Copyright � April 17, 2002
by: Sebastian Molnar
Abstract.
Bacillus anthracis is primarily a pathogen of livestock, but it can also infect humans. Anthrax has been a source of fear with its potential use as a biological weapon. With the threat of bioterrorism, research directed towards understanding the mechanisms and regulation of virulence is crucial in developing therapies, such as vaccines, antibiotics, and antitoxins. Full virulence of anthrax requires two plasmids: pXO1 (182kb) containing toxin genes (pag, lef, cya) and a germination operon (gerX), and pXO2 (92kbp) containing genes for capsule synthesis (capB, A, C and dep). The components of anthrax toxin have been well studied. Edema factor (EF) is an adenylate cyclase that drastically increases cAMP levels in the host cell and causes local edema in the skin. Lethal factor (LF) is a zinc-metalloprotease that cleaves proteins involved in signaling pathways (i.e. MAPKKs) and is the primary cause of death of the host. The protective antigen (PA) binds EF or LF and mediates their translocation into the host cell. The mechanism of translocation is unclear.
The factors involved in regulation of virulence have been studied to some extent. AtxA is essential for activating toxin gene expression. However, the molecular mechanism by which it regulates is unknown, since AtxA does not contain any known DNA-binding motifs. It is known that AtxA and a related protein, AcpA, require a cis-element 50 to 70 base pairs upstream of the capB promoter for mediating its regulation. One objective of this proposal is to characterize the role of AtxA in regulating the atxA regulon. This will be done through identifying DNA sequences essential for regulation (e.g. via SAAB assays) and through protein-protein interaction assays (e.g. two-hybrid, TAP). Various mutagenesis analyses (deletion analysis, site-directed mutagenesis, error-prone PCR) will also be used to identify functional domains and essential amino acid residues in AtxA and in other proteins identified to play a role in the regulation process. Gel filtration, and analytical ultracentrifugation will be used to determine the stoichiometry of �active� AtxA and AcpA, which can be coupled with x-ray crystallography to precisely determine functional domains.
The second objective is to identify genes (e.g. mutants obtained in the above studies, or with random transposon mutagenesis) involved in particular stages of infection. Insertions of Tn917-LTV3 into promoters regulated by AtxA will be screened for inability to �release� from macrophages. This is an important step in pathogenesis as it allows bacilli to enter the blood stream of the infected animal and cause further infections.
Finally, an investigation of global regulatory systems on the atxA regulon will be carried out on various regulator-knockout strains to study relative virulence gene expression. This will be done using real-time PCR, which can quantitate expression levels of many samples over several time points. These studies will lead to a greater understanding of pathogenicity regulation in B. anthracis.
Introduction.
1: Historical Background
The Gram-positive spore-forming soil bacterium Bacillus anthracis is the causative agent of anthrax and can infect livestock and humans [1]. B. anthracis was first discovered in 1850 by a French parasitologist, Casimir-Joseph Davaine, and was later studied by Robert Koch and by Louis Pasteur [2]. By 1876, Koch developed methods for culturing B. anthracis [2]. In 1881, Pasteur began a large-scale program for the vaccination of livestock against anthrax [3]. Anthrax is therefore considered to have been instrumental in the founding of Bacteriology and Immunology [4,5]. In the1950s and 1960s, three components (i.e. protective antigen, edema factor, and lethal factor) of the anthrax toxin were identified [6-8]. Subsequent research on the anthrax toxin has uncovered some details into its molecular mode of action.
Although anthrax was originally discovered as an animal pathogen, and the development of vaccines was intended for the protection of livestock against the disease, humans are also susceptible. Because of this, anthrax has become a source of fear with its potential use as a biological weapon. Anthrax is one of the top four microbial agents considered to be likely candidates used as biological weapons, ranking only second to smallpox [9-11]. Despite the 1972 Biological Weapons Convention, several nations, including the USSR and Iraq, as well as autonomous groups, such as the Japanese cult Aum Shinrikyo, are known to maintain stores of biological weapons [11-13]. In 1979, an outbreak occurred due to an accidental explosion that released an aerosol of anthrax spores from a military factory near Sverdlovsk (now Ekaterinburg, Russia), a city of 1.2 million people [14]. Although conflicting reports obscure the number of people actually infected with anthrax, at least 77 people had been infected and 66 deaths occurred due to inhalational infection [14,15]. From the Gulf War, reports revealed that Iraq developed an extensive bioweapons program, having stockpiles of botulinum and anthrax warheads [9,13]. Of more recent public concern, letters deliberately contaminated with anthrax spores had been sent in the mail by terrorists shortly after the September 11, 2001 attacks on the USA [16-18]. This resulted in 11 cases of inhalational anthrax, 12 cases of cutaneous anthrax, and four deaths [16,18-21]. Acts of terrorism are difficult to predict or prevent. With the reality of bioterrorism, understanding the mechanisms of anthrax toxin and of B. anthracis virulence, as well as developing rapid and effective treatments for the disease, should be of particular importance in pathogen research.
