Chapter 11:  Failures of Host Defense Mechanisms

11-1. Antigenic variation allows pathogens to escape from immunity

One way in which an infectious agent can evade immune surveillance is by altering its antigens; this is particularly important for extracellular pathogens, against which the principal defense is the production of antibody against their surface structures. There are three ways in which antigenic variation can occur. First, many infectious agents exist in a wide variety of antigenic types. There are, for example, 84 known types of Streptococcus pneumoniae, an important cause of bacterial pneumonia. Each type differs from the others in the structure of its polysaccharide capsule. The different types are distinguished by serological tests and so are often known as serotypes. Infection with one serotype of such an organism can lead to type-specific immunity, which protects against reinfection with that type but not with a different serotype. Thus, from the point of view of the adaptive immune system, each serotype of S. pneumoniae represents a distinct organism. The result is that essentially the same pathogen can cause disease many times in the same individual (Fig. 11.1).

A second, more dynamic mechanism of antigenic variation is seen in the influenza virus. At any one time, a single virus type is responsible for most infections throughout the world. The human population gradually develops protective immunity to this virus type, chiefly by directing neutralizing antibody against the major surface protein of the influenza virus, its hemagglutinin. Because the virus is rapidly cleared from individual hosts, its survival depends on having a large pool of unprotected individuals among whom it spreads very readily. The virus might therefore be in danger of running out of potential hosts if it had not evolved two distinct ways of changing its antigenic type (Fig. 11.2).

The first of these, antigenic drift, is caused by point mutations in the genes encoding hemagglutinin and a second surface protein, neuraminidase. Every 2–3 years, a variant arises with mutations that allow the virus to evade neutralization by antibodies in the population; other mutations affect epitopes that are recognized by T cells and, in particular, CD8 T cells, so that cells infected with the mutant virus also escape destruction. Individuals who were previously infected with, and hence are immune to, the old variant are thus susceptible to the new variant. This causes an epidemic that is relatively mild because there is still some cross-reaction with antibodies and T cells produced against the previous variant of the virus, and therefore most of the population have some level of immunity (see Section 10-25).

Major influenza pandemics resulting in widespread and often fatal disease occur as the result of the second process, which is termed antigenic shift. This happens when there is reassortment of the segmented RNA genome of the influenza virus and related animal influenza viruses in an animal host, leading to major changes in the hemagglutinin protein on the viral surface. The resulting virus is recognized poorly, if at all, by antibodies and by T cells directed against the previous variant, so that most people are highly susceptible to the new virus, and severe infection results.

The third mechanism of antigenic variation involves programmed rearrangements in the DNA of the pathogen. The most striking example occurs in African trypanosomes, where changes in the major surface antigen occur repeatedly within a single infected host. Trypanosomes are insect-borne protozoa that replicate in the extracellular tissue spaces of the body and cause sleeping sickness in humans. The trypanosome is coated with a single type of glycoprotein, the variant-specific glycoprotein (VSG), which elicits a potent protective antibody response that rapidly clears most of the parasites. The trypanosome genome, however, contains about 1000 VSG genes, each encoding a protein with distinct antigenic properties. Only one of these is expressed at any one time by being placed into an active ‘expression site’ in the genome. The VSG gene expressed can be changed by gene rearrangement that places a new VSG gene into the expression site (Fig. 11.3). So, by having their own system of gene rearrangement that can change the VSG protein produced, trypanosomes keep one step ahead of an immune system capable of generating many distinct antibodies by gene rearrangement. A few trypanosomes with such changed surface glycoproteins thus evade the antibodies made by the host, and these soon grow and cause a recurrence of disease (see Fig. 11.3, bottom panel). Antibodies are then made against the new VSG, and the whole cycle repeats. This chronic cycle of antigen clearance leads to immune-complex damage and inflammation, and eventually to neurological damage, finally resulting in coma. This gives African trypanosomiasis its common name of sleeping sickness. These cycles of evasive action make trypanosome infections very difficult for the immune system to defeat, and they are a major health problem in Africa. Malaria is another major disease caused by a protozoan parasite that varies its antigens to evade elimination by the immune system.

Antigenic variation also occurs in bacteria: DNA rearrangements help to account for the success of two important bacterial pathogens—Salmonella typhimurium, a common cause of salmonella food poisoning, and Neisseria gonorrhoeae, which causes gonorrhea, a major sexually transmitted disease and an increasing public health problem in the United States. S. typhimurium regularly alternates its surface flagellin protein by inverting a segment of its DNA containing the promoter for one flagellin gene. This turns off expression of the gene and allows the expression of a second flagellin gene, which encodes an antigenically distinct protein. N. gonorrhoeae has several variable antigens, the most striking of which is the pilin protein, which, like the variable surface glycoproteins of the African trypanosome, is encoded by several variant genes, only one of which is active at any given time. Silent versions of the gene from time to time replace the active version downstream of the pilin promoter. All of these mechanisms help the pathogen to evade an otherwise specific and effective immune response.

11-2. Some viruses persist in vivo by ceasing to replicate until immunity wanes

Viruses usually betray their presence to the immune system once they have entered cells by directing the synthesis of viral proteins, fragments of which are displayed on the surface MHC molecules of the infected cell, where they are detected by T lymphocytes. To replicate, a virus must make viral proteins, and rapidly replicating viruses that produce acute viral illnesses are therefore readily detected by T cells, which normally control them. Some viruses, however, can enter a state known as latency in which the virus is not being replicated. In the latent state, the virus does not cause disease but, because there are no viral peptides to flag its presence, the virus cannot be eliminated. Such latent infections can be reactivated and this results in recurrent illness.

Herpes viruses often enter latency. Herpes simplex virus, the cause of cold sores, infects epithelia and spreads to sensory neurons serving the area of infection. After an effective immune response controls the epithelial infection, the virus persists in a latent state in the sensory neurons. Factors such as sunlight, bacterial infection, or hormonal changes reactivate the virus, which then travels down the axons of the sensory neuron and reinfects the epithelial tissues (Fig. 11.4). At this point, the immune response again becomes active and controls the local infection by killing the epithelial cells, producing a new sore. This cycle can be repeated many times. There are two reasons why the sensory neuron remains infected: first, the virus is quiescent in the nerve and therefore few viral proteins are produced, generating few virus-derived peptides to present on MHC class I; second, neurons carry very low levels of MHC class I molecules, which makes it harder for CD8 T cells to recognize infected neurons and attack them. This low level of MHC class I expression might be beneficial, as it reduces the risk that neurons, which regenerate very slowly if at all, will be attacked in appropriately by CD8 T cells. It also makes neurons unusually vulnerable to persistent infections. Another example of this is provided by herpes zoster (or varicella zoster), the virus that causes chickenpox. This virus remains latent in one or a few dorsal root ganglia after the acute illness is over and can be reactivated by stress or immunosuppression to spread down the nerve and reinfect the skin. The reinfection causes the reappearance of the classic rash of varicella in the area of skin served by the infected dorsal root, a disease commonly called shingles. Herpes simplex reactivation is frequent, but herpes zoster usually reactivates only once in a lifetime in an immunocompetent host.

The Epstein-Barr virus (EBV), yet another herpes virus, enters latency in B cells after a primary infection that often passes without being diagnosed. In a minority of infected individuals, the initial acute infection of B cells is more severe, causing a disease known as infectious mononucleosis or glandular fever. EBV infects B cells by binding to CR2 (CD21), a component of the B-cell co-receptor complex. The infection causes most of the infected cells to proliferate and produce virus, leading in turn to the proliferation of antigen-specific T cells and the excess of mononuclear white cells in the blood that gives the disease its name. The infection is controlled eventually by specific CD8 T cells, which kill the infected proliferating B cells. A fraction of B lymphocytes become latently infected, however, and EBV remains quiescent in these cells. Latently infected cells express a viral protein, EBNA-1, which is needed to maintain the viral genome, but EBNA-1 interacts with the proteasome (see Section 5-3) to prevent its own degradation into peptides that would elicit a T-cell response.

Latently infected B cells can be isolated by taking B cells from individuals who have apparently cleared their EBV infection and placing them in tissue culture: in the absence of T cells, the latently infected cells that have retained the EBV genome transform infected B cells sometimes undergo malignant transformation, giving rise to a B-cell lymphoma called Burkitt's lymphoma (see Section 7-33). This is a rare event, and it seems likely that a crucial part of this process is a failure of T-cell surveillance. Further support for this hypothesis comes from the increased risk of EBV-associated B-cell lymphomas developing in patients with acquired and inherited immuno-deficiencies of T-cell function (see Sections 11-15 and 11-26).

11-3. Some pathogens resist destruction by host defense mechanisms or exploit them for their own purposes

Some pathogens induce a normal immune response but have evolved specialized mechanisms for resisting its effects. For instance, some bacteria that are engulfed in the normal way by macrophages have evolved means of avoiding destruction by these phagocytes; indeed, they use macrophages as their primary host. Mycobacterium tuberculosis, for example, is taken up by macrophages but prevents the fusion of the phagosome with the lysosome, protecting itself from the bactericidal actions of the lysosomal contents.

Other microorganisms, such as Listeria monocytogenes, escape from the phagosome into the cytoplasm of the macrophage, where they can multiply readily. They then spread to adjacent cells in tissues without emerging from the cell into the extracellular environment. They do this by hijacking the host cytoskeletal protein actin, which assembles into filaments at the rear of the bacterium. The actin filaments drive the bacteria forward into vacuolar projections to adjacent cells; these vacuoles are then lysed by the Listeria, releasing the bacteria directly into the cytoplasm of the adjacent cell. In this way they avoid attack by antibodies. Cells infected with L. monocytogenes are, however, susceptible to killing by cytotoxic T cells. The protozoan parasite Toxoplasma gondii can apparently generate its own vesicle, which isolates it from the rest of the cell because it does not fuse with any cellular vesicle. This might actually enable T. gondii to avoid making peptides derived from its proteins accessible for loading onto MHC molecules, and thus remain invisible to the immune system.

Two prominent spirochetal infections, Lyme disease and syphilis, avoid elimination by antibodies through less well understood mechanisms and establish a persistent and extremely damaging infection in tissues. Lyme disease is caused by the spirochete bacterium Borrelia burgdorferi, whereas syphilis, the more widespread and much the better understood of the two diseases, is caused by Treponema pallidum. T. pallidum is believed to avoid recognition by antibodies by coating its surface with host molecules until it has invaded tissues such as the central nervous system, where it is less easily reached by antibodies.

Finally, many viruses have evolved mechanisms to subvert various arms of the immune system. These range from capturing cellular genes for cytokines or cytokine receptors, to synthesizing complement-regulatory molecules, inhibiting MHC class I synthesis or assembly, or producing decoy proteins that mimic so-called TIR domains that we learned about in Section 6-15. This area is one of the most rapidly expanding areas in the field of host-pathogen relationships. Examples of how members of the herpes and poxvirus families subvert host responses are shown in Fig. 11.5.

