Figure 1. (a) Representation of the adenovirus virion. The Ad virion is an icosahedron with protrusions, called fibre, attached to penton base at each of the 12 vertices. The capsid protein that forms the major component of the 20 facets is called hexon. A dozen or so additional proteins make up the capsid and core of the virion. Approximately 15% of the molecular mass of the particle comprises DNA packaged as a linear double-stranded molecule. (b) Organization of the viral genome (100 map units (mu)=36 kb). Promoters are indicated by square brackets. Transcription from the major late promoter at 16 mu generates a single long transcript that is spliced into late mRNAs as indicated. 1, 2, 3 and x, y, z represent leader RNAs attached to various late messages. Virus-associated (VA) RNAs are RNA polymerase III transcripts initiating around 29 mu. The mRNA for protein IVa2 is synthesized at intermediate times from a promoter at 16 mu.

Figure 2. Adenovirus DNA replication. (a) Initiation of Ad DNA replication is protein primed and can occur at either end of the viral DNA. Viral DNA replication requires cellular proteins called ORP-A, NF I and NFIII (or Oct-1) in addition to the viral E2B region-coded DNA polymerase (pol) and preterminal protein (pTP). A serine residue in pTP becomes covalently linked to a deoxycytidine monophosphate (dCMP) residue in a reaction catalysed by the virus-coded pol. The 3¢ hydroxyl group of dCMP then serves as a primer for DNA synthesis along one strand of the viral DNA. (b) After initiation of DNA replication at the ends of the viral DNA, synthesis proceeds by a strand-displacement mechanism (i,ii). The viral E2A region-coded DNA binding protein (DBP) is essential for viral DNA replication and binds to single-stranded viral DNA. A fully displaced single strand (iii) can form a hairpin structure (iv) in which the inverted terminal repeats (ITRs) anneal to form a duplex which is identical to the end of double-stranded viral DNA. The duplex portion of the hairpin structure can serve as a site for another initiation reaction (v) that can then complete the replication cycle to generate two duplex daughter molecules.

Figure 3. Construction of Ad vectors by site-specific recombination following cotransfection of 293 cells. The method depends on formation of an infectious viral DNA molecule by recombination between two noninfectious plasmids. The genomic plasmid and the shuttle plasmid are able to replicate in cotransfected cells because inverted terminal repeat (ITR) junctions can serve as origins of adenovirus DNA replication (Graham, 1984 ). However, neither DNA molecule is capable of generating infectious virions – the genomic plasmid because it lacks the packaging signal () and the shuttle plasmid because it does not encode any viral proteins. High-efficiency site-specific recombination between loxP or frt sites (indicated by an open arrowhead) is catalysed respectively by the bacteriophage P1 recombinase, Cre, or by the yeast 2 plasmid-encoded recombinase, FLP. In the example illustrated here (Ng and Graham, 2002 ) the recombinases are expressed from a cassette cloned into the genomic plasmid, but the enzymes can also be expressed from a cassette in the shuttle plasmid or by the cotransfected cells. In any case, the recombinase cassette does not appear in the final vector product.

Figure 4. The Cre/loxP system for generating fully deleted (FD) vectors. 293 cells expressing Cre (293Cre) are coinfected with the FD vector and a helper virus bearing a packaging signal flanked by loxP sites. Cre-mediated excision of the packaging signal () renders the helper virus genome unpackagable, but does not interfere with its ability to provide all of the necessary trans-acting factors for propagation of the FD vector. The titer of the FD vector is increased by serial passage in helper virus-infected 293Cre cells. The FD vector need contain only those Ad cis-acting elements required for DNA replication (inverted terminal repeats (ITRs)) and encapsidation (); the remainder of the genome consists of the desired transgene and non-Ad ‘stuffer’ sequences.