18 
It is thought deletion of the promotor and enhancer sequences within the U3 
region of the 3' LTR should eliminate the ability of a retrovirus provirus, 
even if it did integrate next to a cellular proto-oncogene, to activate that 
gene . ] 
Dr. Anderson said bone marrcw cells rendered malignant by gene therapy vectors 
could probably be recognized in an actual clinical protocol since the transformed 
cells would overgrew in tissue culture. 
Dr. Gottesman asked whether any data exist on gene activation of cells in tissue 
culture by retroviral insertion. Dr. Mulligan said the frequency of activation 
of proto-oncogenes by retroviruses in tissue culture is probably approximately 
10 - ' in tissue culture systems. 
Dr. Gottesman asked whether there is any evidence a human gene therapy vector 
might integrate at a speci'^'c site. 
The working group said no current data demonstrate site specific integration 
of vectors in mammals. Horrologous site-specific integration occurs 
at a very low level, when it occurs at all, in mammals. Retroviruses insert 
at random sites in the chromosome since retroviral integration does not depend 
on homology. 
Dr. Gottesman asked whether deletion mutation would result with a higher 
frequency upon integration than would insertional activation. 
Dr. Anderson said deletion mutation is not observed on tissue culture plates 
with 107 cells per plate. The frequency of deletion mutation may be below 
this detection limit. Several investigators have looked for deletion mutation 
and have not been able to demonstrate it. 
Dr. Landy asked whether evidence suggests tumor formation can result fran gene 
inactivation. Dr. Mulligan said human retinoblastoma is a genetic disease 
in which it is believed that the inactivation of both alleles of a gene play 
some role in tumor formation. Deletion of the alleles of a gene which may be 
an anti-oncogene has been observed in some of these tumors. 
[Rapporteur's Note: Inactivation of genes by provirus integration has been 
observed in both somatic cells in culture as well as in germ cells _in vivo . 
For example, in sane in vitro experiments, it was shown that the v -src gene of 
the Rous sarcoma virus tranformed rat cell line B31 could be inactivated by 
insertion of a MoMLV provirus insertion. In addition, infection of a pre- 
implantation mouse embryo has led to insertion of a MoMULV provirus into the 
collagen gene. In the case of the mutated B31 cell line, reconstitution of 
sre gene expression was associated with excision of most of the MoMLV pra/irus. 
Dr. Gottesman asked whether gene therapy vectors could recombine with endogenous 
sequences to generate an infectious virus containing human cloned DNA. Could 
a recombination involving foreign cloned DNA, viral replication, and encapsida- 
tion functions result in a new virus? 
[457] 
