Fighting anthrax with flies

Anthrax is an infectious disease caused by the spore-forming bacterium Bacillus anthracis (1, 2). The endospores of the bacterium are remarkably resistant to physical stress and highly infectious (2). In October 2001, the attacks on the World Trade Center were followed by an outbreak of anthrax in the United States after the intentional release of spores through contaminated letters, leading to widespread panic and, eventually, the deaths of five people (1). These cases intensified the fight against the disease and prompted a more thorough examination of the mechanisms underlying the pathogenicity of threatening organisms, such as B. anthracis. In this issue of PNAS, Guichard et al. (3) observe that anthrax toxicity factors in Drosophila melanogaster act against similar cellular targets as those that have been identified in mammals, indicating that Drosophila may help fight disease and terror.

The major challenge that is posed by anthrax is that the symptoms cannot be easily distinguished from influenza-like illness. Therefore, critical time might pass before an effective treatment can be applied. The anthrax toxicity results from the activity of three secreted polypeptides, the protective antigen (PA), lethal factor (LF), and edema factor (EF) (1, 2, 4). The PA binds to specific receptors on the surface of target cells. After its proteolytic processing, it polymerizes to form a heptameric pore in the membrane, facilitating binding and translocation of the enzymatically active LF and EF toxins within the target cell (1, 2, 4). LF is a Zn²⁺ metalloprotease that has been shown to cleave six of the seven known human mitogen-activated protein kinase kinases (MAPKKs) in their proline-rich regulatory domain. As a result, MAPKKs are unable to bind, phosphorylate and subsequently activate their substrates, the downstream mitogen-activated protein kinases (MAPK) (1, 2, 4). EF is a Ca²⁺/calmodulin-dependent adenylate cyclase (AC) that converts ATP into cAMP at a rate 1,000-fold higher than the rate of the endogenous AC enzyme (1, 2, 4).

LF and EF Toxic Action is Drosophila melanogaster

Guichard et al. (3) assess whether LF and EF affect MAPKK pathways in Drosophila. They also monitor readouts of pathways whose signaling levels should be altered by modifying cAMP concentrations. They initiate their investigations by showing that, in an in vitro biochemical assay, LF detectably cleaves two of the fly MAPKKs: Hemipterous (Hep) and Licorne (Lic). A third MAPKK, Downstream-of-Raf (Dsor1), is surmised to be cleaved based on sequence homology but is not shown to be cleaved. This finding sets the stage for an in vivo analysis of the toxins.

Ectopic expression of LF is followed by monitoring Hep and Dsor1 activity in transgenic strains of Drosophila. Hep is known to function in the Jun N-terminal kinase (JNK) signaling pathway, which controls dorsal closure in the embryo (5) and proper formation of the adult thorax (6). Overexpression of LF in the embryo results in malformed embryos with defects in dorsal closure. LF overexpression in a specific region of the imaginal discs also impairs the proper formation of the thorax of the adult fly. Finally, concurrent ectopic expression of LF and Hep in the wing suppresses the phenotypes that are conferred by Hep ectopic expression alone. Thus, in every tissue tested so far, LF overexpression antagonizes endogenous and ectopically activated JNK signaling, indicating that the LF enzymatic activity is directed against the Hep kinase activity in vivo.

Dsor1 functions in epidermal growth factor receptor (EGFR) signaling (7), which regulates proliferation and vein formation in the wing (8). LF overexpression in the wing results in small elongated wings, similar to overexpression phenotypes of dominant negative forms of EGFR, indicating that EGFR signaling is inhibited. Overexpressed LF interacts genetically specifically with recessive mutations in genes that further reduce EGFR signaling, such as Dsor1, leading to a vein loss phenotype. LF suppresses the activation of the EGFR pathway when two components that lie upstream of Dsor1 were overexpressed. However, overexpression of a component that lies downstream of Dsor1 is not suppressed by LF. Therefore, LF is likely to inhibit Dsor1 activity in vivo.

