REVIEW: Mechanisms of Bacterial Pathogenesis and Targets for Vaccine Design
With most pathogenic and etiologic agents of deadly diseases acquiring resistance to currently used drugs, the development and formulation of vaccines against predominant infectious diseases has taken centre stage. Our review aims, based on our understanding and knowledge of the process of pathogenesis and its intricacies in particular, to throw light on the aspects of vaccine development. The basic platform for formulating a vaccine involves deciphering the kinds of immune responses to the various antigenic factors of the pathogen. Other particulars, such as immunological memory (primarily mediated by antibodies post-natally), immunopathology of infections, and the selective balance of Th1-Th2 responses are also considered grounding factors for the construction of "immunity eliciting vaccines". The nature of vaccines (either humoral antibody immunity inducing or cell-mediated immunity inducing) depends on the location (extracellular or intracellular) and the expression of the antigens selected for incorporation. To maintain the synergism between the kind of immunity conferred by the vaccines and the cellular location of the included antigens, new findings are gathered about the virulence factors such as toxins, adhesins, invasins (mostly enzymes), anti-apoptotic factors, anti-phagocytic factors, and many more molecules that aid in pathogenesis and invasiveness. The concept of an all-in-one vaccine versus a one-in-one vaccine is also discussed. Bioinformatics tools and algorithms like BLAST and FASTA are nowadays being adopted as a means to identify and detect common target antigens against Gram negative bacterial pathogens. One of the more recently identified and well-studied gram negative bacterial candidates has been Choline phosphorylase. Homology searches have thus facilitated the discovery of potential candidates for an all-in-one vaccine.
When a microbe enters the body, the immune system responds through a diverse set of mechanisms in an attempt to eliminate the infectious agent. These immune responses can be segregated into two compartments, namely the innate and adaptive immune systems. The innate response relies on immediate recognition of antigenic structures common to many microbes by a selected set of immune cells with rapid effector function [Hoffman et al. 1999, Medzhitov and Janeway 1997a, Medzhitov and Janeway 2000, Medzhitov and Janeway 1997b]. In contrast, the adaptive immune response is made up of B and T lymphocytes that have unique receptors specific to various microbial antigens. These antigen-specific receptors are encoded by genes generated during a complex of gene rearrangement that occurs during the course of lymphocyte development. As each B and T lymphocyte contains a unique antigenic receptor, it allows for a large and diverse population of cells capable of recognizing a wide spectrum of pathogens. Although the size and diversity of the lymphocyte repertoire make it likely that there is an antigen, a specific lymphocyte for any given pathogen, the frequency of these cells can be extremely low and normally will not be sufficient to protect the host against a primary infection. After antigenic stimulation, there is activation and expansion of these antigen-specific cells. It is this process of clonal selection and the ultimate perpetuation of these antigen-specific memory cells that protects against a secondary infection. This primary immune response to a foreign antigen takes a week to develop, during which time infecting microbes can replicate within the host body. Once primed, the secondary immune response is more rapid and robust. Hence, the goal of vaccination is to enhance the number of antigen-specific B and T cells against a given pathogen.
On the other hand, the innate immune response, which includes phagocytic cells, antimicrobial peptides, and the complement system, has been viewed to be primarily involved in the initial defense against infection. However, several studies show that innate immune response has an additional role in regulating the adaptive immune response. Innate immune recognition of infectious pathogens is mediated by germline encoded receptors that recognize a relatively limited number of highly conserved microbial structures termed as Pathogen Associated Molecular Patterns (PAMPs) [Hoffman et al. 1999; Medzhitov and Janeway 2000]. These interact with Pattern Recognition Receptors (PRRs), called toll-like receptors [Lemaitre et al. 1996; Medzhitov et al. 1997], which are expressed on various cells of the innate immune response, including the major antigen-presenting cells like dendritic cells and macrophages. Studies show that pathogenic stimuli such as lipopolysaccharides, mycobacterial antigens, and specific sequences contained within bacterial DNA can bind to these toll-like receptors and stimulate the secretion of cytokines and chemokines. This innate recognition pathway leads to an increase in APC effector function. Stimulated APCs then play a key role in the initiation of the adaptive immune response. Thus, vaccines that can specifically activate the innate response could have a profound effect on the initial adaptive immune response generated by immunization. Observations from the interaction of innate and adaptive immunity will lead to a greater focus on how vaccines target APCs and stimulate the innate immune response to optimize the adaptive cellular immune response.
A host which survived an infection or was primed with an antigen, will react more rapidly and with higher titers of antibodies or T cells to a second infection or antigen exposure. It is often explained as a special quality of individual T or B cells: they have acquired special "memory" characteristics when compared in vitro to naïve cells or activated effector lymphocytes. This memory status correlates with increased precursor frequencies and enables the system to respond quickly and efficiently to a second exposure. The nature of the memory status correlates with the acquisition of numerous surface molecules on lymphocytes, but overall memory is still poorly understood. An alternative possibility is that immunity depends on a low-level antigen driven immune response keeping T cells activated and maintaining protective antibody titers. Therefore, this would mean that protection by immunological memory disappears when the antigen disappears. Obviously, these two views, inherent special quality versus an antigen-driven process, differ fundamentally. It is very important to understand how protective memory functions over time against infectious agents or tumors so that we can improve vaccines.
Children that have been infected with measles, pox, polio, or numerous other viruses are subsequently resistant to the same infection. Many years of research have been spent on immunological memory. The period of life before and after birth may, from an evolutionary point of view, be the key to understanding immunological memory and immunity. During this period of physiological immuno-competence, adoptively transferable maternal immunological memory is essential for the survival of the offspring and of the species. We may speculate that a naive adult host who does not survive a first infection will not need immunological memory, whereas a host having survived an infection is fit to survive another infection with the same agent later. Other, but less directly life-limiting, benefits of functional immunological memory include improved fitness and herd immunity.
Immunological Memory Conferred By Neutralizing Antibodies
Passively acquired antibody-mediated protection is absolutely required pre- and post-natally. Therefore, transferable immunity is probably an essential pre-condition in vertebrates for maturation of the immune system. Co-evolution of the infectious agents and of the MHC polymorphisms has prevented easy selection of highly cytopathic mutants capable of evading MHC-restricted T-cell recognition. On the other hand, MHC polymorphism has endangered immunological maternal-fetal relationships during ontogeny. The danger of graft versus host or host versus graft reactions between mother and offspring is reduced by the lack of MHC antigen expression in the placental contact areas, by general immunosuppression of the mother, and by virtual complete immunodeficiency of the offspring until birth. Protective antibodies in the serum of the mother are passively transmissible, soluble forms of immunological experience. They protect the offspring for as long as it needs to develop its own T-cell competence and generate its own T-helper-cell-dependent protective and long-lived neutralizing IgG antibody responses.
