Toll-like Receptors

The Toll-like receptors (TLRs), an evolutionarily conserved family of pattern recognition receptors expressed on many cell types, recognize products of a wide variety of microbes. Toll was originally identified as a Drosophila gene involved in establishing the dorsal-ventral axis during embryogenesis of the fruit fly, but subsequently it was discovered that the Toll protein also mediated antimicrobial responses in these organisms. This discovery led to the identification of mammalian homologues of Toll, which were named Toll-like receptors. There are 9 different functional TLRs in humans, named TLR1 to TLR9 (Fig. 4-2). The TLRs are type I integral membrane glycoproteins that contain leucine-rich repeats flanked by characteristic cysteine-rich motifs in their extracellular regions, which are involved in ligand binding, and a Toll/IL-1 receptor (TIR) homology domain in their cytoplasmic tails, which is essential for signaling. TIR domains are also found in the cytoplasmic tails of the receptors for the cytokines IL-1 and IL-18, and similar signaling pathways are engaged by TLRs, IL-1, and IL-18.

FIGURE 4–2 Structure, location, and specificities of mammalian TLRs.

Note that some TLRs are expressed in endosomes and some on the cell surface.

Mammalian TLRs are involved in responses to a wide variety of molecules that are expressed by microbial but not by healthy mammalian cells. The ligands that the different TLRs recognize are structurally diverse and include products of all classes of microorganism (see Fig. 4-2). Examples of bacterial products that bind to TLRs are LPS and lipoteichoic acid, constituents of the cell walls of gram-negative bacteria and gram-positive bacteria, respectively, and flagellin, the protein subunit component of the flagella of motile bacteria. Examples of TLR ligands produced by viruses are double-stranded RNAs, which compose the genomes of some viruses and are generated during the life cycle of most RNA viruses but are not produced by eukaryotic cells, and single-stranded RNAs, which are distinguished from cellular cytoplasmic single-stranded RNA transcripts by their location within endosomes and by their high guanosine and uridine content. Fungal mannose polysaccharides (mannans) are also TLR ligands.

TLRs are also involved in response to endogenous molecules whose expression or location indicates cell damage. Examples of host molecules that engage TLRs include heat shock proteins (HSPs), which are chaperones induced in response to various cell stresses, and high-mobility group box 1 (HMGB1), an abundant DNA-binding protein involved in transcription and DNA repair. Both HSPs and HMGB1 are normally intracellular but may become extracellular when released from injured or dying cells. From their extracellular location, they activate TLR2 and TLR4 signaling in dendritic cells, macrophages, and other cell types.

The structural basis of TLR specificities resides in the multiple extracellular leucine-rich modules of these receptors, which bind directly to PAMPs or to adaptor molecules that bind the PAMPs. There are between 16 and 28 leucine-rich repeats in TLRs, and each of these modules is composed of 20 to 30 amino acids that include conserved LxxLxLxxN motifs (where L is leucine, x is any amino acid, and N is asparagine) and amino acid residues that vary between different TLRs. The ligand-binding variable residues of the modules are on the convex surface formed by α helices and β turns or loops. These repeats contribute to the ability of some TLRs to bind hydrophobic molecules such as bacterial LPS. Ligand binding to the leucine-rich domains causes physical interactions between TLR molecules and the formation of TLR dimers. The repertoire of specificities of the TLR system is extended by the ability of TLRs to heterodimerize with one another. For example, dimers of TLR2 and TLR6 are required for responses to peptidoglycan.

Specificities of the TLRs are also influenced by various non-TLR accessory molecules. This is best defined for the TLR4 response to LPS. LPS first binds to soluble LPS-binding protein in the blood or extracellular fluid, and this complex serves to facilitate delivery of the LPS to the surface of the responding cell. An extracellular protein called MD2 (myeloid differentiation protein 2) binds to the lipid A component of LPS, forming a complex that then interacts with TLR4 and initiates signaling. Another protein called CD14 is also required for efficient LPS-induced signaling. CD14 is expressed by most cells (except endothelial cells) as a soluble protein or as a glycophosphatidylinositol-linked membrane protein. Both CD14 and MD2 can also associate with other TLRs. Thus, different combinations of accessory molecules in TLR complexes may serve to broaden the range of microbial products that can induce innate immune responses.

TLRs are found on the cell surface and on intracellular membranes and are thus able to recognize microbes in different cellular locations (see Fig 4-2). TLRs 1, 2, 4, 5, and 6 are expressed on the plasma membrane, where they recognize various PAMPs in the extracellular environment. Some of the most potent microbial stimuli for innate immune responses bind to these plasma membrane TLRs, such as bacterial LPS and lipoteichoic acid, which are recognized by TLRs 2 and 4, respectively. In contrast, TLRs 3, 7, 8, and 9 are mainly expressed inside cells on endoplasmic reticulum and endosomal membranes, where they detect several different nucleic acid ligands (see Fig. 4-2). Some of these nucleic acids are much more abundantly expressed by microbes than by mammals, such as double-stranded RNA, which is made by RNA viruses and binds to TLR3, and unmethylated CpG motifs common in prokaryotic DNA, which bind to TLR9. Single-stranded RNA, which binds to TLR8, and single- or double-stranded DNA, which binds to TLR9, are not uniquely expressed by microbes, but the relative specificity of these TLRs for microbial products is linked to their endosomal location. Host cell RNA and DNA are not normally present in endosomes, but microbial RNA and DNA may end up in endosomes of neutrophils, macrophages, or dendritic cells when the microbes are phagocytosed by these cells. Furthermore, host DNA from cells that have died because of infection or other causes may end up in the endosomes of the phagocytes. In other words, TLRs 3, 7, 8, and 9 may distinguish healthy self from foreign or unhealthy self on the basis, in part, of the cellular location of the nucleic acids they bind. A protein in the endoplasmic reticulum called UNC-93B is required for the endosomal localization and proper function of TLR3, 7, 8 and 9.

TLR recognition of microbial ligands results in the activation of several signaling pathways and ultimately transcription factors, which induce the expression of genes whose products are important for inflammatory and antiviral responses (Fig. 4-3). The signaling pathways are initiated by ligand binding to the TLR at the cell surface or in the endoplasmic reticulum or endosomes, leading to dimerization of the TLR proteins. Ligand-induced TLR dimerization is predicted to bring the TIR domains of the cytoplasmic tails of each protein close to one another. This is followed by recruitment of TIR domain–containing adaptor proteins, which facilitate the recruitment and activation of various protein kinases, leading to the activation of different transcription factors. The major transcription factors that are activated by TLR signaling pathways are nuclear factor κB (NF-κB), activation protein 1 (AP-1), interferon response factor 3 (IRF3), and IRF7. NF-κB and AP-1 stimulate the expression of genes encoding many of the molecules required for inflammatory responses including inflammatory cytokines (e.g., TNF and IL-1), chemokines (e.g., CCL2 and CXCL8), and endothelial adhesion molecules (e.g., E-selectin) (discussed later). IRF3 and IRF7 promote production of type I interferons (IFN-α and IFN-β), important for anti-viral innate immune responses.

FIGURE 4–3 Signaling functions of TLRs.

TLRs 1, 2, 5, and 6 use the adaptor protein MyD88 and activate the transcription factors NF-κB and AP-1. TLR3 uses the adaptor protein TRIF and activates the IRF3 and IRF7 transcription factors. TLR4 can activate both pathways. TLRs 7 and 9 in the endosome use MyD88 and activate both NF-κB and IRF7 (not shown).

