There are four main types of bacterial cell wall

The four main types of bacterial cell wall (Fig. 14.1) belong to the following groups.

Fig. 14.1 Bacterial cell walls

Different immunological mechanisms have evolved to destroy the cell wall structure of the different groups of bacteria. All types have an inner cell membrane and a peptidoglycan wall. Gram-negative bacteria also have an outer lipid bilayer in which lipopolysaccharide (LPS) is embedded. Lysosomal enzymes and lysozyme are active against the peptidoglycan layer, whereas cationic proteins and complement are effective against the outer lipid bilayer of the Gram-negative bacteria. The compound cell wall of mycobacteria is extremely resistant to breakdown, and it is likely that this can be achieved only with the assistance of the bacterial enzymes working from within. Some bacteria also have fimbriae or flagellae, which can provide targets for the antibody response. Others have an outer capsule, which renders the organisms more resistant to phagocytosis or to complement. The components indicated with an asterisk (*) are recognized by the innate immune system as a non-specific ‘danger’ signal that selectively boosts some aspects of immune activity. (Gram staining is a method that exploits the fact that crystal violet and iodine form a complex that is more abundant on Gram-positive bacteria. The complex easily elutes from Gram-negative bacteria.)

The outer lipid bilayer of Gram-negative organisms is of particular importance because it is often susceptible to lysis by complement. However, killing of most bacteria usually requires uptake by phagocytes. The outer surface of the bacterium may also contain fimbriae or flagellae, or it may be covered by a protective capsule. These can impede the functions of phagocytes or complement, but they also act as targets for the antibody response, the role of which is discussed later.

Pathogenicity varies between two extreme patterns

The two extreme patterns of pathogenicity are:

Fig. 14.2 Mechanisms of immunopathogenicity

(1) Some bacteria cause disease as a result of only a single toxin (e.g. Corynebacterium diphtheriae, Clostridium tetani) or because of an ability to attach to epithelial surfaces (e.g. in group A streptococcal sore throat). Immunity to such organisms may require only antibody to neutralize this critical function. (2) At the other extreme there are organisms that are not toxic, and cause disease by invasion of tissues and sometimes cells, where damage results mostly from the bulk of organisms or from immunopathology (e.g. lepromatous leprosy). Where organisms invade cells, they must be destroyed and degraded by the cell-mediated immune response. (3) Most organisms fall between the two extremes, with some local invasiveness assisted by local toxicity and enzymes that degrade extracellular matrix (e.g. Staphylococcus aureus, Clostridium perfringens). Antibody and cell-mediated responses are both involved in resistance and the latter is a major cause of antibiotic-induced colitis and diarrhea.

However, most bacteria are intermediate between these extremes, having some invasiveness assisted by some locally acting toxins and spreading factors (tissue-degrading enzymes).

Corynebacterium diphtheriae and Vibrio cholerae are examples of organisms that are toxic, but not invasive. Because their pathogenicity depends almost entirely on toxin production, neutralizing antibody to the toxin is probably sufficient for immunity, though antibody binding to the bacteria and so blocking their adhesion to the epithelium could also be important.

In contrast, the pathogenicity of most invasive organisms does not rely so heavily on a single toxin, so immunity requires killing of the organisms themselves.

The first lines of defense do not depend on antigen recognition

The body’s first line of defense against pathogenic bacteria consists of simple barriers to the entry or establishment of the infection. Thus, the skin and exposed epithelial surfaces have non-specific or innate protective systems, which limit the entry of potentially invasive organisms.

Intact skin is impenetrable to most bacteria. Additionally, fatty acids produced by the skin are toxic to many organisms. Indeed, the pathogenicity of some strains correlates with their ability to survive on the skin. Epithelial surfaces are cleansed, for example, by ciliary action in the trachea or by flushing of the urinary tract.

Many bacteria are destroyed by pH changes in the stomach and vagina, both of which provide an acidic environment. In the vagina, the epithelium secretes glycogen, which is metabolized by particular species of commensal bacteria, producing lactic acid.

The second line of defense is mediated by recognition of bacterial components

If organisms do enter the tissues, they can be combated initially by further elements of the innate immune system. Numerous bacterial components are recognized in ways that do not rely on the antigen-specific receptors of either B cells or T cells. These types of recognition are phylogenetically ancient ‘broad-spectrum’ mechanisms that evolved before antigen-specific T cells and immunoglobulins, allowing protective responses to be triggered by common microbial components bearing so-called ‘pathogen-associated molecular patterns’ (PAMPs), recognized by ‘pattern recognition molecules’ of the innate immune system (see Chapter 6).

Many organisms, such as non-pathogenic cocci, are probably removed from the tissues as a consequence of these pathways, without the need for a specific adaptive immune reaction. Figure 14.3 shows some of the microbial components involved and the host responses that are triggered.

Fig. 14.3 Protective mechanisms not involving antigen-specific B or T cells

Several common bacterial PAMPs are recognized by molecules present in serum and by receptors on cells. These recognition pathways result in activation of the alternative complement pathway (factors C3, B, D, P), with consequent release of C3a and C5a; activation of neutrophils, macrophages, and NK cells; triggering of cytokine and chemokine release; mast cell degranulation, leading to increased blood flow in the local capillary network; and increased adhesion of cells and fibrin to endothelial cells. These mechanisms, plus tissue injury caused by the bacteria, may activate the clotting system and fibrin formation, which limit bacterial spread.

The immune system has selected these structures for recognition because they are not only characteristic of microbes, but are essential for their growth and cannot be easily mutated to evade discovery (though, as might be predicted, there are increasing examples of pathogen strategies that attempt to subvert this process).

It is interesting to note that the ‘Limulus assay’, which is used to detect contaminating lipopolysaccharide (LPS) in preparations for use in humans, is based on one such recognition pathway found in an invertebrate species. In Limulus polyphemus (the horseshoe crab), tiny quantities of LPS trigger fibrin formation, which walls off the LPS-bearing infectious agent.

