Exotoxins

Exotoxins are a group of soluble proteins that are secreted by the bacterium, enter host cells, and catalyze the covalent modification of a host-cell component(s) to alter the host-cell physiology. Both gram-negative and gram-positive bacteria produce exotoxins. A specific bacterial pathogen may produce a single exotoxin or multiple exotoxins. Each exotoxin possesses a unique mechanism of action, which is responsible for the elicitation of a unique pathology. Thus, the role of exotoxins in bacterial pathogenesis is unique to each exotoxin. Corynebacterium diphtheriae produces diphtheria toxin, which is responsible for the systemic pathology associated with diphtheria, whereas Vibrio cholerae produces cholera toxin, which is responsible for the diarrheal pathology associated with cholera. Exotoxins vary in their cytotoxic potency, with the clostridial neurotoxins being the most potent exotoxins of humans. 

Exotoxins also vary with respect to the host that can be intoxicated. Exotoxin A of Pseudomonas aeruginosa can intoxicate cells from numerous species, whereas other toxins, such as diphtheria toxin are more restricted in the species that can be intoxicated. Some bacterial toxins, such as pertussis toxin, can intoxicate numerous cell types, whereas other toxins, such as the Clostridial neurotoxins, show a specific tropism and intoxicate only cells of neuronal origin. Bacterial exotoxins catalyze specific chemical modifications of hostcell components, such as the ADPribosylation reaction catalyzed by diphtheria toxins or the deamidation reaction catalyzed by the cytotoxic necrotizing factor produced by Escherichia coli. These chemical modifications may either inhibit or stimulate the normal action of the target molecule to yield a clinical pathology. Bacterial exotoxins possess an AB structure–function organization, in which the A domain represents the catalytic domain and the B domain comprises the receptor-binding domain and the translocation domain. The translocation domain is responsible for the delivery of the catalytic A domain into an intracellular compartment of the host cell. Many bacterial exotoxins can be chemically modified to toxoids that no longer expresses cytotoxicity, but may retain immunogenicity. Studies have shown that bacterial toxins can also be genetically engineered to toxoids, which may lead to a wider range of vaccine products. Exotoxins have also been used as therapeutic agents to correct various disorders, including the treatment of muscle spasms by botulinum toxin. Nontoxic forms of exotoxins have been used as carriers for the delivery of heterologous molecules to elicit an immune response and as agents in the development of cell-specific chemotherapy. In addition, bacterial toxins have been used as research tools to assist in defining various eukaryotic metabolic pathways, such as G-protein-mediated signal transduction. 

I. CLASSIFICATION OF EXOTOXINS 

Exotoxins are soluble proteins produced by microorganisms that can enter a host cell and catalyze the covalent modification of a cellular component(s) to alter the host-cell physiology. The term ‘’host cell’’ refers to either vertebrate cells or cells of lower eukaryotes, such as protozoa because some bacterial exotoxins intoxicate a broad range of host cells. The recognition that some pathogenic bacteria produced soluble components that were capable of producing the pathology associated with a particular disease was determined in the late nineteenth century. Roux and Yersin observed that culture filtrates of Corynebacterium diphtheriae were lethal in animal models and that the pathology elicited by the culture filtrate was similar to that observed during the infection by the bacterium. Subsequent studies isolated a protein, diphtheria toxin, from the toxic culture filtrates and observed that the administration of purified diphtheria toxin into animals was sufficient to elicit the pathology ascribed to diphtheria. Diphtheria toxin is a prototype exotoxin and has been used to identify many of the biochemical and molecular properties of bacterial exotoxins. The ability of a bacterial pathogen to cause disease frequently requires the production of exotoxins, but the mere ability to produce a toxin is not sufficient to cause disease. 

