Historical Perspectives

“Rubror et tumor cum calore et dolore,” or redness and swelling with heat and pain, were recognized as the four cardinal signs of inflammation by Cornelius Celsus (30 BCE to 38 CE). It was not established until the late 1800s to early 1900s that the body used both cellular and humoral components to identify and destroy microbes. Through the work of such pioneers as Elie Metchnikoff, Paul Ehrlich, and Almoth Roth, it was demonstrated that certain humoral factors called opsonins, later identified as antibodies and complement products, rendered bacteria more susceptible to ingestion and killing by phagocytic cells. Antibodies together with complement could also kill bacteria directly.

Traditionally, the immune system has been divided into innate and adaptive components. Clonal expansion of lymphocytes in response to infection is absolutely critical to the development of the immune response. However, it takes 3 to 5 days for clonal expansion to produce sufficient numbers of “effector” cells. Clearly, this is more than enough time for a pathogen to damage the host. The innate immune system is fundamental in eliminating the infection, and if not, then controlling it, until the adaptive immune responses eliminate it. If the innate and adaptive immune responses are “adequate,” the infection remains localized. If not, then the systemic response to infection, or “sepsis” results (see Chapters 90 and 103). What has become increasingly clear is that the adaptive and innate immune systems can each affect the functioning of the other.¹

In the past 25 years, new concepts that are fundamental to the care of critically ill patient have arisen directly from our understanding of the host’s response to infection. (1) Inflammatory reactions first characterized as a response to infection are the foundation of a number of other pathogenic mechanisms, such as ischemia/reperfusion injury, direct trauma, drug-induced injury, inhalational injury, and multiple-system organ dysfunction. (2) The response to infection may lead to further injury of the host. Alternatively, an exuberant antiinflammatory response may likewise be deleterious, resulting in immune suppression and fibrosis (see Chapter 104). (3) There is a direct interaction between the neuroendocrine axis and inflammation (see Chapter 102). (4) There is a fundamental interrelationship among endothelium, inflammation, and coagulation, and efforts to intervene in one will likely have effects in others (see Chapter 101).

Innate Immune Versus Adaptive Immune Response

The innate immune system is phylogenetically ancient. Innate immune recognition is genetically predetermined (i.e., by germline-encoded receptors). Thus these receptors have evolved by natural selection, and have defined specificities for infectious organisms. In contrast, in the adaptive immune system, the T-cell and B-cell receptors are somatically generated in a way that gives each lymphocyte a unique structural receptor. Because the T- and B-cell receptors are not encoded in a germline, they are not predetermined to recognize any particular antigen. No matter how useful these “receptors” become, they cannot be passed down to the next generation. Although there are potentially a large number of variants of germline-encoded receptors (perhaps in the hundreds), there are 10¹⁴ and 10¹⁸ different somatically generated immunoglobulin receptors and T-cell receptors, respectively.¹ Microbes are heterogenous and can mutate at very high rates. This can be handled appropriately by the adaptive immune system. This heterogeneity, however, represents more of a challenge for the innate immune system. As such, a different strategy, phylogenetically older than the adaptive system, evolved. The innate immune system has developed receptors that recognize “pathogen-associated molecular patterns” (PAMPs). PAMPs are highly conserved structures present in a large group of microorganisms. The common features of PAMPs are (1) PAMPs are only produced by microbial pathogens, not hosts; (2) structures recognized by the innate immune system are usually essential for survival or pathogenicity of the microorganism; and (3) PAMPs are usually invariant structures shared by an entire class of pathogen. The best known examples of PAMPs are bacterial lipopolysaccharide (LPS), peptidoglycan, lipoteichoic acid (LTA), mannans, bacterial DNA, double-stranded RNA, and glucans.²

Pattern Recognition Receptors

In the innate immune system, pathogen recognition molecules are called pattern recognition receptors (PRRs) and belong structurally to several families of proteins: leucine-rich repeat domains, calcium-dependent lectin domains, and scavenger receptor domains. Functionally, PRRs can be divided into three classes: secreted, endocytic, and signaling. Secreted PRRs function as “opsonins.” They bind to microbial cell walls, flagging them for recognition by the complement system and by phagocytes. Members of the secreted PRRS include C-reactive protein (CRP), mannose binding lectin, and the ficolins.^3-5 These molecules recognize sugar residues that are rich on microbial surfaces and can function directly as opsonins, promoting phagocytosis. Alternatively, they can also function indirectly by activating the classical complement pathway (CRP) or the lectin-dependent pathway (ficolin and mannose-binding lectin).⁴

Endocytic pattern-recognition receptors occur on phagocytes and can mediate the uptake and delivery of pathogens into lysosomes. Once in the lysosomes, the microbe can be destroyed. The macrophage mannose receptor recognizes carbohydrates with large numbers of mannoses (characteristic of microorganisms) and mediates their phagocytosis by macrophages. The macrophage scavenger receptor binds to bacterial cell walls and effectively clears them from the circulation. Complement receptor 3 (also known as Mac-1 and CD11b/CD18), in addition to recognizing microbes opsonized with complement, can function by binding mannose molecules directly, thus also functioning as an endocytic pattern-recognition receptor.² Dectin-1 present on macrophages is the endogenous receptor for β-glucan, a major cell wall component of budding yeast. Mutations in dectin 1 lead to mucocutaneous fungal disease, though fungal phagocytosis and killing is normal in affected individuals.⁶

Soluble Components of Immunity

Complement System

The complement system is critically positioned to participate in both the innate and adaptive immune response.²⁶ However, it is also critical for the disposal of immune complexes from tissues and clearance of apoptotic cells.⁴ The three pathways in which complement may become activated are the classic, alternative, and mannose-binding lectin pathway. Although complement activation initiates differently in each, all three converge at the cleavage of C3 (Figure 91-1). The classic pathway is initiated by the binding of the C1 complex (consisting of C1q, C1r, and C1s) to antibodies bound to antigen on the cell wall. The mannose-binding lectin pathway is initiated by binding of the complex of mannose-binding lectin and the mannose-binding lectin-associate proteases 1 and 2 (MASP1 and MASP2) to arrays of mannose groups on the surface of the bacterial cell wall. Both MASP2 and C1q then function similarly in forming the C4bC2a complex, representing the convertase for C3. Bacterial products, LPS, yeast cell wall particles, and aggregated antibody, including immunoglobulin (Ig)A and IgE, are all capable of activating the alternative pathway. The alternative pathway is initiated via low-grade cleavage of C3 in plasma to C3b. C3b binds to the hydroxyl groups on cell surface carbohydrates. C3b binds factor B to form a C3bB complex, which is activated by factor D, forming C3bBb and stabilized with properdin. C3bBb then functions as an alternative convertase for C3, that cleaves many molecules of C3 to C3b. C3b binds to hydroxyl groups on the microorganism around the area of complement activation. C3a, an anaphylatoxin, is released by the C3 convertase. C3b can also bind to the C3 convertase to form C5 convertase. The C5 convertase releases the anaphylatoxin C5a and initiates the formation of the membrane attack complex (MAC), C5b6789.⁴ The membrane attack complex created inserts into the cell membrane, creates large pores, and leads to osmotic lysis of the target.

Figure 91–1 The three activation pathways of complement: the mannose-binding lectin, the classic, and alternative pathways. The three pathways converge at the point of cleavage of C3. The mannose-binding lectin (MBL) pathway is initiated by binding of the complex of the MBL and MBL-associated proteases 1 and 2 (MASP1, MASP2) to arrays of mannose on the bacterial cell wall. MASP2 then activates first C4, then C2 to form the C3 convertase, C4b2a. The classic pathway is initiated by the binding of the C1 complex (which consists of C1q, two molecules of C1r, and two molecules of C1s) to antibodies bound to antigen on the surface of the bacterial cell wall. C1s, similar to MASP2, then activates C4 followed by C2 to form C4b2a (the C3 convertase). The alternative pathway is initiated by covalent binding of a small amount of C3b to hydroxyl groups on the cell surface carbohydrates and is activated by low-grade C3 in plasma. This C3b binds factor B to form C3bB, which is then activated by factor D to then form C3bBb, the alternative pathways C3 convertase. Properdin stabilizes this activation step. The C3 convertase then cleaves many molecules of C3 to C3b and C3a (an anaphylatoxin). C3b binds covalently around the site of complement activation, some of which binds to the C4b and C3b of the classic and alternative pathway C3 convertase, respectively, to form the C5 convertase. The C5 convertase cleaves C5 to C5a (an anaphylatoxin) and C5b, which initiates the formation of the membrane attack complex (MAC).