LPS is the dominant activator of innate immunity in Gram-negative bacterial infection

Injection of pure LPS into mice or even humans is sufficient to mimic most of the features of acute Gram-negative infection, including massive production of proinflammatory cytokines, such as IL-1, IL-6, and tumor necrosis factor (TNF), leading to severe shock.

Q. How does release of proinflammatory cytokines cause shock?

A. These proinflammatory cytokines act directly on endothelium to increase vascular adhesiveness, and indirectly activate other plasma enzyme systems to release vasoactive peptides and amines leading to a drop in blood pressure.

Recognition of LPS is a complex process involving molecules that bind LPS and pass it on to cell membrane-associated receptors on leukocytes, and endothelial and other cells, which initiate this proinflammatory cascade (Fig. 14.4).

Fig. 14.4 Effects of lipopolysaccharide

LPS released from Gram-negative bacteria becomes bound to LPS-binding protein (LBP) which promotes transfer of LPS to either soluble CD14 (sCD14) or to a GPI-linked membrane form of the protein (mCD14) expressed on neutrophils and macrophages and to a lesser extent on epithelial and endothelial cells. LBP then dissociates and transfers LPS to the TLR4/MD2 complex, allowing TLR4 to transduce intracellular signals that increase release of many proinflammatory cytokines including TNFα and IL-6, as well as type I IFN and IL-1. These in turn activate endothelial cells, increasing adhesion molecule expression and also drive the acute phase response in the liver. One product of the acute phase response is further production of LBP and sCD14.

Binding of LPS to TLR4 is a critical event in immune activation. TLR4 knockout mice are resistant to LPS-induced shock and there is some evidence that polymorphisms in human TLR4 may influence the course of infection with these bacteria.

The LBP and CD14, which bind LPS, are also involved in recognition of lipid-containing bacterial components from mycoplasmas, mycobacteria, and spirochetes.

Complement is activated via the alternative pathway

Complement activation can result in the killing of some bacteria, particularly those with an outer lipid bilayer susceptible to the lytic complex (C5b–9).

Perhaps more importantly, complement activation releases C5a, which attracts and activates neutrophils and causes degranulation of mast cells (see Chapter 3). The consequent release of histamine and leukotriene (LTB₄) contributes to further increases in vascular permeability (see Fig. 14.3).

Opsonization of the bacteria, by attachment of cleaved derivatives of C3, is also critically important in subsequent interactions with phagocytes.

Release of proinflammatory cytokines increases the adhesive properties of the vascular endothelium

The rapid release of cytokines such as TNF and IL-1 (see Fig. 14.4) from macrophages increases the adhesive properties of the vascular endothelium and facilitates the passage of more phagocytes into inflamed tissue. Combined with the release of chemokines such as CCL2, CCL3, and CXCL8 (see Chapter 6), this directs the recruitment of different leukocyte populations.

Epithelial cells, neutrophils, and mast cells are also important sources of proinflammatory cytokines.

IL-1, TNF, and IL-6 also initiate the acute phase response, increasing the production of complement components as well as other proteins involved in scavenging material released by tissue damage and, in the case of CRP, an opsonin for improving phagocytosis of bacteria.

When NK cells are stimulated by the phagocyte-derived cytokines IL-12 and IL-18 they rapidly release large quantities of interferon-γ (IFNγ). This response happens within the first day of infection, well before the clonal expansion of antigen-specific T cells, and provides a rapid source of IFNγ to activate macrophages. This T cell-independent pathway helps to explain the considerable resistance of mice with SCID (severe combined immune deficiency, a defect in lymphocyte maturation) to infections such as with Listeria monocytogenes. In mice, CD1d-restricted NK T cells also secrete IFNγ in response to IL-12 and IL-18 and other ligands, and help to further activate both NK cells and macrophages.

Pathogen recognition generates signals that regulate the lymphocyte-mediated response

The signals generated following the recognition of pathogens not only generate a cascade of innate immune events, but also regulate the development of the appropriate lymphocyte-mediated response.

Dendritic cells (DCs) are crucial for the initial priming of naive T cells specific for bacterial antigens. Contact with bacteria in the periphery induces immature DCs to migrate to the draining lymph nodes and augments their antigen-presenting ability by increasing their:

Some of this DC activation occurs secondary to their production of cytokines such as type I IFN.

Activated macrophages also act as antigen-presenting cells (APCs), but probably function more at the site of infection, providing further activation of effector rather than naive T cells. Following initial T cell activation by dendritic cells, B cells are also able to act as APCs during B cell–T cell cooperation and are essential for the protective action of polysaccharide-conjugate vaccines in children against encapsulated bacteria such as S. pneumoniae and H. influenzae.

Binding of bacterial components to pattern recognition receptors such as TLRs induces a local environment rich in cytokines such as IFNγ, IL-12, and IL-18, which promote T cell differentiation down the TH1 rather than TH2 pathway.

Immunologists have made use of these effects for many decades (even without knowing their true molecular basis) in the use of adjuvants in vaccination. ‘Adjuvant’ is derived from the Latin adjuvare, to help. When given experimentally, soluble antigens evoke stronger T and B cell-mediated responses if they are mixed with bacterial components that act as adjuvants. Components with this property are indicated in Figure 14.1. This effect probably reflects that the antigen-specific immune response evolved in a tissue environment that already contained these pharmacologically active bacterial components.

With the exception of proteins such as flagellin, which itself stimulates TLR5 and is also a strong T cell immunogen, the response to pure bacterial antigens, injected without adjuvant-active bacterial components, is essentially an artificial situation that does not occur in nature.

The best known adjuvant in laboratory use, complete Freund’s adjuvant, consists of killed mycobacteria suspended in oil, which is then emulsified with the aqueous antigen solution.

