Studies continue to analyze and engineer the anthrax toxin proteins, and other proteins of interest, for use in therapeutic agents and as tools for cell biology. While we successfully produce many proteins in our improved Bacillus anthracis expression system, some heterologous proteins do not work well in this system. To obtain the necessary amounts of protein, we therefore employ established E. coli expression systems, sometimes also having genes chemically synthesized to have optimized codon bias. A key use of our knowledge of anthrax toxin structure is in design of tumor-targeting agents. In prior work we generated toxin variants that depend on activation by two cancer cell surface proteases, matrix metalloproteases (MMP) and urokinase plasminogen activator (uPA). These agents contain two separate proteins, each activated by one of the two proteases. The design requires that the two mutated and inactive proteins must combine to achieve activity. We have designated these as intercomplementing protective antigen drugs, or IC-PA. By requiring activation by two proteases, activity toward normal tissues is greatly decreased, and therapeutic indices are increased. During 2018, we focused on analysis of the IC3-PA construct. This uses three features that separately contribute to tumor specificity. The L1 and U2 constructs include sequences making the proteins dependent on the MMP and uPA proteases, respectively. The R200A and I207R mutations impose the requirement that the two proteins combine (complement) to make an active protein-conducting channel. Finally, the I656Q mutation makes the proteins highly selective for CMG2, one of the two toxin receptors, and the one which mouse genetic studies shows is key to tumor targeting. Thus, the IC3-PA agent combines the proteins designated PA-L1-I207R-I656Q and PA-U2-R200A-I656Q. A key advance during 2018 was the incorporation of an improved LF variant as the effector component. Work by others showed that the specificity of LF for proteolytic cleavage of its substrates is controlled by an exosite, a region of LF far from its catalytic center. One key exosite residue is W271. Importantly, it was found that the W271A mutant retains specificity for MEKs 1 and 2 but has greatly decreased activity toward other MEKs. Because the antitumor activity of our agents appears to depend on cleavage of MEKs 1 and 2, we considered it likely that use of W271A would preserve anti-tumor activity while decreasing toxicity to normal tissues. This was confirmed by showing that IC3-PA + LF W271A has an increased therapeutic index of >15 in treatment of mouse tumors. A key requirement for extended use of these protein toxins is to prevent development of neutralizing antibodies. We have continued to explore the combined use of pentostatin and cyclophosphamide to eliminate B and T cells while maintaining granulocytes. In addition, we have begun studies using a mouse monoclonal antibody that specifically eliminates mouse B cells, and have confirmed that it is effective. In our continuing outreach to potential partners, both academic and commercial, we have devoted considerable effort to emphasizing the unique nature of our agent. Our work has shown our agent to target the host-derived neovasculature that tumors induce to obtain adequate nutrition. The consequence is that our agent is expected to be active against all types of solid tumors. Furthermore, the genetic makeup of the tumor will be irrelevant, and it is unlikely that resistance could develop, as so often happens when drugs are targeted to mutable signal transduction pathways in the tumor cells themselves. We have continued to work actively with the NIAID technology transfer staff and others to explore ways to move our agent into clinical use.
Showing the most recent 10 out of 31 publications
