Further studies on the replication of adenovirus type 2 DNA

Abstract

Monolayer cultures of KB cells infected with adenovirus type 2 (Ad2) were subjected to short pulses of tritiated thymidine, at the time of maximal viral DNA synthesis. Nascent viral DNA was selectively extracted from the cells, purified, and fractionated into partly single-stranded (or replicating) molecules and double-stranded (or completed) molecules. Completed molecules were cleaved by the restriction endonucleases from either Escherichia coli (EcoRI), Haemophilus influenzae (Hind III), Haemophilus parainfluenzae (Hpa I), or by the latter two enzymes in succession, and the resulting fragments were separated by electrophoresis through agarose–acrylamide slab gels. In the case of replicating molecules, the double-stranded fragments generated after treatment with the same enzymes were first separated from the partly single-stranded fragments by chromatography on benzoylated–naphthoylated DEAE (BND) cellulose before being subjected to electrophoresis. The relative yields of tritium of the various fragments resolved by the gels were then determined. As already described in several reports, the analysis of the completed molecules obtained after the shorter pulses revealed a strong preferential labeling of the fragments derived from the molecular ends, as expected if both of these ends contained termini for replication. The analysis of double-stranded fragments from replicating molecules revealed a predominant labeling of the central portion of the genome which could be reconciled with initiation occurring either centrally or at both molecular ends. In addition, two gradients of labeling, ascending toward each molecular end and encompassing the outer quarters of the molecule, were noted which slowly receded when the pulse length was increased. The latter finding, which is in agreement with our earlier results, could reflect an overrepresentation of molecules in the late stages of replication in the pool of nascent DNA, such as one would expect if replication slows down when it nears the termination site(s). This interpretation is consistent with the kinetics of labeling of the various portions of the genome observed in the pool of completed molecules. Because of this complication, however, the distribution of label in replicating molecules cannot be used to ascertain unambiguously the location of the origin(s) of replication on the Ad2 DNA molecule.