C57BL/6 mice. Mice were inoculated with 1x10^ MCA 205 tumor cells mixed with either 
lxlO 6 control 3T3-LNL6 expressing cells or with lxlO 6 LNL6 vector producing PA317 
cells. The cells were mixed just prior to subcutaneous injection. Tumors measured twice 
weekly. 
To minimize the possibility of contamination of the recovered tumor cells with 3T3 and 
PA317 cells, we waited 4 weeks before excising the growing tumors. The excised tumors 
were minced and digested into a single cell suspension. Tumor cells from each group were 
cultured for two weeks. The cultured cells were then placed in a clonogenic assay as shown 
in Table 1. (Appendix A: Reprint 1). The number of colonies that grew in medium without 
G418 were similar in each group with a cloning efficiency of 15-20%. There were no 
colonies with resistance to G418 in the tumor plus control 3T3-LNL6 expressing cell group. 
While the tumor plus VPC group had a mean of 62% ± 15 (range 55-73) G418-resistant 
colonies. An assay for NPT activity, the enzyme produced by NeoR, was positive for all 
G418-selected tumors in the VPC group. In control mice, PA317/LNL6 producer cells alone 
produces a transient tumor, that is rejected in 7-10 days. Since the evaluation of tumors in 
the experimental group was performed at 4 weeks, it is very unlikely that these G418 
resistant cells are the producer cells. The lack of G4 18-resistance in controls and the lack of 
LNL6 vector production by the recovered G418-selected cells suggests that the PA317/LN6 
producer line is not responsible for the G4 18-resistance. 
2. In Vivo Transduction of the E. coli LacZ (f j -galactosidase gene) Vector(GlBgSvNa) into the 
9L Rat Brain Tumor. To evaluate the in vivo transduction dynamics within brain tumors, we 
have used the GlBgSvNa.29 vector. GlBgSvNa.29 (produced by Genetic Therapy Inc., 
GTI) has a titer of 1-5X10 6 cfu/ml. This vector contains NeoR and the E.coli derived gene 
LacZ which encodes for the production of the enzyme P-galactosidase (B-GAL) (19). The B- 
GAL expression can be detected using an X-GAL histochemical stain. Staining the brain 
with X-GAL turns B-GAL expressing cells blue. This results when an indolyl is liberated 
from X-GAL by the action of the B-GAL enzyme. Subsequent oxidization and self-coupling 
forms an indigo blue derivative. The vector containing cells can thus be discriminated from 
unmodified cells and then be enumerated with light microscopy. Rats were inoculated with 
4xl0 4 9L gliosarcoma cells into the right cerebral hemisphere by stereotaxic guidance. Seven 
days later, 3xl0 6 PA317/GlBgSvNa VPC were injected into the tumor bearing and non- 
tumor bearing hemispheres using the same stereotaxic coordinates. Five, 9, and 14 days after 
injection of the VPC, the rats were sacrificed by an intracardiac injection of formaldehyde to 
fix the brain. The brains were stained with X-GAL. Control rats received GlBgSvNa.29 
transduced, G418 selected 3T3 non-producer cells. The VPC and the control GlBgSvNa.29 
transduced 3T3 cells were 100% positive by X-GAL staining prior to injection. 
In this experiment, we have shown that the injection of VPC led to transduction of 10-50% 
of the tumor cells in situ. Figure 2 (Appendix A: Reprint 2) presents the microscopic 
sections taken from the rats on day 7 and 14. The injected VPC disappeared from the 
injection site after day 14. There was a clear cut delineation between the transduced tumor 
cells and normal brain tissue. Except for the possibility of a rare transduced endothelial cell 
Recombinant DNA Research, Volume 17 
[155] 
