of these G418-resistant clones for h-ADA activity and/or vector DMA sequences 
could not be performed because of the limited lifetime and number of available 
cells. 
The bone marrow cells of two additional animals #77 and #78 were infected 
with a modified protocol: the viral supernatant was at pH 7.4 rather than 6.8 
and the polybrene was mixed with the supernatant before addition of the marrow 
cells rather than afterwards. Although both animals reconstituted rapidly (see 
Figure 2) , animal #77 showed no evidence of h-ADA activity and animal #78 had 
an h-ADA level equivalent to only 0.05% of endogenous monkey activity (Tahle 
1) . In neither case were vector sequences or NFT activity detected. 
Preliminary analysis using the modified protocol to measure gene transfer into 
marine CFU-S suggests that both modifications decrease - the number of spleen 
foci that acquire vector DMA sequences (data not shown) . Whether these 
modifications can account for the primate results is still not clear. 
Discussion 
These studies were undertaken to establish a large animal model to 
evaluate protocols which might ultimately be used for gene therapy in humans. 
We have utilized an amphotropic retroviral vector containing the h-ADA cDNA, 
SAX, to infect the marrow of cynomolgus primates for subsequent infusion back 
into donor animals following total bodhf irradiation. Our studies with primates 
have demonstrated that adequate numbers of autologous bone marrow cells will 
survive after in vitro culture and exposure to a retroviral vector to provide 
the animal with full hematopoietic reconstitution and long term survival. The 
reconstituted animals have also provided evidence for a lew level of successful 
gene transfer together with expression of the inserted genes. 
As detailed in Figures 2 and 3 and Table 1, the method of infection of 
bone marrow cells is an important variable. In previous studies with the mouse. 
Recombinant DNA Research, Volume 12 [289] 