Here, we argue that the pre-existence or co-evolution of transmissible antibodies has offered a basis to develop MHC polymorphism and MHC-restricted-T-cell-mediated immunity. This would signify that the development of cytopathic agents that could not be controlled efficiently by adoptively transferred antibodies during this critical period of immuno-competence would not have been possible because such infections would have endangered the survival of the species. Adoptively transferable immunological experience by antibodies from the mother to the immuno-competent offspring can be seen in agammaglobulinemic patients, such as mice and newborn calves. Infants incapable of generating their own immunoglobulins will be protected by maternal antibodies for the first 3-9 months after birth. Serum IgG antibody is transferred via the placenta to the serum in humans. Importantly, human active antibodies are active within the gut and influence the gut flora, at least before weaning.
Immunity by T Cells
Immunological experience transferred via antibodies from mother to offspring is very crucial for survival. Neutralizing antibody responses against related but serotypically distinct viruses are limited, not by primed T helper cells or CTLs, but by the precursor frequency of the specific B cells. Memory T cells cannot be transmitted from the mother to the offspring because of mutual immunological rejection. So what is the cell mediated memory for? Two aspects must be discussed here: (1) the role of specific T-cell-mediated protective immunity, and (2) the important role of immunity that depends on on-going local infections, which includes the so-called specific infection immunity and less-specific concomitant immunity. Immunological memory cannot be sustained by IgM antibody because of its very short half-life of only 1-2 days. In addition, because of the lack of receptors and its large molecular size, IgM cannot be transmitted to offspring via placenta or milk. The switch from IgM to IgG requires primed T helper cells and IgG has a half-life of about three weeks. Additionally, IgG is more readily diffusible and transportable via various Fc receptors, and this includes transport to the offspring. As stated earlier, adoptively transferred maternal antibody memory is the key to species survival, and therefore it is not surprising that all vaccines that provide efficient protection working today do so via neutralizing antibodies.
Mechanism of Protective Memory Maintained by Infectious Agents and Vaccines
Sufficiently high neutralizing protective antibody titres are essential for the survival of the offspring, as well as the species, in case of a secondary infection. Maintenance of high neutralizing antibody titers may be achieved via the following: (1) Re-exposure to the antigen from external sources, a route typically used by poliovirus. Usually, the spread of the Sabin vaccine strains within households, schools, or via public swimming pools keeps immunity boosted. (2) Re-exposure from antigen sources within the host. This mechanism is essential for understanding the immunity against Tuberculosis, HBV, HIV, many parasites, and also against measles virus. Measles virus persists in the host not in a replication-competent form, but as a crippled virus apparently often missing a functional matrix protein. From this point of view, SSPE presents an extreme form of persistence of measles virus in the central nervous system and tissues. (3) Antibody-antigen complexes on follicular dendritic cells are maintained for long periods of time and boost both antigen-specific B cells directly as well as T helper cells indirectly. Because cross priming and cross processing of inert antigens can only exceptionally access the MHC class I pathway, even on the dendritic cells, these antigen depots in general are neither capable of maintaining activated CD8 T cells nor, as a beneficial consequence, are they reduced or eliminated by CTLs. In the absence of antigen boosts, antibody responses will eventually dwindle.
Antigen Dependency of Protective Memory
B cells cannot become antibody producing plasma cells in the absence of antigens. To receive signals from specific T helper cells, B cells process antigens bound to its surface Ig receptors in order to present the relevant peptides on MHC class II on their surface. This process is necessary for B cells' maturation to plasma cells, but it is not enough to prime naïve T cells. T helper cells are only efficiently induced in secondary lymphatic organs by antigen-presenting cells, including dendritic cells offering helper peptides via MHC class II molecules. After priming, the increased precursor frequencies of specific T and B cells are readily demonstrated in humans, but primed T or BV cells without specific antigen are not protective by themselves, as shown by adoptive transfer experiments. Protection requires pre-existing high neutralizing antibody titers that are antigen-dependent on B cells that mature to plasma cells. Some experiments have suggested that plasma cells have a very long half-life, up to 150-300 days. However this evidence is unfortunately flawed because antibody responses against non-protective antigens composed of multiple undefined determinants have been evaluated instead of neutralizing antibody responses, which are specific for the tip of viral and bacterial glycoproteins. The results show that increased antibody titers depend upon antigen-driven plasma cell maturation and responses. In fact, protective antibody titers usually decrease over time (example: diphtheria, tetanus toxins, and measles vaccines). All these observations strongly suggest that the maintenance of protective or neutralizing antibody titers is antigen dependent.
T-Cell Independent Antigens and Conjugate Vaccines
Some antigens can directly stimulate B-cell proliferation and antibody secretion. This process of T-cell independent B-cell activation takes place in response to highly polymerized antigens that can cause extensive cross linking of B-cell receptors. An important example of this occurs with bacterial polysaccharides, which are repetitive carbohydrate determinants on the surface of some bacteria. The cellular events and outcomes associated with T-cell independent B-cell activation are distinct from T-cell stimulation and hence have important implications for vaccine development. T-cell independent B-cell stimulation produces a predominance of IgM and IgG2 antibody secreting cells, and the production of antigen-specific memory B cells is inefficient. Thus, even repeated immunization with a capsular polysaccharide produces low levels of antibodies and poor memory of antibody response upon challenge. This has hindered the vaccine development process for pathogens like H.influenzae, S.pneumoniae, etc. [Heinzel 2000; Ward and Zangwill 1999; Fedson et al. 1999].
To overcome these inherent limitations of polysaccharide antigens, new vaccine formulations have coupled bacterial polysaccharides to proteins that produce strong T-cell dependent antibody responses. This allows a specific B cell to recognize the polysaccharide antigen and endocytose the coupled antigen complex. Some of the resultant helper peptides are loaded onto MHC class II on the surface of the B cells and provide for a cognate interaction between the antigen-specific CD4+ helper T cell and B cell. This approach has led to the development and licensure of highly effective vaccines against H.influenzae that couple the bacterial capsular polysaccharide to known T helper antigens, such as diphtheria or tetanus toxoid. [Ward and Zangwill 1999; Fedson et al. 1999].