Different combinations of adaptors and signaling intermediates are used by different TLRs, which is the basis for common and unique downstream effects of the TLRs. For example, cell surface TLRs that engage the adaptor MyD88 lead to NF-κB activation, and TLR signaling that uses the adaptor called TRIF (TIR domain–containing adaptor inducing IFN-β) leads to IRF3 activation. All TLRs except TLR3 signal through MyD88 and are therefore capable of activating NF-κB and inducing an inflammatory response. TLR3 signals through TRIF and therefore activates IRF3 and induces expression of type I interferons. TLR4 signals through both MyD88 and TRIF and is able to induce both types of responses. Endosomal TLRs 7 and 9, which are most highly expressed in plasmacytoid dendritic cells, signal through a MyD88-dependent, TRIF-independent pathway that activates both NF-κB and IRF4. Therefore, TLR7 and TLR9, like TLR4, induce both inflammatory and antiviral responses. Details of NF-κB activation are discussed in Chapter 7.

Cytosolic Receptors for PAMPs and DAMPs

In addition to the membrane-bound TLRs, which sense pathogens outside cells or in endosomes, the innate immune system has evolved to equip cells with pattern recognition receptors that detect infection or cell damage in the cytoplasm (see Fig. 4-1 and Table 4-3). The two major classes of these cytoplasmic receptors are NOD-like receptors and RIG-like receptors. These cytoplasmic receptors, like TLRs, are linked to signal transduction pathways that promote inflammation or type I interferon production. The ability of the innate immune system to detect infection in the cytoplasm is important because parts of the normal life cycles of some microbes, such as viral gene translation and viral particle assembly, take place in the cytoplasm. Some bacteria and parasites have mechanisms to escape from phagocytic vesicles into the cytoplasm. Microbes can produce toxins that create pores in host cell plasma membranes, including endosomal membranes, through which microbial molecules can enter the cytoplasm. These pores can also result in changes in the concentration of endogenous molecules in the cytoplasm, which are reliable signs of infection and damage and are detected by the cytoplasmic receptors.

NOD-like Receptors

NOD-like receptors (NLRs) are a family of more than 20 different cytosolic proteins, some of which sense cytoplasmic PAMPs and DAMPs and recruit other proteins to form signaling complexes that promote inflammation. This family of proteins is named after NOD (nucleotide oligomerization domain–containing protein). Typical NLR proteins contain at least three different domains with distinct structures and functions. These include a leucine-rich repeat domain that senses the presence of ligand, similar to the leucine-rich repeats of TLRs; a NACHT (neuronal apoptosis inhibitory protein [NAIP], CIITA, HET-E, and TP1) domain, which allows NLRs to bind to one another and form oligomers; and an effector domain, which recruits other proteins to form signaling complexes. There are three NLR subfamilies, the members of which use different effector domains to initiate signaling, called CARD, Pyrin, and BIR domains. NLRs are found in a wide variety of cell types, although some NLRs have restricted tissue distributions. Some of the best studied NLRs are found in immune and inflammatory and epithelial barrier cells.

NOD1 and NOD2, members of the CARD domain–containing NOD subfamily of NLRs, are expressed in the cytoplasm of several cell types including mucosal epithelial cells and phagocytes, and they respond to bacterial cell wall peptidoglycans. NOD2 is particularly highly expressed in intestinal Paneth cells, where it stimulates expression of antimicrobial substances called defensins in response to pathogens. NOD1 recognizes substances derived mainly from gram-negative bacteria, whereas NOD2 recognizes a distinct molecule called muramyl dipeptide from both gram-negative and gram-positive organisms. These peptides are released from intracellular or extracellular bacteria; in the latter case, their presence in the cytoplasm requires specialized mechanisms of delivery of the peptides into host cells. These mechanisms include type III and type IV secretion systems, which have evolved in pathogenic bacteria as a means of delivering toxins into host cells. When oligomers of NODs recognize their peptide ligands, including bacterial toxins, a conformational change occurs that allows the CARD effector domains of the NOD proteins to recruit multiple copies of the kinase RIP2, forming a signaling complex that has been called the NOD signalosome. The RIP2 kinases in these complexes activate NF-κB, which promotes inflammatory gene expression, similar to TLRs that signal through MyD88, discussed earlier. Both NOD1 and NOD2 appear to be important in innate immune responses to bacterial pathogens in the gastrointestinal tract, such as Helicobacter pylori and Listeria monocytogenes. There is great interest in the finding that certain NOD2 polymorphisms increase the risk for an inflammatory disease of the bowel called Crohn’s disease, probably because of a defective innate response to commensal and pathogenic organisms in the intestine. Also, mutations of NOD2 that cause increased NOD signaling lead to a systemic inflammatory disease called Blau’s syndrome.

The NLRP subfamily of NLRs respond to cytoplasmic PAMPs and DAMPs by forming signaling complexes called inflammasomes, which generate active forms of the inflammatory cytokine IL-1 (Fig. 4-4). There are 14 NLRPs (NLR family, pyrin-domain-containing proteins), most of which share a Pyrin effector domain, named after the Greek root pyro, meaning heat, because it was first identified in a mutated gene that is associated with an inherited febrile illness. Inflammasomes containing only three of these NLRPs have been well studied, notably IPAF/NLRC4, NLRP3, and NLRP1. When these NLRPs are activated by the presence of microbial products or changes in the amount of endogenous molecules or ions in the cytoplasm, they bind other proteins through homotypic interactions between shared structural domains, thereby forming the inflammasome complex. For example, after binding of a ligand, multiple identical NLRP3 proteins interact to form an oligomer, and individual NLRP3 proteins in the oligomer each bind an adaptor protein called ASC. The adaptors then bind an inactive precursor form of the enzyme caspase-1 through interactions of caspase-recruitment domains on both proteins. Caspases are proteases with cysteine residues in their active site that cleave substrate proteins at aspartate residues. Caspase-1 becomes active only after recruitment to the inflammasome complex. Although several other caspases participate in a form of cell death called apoptosis (see Chapter 14), the main function of caspase-1 is to cleave the inactive cytoplasmic precursor forms of two homologous cytokines called IL-1β and IL-18. Caspase-1 cleavage generates active forms of these cytokines, which then leave the cell and perform various proinflammatory functions. We will describe the action of these cytokines and the inflammatory response in detail later in the chapter. Suffice it to say here that the inflammation induced by IL-1 serves a protective function against the microbes that incite the formation of the inflammasome. When inflammasome activity is abnormally stimulated, the abundant IL-1 that is produced can cause tissue damage. For example, some of the hereditary periodic fevers (also called autoinflammatory syndromes), which are rare diseases characterized by repeated bouts of fever, inflammation, and tissue destruction, are caused by gain-of-function mutations of the NLRP3 gene, and IL-1 antagonists are very effective in treating these diseases.

FIGURE 4–4 The inflammasome.

The activation of the NLRP3 inflammasome, which processes pro–IL-1β to active IL-1, is shown. Inflammasomes with other NLRP proteins function in a similar way. Pro–IL-1β expression is induced by various PAMPs or DAMPs through pattern recognition receptor signaling, such as a TLR, as shown. CPPD, calcium pyrophosphate dihydrate; MSU, monosodium urate.

  Page 64 

  Page 65 

NLRP-inflammasome responses are induced by a wide variety of cytoplasmic stimuli, including microbial products, environmentally or endogenously derived crystals, and reduction in cytoplasmic potassium ion (K⁺) concentrations, that are often associated with infections and cell stress (see Fig. 4-4). Microbial products that activate NLRP-inflammasomes include bacterial molecules such as flagellin, muramyl dipeptide, LPS, and pore-forming toxins as well as bacterial and viral RNA. Crystalline substances are also potent activators of inflammasomes, and these crystals can be derived from the environment, such as asbestos and silica, or they can be endogenously derived from dead cells, such as monosodium urate and calcium pyrophosphate dehydrate. Another endogenous stimulus of inflammasome activation is extracellular ATP, perhaps released from dead cells and transported into the cytoplasm of the responding cell. The structural diversity of the agents that activate the inflammasome suggests that they do not directly bind to NLRP proteins but may act by inducing a smaller set of changes in endogenous cytoplasmic conditions that activate the NLRPs. Reduced cytoplasm potassium ion concentrations may be one such common mechanism because reductions in cellular K⁺ induced by some bacterial pore-forming toxins can activate inflammasomes, and many of the other known inflammasome activators cause increased K⁺ efflux from cells. Another common mechanism implicated in inflammasome activation is the generation of reactive oxygen species, which are toxic free radicals of oxygen that are often produced during cell injury. A type of inflammasome that uses a protein called AIM2 (absent in melanoma-2) rather than an NLRP-family protein, recognizes cytosolic dsDNA.

