Hepadnaviruses are small DNA viruses that replicate their genomes by reverse transcription of an RNA intermediate 
. The viral genomes are in a relaxed circular conformation that is stabilized by cohesive overlaps created by the juxtaposition of the 5′ ends of the two DNA strands 
. Hepadnaviruses are enveloped viruses that primarily infect hepatocytes by a pH-independent pathway that is still incompletely understood. Following uncoating of the viral envelope, core particles are released into the cytoplasm and eventually enter nuclear pores and perhaps the nucleus, disassemble and release RC DNA 
. Within a few hours after an infection, CCC DNA derived from RC DNA in virions can be detected in nuclei of infected hepatocytes 
. During early stages of infection, additional CCC DNA is produced from newly synthesized RC DNA present in cytoplasmic core particles by an intracellular amplification pathway 
. As a consequence of this mechanism, infected cells harbor between 5–30 copies of CCC DNA and remain persistently infected even in the presence of antiviral therapies that inhibit the RT (i.e. ref. 
CCC DNA synthesis requires the removal of a 18 nucleotide-long RNA primer from the 5′ end of plus strand DNA and the reverse transcriptase from the 5′ end of minus strand DNA 
. In addition, one or both ends of minus strand DNA have to be trimmed to remove all or some of the sequences in the 9 nucleotide-long terminal redundant r5 and r3 segments. The final step in CCC DNA synthesis is the ligation of the 5′ and 3′ ends of the two DNA strands. (). The exact sequence of events and the enzymatic activities leading to CCC DNA synthesis have not yet been described.
Two models can explain the formation of CCC DNA (). The first (model 1) predicts that the reverse transcriptase performs a cleavage-ligation reaction to synthesize the minus strand of CCC DNA, which then could serve as a template for the repair of plus strand DNA. For this reaction, the RT would have to hydrolyze the phosphodiester bond at the 5′ end of the 3′r region and use the released energy for a transesterification reaction resulting in the dissociation of the RT from the 5′ end and the ligation of the two ends of minus strand DNA. A similar biochemical reaction is carried out by the A protein of bacteriophage ΦX174 during rolling circle DNA replication 
. It has been suggested that an RC DNA form lacking RT at the 5′ end of minus strand DNA might be a precursor for CCC DNA formation essentially as predicted by model 1 
. The second model (model 2) predicts that a cellular DNA endonuclease cleaves minus strand DNA downstream of the 5′ end and that a cellular DNA polymerase extends the 3′ end using plus strand DNA as a template followed by the ligation of the free ends. Thus, the second model would occur independently of a viral enzymatic activity and as a consequence would depend entirely on cellular DNA repair enzymes.
The low efficiency of CCC DNA formation in tissue culture cells and the lack of permissive in vitro systems to recapitulate the conversion of RC to CCC DNA, hampered efforts to investigate the first step critical in hepadnaviral DNA synthesis. As described in this report, we have exploited information about the priming reaction required for reverse transcription of HBV genomes for a genetic analysis of CCC DNA formation that permitted a distinction between the two models proposed above for CCC DNA synthesis.