11-4. Immunosuppression or inappropriate immune responses can contribute to persistent disease

Many pathogens suppress immune responses in general. For example, staphylococci produce toxins, such as the staphylococcal enterotoxins and toxic shock syndrome toxin-1, that act as superantigens. Superantigens are proteins that bind the antigen receptors of very large numbers of T cells (see Section 7-26), stimulating them to produce cytokines that cause significant suppression of all immune responses. The details of this suppression are not understood. The stimulated T cells proliferate and then rapidly undergo apoptosis, leaving a generalized immunosuppression together with the deletion of many peripheral T cells.

Many other pathogens cause mild or transient immunosuppression during acute infection. These forms of suppressed immunity are poorly understood but important, as they often make the host susceptible to secondary infections by common environmental microorganisms. A crucially important example of immune suppression follows trauma, burns, or even major surgery. The burned patient has a clearly diminished capability to respond to infection, and generalized infection is a common cause of death in these patients. The reasons for this are not fully understood.

Measles virus infection, in spite of the widespread availability of an effective vaccine, still accounts for 10% of the global mortality of children under 5 years old and is the eighth leading cause of death worldwide. Malnourished children are the main victims and the cause of death is usually secondary bacterial infection, particularly pneumonia caused by measles-induced immunosuppression. The immunosuppression that follows measles infection can last for several months and is associated with reduced T- and B-cell function. There is reduced or absent delayed-type hypersensitivity and, during this period of acquired immunodeficiency, children have markedly increased susceptibility to mycobacterial infection, reflecting the important role of macrophage activation by T_H1 cells in host defense against mycobacteria. An important mechanism for measles-induced immunosuppression is the infection of dendritic cells by measles virus. Infected dendritic cells cause unresponsiveness of T lymphocytes by mechanisms that are not yet understood, and it seems likely that this is the proximate cause of the immunosuppression induced by measles virus.

The most extreme case of immune suppression caused by a pathogen is the acquired immune deficiency syndrome caused by infection with HIV. The ultimate cause of death in AIDS is usually infection with an opportunistic pathogen, a term used to describe a microorganism that is present in the environment but does not usually cause disease because it is well controlled by normal host defenses. HIV infection leads to a gradual loss of immune competence, allowing infection with organisms that are not normally pathogenic.

Leprosy, which we discussed in Section 8-13, is a more complex case, in which the causal bacterium, Mycobacterium leprae, is associated either with the suppression of cell-mediated immunity or with a strong cell-mediated antibacterial response. This leads to two major forms of the disease—lepromatous and tuberculoid leprosy. In lepromatous leprosy, cell-mediated immunity is profoundly depressed, M. leprae are present in great profusion, and cellular immune responses to many antigens are suppressed. This leads to a phenotypic state in such patients called anergy, here meaning the absence of delayed-type hypersensitivity to a wide range of antigens unrelated to M. leprae. In tuberculoid leprosy, by contrast, there is potent cell-mediated immunity with macrophage activation, which controls but does not eradicate infection. Few viable microorganisms are found in tissues, the patients usually survive, and most of the symptoms and pathology are caused by the inflammatory response to these persistent microorganisms (Fig. 11.6). The difference between the two forms of disease might lie in a difference in the ratio of T_H1 to T_H2 cells, and this is thought to be caused by cytokines produced by CD8 T cells, as we learned in Section 10-6.

11-5. Immune responses can contribute directly to pathogenesis

Tuberculoid leprosy is just one example of an infection in which the pathology is caused largely by the immune response. This is true to some degree in most infections; for example, the fever that accompanies a bacterial infection is caused by the release of cytokines by macrophages. One medically important example of immunopathology is the wheezy broncheolitis caused by respiratory syncytial virus (RSV). Broncheolitis caused by RSV is the major cause of admission of young children to hospital in the Western world, with as many as 90,000 admissions and 4500 deaths each year in the United States alone. The first indication that the immune response to the virus might have a role in the pathogenesis of this disease came from the observation that young infants vaccinated with an alum-precipitated killed virus preparation suffered a worse disease than unvaccinated children. This occurred because the vaccine failed to induce neutralizing antibodies but succeeded in producing T_H2 cells. On infection, the T_H2 cells released interleukin (IL)-3, IL-4, and IL-5, which induced bronchospasm, increased mucus secretion, and tissue eosinophilia. Mice can be infected with RSV and develop a disease similar to that seen in humans.

Another example of a pathogenic immune response is the response to the eggs of the schistosome. Schistosomes are parasitic worms that lay eggs in the hepatic portal vein. Some of the eggs reach the intestine and are shed in the feces, spreading the infection; others lodge in the portal circulation of the liver, where they elicit a potent immune response leading to chronic inflammation, hepatic fibrosis, and eventually liver failure. This process reflects the excessive activation of T_H1 cells, and can be modulated by T_H2 cells, IL-4, or CD8 T cells, which can also act by producing IL-4.

In the case of the mouse mammary tumor virus (MMTV), a retrovirus that causes mammary tumors in mice, the immune response is required for the infective cycle of the pathogen (Fig. 11.7). MMTV is transferred from the mother's mammary gland to her pups in milk. The virus then enters the B lymphocytes of the new host, where it must replicate to be transported to the mammary epithelium to continue its life cycle. As it is a retrovirus, however, MMTV can replicate only in dividing cells. The virus ensures that infected B cells will proliferate by causing them to express on their surface a superantigen encoded within the MMTV genome. This superantigen enables the B cells to bypass the requirement for specific antigen and stimulate large numbers of CD4 T cells with the appropriate T-cell receptor V_β domain (see Section 5-15), causing them to produce cytokines and express CD40 ligand, which in turn stimulates the B cells to divide. The virus can then replicate in the B cells and infect the host's mammary epithelial cells.

One way to block this cycle of transmission is by deleting the particular subset of T cells carrying the V_β domain recognized by the viral superantigen. This has been done experimentally by taking mice that are normally susceptible to a particular MMTV virus, and using the superantigen gene from this virus to construct transgenic mice. As we learned in Section 7-26, superantigens that are expressed in the thymus induce the clonal deletion of developing T cells. Thus the expressed transgene induced the loss of T cells bearing the appropriate V_β domains. The B cells in these transgenic mice could be infected by the MMTV virus but could not activate any of the remaining T cells. Thus the infected B cells were not stimulated to divide, and could not support MMTV replication. Consequently, the transgenic mice, unlike their nontransgenic littermates, were unable to transmit the relevant strain of MMTV.

This mode of protection against MMTV might explain the finding that most mouse strains have MMTV genomes stably integrated into their DNA. These defective endogenous retroviruses have lost certain essential genes and are unable to produce virions, but they have retained the genes encoding their superantigens, which are expressed on the cells of the host. Although a section of the T-cell repertoire is lost as a result of carrying these endogenous retroviruses, the mice are protected against infection with nondefective MMTV encoding the same superantigen. There are several different strains of MMTV whose superantigens bind to different V_β domains, and these are matched by different endogenous MMTV strains. Mice containing different endogenous MMTV genomes delete different parts of their T-cell receptor repertoire, reducing the risk that whole mouse populations will be susceptible to a given MMTV strain. No human diseases dependent on such mechanisms have yet been described.

Summary

Infectious agents can cause recurrent or persistent disease by avoiding normal host defense mechanisms or by subverting them to promote their own replication. There are many different ways of evading or subverting the immune response. Antigenic variation, latency, resistance to immune effector mechanisms, and suppression of the immune response all contribute to persistent and medically important infections. In some cases, the immune response is part of the problem; some pathogens use immune activation to spread infection, others would not cause disease if it were not for the immune response. Each of these mechanisms teaches us something about the nature of the immune response and its weaknesses, and each requires a different medical approach to prevent or to treat infection.

11-6. A history of repeated infections suggests a diagnosis of immunodeficiency

Patients with immune deficiency are usually detected clinically by a history of recurrent infection. The type of infection is a guide to which part of the immune system is deficient. Recurrent infection by pyogenic bacteria suggests a defect in antibody, complement, or phagocyte function, reflecting the role of these parts of the immune system in host defense against such infections. By contrast, a history of recurrent viral infections is more suggestive of a defect in host defense mediated by T lymphocytes.

To determine the competence of the immune system in patients with possible immunodeficiency, a battery of tests is usually conducted (Fig. 11.8); these focus with increasing precision as the nature of the defect is narrowed down to a single element. The presence of the various cell types in blood is determined by routine hematology, often followed by FACS analysis (see Appendix I, Section A-22) of lymphocyte subsets, and the measurement of serum immunoglobulins. The phagocytic competence of freshly isolated polymorphonuclear leukocytes and monocytes is tested, and the efficiency of the complement system is determined by testing the dilution of serum required for lysis of 50% of antibody-coated red blood cells (this is denoted the CH₅₀).

In general, if such tests reveal a defect in one of these broad compartments of immune function, more specialized testing is then needed to determine the precise nature of the defect. Tests of lymphocyte function are often valuable, starting with the ability of polyclonal mitogens to induce T-cell proliferation and B-cell secretion of immunoglobulin in tissue culture (see Appendix I, Section A-31). These tests can eventually pinpoint the cellular defect in immunodeficiency.

11-7. Inherited immunodeficiency diseases are caused by recessive gene defects

Before the advent of highly effective antibiotic therapy, it is likely that most individuals with inherited immune defects died in infancy or early childhood because of their susceptibility to particular classes of pathogen (Fig. 11.9). Such cases would not have been easy to identify, as many normal infants also died of infection. Thus, although many inherited immunodeficiency diseases have now been identified, the first immunodeficiency disease was not described until 1952. Most of the gene defects that cause these inherited immunodeficiencies are recessive and, for this reason, many of the known immunodeficiencies are caused by mutations in genes on the X chromosome. Recessive defects cause disease only when both chromosomes are defective. However, as males have only one X chromosome, all males who inherit an X chromosome carrying a defective gene will manifest disease, whereas female carriers, having two X chromosomes, are perfectly healthy. Immunodeficiency diseases that affect various steps in B- and T-lymphocyte development have been described, as have defects in surface molecules that are important for T- or B-cell function. Defects in phagocytic cells, in complement, in cytokines, in cytokine receptors, and in molecules that mediate effector responses also occur (see Fig. 11.9). Thus, immunodeficiency can be caused by defects in either the adaptive or the innate immune system. Individual examples of these diseases will be described in later sections.