Immunopathaological Basis for the Design of Vaccines
Parasites have to survive in their vertebrate host during a sufficiently prolonged time in order to achieve their life cycle through successful transmission via insect vectors. In their vertebrate hosts, parasites are often confronted by vigorous effector immune responses that they must subvert somehow in order to survive and be transmitted successfully.
The immune response comprises several components in terms of effector cells, antibodies, and signalling molecules such as cytokines. It is now well-recognized that not all components of the immune response triggered during the process of parasitism have anti-parasitic activities or properties. In this context, one of the strategies devised by many intracellular bacteria and viruses has been to utilize some components of the host immune response to its own benefit. Therefore, an effective vaccine against intracellular pathogens should only induce effector mechanisms ultimately leading to the destruction of the parasites. The vaccine should abstain from triggering immune components of the immune response favoring the survival of the parasites.
Parasites have also devised a variety of ways to ways by which to escape the effector components of the immune response that they have elicited in their mammalian hosts. Therefore, the balance between the components of the immune response with effector functions and those irrelevant for protection on the one hand, and the effector immune response and the escape processes induced by parasites on the other largely determine the outcome of the infection. Thus, elucidation of the mechanisms involved in the progression of parasitic diseases and definition of the relevant immune effector mechanisms are equally important for the identification of potential targets for beneficial intervention and subsequent vaccinisation. Indeed, the specific triggering of potent effector mechanisms and responses without activating the immune responses' components leading to pathology, together with the prevention of the escape mechanisms evolved by parasites, are the aims of vaccinisation.
The importance of immunoregulatory molecules like cytokines in the immunopathology of diseases is worthy enough to be mentioned. Their central role as antiparasite factors and as mediators of pathology has been reinforced by the description of two functionally different CD4+ T cell populations that can be distinguished on the basis of the patterns of cytokines which they produce. In both mice and humans, Th1 cells produce Interleukin-2 (IL-2) and Interferon gamma (IFN gamma), whereas Th2 cells produce IL-3, IL-4, IL-5, IL-10, and IL-13. Th1 cells are responsible for cell-mediated immune reactions, whereas Th2 cells are involved in humoral immunity. Resistance and susceptibility to several infections by many pathogens have been analyzed with respect to the Th1-Th2 paradigm. Resistance to intracellular pathogens is often associated with the development of a polarized Th1 response. Although resistance to extracellular parasites is now known to correlate with the generation of Th2 cells, the respective roles of Th1 or Th2 effector responses in protective immunity against many diseases like Schistosoma, an extracellular pathogen, are not yet fully understood. Furthermore, expression of immunity in some parasitic diseases such as malaria cannot simply be analyzed in view of the development of polarized Th1 or Th2 responses because infection with this parasite is a multifocal process.
Factors Determining Vaccine Efficacy
Successful immunization requires the activation, replication, and differentiation of T and B lymphocytes leading to the generation of memory cells. The vaccines in current use require multiple immunizations to maintain effective immunity. For a variety of bacterial and viral infections, there is a well-defined threshold for the amount of antibody required for protection. Live infection induces a greater frequency of antigen-specific cells than immunizations with proteins, DNA vaccines, or recombinant non-replicating viruses encoding specific antigen. The magnitude of the memory phase is generally determined by the size of the initial clonal burst induced by immunization. Also, the amount of antigen can also affect the qualitative aspect of CD8+ T cells memory effector responses. For diseases which require a higher number of memory CD8+ T cells to mediate protection, it will be desirable to use a vaccine strategy that provides a sufficient antigenic load to maximize the initial burst size. Although antigen is not required to sustain resting memory CD8+ T cells, it may be required to sustain optimal T-cell effector function.
Having seen the process by which the two arms of the immune system work synergistically, we now have to design strategies to induce immunity by the process of vaccinisation. Along this line of thought, the best way to confer immune resistance to a pathogen is to mimic itself or to devise formulations which mimic its characteristics. The vaccines which we are going to consider here are non-genetic vaccines, or in simple terms, the components of these vaccines are either proteins or their subconjugates such as glycoproteins, glycolipids, etc. Carbohydrates and short lipopolysaccharide molecules are also used. The main purpose of these peptide vaccines and their analogues is to confer immunity without inducing disease or in other words, "achieving immunogenicity without causing disease." To achieve this end, a proper understanding of the process of pathogenesis is required at the molecular level. Each pathogen has a particular niche, and accordingly it has a range of virulence factors. These virulence factors fall into the domain of candidate antigens, which are screened for incorporation into a vaccine. An in-depth study and subsequent understanding of the virulence factors will greatly aid and help us in developing and configuring vaccines which are target oriented, immunogenic, and capable of eliciting immunity almost similar to the immunity acquired through natural infection.'
Bacterial pathogens utilize a wide variety of molecules called virulence factors to cause infections. Because these molecules are critical for disease, and because these virulence factors are on the bacterial surface, or secreted out of the pathogen, they make attractive targets for potential vaccine candidates. Capsules, toxins, and adhesins such as pili have been used in the past to develop successful vaccines. More recently, significant advances have been made in understanding the molecular basis of bacterial diseases, significantly increasing the number of potential targets. In order for a virulence factor to be effective, it has to be produced at the right place at the right time. Often this is on the bacterial surface, where it is secreted or even delivered directly into the host cell. Much research has focused on studying the regulation of virulence factors, and defining the conditions under which they are produced. As a general rule, virulence factors are usually tightly regulated and expressed under conditions that mimic that particular stage of an infection. Because of this regulation, many virulence factors have gone undetected until recently. For example, molecules that allow a pathogen to invade epithelial cells of the intestine are usually expressed under conditions that mimic the gut. Virulence factors needed to mediate intracellular survival are again only expressed inside the host cells. If we construct strains, which express these virulence factors constitutively, the strain is usually avirulent, emphasizing the necessity of virulence factor regulation. Thus in order for a vaccine to be effective, we must know when and where the virulence factors are expressed and utilize this knowledge when developing vaccine strategies.
Secretion System in Gram Negative Bacteria
Significant research efforts have been focused on secretion systems, especially in gram-negative bacteria. The gram-negative bacterial cell wall is a complex structure, and movement of virulence factors across two membranes and a periplasm requires several dedicated transport and translocation systems [Finlay and Cossart 1997; Finlay and Falkow 1997; Salyers and Whitt 1994].
Several types of secretion systems have been characterized [Table 1]. Amazingly, at least two of these types of secretion systems (types 2 and 4) seem to be designed not only to transport virulence factors out of the bacterium but also to inject them directly into the host cell. This has led to the unprecedented knowledge that several bacterial pathogens drive specific molecules into host cells, where they can interact directly with host cell signaling systems, the cytoskeleton, vesicular transport, and the apoptotic pathways. However this also implies that, in order to develop a vaccine against these molecules, a cell-mediated immunological strategy is required to access bacterial molecules found within the cytoplasm.
