structural proteins hexon and fiber and low levels of the E2A gene product, 
indicating that they are capable of supporting the full life cycle of Ad5. Cells 
harboring El deleted Ad5 expressed little, if any of the structural proteins hexon 
and fiber; however, the E2a gene was expressed at high levels in a subset of 
transgene containing cells. This suggests that the recombinant virus is prevented 
from transitioning into the late phase of transcription in the absence of Ela and 
Elb. However, a subset of cells are capable of activating transcription from the E2a 
promoter independent of Ela and Elb. The implications of this observation for the 
treatment of patients are unclear. The only significant concern we have is that the 
recipient may generate an immune response to the 72 kd product of E2a. Western 
blot analysis of sera from recipient animals (ferrets and primates) has failed to 
identify antibodies to the 72 kd protein, suggesting that this may not be a problem. 
IV.B.5.C. Recovery of El deleted adenovirus from xenografts 
We have also used the xenograft model to determine if a genetically 
reconstituted human airway is capable of supporting ongoing production of 
recombinant adenovirus. Xenografts were sequentially irrigated for a three week 
period after exposure to adenovirus, and the effluents were analyzed for wild type 
and recombinant virus. Wild type virus was never detected in the effluents and the 
concentration of recombinant virus dropped to undetectable levels (< l(r/ml) 
within 14 days of initial infection. 
We have since expanded these experiments to include more grafts and 
to increase the sensitivity of the virus assay to one particle per ml of effluent. A 
similar pattern emerged from these additional experiments in that wild type virus 
was never detected in the effluents and the amount of recombinant virus dropped at 
least four logs within one week of infection. However, in some of the grafts low 
concentrations of recombinant virus (10-l(r/ml) were detected in effluents during 
the second phase of the experiment (7-24 days). Interestingly, there was a direct 
relationship between the efficiency of stable genetic reconstitution and recovery of 
virus in that the highest quantities of virus (l(r particles/ml of effluent) were 
recovered from grafts that were most efficiently reconstituted with transgene (5-20% 
of cells), whereas virus that diminished to undetectable levels were from the 
effluents of grafts that were poorly reconstituted with transgene (<0.01% of cells). 
Mechanisms responsible for recovery of virus in this model are a matter of 
speculation. It is possible that the virus recovered in the effluents represents 
residual virus from the initial infusion. An alternative explanation is that the 
genetically reconstituted xenografts are supporting low levels of virus production. 
Replication of Ela deleted viruses has been described in vitro. One potential 
mechanism to account for the presence of virus in effluents is that an occasional cell 
in the xenograft overcomes the block in adenoviral replication, leading to its death 
and the production of some recombinant virus. The life cycle of group C viruses 
such as Ad5 in the context of full El expression is extremely efficient and results in 
the production of 10,000 virions per infected cells. Progression of the full Ad lytic 
cycle in only one to two cells per xenograft per week could account for the steady 
state level of recombinant virus detected in the effluents (i.e., 100 to 1,000 
virions/irrigation). Another potential mechanism is that the virus replicates at low 
levels in a large population of infected cells and low quantities of virus are released 
Recombinant DNA Research, Volume 17 