Cholera toxin is the principal virulence factor of Vibrio cholerae. Administration of micrograms of purified cholera toxin to human volunteers elicits a diarrheal disease that mimics the magnitude of the natural infection. Nonetheless, nonvirulent toxinproducing strains of V. cholerae have been isolated and shown to lack specific biological properties, such as motility or chemotaxis. Similarly, although anthrax toxin is the principal toxic component of Bacillus anthracis, nonvirulent toxin-producing strains of B. anthrasis have been isolated and shown to lack the ability to produce a polyglutamic acid capsule. An exception to this generalization is the intoxication elicited by the botulinum neurotoxins, in which ingestion of the preformed toxin is responsible for the elicitation of disease; food poisoning by botulinum neurotoxins is an intoxication, rather than an infection by a toxin-producing strain of Clostridium botulinum. Bacterial exotoxins are classified according to their mechanisms of action. The covalent modifications of host-cell components, which are catalyzed by bacterial exotoxins, include ADP-ribosylation, deamidation, depurination, endoproteolysis, and glucosylation (Table 36.1). Most cellular targets of bacterial exotoxins are proteins, although there are exceptions such as shiga toxin, which catalyzes the deadenylation of ribosomal RNA. In addition to exotoxins, there are several other classes of toxins that are produced by bacterial pathogens, including the pore-forming toxins, type III-secreted cytotoxins, heat-stable enterotoxins, and superantigens. Each of these toxins fails to perform one of the properties associated with exotoxins. The pore-forming toxins are not catalytic in their action, but instead disrupt cell physiology through the formation of pores in the host-cell plasma membrane. The type III-secreted cytotoxins can not enter host cells as soluble proteins, but instead are translocated directly into the host cell by the type III secretion apparatus of the cell-bound bacterium. The heatstable enterotoxin and superantigens do not enter the intracellular compartment of the host cell, and elicit host-cell responses by triggering signal-transduction pathways upon binding to the host-cell membrane. In this article, initial emphasis will be placed on the molecular properties of bacterial exotoxins, with a subsequent description of the general properties of pore-forming toxins, type III-secreted cytotoxins, heat-stable enterotoxins, and superantigens. The pathology elicited by a specific exotoxin results from the catalytic covalent modification of a specific host-cell component. Although diphtheria toxin and cholera toxin are both bacterial ADP-ribosylating exotoxins, the pathogenesis elicited by each exotoxin is unique. This is due to the fact that diphtheria toxin ADP-ribosylates elongation factor-2, resulting in the inhibition of protein synthesis and subsequent cell death, whereas cholera toxin ADP-ribosylates the Gs_ component of the heterotrimeric protein, which stimulates the activity of adenylate cyclase. The stimulation of adenylate cyclase elevates intracellular cAMP and the subsequent secretion of electrolytes and H2O from the cell, resulting in the clinical manifestations of cholera. 

II. GENERAL PROPERTIES OF EXOTOXINS 

A. Genetic organization of exotoxins The genes encoding bacterial exotoxins may be located on the chromosome or located on an extrachromosomal element, such as a plasmid or a bacteriophage. Elegant experiments characterizing diphtheria toxin showed that the gene encoding this exotoxin was located within the genome of the lysogenic _-phage. Although both nonlysogenic and lysogenic strains of C. diphtheriae could establish local upper-respiratorytract infection, only strains of C. diphtheriae lysogenized with a _-phage that encoded diphtheria toxin were capable of eliciting systemic disease. This established a basic property for the pathology elicited by bacteria that produce exotoxins; bacteria establish a localized infection and subsequently produce an exotoxin, which is responsible for pathology distal to the site of infection. Most exotoxins are produced only during specific stages of growth with the molecular basis for the regulation of toxin expression varying with each bacterium. This differential expression often reflects a complex regulation of transcription, including responses to environmental conditions, such as iron. Multisubunit toxins are often organized in operons to allow the coordinate expression of their subunit components. B. Secretion of exotoxins from the bacterium Most bacteria secrete exotoxins across the cell membrane by the type II secretion pathway. The secretion of exotoxins by the type II secretion pathway was predicted by the determination that the amino terminus of mature exotoxins had undergone proteolysis relative to the predicted amino acid sequence. Type II secretion is also called the general secretion pathway. Type II secretion involves the coordinate translation and secretion of a nascent polypeptide across the cell membrane. During the translation of the mRNA that encodes a type II-secreted protein, the nascent polypeptide contains an amino-terminal leader sequence that is targeted to and secreted across the cell membrane. After secretion across the cell membrane, the nascent polypeptide folds into its native conformation and the leader sequence is cleaved by a periplasmic leader peptidase to yield a mature exotoxin. 