The discovery that some crystalline substances are potent inflammasome activators has changed our understanding of certain inflammatory diseases. Gout is a painful inflammatory condition of the joints that has long been known to be caused by deposition of monosodium urate crystals in joints. Based on the understanding that urate crystals activate the inflammasome, there is interest in using IL-1 antagonists to treat cases of severe gout that are resistant to conventional anti-inflammatory drugs. Similarly, pseudogout is caused by deposition of calcium pyrophosphate crystals and inflammasome activation. Occupational inhalation of silica and asbestos can cause chronic inflammatory and fibrotic disease of the lung, and there is also interest in the potential of blocking the inflammasome or IL-1 to treat these diseases.

Other Cell-Associated Pattern Recognition Receptors

Several types of plasma membrane and cytoplasmic receptors other than the classes described before are expressed on the plasma membranes of various cell types and recognize microbial molecules (see Table 4-3). Some of these receptors transmit activating signals, similar to TLRs, that promote inflammatory responses and enhance killing of microbes. Other receptors mainly participate in the uptake of microbes into phagocytes.

Epithelial Barriers

Intact epithelial surfaces form physical barriers between microbes in the external environment and host tissue, and epithelial cells produce antimicrobial chemicals that further impede the entry of microbes (Fig. 4-5). The main interfaces between the environment and the mammalian host are the skin and the mucosal surfaces of the gastrointestinal, respiratory, and genitourinary tracts. These interfaces are lined by continuous layers of specialized epithelial cells that serve many physiologic functions, including preventing the entry of microbes. Loss of the integrity of these epithelial layers by trauma or other reasons predisposes an individual to infections. The protective barrier function is in large part physical. The epithelial cells form tight junctions with one another, blocking passage of microbes between the cells. The outer layer of keratin, which accumulates as surface skin keratinocytes die, serves to block microbial penetration into deeper layers of the epidermis. Mucus, a viscous secretion containing glycoproteins called mucins, is produced by respiratory, gastrointestinal, and urogenital epithelial cells. Mucus physically impairs microbial invasion and facilitates microbe removal by ciliary action in the bronchial tree and peristalsis in the gut. Although these physical barrier properties alone are very important in host defense, other epithelial defense mechanisms have evolved to complement the physical barrier.

FIGURE 4–5 Epithelial barriers.

Epithelia at the portals of entry of microbes provide physical barriers, produce antimicrobial substances, and harbor intraepithelial lymphocytes that are believed to kill microbes and infected cells.

Epithelial cells as well as some leukocytes produce peptides that have antimicrobial properties. Two structurally distinct families of antimicrobial peptides are the defensins and the cathelicidins.

Barrier epithelia contain certain types of lymphocytes, including intraepithelial T lymphocytes, that recognize and respond to commonly encountered microbes. Intraepithelial T lymphocytes are present in the epidermis of the skin and in mucosal epithelia. Various subsets of intraepithelial lymphocytes are present in different proportions, depending on species and tissue location. These subsets are distinguished mainly by the type of T cell antigen receptors (TCRs) they express. Some intraepithelial T lymphocytes express the conventional αβ form of TCR, which is present on most T cells in lymphoid tissues. Other T cells in epithelia express a form of antigen receptor called the γδ receptor that may recognize peptide and nonpeptide antigens. A common characteristic of these T cells is the limited diversity of their antigen receptors compared with most T cells in the adaptive immune system. The intraepithelial T lymphocytes are believed to recognize a limited number of commonly encountered microbial structures (e.g., PAMPs). Intraepithelial lymphocytes may function in host defense by secreting cytokines, activating phagocytes, and killing infected cells.

Phagocytes

Cells that have specialized phagocytic functions, primarily macrophages and neutrophils, are the first line of defense against microbes that breach epithelial barriers. We have introduced these cell types in Chapter 2, and we will discuss many other details of their functions later in this chapter and in other chapters. For now, it is important to know that these phagocytic cells perform two general types of functions in defense against microbes. First, they are able to internalize and kill microbes. Neutrophils and macrophages are particularly good at this function. Second, phagocytes respond to microbes by producing various cytokines that promote inflammation and also enhance the antimicrobial function of host cells at the site of infection. Among the “professional phagocytes,” macrophages are particularly good at this second function. Macrophages are also involved in the repair of damaged tissues, which is another function important in host defense. The essential role that phagocytes play in innate immune defense against microbes is demonstrated by the high rate of lethal bacterial and fungal infections in patients with low blood neutrophil counts caused by bone marrow cancers or cancer therapy and in patients with inherited deficiencies in the functions of phagocytes.

Dendritic Cells

Dendritic cells perform essential recognition and effector roles in innate immunity. We introduced dendritic cells in Chapter 2, and their role in antigen presentation to T cells is discussed in Chapter 6. Recall that this heterogeneous family of bone marrow–derived cells with long dendrite-like cytoplasmic processes are constitutively present in epithelia and most tissues of the body. Because of their placement and morphology, these cells are poised to detect invading microbes. Furthermore, dendritic cells express more different types of TLRs and cytoplasmic pattern recognition receptors than any other cell type, making them the most versatile sensors of PAMPs and DAMPs among all cell types in the body. One particular subset of dendritic cells, called plasmacytoid dendritic cells because their morphology is similar to antibody-producing plasma cells, is the major source of antiviral cytokines, type I interferons, produced in response to viral infections. This feature of plasmacytoid dendritic cells is due in part to the fact that these cells, more than other cell types, abundantly express the endosomal TLRs (TLRs 3, 7, 8, 9) that recognize nucleic acids of viruses that have been internalized into the cell. We will discuss the antiviral actions of type I interferons in more detail later in the chapter.

Dendritic cells are uniquely capable of triggering and directing adaptive T cell–mediated immune responses, and this is dependent on their innate immune responses to microbes. This capability reflects the ability of dendritic cells to take up microbial protein antigens, to transport them to lymph nodes where naive T cells home, and to alter and display the protein antigens in a way that the T cells can recognize. These functions will be discussed in great detail in Chapter 6. Importantly, the innate response of dendritic cells to PAMPs is essential for these functions, which are enhanced by TLR signaling. Furthermore, TLR signaling induces dendritic cell expression of molecules, including costimulatory molecules and cytokines, that are needed, in addition to antigen, for the activation of the naive T cells and their differentiation into effector T cells. Depending on the nature of the microbe that induces the innate response, a dendritic cell will direct naive T cell differentiation into distinct types of effector cells, such as IFN-γ–producing T_H1 cells or IL-17–producing T_H17 cells. The influence of dendritic cells on T cell activation and differentiation will be discussed further in Chapter 9.

