Characterization of an Adenovirus Gene Transfer Vector Containing an E4 Deletion

A new adenovirus-based vector (Ad2/CFTR-1) has been constructed in which the cDNA encoding the cystic fibrosis transmembrane conductance regulator (CFTR), the cystic fibrosis (CF) gene product, replaces the early region 1 coding sequences, E1a and E1b. The virus retains the E3 region. Ad2/CFTR-1 and a related construct encoding beta-galactosidase replicate in human 293 cells which provide E1 gene functions in trans. Replication of these recombinant viruses was not detected in a variety of other cells, although very limited viral DNA synthesis and transcription from the E4 and L5 regions could be measured. These E1-deletion vectors were also deficient in cellular transformation, shut-off of host cell protein synthesis, and production of cytopathic effects, even at high multiplicities of infection. Ad2/CFTR-1 produced CFTR protein in a variety of cells including airway epithelia from CF patients. Expression of functional CFTR protein in a CF airway epithelial monolayer was detected by correction of the Cl- transport defect characteristic of CF. Surprisingly low multiplicities of infection (0.1 moi) were sufficient to generate CFTR Cl- current across a CF epithelial monolayer in vitro. These data, together with the lack of obvious toxicity, suggest that Ad2/CFTR-1 should be suitable for CF gene therapy.

Binding of the mammalian transcription factor E2F to the adenovirus E2a early promoter is modulated through interaction with the viral E4-6/7 protein. E4-6/7 induces the cooperative and stable binding of E2F in vitro to two correctly spaced and inverted E2F binding sites in the E2a promoter (E2F induction) by physical interaction in the protein-DNA complex. The E2a promoter is transactivated in vivo by the E4-6/7 product. The C-terminal 70 amino acids of E4-6/7 are necessary and sufficient for induction of E2F binding and for transactivation. To assess the mechanism(s) of E2a transactivation and the induction of cooperative E2F binding by the E4-6/7 protein, we have analyzed a series of point mutants in the functional C-terminal domain of E4-6/7. Two distinct segments of E4-6/7 are required for interaction with E2F. Additionally, and E4-6/7 mutant with a phenylalanine-to-proline substitution at amino acid 125 (F-125-P) efficiently interacts with E2F but does not induce E2F binding to the E2a promoter and is defective for transactivation. Induction of E2F stable complex formation at the E2a promoter by the F-125-P mutant protein is restored by divalent E4-6/7-specific monoclonal antibodies, but not a monovalent Fab fragment, or by appending a heterologous dimerization domain to the N terminus of the mutant protein. These and other data support the involvement of E4-6/7 dimerization in the induction of cooperative and stable E2F binding and transactivation of the E2a promoter. We present evidence that at least two cellular components are involved in E2F DNA binding activity and that both are required for E2F induction by the E4-6/7 product. The recently cloned E2F-related activities E2F-1 and DP-1 individually bind to an E2F binding site weakly, but when combined generate an activity that is indistinguishable from endogenous cellular E2F. Recombinant E2F-1, DP-1, and E4-6/7 are sufficient to form the induced E2F complex at the E2a promoter.