(Modified from Walport MJ: Complement. First of two parts, N Engl J Med 344:1058, 2001.)

Thus the complement system amplifies the initial response to the organism. In addition to lysing the target organism, opsonization with complement fragments C3b/C4b and C3bi (fragment C3b) occurs, resulting in phagocytosis by neutrophils, monocytes, and macrophages. Binding occurs through specific complement receptors, CR1 for C3b/C4b (CD35) and CR3 for C3bi (CD11b/CD18, also known as Mac-1 and α_Mβ2). The anaphylatoxins C3a, C4a, and C5a produced are low molecular weight, biologically active peptides defined by their actions on small blood vessels, smooth muscle, mast cells, and peripheral blood leukocytes. Blood vessels, smooth muscle cells, basophils, and mast cells respond to all three anaphylatoxins. Neutrophils, monocytes, and macrophages respond to C5a. The anaphylatoxins promote edema and increase vascular permeability through the release of histamine from mast cells and through the local production of vasodilatory prostaglandins such as prostaglandin E₂ (PGE₂) and edemogenic leukotrienes (LT) C₄, D₄, and E₄. C5a is a powerful activator of granulocyte function including chemotaxis, degranulation, and increased oxidative metabolism. C5a actions occur through its seven-transmembrane–spanning, G protein–coupled receptor, C5aR.²⁷ In animal models, C5a/C5aR is critical to the development of sepsis and multiple organ failure, and potentiates many early response cytokines and coagulation. C5aR is present not only on leukocytes, but also other tissue, including the brain, kidney, gastrointestinal tract.^27,28

Complement activation also occurs after oxidative stress such as occurs with ischemia/reperfusion. Complement activation is an early event after injury and the inhibition of complement activation or its components offer tissue protection after reperfusion.²⁹ Finally, complement proteins transduce various cell signals. Complement can activate B and T cells. It can regulate apoptosis of various cell types (see Chapter 100).^20,26

Immunoglobulin

The different immunoglobulins (Igs) secreted by B cells (IgG, IgA, IgM, IgD, and IgE) are known as isotopes. IgD plays little role, if any, in containment of microorganisms. Ig isotypes can be divided into subclasses, variants that show slight structural differences but are sufficiently alike structurally to be essentially identical to other members of the isotype class. There are four IgG subclasses: IgG1, IgG2, IgG3, and IgG4. As seen in Table 91-1, isotype subclasses have specialized roles in the immune response. For example, IgG2 has a major role in the formation of carbohydrate antibodies, but has poor complement fixation characteristics. The antibody system consists of Ig present in serum, protecting the blood and tissue spaces, and that present in the secretory system, lining the gastrointestinal and respiratory tracts and present in tears. The serum component is mostly IgG (85%), with lesser amounts of IgA and IgM. The secretory system consists mostly of secretory IgA (85%), which is structurally different than serum IgA, and lesser amounts of IgG and IgM. All Igs have a basic four-chain structure composed of an identical pair of heavy (H) and light (L) chains. The H-L pairs are held together by interchain disulfide bonds and noncovalent forces. The binding site for antigen is formed by one H and one L chain. IgM is a polymer of five four-chain units, and secretory IgA is a dimer. Polymeric forms (Figure 91-2) possess a J chain that is synthesized with the H and L chains and serves to stabilize the sulfhydryl groups during polymerization. At regular intervals along the peptide chain a disulfide bond forms an intrachain loop, known as the Ig domain. This motif is repeated among immunoglobulins, T-cell receptors, adhesion proteins, and histocompatibility antigens. Proteins with these Ig domains share close homology structurally and functionally and suggest a common evolutionary origin. They are termed members of the Ig superfamily (IgSF), and all are involved in cell interaction processes associated with recognition.

Table 91–1 Biologic Properties of Ig Classes and IgG Subclasses

Figure 91–2 Basic immunoglobulin (Ig) structure. Sites of cleavage for papain or pepsin are shown with resulting fragments. The intrachain disulfide bond forms the region of the Ig domain. Insets show polymeric immunoglobulins, secretory IgA (left) and IgM (right). J chain is common to all polymeric forms; SC is the secretory component. V_L, Variable domain of light chain; C_L, constant domain of light chain; V_H, variable domain of heavy chain; C_H, constant domains of heavy chain.

The Ig molecule is divided into regions, the amino acid sequence of which is similar, such as regions needed for complement fixation or attachment to receptors on leukocytes. However, other regions are highly variable; those that bind to antigen have the highest divergency and are collectively known as the hypervariable region. Thus each Ig chain can be divided into constant and variable regions. The H chain has three constant regions and one variable region; the L chain has one variable region and one constant region. The hypervariable region of the L and H chain is tightly apposed and forms the combining site for antigen (see Figure 91-2). The digestion of Ig with papain and pepsin generates fragments with varying biologic capability. The Fc region of the Ig molecule accounts for its isotypic biologic capability. This is the region critical for complement fixation and recognition by Fc receptors on the leukocytes. The Fab region provides for specific unique antigen-antibody interactions. The F(ab′)2 fragment is formed with pepsin cleavage; the affinity for antigen is twice as great as Fab alone.

Contact Activation System

There are four major plasma protein systems that contribute to the host’s defense and participate in the development of inflammatory tissue injury: the complement system (previously discussed), the contact activation system (also known as Hageman factor or intrinsic coagulation system), the extrinsic coagulation cascade, and the fibrinolytic system (see Chapter 80). The contact activation system is critical to host defense and control of local blood flow at sites of injury. Hageman factor (Factor XII) is activated spontaneously (XIIa) on contact with negatively charged surfaces, such as lipid A of LPS and vascular basement membranes (Figure 91-3). High-molecular-weight kininogen (HMWK), prekallikrein, and factor XI circulate in the plasma as complexes. Factor XIIa will activate factor XI and cleave prekallikrein to kallikrein. Kallikrein will then cleave HMWK to bradykinin. The kallikrein-kinin system also encompasses the tissue (or glandular) kallikrein-kinin system (see Figure 91-3). Tissue kallikrein is immunologically distinct from plasma kallikrein and present throughout the body as an inactive “pro-” substance. In the presence of intracellular enzymes, plasmin, or plasma kallikrein, tissue kallikrein is produced, secreted, and active in the tissue where it is made. Tissue kallikrein can then cleave HMWK to bradykinin directly, or it can cleave low-molecular-weight kininogen (LMWK) or tissue kininogen (T-kininogen) to kallidin that is then directly converted to bradykinin. In the plasma, 80% of kininogen is low molecular weight.

Figure 91–3 Contact activation system and prostanoid production. Bradykinin can be produced by either the plasma kallikrein (top left) or tissue (glandular) kallikrein system (bottom left). Once formed, bradykinin can activate phospholipase A₂, and thus drive prostanoid production. In addition, it can potentiate the affects of the prostaglandins and vice versa. Bradykinin is metabolized into inactive peptides via several steps, the last of which is enzymatically driven by angiotensin-converting enzyme (ACE). HMW, High molecular weight; LMW, low molecular weight; PAF, platelet activating factor; COX-1, cyclooxgenase-1; COX-2, cyclooxygenase-2; 5LOX, 5-lipoxygenase; LT, leukotriene; TXA₂, thromboxane A₂; PG, prostaglandin.

(Modified from Ueno A, Oh-ishi S: Roles for the kallikrein-kinin system in inflammatory exudation and pain: lessons from studies on kininogen-deficient rats, J Pharmacol Sci 93:1, 2003.)

Bradykinin is an exceedingly potent vasoactive peptide.³¹ It can cause venous dilation, increased vascular permeability, hypotension, bronchoconstriction, and activation of phospholipase A₂. Phospholipase A₂ releases arachidonic acid from cell membrane and initiates the production of both proinflammatory and antiinflammatory phospholipids-derived products (see subsequent section and Figure 91-3). Bradykinin is metabolized by angiotensin-converting enzyme (ACE) to inactive peptides. Bradykinin, along with prostanoids, stimulates the pain response through polymodal receptors and C-fibers (capsaicin sensitive).³² Bradykinin has been shown to be responsible for the four signs of inflammation: heat, redness, swelling, and pain. The bradykinin effect is enhanced by simultaneous production of prostanoids. Prostanoids, particularly along with the low pH of exudates, inhibit the activity of kininases such as ACE.