Some gram-negative bacteria export the assembled exotoxin from the periplasm into the external environment via a complex export apparatus. While the heat-labile enterotoxin of Escherichia coli remains localized within the periplasmic space, V. cholerae and Bordetella pertussis assemble their respective exotoxins, cholera toxin and pertussis toxin, within the periplasm and then transport the mature exotoxin into the external environment. Although the multiple protein components of the export apparatus have been identified, the mechanism for export across the outer membrane remains to be resolved. C. Bacteria produce and secrete exotoxins as proenzymes Although one property of a bacterial exotoxin is the ability to intoxicate sensitive cells, early biochemical studies observed that in vitro many bacterial exotoxins possessed little intrinsic catalytic activity. These perplexing observations were resolved with the determination that bacteria produce and secrete exotoxins as proenzymes, which must be activated (processed) to express catalytic activity in vitro. Because exotoxins intoxicate sensitive cells, the requirements for in vitro activation reflect the activation steps in vivo. Each exotoxin requires specific conditions for activation, including proteolysis, disulfide-bond reduction, or association with a nucleotide or a eukaryotic accessory protein. Some activation processes result in the release of the catalytic A domain from the B domain, whereas other activation processes appear to result in a conformational change in the catalytic A domain, rendering it catalytically active. Some exotoxins require sequential activation steps. Diphtheria toxin is activated by limited proteolysis, followed by disulfidebond reduction (see Fig. 36.1). 

The determination of the activation mechanism of exotoxins has also provided insight into several physiological pathways of host cells. The eukaryotic protein, ARF (ADP-ribosylation factor), which activates cholera toxin in vitro, was subsequently shown to play a central role in vesicle fusion within the eukaryotic cell. The ability of a host-cell extract to activate cholera toxin is often used as a sign of the presence of ARF. Similarly, the characterization of the mechanisms that pertussis toxin and cholera toxin use to intoxicate eukaryotic cells has provided insight into the pathways for eukaryotic G protein-mediated signal transduction. The ability of pertusis toxin to inhibit the action of a ligand in the stimulation of a signaltransduction pathway is often used to implicate a role for G proteins in that signaling pathway. D. AB Structure–function properties of exotoxins Most bacterial exotoxins possess AB structure–function properties. (see Fig. 36.2). The A domain is the catalytic domain, whereas the B domain includes the translocation and binding domains of the exotoxin. Exotoxins are organized into one of several general types of AB organization. The simplest AB organization is represented by the diphtheria toxin, in which the A domain and B domain are contained in a single protein. Diphtheria toxin is the prototype for this class of AB exotoxin. Diphtheria toxin is a 535-amino-acid protein in which the amino terminus constitutes the ADP-ribosyltransferase domain and the carboxyl terminus comprises the translocation domain and receptorbinding domain. The AB5 exotoxins are composed of six proteins that are noncovalently associated as an oligomer. Cholera toxin is the prototype for the AB5 exotoxin. The A domain of cholera toxin constitutes the ADP-ribosyltransferase domain, whereas the B5 domain is composed of five identical proteins, forming a pentamer. This is organized into a ring structure, on which the A domain is positioned. The five proteins that make up the B domain may be identical, as is the case for cholera toxin and the heat-labile enterotoxin of E. coli, or may be different proteins that form a nonsymmetrical ring structure, as observed with the B oligomer of pertussis toxin. The third class of AB exotoxin is composed of proteins that are not associated in solution, but that do associate following the binding and processing of the B domain to the host cell. C2 toxin is an example of this class of A-B exotoxin. C2 toxin is a bipartile exotoxin composed of a protein that encodes the catalytic A domain and a separate protein that encodes the B domain. 