Interaction of bacterial pathogens with host cells is particularly characterized by factors which are located on the bacterial surface or are secreted into the extracellular space. Although the secreted bacterial proteins are numerous and diverse, and exhibit a wide variety of functions that include proteolysis, haemolysis, cytotoxicity, and protein phosphorylation and dephosphorylation, only a few pathways exist by which these proteins are transported from the bacterial cytoplasm to the extracellular space. Thus, four pathways of protein secretion (type I to IV) have been described in gram-negative bacteria [Fath and Kolter 1993; Finlay and Falkow 1997; Salmond and Reeves 1993; Van Gijsegem et al. 1993].
A fifth system for macromolecular secretion is involved in conjugal transfer of plasmids, T-DNA transfer by Agrobacterim tumefaciens, and secretion of Bordetella pertussis toxin. The last system, which could be named type V secretion, is the least well-characterized and comprises only three members so far [Winans et al. 1996].
In this context, the term "secretion" is used to describe the active transport of proteins from the cytoplasm across the inner and outer membranes into the bacterial supernatant or onto the surface of the bacterial cell. Secretion is distinguished from export, which refers to the transport of proteins from the cytoplasm into the periplasmic space [Pugsley 1993; Salmond and Reeves 1993].
Type I Sec-Independent Pathway
In contrast to the type II and IV secretion pathways, type I and type III secretion are independent of the sec system and thus do not involve amino-terminal processing of the secreted proteins. Furthermore, protein secretion via the latter pathways occurs in a continuous process without the distinct presence of periplasmic intermediates. Type I secretion [Fath and Kolter 1993; Wandersman 1996] is exemplified by the E. coli alpha-hemolysin secretion system. Other members of this group are the adenylate cyclase secretion system of B. pertussis, leukotoxin secretion by Pasteurella haemolytica, and the protease secretion systems from P. aeruginosa and Erwinia chrysanthemi. Type I secretion requires three secretory proteins: an inner membrane transport ATPase (termed ABC protein for ATP-binding cassette), which provides the energy for protein secretion; an outer membrane protein, which is exported via the sec pathway; and a membrane fusion protein, which is anchored in the inner membrane and spans the periplasmic space. The genes encoding the secretion apparatus and the secreted protein are usually clustered.
The proteins which are secreted via the type I pathway are not subject to proteolytic cleavage, and the secretion signal is located within the carboxyl-terminal 60 amino acids of the secreted protein. The secretion signal appears to be specific for subfamilies of the secretion system; i.e., the proteases are only poorly secreted via the hemolysin system and vice versa. The nature of the protease family secretion signal may be mainly conformational [Wandersman 1996], while for E. coli alpha-hemolysin several dispersed key residues which are essential irrespective of a specific secondary structure and could facilitate recognition by the secretion apparatus have been identified [Chervaux and Holland 1996].
Type II and Type IV Sec-Dependent Secretion Pathways
Type II and IV protein secretion pathways involve a separate step of transport across the inner membrane prior to transport across the cell envelope. While these pathways differ in the way in which the proteins are transported across the outer membrane, export to the periplasm occurs via the sec system in both cases. (Secretion of pertussis toxin by B. pertussis, which may be categorized as type V secretion, also involves the sec pathway.) A signature of sec-dependent protein export is the presence of a short (about 30 amino acids), mainly hydrophobic amino-terminal signal sequence in the exported protein. The signal sequence aids protein export and is cleaved off by a periplasmic signal peptidase when the exported protein reaches the periplasm. In E. coli, the sec pathway comprises a number of inner membrane proteins (SecD to SecF, SecY), a cytoplasmic membrane-associated ATPase (SecA) that provides the energy for export, a chaperone (SecB) that binds to pre-secretory target proteins, and the periplasmic signal peptidase. A number of accessory proteins are also required for normal function [Murphy and Beckwith 1996; Pugsley 1993].
In type II secretion, transport across the outer membrane requires an additional set of inner and outer membrane proteins. In the case of pullulanase secretion by Klebsiella oxytoca, the best-studied example of type II secretion, 14 additional secretion factors, which are encoded by a continuous gene cluster, are necessary and sufficient for secretion. At least seven of these proteins are located in the cytoplasmic membrane, while PulS and PulD are outer membrane proteins [Pugsley 1993]. Other examples of type II secretion [Hobbs and Mattick 1996] include the out pathway of Erwinia spp. for the secretion of pectic enzymes and cellulases, the xcp-encoded secretion of elastase, exotoxin A, phospholipase C, and other proteins by Pseudomonas aeruginosa, amylase and protease secretion by Aeromonas hydrophila exe, and secretion of polygalacturonase and other proteins by Xanthomonas campestris xps. Thus, type II secretion is the primary pathway for the secretion of extracellular degradative enzymes by gram-negative bacteria. Furthermore, parts of the type II secretion pathway have homologs in other transport systems; for example in the secretion and assembly of N-methyl-Phe (type 4) pili of P. aeruginosa and other bacteria and in DNA transfer systems of Haemophilus influenzae and Bacillus subtilis [Hobbs and Mattick 1996]. Most notably, however, the outer membrane component of the pullulanase secretion system, PulD, is conserved in a variety of gram-negative protein transport systems.
The type IV secretion pathway [Finlay and Falkow 1997] comprises a group of so-called auto-transporters, including gonococcal immunoglobulin A and other proteases, the vacuolating cytotoxin of Helicobacter pylori, a family of outer membrane proteins in B. pertussis, and the secreted proteins SepA and EspC from S. flexneri and EPEC, respectively. As in type II secretion, these proteins are exported from the cytoplasm via the sec pathway, involving the cleavage of an amino-terminal signal peptide. However, the information required for transport across the outer membrane resides entirely within the secreted protein. Apparently, these auto-transporters form a pore in the outer membrane through which they pass, and autoproteolytic cleavage releases the proteins into the supernatant.
Type III Secretion System
Genetic analyses of bacterial virulence factors has shown that pathogens are distinguished from their nonpathogenic relatives by the presence of specific pathogenicity genes, often organized in so-called pathogenicity islands, clusters of genes which apparently have been acquired during evolution via horizontal genetic transfer. Thus, distantly related pathogens have turned out to harbor closely related virulence genes. This point has become particularly apparent for a set of approximately 20 genes which together encode a pathogenicity mechanism termed type III secretion. Type III secretion enables gram-negative bacteria to secrete and inject pathogenicity proteins into the cytosol of eukaryotic host cells. Fascinatingly, while the type III secretion apparatus is conserved in pathogens as distantly related as Yersinia and Erwinia, the secreted proteins differ entirely; illustrating how one bacterial pathogenicity mechanism can give rise to a multitude of diseases that range from bubonic plague in humans to fire blight in fruit trees.
