Antibodies offer a means to harness the extreme potency of bacterial protein toxins-capable of killing both healthy and transformed cells at picomolar concentrations-though no methods are currently available to link these toxins to high-affinity, bivalent mAbs. Site-specific attachment is critical in this application, as both proteins contain regions that cannot be perturbed in order to maintain tumor targeting and cytotoxicity. In principle, this could be accomplished by producing a mAb-toxin as a recombinant fusion protein. However, this is not possible with current technology, as the eukaryotic cells required to produce mAbs are killed by expression of protein toxins. We have characterized two reactions catalyzed by the bacterial enzyme sortase A (SrtA) that allow the mAb and the toxin to be produced separately by effective, proven methods, then site- specifically joined in vitro. We believe that this will provide a general mechanism to conjugate protein toxins such as diphtheria toxin and pseudomonas exotoxin A to mAbs independent of their target antigen, providing a means to generate antibody-toxin conjugates against a variety of tumor antigens for which mAbs are already available.
The mechanism of action of protein toxins offers a unique therapeutic modality for the treatment of cancer. Both diphtheria toxin and pseudomonas exotoxin A inactivate protein synthesis in any eukaryotic cell that internalizes the toxin. Because this method of cell killing is orthogonal to the majority of anti-cancer therapeutics, including existing antibody drug conjugates, it offers a complementary treatment strategy that may synergize with existing therapeutics. However, application of protein toxins has been limited, primarily because most immunotoxins in preclinical development are genetic-level fusions to cytokines or monovalent scFv antibody fragments. Though they can be produced in prokaryotic cells, cytokine- and scFv-linked toxins are hampered by their dramatically lower binding strength and reduced target selectivity compared to full- length, bivalent mAbs. Our enzyme-mediated methods for conjugating toxins to full-length mAbs provide mechanisms to use high-affinity tumor targeting agents that will mitigate side effects caused by the non- specific exposure of healthy, toxin-sensitive tissues to the conjugates.
