Abstract 5703: Efficient targeted delivery of protein toxins using self-manufacturing nanoparticles (minicells) derived from bacteria

Abstract Significant progress has been made with targeted delivery of small molecule cancer therapeutics. The key to clinical success has been the use of extremely potent drugs conjugated to targeting antibodies. There are also classes of very potent and promising therapeutic macromolecules, including protein toxins and RNAi. However, there are several challenges unique to macromolecular payloads that cannot be adequately addressed when using antibodies alone for targeting. These include manufacturing complexity, blood-based catabolism, low payload capacity, poor loading efficiency, and inefficient endosomal escape. Several of these issues have also impeded clinical success with antibody-decorated liposomes. We are addressing these issues by using self-manufacturing nanoparticles, or minicells, derived from E. coli. Minicells are spherical, non-viable ∼400 nm nanoparticles containing all bacterial components, except the chromosome. Minicells bud off the end of rod-shaped bacteria as a result of unequal cell division after induced overexpression of a gene involved in septum formation. We have developed procedures for purification of minicells, including removal of parental bacteria. Our minicell-producing bacterial strains are genetically engineered to prevent growth outside the laboratory. In addition, we have introduced a lipopolysaccharide biosynthesis mutation, which significantly attenuates endotoxin-related inflammatory responses. We have produced minicells with surface expression of the beta1 integrin-targeting protein invasin. The minicells also contain a bacterially-expressed cholesterol-dependent endosomal membrane disrupting protein, such as listeriolysin O (LLO) or perfringolysin O (PFO), and a potent protein toxin, such as the catalytic domain of diphtheria toxin (DT). Invasin binds with high affinity to certain β1-containing integrins, including α5β1, which is overexpressed in tumor vasculature and several solid tumors. Surface expression of invasin on minicells led to binding and endocytosis into human endothelial cells (HUVECs) and tumor cells lines that express α5β1 integrin. Despite significant internalization, invasin-expressing minicells were not toxic to mammalian cells at doses of ≥10,000 minicells per cell. Inclusion of LLO or PFO and DT led to targeted killing of HUVECs and tumor cells (melanoma, colorectal, pancreatic, lung and hepatocellular carcinoma) with IC50s in the range of 50-500 minicells per mammalian cell. In vivo evaluation of minicell toxicity, immunogenicity, tissue distribution and efficacy is in progress. In summary, our minicell-based targeted delivery platform for macromolecular therapeutics reduces manufacturing time and complexity, significantly increases payload capacity, protects payload from blood-based catabolism, and provides a critically enabling endosomal escape mechanism. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 5703. doi:1538-7445.AM2012-5703