The A domain protein of C2 toxin ADPribosylates actin. The B domain protein of C2 binds to sensitive cells and is nicked by a eukaryotic protease. The processed B components oligomerize and are then capable of binding either of the A domain proteins. A new class of toxin organization has recently been recognized in which the A domain is a protein and the bacterium is directly responsible for its delivery into the cell. The bacterium binds to the eukaryotic cell and uses a type III secretion apparatus to deliver cytotoxins, also called effector proteins, into the intracellular compartment of the cell. The YopE cytotoxin of Yersinia is the prototype of this group of toxins. These cytotoxins do not conform to the strict definition of exotoxins because the purified cytotoxins cannot directly enter to modify host-cell physiology and are not included with the family of bacterial exotoxins. They are termed ‘’type III secreted cytotoxins’’, and are described later. Although the A domain possesses the catalytic activity of the exotoxin, the B domain possesses two specific functions, receptor binding and translocation capacity. Each exotoxin uses a unique host-cell surface component as a receptor. The cell surface receptor for each exotoxin may be specific. The cell surface receptor for cholera toxin is the ganglioside, GM1, whereas diphtheria toxin binds directly to the epidermal growth factor precursor. In contrast, the binding of pertussis toxin appears to be less specific, as pertussis toxin is able to bind numerous cell surface proteins. The ability to bind its cell surface receptor is an absolute requirement for an exotoxin to intoxicate a host cell because the deletion of the receptor-binding domain renders the exotoxin essentially noncytotoxic. After binding to the cell surface, some exotoxins are proteolytically processed or are processed during endocytic vesicle transport. The second function of the B domain includes translocation capacity, which is responsible for the delivery of the A domain across the cell membrane. 

The presence of a translocation domain was predicted from early structure–function studies of diphtheria toxin, which showed that in addition to the catalytic domain and receptor-binding domain, a third function was required for the efficient expression of cytotoxicity. This third function was subsequently shown to correspond to a region of diphtheria toxin that had the propensity to interact with membranes. The crystal structure of diphtheria toxin revealed the presence of three distinct domains, representing the catalytic, translocation, and receptor-binding functions. E. Exotoxins enter host cells via distinct pathways Although A domain translocation is one of the least understood aspects of the intoxication process of exotoxins, there are several general themes that are involved in translocation of the A domain across the cell membrane. One translocation mechanism uses a pH gradient within the endosome to stimulate protein conformational changes in the B domain, making it competent to interact with the endocytic vesicle. After insertion into the endocytic membrane, the B domain generates a pore that is believed to be involved in the translocation of the A domain across the vesicle membrane in an unfolded form. After the translocation across the endocytic membrane, the A domain refolds to its native conformation. Subsequent to translocation of the A domain across the vesicle membrane, reduced glutathione may reduce the disulfide that connects the Adomain with the B domain, and release the Adomain into the cytoplasm. The potency and catalytic potential of exotoxins was demonstrated by the observation that the introduction of one molecule of the catalytic domain of diphtheria toxin into the intracellular cytoplasm was sufficient to inhibit host-cell physiology, resulting in cell death. Other toxins, such as cholera toxin and exotoxin A of Pseudomonas aeruginosa, appear to use retrograde transport to enter the interior regions of the cell. Movement appears to occur through retrograde transport from the endosome to the Golgi apparatus and ultimately to the endoplasmic reticulum. Many exotoxins that are ultimately delivered to the endoplasmic reticulum possess a KDEL (Lys-Asp-Glu-Leu)-like retention signal sequence on their carboxyl terminus. 

Although the details for the actual transport pathway remain to be determined, studies with chimeric proteins have shown that the introduction of a KDEL retention sequence is sufficient to retrograde transport a protein, which is normally delivered only to the early endosome, into the endoplasmic reticulum. Thus, there is physiological precedence for the use of the KDEL sequence to retrograde transport exotoxins toward the endoplasmic reticulum. One of the basic questions concerning the intoxication process of these exotoxins is the actual mechanism of translocation and whether or not eukaryotic proteins assist in the translocation process. F. Covalent modification of host-cell components by exotoxins Exotoxins use several unique mechanisms to covalently modify host-cell components. The major classes of reactions are the covalent addition of a chemical group to the target protein, the cleavage of a chemical group from a target protein, or the endoproteolytic cleavage of a peptide bond of the target protein. The ADP-ribosylation of host proteins is the prototype mechanism of action of bacterial exotoxins. Numerous bacterial exotoxins catalyze the ADPribosylation of specific host proteins and elicit physiological changes. In the ADP-ribosylation reaction, exotoxins use the oxidized form of nicotinamide adenine dinucleotide (NAD) as the substrate, and transfer the ADP-ribose portion of NAD to a specific amino acid via an N-glycosidic linkage of ADP-ribose onto the host target protein. The specific type of amino acid that is ADP-ribosylated within the target protein varies with the specific exotoxin. ADP-ribosylation may either inactivate or stimulate the activity of the target protein. 