New-generation adjuvants based on bacterial components (and safe to use in humans, unlike Freund’s adjuvant) include synthetic TLR activators such as CpG motifs and monophosphoryl lipid A (MPL) as well as recombinant cytokines such as IL-12, IL-1, and IFNγ. Identifying the best adjuvant for inclusion in a vaccine is arguably as important as the choice of antigens and is dramatically illustrated in the RTS,S malaria vaccine – a product which was not effective until reformulated with a new MPL based adjuvant.

Pathogenic bacteria may avoid the effects of antibody

Neisseria gonorrhoeae is an example of a pathogenic bacterium that uses several immune evasion strategies (Fig. 14.8) and humans can be repeatedly infected with N. gonorrhoeae with no evidence of protective immunity.

Fig. 14.8 Mechanisms used by Neisseria gonorrhoeae to avoid the effects of antibody

N. gonorrhoeae is an example of a bacterium that uses several strategies to avoid the damaging effects of antibody. First, it fails to evoke a large antibody response, and the antibody that does form tends to block the function of damaging antibodies. Second, the organism secretes an IgA protease to destroy antibody. Third, blebs of membrane are released, and these appear to adsorb and so deplete local antibody levels. Finally, the organism uses three strategies to alter its antigenic composition: (i) the LPS may be sialylated, so that it more closely resembles mammalian oligosaccharides and promotes rapid removal of complement; (ii) the organism can undergo phase variation, so that it expresses an alternative set of surface molecules; (iii) the gene encoding pilin, the subunits of the pilus, undergoes homologous recombination to generate variants. N. gonorrhoeae also impairs T cell activation by engaging a co-inhibitory receptor CEACAM-1 on the lymphocyte surface by one of its OPA proteins.

Antibodies may also be important for effective immunity against some intracellular bacteria such as Legionella and Salmonella spp.

Pathogenic bacteria can avoid the detrimental effects of complement

Some bacterial capsules are very poor activators of the alternative pathway (Fig. 14.9).

Fig. 14.9 Avoidance of complement-mediated damage

Bacteria avoid complement-mediated damage by a variety of strategies. (1) An outer capsule or coat prevents complement activation. (2) An outer surface can be configured so that complement receptors on phagocytes cannot obtain access to fixed C3b. (3) Surface structures can be expressed that divert attachment of the lytic complex (MAC) from the cell membrane. (4) Membrane-bound enzyme can degrade fixed complement or cause it to be shed. (5) The outer membrane can resist the insertion of the lytic complex. (6) Secreted decoy proteins can cause complement to be deposited on them and not on the bacterium itself.

For other bacteria, long side chains (O antigens) on their LPS may fix C3b at a distance from the otherwise vulnerable lipid bilayer. Similarly, smooth-surfaced Gram-negative organisms (Escherichia coli, Salmonella spp., Pseudomonas spp.) may fix but then rapidly shed the C5b–C9 membrane lytic complex.

Other organisms exploit the physiological mechanisms that block destruction of host cells by complement. When C3b has attached to a surface it can interact with factor B leading to further C3b amplification or it can become inactivated by factors H and I. Capsules rich in sialic acid (as host cell membranes are) seem to promote the interaction with factors H and I.

Neisseria meningitidis, E. coli K1, and group B streptococci all resist complement attachment in this way.

The M protein of group A streptococci acts as an acceptor for factor H, thus potentiating C3bB dissociation. These bacteria also have a gene for a C5a protease.

Bacterial components attract phagocytes by chemotaxis

Unlike neutrophils, which in the uninfected host are found almost entirely in the blood, resident macrophages are constitutively present in tissues where exposure to pathogens first occurs (such as alveolar macrophages in the lung and Kupffer cells in the liver). These macrophages have some killing activity, but invariably need to be supplemented by recruitment of neutrophils and/or monocytes across the blood vessel wall. Phagocytes are attracted by:

The cellular composition of this inflammatory response varies according to the pathogen and the time since infection. For instance:

The choice of receptors is critical

The choice of receptors used for attachment of the phagocyte to the organism is critical and will determine:

The binding can be mediated by lectins on the organism (e.g. on the fimbriae of E. coli), but receptors on the phagocyte are the most important. These either bind directly to the bacterium or indirectly via host complement and antibody deposited on the bacterial surface (opsonization).

Complement can also be fixed by MBL present in serum, which can itself bind to C1q receptors and CR1.

Additionally, Fc receptors on the phagocyte (FcγRI, FcγRII, and FcγRIII, see Chapter 3) bind antibody that has coated bacteria (see Fig. 14.7), whereas various integrins can bind fibronectin and vitronectin opsonized particles.

Phagocytic cells have many killing methods

Killing of bacteria and fungi occurs most efficiently when the organisms have been internalized by the phagocyte and are now within a host membrane bound phagosome. This confinement helps to deliver antimicrobial molecules to the organism at high concentrations and reduces collateral damage to the host. Maturation of the phagosome into a killing zone occurs by acquisition of microbicidal mediators following fusion with other intracellular vesicles such as lysosomes.

The killing pathways of phagocytic cells can be oxygen dependent, with the generation of reactive oxygen intermediates, or oxygen independent (see Chapter 7). In neutrophils, the oxidative burst may also act indirectly, by promoting the flux of K⁺ ions into the phagosome and activating microbicidal proteases.

A second oxygen-dependent pathway involves the creation of nitric oxide (NO^•) from the guanidino nitrogen of L-arginine. This in turn leads to further toxic substances such as the peroxynitrites, which result from interactions of NO^• with the products of the oxygen reduction pathway.