Cytokines

Cytokines are signaling proteins secreted by cells that affect the functional properties of other cells of the same organism. The cytokine family includes lymphokines, chemokines, interleukins, and interferons. Unlike circulating hormones, cytokines travel short extracellular distances before interacting with target cell surface receptors in a paracrine or autocrine manner. Cytokines can be detected in serum samples, particularly during times of maximal production, as occurs in sepsis. Cytokines as a group are low-molecular-weight (<80 kDa) proteins. They interact with high-affinity cell surface receptors specific for each cytokine. Their cell surface binding ultimately leads to changes in the pattern of protein synthesis and/or altered cell behavior. They often have multiple-overlapping cell regulatory functions. Many cytokines are produced early in infection, whereas others are produced at later stages.

Interleukin-1 and Tumor Necrosis Factor

IL-1 is a phylogenetically old molecule that predates the evolution of lymphocytes and immunoglobulin. Its activity extends beyond immune function. IL-1 is produced by a wide variety of cells, including macrophages, endothelial cells, epithelial cells, and vascular smooth muscle cells. There are two separate forms of IL-1, IL-1α, and IL-1β. In contrast, TNF-α is produced by cells primarily of the innate immune system, including monocytes/macrophages, NK cells, mast cells, and neutrophils under specific conditions. TNF is also produced by other cell types under conditions of stress. For example, TNF is produced by cardiac myocytes and is implicated in both acute and chronic congestive heart failure as well as in the cardiomyopathy associated with sepsis. TNF-β (also known as lymphokine) is produced by T lymphocytes, but occasionally antigen-activated T cells may also produce TNF-α. Both TNF and IL-1 are produced as small precursor molecules or “pro” molecules that are cleaved by IL-1β converting enzyme in the inflammasome (by caspase 1) and TNF-α converting enzyme, ADAM17 (a disintegrin and metalloproteinase), respectively. Once cleaved, these proteins are then excreted. It should be noted that caspase-1 (in the inflammasome) and ADAM17 are multifunctional proteinases and have activity on other interleukins and inflammatory molecules. IL-1 is the only cytokine with a natural inhibitor, IL-1 receptor antagonist (IL-1RA) produced by the same cells that produce IL-1. IL-1RA functions to downregulate the proinflammatory effects of IL-1. IL-1 and TNF-α are the early major mediators of gram-negative endotoxin shock (see Chapter 103). As outlined earlier in this chapter, signaling of cells by TNF and IL-1 occurs at least in part through NF-κB, and consequently share a similarity in receptor function and signaling molecules.

For routine infection and injury, IL-1β and TNF-α are transiently expressed and secreted. Their activities are modulated by coproduction of naturally occurring antiinflammatory cytokines, such as IL-10 and IL-1RA. The levels of these cytokines fall off rapidly; thus production is tightly regulated. Dysregulated cytokine production, such as occurs in chronic diseases such as rheumatoid arthritis and inflammatory bowel disease, has lead to the development of anticytokine therapy such as human recombinant IL-1RA (Anakinra), human/murine chimeric monoclonal antibody against TNF-α (Infliximab), and recombinant fusion protein composed of the extracellular binding domain of TNF receptor II and human IgG1 (Etanercept). With the development of these products, there is increasing evidence that neutralization of TNF-α is associated with increased risk of opportunistic infections, including mycobacterial diseases. Blockade of IL-1 using IL-1RA appears at present to be safe.³⁵ Using such blocking agents in animal models it appears that neutralization or gene deletion of TNF-α is associated with reduction of host defense in models of live gram-positive or gram-negative infections as well as infection by intracellular microbes such as Salmonella and Listeria. Absence of IL-1RA can also result in decreased resistance to Listeria or gram-positive bacteria. TNF and IFN-γ (discussed in the following paragraphs) are required for defense against infection caused by Mycobacterium tuberculosis.³⁵

Il-17 is produced by an effector helper T cell, Th17, as well as natural killer (NK) cells and NK-T cells (combined properties of both NK and T cells). Th17-driven inflammation is typified by neutrophils predominating the inflammatory response. A number of pathogens induce Th17 response, including Citrobacter, Klebsiella pneumonia, Mycobacterial tuberculosis, and fungi such as Candida albicans.³⁶ To eliminate the body of fungi and certain extracellular pathogens requires inflammation driven by Th17. Most cell types have IL-17 receptors, and signaling through this receptor produces IL-6, TNF, IL-8, antimicrobial peptides, and matrix metalloproteinases. IL-17 is also implicated in several inflammatory conditions, including rheumatoid arthritis, psoriasis, multiple sclerosis, and inflammatory bowel disease.³⁶

IL-18 is a member of the IL-1 family of ligands. It has unique characteristics in that only in conjunction with IL-12 is IFN-γ produced by activated T cells and by LPS-stimulated macrophages. IL-18 is implicated in the development of endotoxic shock and in the myocardial depression that occurs in these models.³⁷ As a key mediator of IFN-γ production, IL-18 serves an important role in controlling infections from Salmonella, Cryptococcus, toxoplasma, Candida, and Mycobacterium organisms, often through its modulation of the production of nitric oxide.^37,38

IL-12 is produced by monocytes, macrophages, dendritic cells, neutrophils, and to a lesser extent B cells. IL-12 effects are primarily on T and NK cells. The responses by T cells and NK cells include increased proliferation, increased IFN-γ production, increased cytotoxic activity (cytotoxic T lymphocytes and NK cells), and for T cells, polarization toward a Th1 phenotype (see following discussion). Patients with defects in IL-12 or IL-12 receptor have increased susceptibility to mycobacterial and salmonella infections. Neonates have diminished IL-18 and IL-12 production contributing to inadequate IFN-γ production and increased susceptibility to infections.^39,40

IL-6 is one member of the IL-6 family of cytokines. It is produced by many cell types including the cardiac and skeletal myocytes, but mainly by macrophages and monocytes, adipose cells, endothelial cells, T lymphocytes, mast cells, and osteoclasts. Many cell types respond to IL-6, including hepatocytes. IL-6 is the factor most directly responsible for the production of acute phase reactants. Its production is somewhat delayed compared to IL-1 and TNF, IL-6 levels remain elevated longer than the other two. IL-6 deficient animals have an increased susceptibility to Listeria monocytogenes, Streptococcus pneumoniae, Escherichia coli, Candida albicans, and mycobacterial infections.^35,41

IL-10, like IL-6, is a pleiotropic cytokine that exerts both immunosuppressive and immunostimulatory effects. It is produced primarily by CD4⁺ T cells, B cells, and other cell types. Its predominant suppressive effect is to inhibit Th1 (IFN-γ and IL-2) cytokine production. It downregulates MHC II expression, thus limiting its interaction with T cells and NK cells. It also inhibits cytokines involved in inflammatory responses, including prostaglandin E₂, TNF-α, IL-1, IL-6, and IL-8. IL-10 induces the production of IL-1RA. IL-10 can upregulate FCRγI on monocytes, macrophages, and NK cells, enhancing antibody-mediated cellular cytotoxicity.

IL-2 is produced by CD4 T cells when they encounter foreign peptide-MHC complex on antigen-presenting cells. This cytokine is critical for adaptive immune response in that it has both an autocrine and paracrine function, triggering T cells to undergo multiple rounds of proliferation and differentiate into effector T cells.

Interferons and Other Soluble Products

IFN-α and IFN-β are known as type I interferons. IFN-α is produced by monocytes and macrophages, whereas IFN-β is produced by fibroblasts and other cell types. The major stimuli for type I interferon are viral infections; they also respond to T-cell–derived factors in adaptive immune responses. Type I IFNs inhibit viral replication, and patients with insufficient production suffer from severe progressive or fulminant viral disease. These IFNs inhibit cell proliferation, enhance the lytic potential of NK cells, and increase class I HLA while decreasing class II HLA. IFN-γ is a type II IFN and is produced primarily by CD4⁺, CD8⁺, and NK cells (Figure 91-4). IFN-γ has antiviral and antiproliferative activity. It upregulates class I and II HLA expression, thus enhancing cellular toxicity and antigen presentation, respectively. IFN-γ activates monocytes and macrophages as well as neutrophils, resulting in enhanced killing of intracellular organisms, including mycobacteria and listeria. In addition, IFN-γ induces inducible nitric oxide synthase in macrophages, resulting in the generation of nitric oxide, a critical component for bactericidal function. Animals with a deficiency of IFN-γ have decreased survival in response to salmonella and mycobacterial infections. Patients with complete loss of IFNγ receptors have severe, early life infections with salmonella and viral infections (including respiratory syncytial virus, parainfluenza, herpes simplex virus, and cytomegalovirus) with a high mortality rate.^35,40,42 IFN-γ is a critical cytokine at the interface of the adaptive and innate immune system because of its function on NK cells and ultimately monocytes/macrophages.