Secretion of bacterial pathogenicity proteins by the type III pathway and their injection into the cytosol of animal or plant cells initiates a sophisticated "biochemical cross-talk" between pathogen and host. The injected proteins often resemble eukaryotic factors with signal transduction functions and are capable of interfering with eukaryotic signalling pathways. Redirection of cellular signal transduction may result in disarmament of host immune responses or in cytoskeletal reorganization, establishing subcellular niches for bacterial colonization and facilitating a highly adapted pathogenic strategy of "stealth and interdiction" of host defense communication lines.
Like the type I secretion pathway, type III secretion is independent of the sec system. (Assembly of the type III secretion apparatus, however, probably requires the sec pathway, since several components of the type III secretion apparatus carry sec-characteristic amino-terminal signal sequences.) The type III secretion apparatus is composed of approximately 20 proteins, most of which are located in the inner membrane, and type III secretion requires a cytoplasmic; probably membrane-associated ATPase. Interestingly, most of the inner membrane proteins are homologous to components of the flagellar biosynthesis apparatus of both gram-negative and gram-positive bacteria, while an outer membrane protein of the type III secretion apparatus is homologous to PulD, the outer membrane secretion of the type II secretion pathway. Although type III secretion does not include distinct periplasmic intermediates of the secreted proteins, transport through the inner membrane is genetically separable from secretion through the outer membrane, since a mutant of the outer membrane PulD homolog of P. syringae was shown to accumulate considerable amounts of a secreted protein in the periplasm [Charkowski et al. 1997]. As in type I and type II secretion, the genes encoding the type III secretion apparatus are clustered.
As in type I secretion, the proteins secreted via the type III pathway are not subjected to amino-terminal processing during secretion. The signal for secretion has long been thought to reside within the amino-terminal 15 to 20 amino acids of the secreted proteins, since this region is necessary for secretion and suffices to direct the secretion of hybrid fusion proteins. However, the amino-terminal sequences of proteins secreted via the type III pathway do not share any recognizable structural similarities that could function as a common secretion signal, and exhaustive mutational analysis of some secreted proteins has revealed a high degree of tolerance for sequence changes within the amino terminus without loss of secretion. Therefore, it has recently been proposed that the secretion signal resides in the 5′ region of the mRNA which encodes the secreted proteins [Anderson and Schneewind 1997]. Interestingly the secreted proteins require small cytoplasmic proteins with chaperone functions to protect the secreted factors from premature interaction with other components of the secretion system. In contrast to type I secretion, which is a true secretory system in that the secreted enzymes are active in the extracellular space, type III secretion systems appear to be dedicated machineries for the translocation of pathogenicity proteins into the cytosol of eukaryotic cells. Accordingly, protein secretion,at least in some cases,is regulated by contact with the surface of a target cell. In accordance with the homology of the type III secretion apparatus to flagellar biosynthesis factors, some type III secretion systems assemble supermolecular structures on the bacterial surface, which could be involved in protein translocation into eukaryotic cells [Ginocchio 1994; Roine et al. 1997].FUNCTIONAL CLASSIFICATION OF VIRULENCE FACTORS:
Despite their large number, most virulence factors fall into one of a few categories when classified according to function. These include the following:
A. Toxins, which are secreted bacterial molecules that poison host cells by a variety of mechanisms following uptake by the host cell.
B. Adhesins, which are bacterial surface structures that allow pathogens to adhere to host cells.
C. Invasins, which are bacterial surface molecules that trigger host cells to enable bacterial uptake into them.
D. Virulence Factors, which mediate intracellular survival within host cells.
E. Anti Phagocytic Factors, often encoded by cell surface structures such as capsules but including virulence factors injected directly into host cells to paralyze phagocytosis.
F. Host Cell Apoptosis Altering Factors, which either trigger or inhibit apoptosis.
Toxins are perhaps the best characterized virulence factors, because they are most often the easiest to purify by concentrating the supernatant, and because they have distinct phenotypes. They have also been used extensively in toxoid vaccines, as well as adjuvants. However as virulence factors, they often have essential roles in altering and often killing host cells, usually by an enzymatic process [Scihavo and van Der Goot 2001; Steele-Mortimer et al. 2000]. Exotoxins are usually proteins that are secreted by bacteria out into the supernatant. Endotoxin is lipolysaccharide (LPS), which is principally carbohydrate, not protein, non-enzymatic, and not normally secreted. Because LPS is a critical component of all gram-negative bacteria, generally it is not considered a bacterial virulence factor, although it stimulates toxic effects in the host by activating immune responses and cytokine production.
Exotoxins come in several types and specificities. They usually have two general functions, which can often be uncoupled: the ability to bind to a host cell receptor, and enzymatic activity. For example, cholera toxin [Figure 1], labile toxin, and many others have five binding (B) subunits coupled to the active enzymatic subunit (A). Several vaccines are made from the B subunits without the catalytic subunit, ensuring no toxin activity. The B subunits bind to host cell molecules such as carbohydrates (for example, cholera binds the GM1 ganglioside, a glycolipid on the intestinal surfaces). Following binding, the A subunit then enters the cell, which is facilitated by the B subunits. Once inside the cell, the A subunit modifies the host cell molecules.
The AB Toxin Complex of Vibrio Cholerae: An Example
To induce disease, CT released into the intestinal lumen must enter the intestinal epithelial cell at the apical membrane and eventually activate epithelial adenylyl cyclase at the cytoplasmic surface of the basolateral membrane. This event, likely amplified by interaction of toxin with subepithelial enteric nerves and the effect of local neurotransmitter release, leads to the massive intestinal salt and water secretory response characteristic of cholera. The microbe V. cholerae, however, does not invade the intestinal mucosa or directly assist the delivery of toxin into the cell cytoplasm by other mechanisms (such as by type III secretion). Thus the molecular determinants that drive entry of CT into the host intestinal epithelial cell are encoded entirely within the structure of the fully assembled and folded protein toxin itself.
