4.4.5 Literature: Evidence against viral rescue or recombination: 
Can human papilloma viruses rescue a bovine, incomplete virus? 
In principle the possibility cannot be denied. There is however extensive evidence that such an 
event is exceedingly unlikely and maybe not possible in vivo. In the following the evidence for this 
statement is cited: 
a) complete virus particles have only been observed in stratified squamous epithelial cells, in vivo. 
Other cell types have never been seen to harbor virus particles or produce virus in vitro (see above). 
For virus particles to form, LI and L2 must be expressed. This occurs only in terminally differentiated 
epithelial cells. Small cell lung carcinoma cells are neuroendocrine cells. They are not likely to harbor 
virus or to be able to produce virus, even if they contained the complete BPV1 or a recombined 
genome. 
b) Papilloma viruses are extremely species specific. Never has virus production been observed in 
a heterologous host. BPV1 can only be produced from bovine warts. Even though BPV1 can cause 
tumors in hamsters (non epithelial tissues), hamsters will not produce BPV1 particles. Many 
laboratories have attempted to produce papilloma virus particles in heterologous hosts without success. 
It is therefore almost inconceivable that BPV1 could be produced in human host. Studies of 
slaughterhouse butchers’ warts have not revealed any unusual distribution of viral subtypes or any 
evidence for BPV-HPV recombination, even though the incidence of warts in one slaughterhouse, using 
less automated techniques, was increased (97). 
c) Many attempts have been made at pseudotyping papillomaviruses; again all attempts were without 
success. Even in papilloma viruses within the same species recombination has not been observed. It 
is not uncommon for humans and other species to harbor more than one papilloma subtype. 
In no case has recombination between subtypes been observed. Reports exist of several types of 
HPV virus within the same lesion; however recombination could not be documented. Finally, in 
Epidermodysplasia verruciformis many human papilloma virus subtypes have been observed in the same 
individuals. Again no single event of recombination could be documented (98, 99). Pseudotypes 
(generated by recombination) are unknown in the field of papilloma virus despite extensive search. 
Thus a recombination between bovine and human virus appears exceedingly unlikely. 
d) The pBMG-Neo vector containing IL2 contains 14.8kb DNA. Even though the packaging 
requirements for papilloma viruses are not well understood, the structure of the assembled LI capsid 
protein (ref 96) would appear to make packaging requirements relatively stringent. It is unlikely that 
the 14.8 kp BPV1 plasmid plus recombined LI and L2 (for the hypothetical argument of recombination) 
could meet the packaging requirements, having twice the genome size of the natural virus; the genome 
size of the complete BPV1 is 8kb. 
Given the constraints of host specificity and of virus production in vitro and the lack of any 
observed and documented recombination in vivo and in vitro , it is not practical to devise screening 
assays for events that have never been observed despite intensive search. This is tantamount to saying 
that the risk of virus production is minimal (or non existent) and cannot be determined with current 
assays. 
What then is the best way of guarding against this minimal remaining risk? 
We propose to screen all cells to be transfected by Southern blotting for human papilloma viruses. 
This will eliminate the majority of occurrences of HPV infection and further minimize the minimal risk 
of bovine-human PV recombination. 
[256] 
Recombinant DNA Research, Volume 18 