Figure 91–4 Secreted cytokines and effects on adaptive (T cell) and innate (dendritic cell/mononuclear phagocyte) and NK cells. NK cells activate mononuclear phagocytes and dendritic cells via IFN-γ secretion. With activation, mononuclear phagocytes demonstrate increased phagocytic and antimicrobial actions against intracellular pathogens. The phagocytes secrete cytokines which activate the NK cells, inducing IFN-γ production. IFN-γ from NK cells, and IFN αβ, IL-12, IL-18 from the mononuclear phagocytes/dendritic cells in turn activate the T cell along the TH-1 pathway. Activation of the T cell produces IL-2 and IFN-γ, which affects the mononuclear phagocyte/dendritic cell, NK cell, other T cells. Activation of the mononuclear phagocyte/dendritic cell results in production of TNF that has both autocrine and paracrine effects.

(Modified from Douglas S, Kapur R, Merrill JD, et al: In Stiehm ER, Ochs HD, Winkelstein JA, editors: Immunologic disorders of infants and children, Philadelphia, 2004, Saunders Elsevier.

Macrophage migration inhibitory factor (MIF) has been identified for more than 40 years, though its function has only been well defined in the past 15 years with the cloning of human MIF cDNA. Although T cells were initially believed to be the main source, monocytes, eosinophils, basophils, dendritic cells, B cells, mast cells, and neutrophils all express MIF. In contrast to other cytokines, MIF is secreted and stored in intracellular pools and therefore does not require de novo protein synthesis before secretion. MIF has a very broad tissue distribution and is expressed by cells and tissues that are in direct contact with the host’s natural environment as well as organs involved in the stress response (hypothalamus, pituitary, and adrenal glands). MIF is implicated in both gram-negative and gram-positive infections. MIF-deficient animals have increased susceptibility to low-dose inoculum of salmonella and Escherichia coli. In contrast, blockade of MIF results in improved survival in animals treated with high doses of E. coli, after cecal ligation and puncture, or after bacterial superantigen challenge. Recent work supports that MIF also has isomerase activity and contributes to chronic inflammation seen in inflammatory bowel disease and cancer.^43,44

High mobility group box 1 (HMGB1) has been identified as a late mediator of sepsis. Like MIF, antibodies to HMGB1 can be given many hours after the induction of sepsis and improve survival in animal models.^43,45 Also similar to MIF, HMGB1 is released from the cytoplasmic pool. Intracellular HMGB1 has diverse functions, including nucleosomal structure and function, as well as binding of transcription factors to their cognate DNA sequences. However, in systemic concentrations it has diverse proinflammatory responses mirroring the “late” effects of systemic inflammation and as such is often grouped with DAMPS (reviewed elsewhere⁴⁵). Macrophages and neutrophils, when stimulated with LPS and C5a, respectively, release large amounts of HMGB1 into the culture medium. Systemic HMGB1 accumulation occurs in mice 8 hours after LPS administration, long after TNF and IL-1β levels have decreased. Humans with sepsis have elevated levels of HMGB1, and those whose levels were most elevated were at highest risk of dying.⁴⁶ In a mouse model of cecal ligation and puncture in which mortality rate was 75%, treatment with anti-HMGB1 monoclonal antibody 24 hours after injury decreased the mortality rate to 25%. It had no effect on recovery of bacterial counts from the spleens of these animals and thus did not appear to affect bacterial clearance.⁴⁷ A derivative of a Chinese herb, tanshinone IIA TSN IIA-SS, at concentrations (100 μmol/L) completely abrogated LPS-induced HMGB1 release.⁴⁸ This needs validation in other animal studies.

Chemokines are structurally and functionally related inflammatory cytokines with the ability to stimulate the chemotactic migration of distinct sets of cells, including neutrophils, monocytes, lymphocytes, dendritic cells, macrophages, fibroblasts, stem cells, and smooth muscle cells. The chemokine family is the largest family of cytokines, and although their main function is characterized as “chemotaxis,” or directing migration through a concentration gradient, they also have a number of other functions, including cell activation, signaling, effects on angiogenesis and tumorigenesis, as well as immune cell polarization. To date there are 40+ identified chemokines.⁴⁹ They are small (∼8 to 14 kDa), mostly basic molecules. They function through unique receptors that are G protein coupled and are seven membrane spanning (i.e., have seven transmembrane domains).⁵⁰ The cytoplasmic domains of the receptors are critical for cell signaling and function. Chemokines may use more than one receptor for function. Chemokines are central to the process of extravasation of leukocytes that includes multiple steps involving interactions of adhesion molecules and the chemoattractant function of these proteins.⁵¹ Chemokines are defined by structure, not function, and can be divided into two large and two small subgroups depending on the number and arrangement of conserved cysteines. The subgroups are CC, CXC, C, and CX3C. It should be appreciated that in the past 10 years, chemokines have been designated by their subgroup followed by ligand number; for example, IL-8 is CXCL8. Most chemokines are classified into two main groups according to function: (1) those concerned with hemostasis that are constitutively expressed and coordinate leukocyte trafficking during hematopoiesis and those with lymphocyte recirculation, and (2) those concerned with inflammation and tissue injury. Chemokines are produced by hematopoietic cells themselves as well as by endothelial cells, epithelial cells, and cells arising from the mesoderm, including fibroblasts, myocytes, hepatocytes, and lymphatic cells. Table 91-2 lists a select group of cytokines involved in infection and inflammation.

Table 91–2 Chemokine/Chemokine Receptor Families (Partial List)

Name	Original Ligand Name∗	Chemokine Receptor
CXCL1	GRO-α/MGSA-α	CXCR2>CXCR1
CXCL2	GRO-β/MGSA-β	CXCR2
CXCL3	GRO-γβ/MGSA-γ	CXCR2
CXCL5	ENA-78	CXCR2
CXCL7	NAP-2	CXCR2
CXCL8	IL-8	CXCR1, CXCR2
CXCL9	Mig	CXCR3
CXCL10	IP-10	CXCR3
CXCL12	SDF-1	CXCR4
C CHEMOKINE/RECEPTOR FAMILY
XCL-1	Lymphotactin	XCR1
CX3C CHEMOKINE/RECEPTOR FAMILY
CX3CL1	Fractalkine	CX3CR1
CC CHEMOKINE/RECEPTOR FAMILY
CCL2	MCP-1	CCR2
CCL3	MIP-1α	CCR1,CCR5
CCL4	MIP-1β	CCR5
CCL5	RANTES	CCR1, CCR3, CCR5
CCL8	MCP-2	CCR3
CCL11	Eotaxin	CCR3
CCL19	MIP-3β	CCR7
CCL21	6Ckine	CCR7
CHEMOKINE RECEPTORS AND CELLULAR DISTRIBUTION
XCR1	Lymphotactin	T,B, NK
CXCR1	IL-8, GRO-α	N,M, T, NK, En, Ms, Bs
CXCR2	IL-8, GRO-α,-β-γ, NAP-2, ENA-78	N,M,T,NK, Ms, As, Nn, En
CXCR3	IP-10, Mig	Activated T
CXCR4	SDF-1	Myeloid, T, B, Ep, En, DC
CX3CR1	Fractalkine	NK, M, T
CCR1	RANTES, MIP-1α, MCP-2, MCP-3	N, M, T, NK, B, Ms, As, Nn
CCR2	MCP-1	M, T, B, Bs
CCR3	RANTES, eotaxin	Eo, Bs, T
CCR5	RANTES, MIP-1α, MIP-1β, MCP-2	T, M, Mφ
CCR7	MIP-3β, 6Ckine	T, B, DC

GRO, Growth regulating peptide; MGSA, melanocyte growth stimulating activity; ENA, epithelial-derived neutrophil attractant; NAP, neutrophil activating peptide; IL-8, interleukin-8; IP-10, γ-interferon-induced peptide 10; SDF, stroma derived factor; MCP, monocytes chemotactic peptide; MIP, macrophage inflammatory peptide; RANTES, regulated on activation, normal T cell expressed and secreted; T, T cell ; B, B cell; NK, natural killer cell; M, monocyte/macrophage; N, neutrophil; Ms, mast cell; Bs, basophil; As, astrocyte; Nn, neuron; En, endothelium, Eo, eosinophils; DC, dendritic cell; Ep, epithelial cell; Mφ macrophage.