The crystal structures of CT, and the closely related Escherichia coli heat-labile enterotoxin LTI responsible for "traveler's diarrhea," have been solved. Both toxins belong to the AB5 subunit family that also includes Shiga/verotoxin (responsible for hemolytic uremic syndrome) and pertussis toxin. In CT, five identical peptides (~11 kDa) assemble into a highly stable pentameric ring termed the B subunit (~55 kDa). The B subunit exhibits specific and high-affinity binding to the oligosaccharide domain of ganglioside GM1 and functions to tether the toxin to the plasma membrane of host cells. B subunit binding to GM1 is stoichiometric, with one B subunit pentamer cross-linking five GM1 gangliosides at the cell surface. The specificity and stability of toxin binding to GM1 dictate toxin function, likely by affecting toxin trafficking into the cell. The single A subunit is comprised of two major structural domains termed the A1 and A2 peptides. The A1 and A2 peptides are linked by an exposed loop containing a serine protease-sensitive "nick" site and a single disulfide bond. The A2 peptide (~5 kDa) tethers the A1 peptide to the B subunit and contains a COOH-terminal motif that protrudes below the pentameric B subunit on the side that binds GM1 at the cell surface. This motif is known to be a sorting signal that allows endogenous luminal endoplasmic reticulum (ER) proteins of the eukaryotic cell to be retrieved efficiently from post-ER compartments. The A1 peptide (~22 kDa) is the enzymatically active subunit that must eventually dissociate from the B subunit, translocate across a cellular membrane, and act inside the cell to activate adenylyl cyclase by catalyzing the ADP-ribosylation of the heterotrimeric GTPase Gs .
The B subunit, unlike the A1 peptide, does not translocate across cell membranes. Rather, the B subunit remains membrane associated (presumably bound to GM1) and eventually moves back out the secretory pathway by vesicular traffic to the cell surface. Thus the B subunit (and a small fraction of holotoxin) can move from its original site of binding on the apical (or mucosal) cell surface to the basolateral (or serosal) cell surface by first moving through Golgi cisternae and possibly ER. We have termed this process "indirect" transcytosis. This ability of CT to breach the epithelial barrier by crossing through epithelial cells within transport vesicles may contribute to the potent effects of orally delivered CT on mucosal and systemic immune responses.
To invade the intestinal cell, CT co-opts the machinery for membrane traffic endogenous to the host epithelial cell itself. CT is not a pore-forming toxin. Rather, CT enters polarized epithelial cells through a complex pathway involving apical endocytosis and retrograde membrane traffic through Golgi cisternae to ER. It is currently believed that CT must enter the ER for the A1 peptide to unfold and translocate into the cytosol. After membrane translocation, the A1 peptide may move to the adenylyl cyclase complex on the cytoplasmic surface of the basolateral membrane by diffusion through the cytosol (if the A1 peptide breaks away from the membrane after translocation) or by vesicular transport back out the secretory pathway (if the A1 peptide remains membrane-associated).
The cholera toxin [Figure 1] affects adenylate cyclase. On exposure to small bowel epithelial cells, each B subunit binds to a receptor on the gut epithelium. Following binding, the A and A2 moieties migrate through the epithelial cell membrane. The A subunit is an ADP-ribosyl transferase that catalyzes the transfer of ADP ribose moiety (ADPR) from NAD to a guanosine triphosphate (GTP)-binding protein that regulates adenylate cyclase activity. The ADP-ribosylation of GTP binding protein inhibits the GTP turnoff reaction and causes a sustained increase in adenylate cyclase activity. This increase in activity results in excess secretion of isotonic fluid into the intestine, which results in diarrhea.
The petrussis toxin modifies heterotrimeric G proteins. The pertussis toxin, PTx, is a protein that mediates both the colonization and toxemic stages of the disease. PTx is a two component, A+B bacterial exotoxin. The A subunit (S1) is an ADP ribosyl transferase. The B component, composed of five polypeptide subunits, binds to specific carbohydrates on cell surfaces. PTx is transported from the site of growth of the Bordetella to various susceptible cells and tissues of the host. Following binding of the B component to host cells, the A subunit is inserted through the membrane and released into the cytoplasm in a mechanism of direct entry. The A subunit gains enzymatic activity and transfers the ADP ribosyl moiety of NAD to the membrane-bound regulatory heterotrimeric Gi protein that normally inhibits the eukaryotic adenylate cyclase. The Gi protein is inactivated and cannot perform its normal function to inhibit adenylate cyclase. The conversion of ATP to cyclic AMP cannot be stopped and intracellular levels of cAMP increase. This has the effect to disrupt cellular function, and in the case of phagocytes, to decrease their phagocytic activities such as chemotaxis, engulfment, the oxidative burst, and bactericidal killing. Systemic effects of the toxin include lymphocytosis and alteration of hormonal activities that are regulated by cAMP, such as increased insulin production (resulting in hypoglycemia) and increased sensitivity to histamine (resulting in increased capillary permeability, hypotension, and shock). PTx also affects the immune system in experimental animals. B cells and T cells that leave the lymphatics show an inability to return. This alters both AMI and CMI responses and may explain the high frequency of secondary infections that accompany pertussis (the most frequent secondary infections during whooping cough are pneumonia and otitis media). Although the effects of the pertussis toxin are dependent on ADP ribosylation, it has been shown that mere binding of the B oligomer can elicit a response on the cell surface such as lymphocyte mitogenicity, platelet activation, and production of insulin effects.
Comparison between Cholera Toxin and Pertussis Toxin (Ptx) In Their Ability to Interfere With the Regulation of the Eukaryotic Adenylate Cyclase Complex.
Normal regulation of adenylate cyclase activity in mammalian cells (Adenylate cyclase (AC)) is activated normally by a stimulatory regulatory protein (Gs) and guanosine triphosphate (GTP). However, the activation is normally brief because an inhibitory regulatory protein (Gi) hydrolyzes the GTP [Equation 1].
Adenylate cyclase is activated by cholera toxin. The cholera toxin A1 fragment catalyzes the attachment of ADP-Ribose (ADPR) to the regulatory protein Gs, forming Gs-ADPR from which GTP cannot be hydrolyzed. Since GTP hydrolysis is the event that inactivates adenylate cyclase (AC), the enzyme remains continually activated [Equation 2].
Adenylate cyclase activated by pertussis toxin. The pertussis A subunit transfers the ADP ribosyl moiety of NAD to the membrane-bound regulatory protein Gi that normally inhibits the eukaryotic adenylate cyclase. The Gi protein is inactivated and cannot perform its normal function to inhibit adenylate cyclase. The conversion of ATP to cyclic AMP cannot be stopped [Equation 3].