More recently, the use of gene knockout techniques in mice has allowed the creation of many immunodeficient states that are adding rapidly to our knowledge of the contribution of individual molecules to normal immune function. Nevertheless, human immunodeficiency disease is still the best source of insight into normal pathways of host defense against infectious diseases in humans. For example, a deficiency of antibody, of complement, or of phagocytic function each increases the risk of infection by certain pyogenic bacteria. This shows that the normal pathway of host defense against such bacteria is the binding of antibody followed by fixation of complement, which allows the uptake of opsonized bacteria by phagocytic cells. Breaking any one of the links in this chain of events leading to bacterial killing causes a similar immunodeficient state.

The study of immunodeficiency also teaches us about the redundancy of mechanisms of host defense against infectious disease. The first two humans to be discovered with a hereditary deficiency of complement were healthy immunologists. This teaches us two lessons. The first is that there are multiple protective immune mechanisms against infection; for example, although there is abundant evidence that complement deficiency increases susceptibility to pyogenic infection, not every human with complement deficiency suffers from recurrent infections. The second lesson concerns the phenomenon of ascertainment artifact. When an unusual observation is made in a patient with disease, there is a temptation to seek a causal link. However, no one would suggest that complement deficiency causes a genetic predisposition to becoming an immunologist. Complement deficiency was discovered in immunologists because they used their own blood in their experiments. If a particular measurement is made only in a highly selected group of patients with a particular disease, it is inevitable that the only abnormal results will be discovered in patients with that disease. This is an ascertainment artifact and emphasizes the importance of studying appropriate controls.

11-8. The main effect of low levels of antibody is an inability to clear extracellular bacteria

Pyogenic, or pus-forming, bacteria have polysaccharide capsules that are not directly recognized by the receptors on macrophages and neutrophils that stimulate phagocytosis. They therefore escape immediate elimination by the innate immune response and are successful extracellular pathogens. Normal individuals can clear infections by such bacteria because antibody and complement opsonize the bacteria, making it possible for phagocytes to ingest and destroy them. The principal effect of deficiencies in antibody production is therefore a failure to control this class of bacterial infection. In addition, susceptibility to some viral infections, most notably those caused by enteroviruses, is also increased, because of the importance of antibodies in neutralizing infectious viruses that enter the body through the gut (see Chapter 10).

The first description of an immunodeficiency disease was Ogden C. Bruton's account, in 1952, of the failure of a male child to produce antibody. As this defect is inherited in an X-linked fashion and is characterized by the absence of immunoglobulin in the serum, it was called Bruton's X-linked agammaglobulinemia (XLA). The absence of antibody can be detected using immunoelectrophoresis (Fig. 11.10). Since then, many more diseases of antibody production have been described, most of them the consequence of failures in the development or activation of B lymphocytes. Infants with these diseases are usually identified as a result of recurrent infections with pyogenic bacteria such as Streptococcus pneumoniae.

The defective gene in XLA is now known to encode a protein tyrosine kinase called Btk (Bruton's tyrosine kinase) which is a member of the recently described family called Tec kinases (see Section 6-10). This protein is expressed in neutrophils as well as in B cells, although only B cells are defective in these patients, in whom B-cell maturation halts at the pre-B-cell stage. Thus it is likely that Btk is required to couple the pre-B-cell receptor (which consists of heavy chains, surrogate light chains, and Igα and Igβ) to nuclear events that lead to pre-B-cell growth and differentiation (see Section 7-9). In patients with Btk deficiencies, some B cells mature despite the defect in the signaling kinase, suggesting that signals transmitted by these kinases are not absolutely required.

As the gene responsible for XLA is found on the X chromosome, it is possible to identify female carriers by analyzing X-chromosome inactivation in their B cells. During development, female cells randomly inactivate one of their two X chromosomes. Because the product of a normal btk gene is required for normal B-lymphocyte development, only cells in which the normal allele of btk is active can develop into mature B cells. Thus, in female carriers of mutant btk genes, all B cells have the normal X chromosome as the active X. By contrast, the active X chromosomes in the T cells and macrophages of carriers are an equal mixture of the normal and btk mutant X chromosomes. This fact allowed female carriers of XLA to be identified even before the nature of btk was known. Nonrandom X inactivation only in B cells also demonstrates conclusively that the btk gene is required for normal B-cell development but not for the development of other cell types, and that Btk must act within B cells rather than on stromal cells or other cells required for B-cell development (Fig. 11.11).

The commonest humoral immune defect is the transient deficiency in immunoglobulin production that occurs in the first 6–12 months of life. The newborn infant has initial antibody levels comparable to those of the mother, because of the transplacental transport of maternal IgG (see Chapter 9). As the transferred IgG is catabolized, antibody levels gradually decrease until the infant begins to produce useful amounts of its own IgG at about 6 months of age (Fig. 11.12). Thus, IgG levels are quite low between the ages of 3 months and 1 year and active IgG antibody responses are poor. In some infants this can lead to a period of heightened susceptibility to infection. This is especially true for premature babies, who begin with lower levels of maternal IgG and also reach immune competence later after birth.

The most common inherited form of immunoglobulin deficiency is selective IgA deficiency, which is seen in about 1 person in 800. Although no obvious disease susceptibility is associated with selective IgA defects, they are commoner in people with chronic lung disease than in the general population. Lack of IgA might thus result in a predisposition to lung infections with various pathogens and is consistent with the role of IgA in defense at the body's surfaces. The genetic basis of this defect is unknown but some data suggest that a gene of unidentified function mapping in the class III region of the MHC could be involved. A related syndrome called common variable immunodeficiency, in which there is usually a deficiency in both IgG and IgA, also maps to the MHC region.

People with pure B-cell defects resist many pathogens successfully. However, effective host defense against a subset of extracellular pyogenic bacteria, including staphylococci and streptococci, requires opsonization of these bacteria with specific antibody. These infections can be suppressed with antibiotics and periodic infusions of human immunoglobulin collected from a large pool of donors. As there are antibodies against many pathogens in this pooled immunoglobulin, it serves as a fairly successful shield against infection.

11-9. T-cell defects can result in low antibody levels

Patients with X-linked hyper IgM syndrome have normal B- and T-cell development and high serum levels of IgM but make very limited IgM antibody responses against T-cell dependent antigens and produce immunoglobulin isotypes other than IgM and IgD only in trace amounts. This makes them highly susceptible to infection with extracellular pathogens. The molecular defect in this disease is in the CD40 ligand expressed on activated T cells, which therefore cannot engage the CD40 molecule on B cells; the B cells themselves are normal. We learned in Chapter 9 that CD40 ligand is critical in the T-cell dependent activation of B-cell proliferation and these patients show that CD40 ligand is also essential for the induction of the isotype switch and formation of germinal centers (Fig. 11.13). There are also defects in cell-mediated immunity in these individuals. For example, they are susceptible to infection with the opportunistic lung pathogen Pneumocystis carinii, which is normally killed by activated macrophages. The susceptibility is thought to be due, at least in part, to the inability of the T cells to deliver an activating signal to infected macrophages by engaging the CD40 expressed on these cells (see Section 8-29). A defect in T-cell activation could also contribute to the profound immunodeficiency suffered by these patients, as studies on mice that lack CD40 ligand have revealed a failure of antigen-specific T cells to expand in response to primary immunization with antigen.

In XLA, the hunt for the cause of the disease led to the discovery of a previously unidentified gene product. In X-linked hyper IgM syndrome, the gene for CD40 ligand was cloned independently and only then identified as the defective gene in this disorder. Thus, inherited immunodeficiencies can either lead us to new genes or help us to determine the roles of known genes in normal immune system function.

11-10. Defects in complement components cause defective humoral immune function

Not surprisingly, the spectrum of infections associated with complement deficiencies overlaps substantially with that seen in patients with deficiencies in antibody production. Defects in the activation of C3, and in C3 itself, are associated with a wide range of pyogenic infections, emphasizing the important role of C3 as an opsonin, promoting the phagocytosis of bacteria (Fig. 11.14). In contrast, defects in the membrane-attack components of complement (C5–C9) have more limited effects and result exclusively in susceptibility to Neisseria species. This indicates that host defense against these bacteria, which are capable of intracellular survival, is mediated by extracellular lysis by the membrane-attack complex of complement. Accurate data from large population studies in Japan, where endemic N. meningitidis infection is rare, show that the risk each year to a normal person of infection with this organism is approximately 1/2,000,000. This compares with a risk of 1/200 in the same population to a person with inherited deficiency of one of the membrane-attack complex proteins—a 10,000-fold increase in risk compared to a person with normal complement activity. The early components of the classical complement pathway are particularly important for the elimination of immune complexes and apoptotic cells, which can cause significant pathology in autoimmune diseases such as systemic lupus erythematosus. This aspect of inherited complement deficiency is discussed in Chapter 13.

Another set of diseases are caused by defects in complement control proteins (see Section 2-14). People lacking decay-accelerating factor (DAF) and CD59, which protect a person's own cell surfaces from complement activation, destroy their own red blood cells. This results in the disease paroxysmal nocturnal hemoglobinuria, as we learned in Chapter 2. A more striking consequence of the loss of a regulatory protein is seen in patients with C1-inhibitor defects. C1-inhibitor irreversibly inhibits the activity of several serine proteinase enzymes. These include two enzymes that participate in the contact activation system, factor XIIa (activated Hageman factor) and kallikrein, in addition to the two enzymes that together initiate the classical pathway of complement, C1r and C1s. Deficiency of C1-inhibitor leads to failure to regulate these two pathways. Their unregulated activity results in the excessive production of vasoactive mediators that cause fluid accumulation in the tissues and epiglottal swelling that can lead to suffocation. These mediators are bradykinin, produced by the cleavage of high molecular weight kininogen by kallikrein and the C2 kinin, produced by the activity of C1s on C2a. This syndrome is called hereditary angioneurotic edema.

11-11. Defects in phagocytic cells permit widespread bacterial infections

Defects in the recruitment of phagocytic cells to extravascular sites of infection can cause serious immunodeficiency. Leukocytes reach such sites by emigrating from blood vessels in a tightly regulated process consisting of three stages. The first is the rolling adherence of leukocytes to endothelial cells, through the binding of a fucosylated tetrasaccharide ligand known as sialyl-Lewis^x to E-selectin and P-selectin. Sialyl-Lewis^x is expressed on monocytes and neutrophils, whereas E-selectin and P-selectin are expressed on endothelium activated by mediators from the site of inflammation. The second stage is the tight adherence of the leukocytes to the endothelium through the binding of leukocyte β₂ integrins such as CD11b:CD18 (Mac-1:CR3) to counterreceptors on endothelial cells. The third and final stage is the transmigration of leukocytes through the endothelium along gradients of chemotactic molecules originating from the site of tissue injury. Neutrophil recruitment is illustrated in Fig. 2.36 and described in more detail in Section 2-22.