∗ All ligand names are human chemokines.

Modified from Murdoch C, Finn A: Chemokine receptors and their role in inflammation and infectious diseases, Blood 95:3032, 2000; and Zlotnik A, Yoshie O: Chemokines: a new classification system and their role in immunity, Immunity 12:121, 2000.

Granulocyte colony stimulating factor (G-CSF) and granulocyte-macrophage stimulating factor (GM-CSF) were initially identified by their ability to induce granulocytopoiesis and monocytopoiesis; however, they have marked effects on neutrophil and monocyte function. Although both G-CSF and GM-CSF are produced by bone marrow stromal cells, they are also produced by activated monocytes/macrophages and fibroblasts. G-CSF is also produced by epithelial cells of the gut and lung in response to inflammation. GM-CSF is produced by T lymphocytes and NK cells. G-CSF enhances the physiologic activation of mature neutrophils, whereas GM-CSF stimulates the functional activity of neutrophils, eosinophils, and monocytes/macrophages.^52,53 GM-CSF is critically involved in the normal surfactant turnover, and alveolar proteinosis is an autoimmune disease targeting GM-CSF.⁵⁴ Critical to the immune response is that production of G-CSF and GM-CSF results in increased neutrophil survival. Neutrophils usually have a half-life of less than 12 hours before removal from the circulation and undergo apoptosis. However, exposure to G-CSF or GM-CSF decreases apoptosis, leading to prolonged survival of circulating neutrophils and those that have moved to a site of infection. Removal of these cytokines, such as occurs in the resolution phase of inflammation, leads to induction of apoptosis and increased clearance of neutrophils. G-CSF and GM-CSF are approved for use in a variety of hematologic pathologies.⁵³ G-CSF and GM-CSF have been proposed for use in the nonneutropenic critically ill adult and neonatal population. However, studies to date do not support their routine use as either a treatment of established systemic infection or as prophylaxis to prevent systemic infection in high-risk individuals.^55-57

Nitric oxide (NO) is a stable, free radical gas. Extensive work over the past 20 years has converged to establish NO as a major messenger molecule regulating immune function and blood vessel dilation as well as a neurotransmitter. NO is formed from arginine by the enzyme nitric oxide synthase (NOS). NOS-2 or inducible NOS (iNOS) is present in many tissues, whereas NOS-1 and NOS-3 are primarily present in neuronal and endothelial cells, respectively. NOS-1 and NOS-3 are present in low amounts and generate NO for neurotransmission and vasodilation. In contrast, NOS-2 is induced by microbial peptides and inflammatory cytokines and serves as a major bactericidal and tumoricidal agent. NO is critical in adaptive immunity to intracellular pathogens such as Mycobacterium tuberculosis and Listeria monocytogenes. NO enhances the activity of NK cells, γδ T cells, and macrophages. The condensation of NO and the reactive oxygen metabolite superoxide, O₂⁻, results in the production of peroxynitrite (OHNOO⁻), which decays to the highly reactive hydroxyl radical (OH) and nitrogen dioxide. These agents contribute in part to the killing of the microorganisms in tightly regulated structurally “isolated” areas of phagocytes, the phagolysosomes, discussed below.

Cellular Components of Immunity

The cellular components of immunity have traditionally been divided into innate and adaptive immunity, but such distinctions have become increasingly blurred and critical overlap occurs. For example, the dendritic cells and other antigen presenting cells, such as macrophages and monocytes, are part of the innate immune system; nonetheless, they directly drive adaptive immunity.

The cells of the immune system can be defined by the surface antigens they display. These surface antigens are denoted by the CD nomenclature and refer to cluster designation. These antigens denote the lineage and often the functional capacity of a cell. Surface antigens are revealed by using monoclonal antibodies, commonly with flow cytometry.

Lymphocyte types include the T cells, B cells, NK cells, and the NK T cell (NKT), which exhibits features of both NK and T cells. T cells got their name because the vast majority arise from the thymus. They mediate antigen-specific cellular immunity and play a critical role in facilitating antigen-specific B cell–dependent humoral immunity. B cells are the subset of lymphocytes that synthesize, express Ig on their surface, and differentiate to plasma cells that produce Ig. B cells arise from the bone marrow in humans and other mammals or bursa in birds.⁵⁸

The major T-cell subsets are the CD4⁺ helper/inducer and the CD8⁺ suppressor/killer cells. Nearly all T cells bear a T-cell receptor composed of an α- and β-chain; they also express CD4 or CD8 coreceptors. Nearly all the αβ T cells recognize protein antigen in the form of peptide fragments bound to classic MHC molecules (MHC class I or MHC class II). The CD4 molecule augments binding to antigens presented in association with MHC II antigens, whereas CD 8 molecules are necessary for antigen binding to MHC I. The immunologic synapse is the highly ordered junction that forms between the APC, such as the dendritic cells or tissue macrophages, and the T cell during antigenic stimulation. The structure resembles a doughnut in which the T-cell receptor-peptide-MHC, CD3, CD4, or CD8 is in the center of the synapse. The costimulatory molecules CD2, CD28, CD54 (intercellular adhesion molecule-1 [ICAM-1]) on the T cell bind to their respective ligands LFA-3 (CD 58), CD80-CD86, LFA-1 (CD11a/CD18) on the APC on the perimeter of synapse (Figure 91-5).⁵⁹

Figure 91–5 Immunologic synapse. Protein antigen bound to the major histocompatibility complex I or II (MCH I or II) is presented by the antigen presenting cell (APC) to the T-cell receptor. The T-cell receptor/CD3 complex is composed of the two chains of the T cell receptor (α and β) and the 6 subunits of the CD3(a γ and a δ subunit and two each of the ε and ζ subunits). CD8 or CD4 interacts with the MHC I or MHC II/antigen complex, respectively. Engagement of the T-cell receptor can then signal the T cell to activate a number of activities, including transcription. However, for activation to occur, there must be additional signals through costimulatory molecules on the T cell and APC. The T-cell receptor/CD3/CD4 (or CD8) complex is centered in the middle of the immunologic synapse, while the costimulatory molecules are in the periphery.

The CD4⁺ naive cells, which are those never exposed to antigen, can differentiate into effector cells expressing specific patterns of cytokines. Originally designated as Th1 and Th2 based on the cytokines they produce, there has been an explosion in the past 15 years delineating additional subtypes. Most of these cytokines produced by T cells are secreted, but some can be expressed on the cell surface. IL-12 and IFN-γ drive T cells into the Th1 pathway. IFN-γ is the signature cytokine produced by Th1 cells, but the Th1 cells also produce substantial amounts of IL-2, TNF-α, and TNF-β. The Th1 response is considered proinflammatory. IL-4 drives T cells into the Th2 pathway. Th2 cells also produce large amounts of IL-4 along with IL-5, IL-9, and IL-13. IL-4 and IL-13 (along with IL-10 produced by monocytes/macrophages) are considered antiinflammatory or immunosuppressive. The Th2 response is critical for eosinophil function and is important in the development of IgE responses and the killing of parasites.

Th17 cells, which produce IL-17, result from activation of naive T cells by TGF-β plus an inflammatory cytokine (IL-21 alone, IL-6 + IL-23, or IL6 + IL-21).³⁶ IL-23 is produced by monocytes and dendritic cells. IL-21 is produced by Th17 cells and thus serves an amplification loop for Th17 differentiation. Important to highlight is that TGF-β is considered a prototypic antiinflammatory cytokine critical for the healing. It also serves as a warning of categorizing a particular molecule as either “pro” or “anti” inflammatory solely, which is often proven to be too simplistic as our understanding of immunity progresses. Th17 is critical for the control of extracellular cytokines, induces the destruction of matrix, and often synergizes with TNF and IL-1.³⁶

TGF-β is produced by many cells. In the absence of IL-6, it will induce CD4⁺ cells into a regulatory T cell, known as a T-reg. T-reg cells are identified by a transcription factor, Foxp3+ (Forkhead box P3 transcription factor) and carry the IL-2 receptor CD25. Presence of Foxp3+ is implicated in the development of autoimmunity, allergy, and rejection in transplant medicine and suppression of immune responses to cancer.⁶⁰ There are several different types of T-regs. All require cell-to-cell contact for immune suppression.