Diphtheria toxin ADP-ribosylates elongation factor-2 on a post-translationally modified histidine residue called diphthamide. ADP-ribosylated elongation factor-2 is unable to perform its translocation of nascent polypeptides in the ribosome, which results in the inhibition of protein synthesis and subsequent cell death. In contrast, cholera toxin ADP-ribosylates the Gs_ component of a heterotrimeric G protein. ADP-ribosylated Gs_ is locked in an active conformation, which results in the stimulation of adenylate cyclase and the subsequent elevation of intracellular cAMP. Likewise, deamidation of Gln63 of RhoA by E. coli cytotoxic necrotizing factor (CNF) results in a constitutively active RhoA protein. Note that although most host targets for exotoxins are proteins, Shiga toxin catalyzes the deadenylation of a specific adenine on 28S RNA. Recall that each exotoxin modifies a specific host-cell component, which is responsible for the specific pathology elicited by that exotoxin. Although there are no absolute rules for the types of proteins targeted for covalent modification, the most frequent targets are nucleotide-binding proteins that are involved in signaltransduction pathways, including both the heterotrimeric G-proteins and the small-molecular-weight GTP-binding proteins of the Ras superfamily. It is not clear whether this class of host protein is targeted for modification due to the presence of a common structural motif, or to its critical role in host-cell metabolism. G. Molecular and structural properties of bacterial exotoxins Early biochemical studies provided significant advances in defining the structure–function properties of exotoxins, resolving many of the exotoxin mechanisms of action and developing the concept that exotoxins have AB organization. Molecular genetics and structural biology have extended the earlier studies and provided a more detailed understanding of the biochemical and molecular relationships among the exotoxins. The biochemical characterization of diphtheria toxin and exotoxin A (ETA) of Pseudomonas aeruginosa showed that these two exotoxins catalyzed kinetically identical reactions during the ADP-ribosylation of elongation factor-2. 

In addition, both diphtheria toxin and ETAwere shown to possess an active site glutamic acid, which was subsequently shown to be a signature property of exotoxins that catalyze the ADP-ribosyltransferase reaction. These observations predicted that ADP-ribosylating exotoxins would possess considerable primary amino acid homology. Thus, the determination that the genes encoding diphtheria toxin and ETA shared little primary amino acid homology was unexpected. This paradox was resolved after the analysis of the threedimensional structures of ETA and the heat-labile enterotoxin (LT) of E. coli, and subsequently confirmed with diphtheria toxin. The three-dimensional structures of ETA and LT showed little similarity in their respective receptor-binding domains and translocation domains; however, the catalytic domains of ETAand LT, which are composed of seven discontinuous regions of each protein, could be superimposed on each other despite possessing homology at only three of the 43 amino acids. One of the homologous amino acids in ETA and LT was the signature active site glutamic acid. This was a remarkable finding because ETA and LT ADP-ribosylate different host target proteins and possess different AB organization. Acommon theme has evolved for describing the structure– function properties of this family of bacterial exotoxins in which the ADP-ribosylating exotoxins possess a conserved three-dimensional structure in their active sites, despite the lack of primary amino acid homology. These findings have provided a framework for the study of other classes of exotoxins produced by divergent groups of bacteria. 

III. CONVERSION OF EXOTOXINS INTO TOXOIDS 

A. Chemical detoxification of bacterial exotoxins Shortly after the determination that toxic components were associated with bacterial pathogens, several studies showed that cell extracts or cell cultures of a pathogen could be treated with chemical denaturants, such as formalin, to produce nontoxic immunogenic material that could prevent the disease associated with that pathogen. In the case of diphtheria toxin and tetanus toxin, chemical modification with formalin produced toxoids that were used as acellular vaccines in large-scale immunizations. This resulted in a remarkable decrease in the incidence of both diphtheria and tetanus within the populations that were immunized. In areas where these toxoids are not administered, diphtheria and tetanus remain clinically important diseases. In addition to formalin, other chemicals have been used to detoxify bacterial exotoxins, including glutaraldehyde and hydrogen peroxide. In contrast, the chemical toxoiding of other exotoxins, such as cholera toxin and pertussis toxin, has been more difficult because the treatment of these toxins with denaturants often results in a reduction of immunogenicity. Thus, there is a need to develop alternative strategies for eliminating the cytotoxicity of certain exotoxins without compromising their immunogenicity. B. Genetic detoxification of bacterial exotoxins Developments in genetic engineering have provided an opportunity to produce recombinant forms of bacterial exotoxins that possess greatly reduced toxicity, but retain immunogenicity. The use of genetic engineering to develop a toxoid of pertussis toxin has been successful. The whole-cell pertussis vaccine is composed of a chemically treated preparation of Bordetella pertussis, which is effective in the elicitation of a protective immune response after mass immunization. However, the whole-cell pertussis vaccine is acutely reactive when administered to children. Pertussis toxin, a primary virulence determinant of B. pertussis, is an exotoxin that ADP-ribosylates the Gi_ component of heterotrimeric G proteins and effectively uncouples signal transduction between the G protein-coupled receptor and the G protein. Genetically engineered forms of pertussis toxin have been produced that possess essentially no catalytic activity or cytotoxicity, but that maintain native conformation and elicit a protective immune response when used as an immunogen. These recombinant noncytotoxic forms of pertussis toxin have been engineered with multiple mutations in their active site, virtually eliminating the risk of reversion to a cytotoxic form. Similar strategies are being applied to other bacterial exotoxins with the goal of engineering acellular vaccine candidates. 