Deficiencies in the molecules involved in each of these stages can prevent neutrophils and macrophages from reaching sites of infection to ingest and destroy bacteria. Reduced rolling adhesion has been described in patients with a lack of sialyl-Lewis^x caused by a deficiency in the fucosylation pathway responsible for its biosynthesis. Similarly, deficiencies in the leukocyte integrin common β₂ subunit CD18 have been identified and these prevent leukocyte migration to sites of infection because of a lack of tight leukocyte adhesion, causing the leukocyte adhesion deficiency syndrome. All these deficiencies lead to infections that are resistant to antibiotic treatment and that persist despite an apparently effective cellular and humoral adaptive immune response. A deficiency of neutrophils (neutropenia) associated with chemotherapy, malignancy, or aplastic anemia is also associated with a similar spectrum of severe pyogenic bacterial infections.

Most of the other known defects in phagocytic cells affect their ability to kill intracellular and/or ingested extracellular bacteria (Fig. 11.15). In chronic granulomatous disease, phagocytes cannot produce the superoxide radical and their antibacterial activity is thereby seriously impaired. Several different genetic defects, affecting any one of the four constituent proteins of the NADPH oxidase system, can cause this. Patients with this disease have chronic bacterial infections, which in some cases lead to the formation of granulomas. Deficiencies in the enzymes glucose-6-phosphate dehydrogenase and myeloperoxidase also impair intracellular killing and lead to a similar, although less severe, phenotype. Finally, in Chediak-Higashi syndrome, a complex syndrome characterized by partial albinism, abnormal platelet function, and severe immunodeficiency, a defect in a gene encoding a protein involved in intracellular vesicle formation causes a failure to fuse lysosomes properly with phagosomes; the phagocytes in these patients have enlarged granules and impaired intracellular killing.

11-12. Defects in T-cell function result in severe combined immunodeficiencies

Although patients with B-cell defects can deal with many pathogens adequately, patients with defects in T-cell development are highly susceptible to a broad range of infectious agents. This demonstrates the central role of T cells in adaptive immune responses to virtually all antigens. As such patients make neither specific T-cell dependent antibody responses nor cell-mediated immune responses, and thus cannot develop immunological memory, they are said to suffer from severe combined immunodeficiency (SCID).

Several different defects can lead to the SCID phenotype. In X-linked SCID, which is the commonest form of SCID, T cells fail to develop because of a mutation in the common γ chain of several cytokine receptors, including those for the interleukins IL-2, IL-4, IL-7, IL-9, and IL-15. We will examine this defect further in Section 11-13. The commonest forms of autosomally inherited SCID are due to adenosine deaminase (ADA) deficiency and purine nucleotide phosphorylase (PNP) deficiency. These enzyme defects affect purine degradation, and both result in an accumulation of nucleotide metabolites that are particularly toxic to developing T cells. B cells are also somewhat compromised in these patients.

One class of SCID individuals lack expression of all MHC class II gene products on their cells. This condition is also referred to as the bare lymphocyte syndrome as MHC class II molecules are not expressed on lymphocytes or thymic epithelial cells. As the thymus in such individuals lacks MHC class II molecules, CD4 T cells cannot be positively selected and therefore few develop. The antigen-presenting cells in these individuals also lack MHC class II molecules and so the few CD4 T cells that do develop cannot be stimulated by antigen. In these individuals, MHC class I expression is normal and CD8 T cells develop normally. However, such people suffer from severe combined immunodeficiency, illustrating the central importance of CD4 T cells in adaptive immunity to most pathogens. The syndrome is caused not by mutations in the MHC genes themselves, but by mutations in one of several different genes encoding gene-regulatory proteins that are required for the transcriptional activation of MHC class II promoters. Four complementing gene defects (known as Groups A, B, C, and D) have been defined in patients who fail to express MHC class II molecules, which implies that at least four different genes are required for normal MHC class II gene expression. One of these, named the MHC class II transactivator, or CIITA, is the gene mutated in Group A. The genes mutated in Groups B, C, and D are named RFXANK, RFX5, and RFXAP. These genes encode three proteins that are components of a multimeric transcriptional complex, RFX, which binds a sequence named an X box, present in the promoter of all MHC class II genes.

In contrast, a more limited immunodeficiency, associated with chronic respiratory bacterial infections and skin ulceration with vasculitis, has been observed in a small number of patients showing almost complete absence of cell-surface MHC class I molecules. This condition has been labeled bare lymphocyte syndrome (MHC class I). Affected individuals have normal levels of mRNA encoding MHC class I molecules and normal production of MHC class I proteins, but these proteins reach the cell surface in severely reduced numbers. The defect was shown to be similar to that in the TAP mutant cells mentioned in Section 5-2 and, indeed, affected patients have been found to have mutations in the TAP1 or TAP2 genes that encode the two subunits of the peptide transporter. The absence of MHC class I molecules at the cell surface leads to a lack of CD8 T cells expressing the α:β T cell receptor, but these patients do have CD8 T cells that bear the γ:δ receptor. It is surprising that they are not abnormally susceptible to viral infections, given the key role of MHC class I presentation and cytotoxic CD8 α:β T cells in the control of viral infections. However, there is evidence for TAP-independent pathways of antigen presentation by MHC class I molecules of certain peptides. The clinical phenotype of TAP1- and TAP2-deficient patients illustrates that these pathways may be sufficient to allow the control of viral infections.

Another set of defects leading to SCID are those that cause failures of DNA rearrangement in developing lymphocytes. For example, defects in either the RAG-1 or RAG-2 genes result in the arrest of lymphocyte development because of a failure to rearrange the antigen receptor genes. Thus there is a complete lack of T and B cells in mice with genetically engineered defects in the RAG genes, and in patients with autosomally inherited forms of SCID who lack a functional RAG protein. There are other patients with mutations in either the RAG-1 or RAG-2 genes who can nonetheless make a small amount of functional RAG protein, allowing a small amount of V(D)J recombination activity. They suffer from a distinctive and severe disease called Omenn's syndrome, in which, in addition to increased susceptibility to multiple opportunistic infections, there are also clinical features very similar to graft-versus-host disease (see Section 13-21) with rashes, eosinophilia, diarrhea, and enlargement of the lymph nodes. Normal or elevated numbers of T cells, all of which are activated, are found in these unfortunate children. A possible explanation for this phenotype is that very low levels of RAG activity allow some limited T-cell receptor gene recombination. However, no B cells are found and it may be that B cells have more stringent requirements for RAG activity. The T cells that are produced in these patients show an abnormal and highly restricted receptor repertoire, both in the thymus and in the periphery, where they have undergone clonal expansion and activation. The clinical features strongly suggest that these peripheral T cells are autoreactive and responsible for the graft-versushost phenotype.

Another group of patients with autosomal SCID have a phenotype very similar to that of a mutant mouse strain called scid; scid mice suffer from an abnormal sensitivity to ionizing radiation as well as from severe combined immuno-deficiency. They produce very few mature B and T cells, as there is a failure of DNA rearrangement in their developing lymphocytes; only rare VJ or VDJ joints are seen and most of these have abnormal features. The underlying defect has now been shown to be in the enzyme DNA-dependent protein kinase (DNA-PK), which binds to the end of the double-stranded breaks that occur during the process of antigen receptor gene rearrangement. These ends are found as DNA hairpin structures in the immature thymocytes of scid mice. Thus, it seems likely that DNA-PK is involved in resolving the hairpin structure (see Section 4-5).

Other defects in DNA repair and metabolizing enzymes are associated with a combination of immunodeficiency, increased sensitivity to the damaging effects of ionizing radiation, and cancer development. One example is Bloom's syndrome, a disease caused by mutations in a DNA helicase enzyme, which unwinds DNA. Another is ataxia telangiectasia (AT), in which the underlying defect is in a protein called ATM, which contains a kinase domain thought to be involved in intracellular signaling in response to DNA damage. Because repair of double-stranded DNA breaks and lymphocyte division are central to the function of the adaptive immune system, it is not surprising that defects such as these are associated with the development of immunodeficiency.

Finally, in patients with DiGeorge's syndrome the thymic epithelium fails to develop normally. Without the proper inductive environment T cells cannot mature, and both T-cell dependent antibody production and cell-mediated immunity are absent. Such patients have some serum immuno-globulin and variable numbers of B and T cells. As with all the severe combined immuo-deficiency diseases, it is the defect in T cells that is crucial. These diseases abundantly illustrate the central role of T cells in virtually all adaptive immune responses. In many cases B-cell development is normal, yet the response to nearly all pathogens is profoundly impaired.

11-13. Defective T-cell signaling, cytokine production, or cytokine action can cause immunodeficiency

As we learned in Chapter 8, virtually all adaptive immune responses require the activation of antigen-specific T lymphocytes and their differentiation into cells producing cytokines that act on specific cytokine receptors. Several gene defects have been described that interfere with these processes. Thus, patients who lack CD3γ chains have low levels of surface T-cell receptors and defective T-cell responses. Patients making low levels of mutant CD3ε chains are also deficient in T-cell activation. Patients who make a defective form of the cytosolic protein tyrosine kinase ZAP-70, which transmits signals from the T-cell receptor (see Section 6-9) have recently been described. Their CD4 T cells emerge from the thymus in normal numbers, whereas CD8 T cells are absent. However, the CD4 T cells that mature fail to respond to stimuli that normally activate via the T-cell receptor and the patients are thus very immunodeficient.

Another group of patients show a lack of IL-2 production upon receptor ligation, and these patients have a severe immunodeficiency; however, T-cell development is normal in these individuals, as it is in mice in which mutations have been made in their IL-2 genes by gene knockout (see Appendix I, Section A-47). These IL-2-negative patients have heterogeneous defects; some of them fail to activate the transcription factor NFAT (see Section 6-11), which induces the transcription of several cytokine genes in addition to the IL-2 gene. This might explain why their immunodeficiency is more profound than that of mice whose IL-2 gene has been disrupted. IL-2-deficient mice can mount adaptive immune responses through an IL-2-independent pathway, possibly involving the cytokine IL-15, which shares many activities with IL-2; nevertheless, they are susceptible to a variety of infectious agents.

In contrast to the normal development of T cells in patients deficient in IL-2, there is a failure of T-cell development in patients with X-linked severe combined immunodeficiency (X-linked SCID), which is caused by a defect in the γ chain of the IL-2 receptor. Thus, this disease showed that the common γ chain (γ_c) must be important in T-cell development for reasons unrelated to IL-2 binding or IL-2 responses. The demonstration that the IL-2 receptor γ_c chain is also part of other cytokine receptors, including the IL-7 receptor, helps to explain its role in early T-cell development. The γ chain seems to function in transducing the signal from this group of receptors and interacts with a kinase, JAK3 kinase, which is known to be defective in patients with an autosomally inherited immunodeficiency similar in phenotype to X-linked SCID.