Natural Killer Cells

Natural killer (NK) cells are lymphocytes distinct from T and B cells that play important roles in innate immune responses mainly against intracellular viruses and bacteria. The term natural killer derives from the fact that these cells are capable of performing their killing function without a need for clonal expansion and differentiation, which is required for effector responses of the immune system’s other killer cells, the cytotoxic T lymphocytes (CTLs). NK cells constitute 5% to 15% of the mononuclear cells in the blood and spleen. They are rare in other lymphoid organs but are concentrated in certain organs such as the liver and gravid uterus. NK cells arise from bone marrow precursors and appear as large lymphocytes with numerous cytoplasmic granules. NK cells do not express highly diverse, clonally distributed antigen receptors typical of B and T cells. Rather, they use germline DNA-encoded receptors, discussed later, to distinguish pathogen-infected from healthy cells. They can be identified in the blood by expression of CD56 and the absence of CD3, two membrane proteins often found together on activated CTLs.

Recognition of Infected and Stressed Cells by NK Cells

NK cells distinguish infected and stressed cells from healthy cells, and NK cell activation is regulated by a balance between signals that are generated from activating receptors and inhibitory receptors. There are several families of these receptors (Fig. 4-6), some members of which we will discuss later. These receptors recognize molecules on the surface of other cells and generate activating or inhibitory signals that promote or inhibit NK responses. In general, the activating receptors recognize ligands on infected and injured cells, and the inhibitory receptors recognize healthy normal cells. When an NK cell interacts with another cell, the outcome is determined by the integration of signals generated from the array of inhibitory and activating receptors that are expressed by the NK cell and that interact with ligands on the other cell. Because of the stochastic nature of their expression, there is significant diversity in the array of activating and inhibitory receptors that different NK cells express in any one individual. The result of this is that an individual’s NK cells will respond to different types of microbes or infected cells. Furthermore, the genes encoding many of these receptors are polymorphic, meaning that there are several variants of the genes in the population, so that one person will express a slightly different form of the receptors than another person.

FIGURE 4–6 Functions of activating and inhibitory receptors of NK cells.

A, Activating receptors of NK cells recognize ligands on target cells and activate protein tyrosine kinase (PTK), whose activity is inhibited by inhibitory receptors that recognize class I MHC molecules and activate protein tyrosine phosphatases (PTP). NK cells do not efficiently kill class I MHC–expressing healthy cells. B, If a virus infection or other stress inhibits class I MHC expression on infected cells and induces expression of additional activating ligands, the NK cell inhibitory receptor is not engaged and the activating receptor functions unopposed to trigger responses of NK cells, such as killing of target cells and cytokine secretion. C. Cells stressed by infection or neoplastic transformation may express increased amounts of activating ligands, which bind NK cell activating receptors and induce more tyrosine phosphorylation than can be removed by inhibitory receptor associated phosphatases, resulting in killing of the stressed cells. Structural details and ligands of inhibitory and activating NK cell receptors are shown in Figure 4-7.

Most NK cells express inhibitory receptors that recognize class I major histocompatibility complex (MHC) molecules, which are cell surface proteins normally expressed on almost all healthy cells in the body (Fig. 4-7). A major function of class I MHC molecules, distinct from their role in regulating NK cell activation, is to display peptides derived from cytoplasmic proteins, including microbial proteins, on the cell surface for recognition by CD8⁺ T lymphocytes. We will describe the structure and function of MHC molecules in relation to CD8⁺ T cell antigen recognition in Chapter 6. For now, it is important to understand that NK cells use fundamentally different types of receptors than do T cells to recognize class I MHC molecules. Unlike T cells, many of the NK receptors for class I MHC respond by inhibiting NK activation. This is useful because normal cells express class I MHC molecules, and many viruses and other causes of cell stress lead to a loss of cell surface expression of class I MHC. Thus, NK cells interpret the presence of class I MHC molecules as markers of normal, healthy self, and their absence is an indication of infection or damage. Conversely, NK cells will not receive inhibitory signals from infected or stressed cells. At the same time, the NK cells are likely to receive activating signals from the same infected cells through activating receptors. The net result will be activation of the NK cell to secrete cytokines and to kill the infected or stressed cell. This ability of NK cells to become activated by host cells that lack class I MHC has been called recognition of missing self.

FIGURE 4–7 Structure and ligands of activating and inhibitory receptors of NK cells.

Examples of inhibitory and activating NK cell receptors and their ligands. CD16 and the natural cytotoxic receptors (NCRs) associate with ζ chain homodimers, FcεRIγ homodimers, or ζ-FcεRIγ heterodimers. There are multiple different KIRs, with varying ligand specificities.

  Page 69 

Inhibitory receptors of NK cells share the common feature of a structural motif in their cytoplasmic tails, called an immunoreceptor tyrosine-based inhibition motif (ITIM), which engages molecules that block the signaling pathways of activating receptors (see Figs. 4-6 and 4-7). ITIMs contain tyrosine residues that are phosphorylated on ligand binding to the inhibitory receptor. This leads to the recruitment and activation of phosphatases, which remove phosphates from several signaling proteins or lipids generated by the signaling pathways downstream of NK activating receptors. The end result is blocking of the signaling functions of activating receptors. ITIMs are found in cytoplasmic tails of other receptors besides NK inhibitory receptors, and their structure and signaling functions are discussed in more detail in Chapter 7.

The largest group of NK inhibitory receptors are the killer cell immunoglobulin-like receptors (KIRs), which are members of the immunoglobulin (Ig) superfamily. Members of this family all contain a structural domain called an Ig fold, first identified in antibody (also known as Ig) molecules, discussed in Chapter 5. KIRs bind a variety of class I MHC molecules. A second important group of NK inhibitory receptors belong to the C-type lectin family, which includes proteins with carbohydrate-binding properties, as discussed earlier. One of these receptors is a heterodimer called CD94/NKG2A, which recognizes a class I MHC molecule called HLA-E. Interestingly, HLA-E displays peptides derived from other class I MHC molecules, so in essence, CD94/NKG2A is a surveillance receptor for several different class I MHC molecules. A third family of NK inhibitory receptors, called the leukocyte Ig-like receptors (LIRs), are also Ig superfamily members that bind class I MHC molecules, albeit with lower affinity than the KIRs, and are more highly expressed on B cells than on NK cells.

  Page 70 

Activating receptors on NK cells recognize a heterogeneous group of ligands, some of which may be expressed on normal cells and others of which are expressed mainly on cells that have undergone stress, are infected with microbes, or are transformed. The molecular features of the ligands for many of these receptors are not well characterized. The induced expression of ligands on unhealthy cells that bind to activating receptors on NK cells may lead to signals that overwhelm the signals from inhibitory receptors, especially if class I MHC is also reduced or lost on the unhealthy cell (see Fig. 4-6).

Most activating NK receptors share the common feature of a structural motif in their cytoplasmic tails, called an immunoreceptor tyrosine-based activation motif (ITAM), which engages in signaling events that promote target cell killing and cytokine secretion (see Fig. 4-7). In some of these receptors, a single polypeptide chain contains the ITAM as well as the extracellular ligand-binding portion. In other receptors, the ITAMs are in separate polypeptide chains, such as FcεRIγ, ζ, and DAP12, that do not bind ligand but are noncovalently associated with the ligand-binding chain. ITAMs are also found in cytoplasmic tails of other multichain signaling receptors in the immune system, including the antigen receptors on T and B cells. After ligand binding to the NK cell activating receptors, tyrosine residues within the ITAMs become phosphorylated by cytoplasmic kinases, other protein kinases are recruited to the modified ITAMs and become activated, and these kinases contribute to further signaling by phosphorylating additional proteins. The structure and signaling functions of ITAMs are discussed in more detail in Chapter 7.

