Abstract
Tumor necrosis factor (TNF) is a highly effective cytokine with considerable anti tumor activity not only because of direct tumoricidal effects but also due to its strong inflammatory and anti-angiogenic properties. However, while impressive clinical responses were obtained by locoregional application, e.g. in isolated limb perfusion, systemic application of TNF causes intolerable side effects. Therefore, multiple attempts have aimed for targeted delivery of TNF to maximize therapeutic efficiency and sparing normal tissue from detrimental effects, e.g. by generating antibody-derived fusion proteins with human TNF replacing the IgG1 CH2/CH3 Fc domain. However, a highly selective and thus clinically applicable drug for stimulation of TNF-receptors (TNF-R) at the tumor site with minimal unspecific TNF-R stimulation is still not available. Here we report the generation of a bispecific EpCAM X TNFR1 Fab2 antibody (bsFab2). The Fab fragments of an agonistic antibody to mouse TNFR1 were chemically hybridized to anti-EpCAM Fab fragment which served as a targeting molecule on EpCAM transfected mouse melanoma cells. Bound to such cells the bsFab2 exerted strong TNF activity as measured by killing of TNF sensitive WEHI-164 cells at concentrations as low as 0.1ng/ml. In the presence of untransfected tumor cells weak killing was detectable only at antibody concentrations as high as 1000ng/ml. Even more, killing of WEHI-164 cells in the absence of tumor target cells was undetectable. This demonstrates that TNF activity is induced by the bsFab2 in a target cell restricted manner. Target cell restriction was also confirmed when caspase activation rather than cell death was measured. We also demonstrate that the bsFab2 is able to induce target antigen restricted cellular inflammatory responses like monocyte/macrophage cytokine production. Next we investigated the activity of the bsFab2 in the B16 melanoma mouse model with syngeneic, immunocompetent mice. C57Bl/6 mice were inoculated 2500 EpCAM transfected B16 melanoma cells and treated with 6x20μg of the bsFab2. While all of six treated mice survived without developing a tumor, in the control group five of six mice died within 10 weeks. No significant side effects were observed. Our data demonstrate that the TNF activity of the EpCAM X TNFR1-bsFab2 is strictly depending on the presence of EpCAM positive tumor cells, stimulates cellular inflammatory responses in immediate vicinity of target antigen expressing tumors and provide the rationale for further studies evaluating the clinical potential of selective TNFR activation with bispecific antibodies in humans.
Author notes
Corresponding author