multiple cell types, including bronchial epithelial cells, monocytes, and endothelial cells. As a growth factor, GM-CSF promotes an increase in the number of circulating polymorphonuclear nuclear cells (PMNs). However, in addition, GM-CSF has a number of IFN-γ–like features, including the use of JAK-STAT signaling pathways. In both in vitro and in vivo studies, treatment with GM-CSF primes monocytes to produce more proinflammatory cytokines, such as TNF, in response to LPS.

A randomized trial of adjuvant treatment with recombinant GM-CSF in neonates with sepsis and neutropenia has shown that survival is significantly improved in the group treated with the cytokine–growth factor.¹⁵ Similarly, in a single-center randomized controlled trial (RCT), adjuvant treatment with recombinant GM-CSF significantly shortened hospital stay and decreased the number of infectious complications in patients with intra-abdominal sepsis.¹⁶ A more recent multicentric RCT has suggested that adjuvant treatment with GM-CSF can improve outcome for selected patients with sepsis.¹⁷ This study randomized 38 patients with severe sepsis and evidence of sepsis-induced immunosuppression to treatment with GM-CSF or placebo for 8 days. Although survival was similar in both groups, the GM-CSF–treated patients required mechanical ventilation and care in an ICU for a significantly shorter period of time

Crohn’s disease is a chronic inflammatory disorder of the gastrointestinal (GI) tract. Treatment with corticosteroids often ameliorates symptoms of the disease, but chronic administration of corticosteroids is associated with many adverse side effects. Accordingly, clinicians and scientists are actively seeking better approaches to treat Crohn’s disease. Because there is considerable evidence that Crohn’s disease may result, at least in part, from impaired innate immunity (e.g., caused by a mutation in the NOD2 gene),¹⁸ recombinant GM-CSF might be a therapeutic option for this condition. This hypothesis has been supported by the results from two RCTs, which showed that therapy with GM-CSF can induce remission in the absence of treatment with corticosteroids.^19,20

Interleukin-1 and Tumor Necrosis Factor

IL-1 and TNF are structurally dissimilar pluripotent cytokines. Although these compounds bind to different cellular receptors, their multiple biologic activities overlap considerably. For example, in vitro, both cytokines are capable of activating endothelial cells, leading to increased expression of cell surface adhesion molecules, such as intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1), which play important roles in the process whereby neutrophils extravasate from circulation into tissues at the site of infection and/or inflammation. Similarly, incubating cultured monocytes, neutrophils, endothelial cells, hepatocytes, mesangial cells, articular chondrocytes, or synovial fibroblasts with IL-1 or TNF leads to secretion of a chemokine, IL-8 (see later), which is important for recruiting neutrophils into inflammatory foci. Recombinant forms of IL-1β and TNF have been available for many years. Table 4-3 summarizes some of the biologic effects, which are observed when human subjects are injected with recombinant IL-1β or TNF. The information in this table should convince the reader that many of the features associated with the systemic inflammatory response syndrome (SIRS), such as increased circulating leukocyte count and fever, can be reproduced by injecting subjects with the alarm phase cytokines, IL-1β or TNF. Through their ability to potentiate the activation of helper T cells, IL-1 and TNF can promote almost all types of humoral and cellular immune responses. Furthermore, both these cytokines are capable of activating neutrophils and macrophages and inducing the expression of many other cytokines and inflammatory mediators. Many of the biologic effects of IL-1 or TNF are greatly potentiated by the presence of the other cytokine.

Table 4-3 Partial List of Physiologic Effects Induced by Infusing Interleukin-1 or Tumor Necrosis Factor Into Human Subjects

Interleukin-1 and the Interleukin-1 Receptor

IL-1 was first described as a lymphocyte-activating factor produced by stimulated macrophages. IL-1 is not a single compound, but rather a family of three distinct proteins, IL-1α, IL-1β, and IL-1 receptor antagonist (IL-1RA), which are products of different genes located close to one another on the long arm of human chromosome 2. The genes for the two receptors for IL-1, IL-1RI and IL-1RII, are also located on chromosome 2. IL-1α and IL-1β are peptides composed of 159 and 153 amino acids, respectively. Although IL-1α and IL-1β are structurally distinct—only 26% of their amino acid sequences are homologous—the two compounds are almost identical from a functional standpoint. IL-1RA, the third member of the IL-1 family of proteins, is biologically inactive but competes with IL-1α and IL-1β for binding to IL-1 receptors on cells and thereby functions as a competitive inhibitor to limit IL-1–mediated effects.

IL-1 is synthesized by a wide variety of cell types, including monocytes, macrophages, B lymphocytes, T lymphocytes, NK cells, keratinocytes, dendritic cells, fibroblasts, neutrophils, endothelial cells, and enterocytes. Compounds that can trigger the production of IL-1 by monocytes, macrophages, or other cell types include PAMPs such as LPS (from gram-negative bacteria), lipoteichoic acid (from gram-positive bacteria), and zymosan (from yeast). Production of IL-1 can also be stimulated by other cytokines, including TNF, GM-CSF, and IL-1 itself.