As in all serious T-cell deficiencies, X-linked SCID patients do not make effective antibody responses to most antigens, although their B cells seem normal. However, as the gene defect is on the X chromosome, one can determine whether the lack of B-cell function is solely a consequence of the lack of T-cell help by examining X-chromosome inactivation (see Section 11-7) in B cells of unaffected carriers. The majority of naive IgM-positive B cells from female carriers of X-linked SCID have inactivated the defective X chromosome rather than the normal one, showing that B-cell development is affected by, but not wholly dependent on, the common γ chain. However, mature memory B cells that have switched to isotypes other than IgM have inactivated the defective X chromosome almost without exception. This might reflect the fact that the IL-2 receptor γ chain is also part of the IL-4 receptor. Thus, B cells that lack this chain will have defective IL-4 receptors and will not proliferate in T-cell-dependent antibody responses. X-linked SCID is so severe that children who inherit it can survive only in a completely pathogen-free environment, unless given antibodies and successfully treated by bone marrow transplantation. A famous case in Houston became known as the ‘bubble baby’ because of the plastic bubble in which he was enclosed to protect him from infection.

Wiskott-Aldrich syndrome (WAS) is a disease that has shed new light on the molecular basis of T-cell signaling and its importance for immune function. The disease affects platelets and was first described as a blood-clotting disorder, but it is also associated with immunodeficiency due to impaired T-cell function, reduced T-cell numbers, and a failure of antibody responses to encapsulated bacteria. WAS is caused by a defective gene on the X chromosome, encoding a protein called WAS protein (WASP). This protein has been shown to bind Cdc42, a small GTP-binding protein that is known to regulate the organization of the actin cytoskeleton and to be important for the effective collaboration of T and B cells. WASP might have a role in regulating changes in the actin cytoskeleton in response to external stimuli. It has the ability to bind SH3 domains, which, as we saw in Chapter 6, have an affinity for amino acid sequences rich in proline that are found on some proteins of intracellular signaling pathways. In WAS patients, and in mice whose WASP gene has been knocked out, T cells fail to respond normally to mitogens or to the cross-linking of surface receptors. Cytotoxic T-cell responses are also impaired, and T-cell help for B-cell responses to polysaccharide antigens is lacking. WASP is expressed in all hematopoietic cell lineages and is likely to be a key regulator of lymphocyte and platelet development and function.

11-14. The normal pathways for host defense against intracellular bacteria are illustrated by genetic deficiencies of IFN-γ and IL-12 and their receptors

A small number of families have been identified containing several individuals who suffer from persistent and eventually fatal attacks by intracellular pathogens, especially mycobacteria and salmonellae. Typically these patients suffer from the ubiquitous, environmental nontuberculous strains of myco-bacteria, such as Mycobacterium avium. They may also develop disseminated infection after vaccination with Mycobacterium bovis bacillus Calmette-Guérin, the strain of M. bovis that is used as a live vaccine against M. tuberculosis. The molecular bases of the susceptibility to these infections are null mutations in one of the following genes: IL-12, the IL-12 receptor β1 chain, or either of the two protein subunits, R1 and R2, of the receptor for IFN-γ. Similar susceptibility to intracellular bacterial infection is seen in mice with induced mutations in these same genes and also in mice lacking TNF-α or the TNF p55 receptor gene. All these genes must therefore play a critical part in the normal mechanisms of host defense against infection by these intracellular bacteria.

Mycobacteria and salmonellae enter dendritic cells and macrophages, where they can reproduce and multiply. At the same time they provoke an immune response that involves several stages and eventually controls the infection with the help of CD4 T cells (Fig. 11.16). First, lipoproteins from the surface of the bacteria ligate receptors on macrophages and dendritic cells as they enter the cells. These receptors include the Toll-like receptors (see Section 2-16), particularly TLR-2, and their ligation stimulates nitric oxide (NO) production within the cells, which is toxic to the bacteria. Signaling by these Toll-like receptors also stimulates the release of IL-12 which, in turn, stimulates CD4 T cells to release IFN-γ and TNF-α. These cytokines activate and recruit more mononuclear phagocytic cells to the site of infection, resulting in the formation of granulomas. The key role of IFN-γ in activating macrophages to kill intracellular bacteria is illustrated dramatically by the failure to control infection in patients who are genetically deficient in either of the two subunits of this receptor. In the total absence of IFN-γ receptor expression, granuloma formation is much reduced, showing a role for this receptor in the development of granulomas. In contrast, if the underlying mutation is associated with the presence of low levels of functional receptor, granulomas form, but the macrophages within the granulomas are not sufficiently activated to be able to control the division and spread of the mycobacteria. It is important to appreciate that this cascade of cytokine reactions is occurring in the context of cognate interactions between the macrophages and dendritic cells harboring the intracellular bacteria and antigen-specific CD4 T cells. T-cell receptor ligation and co-stimulation of the phagocyte by, for example, CD40–CD40 ligand interaction are important components that augment the capacity of T cells to effectively activate the infected phagocytes to kill the intracellular bacteria (see Sections 8-26 and 8-28).

11-15. X-linked lymphoproliferative syndrome is associated with fatal infection by Epstein-Barr virus and with the development of lymphomas

Epstein-Barr virus is a herpes virus that infects the majority of the human race and remains latent in B cells throughout life after primary infection. EBV infection can transform B lymphocytes and is used as a technique for immortalizing clones of B cells in the laboratory. This does not normally happen in vivo in humans because EBV infection is actively controlled and maintained in a latent state by cytotoxic T cells with specificity for B cells expressing EBV antigens (see Section 11-2). In the presence of T-cell immunodeficiency, this control mechanism can break down and a potentially lethal B-cell lymphoma may develop. One of the situations in which this occurs is the rare immuo-deficiency, X-linked lymphoproliferative syndrome, which results from mutations in a gene named SH2-domain containing gene 1A (SH2D1A). Boys with this deficiency typically develop overwhelming EBV infection during childhood, and sometimes lymphomas. EBV infection in this condition is usually fatal and is associated with necrosis of the liver. Thus SH2D1A must play a vital, nonredundant role in the normal control of EBV infection.

The function of SH2D1A is partly understood. The SH2 domain of the protein interacts with the cytoplasmic tails of two transmembrane receptors, SLAM and 2B4, which are structurally homologous to each other, and to the T-cell adhesion molecule CD2 (see Section 8-4). SLAM (signaling lymphocyte activation molecule) is expressed on activated T cells, whereas 2B4 is found on T cells, B cells, and NK cells. Activation of these receptors initiates a signaling pathway by the recruitment of the tyrosine phosphatase, SHP-2 (see Section 6-14). It appears that the function of SH2D1A is to inhibit the recruitment of SHP-2 and thereby to inhibit cellular activation by SLAM and 2B4. There are two hypotheses to explain the pathogenesis of the fatal EBV infection seen in children with defects in SH2D1A. The first is that failure of T cells to kill B cells expressing antigens from multiplying EBV allows uncontrolled infection. The second is that B cells presenting EBV peptides uncontrollably activate T cells and that cytotoxic T cells cause tissue necrosis and death. Some cases of lymphoma in young boys have now been found associated with mutations in the SH2D1A gene in the absence of any evidence of EBV infection. This raises the possibility that SH2D1A may be a tumor suppressor gene in its own right, in addition to controlling a virus that can contribute to tumor formation.

11-16. Bone marrow transplantation or gene therapy can be useful to correct genetic defects

It is frequently possible to correct the defects in lymphocyte development that lead to the SCID phenotype by replacing the defective component, generally by bone marrow transplantation. The major difficulties in these therapies result from MHC polymorphism. To be useful, the graft must share some MHC alleles with the host. As we learned in Section 7-20, the MHC alleles expressed by the thymic epithelium determine which T cells can be positively selected. When bone marrow cells are used to restore immune function to individuals with a normal thymic stroma, both the T cells and the antigen-presenting cells are derived from the graft. Therefore, unless the graft shares at least some MHC alleles with the recipient, the T cells that are selected on host thymic epithelium cannot be activated by graft-derived antigen-presenting cells (Fig. 11.17). There is also a danger that mature, post-thymic T cells in donor bone marrow might recognize the host as foreign and attack it, causing graft-versus-host disease (GVHD) (Fig. 11.18, top panel). This can be overcome by depleting the donor bone marrow of mature T cells. Bone marrow recipients are usually treated with irradiation that kills their own lymphocytes, thus making space for the grafted bone marrow cells and minimizing the threat of host-versus-graft disease (HVGD) (Fig. 11.18, third panel). In patients with the SCID phenotype, however, there is little problem with the host response to the transplanted bone marrow, as the patient is immunodeficient.

Now that specific gene defects are being identified, a different approach to correcting these inherited immune deficiencies can be attempted. The strategy involves extracting a sample of the patient's own bone marrow cells, inserting a normal copy of the defective gene into them, and returning them to the patient by transfusion. This approach, called somatic gene therapy, should correct the gene defect. Moreover, in immunodeficient patients, it might be possible to reinfuse the bone marrow into the patient without the usual irradiation used to suppress the recipient's bone marrow function. There is no risk of graft-versus-host disease in this case, although the host might respond to the replaced gene product and reject the engineered cells. Although this kind of approach is theoretically attractive, efficient transfer of genes into bone marrow stem cells is technically difficult and has been achieved only in mouse models. The first trials of gene therapy for correcting immunodeficiency, such as the treatment of a child with ADA deficiency at the National Institute of Health (NIH) in 1990, used the patient's lymphocytes as the vehicle for gene introduction. However, because most lymphocytes divide regularly, thus diluting out the new gene, the treatment had to be repeated regularly. In another study, bone marrow stem cells, obtained from cord blood from three patients with ADA deficiency were transduced with the ADA gene and reinfused. At the age of 4 years, these children expressed up to 10% of normal ADA levels only in T cells and not in other bone marrow-derived cells, and they remained immunodeficient in the absence of treatment with ADA enzyme replacement. More recently, however, there has been a successful attempt at correcting the phenotype of two X-linked SCID patients using a Moloney retrovirus-derived construct containing the γ_c chain to infect bone marrow stem cells.

Summary

Genetic defects can occur in almost any molecule involved in the immune response. These defects give rise to characteristic deficiency diseases, which, although rare, provide a great deal of information about the development and functioning of the immune system in normal humans. Inherited immuo-deficiencies illustrate the vital role of the adaptive immune response and T cells in particular, without which both cell-mediated and humoral immunity fail. They have provided information about the separate roles of B lymphocytes in humoral immunity and of T lymphocytes in cell-mediated immunity, the importance of phagocytes and complement in humoral and innate immunity, and the specific functions of several cell-surface or signaling molecules in the adaptive immune response. There are also some inherited immune disorders whose causes we still do not understand. The study of these diseases will undoubtedly teach us more about the normal immune response and its control.