Naive CD8⁺ cells are not effective killer cells. However, after activation with antigen in the context of MHC class I by APCs in the presence of IL-2 and IL-12, they differentiate quickly into CD8⁺ cytotoxic cells. These cells express perforin, granzymes, and Fas ligand and produce effector cytokines including TNF-α and IFN-γ. Perforins introduce pores into the target cell through which granzymes can enter, leading to the triggering of apoptosis and cell death. Alternatively, the cytotoxic T cell upregulates the Fas ligand (CD95L) that engages Fas (CD95) on the target cell, resulting in delivery of death signal culminating in apoptosis (Figure 91-6).

Figure 91–6 Mechanisms of antigen-specific MHC I–restricted T cell–mediated cytotoxicity. The engagement of the αβ T-cell receptor (TCR) in the CD3/CD8 complex of the T cell by antigenic peptide bound to MHC I on the target cell leads to T-cell activation and target cell death. A, Cytotoxicity occurs via the release of the contents of the cytotoxic granules from the T cell, including perforin and granzyme. A perforin pore is introduced into the target cell membrane by which granzymes can enter the target cell, leading to the triggering of apoptosis and cell death. B, Activation of the T cells leads to upregulation of the Fas-ligand (CD95L), which engages Fas (CD95) on the target cell, resulting in the delivery of a death signal culminating in apoptosis. Both of these mechanisms are also used by NK (natural killer) T cells and NK cells.

(Modified from Lewis DB, Wenwei T: The physiologic immunodeficiency of immaturity. Stiehm ER, Ochs HD, Winkelstein JA, editors: Immunologic disorders of infants and children, Philadelphia, 2004, Saunders Elsevier.)

A small proportion of T cells in the circulation have γδ T-cell receptors that are not restricted to antigen recognition bound to either MHC I or MHC II. Thus γδ T cells do not have either CD4 or CD8 molecules on their surface. The γδ T cell recognizes either stress-induced or nonclassic MHC molecules directly or nonpeptide antigens, such as host or pathogen-derived lipids bound to these nonclassic MHC molecules. The γδ T cell is primarily located in epithelial tissues in certain species and performs effector functions that protect the host from infections and malignancy and maintains tissue integrity. These cells also play a critical role in regulating the immune response, leading to resolution of infection and inflammation.⁶¹

NK cells are large granular lymphocytes with innate immune function. They play a critical role in the early host defense against viral, bacterial, and other infections as well as cancer. NK cells recognize their targets through unique NK receptors and are able to recognize self-MHC class I or class I–like molecules that inhibit or enhance NK function. Their phenotype is characterized by the expression of the CD56 surface antigen and the lack of CD3. NK cells produce IFN-γ, TNF-α, IL-10, and GM-CSF. They exhibit spontaneous cytotoxic activity against virus-infected cells and mediate antibody-dependent cell cytotoxicity through FCRγIII (CD16). Cytotoxicity is the major effector function of NK cells. They bridge the innate and adaptive immune response (Figure 91-6).^62,63 The NKT cells express both CD56 and CD3 T-cell receptor and thus share receptor structures of both conventional NK and T cells. NKT cells are potentially capable of very rapid secretion of large amounts of Th1 or Th2 cytokines but also contain perforin. Both NK and NKT cells use the same mechanisms as the cytotoxic CD8⁺ T cell for killing (i.e., perforin/granzyme cytotoxicity and Fas-ligand/fas cytotoxicity) (see Figure 91-6). The control and resolution of viral infections require the eliminate of the source of the virus, hence the destruction of the virus-infected cell before progeny virus is produced. This is mediated during the innate phase of the immune response by NK cells. In most cases, resolution of active viral infection ultimately requires the development of antigen-specific T lymphocytes, the majority of which are MHC class I–restricted CD8⁺ T cells, although MHC class II CD4⁺ cells and γδ T cells may also mediate cytotoxicity.

Phagocytic cells include neutrophils, eosinophils, monocytes/macrophages, and dendritic cells. They have in common a number of different properties that are of prime importance to the host inflammatory response. Neutrophils, eosinophils, and monocytes/macrophages share the ability to phagocytose foreign material, release granule constituents, secrete inflammatory mediators and regulators, and synthesize reactive oxygen products through a unique reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase enzyme system present on the cell membrane. Neutrophils, eosinophils and basophils are all polymorphonuclear leukocytes (PMNLs). Neutrophils and eosinophils share similar mechanisms of cell migration, phagocytosis, and pathogen killing. All three cell types have segmented nuclei and contain granules, although their granule content varies. Neutrophils are the host’s main defense against bacterial and fungal infections. Eosinophils are important to the control of parasitic infections. Neutrophils may remain in the storage pool of the bone marrow for up to 5 days. Released from the bone marrow into the blood, about half the neutrophils circulate for about 10 hours; the other half remains in a marginated pool, so they are not accessible to phlebotomy. This marginated pool is thought to be in the spleen, along vessel walls, and in the lung microcirculation. Cells can be mobilized from this marginated pool by infection/inflammation and stress. Once circulating neutrophils migrate into the tissue they survive for 1 to 2 days, likely longer in the presence of G-CSF and GM-CSF.

Monocytes, macrophages, and dendritic cells are part of the mononuclear phagocyte system. Although these cells share characteristics with neutrophils and eosinophils, they also have unique properties, including antigen processing and interaction with lymphocytes in the generation of the immune response and extracellular killing of tumor cells. Monocytes are released from the bone marrow, circulate for 1 to 4 days, then migrate into the tissues. Three fourths of the circulating monocytes are localized to blood vessel walls in a marginated pool. Monocytes emigrate into tissue to replace resident macrophages and are either “free” or “fixed.” Free macrophages are found in pleural, synovial, peritoneal, alveolar spaces and in inflammatory sites. Fixed macrophages are generally less motile and include those in the splenic sinusoids, Kupffer cells (liver), bone marrow reticulum, lamina propria of the gastrointestinal tract, lymph node reticulum, osteoclasts in the bone, and as microglia (in the central nervous system).⁶³ These macrophages are heterogenous in their phenotype and function. It has been hypothesized that macrophages have functional patterns such as seen with T cells (i.e., Th1- or Th2-driven phenotypes). However, work supports that monocytes and tissue macrophages develop their phenotypic function in response to changes in the microenvironment in which they are located rather than representing particular populations of monocytes that have been recruited there.¹⁵

The localization of phagocytes to a site of infection is reviewed in a subsequent section. However, after localization occurs, recognition of the pathogen by the phagocyte occurs through fragment receptors (FcRs; see previous section) and complement receptor or other pattern recognition receptors present on the phagocyte. On encountering a particle/pathogen, the appropriate receptors are activated and the phagocyte membrane ruffles. The phagocyte then assumes a bipolar configuration, with the formation of a “head,” or pseudopod, and “tail,” or uropod. The pseudopod surrounds a particle and fuses at its distal end to form a phagolysosome, thus internalizing the particle and a portion of the plasma membrane. The pseudopodia only advance over the portion of the particle or pathogen that is “opsonized” or where there are molecular patterns that fit the appropriate receptor on the phagocyte. Granules present in phagocytes then join this newly formed vacuole and discharge their contents within seconds. Neutrophils have at least three types of granules containing microbial enzymes, myeloperoxidase and lysozyme, proteases, cationic proteins, BPI, and defensins and acid hydrolases. They also contain molecules critical for adhesion and locomotion, such as Mac-1 (CD11b/CD18). Release of myeloperoxidase from primary granules is important in oxygen-dependent microbial killing. Release of other granule constituents, such as lysozyme, lactoferrin, defensins, and BPI, is of critical importance in decreasing the pH of the phagolysosome and in oxygen-independent microbial killing. Phagocytes are responsive to environmental cytokines such as TNF-α, IFN-γ, and chemokines that can prime the phagocyte for increased killing. Phagocytes also produce a large number of products in response to bacterial challenge. Monocytes/macrophages are a rich source of TNF-α and IL-1; even neutrophils produce TNF-α and IL-1, though in lesser amounts. Each cell produces chemokines that attract other leukocytes to the inflammatory focus. Phagocytes can also produce a number of lipid mediators that further stimulate the innate immune response.

