Anthrax toxin protective antigen protein (PA, 83 kDa) binds to receptors on the surface of mammalian cells, is cleaved by the cell surface protease furin, and then captures either of the two other toxin proteins, lethal factor (LF, 90 kDa) or edema factor (EF, 89 kDa). The PA-LF and PA-EF complexes enter cells by endocytosis and LF and EF translocate to the cytosol. EF is a calcium- and calmodulin-dependent adenylyl cyclase that causes large and unregulated increases in intracellular cAMP concentrations. LF is a metalloprotease that cleaves several mitogen-activated protein kinase kinases (MEKs) Entry of anthrax toxin into cells depends on two related cell surface receptors, tumor endothelium marker 8 (TEM8) and capillary morphogenesis gene product 2(CMG2). TEM8 was initially identified as a protein upregulated in colon cancers. CMG2 has substantial sequence similarity to this candidate tumor marker. The tissue distribution and the relative importance of the two toxin receptors in toxin action are not well understood. During 2010 we characterized the responsiveness of various types of animal cells to anthrax toxin using a previously described reporter protein. This fusion protein consists of LF residues 1-254 fused to beta-lactamase. Delivery of this fusion into the cytosol of cells, which is dependent on PA and PA receptor function, can be detected with a fluorescent beta-lactamase substrate, CCF2/AM. One attractive feature of this widely used reporter system is that the substrate cannot be cleaved by beta-lactamase until it has entered cells and been deacylated to CCF2. The change in fluorescence upon cleavage can be measured by flow cytometry. We used this system to examine the relative abilities of different primary immune cells to internalize the reporter protein. All the mouse cells examined could take up the fusion protein, but macrophages, dendritic cells, and B cells had a higher uptake capacity than T cells. These findings may have implications in the anti-inflammatory effects of the anthrax toxins during establishment of an infection. During 2010 we worked with collaborators at Catholic University and elsewhere in creation and testing of novel fusion proteins containing key domains of both LF and PA. Knowledge about the interactions of PA, LF, and the CMG2 receptor was used to design fusions that would efficiently induce neutralizing antibodies. For example, a fusion of LF residues 1-254 to PA (LFn-PA) was able to mimic a mixture of the proteins in inducing antibodies to both proteins. Furthermore, this protein successfully immunized rabbits and protected them against intranasal challenge with fully virulent B. anthracis. This vaccine candidate may have advantages over vaccines based on a single toxin protein. In other collaborative work during 2010, we studied the mechanism of action of the previously identified 1G3 mouse monoclonal antibody, a potent anti-PA neutralizing antibody. The 1G3 antibody was known to bind only to the PA oligmer, and not to uncleaved PA83, and to compete with LF for binding to the oligomer. Thus, it was considered probable that 1G3 binds to the same site on the PA oligomer as LF. Of equal interest, this antibody is the only neutralizing mouse anti-PA antibody reported to be dependent on engaging cell surface Fc receptors to cause neutralization. This surprising property is shared with the human antibody MDX-1303 that is being developed by a company for therapeutic use. Our new studies of 1G3, done in collaboration with electron microscopists, shows that this antibody induces formation of a PA63 supercomplex made of what appears to be two linked oligomers. This distortion of the native oligomer may destroy binding sites for LF and EF and thereby account for the ability of the antibody to neutralize the toxin. Furthermore, we find that, unlike previous reports, the antibody performs this reorganization of the PA oligomer and neutralizing function in an Fc-independent manner. This was the first report of an antibody causing severe perturbation of the PA63 oligomeric structure. Since the original report of Fc dependent toxin neutralization also found that polyclonal neutralizing antibody preparations also required Fc engagement, these findings with a mouse monoclonal antibody may help to explain these unique and unexpected findings.
Showing the most recent 10 out of 31 publications