11-17. Most individuals infected with HIV progress over time to AIDS

Many viruses cause an acute but limited infection inducing lasting protective immunity. Others, such as herpes viruses, set up a latent infection that is not eliminated but is controlled adequately by an adaptive immune response. However, infection with HIV seems rarely, if ever, to lead to an immune response that can prevent ongoing replication of the virus. Although the initial acute infection does seem to be controlled by the immune system, HIV continues to replicate and infect new cells.

The initial infection with HIV generally occurs after transfer of body fluids from an infected person to an uninfected one. The virus is carried in infected CD4 T cells, dendritic cells, and macrophages, and as a free virus in blood, semen, vaginal fluid, or milk. It is most commonly spread by sexual intercourse, contaminated needles used for intravenous drug delivery, and the therapeutic use of infected blood or blood products, although this last route of transmission has largely been eliminated in the developed world where blood products are screened routinely for the presence of HIV. An important route of virus transmission is from an infected mother to her baby at birth or through breast milk. In Africa, the perinatal transmission rate is approximately 25%, but this can largely be prevented by treating infected pregnant women with the drug zidovudine (AZT) (see Section 11-23). Mothers who are newly infected and breastfeed their infants transmit HIV 40% of the time, showing that HIV can also be transmitted in breast milk, but this is less common after the mother produces antibodies to HIV.

Primary infection with HIV is probably asymptomatic in 50% of cases but often causes an influenza-like illness with an abundance of virus in the peripheral blood and a marked drop in the numbers of circulating CD4 T cells. This acute viremia is associated in virtually all patients with the activation of CD8 T cells, which kill HIV-infected cells, and subsequently with antibody production, or seroconversion. The cytotoxic T-cell response is thought to be important in controlling virus levels, which peak and then decline, as the CD4 T-cell counts rebound to around 800 cells μl^-1 (the normal value is 1200 cells μl^-1). At present, the best indicator of future disease is the level of virus that persists in the blood plasma once the symptoms of acute viremia have passed.

Most patients who are infected with HIV will eventually develop AIDS, after a period of apparent quiescence of the disease known as clinical latency or the asymptomatic period (Fig. 11.20). This period is not silent, however, for there is persistent replication of the virus, and a gradual decline in the function and numbers of CD4 T cells until eventually patients have few CD4 T cells left. At this point, which can occur anywhere between 2 and 15 years or more after the primary infection, the period of clinical latency ends and opportunistic infections begin to appear.

The typical course of an infection with HIV is illustrated in Fig. 11.21. However, it has become increasingly clear that the course of the disease can vary widely. Thus, although most people infected with HIV go on to develop AIDS and ultimately to die of opportunistic infection or cancer, this is not true of all individuals. A small percentage of people seroconvert, making antibodies against many HIV proteins, but do not seem to have progressive disease, in that their CD4 T-cell counts and other measures of immune competence are maintained. These long-term nonprogressors have unusually low levels of circulating virus and are being studied intensively to determine how they are able to control their HIV infection. A second group consists of seronegative people who have been highly exposed to HIV yet remain disease-free and virus-negative. Some of these people have specific cytotoxic lymphocytes and T_H1 lymphocytes directed against infected cells, which confirms that they have been exposed to HIV or possibly noninfectious HIV antigens. It is not clear whether this immune response accounts for clearing the infection, but it is a focus of considerable interest for the development and design of vaccines, which we will discuss later. There is a small group of people who are resistant to HIV infection because they carry mutations in a cell-surface receptor that is used as a co-receptor for viral entry, as we will see below.

We will return to discuss in more detail the interactions of HIV with the immune system and the prospects for manipulating them later in this chapter, but before doing so we must describe the viral life cycle and the genes and proteins on which it depends. Some of these proteins are the targets of the most successful drugs in use at present for the treatment of AIDS.

11-18. HIV is a retrovirus that infects CD4 T cells, dendritic cells, and macrophages

HIV is an enveloped retrovirus whose structure is shown in Fig. 11.22. Each virus particle, or virion, contains two copies of an RNA genome, which are transcribed into DNA in the infected cell and integrated into the host cell chromosome. The RNA transcripts produced from the integrated viral DNA serve both as mRNA to direct the synthesis of the viral proteins and later as the RNA genomes of new viral particles, which escape from the cell by budding from the plasma membrane, each in a membrane envelope. HIV belongs to a group of retroviruses called the lentiviruses, from the Latin lentus, meaning slow, because of the gradual course of the diseases that they cause. These viruses persist and continue to replicate for many years before causing overt signs of disease.

The ability of HIV to enter particular types of cell, known as the cellular tropism of the virus, is determined by the expression of specific receptors for the virus on the surface of those cells. HIV enters cells by means of a complex of two noncovalently associated viral glycoproteins, gp120 and gp41, in the viral envelope. The gp120 portion of the glycoprotein complex binds with high affinity to the cell-surface molecule CD4. This glycoprotein thereby draws the virus to CD4 T cells and to dendritic cells and macrophages, which also express some CD4. Before fusion and entry of the virus, gp120 must also bind to a co-receptor in the membrane of the host cell. Several different molecules may serve as a co-receptor for HIV entry, but in each case they have been identified as chemokine receptors. The chemokine receptors (see Chapters 2 and 10) are a closely related family of G protein-coupled receptors with seven transmembrane-spanning domains. Two chemokine receptors, known as CCR5, which is predominantly expressed on dendritic cells, macrophages, and CD4 T cells, and CXCR4, expressed on activated T cells, are the major co-receptors for HIV. After binding of gp120 to the receptor and co-receptor, the gp41 then causes fusion of the viral envelope and the plasma membrane of the cell, allowing the viral genome and associated viral proteins to enter the cytoplasm.

There are different variants of HIV, and the cell types that they infect are determined to a large degree by which chemokine receptor they bind as co-receptor. The variants of HIV that are associated with primary infections use CCR5, which binds the CC chemokines RANTES, MIP-1α, and MIP-1β (see Chapter 2), as a co-receptor, and require only a low level of CD4 on the cells they infect. These variants of HIV infect dendritic cells, macrophages, and T cells in vivo. However, they are often described simply as ‘macrophage-tropic’ because they infect macrophage but not T-cell lines in vitro and the cell tropism of different HIV variants was originally defined by their ability to grow in different cell lines.

In contrast, ‘lymphocyte-tropic’ variants of HIV infect only CD4 T cells in vivo and use CXCR4, which binds the CXC chemokine stromal-derived factor-1 (SDF-1), as a co-receptor. The lymphocyte-tropic variants of HIV can grow in vitro in T-cell lines, and require high levels of CD4 on the cells that they infect.

It appears that macrophage-tropic isolates of HIV are preferentially transmitted by sexual contact as they are the dominant viral phenotype found in newly infected individuals. Virus is disseminated from an initial reservoir of infected dendritic cells and macrophages and there is evidence for an important role for mucosal lymphoid tissue in this process. Mucosal epithelia, which are constantly exposed to foreign antigens, provide a milieu of immune system activity in which HIV replication occurs readily. Infection of CD4 T cells via CCR5 occurs early in the course of infection and continues to occur, with activated CD4 T cells accounting for the major production of HIV throughout infection. Late in infection, in approximately 50% of cases, the viral phenotype switches to a T-lymphocyte-tropic type that utilizes CXCR4 co-receptors, and this is followed by a rapid decline in CD4 T-cell count and progression to AIDS.

11-19. Genetic deficiency of the macrophage chemokine co-receptor for HIV confers resistance to HIV infection in vivo

Further evidence for the importance of chemokine receptors in HIV infection has come from studies in a small group of individuals with high-risk exposure to HIV-1 but who remain seronegative. Cultures of lymphocytes and macrophages from these people were relatively resistant to macrophage-tropic HIV infection and were found to secrete high levels of RANTES, MIP-1α and MIP-1β in response to inoculation with HIV. In other experiments, the addition of these same chemokines to lymphocytes sensitive to HIV blocked their infection because of competition between these CC chemokines and the virus for the cell-surface receptor CCR5.

The resistance of these rare individuals to HIV infection has now been explained by the discovery that they are homozygous for an allelic, nonfunctional variant of CCR5 caused by a 32-base-pair deletion from the coding region that leads to a frameshift and truncation of the translated protein. The gene frequency of this mutant allele in Caucasoid populations is quite high at 0.09 (meaning that about 10% of the Caucasoid population are heterozygous carriers of the allele and about 1% are homozygous). The mutant allele has not been found in Japanese or black Africans from Western or Central Africa. Heterozygous deficiency of CCR5 might provide some protection against sexual transmission of HIV infection and a modest reduction in the rate of progression of the disease. In addition to the structural polymorphism of the gene, variation of the promoter region of the CCR5 gene has been found in both Caucasian and African Americans. Different promoter variants were associated with different rates of progression of disease.

These results provide a dramatic confirmation of experimental work suggesting that CCR5 is the major macrophage and T-lymphocyte co-receptor used by HIV to establish primary infection in vivo, and offers the possibility that primary infection might be blocked by therapeutic antagonists of the CCR5 receptor. Indeed, there is preliminary evidence that low molecular weight inhibitors of this receptor can block infection of macrophages by HIV in vitro. Such low molecular weight inhibitors might be the precursors of useful drugs that could be taken by mouth. Such drugs are very unlikely to provide complete protection against infection, as a very small number of individuals who are homozygous for the nonfunctional variant of CCR5 are infected with HIV. These individuals seem to have suffered from primary infection by CXCR4-using strains of the virus.

11-20. HIV RNA is transcribed by viral reverse transcriptase into DNA that integrates into the host cell genome

One of the proteins that enters the cell with the viral genome is the viral reverse transcriptase, which transcribes the viral RNA into a complementary DNA (cDNA) copy. The viral cDNA is then integrated into the host cell genome by the viral integrase, which also enters the cell with the viral RNA. The integrated cDNA copy is known as the provirus. The infectious cycle up to the integration of the provirus is shown in Fig. 11.23. In activated CD4 T cells, virus replication is initiated by transcription of the provirus, as we will see in the next section. However, HIV can, like other retroviruses, establish a latent infection in which the provirus remains quiescent. This seems to occur in memory CD4 T cells and in dormant macrophages, and these cells are thought to be an important reservoir of infection.