Oxygen-independent killing mechanisms may be more important than previously thought. Many organisms can be killed by cells from patients with chronic granulomatous disease (CGD), which cannot produce reactive oxygen intermediates, and from patients with myeloperoxidase (MPO) deficiency, which cannot produce hypohalous acids. Some of this killing may be due to NO^•, but many organisms can be killed anaerobically, so other mechanisms must exist. Some have been identified and are discussed below. Indeed, many innate immune factors have actually evolved to work optimally in the hypoxic environment of infected tissues. Under conditions of low oxygen tension and pH, phagocytes specifically upregulate genes which contain hypoxic-response elements, resulting in increased phagocytic activity, a longer life span and the production of antimicrobial molecules and inflammatory cytokines.

Some cationic proteins have antibiotic-like properties

The defensins (Fig. 14.10) are cysteine- and arginine-rich cationic peptides of 30–33 amino acids found in phagocytes such as neutrophils, where they comprise 30–50% of the granule proteins.

Fig. 14.10 Cationic host defense peptides in immunity to fungi and bacteria

Numerous cationic host defense peptides are produced by neutrophils, monocytes, epithelia, and mast cells. Their synthesis is usually constitutive but is also enhanced by proinflammatory cytokines such as TNF, IL-1, IL-22, and IFNγ generated following infection. Originally defined by their direct killing activity against pathogens, they are now known also to act on immune cells, having multiple immunomodulatory effects on inflammation and adaptive immunity.

Q. Which other cells secrete antimicrobial peptides?

A. Paneth cells of the intestine (see Fig. 12.11) and airway epithelial cells (i.e. sites of primary contact with pathogens).

Defensins evolved early in evolution and similar molecules are found in insects. They act by integrating into microbial lipid membranes (in some cases forming ion-permeable channels) and disrupting membrane function and structure, resulting in lysis of the pathogen. Defensins can act both inside and outside of host cells and kill organisms as diverse as Staphylococcus aureus, Pseudomonas aeruginosa, E. coli, as well as fungi such as Cryptococcus neoformans.

Defensins also have important immunostimulatory properties including:

• promoting chemotaxis and phagocytosis;

• regulating cytokine production; and

• acting as adjuvants for adaptive immunity by promoting multiple facets of dendritic cell function including antigen uptake, processing and presentation as well as their migration and maturation.

Other antibacterial peptides include the cathelicidins (which can kill Mycobacterium tuberculosis under the regulation of Vitamin D) and protegrins, which can bind LPS and also form membrane pores.

There are also cationic proteins with different pH optima, including cathepsin G and azurocidin, both of which are related to elastase, but have activity against Gram-negative bacteria – this is unrelated to their enzyme activity.

Neutrophils (and possibly mast cells and eosinophils) possess an extracellular mechanism of microbicidal activity by release of so called neutrophil extracellular traps (NETs). This involves the release of chromatin, histones, and antimicrobial proteins which bind and kill bacteria and fungi as well as generating a barrier against the spread of infection.

Macrophage killing can be enhanced on activation

Unlike neutrophils, which have a short life span but are efficient killers even in their normal state, macrophages are long-lived cells that without appropriate activation can actually provide a haven for microbial growth.

Macrophage activation occurs most effectively by the combination of exposure to:

Optimal activation of macrophages is dependent on TH1 CD4 T cells

Microbial products can directly activate monocytes and resident macrophages to secrete proinflammatory cytokines and thus initiate the immune process. However, complete activation, including the ability to kill intracellular microbes, requires the action of IFNγ. IFNγ knockout mice are extremely susceptible to infection and children with deficiencies in either the IFNγ receptor or the cytokines necessary for its production (such as IL-12, IL-18, and IL-23) have increased susceptibility to intracellular bacteria such as Salmonella spp., and mycobacteria including bacille Calmette–Guérin (BCG).

IFNγ is so potent because it enhances several different microbicidal pathways, including both the respiratory burst and the generation of NO^•.

As described above, NK cells, NK T cells, and even macrophages themselves can produce IFNγ during the innate immune response. However, the additional actions of antigen-specific T cells are necessary for optimal cell-mediated immunity.

The most important source of IFNγ during the adaptive immune response to intracellular bacteria is from TH1 CD4⁺ T cells (Fig. 14.11).

Fig. 14.11 Overview of CD4⁺ T cell-mediated immunity to bacteria and fungi

Naive CD4⁺ T cells are stimulated by class II MHC positive antigen-bearing dendritic cells (DCs) via the TCR, in conjunction with co-stimulatory molecules such as CD80/86 and CD28, which induce T cell activation and proliferation. Differentiation into TH1, TH17, or TH2 effector cells is strongly influenced by the cytokine environment during this interaction – microbial pattern recognition events that favor production of IL-12 promote TH1 development, low IL-12 favors TH2 responses, whereas combinations of IL-6, IL-1 TGFβ, IL-21, and IL-23 are required for development and maintenance of TH17 responses. TH17 cells mediate their biological activities via secretion of IL-17 which induces G-CSF and CXC family chemokines to enhance neutrophil differentiation and recruitment, and of IL-22 which induces antimicrobial peptides such as defensins and mucins. Although not shown here, conditions with high levels of IL-10 or TGFβ can induce regulatory T cells, rather than effector (i.e. TH1, TH2, or TH17) subsets. Optimal T cell help for either B cells or macrophage responses involves T cell-derived cytokines and direct cell contact. TH1 cells also promote opsonizing antibody production of high affinity (Fig. 9.8), which complements their activation of phagocytes by IFNγ, but the main activators of antibody synthesis are THz cells, themselves induced by IL-4, producing IL-4, 5, 6, 10 and 13 (Fig. 9.7).

Patients who have AIDS and a reduced CD4 T cell number and function have dramatically increased susceptibility to M. tuberculosis, as well as Mycobacterium avium and atypical salmonella.

As mentioned above, many bacterial components activate the TLR pattern recognition receptors, ensuring the preferential expression of TH1 rather than TH2 CD4⁺ T cell responses in most cases.

TH1 T cells provide both IFNγ for macrophage activation and B cell help to produce IgG subclasses for opsonization of bacteria, rather than the eosinophilia and IgE responses typical of helminth infections.