2: Anthrax Infection and Treatment
There are three modes of anthrax infection in humans -- inhalational, cutaneous, and gastrointestinal -- each one differing in severity [extensively reviewed in ref. 5]. The names of these infection modes, however, refer only to their point of entry into the body. For example, the victims of the Sverdlovsk incident with inhalational infection showed signs of gastrointestinal lesions [15]. The inhalation of anthrax spores is the most severe form of infection, resulting in fatalities of more than 80% of cases if left untreated [5,14,22]. Death can occur within a few days [11,16]. While the inhalational form is rare, cutaneous anthrax is the most common mode of infection (acquired by handling infected animals and animal products), but is much less severe than inhalational anthrax5. Symptoms include local edema in the skin after spore germination, with fatalities occurring up to 20% without antibiotic treatment [11]. No deaths occurred in victims from the Sverdlovsk outbreak due to cutaneous infection [14]. Gastrointestinal infection, which can be fatal, is considered rare and occurs through eating insufficiently cooked meat from infected animals [4,5]. Anthrax meningitis is a rare complication of infection, and is nearly always fatal [5]. Considering these modes of infection, it is possible to improve current (or develop novel) therapies and preventative strategies to treat the disease and avoid outbreaks.
Infection through B. anthracis spores involves several stages, and a model is shown in figure-1. These include (1) engulfment by the macrophage, (2) germination of the endospore, (3) escape from the phagolysosome, (4) survival in the macrophage intracellular environment, (5) release from the macrophage, and (6) bacterial growth in the host organism leading to further infection. There is a consensus that spores germinate and survive within macrophages [23,24]. Anthrax toxin production was shown to occur immediately after germination of the endospore, which may be required for enhancing bacterial survival in the macrophage [24,25]. More details on the toxin components will be discussed in the next section.
Vaccines, antibiotics, and antitoxin treatments are the current means of dealing with anthrax infection [5,11,26-29]. The evolution of microbial resistance to antibiotics is an ongoing concern in the treatment of infections. Humans have had a major impact on the emergence of resistance due to widespread antibiotic use [30,31]. For example, Staphylococcus aureus is a major human pathogen that has acquired methicillin-resistance independently by lateral gene transfer in several strains [32,33]. Vaccine and antibiotic resistance have also been detected in some strains of B. anthracis [22,34,35]. Strong selection pressures introduced by humans may favor the maintenance of antibiotic resistance genes and their horizontal transfer via mobile genetic elements, such as plasmids, transposons, bacteriophages, and integrons [36,37]. Methods to control the evolution of antibiotic resistance and to improve the effectiveness of current treatments have therefore been suggested [31,38-42]. Understanding the molecular mechanisms involved in anthrax virulence and intoxication will be useful for targeting specific therapies when infection occurs. A partial sequence of the Bacillus anthracis genome is available from The Institute for Genomic Research (TIGR; http://www.tigr.org) and is currently being annotated [43]. Completion of the B. anthracis sequencing project (including up to 20 strains) will undoubtedly provide a wealth of information for such studies in anthrax treatments. The available sequence data will also be useful for this research proposal, which will focus on understanding the mechanisms of virulence and gene regulation in Bacillus anthracis.
3: Anthrax Toxin
The anthrax toxin is composed of three proteins: protective antigen (PA), lethal factor (LF), and edema factor (EF) [reviewed in ref. 44]. Only PA elicits an immune response, which is enhanced in the presence of EF8. Interestingly, the overall immunogenicity of PA and EF is diminished with addition of LF8. However, when injected alone, none of these proteins are lethal. LF and EF each require the presence of PA to mediate their toxic effects on the host cell [7]. Therefore, anthrax has been classified as an AB-toxin: the A moiety (i.e. either LF or EF) has enzymatic activity leading to changes in cellular physiology or death, and the B moiety (i.e. PA) binds the A component to cell-surface receptors and is involved in translocating the toxins into the cytoplasm [45]. PA-LF (lethal toxin) is responsible for causing death in the infected organism, and PA-EF (edema toxin) causes local edema in the skin [7,46]. It is possible that the lethal and edema toxins act synergistically in the host [47].