The entire HIV genome consists of nine genes flanked by long terminal repeat sequences (LTRs), which are required for the integration of the provirus into the host cell DNA and contain binding sites for gene regulatory proteins that control the expression of the viral genes. Like other retroviruses, HIV has three major genes—gag, pol, and env. The gag gene encodes the structural proteins of the viral core, pol encodes the enzymes involved in viral replication and integration, and env encodes the viral envelope glycoproteins. The gag and pol mRNAs are translated to give polyproteins—long polypeptide chains that are then cleaved by the viral protease (also encoded by pol) into individual functional proteins. The product of the env gene, gp160, has to be cleaved by a host cell protease into gp120 and gp41, which are then assembled as trimers into the viral envelope. As shown in Fig. 11.24, HIV has six other, smaller, genes encoding proteins that affect viral replication and infectivity in various ways. We will discuss the function of two of these—Tat and Rev—in the following section.

11-21. Transcription of the HIV provirus depends on host cell transcription factors induced upon the activation of infected T cells

The production of infectious virus particles from an integrated HIV provirus is stimulated by a cellular transcription factor that is present in all activated T cells. Activation of CD4 T cells induces the transcription factor NFκB, which binds to promoters not only in the cellular DNA but also in the viral LTR, thereby initiating the transcription of viral RNA by the cellular RNA polymerase. This transcript is spliced in various ways to produce mRNAs for the viral proteins. The Gag and Gag-Pol proteins are translated from unspliced mRNA; Vif, Vpr, Vpu, and Env are translated from singly spliced viral mRNA; Tat, Rev, and Nef are translated from multiply spliced mRNA. At least two of the viral genes, tat and rev, encode proteins, Tat and Rev respectively, that promote viral replication in activated T cells. Tat is a potent transcriptional regulator, which functions as an elongation factor that enables the transcription of viral RNA by the RNA polymerase II complex. Tat contains two binding sites, contained in one domain, named the transactivation domain. The first of these allows Tat to bind to a host cellular protein, cyclin T1. This binding reaction promotes the binding of the Tat protein through the second binding site in its transactivation domain to an RNA sequence in the LTR of the virus known as the transcriptional activation region (TAR). The consequence of this interaction is to greatly enhance the rate of viral genome transcription, by causing the removal of negative elongation factors that block the transcriptional activity of RNA polymerase II. The expression of cyclin T1 is greatly increased in activated compared with quiescent T lymphocytes. This, in conjunction with the increased expression of NFκB in activated T cells, may explain the ability of HIV to lie dormant in resting T cells and replicate in activated T cells (Fig. 11.25).

Eukaryotic cells have mechanisms to prevent the export from the cell nucleus of incompletely spliced mRNA transcripts. This could pose a problem for a retrovirus that is dependent on the export of unspliced, singly spliced, and multiply spliced mRNA species in order to translate the full complement of viral proteins. The Rev protein is the viral solution to this problem. Export from the nucleus and translation of the three HIV proteins encoded by the fully spliced mRNA transcripts, Tat, Nef, and Rev, occurs early after viral infection by means of the normal host cellular mechanisms of mRNA export. The expressed Rev protein then enters the nucleus and binds to a specific viral RNA sequence, the Rev response element (RRE). Rev also binds to a host nucleocytoplasmic transport protein named Crm1, which engages a host pathway for exporting mRNA species through nuclear pores into the cytoplasm.

When the provirus is first activated, Rev levels are low, the transcripts are translocated slowly from the nucleus, and thus multiple splicing events can occur. Thus, more Tat and Rev are produced, and Tat in turn ensures that more viral transcripts are made. Later, when Rev levels have increased, the transcripts are translocated rapidly from the nucleus unspliced or only singly spliced. These unspliced or singly spliced transcripts are translated to produce the structural components of the viral core and envelope, together with the reverse transcriptase, the integrase, and the viral protease, all of which are needed to make new viral particles. The complete, unspliced transcripts that are exported from the nucleus late in the infectious cycle are required for the translation of gag and pol and are also destined to be packaged with the proteins as the RNA genomes of the new virus particles.

11-22. Drugs that block HIV replication lead to a rapid decrease in titer of infectious virus and an increase in CD4 T cells

Studies with powerful drugs that completely block the cycle of HIV replication indicate that the virus is replicating rapidly at all phases of infection, including the asymptomatic phase. Two viral proteins in particular have been the target of drugs aimed at arresting viral replication. These are the viral reverse transcriptase, which is required for synthesis of the provirus, and the viral protease, which cleaves the viral polyproteins to produce the virion proteins and viral enzymes. Inhibitors of these enzymes prevent the establishment of further infection in uninfected cells. Cells that are already infected can continue to produce virions because, once the provirus is established, reverse transcriptase is not needed to make new virus particles, while the viral protease acts at a very late maturation step of the virus, and inhibition of the protease does not prevent virus from being released. However, in both cases, the released virions are not infectious and further cycles of infection and replication are prevented.

Because of the great efficacy of the protease inhibitors, it is possible to learn much about the kinetics of HIV replication in vivo by measuring the decline in viremia after the initiation of protease inhibitor therapy. For the first 2 weeks after starting treatment there is an exponential fall in plasma virus levels with a half-life of viral decay of about 2 days (Fig. 11.26). This phase reflects the decay in virus production from cells that were actively infected at the start of drug treatment, and indicates that the half-life of productively infected cells is similarly about 2 days. The results also show that free virus is cleared from the circulation very rapidly, with a half-life of about 6 hours. After 2 weeks, levels of virus in plasma have dropped by more than 95%, representing an almost total loss of productively infected CD4 lymphocytes. After this time, the rate of decline of plasma virus levels is much slower, reflecting the very slow decay of virus production from cells that provide a longer-lived reservoir of infection, such as dendritic cells and tissue macrophages, and from latently infected memory CD4 T cells that have been activated. Very long-term sources of infection might be CD4 memory T cells that continue to carry integrated provirus, and virus stored as immune complexes on follicular dendritic cells. These very long-lasting reservoirs of infection might prove to be resistant to drug therapy for HIV.

These studies show that most of the HIV present in the circulation of an infected individual is the product of rounds of replication in newly infected cells, and that virus from these productively infected cells is released into, and rapidly cleared from, the circulation at the rate of 10⁹ to 10¹⁰ virions every day. This raises the question of what is happening to these virus particles: how are they removed so rapidly from the circulation? It seems most likely that HIV particles are opsonized by specific antibody and complement and removed by phagocytic cells of the mononuclear phagocyte system. Opsonized HIV particles can also be trapped on the surface of follicular dendritic cells, which are known to capture antigen:antibody complexes and retain them for prolonged periods (see Chapters 9 and 10).

The other issue raised by these studies is the effect of HIV replication on the population dynamics of CD4 T cells. The decline in plasma viremia is accompanied by a steady increase in CD4 T lymphocyte counts in peripheral blood: what is the source of the new CD4 T cells that appear once treatment is started? It seems highly unlikely that they are the recent progeny of stem cells that have developed in the thymus, because CD4 T cells are not normally produced in large numbers from the thymus even at its maximum rate of production in adolescents. Some investigators believe that these cells are emerging from sites of sequestration and add little to the total numbers of CD4 T cells in the body, whereas others advocate their origin from mature CD4 T cells that replicate, and argue that the production of such cells is an ongoing process that compensates for the continual loss of productively infected CD4 T cells.

11-23. HIV accumulates many mutations in the course of infection in a single individual and drug treatment is soon followed by the outgrowth of drug-resistant variants of the virus

The rapid replication of HIV, with the generation of 10⁹ to 10¹⁰ virions every day, coupled with a mutation rate of approximately 3 × 10^-5 per nucleotide base per cycle of replication, leads to the generation of many variants of HIV in a single infected patient in the course of one day. Replication of a retroviral genome depends on two error-prone steps. Reverse transcriptase lacks the proofreading mechanisms associated with cellular DNA polymerases, and the RNA genomes of retroviruses are therefore copied into DNA with relatively low fidelity; the transcription of the proviral DNA into RNA copies by the cellular RNA polymerase is similarly a low-fidelity process. A rapidly replicating persistent virus that is going through these two steps repeatedly in the course of an infection can thereby accumulate many mutations, and numerous variants of HIV, sometimes called quasi-species, are found within a single infected individual. This very high variability was first recognized in HIV and has since proved to be common to the other lentiviruses.

As a consequence of its high variability, HIV rapidly develops resistance to antiviral drugs. When antiviral drugs are administered, variants of the virus that carry mutations conferring resistance to their effects emerge and expand until former levels of plasma virus are regained. Resistance to some of the protease inhibitors appears after only a few days (Fig. 11.27). Resistance to some of the nucleoside analogues that are potent inhibitors of reverse transcriptase develops in a similarly short time. By contrast, resistance to the nucleoside zidovudine (AZT), the first drug to be widely used for treating AIDS, takes months to develop. This is not because AZT is a more powerful inhibitor, but because resistance to zidovudine requires three or four mutations in the viral reverse transcriptase, whereas a single mutation can confer resistance to the protease inhibitors and other reverse-transcriptase inhibitors. As a result of the relatively rapid appearance of resistance to all known anti-HIV drugs, successful drug treatment might depend on the development of a range of antiviral drugs that can be used in combination. It might also be important to treat early in the course of an infection, thereby reducing the chances that a variant virus has accumulated all the necessary mutations to resist the entire cocktail. Current treatments follow this strategy and use combinations of viral protease inhibitors together with nucleoside analogues (see Fig. 11.26).

11-24. Lymphoid tissue is the major reservoir of HIV infection

Although viral load and turnover are usually measured by detecting the viral RNA present in viral particles in the blood, the major reservoir of HIV infection is in lymphoid tissue, in which infected CD4 T cells, monocytes, macrophages, and dendritic cells are found. In addition, HIV is trapped in the form of immune complexes on the surface of follicular dendritic cells. These cells are not themselves infected but may act as a store of infective virions.

HIV infection takes different forms within different cells. As we have seen, more than 95% of the virus that can be detected in the plasma is derived from productively infected cells, which have a very short half-life of about 2 days. Productively infected CD4 lymphocytes are found in the T-cell areas of lymphoid tissue, and these are thought to succumb to infection in the course of being activated in an immune response. Latently infected memory CD4 cells that are activated in response to antigen presentation also produce virus. Such cells have a longer half-life of 2 to 3 weeks from the time that they are infected. Once activated, HIV can spread from these cells by rounds of replication in other activated CD4 T cells. In addition to the cells that are infected productively or latently, there is a further large population of cells infected by defective proviruses; such cells are not a source of infectious virus.

Macrophages and dendritic cells seem to be able to harbor replicating virus without necessarily being killed by it, and are therefore believed to be an important reservoir of infection, as well as a means of spreading virus to other tissues such as the brain. Although the function of macrophages as antigen-presenting cells does not seem to be compromised by HIV infection, it is thought that the virus causes abnormal patterns of cytokine secretion that could account for the wasting that commonly occurs in AIDS patients late in their disease.