Neurotoxins such as botulinum toxin cleave molecules essential for neurotransmitter release. Botulinum toxin binds with high affinity to peripheral cholinergic nerve endings, such as those at the neuromuscular junction and in the autonomic nervous system. The toxin proceeds through a complex sequence of events that culminates in blockade of acetylcholine release. This means that toxin action at the neuromuscular junction can cause weakness and even complete paralysis. Similarly, patients may have many signs of parasympathetic and sympathetic dysfunction attributable to toxin action on autonomic nerves. The past several years have been exciting for botulinum toxin research because investigators have discovered the mechanism of toxin action on cholinergic cells. After binding with high affinity to receptors on nerve endings, the toxin penetrates the cell membrane by receptor-mediated endocytosis and then crosses the endosome membrane by pH-dependent translocation. When it reaches the cytosol, the toxin acts as a zinc-dependent endoprotease to cleave polypeptides that are essential for exocytosis. In the absence of these peptides, nerve impulses can no longer trigger the release of acetylcholine
Another class of toxins is capable of inserting directly into the host cell membranes, thereby forming a pore and causing lysis of the host cell. For example, streptolysin O, Staphylococcus alpha-toxin and Escheria coli hemolysin all cause pore formation in host cells by assembling into oligomeric pores in the cell membrane. Because of the complexity of gram-negative envelopes (two membranes and a periplasm), most gram-negative toxins also have a dedicated secretion pathway to enable secretion. For example, hemolysin uses a type 1 secretion system, whereas petrussis toxin uses a type 4 system.
The spectrum of host cells targeted by the toxins is extensive, including many molecules involved in cytoskeletal function, vesicular transport, protein synthesis, etc. The vast diversity of toxin targets is beginning to be realized. In addition to this, we are realizing that these toxins make excellent probes of normal cellular function [Scihavo and van Der Goot 2000].
The ability to adhere to the host cell is usually central to a pathogen's ability to cause a disease. Even pathogens which rely mainly upon a toxin-mediated event such as cholera require the organism to adhere to cell surfaces, thereby allowing bacterial growth, and toxin production and secretion. Thus blockage of adherence factors by vaccinating against adhesins makes a logical vaccine target. Veterinarian vaccines against E.coli fimbriae have been very effective against calf and pig scours.
Like toxins, there are many types of adhesins [Hultgren et al. 1993]. Because of the necessity to be on the bacterial surface to access host cells, many adhesins are at the tip of hair like projections called pili or fimbriae. This presumably allows surface exposure, as well as flexibility upon initial contact, much like a harpoon on the end of a line. Some fimbriae, like those of uropathogenic E. coli (which causes urinary tract infections), have a dedicated tip adhesin. However, others such as Pseudomonas aeruginosa pili utilize the same protein as that which makes the pilus; however because of pilin stacking, the adhesive domain is exposed on the subunits that are on the tip, which are otherwise buried within the stalk. From a vaccine perspective, both the pilus tip and stalk are potential targets, as antibodies adherent to the tips would block adherence, whereas antibodies coating the stalk would enhance complement lysis and bacterial clearance. Like flagella and other surface organelles, adhesins have a complex base structure embedded in the outer membrane, plus several other chaperones and secretins that assist in pilus assembly.
In contrast to fimbriae, there are also many adhesins that are part of the bacterial surface, often embedded in the outer membrane. Some examples of these include petrussis filamentous hemagglutinin (FHA, binds several receptors), the opacity proteins (Opa) of Neisseria, which bind carcinoembryogenic antigens (CEA), and also several extracellular matrix-binding proteins of gram-positive pathogens. The receptors to which adhesins bind are varied but, not surprisingly, are usually cell surface molecules, including glycoproteins, glycolipids, and proteins. Generally most adhesins target carbohydrates, which seem to give them much tissue and cell specificity if needed. E.coli type 1 pili bind mannose, which gives them a broad range of receptors given the high presence of mannose on the cell surfaces. In contrast, uropathogenic E.coli binds globosides in a certain conformation only, which seems to specifically target the bladder epithelium.
Most pathogens synthesize many adhesins whose expression is regulated differently. For example, Salmonella has at least five different adhesins that seem to contribute to adherence to the intestinal epithelium. Not surprisingly, mutation of a single adhesin does not seem to have any effect in the virulence degree, yet mutation of several adhesins decreases virulence significantly. Because of this complexity and redundancy, definition of the contribution to virulence of particular adhesins has proven difficult. However, there is usually a correlation with the presence of particular adhesins among clinical isolates and disease severity.
A fascinating yet unexpected mechanism of adherence has been described for enteropathogenic E.coli (EPEC) and enterohaemmorhagic E.coli (EHEC) [Kenny et al. 1997]. Both of these significantly cause diarrhea, with EHEC also causing hemolytic uremic syndrome (HUS, commonly called Hamburger's disease). These pathogens utilize a type 3-secretion system to inject a bacterial molecule, Tir, directly into the host cell membrane [Figure 2]. Tir then functions as a receptor for an outer membrane molecule, Intimin. Tir-Intimin interactions bind these bacteria tightly to intestinal cells and also cause host cell cytoskeleton to accumulate in pedestals beneath adherent organisms, a process critical for disease. This is the first example of the pathogen injecting its own receptor into the host cells, and Tir and Intimin make particularly strong vaccine candidates, especially because one can vaccinate against both the adhesin and the intimin.
Although most bacterial pathogens function as extracellular pathogens, several of the more serious pathogens actually reside inside host cells [Meresee et al. 1999]. Some examples include Salmonella, Mycobacterium, Listeria, and Chlamydia species. Although perhaps the easiest way to enter a phagocytic cell is to be passively taken up through phagocytosis, most pathogens have derived specialized mechanisms for driving themselves into host cells, including non-phagocytic cells. This facilitates entry and penetration through the intestinal epithelial cells, and often places the microbe in a privileged and protected niche inside the host cell, thereby avoiding phagosome-lysosome fusion events.
There are two mechanisms for pathogen directed endocytosis (invasion into the host cells). The first involves a zipper-like uptake mechanism, whereby a bacterial surface molecule binds tightly to a receptor. The bacterial molecule is expressed over the entire bacterial surface, thereby allowing sequential binding of additional molecules to the host receptors and "zippering" the pathogen into the host cell. This process requires the host cell actin cytoskeleton, although significant rearrangements of the host cell surface are not obvious. Yersinia enterocolitica and Yersinia pseudotuberculosis both encode an outer membrane protein called Invasin, which binds to host cell surface Beta-1integrins (which normally bind matrix molecules such as fibronectin) very tightly, facilitating invasion. Expression of Invasin in non-pathogenic E.coli enables these normally noninvasive organisms to enter the non-phagocytic cells, such as epithelial cells. The gram-positive food-borne pathogen Listeria monocytogenes also has a family of invasions called Internalins (Inl). InlA binds to E cadherins on the cell surface, mediating bacterial uptake via the cadherin-cytoskeletal linkage, whereas InlB binds to at least two host cell molecules to direct uptake.
