producer will be performed using an ecotropic DFG producer. A similar method has 
resulted in an enhancement of the titers of many of the MFG-based vectors without the 
detection of replication-competent virus. The best producer of DFG-IRAP-tk as 
determined by IRAP expression and tk resistant colonies will also be titered by Southern 
analysis to confirm the integrity of the provirus structure after infection. 
Detection of replication competent virus: To determine if the producer lines 
contained replication competent virus, a BAG mobilization assay was, and will be, used. 
Briefly, we have established a 3T3 population infected with the BAG retrovirus, 
expressing both LacZ and neor, and G418 selected. Supernatant from the CRIP 
producer was added to the 3T3-BAG cells and cocultured through several passages in 
the presence of polybrene. Supernatants from these cocultures were used to infect 3T3 
target cells and the infected cells selected in G418. As a positive control, supernatant 
from a PA317 producer which had a low level of helper virus was used. The MFG-IRAP 
and MFG-tk producers were unable to mobilize BAG: no G418 resistant colonies were 
detected. 
Expression of two genes in MFC (DFG vector): The strategy employed to express 
IRAP and tk genes in the same MFG vector is to use an internal ribosome entry site in 
the context of a polycistronic message. We previously have demonstrated that two genes 
can be expressed efficiently in MFG using a polycistronic message with one, and even 
two, IRES sequences. To examine the level of expression of two genes in MFG, IRAP 
and human growth hormone were used initially since their expression is easily assayed in 
a quantitative manner by ELISA and RIA respectively. A restriction map of DFG- 
IRAP-GH is shown in fig. 10. The level of IRAP and growth hormone expression from 
the DFG-IRAP-GH vector was compared to the level of expression of each protein from 
MFG-IRAP and MFG-GH vectors. BOSC23 cells were transiently transfected with the 
three plasmids and the supernatants harvested 48-72 hours post-transfection. 3T3 cells 
were infected with 5 ml of the BOSC23 supernatants in the presence of 8 g of polybrene. 
72 hours post-infection the level of growth hormone and IRAP was determined in the 
media of the infected cells. As is shown in table 1, the level of expression of the two 
genes in the DFG vector were roughly equivalent to their levels of expression in the 
single MFG vectors. However, it appears as if expression of the gene 3' to theTRES is 
expressed at a slightly lower level. Therefore, the vector carrying the HSV tk gene were 
constructed with the tk gene 3' to the IRES sequence. 
(ii) Ex Vivo Gene Transfer to Rabbit Knee Joint 
New Zealand White rabbits weighing 5-61bs are anesthetized by i.v. injection of 
1.5ml of Nembutal, one knee joint shaved and a partial synovectomy performed by sharp 
dissection. Cells released from the synovial sample by digestion with clostridial 
collagenase (0.2% w/v; 3h; 37°) and seeded into 25cm^ plastic culture flasks in the 
presence of 4mls of Ham's Fjj medium containing 10% fetal bovine serum. When 
confluent, the cells are recovered by treatment with trypsin (0.25% w/v) and seeded at a 
1:2 split ratio into further flasks. After 3 such passages any monocytes, macrophages. 
[436] 
Recombinant DNA Research, Volume 19 