Although many cell types express genes for both IL-1α and IL-1β, most cells produce predominantly one form of the cytokine. For example, human monocytes produce mostly IL-1β, whereas keratinocytes produce predominantly IL-1α. The two forms of IL-1 are both initially synthesized as 31-kDa precursors (pro–IL-1α and pro–IL-1β), which are then modified post-translationally to create the carboxyl terminal 17-kDa peptide forms of the mature cytokines. IL-1α is stored in the cytoplasm as pro–IL-1α or, after being phosphorylated or myristoylated, in a membrane-bound form. Whereas both pro–IL-1α and membrane-bound IL-1α are biologically active, pro–IL-1β is devoid of biologic activity. Pro–IL-1α is converted to the mature peptide by calpain and other nonspecific extracellular proteases. Pro–IL-1β is cleaved to its mature active form by a specific intracellular cysteine protease called IL-1β converting enzyme (ICE) or caspase-1. Like IL-1β, ICE–caspase-1 is stored in cells in an inactive form and must be proteolytically cleaved to become enzymatically active.

Transgenic mice deficient in ICE–caspase-1 are resistant to endotoxic shock and manifest an impaired ability to mount a local inflammatory response to intraperitoneal zymosan, a known inducer of sterile peritonitis. In contrast, ICE–caspase-1 knockout mice manifest increased susceptibility to infections caused by various pathogens, including E. coli, Shigella flexneri, Salmonella typhimurium, Listeria monocytogenes, and Candida albicans. Taken together, these data suggest that ICE-dependent processes, including secretion of the mature forms of IL-1β and the related cytokine, IL-18 (see later), are important for host defense against microbial infection but also are crucial for the pathologic manifestations of poorly controlled inflammation.²¹ Various ICE-like enzymes, the caspases, have been identified as being important mediators of the process of programmed cell death, or apoptosis. A special form of apoptosis, called pyroptosis, can occur within minutes after macrophages are infected with certain intracellular pathogens. Pyroptosis is an ICE-dependent process.

The activation of ICE–caspase-1 can be triggered in cells by the formation of a molecular complex called the inflammasome.²¹ Inflammasomes are oligomeric complexes, which are composed of ICE–caspase-1 as well as various members of the NLR family of PRRs called NALPs (NACHT domain leucine-rich repeat and PYD-containing protein) and an adapter protein called ASC (apoptosis-associated specklike protein containing a CARD). Assembly of the inflammasome, which in many cases is triggered when an NLR family member senses the presence of PAMP molecules and ultimately leads to ICE–caspase-1 activation and secretion of IL-1β (and IL-18). Inflammasomes that contain a particular NALP (NALP3), can activate ICE–caspase-1 in response to a wide variety of unrelated compounds, including certain toxins, high concentrations of adenosine triphosphate (ATP), and crystals of monosodium urate (the mineral-like structures that are associated with gout). Alum, the adjuvant used in most vaccines to enhance immune responses to antigens, also has been shown to induce activation of the NALP3 inflammasome. All these compounds can lead to ICE–caspase-1 activation and secretion of IL-1β and the related cytokines, IL-18 and IL-33.

The mature 17-kDa form of IL-1β lacks a secretory signal peptide and is not secreted via the classic exocytic pathway used for the secretion of most proteins (including most other cytokines) from cells. ICE-dependent processing of pro–IL-1β and the secretory step appear to occur at the same time. Secretion of the leaderless mature peptide apparently occurs through the action of a specific transporter called ABC1, which can be inhibited by the oral hypoglycemic agent glyburide.

Similar to the other members of the IL-1 family, IL-1RA can be produced by a variety of cell types. However, unlike IL-1α and IL-1β, IL-1RA is synthesized with a leader peptide that allows normal secretion of the protein. A specialized form of IL-1RA, intracellular IL-1RA, is synthesized without a leader peptide sequence and therefore accumulates intracellularly in certain cell types. In some tissues, such as intestinal epithelium, the formation of intracellular IL-1RA may serve a counterregulatory function to limit inflammation and thereby confer mucosal protection. Moreover, an imbalance between the production of IL-1 and IL-1RA may promote the development of chronic inflammation in certain pathologic conditions, such as Crohn’s disease. Cellular production of IL-1 and IL-1RA is differentially regulated. Certain cytokines, notably IL-4, IL-10, and IL-13, serve as anti-inflammatory mediators, in part by promoting the synthesis of IL-1RA. IL-6, although not usually considered an anti-inflammatory cytokine, is also capable of triggering the production of IL-1RA.