The anthrax toxin components have been sequenced and cloned [1], which allows for genetic manipulation [46,47]. The crystal structures of PA, LF, and EF have also been determined [48-50]. All three components are encoded on the pXO1 plasmid by pagA, lef, and cya, respectively, and are located in a region identified as a pathogenicity island [51]. Pathogenicity islands are DNA regions containing virulence genes and are associated with transposases, insertion sequences, transposons, or prophages, which suggests their acquisition through lateral gene transfer [52]. The pXO1 pathogenicity island also contains atxA -- a trans-activator that positively regulates the expression of toxin genes on pXO1 and a capsule biosynthesis gene (capB) on pXO2 in response to CO2 or bicarbonate [53,54]. Presumably, CO2/bicarbonate acts as a signal to trigger expression of toxin genes upon infection of mammalian cells.
Although the major components involved in anthrax intoxication have been identified, some details of virulence and toxin production (e.g. regulation) still need to be investigated. A general model for anthrax toxin uptake will be discussed below, along with a description of what is currently known about the individual toxin components.
3.1: A model for anthrax toxin uptake and the role of protective antigen
Several steps are involved in anthrax intoxication and a model for toxin uptake (figure 2) has been suggested [1,48]. Upon infection, protective antigen (83 kDa) binds to a cell surface receptor and is then cleaved by a furin-like protease [55-58]. Furin is a protease that is involved in a number of different signaling and transport pathways [reviewed in ref 59], and proteolytic cleavage using host-cell proteases appears to be a general strategy exploited by other bacteria for exotoxin activation [57]. Mutagenesis or deletion of the furin cleavage site in PA prevents anthrax toxicity [58,60]. The crystal structure for PA is shown in figure 3. Proteolytic cleavage at domain 1' releases an N-terminal 20 kDa fragment (PA20) from PA leaving a C-terminal 63 kDa fragment (PA63) attached to the cell surface receptor [48,60]. Removal of the PA20 fragment allows for the heptamerization of PA63 via domain 2 and exposes a binding site at domain 3 for EF or LF [48,61,62. Domain 4 of PA binds to the cell receptor. The receptor itself has been identified as a type I membrane protein and has an extracellular von Willebrand factor A (VWA) domain for binding PA [33,36]. A solubilized VWA domain from ATR was shown to inhibit toxin action, which may provide an antitoxin treatment [55].
The protective antigen mediates transport of EF or LF into the cytoplasm of the host-cell [7]. Internalization of the cell-surface bound PA63-LF or PA63-EF complexes occurs through receptor-mediated endocytosis and is followed by endosomal acidification [48,63,64]. Acidification induces a conformational change in heptameric PA63, which leads to the formation of a 14-stranded beta-barrel pore inserted into the membrane [48,62,64,65]. The pore structure of PA63 is shown in figure 4. Although each PA monomer can potentially bind to one EF or LF molecule, it was demonstrated that only three ligands of LF or EF can actually bind at one time due to physical occlusion at adjacent binding sites [66]. Point mutations in beta-barrel loops of PA were shown to abolish translocation of EF and LF, without interfering with receptor and toxin binding [67]. Some details of the translocation mechanism, such as how the enzyme moieties actually cross a membrane, remain uncertain [65,66]. Overall, PA is involved in directly binding to EF or LF and to the cell receptor, and it has the role of translocating these virulence factors into the host-cell cytoplasm.
3.2: Edema Factor
An edematous response (swelling by the excessive accumulation of serous fluid in connective tissue) in the skin of infected animals is elicited by the edema toxin [7]. Edema toxin may also cause death in a small portion of infected animals [7]. The effects of EF can be reduced by the presence of LF [7,68]. EF was demonstrated to be an adenylate cyclase that functions in the cytoplasm and increases cAMP concentrations in the host cell by 200-fold [68].
The crystal structure of EF has been determined [50] and is shown in figure 5. EF is synthesized and secreted from B. anthracis in an inactive form. Upon entering the cytoplasm, calmodulin from the host-cell binds EF and activates the toxin [68]. This was suggested by in vitro studies where EF activity was retained by replacing CHO cell lysate with calmodulin, and inhibited with the addition of calcium chelators [68]. Calmodulin (CaM) is a ubiquitous calcium-binding protein in eukaryotes that is involved in a variety of cellular processes, by binding to and activating proteins in signaling pathways [69]. It has been suggested that EF binds CaM irreversibly under conditions of low calcium levels, and therefore prevents CaM from activating other proteins, such as enzymes involved in the breakdown of cAMP to ATP [70]. Calmodulin (CaM) is required to activate EF and does so by inducing a significant conformational change upon binding [50,71,72]. A helical region and three switches (A, B, and C) on EF contact CaM directly (figure 5b). The catalytic core consists of two domains, with a single magnesium ion held by two aspartate residues in the active site50. Activation by CaM enables EF to bind ATP at the active site; without CaM, ATP-binding is severely reduced [72-74].