The respiratory burst refers to the coordinated consumption of oxygen and production of metabolites that occur when phagocytes are exposed to appropriate stimuli. These events underlie all oxygen-dependent killing by phagocytes. Defects in the respiratory burst mechanisms result in the disorder chronic granulomatous disease. The NADPH oxidase system is a transmembrane electron transport system in which NADPH, the primary electron donor on the cytoplasmic side of the membrane, reduces oxygen in the extracellular fluid or within the phagolysosome to form O₂^– (superoxide). In turn two molecules of O₂^– spontaneously or enzymatically (through superoxide dismutase) generate hydrogen peroxide (H₂O₂). Although both O₂^– and H₂O₂ can directly injure bacteria, the oxidants that are formed from them are primarily responsible for microbicidal action. Myeloperoxidase released into the phagolysosome will enzymatically form hypohalous acids from halide anion. Hypohalous acids such as hypochlorous acid (HOCL) are extremely potent antimicrobials. These agents can then alternatively react with ambient amines (RNH₂) to form N-chloramines (RNHCL). RNHCL are lipophilic oxidizing and chlorinating agents that readily penetrate cellular membranes. The toxic effects of HOCL and RNHCL include sulfhydryl oxidation; hemoprotein inactivation; protein, amino acid, and DNA degradation; and inactivation of essential metabolic cofactors. Oxyradicals, in particular hydroxyl radical (HO), are some of the most powerful oxidizing substances known. In the presence of Fe^3–, O₂^–, and H₂O₂ combine to form HO. Presence of NOS-2 in phagocytes, in particular the macrophages, provides a source of NO that can react with O₂^–, producing peroxynitrite (ONOO^–), another toxic and powerful oxidant.⁶⁴

Neutrophils from patients with chronic granulomatous disease retain some of the antimicrobial activity of normal neutrophils despite the inability to produce oxygen species. This is due to the many endogenous antimicrobials present in the granules that are critical to the killing of microbes. These antimicrobials, discussed earlier in the chapter, include defensins, bacterial permeability increasing protein, lactoferrin, and lysozyme. A number of other proteases, hydrolases, and nucleases in the granules of phagocytes, although not directly microbicidal, act synergistically with the antimicrobial agents to contribute to killing. These products can also result in host injury if released from the neutrophil. Neutrophil elastase, collagenase, and gelatinase (also known as matrix metalloproteinase-8, -9, respectively) can hydrolyze key components of the extracellular matrix. Neutrophil elastase not only can degrade almost all components of the extracellular matrix, but can also cleave a variety of key plasma proteins such as immunoglobulins, complement proteins, and clotting factors. The activity of the elastase outside the cells is regulated primarily by α₁-proteinase inhibitor. Neutrophil elastase can mediate injury outside the neutrophil when α₁-proteinase inhibitor is inactivated by oxidants.⁶⁵

Dendritic cells (DCs) are a distinct lineage of migratory leukocytes. They initiate the primary immune response yet are part of the innate immune system. DCs exist in most tissues of the reticuloendothelial system. They are prominent in tissues exposed to the external environment with frequent exposure to foreign antigen. DCs from different tissues have varying cell membrane markers and functions and include thymic DCs, interstitial DCs (heart, lung, kidney, intestine), interdigitating DCs (lymph nodes), and Langerhans cells (epidermis). The features these DCs have in common are that they originate from bone marrow CD34⁺ stem cells and migrate through the bloodstream to tissues to become immature DCs. Immature DCs take up antigen or respond to environmental cues through pattern recognition receptors and other receptor and nonreceptor mechanisms. This then results in maturation of the DCs and production of unique sets of cytokines and receptors. Depending on the specific environmental signal, DCs will mature into different clones.^66,67

Immature DCs migrate to and remain in the periphery, where they express low levels of MCH I and II. These immature DCs will then mature either after phagocytosis of foreign antigen or by activation of one of its receptors (e.g., TLR). The mature DC will then migrate to the local lymphoid organ; as they do, they increase cell surface expression of MHC as well as T costimulatory molecules (i.e., CD80-CD86) and begin to secrete specific cytokines. Mature DCs lose the ability to phagocytose. In the lymphoid organ, the DC will present antigen to the CD4⁺ or CD8⁺ T-cell receptor via MHC II or I, respectively. The DC costimulatory molecules CD80 or CD86 must engage their ligand, CD28, on the T cell for full activation (Figure 91-6). It should be noted that for the initial immune response, the DC (or APC) must be in geographic proximity of the CD8⁺ and CD4+ cell. Alternatively, the DC or APC may be preconditioned by an activated CD4⁺ cell that is then able to activate naive CD8⁺ cells to become cytotoxic T cells.^66,67 Not all macrophages function well as antigen presenting cells; elicited peritoneal macrophages do, but alveolar macrophages and Kupffer cells do not.¹⁵

Leukocyte Localization

In response to infection and in normal immune surveillance all leukocytes must travel from their sites of production to the point at which their function is required. There is a multiple-step process for localization to occur. Permutations in this process exist in specialized vascular beds such as the lung, liver, and kidney.^69,70

In general, the multistep process begins by activation of the postcapillary venular endothelial surface by inflammatory cytokines such as IL-1, TNF, IL-4, and IFNγ. The surface transforms from a nonadhesive surface to one that is proadhesive through the expression of specific ligands. These ligands will recognize their cognate receptors on the circulating effector leukocytes (and platelets) (i.e., neutrophils, eosinophils, monocytes, NK cells, activated T cells). Endothelial ligands that are upregulated include members of the selectin family and IgSF (Table 91-3). Selectins are responsible for the initial capture of the leukocyte from the free-flowing stream as well as rolling on the endothelial surface of the IgSF that is critical for leukocyte slowing, arrest, and migration on the cell surface (Figure 91-7). The leukocyte ligand for E- and P-selectin is PSGL-1 (CD162, P-selectin glycoprotein ligand 1). The leukocyte receptors for the IgSF are the β2 integrins, in particular Mac-1 (CD11b/CD18), LFA-1 (CD11a/CD18), and the β1 integrin VLA-4 (CD49d/CD29). The leukocyte integrins are heterodimers composed of an α and β subunit. The β subunit may be shared by multiple members of a subfamily, whereas the α subunit confers specificity. Mac-1 functions not only in leukocyte recruitment, but also as a receptor for complement fragment C3bi; hence its alternative name of complement receptor 3. LFA-1 also functions as coinducer of the immune response in the immunologic synapse as discussed above (Table 91-3). The endothelial surface also secretes a number of chemokines and other activating substances, such as PAF, that activate the leukocyte and mediate the transition from rolling to arrest. After the leukocyte has arrested, it polarizes, then “crawls” and emigrates through the endothelial lining of the vessel. This emigration, known as diapedesis, is in response to activating agents (chemokines, LTB₄) released by cells present in the subendothelial matrix, released bacterial products (N-formyl peptides), or through complement activation (C5a). This process is also dependent on leukocyte integrins that recognize IgSF on the endothelial cells necessary for transendothelial migration (see Figure 91-7). Locomotion through the subendothelial matrix requires additional leukocyte integrins (VLA-1, -2, -3, 5, -6) that recognize matrix proteins, including fibronectin, collagen, vitronectin, and vimentin. What effector cell is recruited and the tissue to which it is recruited depend on the adhesion molecules present on the endothelial surface and the effector cells as well as the “signals” released or presented at the endothelial surface and in the subendothelial region.^71-73

Figure 91–7 Leukocyte localization. A, Leukocytes are captured from the free-flowing stream (tether) and roll on the endothelial lining of the blood vessel. This interaction is mediated by all three members of the selectin family. The leukocyte slows, arrests, and changes shape (polarizes). Integrins and their ligands, the immunoglobulin superfamily (IgSF), mediate these steps. The cells then crawl, or diapedese, over the surface of the endothelium until they migrate through the endothelium. The integrins, members of the IgSF, and CD99 all have a role in this response. A leukocyte also may be tethered by adherent platelets or adherent leukocytes. Platelets through β3 integrins and GPIb-IX-V can bind directly to collagen or fibronectin on exposed basement membrane, or alternatively to von Willebrand factor bound to the endothelium or basement membrane. Platelets can release cytokines that activate leukocytes directly once tethered. Leukocytes can bind to platelets directly through integrins or to fibrinogen that is bound to platelets. B, Model of leukocyte activation leading to arrest. Leukocytes tether to the endothelium expressing P-selectin via P-selectin-glycoprotein-ligand 1 (PSGL1). In the presence of shear, PSGL-1 can activate the leukocyte integrins LFA-1 and Mac-1 (arrow 1). Chemotactic factors expressed on the endothelial surface can also directly activate LFA-1 and Mac-1. LFA-1 and Mac-1 both can bind to endothelial ICAM-1 at domains 1 and 3, respectively. C, Proposed mechanisms for transendothelial migration. With stimulation of the endothelial cell, ICAM-1 is upregulated. JAM-1, which is localized at the interendothelial cleft, is mobilized away from the cleft. With activation of the leukocytes, migration across the vascular endothelial can then occur via LFA-1, PECAM-1, JAM-1, and CD99. Leukocyte LFA-1 can bind to JAM-1 and ICAM-1 on the endothelial surface. Leukocytes can then traverse the interendothelial cleft through sequential transhomophilic interactions of PECAM-1 and CD99. Ig, Immunoglobulin; vWF, von Willebrand factor; JAM, junctional adhesion molecule; PECAM, platelet endothelial adhesion molecule.