There is mutual antagonism between the TH1 and TH2 pathways at the level of both T cell differentiation and also directly on the macrophage:

• IFNγ upregulates induced NO^• synthetase expression; whereas

• IL-4 and -13 promote the expression of arginase, which inhibits NO^• production, reducing the macrophage killing potential and diverting it to a profibrotic phenotype. In some cases these cells are called alternatively activated macrophages.

Other cytokines such as GM-CSF and TNF can also contribute to macrophage activation.

Macrophage activation is also promoted by direct contact with CD4 T cells via CD40–CD40L interactions.

Thus T cell-mediated help for macrophages and B cells shares the common themes of soluble and cell-contact mediated activation by CD4 TH1 cells.

While this functional link between CD4 TH1 cells and macrophages has been known for many years, only recently have we discovered that a different T cell subset (CD4 TH17 cells) mediate a link to neutrophils, the other major phagocyte group in the body. TH17 cells preferentially produce IL-17 and IL-22 and were originally discovered for their role in autoimmune diseases. TH17 cells appear to be particularly important in resistance to extracellular (rather than intracellular) fungi and bacteria at mucosal surfaces. The major biological activity of IL-17 is to increase neutrophil recruitment and differentiation in an indirect manner by acting on epithelial cells to produce CXC chemokines, TNF, IL-6 and G-CSF, whilst IL-22 induces the production of anti-microbial peptides. Since neutrophil responses can also cause pathology if excessive, in different animal models TH17 cells can either be protective or contribute to immune pathology. These cells are also found in humans but to date their importance is not clear.

Successful pathogens have evolved mechanisms to avoid phagocyte-mediated killing

Because most organisms are ultimately killed by phagocytes, it is not surprising that successful pathogens have evolved an array of mechanisms to counteract this risk (Fig. 14.12).

Fig. 14.12 Evasion mechanisms of bacteria (and some fungi)

Evasion mechanisms of bacteria (and some fungi), particularly those that are successful intracellular parasites, have evolved the ability to evade different aspects of phagocyte-mediated killing. (1) Some can secrete repellents or toxins that inhibit chemotaxis. (2) Others have capsules or outer coats that inhibit attachment by the phagocyte (e.g. Streptococcus pneumoniae or the yeast C. neoformans). (3) Others permit uptake, but release factors that block subsequent triggering of killing mechanisms. Once ingested, some, such as M. tuberculosis, inhibit lysosome fusion with the phagosome. They also inhibit the proton pump that acidifies the phagosome, so the pH does not fall. (4) They may also secrete catalase (e.g. staphylococci), which breaks down hydrogen peroxide. (5) Organisms such as M. leprae have highly resistant outer coats. M. leprae surrounds itself with a phenolic glycolipid, which scavenges free radicals. (6) Mycobacteria also release a lipoarabinomannan, which blocks the ability of macrophages to respond to the activating effects of IFNγ. (7) Cells infected with Salmonella enterica, M. tuberculosis, or Chlamydia trachomatis have impaired antigen-presenting function. (8) Several organisms (e.g. Listeria and Shigella spp.) can escape from the phagosome to multiply in the cytoplasm. Finally, the organism may kill the phagocyte via either necrosis (e.g. staphylococci) or induction of apoptosis (e.g. Yersinia spp.).

Infected cells can be killed by CTLs

CD8⁺ cytotoxic T lymphocytes (CTLs) can release intracellular organisms by killing the infected cell. For example, mice become strikingly susceptible to M. tuberculosis if class I MHC genes are knocked out so that antigen-specific CD8⁺ T cells do not develop, and kill infected macrophages.

This is consistent with an essential role for CTLs in resistance to intracellular bacteria, and inducing these responses is now a primary goal of new vaccines against bacteria such as M. tuberculosis as well as other pathogens.

Tissue cells that are not components of the immune system can also harbor bacteria such as M. leprae, invasive Shigella and Salmonella spp., and Rickettsia and Chlamydia spp. These infected cells may also be sacrificed by CTLs.

Dendritic cells appear to be particularly important in the generation of strong CD8 T cell responses to bacteria such as L. monocytogenes and Salmonella spp.

Although antigen processing and presentation via the class I MHC pathway (see Chapter 8) is most efficient for microbial antigens derived from the cytosol, nevertheless, CTLs are also clearly induced by bacteria that never escape the phagosome such as M. tuberculosis, salmonellae, and chlamydiae. This occurs either by cross-presentation of antigens within the same cell or by cross-priming where antigens are released from infected cells undergoing apoptosis and then transferred to nearby DCs for efficient presentation via the MHC I pathway. In some cases, lysis of infected host cells by CTLs can result in killing of the organism inside. This can be due to the action of granulysin – an antibacterial peptide stored in the cytotoxic granules and released during the cytotoxic process.

CTLs can also secrete IFNγ when they recognize infected targets, providing an additional pathway of macrophage activation and protective immunity (Fig. 14.13).

Fig. 14.13 Pathways of CD8 T cell activation and function

Naive CD8 T cells are activated by peptides presented via MHC class I molecules, primarily derived from microorganisms that reside in the cytoplasm, such as viruses and some intracellular bacteria that escape the phagosome such as Listeria spp. Other pathogens that do not escape the phagosome (such as M. tuberculosis) can still induce CD8 T cell responses via cross-priming, in which infected and apoptotic host cells release antigenic fragments that are taken up by dendritic cells (DCs). Effector CD8 T cells (CTLs) provide protection by releasing proinflammatory and macrophage-activating cytokines and killing infected host cells via perforin release and Fas. In some cases, the release of granulysin from the CTL can also result in killing of the pathogen.

Other T cell populations can contribute to antibacterial immunity

In addition to the classical MHC class I- and MHC class II-mediated recognition of bacterial proteins by αβ CD4 and CD8 T cells, other ‘non-conventional’ T cell populations allow the host to respond rapidly to other microbial chemistries.