The importance of IL-1β as a proinflammatory cytokine and IL-1RA as an anti-inflammatory cytokine is emphasized by experiments using transgenic mouse strains deficient in IL-1RA, IL-1α, IL-1β, or both IL-1α and IL-1β (double knockout mice). In these studies, IL-1α knockout mice were able to mount a normal inflammatory response, whereas the IL-1β knockout animals manifested an impaired ability to mount a normal inflammatory response. In contrast, mice functionally deficient in IL-1RA manifested an exaggerated response to a systemic proinflammatory stimulus (intraperitoneal injection of turpentine).

There are two distinct IL-1 receptors, IL-1RI and IL-1RII. IL-1RI is an 80-kDa transmembrane protein with a long cytoplasmic tail. In contrast, IL-1RII, a 60-kDa protein, has only a very short cytoplasmic tail and is incapable of initiating intracellular signaling. As a consequence, IL-1RII is actually a decoy receptor that serves a counterregulatory role by competing with IL-1RI, the fully functional IL-1 receptor, for IL-1 in the extracellular space. IL-1RI is present on a wide variety of cell types, including T cells, endothelial cells, hepatocytes, and fibroblasts. IL-1RII is the predominant IL-1 receptor found on B cells, monocytes, and neutrophils. The extracellular domains of IL-1RI and IL-1RII are shed by activated neutrophils and monocytes. The shed receptors can act as a sink for secreted IL-1 and, thus, along with IL-1RA, represent an important counterregulatory component of the inflammatory response.

IL-1RI is a member of the IL-1R–TLR superfamily of receptors. The cytoplasmic portions of all the members of this superfamily of transmembrane proteins are homologous and are called Toll IL-1 receptor (TIR) domains. In contrast, the extracellular domains fall into two main subdivisions. In one subdivision, the extracellular portion of the molecule contains three immunoglobulin-like regions and is homologous to the structure of IL-1RI. In the other subdivision, which includes the TLRs, the extracellular domain contains leucine-rich repeats.

Because the cytoplasmic TIR domains of the TLRs are homologous to the cytoplasmic region of IL-1RI, it is not surprising that some shared mechanisms are responsible for downstream signaling (Figs. 4-3 and 4-4). In the MyD88-dependent pathway, an adapter protein, myeloid differentiation primary response factor 88 (MyD88), links the receptor to another protein called IL-1 receptor–associated kinase 1 (IRAK-1). On binding of the ligand to the TLR (or IL-1RI), IRAK-1 is phosphorylated and dissociates from the receptor complex, thereby allowing it to interact with another signaling protein, TNF receptor–activated factor 6 (TRAF6). This process results in the activation of nuclear factor κB (NF-κB), a pivotal proinflammatory transcription factor, as well as the phosphorylation signaling cascades involving mitogen-activated protein kinases (MAPKs).

FIGURE 4-3 Simplified representation of the intracellular signal transduction steps, which are initiated by the binding of IL-1 to its receptor. There are two IL-1 receptors, IL-1RI and IL-1RII. Only IL-1RI participates in signal transduction, and signaling via this receptor requires the participation of another transcytoplasmic protein, IL-1RAcP. The interaction of IL-1 with IL-1RI and IL-1RAcP leads to the formation of a trimolecular complex, which in turn results in the docking of yet another protein, IRAK-1. As a result of its interaction with MyD88, IRAK-1 is phosphorylated and activates another signaling protein, TRAF6. The IRAK-TRAF6 complex activates various downstream kinase cascades, ultimately leading to the activation of key transcription factors, such as NF-κB, and transcriptional activation of various IL-1–responsive genes.

FIGURE 4-4 Simplified representation of the intracellular signal transduction steps, which are initiated by the binding of the microbial product, LPS, to TLR4. The interaction of LPS with TLR4 requires several extracellular accessory proteins—LBP, CD14 (a glycophosphoinositol-anchored cell surface receptor), and MD2. After assembly of the extracellular LPS-LBP-CD14-TLR4-MD2 complex, signaling can follow two different pathways. In the more immediate MyD88-dependent signaling pathway, an adapter protein, MyD88, links the intracellular portion of TLR4 to other adaptor proteins called IRAK-1 and IRAK-4. Phosphorylation of IRAK-1 allows it to dissociate from the receptor complex, thereby permitting it to interact with another signaling protein, TRAF6. This process results in the activation of NF-κB, a pivotal proinflammatory transcription factor, as well as signaling cascades involving MAPKs.⁶ In the more delayed MyD88-independent pathway, the adapter proteins, TRIF and TRAM, lead to the activation of the serine-threonine kinase, TANK-binding kinase (TBK) 1, which leads to the activation of the transcription factor, IRF3. After phosphorylation, IRF3 forms a complex with cyclic adenosine monophosphate (cAMP) response element-binding (CREB) protein-binding protein (CREBBP), and this complex translocates to the nucleus, leading to the transcription of the genes for IFN-α and IFN-β, as well as other interferon-induced genes. The association of TRIF with the TIR domain of TLR4 also leads to the activation of NF-κB via pathways, which involve TRAF6 as well as another adapter protein called RIP1 (not shown).