IV. THERAPEUTIC APPLICATIONS OF EXOTOXINS 

One of the most exciting areas of bacterial exotoxin research has been the development of strategies to use exotoxins in therapeutic disciplines. Some therapies use the native cytotoxic form of the exotoxin. Other therapies use either the A or B domain, which is conjugated to a heterologous binding component or to effector elements, respectively, to produce a chimeric molecule with directed properties. Botulinum toxin and tetanus toxin (BT/TT) are each a single protein that is organized as an AB exotoxin. The amino terminus of BT/TT expresses endopeptidase activity and constitutes the A domain, whereas the B domain possesses neuronal cell-specific receptorbinding activity. The specific association of the B domain with neuronal cells is responsible for the clinical manifestation of these neurotoxins. BT/TT appear to enter neuronal cells by receptor-mediated endocytosis and to deliver the A domain to the cytosol, where it catalyzes the endoproteolytic cleavage of host proteins that are involved in vesicle fusion. Studies have shown that botulinum toxin can be introduced into the muscles surrounding the eye to temporarily reduce muscle spasms associated with several clinical disorders. Diphtheria toxin has been used as a carrier to stimulate an immune response against several epitopes. One epitope is polyribitolphosphate, a component of the polysaccharide capsule of Haemophilus influenzae type b (Hib). Early attempts to elicit an effective immune response to purified Hib antigen resulted in the production of a T-cell-independent immune response that did not yield an effective memory. A noncatalytic mutant of diphtheria toxin, CRM197, has been used as a carrier for the Hib epitope. Immunization with the CRM197-Hib conjugate yielded a strong T-dependent immune response. Mass immunization with Hib conjugates has resulted in a dramatic reduction in the number of cases of Hib in the immunized population. Due to their potency, the catalytic A domain of exotoxins have been used in the construction of chimeric immunotoxins that are designed to target cancer cells. Early studies used conjugates that were composed of the Adomain of the diphtheria toxin coupled to an antibody that recognized a cell surface-specific antigen. The Achain of the diphtheria toxin was used in the first generation of immunotoxins because it had been shown to possess impressive cytotoxic potential when introduced into the cytosol of eukaryotic cells. It was estimated that the introduction of a single molecule of the A chain of diphtheria toxin into the cytosol was sufficient to kill that cell. In cell culture, these chimera have proven to be both potent and antigen specific. Ongoing research involves the determination of clinical situations for the use of these chimeras in a therapeutic arena. The B component of anthrax toxin, protective antigen (PA) and a truncated, non-cytotoxic form of one of its A components (LF) has recently been used to deliver epitopes into antigen presenting cells to elicit a cytotoxic lymphocyte (CTL) response. Anthrax toxin is a tripartite toxin composed of three non-associated proteins. After binding to cells, PA is proteolytically processed and undergoes oligomerization to form a heptameric structure on the cell surface. Processed PA is able to bind either LF or edema factor (EF) and the AB complex undergoes receptor-mediated endocytosis in which acidification in the early endosome stimulates the translocation of the A domain into the cell cytosol. In the nontoxic anthrax delivery system, PA is added to antigen-presenting cells with a nontoxic LF–CTL epitope chimera used to deliver the epitope into the host cell for antigen presentation. One of the more attractive aspects of this CTL-epitope delivery system is that small amounts of PA are required to present antigen. 