EF overexpression in the wing affects the spacing between the veins. This patterning phenotype is reminiscent of mutations that attenuate another evolutionarily conserved pathway, the Hedgehog (Hh) pathway (9). A central player in the Hh pathway is PKA, whose inhibitory activity depends on the levels of the endogenous cAMP levels. Upon increase of cAMP, the inhibitory regulatory subunit of PKA (PKAr) dissociates from the catalytic one, and PKA activity is increased. Presumably, overexpression of EF elevates cAMP levels, leading to activation of PKA and thereby reducing Hh signaling. In accordance with this model, overexpression of EF suppresses the lethality and patterning defects that are conferred by overexpression of PKAr, providing indirect evidence that cAMP concentrations are increased when EF is overexpressed.

The Use of Toxins in Drosophila

Anthrax cannot infect Drosophila. Guichard et al. (3) report that there are no receptors onto which PA anthrax toxin can bind encoded in the Drosophila genome. However, it is clear from this study that Drosophila can still be used for studying the effects of the anthrax toxins. Drosophila has been restricted so far in infection studies by a small number of other pathogenic organisms, such as Vibrio cholerae, Staphylococcus aureus (10), and Pseudomonas aeruginosa (11), or fungi such as Aspergillus fumigatus (12) and Cryptococcus neoformans (13). In fact, most of them rely on infection of the host by inoculation or ingestion of living bacteria or fungi. However, directed expression of LF and EF toxins shows that a wide range of toxins and virulent factors can now be analyzed, including those that originate from organisms that cannot infect Drosophila.

In addition, directed expression of toxins in a specific spatiotemporal manner and controllable levels in combination with toxins’ action against specific target molecules provide greater experimental control, flexibility, and specificity in manipulations of toxins’ action in vivo. For example, the N-terminal GTPase-activating (GAP) domain of the exoS exotoxin has been ectopically expressed in the fly eye to confirm its interaction with Rho GTPases (14). Diphtheria toxin (15) and ricin (16) have been helpful tools for ablation of specific cell types in developmental studies, whereas tetanus toxin light chain has been used for blocking neurotransmission (17). Similarly, LF can be exploited for artificially blocking most MAPKK activity (or all, if it is verified that LF acts on Lic and MAPKK4 in vivo), whereas EF can be used for elevating cAMP levels and affect cAMP-dependent signaling. It would also be interesting to test whether EF can modulate neuronal function, because mutations in genes that control cAMP synthesis in Drosophila (rutabaga and dunce) are associated with defects in olfactory conditional learning and in posttetanic potentiation, a simple form of short-term synaptic plasticity, of the neuromuscular synaptic transmission (18).

Most importantly, Guichard et al.’s study (3) shows that LF and EF function on evolutionarily conserved pathways and provides the platform for second-site modifier genetic screens. Overexpression of toxins in tissues dispensable for viability of the fly, such as the eye or the wing, create a sensitized genetic background in which mutations that will suppress and/or enhance the toxins’ effect can be readily identified. This kind of genetic screen, in combination with biochemical approaches and microarrays, will possibly identify additional regulators and unknown in vivo targets of LF and EF activity as well as discover novel players in well known pathways, allowing us to uncover crosstalk points among the pathways that govern cellular physiology.

Furthermore, knowledge of LF and EF functions in Drosophila creates the opportunity of flies becoming a fast, high-throughput, and accurate in vivo model for testing known inhibitors (19) and discovering new drugs against anthrax toxins. The use of anthrax in bioterrorism and the pathology of the disease intensify the need of fast-tracking new and efficient therapeutic agents against the bacterial toxins. In the past, genetic analysis and pharmacological research have been combined in Drosophila successfully. For example, the efficacy of antifungal drugs against Aspergillus virulence has been tested (12), anticonvulsants have been used to dissect aspects of seizure behavior in fly models of epilepsy (20), and new peptides have been tested in fly models of Huntington’s disease (21).

In conclusion, Drosophila is a suitable multicellular model organism that combines complexity, genetic tractability, and a rich, constantly developing arsenal of techniques and reagents to investigate the actions of toxins and pathogens. Although every new finding must be extensively reexamined in vertebrate disease models, the fly will continue to help unravel mechanisms of disease.