T cells bearing γδ (rather than αβ) receptors (see Chapter 5) proliferate in response to bacterial infection.

Some γδ T cells recognize small phospholigands derived from M. tuberculosis and possibly other bacteria, whereas others are triggered in an antigen-independent manner by the presence of pathogen-activated dendritic cells expressing high levels of co-stimulatory molecules and IL-12.

NK T cells are a diverse group of T cells, some of which have an invariant T cell antigen receptor. They recognize not proteins, but hydrophobic antigens, particularly microbial glycolipids such as the lipoarabinomannan from M. tuberculosis, presented via CD1 molecules (see Fig. 5.w3).

Such γδ and NKT cells can have cytotoxic activity and also secrete multiple cytokines including IFNγ and IL-17 (depending on how they are stimulated) giving a potential role in host defense. In animal models of infection these non-conventional T cells can be protective or immunoregulatory, but their relative importance in human immunity is not resolved.

Examples to illustrate the relationship between the nature of an organism, the disease, and immunopathology caused, and the mechanism of immune response that leads to protection, are given in Figure 14.14.

Fig. 14.14 Immunity in some important bacterial infections

This table provides examples of how a knowledge of the organism, and the mechanism of disease, can lead to a prediction of the relevant protective mechanism.

Excessive cytokine release can lead to endotoxin shock

If cytokine release is sudden and massive, several acute tissue-damaging syndromes can result and are potentially fatal.

One of the most severe examples of this is endotoxin (septicemic) shock, when there is massive production of cytokines, usually caused by bacterial products released during septicemic episodes. Endotoxin (LPS) from Gram-negative bacteria is usually responsible, though Gram-positive septicemia can cause a similar syndrome. There can be life-threatening fever, circulatory collapse, diffuse intravascular coagulation, and hemorrhagic necrosis, leading eventually to multiple organ failure (Fig. 14.15).

Fig. 14.15 Endotoxin shock

Excessive release of cytokines, often triggered by the endotoxin (LPS) of Gram-negative bacteria, can lead to diffuse intravascular coagulation with consequent defective clotting, changes in vascular permeability, loss of fluid into the tissues, a fall in blood pressure, circulatory collapse, and hemorrhagic necrosis, particularly in the gut. This figure illustrates some important parts of this pathway at the cellular level. The cytokines TNF and IL-1 cause endothelial cells to express cell adhesion molecules and tissue thromboplastin. These promote adhesion of circulating cells and deposition of fibrin, respectively. Platelet activating factor (PAF) enhances these effects. In experimental models, shock can be blocked by neutralizing antibodies to TNF, and greatly diminished by antibodies to tissue thromboplastin, or by inhibitors of PAF or of nitric oxide production, but these have not been successful clinically. Gram-positive bacteria can induce shock, for example by massive release of cytokines mediated by superantigens. (PDGF, platelet-derived growth factor, produced by both platelets and endothelium.)

Paradoxically, individuals who recover from the initial life-threatening phase often overcompensate and switch from a hyper- to a hyporesponsive phase, in which excessive production of endogenous immune regulators such as IL-10 and TGFβ (and possibly other mechanisms) results in immune paralysis, making them susceptible to secondary infection.

The Schwartzman reaction is a form of cytokine-dependent tissue damage

Schwartzman observed that if Gram-negative organisms were injected into the skin of rabbits, followed by a second dose given intravenously 24 hours later, hemorrhagic necrosis occurred at the prepared skin site. This is known as the Schwartzman reaction (Fig. 14.w1).

Fig. 14.w1 The Schwartzman reaction

In the Schwartzman reaction there is cytokine-mediated tissue damage in a site of previous inflammation. This phenomenon is related to several clinical situations in humans. The first injection into the skin prepares the site by inducing inflammation and upregulating cytokine receptors, which are now the target for systemic cytokines released by the later intravenous injection of endotoxin (LPS).

Many other organisms are now known to ‘prepare’ the skin in the same way, including streptococci, mycobacteria, Haemophilus spp., corynebacteria, and vaccinia virus.

Schwartzman also noted that two intravenous injections 24 hours apart caused a systemic reaction, commonly involving circulatory collapse and bilateral necrosis of the renal cortex. Sanarelli had made similar observations and this is now known as the systemic Schwartzman, or Sanarelli–Schwartzman, reaction.

These reactions can also be accompanied by necrosis in the pancreas, pituitary, adrenals, and gut. There is marked diffuse intravascular coagulation and thrombosis.

Endotoxin (LPS) is the active component of the intravenous ‘triggering’ injection.

Early work implicated endothelial changes, fibrin deposition, neutrophil accumulation and degranulation, and platelets as mediating the damage. This is correct, but it is now clear that tissues are primed by the induction of IFNγ (involving IL-12) derived from either NK cells or NKT cells, whereas TNF is critical in the effector phase of systemic tissue damage.

These phenomena may contribute to the characteristic hemorrhagic rash seen in children with meningococcal meningitis and the systemic effects observed in Gram-negative septic shock.

The Koch phenomenon is necrosis in T cell-mediated mycobacterial lesions and skin test sites

The Koch phenomenon is a necrotic response to antigens of M. tuberculosis, originally demonstrated by Robert Koch in tuberculous guinea pigs (Fig. 14.w2). It may be related to the necrosis that also occurs in the lesions in tuberculosis. It is at least partly due to the release of cytokines into a T cell-mediated inflammatory site (delayed hypersensitivity site). Such sites can be extremely sensitive to the tissue-damaging effects of cytokines, as seen in the Schwartzman reaction, particularly when there is mixed TH1 and TH2 activity.