11-25. An immune response controls but does not eliminate HIV

Infection with HIV generates an adaptive immune response that contains the virus but only very rarely, if ever, eliminates it. The time course of various elements in the adaptive immune response to HIV is shown, together with the levels of infectious virus in plasma, in Fig. 11.28.

Seroconversion is the clearest evidence for an adaptive immune response to infection with HIV, but the generation of T lymphocytes responding to infected cells is thought by most workers in the field to be central in controlling the infection. Both CD8 cytotoxic T cells and T_H1 cells specifically responsive to infected cells are associated with the decline in detectable virus after the initial infection. These T-cell responses are unable to clear the infection completely and can cause some pathology. Nevertheless, there is evidence that the virus itself is cytopathic, and T-cell responses that reduce viral spread should therefore, on balance, reduce the pathology of the disease.

The ability of cytotoxic T lymphocytes to destroy HIV-infected cells is demonstrated by studies of peripheral blood cells from infected individuals, in which cytotoxic T cells specific for viral peptides can be shown to kill infected cells in vitro. In vivo, cytotoxic T cells can be seen to invade sites of HIV replication and they could, in theory, be responsible for killing many productively infected cells before any infectious virus can be released, thereby containing viral load at the quasi-stable levels that are characteristic of the asymptomatic period. The best evidence for the clinical importance of the control of HIV-infected cells by CD8 cytotoxic T cells comes from studies relating the numbers and activity of CD8 T cells to viral load. An inverse correlation was found between the number of CD8 T cells carrying a receptor specific for an HLA-A2-restricted HIV peptide and plasma RNA viral load. Similarly, patients with high levels of HIV-specific CD8 T cells showed slower progression of disease than those with low levels. There is also direct evidence from experiments in macaques infected with simian immunodeficiency virus (SIV) that CD8 cytotoxic T cells control retrovirally-infected cells in vivo. Treatment of infected animals with depleting anti-CD8 monoclonal antibodies was followed by a large increase in viral load.

Mutations that occur as HIV replicates can allow variants of the virus to escape recognition by antibody or cytotoxic T cells and can contribute to the failure of the immune system to contain the infection in the long term. Direct escape of virus-infected cells from killing by cytotoxic T lymphocytes has been shown by the occurrence of mutations of immunodominant viral peptides presented by MHC class I molecules. In other cases, variant peptides produced by the virus have been found to act as antagonists (see Section 6-12) for T cells responsive to the wild-type epitope, thus allowing both mutant and wild-type viruses to survive. Mutant peptides acting as antagonists have also been reported in hepatitis B virus infections, and similar mutant peptides might contribute to the persistence of some viral infections, especially when, as often happens, the immune response of an individual is dominated by T cells specific for a particular epitope.

11-26. HIV infection leads to low levels of CD4 T cells, increased susceptibility to opportunistic infection, and eventually to death

There are three dominant mechanisms for the loss of CD4 T cells in HIV infection. First, there is evidence for direct viral killing of infected cells; second, there is increased susceptibility to the induction of apoptosis in infected cells; and third, there is killing of infected CD4 T cells by CD8 cytotoxic lymphocytes that recognize viral peptides.

When CD4 T-cell numbers decline below a critical level, cell-mediated immunity is lost, and infections with a variety of opportunistic microbes appear (Fig. 11.29). Typically, resistance is lost early to oral Candida species and to Mycobacterium tuberculosis, which shows as an increased prevalence of thrush (oral candidiasis) and tuberculosis. Later, patients suffer from shingles, caused by the activation of latent herpes zoster, from EBV-induced B-cell lymphomas, and from Kaposi's sarcoma, a tumor of endothelial cells that probably represents a response both to cytokines produced in the infection and to a novel herpes virus called HHV-8 that was identified in these lesions. Pneumonia caused by the fungus Pneumocystis carinii is common and often fatal. In the final stages of AIDS, infection with cytomegalovirus or Mycobacterium avium complex is more prominent. It is important to note that not all patients with AIDS get all these infections or tumors, and there are other tumors and infections that are less prominent but still significant. Rather, this is a list of the commonest opportunistic infections and tumors, most of which are normally controlled by robust CD4 T cell-mediated immunity that wanes as the CD4 T-cell counts drop toward zero (see Fig. 11.21).

11-27. Vaccination against HIV is an attractive solution but poses many difficulties

A safe and effective vaccine for the prevention of HIV infection and AIDS is an attractive goal, but its achievement is fraught with difficulties that have not been faced in developing vaccines against other diseases. The first problem is the nature of the infection itself, featuring a virus that proliferates extremely rapidly and causes sustained infection in the face of strong cytotoxic T-cell and antibody responses. As we discussed in Section 11-25, HIV evolves in individual patients by the selective proliferative advantage of mutant virions encoding peptide sequence changes that escape recognition by antibodies and by cytotoxic T lymphocytes. This evolution means that the development of therapeutic vaccination strategies to block the development of AIDS in HIV-infected patients will be extremely difficult. Even after the viremia has been largely cleared by drug therapy, immune responses to HIV fail to prevent drug-resistant virus from rebounding and replicating at pretreatment levels.

The second problem is our uncertainty over what form protective immunity to HIV might take. It is not known whether antibodies, cytotoxic T lymphocyte responses, or both are necessary to achieve protective immunity, and which epitopes might provide the targets of protective immunity. Third, if strong cytotoxic responses are necessary to provide protection against HIV, these might be difficult to develop and sustain through vaccination. Other effective viral vaccines rely on the use of live, attenuated viruses and there are concerns over the safety of pursuing this approach for HIV. Another possible approach is the use of DNA vaccination, a technique that we discuss in Section 14-25. Both of these approaches are being tested in animal models.

The fourth problem is the ability of the virus to persist in latent form as a transcriptionally silent provirus, which is invisible to the immune system. This might prevent the immune system from clearing the infection once it has been established. In summary, the ability of the immune system to clear infectious virus remains uncertain.

However, against this pessimistic background, there are grounds for hope that successful vaccines can be developed. Of particular interest are rare groups of people who have been exposed often enough to HIV to make it virtually certain that they should have become infected but who have not developed the disease. In some cases this is due to an inherited deficiency in the chemokine receptor used as co-receptor for HIV entry, as we explained in Section 11-19. However, this mutant chemokine receptor does not occur in Africa, where one such group has been identified. A small group of Gambian and Kenyan prostitutes who are estimated to have been exposed to many HIV-infected male partners each month for up to 5 years were found to lack antibody responses but to have cytotoxic T lymphocyte responses to a variety of peptide epitopes from HIV. These women seem to have been naturally immunized against HIV.

Although there is no perfect animal model for the development of HIV vaccines, one model system is based on simian immunodeficiency virus (SIV), which is closely related to HIV and infects macaques. SIV causes a similar disease to AIDS in Asian macaques such as the cynomolgus monkey, but does not cause disease in African cercopithecus monkeys such as the African green monkey, with which SIV has probably coexisted for up to a million years. Live attenuated SIV vaccines lacking the nef gene, and hybrid HIV-SIV viruses have been developed to test the principles of vaccination in primates, and both have proved successful in protecting primates against subsequent infection by fully virulent viruses. However, there are substantial difficulties to be overcome in the development of live attenuated HIV vaccines for use in at-risk populations, not least the worry of recombination between vaccine strains and wild-type viruses leading to reversion to a virulent phenotype. The alternative approach of DNA vaccination is being piloted in primate experiments, with some early signs of success.

Subunit vaccines, which induce immunity to only some proteins in the virus, have also been made. One such vaccine has been made from the envelope protein gp120 and has been tested on chimpanzees. This vaccine proved to be specific to the precise strain of virus used to make it, and was therefore useless in protection against natural infection. Subunit vaccines are also less efficient at inducing prolonged cytotoxic T-cell responses.

Finally, there are difficult ethical issues in the development of a vaccine. It would be unethical to conduct a vaccine trial without trying at the same time to minimize the exposure of a vaccinated population to the virus itself. However, the effectiveness of a vaccine can only be assessed in a population in which the exposure rate to the virus is high enough to assess whether vaccination is protective against infection. This means that initial vaccine trials might have to be conducted in countries where the incidence of infection is very high and public health measures have not yet succeeded in reducing the spread of HIV.

11-28. Prevention and education are one way in which the spread of HIV and AIDS can be controlled

The one way in which we know we can protect against infection with HIV is by avoiding contact with body fluids, such as semen, blood, blood products, or milk from people who are infected. Indeed, it has been demonstrated repeatedly that this precaution, simple enough in the developed world, is sufficient to prevent infection, as health-care workers can take care of AIDS patients for long periods without seroconversion or signs of infection.

For this strategy to work, however, one must be able to test people at risk of infection with HIV periodically, so that they can take the steps necessary to avoid passing the virus to others. This, in turn, requires strict confidentiality and mutual trust. A barrier to the control of HIV is the reluctance of individuals to find out whether they are infected, especially as one of the consequences of a positive HIV test is stigmatization by society. As a result, infected individuals can unwittingly infect many others. Balanced against this is the success of therapy with combinations of the new protease inhibitors and reverse transcriptase inhibitors, which provides an incentive for potentially infected people to identify the presence of infection and gain the benefits of treatment. Responsibility is at the heart of AIDS prevention, and a law guaranteeing the rights of people infected with HIV might go a long way to encouraging responsible behavior. The rights of HIV-infected people are protected in the Netherlands and Sweden. The problem in the less-developed nations, where elementary health precautions are extremely difficult to establish, is more profound.

Summary

Infection with the human immunodeficiency virus (HIV) is the cause of acquired immune deficiency syndrome (AIDS). This worldwide epidemic is now spreading at an alarming rate, especially through heterosexual contact in less-developed countries. HIV is an enveloped retrovirus that replicates in cells of the immune system. Viral entry requires the presence of CD4 and a particular chemokine receptor, and the viral cycle is dependent on transcription factors found in activated T cells. Infection with HIV causes a loss of CD4 T cells and an acute viremia that rapidly subsides as cytotoxic T-cell responses develop, but HIV infection is not eliminated by this immune response. HIV establishes a state of persistent infection in which the virus is continually replicating in newly infected cells. The current treatment consists of combinations of viral protease inhibitors together with nucleoside analogues and causes a rapid decrease in virus levels and a slower increase in CD4 T-cell counts. The main effect of HIV infection is the destruction of CD4 T cells, which occurs through the direct cytopathic effects of HIV infection and through killing by CD8 cytotoxic T cells. As the CD4 T-cell counts wane, the body becomes progressively more susceptible to opportunistic infection with intracellular microbes. Eventually, most HIV-infected individuals develop AIDS and die; however a small minority (3–7%), remain healthy for many years, with no apparent ill effects of infection. We hope to be able to learn from these individuals how infection with HIV can be controlled. The existence of such people and other people who have been naturally immunized against infection gives hope that it will be possible to develop effective vaccines against HIV.