  Page 71 

Many of the NK cell activating receptors are members of the C-type lectin or KIR families, which also include inhibitory receptors, as discussed before. Some of the activating receptors appear to bind class I MHC molecules, like the inhibitory receptors, but it is not known how these receptors are preferentially activated by infected or damaged cells. It is also clear that the activating receptors recognize ligands other than classical MHC molecules. One well-studied NK cell activating receptor in the C-type lectin family is NKG2D, which binds class I MHC–like proteins, including MIC-A and MIC-B, that are found on virally infected cells and tumor cells but not normal cells. The NKG2D receptor associates with a signaling subunit named DAP10, which has a signaling motif different from the ITAMs in other activating receptors but also enhances NK cell cytotoxicity against target cells.

Another important activating receptor on NK cells is CD16 (FcγRIIIa), which is a low-affinity receptor for IgG antibodies. Antibody molecules have highly variable antigen-binding ends, and on the opposite end, they have an invariant structure, called the Fc region, that interacts with various other molecules in the immune system. We will describe the structure of antibodies in detail in Chapter 5 but, for now, it is sufficient to know that CD16 binds to the Fc regions of certain types of antibodies called IgG1 or IgG3. CD16 associates with one of three different ITAM-containing signaling proteins (e.g., FcεRIγ, ζ, and DAP12 proteins). During an infection, the adaptive immune system produces IgG1 and IgG3 antibodies that specifically bind to the infecting microbes and their antigens on infected cells, and CD16 on NK cells can bind to the Fc parts of these antibodies. As a result, CD16 generates activating signals, through the associated signaling partners, and the NK cells may kill the infected cells that have been coated with antibody molecules. This process is called antibody-dependent cell-mediated cytotoxicity; it is an effector function of adaptive immunity and will be discussed in Chapter 12 when we consider humoral immunity.

The ability of activating receptors to induce functional responses in NK cells is enhanced by cytokines. The major cytokines of the innate immune system that stimulate NK function are IL-12, IL-15, IL-18, and type I interferons (discussed later). Each of these cytokines enhances the cytotoxic activity of NK cells and the amount of the cytokine IFN-γ the NK cells secrete. IFN-γ has various antimicrobial effects and will be discussed in detail in Chapter 10. In addition, IL-12 and IL-15 are important growth factors for NK cells.

KIR genes are polymorphic, meaning that there are several allelic variants in the human population, and groups of KIR alleles are often inherited together from a single parent. These groups of linked genes are called KIR haplotypes. There are two major KIR haplotypes and some rarer ones. Haplotypes differ in the number of receptors encoded, and some have more or fewer activating receptors than others. Some haplotypes are associated with increased susceptibility to some diseases, including spontaneous abortion and uveitis.

Effector Functions of NK Cells

The effector functions of NK cells are to kill infected cells and to activate macrophages to destroy phagocytosed microbes (Fig. 4-8). The mechanism of NK cell–mediated cytotoxicity is essentially the same as that of CD8⁺ CTLs, which we will describe in detail in Chapter 10. NK cells, like CTLs, have granules containing proteins that mediate killing of target cells. When NK cells are activated, granule exocytosis releases these proteins adjacent to the target cells. One NK cell granule protein, called perforin, facilitates the entry of other granule proteins, called granzymes, into the cytoplasm of target cells. The granzymes are enzymes that initiate a sequence of signaling events that cause death of the target cells by apoptosis. The signaling pathways that cause apoptosis are discussed in Chapter 14. By killing cells infected by viruses and intracellular bacteria, NK cells eliminate reservoirs of infection. Some tumors, especially those of hematopoietic origin, are targets of NK cells, perhaps because the tumor cells do not express normal levels or types of class I MHC molecules.

FIGURE 4–8 Functions of NK cells.

A, NK cells recognize ligands on infected cells or cells undergoing other types of stress and kill the host cells. In this way, NK cells eliminate reservoirs of infection as well as dysfunctional cells. B, NK cells respond to IL-12 produced by macrophages and secrete IFN-γ, which activates the macrophages to kill phagocytosed microbes.

  Page 72 

NK cell–derived IFN-γ serves to activate macrophages, like IFN-γ produced by T cells, and increases the capacity of macrophages to kill phagocytosed bacteria (see Chapter 10). IFN-γ produced by NK cells in lymph nodes can also direct the differentiation of naive T cells into T_H1 cells (see Chapter 9).

NK cells play several important roles in defense against intracellular microbes. They kill virally infected cells before antigen-specific CTLs can become fully active, that is, during the first few days after viral infection. Early in the course of a viral infection, NK cells are expanded and activated by IL-12 and IL-15, and they kill infected cells, especially those that display reduced levels of class I MHC molecules. In addition, IFN-γ secreted by NK cells activates macrophages to destroy phagocytosed microbes. This IFN-γ–dependent NK cell–macrophage reaction can control an infection with intracellular bacteria such as Listeria monocytogenes for several days or weeks and thus allow time for T cell–mediated immunity to develop and eradicate the infection. Depletion of NK cells leads to increased susceptibility to infection by some viruses and intracellular bacteria. In mice lacking T cells, the NK cell response may be adequate to keep infection with such microbes in check for some time, but the animals eventually succumb in the absence of T cell–mediated immunity. NK cells may also be important later in the body’s response to infection by killing infected cells that have escaped CTL-mediated immune attack by reducing expression of class I MHC molecules. Because NK cells can kill certain tumor cells in vitro, it has also been proposed that NK cells serve to kill malignant clones in vivo.

T and B Lymphocytes with Limited Antigen Receptor Specificities

As we will discuss in greater detail in later chapters, most T and B lymphocytes are components of the adaptive immune system and are characterized by a highly diverse repertoire of specificities for different antigens. The diversity of antigen receptors is generated by random somatic recombination of a large set of germline DNA segments as well as modification of nucleotide sequences at the junctions between the recombined segments, yielding unique antigen receptor genes in each lymphocyte clone (see Chapter 8). However, certain subsets of T and B lymphocytes have very little diversity because the same antigen receptor gene DNA segments are recombined in each clone and there is little or no modification of junctional sequences. It appears that these T and B cell subsets recognize structures expressed by many different or commonly encountered microbial species; in other words, they recognize PAMPs. T cell subsets with limited antigen receptor diversity include invariant natural killer T cells (iNKT), γδ T cells, and intraepithelial T cells with αβ TCRs (mentioned earlier). B cell subsets that produce antibodies with a limited set of specificities include B-1 B cells and marginal zone B cells. Although these T and B cells perform similar effector functions as do their more clonally diverse counterparts, the nature of their specificities places them in a special category of lymphocytes that is akin more to effector cells of innate immunity than to cells of adaptive immunity. These special T and B cell subsets are described in Chapters 10 and 11, respectively.

Mast Cells

Mast cells are present in the skin and mucosal epithelium and rapidly secrete proinflammatory cytokines and lipid mediators in response to infections and other stimuli. We introduced mast cells in Chapter 2. Recall that these cells contain abundant cytoplasmic granules containing various inflammatory mediators that are released when the cells are activated, either by microbial products or by a special antibody-dependent mechanism. The granule contents include vasoactive amines (such as histamine) that cause vasodilation and increased capillary permeability, and proteolytic enzymes that can kill bacteria or inactivate microbial toxins. Mast cells also synthesize and secrete lipid mediators (such as prostaglandins) and cytokines (such as TNF). Because mast cells are usually located adjacent to blood vessels (see Fig. 2-1), their released granule contents rapidly induce changes in the blood vessels that promote acute inflammation. Mast cells express TLRs, and TLR ligands can induce mast cell degranulation. Mast cell–deficient mice are impaired in controlling bacterial infections, probably because of impaired innate immune responses. Mast cell products also provide defense against helminths and are responsible for symptoms of allergic diseases. We will return to a detailed discussion of mast cells in relation to allergic diseases in Chapter 19.