Fig. 14.w2 The Koch phenomenon

Robert Koch observed that injection of M. tuberculosis or soluble antigens from M. tuberculosis into the skin of tuberculous guinea pigs resulted in a necrotizing reaction both at the challenge site and in the original tuberculous lesion. This is at least partly due to the fact that the delayed hypersensitivity reaction to mycobacterial antigens can, like the LPS-injected site in the Schwartzman reaction, be very sensitive to the toxicity of cytokines. These may be released locally in the skin test site. A similar reaction is seen in humans who have or have had tuberculosis. Responses to the same antigen in individuals who are skin-test positive as a result of BCG (bacille Calmette–Guérin) vaccination do not usually show necrosis.

Some individuals suffer from excessive immune responses

Why some individuals suffer from excessive immune responses and others control infection with little or no damage is still unclear. The host constantly attempts to regulate immune responses to avoid the scenarios described above. For instance, immunoregulatory cytokines such as IL-10 and transforming growth factor-β (TGFβ) are produced from:

The influence of both ‘natural’ and ‘inducible’ regulatory T cell populations has been reported in infections with bacteria such as L. monocytogenes, Helicobacter spp., Bordetella pertussis, and M. tuberculosis. Their presence may be beneficial to the host in preventing unwanted tissue damage, but can also be subverted by pathogens as a means of immune evasion. For example, filamentous hemagglutinin and adenylate cyclase toxin, two known virulence factors of B. pertussis, promote the induction of IL-10-secreting regulatory T cells in mice, dampening protective TH1 immunity and promoting persistence in the host.

Excessive immune responses can occur during treatment of severe bacterial infections

Paradoxically, excessive immune responses can also occur during treatment of humans with severe bacterial infections.

The simultaneous destruction of large numbers of organisms by antibiotics can cause increased cytokine production and systemic pathology following treatment of relapsing fever, Lyme disease, and tuberculosis.

Also, when patients who have AIDS and underlying infections such as tuberculosis and cryptococcal meningitis are treated with effective antiretroviral drugs (highly active antiretroviral therapy – HAART), the recovering T cells induce an ‘immune reconstitution inflammatory syndrome’ with severe tissue damage.

Heat-shock proteins are prominent targets of immune responses

Heat-shock proteins (HSPs or stress proteins) are found in all eukaryotic and prokaryotic cells, where they have essential roles in the assembly, folding, and transport of other molecules.

Mammalian cells exposed to abnormally elevated temperatures (or to other stresses including infection) express higher levels of these proteins, which reflects their role in the stabilization of protein structure.

Bacteria also increase HSP expression, particularly as they adapt to the intracellular environment of the host.

Immune responses are generated against both mammalian and pathogen-derived HSPs during infection.

Mammalian (and probably also bacterial) HSPs are able to activate innate immunity via multiple cell surface receptors including the α-globulin receptor CD91, CD40, CD14, and TLRs 2 and 4.

HSPs can enhance the presentation of antigens via the class I MHC/CD8 pathway (see Chapter 5) and are themselves immunogenic for both CD4 and CD8 T cells.

Because the amino acid sequences of HSPs are very highly conserved between humans and bacteria there is still considerable speculation about their role in the initiation of autoimmunity, though it seems their contribution is more complicated than simply antigenic mimicry.

The toxicity of superantigens results from massive cytokine release

Certain bacterial components called superantigens bind directly to the variable regions of β chains (Vβ) of antigen receptors on subsets of T cells, and cross-link them to the MHC molecules of APCs, usually outside the normal antigen-binding groove (Fig. 14.16). Between them staphylococci and streptococci have some 21 different superantigens and these molecules can also be found in other bacteria such as mycoplasmas. The full biological significance of this bacterial adaptation is not yet clear – it could be to the organism’s advantage to exhaust or deplete T cells that would otherwise be protective.

Fig 14.16 T cell stimulation by superantigens

Normally antigenic peptides are processed and presented on MHC molecules (left). Superantigens, such as staphylococcal enterotoxins, are not processed but bind directly to MHC class II molecules and Vβ of the TCR (right). Each superantigen activates a distinct set of Vβ-expressing T cells.

One certain effect is the toxicity of the massive release of cytokines (including IL-2, TNFα, and TNFβ, together with IL-1β from activated macrophages) due to the simultaneous stimulation of up to 20% of the entire T cell pool.

The staphylococcal toxins responsible for the toxic shock syndrome (toxic shock syndrome toxin-1 [TSST-1], etc.) operate in this way, though not all shock syndromes caused by staphylococci are the result of T cell activation.

Recent evidence suggests that streptococcal M protein, a known virulence factor of S. pyogenes, forms a complex with fibrinogen, which then binds to β-integrins on neutrophils, causing the release of inflammatory mediators, which also result in massive vascular leakage and shock.

The ‘hygiene hypothesis’

Several groups of diseases, all characterized by defects in the regulation of the immune system, are becoming more common, particularly in developed countries. These diseases include:

This may have many causes but the ‘hygiene hypothesis’ suggests that increasing immunological dysregulation correlates with decreasing exposure to environmental microorganisms. Decreased exposure could be due to hygiene, vaccines, and antibiotic use. This was mostly thought to act on infants as they develop in the first few years of life but there are now indications that this can be preceded by immune programming in utero, where infection and inflammatory stresses acting on the mother (such as aerosol exposure to microbe containing dust) during pregnancy directly influence the immune response of the newborn. If proved correct, the solution would clearly not be the abandonment of the most important achievements of medicine (hygiene, vaccines, antibiotics), but rather the improvement of maternal health and identification of other environmental factors that are lacking from the modern lifestyle so that they can be replaced as vaccines or probiotics.

However, recent data suggest that the correlation does not always hold true – some viral infections (such as respiratory syncytial virus) seem to promote rather than decrease allergy and asthma in animal models and in humans.

