| Immunotoxins are a class of targeted therapeutics designed to treat cancer. They contain two parts: an antibody-based targeting moiety and a toxin. These toxins typically are of plant or bacterial origin. Although the antibody targets the toxin to malignant cells, clinical trials of immunotoxins revealed their utility was still limited by toxicity mediated primarily by non-specific effects of the toxins. We therefore sought to generate an immunotoxin with lowered non-specific toxicity and immunogenicity by substituting human ribonuclease (hRNase) for the toxin moiety. HRNase is a small stable protein that efficiently degrades RNA, leading to cell death. As a component in a targeted therapeutic, its major disadvantage is that it is potently inhibited by the cytosolic protein ribonuclease inhibitor (RI). To circumvent this, I generated a series of RI-resistant hRNase mutants. Two such mutants, hRNaseRDD and K7A hRNaseRDD, were genetically fused to an anti-CD7 single chain Fv and tested for their specific cytotoxicity in vitro. The naturally RI-resistant amphibian RNase, Onconase, was also fused to the single chain Fv to serve as a positive control. Unexpectedly, none of the constructs was significantly cytotoxic. Biochemical and functional assays suggested the fusion proteins were not being efficiently routed to the cytosol, and that the reduced affinity of the mutants for RI was insufficient to yield potent immunotoxins. This was true for both hRNase fusion proteins and their disulfide-linked counterparts. In contrast, Onconase was far more potent as a chemically conjugated immunotoxin than as a fusion protein. Finally, I describe a novel, whole mouse sectioning technique for the detection of minimal residual disease. Leukemic cells in mice treated with the CD7-specific immunotoxin DA7 were detected at the single cell level and found to reside mainly in the central nervous system post treatment. |