Natural Antibodies

Many antibodies with millions of different fine specificities are produced in humoral immune responses by B lymphocytes and their progeny, as part of the adaptive immune system, and we will describe antibodies and B cell responses in detail in later chapters. However, there are subsets of B cells that produce antibodies with only a limited number of specificities without overt exposure to foreign antigens, and these are called natural antibodies. As is typical for other components of innate immunity, natural antibodies are already present before infections, and they recognize common molecular patterns on microbes or stressed and dying cells. Natural antibodies are usually specific for carbohydrate or lipid molecules but not proteins, and most are IgM antibodies, one of several structural classes of Ig molecules (see Chapter 5). A remarkably large proportion of the natural antibodies in humans and mice are specific for oxidized lipids, including phospholipid head groups such as lysophosphatidylcholine and phosphorylcholine, which are found on bacterial membranes and on apoptotic cells but are not exposed on the surface of healthy host cells. Some experimental evidence indicates that the natural antibodies specific for these phospholipids provide protection against bacterial infections and facilitate the phagocytosis of apoptotic cells. The anti-ABO blood group antibodies, another example of natural antibodies, recognize certain glycolipids (blood group antigens) expressed on the surface of many cell types, including blood cells. Blood group antigens and antibodies are important for transplantation but not for host defense and are discussed in Chapter 16.

The Complement System

The complement system consists of several plasma proteins that work together to opsonize microbes, to promote the recruitment of phagocytes to the site of infection, and in some cases to directly kill the microbes (Fig. 4-9). Complement activation involves proteolytic cascades, in which an inactive precursor enzyme, called a zymogen, is altered to become an active protease that cleaves and thereby induces the proteolytic activity of the next complement protein in the cascade. As the cascade proceeds, the enzymatic activities result in tremendous amplification of the amount of proteolytic products that are generated. These products perform the effector functions of the complement system. Other proteolytic cascades include the blood coagulation pathways and the kinin-kallikrein system that regulates vascular permeability.

FIGURE 4–9 Pathways of complement activation.

The activation of the complement system may be initiated by three distinct pathways, all of which lead to the production of C3b (the early steps). C3b initiates the late steps of complement activation, culminating in the production of peptides that stimulate inflammation (C5a) and polymerized C9, which forms the membrane attack complex, so called because it creates holes in plasma membranes. The principal functions of major proteins produced at different steps are shown. The activation, functions, and regulation of the complement system are discussed in much more detail in Chapter 12.

The first step in activation of the complement system is recognition of molecules on microbial surfaces but not host cells, and this occurs in three ways, each referred to as a distinct pathway of complement activation.

FIGURE 4–10 C1, mannose-binding lectin, and ficolin.

These three homologous pentameric proteins can all initiate complement activation on binding to their ligands on cell surfaces. C-type lectin–like globular heads at the end of collagenous-like stalks in the C1q and mannose-binding lectin proteins bind the Fc regions of IgM or mannose on the surface of microbes, respectively. Fibrinogen-like globular heads on ficolin bind N-acetylglucosamine on the surface of microbes. Binding results in conformational changes that activate the serine protease activity of C1r and C1s, associated with C1q, or MASP1 and MASP2, associated with mannose-binding lectin and ficolin.

Recognition of microbes by any of the three complement pathways results in sequential recruitment and assembly of additional complement proteins into protease complexes (see Fig. 4-9). One of these complexes, called C3 convertase, cleaves the central protein of the complement system, C3, producing C3a and C3b. The larger C3b fragment becomes covalently attached to the microbial surface where the complement pathway was activated. C3b serves as an opsonin to promote phagocytosis of the microbes. A smaller fragment, C3a, is released and stimulates inflammation by acting as a chemoattractant for neutrophils. C3b binds other complement proteins to form a protease called C5 convertase that cleaves C5, generating a secreted peptide (C5a) and a larger fragment (C5b) that remains attached to the microbial cell membranes. C5a is also a chemoattractant; in addition, it induces changes in blood vessels that make them leak plasma proteins and fluid into sites of infections. C5b initiates the formation of a complex of the complement proteins C6, C7, C8, and C9, which are assembled into a membrane pore, called the membrane attack complex (MAC), that causes lysis of the cells where complement is activated.

The complement system is an essential component of innate immunity, and patients with deficiencies in C3 are highly susceptible to recurrent, often lethal, bacterial infections. However, genetic deficiencies in MAC formation (the terminal product of the classical pathway) increase susceptibility to only a limited number of microbes, notably Neisseria bacteria, which have thin cell walls that make them especially susceptible to the lytic action of the MAC. The complement system will be discussed in more detail in Chapter 12.

Pentraxins

Several plasma proteins that recognize microbial structures and participate in innate immunity belong to the pentraxin family, which is a phylogenetically old group of structurally homologous pentameric proteins. Prominent members of this family include the short pentraxins C-reactive protein (CRP) and serum amyloid P (SAP) and the long pentraxin PTX3. Both CRP and SAP bind to several different species of bacteria and fungi. The molecular ligands recognized by CRP and SAP include phosphorylcholine and phosphatidylethanolamine, respectively, which are found on bacterial membranes and on apoptotic cells, as discussed earlier. PTX3 recognizes various molecules on fungi, selected gram-positive and gram-negative bacteria, and viruses. CRP, SAP, and PTX3 all activate complement by binding C1q and initiating the classical pathway.

Plasma concentrations of CRP are very low in healthy individuals but can increase up to 1000-fold during infections and in response to other inflammatory stimuli. The increased levels of CRP are a result of increased synthesis by the liver induced by the cytokines IL-6 and IL-1, which are produced by phagocytes as part of the innate immune response. Liver synthesis and plasma levels of several other proteins, including SAP and others unrelated to the pentraxins, also increase in response to IL-1 and IL-6, and as a group these plasma proteins are called acute-phase reactants.

PTX3 is produced by several cell types, including dendritic cells, macrophages, and endothelial cells, in response to TLR ligands and inflammatory cytokines, such as TNF, but it is not an acute-phase reactant. PTX3 is also stored in neutrophil granules and released as neutrophils die. PTX3 recognizes apoptotic cells and certain microorganisms. Studies with knockout mice reveal that PTX3 provides protection against some microbes, including the fungus Aspergillus fumigatus.

Collectins and Ficolins

The collectins are a family of trimeric or hexameric proteins, each subunit of which contains a collagen-like tail connected by a neck region to a calcium-dependent (C-type) lectin head. Three members of this family serve as soluble effector molecules in the innate immune system; these are mannose-binding lectin (MBL) and pulmonary surfactant proteins SP-A and SP-D.

MBL, which is a soluble pattern recognition receptor that binds carbohydrates with terminal mannose and fucose, was discussed earlier in relation to the lectin pathway of complement activation (see Fig. 4-10). MBL can also function as an opsonin by binding to and enhancing phagocytosis of microbes. Recall that opsonins simultaneously bind microbes and a surface receptor on phagocyte membranes, and in the case of MBL, the surface receptor is called the C1q receptor because it also binds C1q. This receptor mediates the internalization of microbes that are opsonized by MBL. The gene encoding MBL is polymorphic, and certain alleles are associated with impaired hexamer formation and reduced blood levels. Low MBL levels are associated with increased susceptibility to a variety of infections, especially in combination with other immunodeficiency states.

Surfactant protein A (SP-A) and surfactant protein D (SP-D) are collectins with lipophilic surfactant properties shared by other surfactants. They are found in the alveoli of the lungs, and their major functions appear to be as mediators of innate immune responses in the lung. They bind to various microorganisms and act as opsonins, facilitating ingestion by alveolar macrophages. SP-A and SP-D can also directly inhibit bacterial growth, and they may activate macrophages. SP-A– and SP-D–deficient mice have impaired abilities to resist a variety of pulmonary infections.

