Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Dr. Frank L. Graham


This thesis describes studies designed to develop and utilize human adenovirus type 5 (Ad5) as a cloning and expression vector in mammalian cells. The experimental approach taken in this work was first to determine the limits on the amount of DNA which could be packaged into virion capsid. To this end I have used mutants of Ad5 derived from insertion of a bacterial transposable element (Tn5) into the left end (the transforming region) of the viral genome. Attempts to rescue various sized derivatives of these Tn5 insertion mutants into full length infectious Ad5 DNA indicated that not more than about 2 kb of insert DNA could be rescued in this way. To extend this limit, a helper-independent Ad5 cloning vector has been constructed in which most of early region 3 (E3), from map co-ordinates 78.5 to 84.7, and essentially all of early region 1 (E1), from 1.0 - 10.6, have been deleted. E3 is non-essential for adenovirus replication in cultured cells and E1 is non-essential when the virus is propagated in 293 cells which constitutively express the E1 gene products. The resulting new virus, d1E1,3 was about 5.5 kb shorter than wt Ad5 and therefore should be able to accept up to 7.5 kb in foreign DNA. To test the usefulness of this vector, the Herpes Simplex Virus type 1 (HSV.1) thymidine kinase gene (tk) along with its regulatory sequences was inserted into the unique Xbai site of d1E1,3 (at map position 78.5/84.7). The resulting recombinant virus, Adtk, expressed the HSV tk at a low level (as compared to levels induced by HSV.1) in infected cells; however, tk expression was markedly enhanced when Adtk infected cells were super-infected with a tk mutant of HSV. Furthermore, the Adtk virus efficiently transformed tk⁻ mouse cells (line K4 and its progenitor the LTA) to the tk⁺ phenotype. At a low efficiency, it was also possible to transform tk⁻ human cells (line 143), and tk⁺ transformants of both mouse and human origin have been established as permanent lines.

Lastly, during construction of d1E1,3, we isolated a variant, d1E1,3-1, which had a direct repeat of viral DNA terminal sequences attached to the left end of the genome. Analysis of this variant with restriction enzymes and by hybridization of Southern blots with specific probes indicated that the extra terminal segment contained the left 2.6%(920 bp) of Ad5 joined to 352 bp of pBR322 which in turn was linked to the left end (minus 21 bp) of d1E1,3. During replication of d1E1,3-1 the extra terminal segment was found to transfer to the right end of the genome resulting in a second variant, d1E1,3-2, having duplicated terminal sequences at both ends of the viral genome. D1E1,3-2, was shown to revert back to d1E1,3-1 at high frequency. Although evidence was obtained indicating that the extra segment could be lost from the left end at low frequency, spontaneous mutants which had lost direct repeats from both ends were never isolated. It was, however, possible to remove the extra terminal repeat of d1E1,3-1 by cleavage with a restriction enzyme and to isolate d1E1,3 containing wt termini. The rearrangements occurring during replication of d1E1,3-1 and d1E1,3-2 may be the consequences of the mode of replication of Ad5 DNA.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Biology Commons