Vaccines and Therapeutics for Anthrax

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications & trials

Abstract

The LF protein is a zinc-dependent metalloprotease that cleaves all but one of the mitogen-activated protein kinase kinases (MEKs, or MKKs), as well as two other targets that we identified. The MKKs do not share strong amino acid sequences at the cleavage site, which led others to recognize that LF’s narrow specificity comes from binding of MKKs to LF through an LF ‘exosite” distant from its catalytic site, whereas binding of the N-terminal MKK region cleaved by LF appears weak. Several amino acid substitutions in the putative LF exosite made it more selective for single MKKs, a finding that suggested opportunities for dissecting the physiological roles of each MKK and for enhancing the specificity of LF as an antitumor agent. To exploit these opportunities, during FY2025 we sought to obtain cryo-EM structures of complexes of LF with several MKKs. Plasmids were created for expression of all seven MKKs and protein was produced for nearly all of them. Isothermal titration calorimetry (ITC) was used to show that LF bound to several of the MKKs with binding constants of about 5 micromolar. (All these studies use a catalytically dead LF.) Analysis by Size-Exclusion Chromatography linked to Multi-Angle Light Scattering (SEC-MALS) confirmed tight binding of LF to several of the MKKs. These findings justified submission of the protein mixture for cryo-EM analysis at the NIAID RTB core. The first attempt was successful in yielding a structure of MKK1 bound to LF. Recently, conditions for producing the complex have been improved and a structure at 2.5 angstroms was obtained. To facilitate further analysis of these complexes, variant proteins were designed and produced that allow facile cross-linking to capture and stabilize the complexes. In parallel, we are developing several biophysical approaches to quantitate the interaction of LF with the MKKs. A full understanding of how anthrax toxin damages host tissues requires that we know the toxin’s binding and uptake in various cells, tissues, and organs. This information will also be required if these toxin proteins are successfully carried forward as therapeutics for cancer. We previously developed a unique reagent for visualizing the uptake and action of anthrax toxin in mice. A fusion protein of LFn to Cre recombinase (LFn-NLS-Cre) was injected into mTmG mice (“Tomato mice”), which ubiquitously express a red fluorescent protein until Cre activity switches expression to a green fluorescent protein. PA + LFn-NLS-Cre allowed toxin action to be visualized at single-cell resolution by confocal microscopy. During the 2025FY, using mice expressing either CMG2 or TEM8 (and the mTmG transgene), we showed that functional CMG2 receptors support toxin uptake in the liver and heart, whereas TEM8 supports uptake in the kidney and spleen. This reagent was also used to demonstrate the specific delivery of LFn-NLS-Cre (and by implication, the native LF) to tumors cells by the PA variants, PA-U2 and PA-L1, which require activation by the tumor-associated proteases urokinase plasminogen activator and matrix metalloproteases, respectively. The LFn-NLS-Cre reagent has many potential in vivo applications since it can deliver Cre activity in a temporally and spatially controlled manner, replacing complex genetic systems for inducing Cre activity, expressing Cre from adenovirus, or breeding to Cre-expressing mice. Our development of anthrax toxin-based anti-tumor agents over the last 20 years has periodically attracted the attention of biotech entrepreneurs. In the latest and most promising example, one such entrepreneur founded a company in 2021 specifically to use the anthrax toxin system for intracellular delivery of protein drugs. We have been advising this company on the properties of the proteins and have supplied small amounts of protein for pre-clinical work. The company’s initial intent was to use LFn-Protein Kinase R (LFn-PKR, which binds double-stranded RNA) to deliver siRNA, but the company has recently chosen to focus on more potent candidates, such as the combination of the MKK1 specific LF W271A protein, delivered to cells by a PA redirected to HER2. We and our former Staff Scientist ShiHui Liu, now a faculty member at the University of Pittsburgh, have provided extensive data, materials, and advice to the company to facilitate those efforts. The company has been successful in raising public and private funds to support the preclinical work to date. In initial studies performed by a contract organization, the toxin greatly extended survival in a mouse orthotopic glioblastoma model. We will continue to supply advice and materials to advance this promising technology, while continuing basic research to increase the efficacy of these agents. Our group has developed expertise in analysis and use of unique monoclonal antibodies directed to nucleic acids. We previously popularized the S9.6 antibody specific for DNA/RNA hybrids, which became an essential reagent for analysis of R-loops, structures that occur naturally at the site of mRNA synthesis initiation, but also accumulate in several pathological processes. We determined the amino acid sequence, produced recombinant IgG and Fab, worked with collaborators to characterize its biophysical properties, and determined the crystal structure of an S9.6 complexed to a short DNA/RNA hybrid. Subsequently, we demonstrated the utility of S9.6 for detection of pathogen nucleic acids. Based on our success with S9.6, during teh 2025FY we extended this approach to produce and characterize the J2 monoclonal antibody that recognizes double-stranded RNA (dsRNA). This antibody is widely used to visualize cells infected with RNA viruses, as well as to identify undesirable dsRNA in RNA-based therapeutics. As was done for S9.6, we obtained crystals of the antibody-dsRNA complex, solved the structure, and have characterized the nature and requirements for binding. This work will be published during this year. During this 2025FY period, we continued studies on delivery of peptides that are expected to bind and inhibit myc, a key transcriptional activator. Myc normally binds via a coiled-coil interaction with a similar protein named Max, and the complex activates expression of many genes. We prepared LFn fusions to peptides that have been reported to disrupt the Myc-Max complex. However, it became evident that coiled-coil peptides will dimerize and cannot be translocated through the PA channel. Other peptides not expected to dimerize were made, but none were found that affected cells. Work on these myc inhibitor peptides will be ceased in favor or other types of work.