Ficolins are plasma proteins that are structurally similar to collectins, possessing a collagen-like domain, but instead of a C-type lectin domain, they have a fibrinogen-type carbohydrate recognition domain (see Fig. 4-10). Ficolins have been shown to bind several species of bacteria, opsonizing them and activating complement in a manner similar to that of MBL. The molecular ligands of the ficolins include N-acetylglucosamine and the lipoteichoic acid component of the cell walls of gram-positive bacteria.

Now that we have discussed the general properties and various components of the innate immune system, including the cells, cellular pathogen recognition receptors, and soluble recognition and effector molecules, we can consider how these various components work to protect against pathogens. The three major ways in which the innate immune system protects against infections is by inducing inflammation, inducing antiviral defense, and stimulating adaptive immunity. Many of these reactions are mediated by cytokines, which serve diverse and important roles in innate immunity (Table 4-4). As we discuss below, these cytokines act mainly close to their site of production (paracrine actions), but some of them can also have distant effects (endocrine actions).

TABLE 4–4 Cytokines of Innate Immunity

The Major Proinflammatory Cytokines TNF, IL-1, and IL-6

One of the earliest responses of the innate immune system to infection and tissue damage is the secretion of cytokines by tissue cells, which is critical for the acute inflammatory response. Three of the most important proinflammatory cytokines of the innate immune system are TNF, IL-1 (which we have mentioned already several times), and IL-6 (see Table 4-4). Tissue macrophages and mast cells are the major source of these cytokines, although other cell types, including endothelial and epithelial cells, can also produce IL-1 and IL-6. We will discuss the major features of these cytokines, focusing mainly on TNF and IL-1, before describing their role in acute inflammation.

Tumor Necrosis Factor

Tumor necrosis factor (TNF) is a mediator of the acute inflammatory response to bacteria and other infectious microbes. The name of this cytokine derives from its original identification as a serum substance (factor) that caused necrosis of tumors, now known to be the result of local inflammation and thrombosis of tumor blood vessels. TNF is also called TNF-α to distinguish it from the closely related TNF-β, also called lymphotoxin. TNF is produced by macrophages, dendritic cells, and other cell types. In macrophages, it is synthesized as a nonglycosylated type II membrane protein and is expressed as a homotrimer, which is able to bind to one form of TNF receptor. The membrane form of TNF is cleaved by a membrane-associated metalloproteinase, releasing a polypeptide fragment, and three of these polypeptide chains polymerize to form a triangular pyramid-shaped circulating TNF protein (Fig. 4-11). The receptor-binding sites are at the base of the pyramid, allowing simultaneous binding of the cytokine to three receptor molecules.

FIGURE 4–11 Structure of the TNF receptor with bound lymphotoxin.

The ribbon structure depicts a top view of a complex of three TNF receptors (TNF-RI) and one molecule of the bound cytokine, revealed by x-ray crystallography. Lymphotoxin is a homotrimer in which the three subunits are colored dark blue. The lymphotoxin homotrimer forms an inverted three-sided pyramid with its base at the top and its apex at the bottom. Three TNF-RI molecules, colored magenta, cyan, and red, bind one homotrimer of lymphotoxin, with each receptor molecule interacting with two different lymphotoxin monomers in the homotrimer complex. Disulfide bonds in the receptor are colored yellow. TNF is homologous to lymphotoxin and presumably binds to its receptors in the same way.

(From Banner DW, et al., Cell: Crystal structure of the soluble human 55 kd TNF receptor–human TNFβ complex: 73:431-445. © Cell Press, 1993).

There are two distinct TNF receptors called type I (TNF-RI) and type II (TNF-RII). The affinities of TNF for its receptors are unusually low for a cytokine, the K_d being only ~1 × 10⁻⁹ M for binding to TNF-RI and approximately 5 × 10⁻¹⁰ M for binding to TNF-RII. Both TNF receptors are present on most cell types. The TNF receptors are members of a large family of proteins called the TNF receptor superfamily, many of which are involved in immune and inflammatory responses. These receptors exist as trimers in the plasma membrane. Cytokine binding to some TNF receptor family members, such as TNF-RI, TNF-RII, and CD40, leads to the recruitment of proteins, called TNF receptor–associated factors (TRAFs), to the cytoplasmic domains of the receptors. The TRAFs activate transcription factors, notably NF-κB and AP-1. Cytokine binding to other family members, such as TNF-RI, leads to recruitment of an adaptor protein that activates caspases and triggers apoptosis. Thus, different members of the TNF receptor family can induce gene expression or cell death, and some can do both (see Chapter 7).

TNF production by macrophages is stimulated by PAMPs and DAMPs. TLRs, NLRs, and RLRs can all induce TNF gene expression, in part by activation of the NF-κB transcription factor. Many different microbial products can therefore induce TNF production. Large amounts of this cytokine may be produced during infections with gram-negative and gram-positive bacteria, which release the TLR ligands LPS and lipoteichoic acid, respectively, from their cell walls. Septic shock, a life-threatening condition caused when bacteria enter the blood stream, is mediated in large part by TNF. We will discuss septic shock later in this chapter.

Recruitment of Leukocytes to Sites of Infection

Recruitment of large numbers of neutrophils, followed by monocytes, from blood into tissues typically occurs as part of the acute inflammatory response to infections and tissue injury. The cytokines TNF, IL-1, and IL-6 and chemokines, which are secreted in the local sites of infection or tissue injury, have multiple effects on vascular endothelial cells, leukocytes, and bone marrow, which together increase the local delivery of cells that can fight infections and repair tissues (see Fig. 3-3, Chapter 3). Leukocyte recruitment was described in Chapter 3 and will be only briefly considered here.

Both TNF and IL-1 induce postcapillary venule endothelial cells to express E-selectin and to increase their expression of ICAM-1 and VCAM-1, the ligands for leukocyte integrins. These changes in endothelial adhesion molecule expression are the result of TNF and IL-1 activation of transcription factors, including NF-κB, leading to new adhesion molecule gene transcription. P-selectin expression is also induced on venular endothelial cells at sites of infection and tissue injury, but in large part, this is due to the effects of histamine and thrombin, which stimulate the rapid mobilization of P-selectin stored in granules in the endothelial cell to the cell surface.

TNF and IL-1 also stimulate various cells to secrete chemokines, such as CXCL1 and CCL2, that bind to receptors on neutrophils and monocytes, respectively, increase the affinity of leukocyte integrins for their ligands, and stimulate directional movement of leukocytes. The result of increased selectin, integrin, and chemokine expression is an increase in neutrophil and monocyte adhesion to endothelial cells and transmigration through the vessel wall. The leukocytes that accumulate in the tissues compose an inflammatory infiltrate. The actions of TNF on endothelium and leukocytes are critical for local inflammatory responses to microbes. If inadequate quantities of TNF are present (e.g., in patients treated with drugs that block TNF or in TNF gene knockout mice), a consequence may be failure to contain infections.

In addition, TNF, IL-1, and IL-6 produced at inflammatory sites may enter the blood and be delivered to the bone marrow, where they enhance the production of neutrophils from bone marrow progenitors, usually acting in concert with colony-stimulating factors. In this way, these cytokines increase the supply of cells that can be recruited to the sites of infection.

Phagocytosis and Killing of Microbes by Activated Phagocytes

Neutrophils and macrophages that are recruited into sites of infections ingest microbes into vesicles by the process of phagocytosis and destroy these microbes (Fig. 4-12). Phagocytosis is an active, energy-dependent process of engulfment of large particles (>0.5 µm in diameter) into vesicles. Phagocytic vesicles fuse with lysosomes, where the ingested particles are destroyed, and in this way, the mechanisms of killing, which could potentially injure the phagocyte, are isolated from the rest of the cell.

FIGURE 4–12 Phagocytosis and intracellular destruction of microbes.