The other mechanism invasive pathogens use to facilitate their uptake involves causing significant activation of the actin cytoskeleton, causing significant membrane ruffling, macropinocytosis, and invasion. Pathogens that use such a mechanism include Salmonella and Shigella species. These pathogens utilize a type 3 secretion system to inject bacterial effectors into host cells to alter host cell small GTPases such as rac, rho, and CDC42, whose activity controls actin polymerization levels. By promoting actin polymerization and interfacing directly with host cell signal transduction pathways, membrane ruffling occurs, which appears as a "splash" on the cell surface prior to invasion. Once these pathogens are inside the host cell, the cellular surface returns to normal, with the pathogen residing within a membrane-bound compartment inside the host cell.
Intracellular Survival Assisting Virulence Factors
Once inside a host cell, invasive pathogens have derived several strategies to avoid the intracellular killing mechanisms that normally follow uptake. These include vacuole acidification and phagosome-lysosome fusion (which delivers the toxic substances and enzymes to vacuoles containing ingested materials). However, this altered targeting within a host cell also contributes to the difficulty of developing vaccines against intracellular pathogens, because they often do not reside within the normal antigen-presenting pathways used for cell-mediated immunity.
A small number of intracellular pathogens appear to thrive within phagolysosomes, apparently oblivious to lysosomal enzymes and low pH. One such pathogen is Coxiella burnetti, the causative agent of Q-fever. How these pathogens overcome the hostile host intracellular defenses remains undefined. A similarly small number of pathogens lyse the vacuolar membrane once they are internalized, releasing them directly into the host cell cytosol where they flourish. Shigella species, L.monocytogenes, and Rickettia have phospholipases that break down the vacuolar membrane. Once free in the cytosol, additional bacterial proteins directly recruit the actin polymerization machinery, which then create a comet tail of cytoskeleton behind the organisms, thereby propelling them through the cytosol and presumably directly into adjacent host cells, all without ever being exposed to the extracellular environment [Finlay and Cossart 1997]. This is the primary reason that L.monocytogenes is used as a classic T-cell mediated immune mechanism, as antibodies would never see the pathogen once it initially enters a host cell. From a vaccine perspective, cell mediated immunity is necessary to control such pathogens, most likely by killing the host cell or by killing the pathogen within the host cell. It is surprising that only a few intracellular pathogens have evolved such a clever and effective intracellular survival mechanism.
Most intracellular pathogens choose to alter the targeting of the membrane vacuole that surrounds the invading organism. Unfortunately, the mechanisms by which the pathogens achieve this are not well characterized [Meresee et al. 1999]. For example, Salmonella species utilize a type 3 secretion system to invade cells and then induce a second type 3 secretion system that delivers effectors into host cells to promote macrophage survival. Following invasion, these bacteria initially traffic along the normal cellular route, but then they diverge prior to lysosomal fusion. Some of the effectors include those that include long filaments associated with the Salmonella containing vacuole, whereas others cause actin to condense around the vacuole at later times, thereby affecting fusion with the lysosomes. The Mycobacterium alters trafficking also, homing to early endosomes and then preventing the vacuolar ATPase from fusing with the phagosome, and thereby preventing vacuole acidification [Charkowski et al. 1997]. Legionella pneumophila and Brucella abortus use a type 4 secretion system for this effect. At present, the type 4 effectors have not been identified, but this area and how pathogens affect intracellular trafficking in general are under intensive investigation. As effector molecules are identified, they may make excellent vaccine candidates for a cell-mediated vaccine.
Antiphagocytotic Virulence Factors
An alternative strategy to avoid being killed by phagocytic cells is to prevent phagocytosis from occurring in the first place. Although only a few pathogens are currently recognized for this ability, their numbers have been increasing as additional pathogens are tested. The best-studied antiphagocytic strategy is that utilized by the Yersinia species [Anderson and Schneewind 1997]. These pathogens also encode a type 3 secretion system, but this pathogen's injected effectors are designed to paralyze phagocytosis. They include a potent tyrosine phosphatase, a serine-threonine kinase, and several actin poisons that collectively and effectively block phagocytosis. Enteropathogenic E.coli employs a similar strategy, utilizing its type 3 secretion system to inject as yet unidentified effectors that disrupt host cell phosphatidyl inositol-3 (PI-3) kinase, which is a signaling pathway essential for phagocytosis. Theoretically, if a vaccine could be designed to neutralize antiphagocytic effectors, pathogens such as these would then be susceptible to phagocytosis and subsequent intracellular killing. One precedent for this is the pneumococcal polysaccharide, which is slimy, making it difficult for phagocytic cells to ingest the organism. When antibodies termed opsonins bind to the capsule, phagocytosis and intracellular killing rapidly ensue.
A theme that has emerged in bacterial pathogenesis is the ability to trigger apoptotic mechanisms that lead to programmed cell death of host cells [Weinrauch and Zychlinsky 1999]. This has a significant effect on the host immune system, as apoptotic death is designed not to incite inflammation. Although well-described for viral pathogenesis, the ability to trigger or block apoptosis by bacterial pathogens is just being recognized. As is the case for most pathogenic mechanisms, triggering of apoptosis usually involves the delivery of effectors into host cells that interfere with normal signaling pathways. A good example of this is the Salmonella species. It was observed that this pathogen triggers apoptosis in macrophages within the liver.
It has also been found that there are at least three distinct mechanisms, two of which employ type 3 secretion systems that Salmonella enterica serovar typhimurium employs to trigger apoptotic death in macrophages [Figure 3]. In addition, it appears to have another mechanism that inhibits apoptotic death in epithelial cells by activating a pro-survival pathway. Whether these will represent potentially useful vaccine candidates remains to be discovered.
The Effect of Parasite Variability in Host-Parasite Relationships
Most cytopathic agents leave little leeway between death and survival. In contrast, a wide range of relationships is possible between low or non-cytopathic infections and the host. Many parameters such as interferon production, and susceptibility of the infection to interferons, influence this equilibrium. In addition, mutation rates of the pathogens play a major role. Variability of antigens permits infectious agents to escape T cells or neutralizing antibodies, in the individual as well as at the population level. The variability of the parasite genome