Whatever their cause, prevention or treatment of these diseases will need to address:

There are four categories of fungal infection

Although some fungi can cause disease in otherwise healthy individuals, severe fungal infections are a growing problem because of the markedly increased numbers of immunologically compromised hosts. Fungal infections are therefore regularly seen in:

These clinical findings point to the key roles of neutrophils and macrophages and the CD4 T cell subsets that regulate their activity (i.e. TH1 and TH17) in antifungal immunity.

Human fungal infections fall into the following four major categories:

Innate immune responses to fungi include defensins and phagocytes

The basic protective features of the skin and normal commensal flora described against bacterial infections above are also important in resistance to fungi.

Defensins have antifungal as well as antibacterial properties, and collectins such as MBL and the surfactant proteins A and D can bind, aggregate, and opsonize fungi for phagocytosis.

Phagocytes, particularly neutrophils (Fig. 14.17) and macrophages, are essential for killing fungi, either by:

Fig. 14.17 Evidence for neutrophil-mediated immunity to mucormycosis

This is a section from the lung of a patient suffering from mucormycosis – an opportunistic infection in an immunosuppressed subject. The inflammatory reaction consists almost entirely of neutrophil polymorphs around the fungal hyphae. The disease is particularly associated with neutropenia (lack of neutrophils). Silver stain. × 400.

(Courtesy of Professor RJ Hay.)

The oxidative burst plays a crucial role in some antifungal responses, as seen in the susceptibility to severe aspergillosis by patients with CGD who have defects in the NADPH oxidase system. However, phagocytes from such patients with defective oxygen reduction pathways nevertheless kill other yeast and hyphae with near normal efficiency, so demonstrating the role of other killing mechanisms (Fig. 14.18). For instance, NO^• and its derivatives are important for resistance to C. neoformans.

Fig. 14.18 Monocyte/macrophage killing of fungi

Many fungi are killed by monocytes or macrophages. Individuals with chronic granulomatous disease (CGD) are highly susceptible to Aspergillus spp. infections whereas myeloperoxidase (MPO) deficiency does not usually lead to opportunistic infection, suggesting that non-oxygen-dependent mechanisms are also important in host defense.

These responses rely on the recognition of PAMPs in the fungal cell wall by either soluble or cell-bound pattern recognition molecules. The TLR family again plays an important role in this process, along with the mannose receptor and complement receptors:

Dectin-1, a C type lectin receptor is widely expressed on myeloid cells of the gut and airway mucosa. Recognition of fungi via this receptor promotes phagocytosis, triggers the respiratory burst and elicits inflammatory cytokine, chemokine and prostaglandin responses. TNF is one of these important cytokines in humans, since individuals given anti-TNF therapy have increased susceptibility to multiple fungal pathogens. Not all of these recognition events are to the host’s advantage, for example binding of Candida albicans mannan via TLR4 induces proinflammatory chemokine responses, whereas ligation of candidal phospholipomannan and glucans with TLR2/dectin-1 generates a strong IL-10 response, which may inhibit the relevant immune response.

T cell-mediated immunity is critical for resistance to fungi

Most fungi are highly immunogenic and induce strong antibody and T cell-mediated immune responses, which can be detected by serology and delayed-type (type IV) hypersensitivity skin reactions (see Chapter 26).

Considerable evidence points to the dominant protective role of TH1 (and perhaps also TH17 T cells) and phagocyte activation, rather than antibody-mediated responses.

Patients with T cell deficiencies, rather than defects in antibody production, are more at risk of disseminated fungal disease, and antibody titers, though useful as an epidemiological tool to determine exposure, do not necessarily correlate with prognosis. Nevertheless, fungi can elicit both protective and non-protective antibodies and the protection afforded by some experimental vaccines can be adoptively transferred by immune sera.

Resistance to most pathogenic fungi (including dermatophytes and most systemic mycoses including C. neoformans, Histoplasma capsulatum, etc., but not Aspergillus spp.) is clearly dependent upon T cell-mediated immunity, particularly CD4⁺ TH1 cells secreting IFNγ and to a lesser extent CD8 T cells (Fig. 14.19). As in the case of bacteria, dendritic cells are necessary for this response and produce IL-12 after engulfing fungi.

Fig. 14.19 Evidence for T cell immunity in chromomycosis

The pigmented fungal cells of chromomycosis (a subcutaneous mycosis) (c) are visible inside giant cells (gc) in the dermis of a patient. The area is surrounded by a predominantly mononuclear cell infiltrate. The basal layer of epidermis (e) is visible at the top of the frame. H & E stain. × 400.

(Courtesy of Professor RJ Hay.)

The clinical relevance of TH1 versus TH2 responses is also clear for some human mycoses, for example:

Children with the primary immunodeficiency hyper IgE syndrome have defects in the production of IFNγ, fail to develop TH17 cells and have increased susceptibility to fungal infections.

An increased level of IL-10 (with concomitant reductions in IFNγ) is also a marker of impaired immunity to systemic mycoses, C. albicans, and in neutropenia-associated aspergillosis.

Fungi possess many evasion strategies to promote their survival

Evasion strategies used by fungi to promote their survival include the following:

Immune responses to fungi are therefore as complex and interesting as those against bacteria and for many infections (such as the subcutaneous mycoses) these responses remain poorly understood.

New immunological approaches are being developed to prevent and treat fungal infections

Unlike many antibiotics, which are directly microbicidal, antifungal drugs need significant assistance from the immune system to be most effective.

Reducing the underlying immunosuppression that leads to susceptibility to fungi is an important goal and generic immunotherapies such as cytokine administration (using IFNγ in patients with CGD and granulocyte colony stimulating factor [G-CSF] therapy to reduce neutropenia in patients with cancer) have had some success. Human antibodies specific for fungal antigens are being tested for their protective effects by passive transfer. In a clinical trial, administration of Aspergillus specific donor CD4 T cells reduced the incidence of invasive aspergillosis in patients undergoing allogeneic bone marrow transplantation. There is also considerable interest in dendritic cell-based vaccine strategies to promote TH1-mediated immunity.