V. OTHER CLASSES OF BACTERIAL TOXINS 

A. Pore-forming toxins The lack of a catalytic A domain differentiates the poreforming toxins from exotoxins. Thus, the pathology associated with pore-forming toxins is due solely to the generation of a pore within the membrane of the host cell. Several bacterial pathogens produce pore-forming toxins, some of which are secreted by a type I secretion pathway. Unlike type II-secreted proteins, the amino terminus of type I-secreted proteins is not processed. Type I-secreted proteins possess a polyglycine signal sequence in the carboxyl terminus of the mature toxin. There are several classes of pore-forming toxins, including members of the hemolysin family of pore-forming toxins, the aerolysin family of pore-forming toxins, and the _-toxin of Staphylococcus aureus. Host-cell specificity differs among pore-forming toxins. The crystal structures of several of the pore-forming toxins have been determined. The molecular events generating a pore in the membrane of a host cell have been proposed for the aerolysin family of pore-forming toxins. Aerolysin is exported by Aeromonas hydrophilia as a monomeric molecule, which binds to the host cell. The monomer is pro-teolytically processed and subsequently undergoes oligomerization. The oligomerized complex is inserted into the membrane and generates a pore in the center of the complex, causing the release of the cytoplasmic components of the host cell. B. Type III-secreted cytotoxins The lack of a B domain differentiates the type IIIsecreted cytotoxins from exotoxins. Thus, the organization of the type III-secreted cytotoxins may be represented as A domains that are specific effector proteins. Type III-secreted cytotoxins are transported directly into the host cells by cell surface-bound bacteria. Type III secretion of bacterial proteins is a recently defined pathway for the delivery of proteins into the cytoplasm of host cells. Type III-secreted proteins were initially recognized by the fact that the secreted mature cytotoxins were unique to proteins secreted by either the type I or II secretion pathways, whereas the amino terminus of type III-secreted proteins is not processed nor is there a polyglycine motif in their carboxyl terminus. Although it is clear that a complete type III secretion apparatus is required for the transport of type III-secreted proteins into the host cytoplasm, the mechanism for the delivery of type IIIsecreted proteins across the host-cell membrane remains to be resolved. 

Numerous bacteria have been shown to possess type III secretion pathways, including members of the genera Escherichia, Pseudomonas, Shigella, Salmonella, and Yersinia. Cytotoxicity elicited by type III-secreted cytotoxins has an absolute requirement for the type III secretion apparatus of the bacterium, as purified forms of the cytotoxins are not toxic to host cells. The A domains of type III-secreted cytotoxins catalyze several unique mechanisms of action, including the depolymerization of the actin cytoskeleton, phosphatase activity, ADP-ribosyltransferase activity, and the stimulation of apoptosis. Each type III secretion apparatus appears capable of delivering numerous type III-secreted proteins into the host cell. C. Heat-stable enterotoxins The inability of the heat-stable enterotoxins to enter the host cell or possess catalytic activity differentiates the heat-stable enterotoxins from exotoxins. Several genera of bacteria produce heat-stable enterotoxins, including Escherichia and Yersinia. The heat-stable enterotoxin a (STa) of E. coli is the prototype toxin of this group. E. coli secretes STa into the periplasm as a 72-amino-acid precursor in which three intramolecular disulfide bonds are formed and processed into a 53-amino-acid form. The 53-amino-acid form of STa is exported into the environment, where a second proteolytic cleavage results in the production of an 18- or 19-amino-acid mature STa molecule. The mature STa binds to a protein receptor on the surface of epithelial cells, which results in an increase in the intracellular concentrations of cGMP. The intracellular increase in cGMP results in a stimulation of chloride secretion and net fluid secretion, resulting in diarrhea. D. Superantigens The inability of superantigens to enter the host cell or possess catalytic activity differentiates the superantigens from exotoxins. Superantigens are soluble proteins of approximately 30 kDa that are secreted by bacteria that possess mitogenic properties. Superantigens are produced by both Streptococcus and Staphylococcus. The superantigens bind to a component of the major histocompatibility complex of T lymphocytes through an antigen-independent mechanism, which stimulates proliferation of a large subset of T lymphocytes.