Chemosensitization by erythropoietin through inhibition of the NF-kappaB rescue pathway.

Two cell lines that exemplify erythropoietin (EPO) receptor-positive tumors, human renal carcinoma cell lines RCC and the myelomonocytic leukemia cell line U937, were investigated for the apoptosis-modulatory potential of EPO. Cells cultured in the presence of EPO exhibited an elevated apoptotic response to cancer chemotherapeutic agents such as daunorubicin (Dauno) and vinblastine (VBL). Chemosensitization by EPO did not involve an increase in p53 activation, yet correlated with enhanced Bax/Bak-dependent mitochondrial membrane perturbation and caspase maturation. In vitro monotherapy with Dauno or VBL induced the degradation of IkappaBalpha, provoked the translocation of NF-kappaB p65/50 to the nucleus and stimulated the expression of an NF-kappaB-activatable reporter gene. All these signs of NF-kappaB activation were perturbed in the presence of EPO. Inhibition of JAK2, one of the receptor-proximal elements of EPO-mediated signal transduction, greatly diminished the EPO-mediated chemosensitization and NF-kappaB inhibition. EPO lost its death-facilitating effects in the presence of an NF-kappaB inhibitor, underscoring the cause-effect relationship between EPO-mediated chemosensitization and NF-kappaB inhibition. Altogether, these results suggest that, at least in a specific subset of tumors, EPO receptor agonists can prevent activation of the NF-kappaB pathway, thereby enhancing the propensity of EPO receptor-positive tumor cells to undergo apoptosis.

In vivo-targeted gene delivery using antibody-based nonviral vector.

Tissue-specific gene transfer remains one of the main challenges to deliver genes into designated and/or disseminated cells. We have previously shown successful gene transfer with a nonviral gene delivery system based on the simple chemical conjugation of plasmid DNA with antibody. However, this approach was hampered by low efficiency due to the poor translocation rate of DNA to the nucleus. To improve this approach, we have modified our vector by introducing noncovalent binding between the antibody and DNA, allowing the possibility to introduce different important molecules. The noncovalent association was achieved with neutravidin and biotinylated components: (1) biotinylated antibodies; (2) a biotinylated hemagglutinin fusogenic peptide of influenza virus to favor endosomal escape; and (3) biotinylated histone H1 to compact, protect, and associate DNA to the complex. We report here that this delivery system can be internalized by tumor cells targeted by a specific monoclonal antibody, permits the protection of the transfected DNA, and allows its subsequent transfer into the nucleus after escape from the endosomal compartment. We also demonstrate that, in vitro, gene transfer with this vector showed much higher reporter activity in cells (15 vs. 0.5%) and a stronger production of murine interleukin 2 as compared with our previous vector. In vivo, a single intravenous injection of the vector containing an antibody directed to the G250 renal cell carcinoma-associated antigen led to beta-galactosidase expression in engrafted tumor bearing G250 but not in G250-negative tumor or in other tissues. Altogether, these results indicate that our antibody-based vector is suitable to promote gene delivery in vitro and in vivo in tumor cells.