(From Mariscalco M: Integrins and cell adhesion molecules. Polin RA, Fox WW, Abman SH, editors: Fetal and neonatal physiology, Philadelphia, 2004, Saunders Elsevier.)

This multistep paradigm is also critical for the homing of lymphocytes to the high endothelial venules, a specialized endothelium of the secondary lymphoid organs. Here naive lymphocytes (rather than activated cells) tether, roll, arrest, and emigrate through the endothelial surface. The naive lymphocytes migrate toward DC or APC that secrete cytokines and chemokines necessary for activation of the T cell through the T-cell receptor.⁷⁴ One critical chemokine is SDF-1 (CXCL12). Its receptor, CXCR4, is present on almost all lymphocytes and many monocytes. The multistep paradigm is not operative in all vascular beds. In lungs the inflammatory cells do not “roll,” but instead are physically trapped in the pulmonary capillary bed and emigrate from this area rather than in the postcapillary venules.⁷⁵ In the liver, leukocytes are physically trapped in the sinusoids. Because the sinusoidal endothelia has large pores, the leukocytes can easily interact with the underlying hepatocytes⁷⁶ The platelet can also function as a surface to which a leukocyte can bind by the activated release from the Weibel-Palade body of P-selectin at sites of injury or inflammation to the endothelium (see Figure 91-7).

An increasing number of genetic defects in leukocyte localization have been identified. Those that involved adhesion receptors critical for neutrophil recruitment result in neutrophilia and severe, recurrent skin abscesses. In patients with leukocyte adhesion deficiency type-1 (LAD-1), there is a selective defect in the expression and/or functional activation of β2 (CD18) integrins. In patients with the rarer defect, LAD-2, there is defective rolling of leukocytes on inflamed endothelium due to lack of fucosylated glycoconjugates on the selectins. Finally, patients with LAD-3 have a functional defect in the activation of β1 and β2-integrin avidity, leading to defective leukocyte arrest on vascular endothelium.^71,77,78

Host Response to Infection: A Summary

There is a coordinated and highly regulated response by the body to microbial infection. As outlined in Figure 91-8, the first defense is local immunity. The epithelial surface functions as a physical barrier. Through the release of antimicrobial peptides from the epithelium and secretory IgA from submucosal plasma cells, microbial burden is decreased. The epithelial cells at the site of infection will produce cytokines and chemokines that regulate the invasion of the area by leukocytes. However, these cytokines will also modulate the submucosal macrophages and plasma cells that are constitutively present. The presence of soluble agents critical for opsonization (including Ig and complement components) permit the efficient phagocytosis of organisms by recruited neutrophils and monocytes as well as resident macrophages. The presence of T cells and plasma cells early in the infection depends on a previous encounter with the organism. If the host has immunologic memory for the microbe, then there will be fairly rapid (i.e., within 1 day) expansion of the memory T cells to effector cells (i.e., cytotoxic T cells) and expansion of memory B cells and differentiation to plasma cells to produce IgG. It should be noted that at each step of the process, the number of bacteria decrease.^76,79

Figure 91–8 Overview of defenses at a mucosal site to a bacterial infection. In this example, the host has been previously exposed to the bacterial pathogen. Thus there is a rapid innate immune response, specific antibody, and T-cell response. The number of bacteria decreases as each level of defense is encountered.

(Modified from Goldman AS: Back to basics: host responses to infection, Pediatr Rev 21:342, 2000.)

If the host must mount a primary immune response to control and eliminate the infection, it may take 3 to 5 days for microbial or viral elimination to occur if local mechanisms or innate immune systems response are inadequate. As outlined in Figure 91-9, the initiation of adaptive immunity for a typical viral infection occurs within 2 days. Between 3 and 4 days, there is the establishment of the adaptive immune response with clonal expansion of CD8⁺ and CD4⁺ effector cells, and eradication of the infection occurs. After eradication of the infection there is contraction of the clonal response, but the continued presence of antibody, residual effector cells, and immunologic memory provide lasting protection against reinfection.

Figure 91–9 Time course of a typical acute infection. A, During period 1, the infectious agent (bacterial or viral) replicates. In period 2, an immune response is initiated when the numbers of pathogen exceed the threshold dose of antigen required for an adaptive immune response. Simultaneously, the pathogen continues to replicate, retarded only by the innate and nonadaptive responses. Immunologic memory is believed to be initiated during this stage. During period 3, after 4 to 5 days, effector arms of the immune response begin to clear the infection. In period 4, the clearance of the infectious agent and the decrease in the antigen dose below the response threshold results in the cessation of the response. The presence of antibody, residual effector cells, and immunologic memory provide lasting protection against reinfection. B, During a viral infection, antigen-specific T cells clonally expand during the first phase in the presence of antigen. Soon after the virus is cleared, the contraction phase follows and the number of antigen-specific T cells decreases due to apoptosis. After the contraction phase, the number of virus-specific T cells stabilizes and can be maintained for great lengths of time (memory phase). Magnitude of the CD4+ T-cell response is less than that of the CD8+ cells, and the contraction phase can be less pronounced. The number of memory CD4+ T cells may decline slowly over time.

(Modified from Huang AYC, Rigby MR: The immune response: generation, regulation and maintenance. Stiehm ER, Ochs HD, Winkelstein JA, editors: Immunologic disorders of infants and children, Philadelphia, 2004, Saunders Elsevier.)

Apoptosis is the process of programmed cell death necessary for the resolution of the inflammatory response (see Chapter 100). As cells begin to die, they express ligands on their surfaces that signal resident tissue macrophages for removal via phagocytosis. During apoptosis the cell’s nucleus and cytoplasm condense, nuclear DNA is degraded into small dense pieces, and marked cytoplasmic vesiculation and blebbing of the plasma membrane occur. In the final stages the cell collapses into multiple fragments (apoptotic bodies). Apoptosis can be triggered by external forces such as interaction of CD95 ligand (Fas ligand) on cytotoxic T lymphocytes or NK cells with CFD95 on a target cell such as a virally infected cell (see Figure 91-6). Apoptosis can also be initiated by signals arising from DNA or mitochondrial damage.^80,81 The signals converge on activation of a family of proteases called caspases that cleave multiple substrates to induce destruction of the cell from within. This process is central to peripheral deletion of excess T and B lymphocytes as the immune response wanes, to clearances of infected cells, and to resolution of inflammation with the removal of emigrated leukocytes.⁸² The final step of apoptosis is the removal via the tissue macrophage.⁸³ In the process of eliminating apoptotic cells, the macrophage produces “immunosuppressive” cytokines such as TGF-β and other cytokines that downregulate the inflammatory response (e.g., IL-10).

It is clear that phagocytic members of the reticuloendothelium system, in particular the Kupffer cells and the NKT cells of the liver, work to trap and kill bacteria in the bloodstream. However, it is also clear that neutrophils are also critical to this process; patients with neutropenia have bacteremia despite having adequate macrophage function. Only recently has there been some understanding as to the mechanism. Neutrophils under select conditions can be activated to cause the release of “neutrophil extracellular traps” (NETS), which are weblike structures of DNA. These NETS contain proteolytic activity that can trap and kill bacteria.⁷⁰ Using intravital microscopy, Clark et al.⁸⁴ demonstrated that platelet TLR4 detects TLR4 ligands in the blood and induces platelet binding to adherent neutrophils. The NETS were present and functional under flow conditions and ensnared bacteria within the vasculature. The formation of the NETS occurred primarily in the liver sinusoids and pulmonary capillaries.⁸⁴

References are available online at http://www.expertconsult.com.