3.3: Lethal Factor
The major cause of death in organisms infected with anthrax is due to the lethal toxin [7]. Lethal toxin specifically targets macrophages and has little effect on other cell types [75]. This is in contrast to edema toxin, which was found to not be cell specific in raising cAMP levels [68]. LF is a zinc-metalloprotease (i.e. a zinc atom is found in the catalytic site) that specifically cleaves members of the MAPKK (mitogen-activated protein kinase kinase) family [28,29,76-78]. The crystal structure of LF has been determined enabling the identification of its functional domains [49]. Domain I is involved in binding to PA, while domains II, III, and IV bind to MAPKK (figure 6). Domain IV is the catalytic centre and contains the zinc ion.
MAPKKs are involved in a number of signaling pathways in eukaryotes, from sporulation in yeast to apoptosis in mammals [reviewed in ref. 79]. LF cleaves MAPKKs involved signaling pathways for the release of cytokines, including TNF-alpha, IL-1 and IL-6, from macrophages [78]. It is thought that, late in the infection process, lysis of macrophages release substantial levels of nitric oxide and TNF-alpha into the bloodstream, which leads to systemic shock and sudden death [75,78]. Certain macrophage lines can survive lethal toxin treatment even though cleavage of MAPKKs occurs. Therefore, the complete pathway leading to death by anthrax toxin remains to be determined.
4: Virulence Regulation and Research Direction
An important issue in pathogenicity research is how virulence genes are regulated. A model of B. anthracis toxin and capsule gene regulation is shown in figure 7. The anthrax toxin structural components on pXO1, and the capsule synthesis gene capB on pXO2, are regulated in response to CO2/bicarbonate by the trans-activator AtxA [80]. AtxA itself is synthesized in response to temperature, but not in response to CO2/bicarbonate [81]. AtxA is essential for toxin gene expression [82]. Other regulators of virulence genes, which also influence expression of AtxA, include PagR and AbrB. The pag operon contains pagR, which is co-transcribed with pagR (i.e. protective antigen) and inhibits the expression of both pagR and atxA [83]. AbrB is a �transition-state regulator� in B. subtilis that regulates many genes and mediates the transition from active growth to stationary phase, and vice versa [84,85]. abrB is present in two copies in B. anthracis, a functional one on the chromosome and apparently non-functional one on pXO1. The chromosomal copy has been shown to repress toxin and atxA expression in B. anthracis, while the plasmid-encoded version does not [85]. Phosphorylated Spo0A -- the major transcription factor involved in sporulation -- represses abrB expression in B. subtilis and B. anthracis [84-86]. Thus, global regulatory networks influence the synthesis of toxin genes, however, this is only poorly understood.
The mechanism of atxA-mediated expression of toxins and capB is unknown -- AtxA does not have any known DNA-binding motifs. AtxA has sequence similarity to AcpA, a trans-activator encoded on pXO2, which also lacks any recognizable DNA-binding motif. AcpA is expressed in response to CO2/bicarbonate.and regulates the polyglutamate capsule synthesis genes, capB, C, A and dep [87]. Deletion analysis has shown that a region 50 to 70 bp upstream of the capB gene on pXO2 is required for its regulation by AcpA and AtxA [80]. The actual sequence necessary for capB expression has not yet been elucidated. Shown with deletion analysis, a DNA region between 54 bp and 111 bp upstream of the pag promoter is required for transcription of pag by AtxA [82]. Uchida et al. (unpublished data) also suggested that AtxA binds upstream of the pag promoter by using a mobility-shift assay. Currently, however, there is no evidence that AtxA binds directly to DNA.
AtxA probably regulates genes other than those for toxins and capsule synthesis -- this has been suggested through the phenotypic effects of various B. anthracis null-mutants. For example, lethal factor and protective antigen are not required for release of bacilli from the macrophage, whereas atxA is essential for this step in infection [23]. Furthermore, atxA may regulate genes that influence growth and sporulation. As noted by Hoffmaster and Koehler [83], although atxA-null mutants have comparable growth rates to the parental Sterne strain in rich media, they have relatively poor growth in minimal media. The atxA-null mutants can also sporulate more efficiently in rich media than the Sterne strain. It is likely that atxA regulates genes that have not yet been identified.
There are several possible explanations for the mechanism of atxA-mediated expression. If AtxA can bind DNA, it may have a novel DNA-binding domain or it may recognize local DNA structure in a manner analogous to histone-like DNA-binding proteins [82]. Since there are no obvious sequence similarities in the promoter regions of the three toxin genes, the latter case seems plausible. An alternative is that AtxA interacts with other proteins involved in a regulatory network for virulence gene expression [54,81]. This proposal is directed towards resolving these issues and understanding the regulatory aspects of virulence gene expression in Bacillus anthracis, with particular focus on AtxA.
Anthrax Research Plan
Anthrax References
[MICROBES and EVOLUTION]
[EVOLUTION and the ORIGINS of LIFE]
|
Get your own Free Home Page