Microbes may be ingested by different membrane receptors of phagocytes; some directly bind microbes, and others bind opsonized microbes. (Note that the Mac-1 integrin binds microbes opsonized with complement proteins, not shown.) The microbes are internalized into phagosomes, which fuse with lysosomes to form phagolysosomes, where the microbes are killed by reactive oxygen and nitrogen species and proteolytic enzymes. iNOS, inducible nitric oxide synthase; NO, nitric oxide; ROS, reactive oxygen species.

Neutrophils and macrophages express receptors that specifically recognize microbes, and binding of microbes to these receptors is the first step in phagocytosis. Some of these receptors are pattern recognition receptors, including C-type lectins and scavenger receptors, which we discussed previously. Pattern recognition receptors can contribute to phagocytosis only of organisms that express particular molecular patterns, such as mannose for the mannose receptor. Phagocytes also have high-affinity receptors for certain opsonins, including antibody molecules, complement proteins, and plasma lectins; these receptors are critical for phagocytosis of many different microbes that are coated with the opsonins. One of the most efficient systems for opsonizing microbes is coating them with antibodies. Recall that antibody molecules have antigen-binding sites at one end, and the other end, the Fc region, interacts with effector cells and molecules of the innate immune system. There are several types of antibodies, which we will discuss in detail in Chapters 5 and 12. Phagocytes express high-affinity Fc receptors called FcγRI specific for one type of antibody called IgG (see Chapter 12). Thus, if an individual responds to an infection by making IgG antibodies against microbial antigens, the IgG molecules bind to these antigens, the Fc ends of the bound antibodies can interact with FcγRI on phagocytes, and the end result is efficient phagocytosis of the microbes. Because many different antibodies may be produced that bind to many different microbial products, antibody-mediated opsonization contributes to the phagocytosis of a broader range of microbes than do pattern recognition receptors. Antibody-dependent phagocytosis illustrates a link between innate and adaptive immunity—antibodies are a product of the adaptive immune system (B lymphocytes) that engage innate immune system effector cells (phagocytes) to perform their protective functions.

Once a microbe or particle binds to receptors on a phagocyte, the plasma membrane in the region of the receptors begins to redistribute and extends a cup-shaped projection around the microbe. When the protruding membrane cup extends beyond the diameter of the particle, the top of the cup closes over, or “zips up,” and pinches off the interior of the cup to form an inside-out intracellular vesicle (see Fig. 4-12). This vesicle, called a phagosome, contains the ingested foreign particle, and it breaks away from the plasma membrane. The cell surface receptors also deliver activating signals that stimulate the microbicidal activities of phagocytes. Phagocytosed microbes are destroyed, as described next; at the same time, peptides are generated from microbial proteins and presented to T lymphocytes to initiate adaptive immune responses (see Chapter 6).

Activated neutrophils and macrophages kill phagocytosed microbes by the action of microbicidal molecules in phagolysosomes (see Fig. 4-12). Several receptors that recognize microbes, including TLRs, G protein–coupled receptors, antibody Fc and complement C3 receptors, and receptors for cytokines, mainly IFN-γ, function cooperatively to activate phagocytes to kill ingested microbes. Fusion of phagocytic vacuoles (phagosomes) with lysosomes results in the formation of phagolysosomes, where most of the microbicidal mechanisms are concentrated. Three types of microbicidal mechanisms are known to be the most important.

When neutrophils and macrophages are strongly activated, they can injure normal host tissues by release of lysosomal enzymes, ROS, and nitric oxide. The microbicidal products of these cells do not distinguish between self tissues and microbes. As a result, if these products enter the extracellular environment, they are capable of causing tissue injury.

Other Functions of Activated Macrophages

In addition to killing phagocytosed microbes, macrophages serve many other functions in defense against infections (Fig. 4-13). Several of these functions are mediated by the cytokines the macrophages produce. We have already described how TNF, IL-1, and chemokines made by phagocytes enhance the inflammatory reactions to microbes and bring in more leukocytes and plasma proteins. Activated macrophages also produce growth factors for fibroblasts and endothelial cells that participate in the remodeling of tissues after infections and injury. The role of macrophages in cell-mediated immunity is described in Chapter 10.

FIGURE 4–13 Effector functions of macrophages.

Macrophages are activated by microbial products such as LPS and by NK cell–derived IFN-γ (described earlier in the chapter). The process of macrophage activation leads to the activation of transcription factors, the transcription of various genes, and the synthesis of proteins that mediate the functions of these cells. In adaptive cell-mediated immunity, macrophages are activated by stimuli from T lymphocytes (CD40 ligand and IFN-γ) and respond in essentially the same way (see Chapter 10, Fig. 10-7).

Systemic and Pathologic Consequences of the Acute Inflammatory Responses

TNF, IL-1, and IL-6 produced during the innate immune response to infection or tissue damage have systemic effects that contribute to host defense and are responsible for many of the clinical signs of infection and inflammatory disease (Fig. 4-14).

FIGURE 4–14 Local and systemic actions of cytokines in inflammation.

TNF, IL-1, and IL-6 have multiple local and systemic inflammatory effects. TNF and IL-1 act on leukocytes and endothelium to induce acute inflammation, and both cytokines induce the expression of IL-6 from leukocytes and other cell types. TNF, IL-1, and IL-6 mediate protective systemic effects of inflammation, including induction of fever, acute-phase protein synthesis by the liver, and increased production of leukocytes by the bone marrow. Systemic TNF can cause the pathologic abnormalities that lead to septic shock, including decreased cardiac function, thrombosis, capillary leak, and metabolic abnormalities due to insulin resistance.

In severe infections, TNF may be produced in large amounts and causes systemic clinical and pathologic abnormalities. If the stimulus for cytokine production is sufficiently strong, the quantity of TNF may be so large that it enters the blood stream and acts at distant sites as an endocrine hormone (see Fig. 4-14). The principal systemic actions of TNF are the following:

A complication of severe bacterial sepsis is a syndrome called septic shock, which may be caused by LPS released from gram-negative bacteria (in which case it is called endotoxin shock) or lipoteichoic acid from gram-positive bacteria. Septic shock is characterized by vascular collapse, disseminated intravascular coagulation, and metabolic disturbances. This syndrome is due to LPS- or lipoteichoic acid–induced TLR signaling leading to the production of TNF and other cytokines, including IL-12, IFN-γ, and IL-1. The concentration of serum TNF may be predictive of the outcome of severe bacterial infections. Septic shock can be reproduced in experimental animals by administration of LPS, lipoteichoic acid, or TNF. Antagonists of TNF can prevent mortality in the experimental models, but clinical trials with anti-TNF antibodies or with soluble TNF receptors have not shown benefit in patients with sepsis. The cause of this therapeutic failure is not known, but it may be because other cytokines elicit the same responses as TNF, an example of redundancy.

Acute inflammation may cause tissue injury because the effector mechanisms that phagocytes use to kill microbes are also highly toxic to host tissues. The proteolytic enzymes and reactive oxygen species produced by phagocytes that accumulate at a site of infection can injure host cells and degrade extracellular matrix if they are generated in large quantities, especially if the microbes resist being killed and continue to stimulate the innate immune responses. In fact, much of the pathology associated with infections is due to the inflammatory responses and not direct toxic effects of the microbes. Acute inflammation also causes tissue damage in the setting of autoimmune diseases, in which case neutrophils and macrophages accumulate and become activated secondarily to stimulation of the adaptive immune system by self antigens (see Chapter 14). As in inflammation induced by infections, TNF, IL-1, IL-6, and IL-12 are the key inducers of inflammation in autoimmune disease. Antagonists against TNF, IL-1, and IL-12 and antibodies against IL-6 receptors are in clinical use or in trials to reduce inflammation in patients with some of these diseases, such as rheumatoid arthritis, inflammatory bowel disease, and psoriasis.