Molecular and Cellular Biology, Aug. 1986, p. 2895-2902 
0270-7306/86/082895-08 $02. 00/0 
Copyright C 1986, American Society for Microbiology 
Vo«. 6, No. 8 
Redesign of Retrovirus Packaging Cell Lines To Avoid 
Recombination Leading to Helper Virus Production 
A. DUSTY MILLER* and CAROL BUTTIMORE 
Fred Hutchinson Cancer Research Center, Seattle, Washington 98104 
Received 5 March 1986/Accepted 5 May 1986 
Retrovirus vectors can be made in the absence of helper virus by using retrovirus packaging cell lines. 
Helper-free virus is critical for a variety of gene transfer studies. The most useful packaging cell lines contain 
helper virus DNA from which the signal required for packaging of the viral RNA genome into virions has been 
deleted. However, we showed that the ability to package virus is conferred at very low frequency to cells 
infected with virus from these packaging cell lines, presumably by low-frequency transmission of the deleted 
virus genome. In addition, these packaging cell lines can interact with some retroviral vectors to yield 
replication-competent virus. We constructed packaging cell lines containing helper virus PNA that had several 
alterations in addition to deletion of the packaging signal. The new packaging cells retained the useful features 
of previously available lines but did not yield helper virus after introduction of any of the vectors tested, and 
transfer of the packaging function was not detected. 
Retroviruses have become important tools for efficient 
transfer of genes into eucaryotic cells. A major part of their 
utility is due to the availability of retrovirus packaging cell 
lines (3, 13, 13, 24, 27) which allow production of replication- 
defective retrovirus vectors in the absence of helper virus. 
Such vectors infect and integrate into cells but cannot 
replicate and spread. These properties make possible a 
variety of studies in which virus spread would make inter- 
pretation of results difficult or impossible. For example, 
helper-free vectors allow study of hemopoietic cell lineage 
relationships in intact animals owing to the presence of one 
to a few unique viral integration sites in each infected cell (3, 
11). The absence of helper virus prevents the occurrence of 
new integration events. An additional important use of 
retroviruses may be in human gene therapy (1), and these 
viruses must be helper free to avoid helper-induced disease 
or virus spread outside the treated patient. 
However, there are several problems with currently avail- 
able retrovirus packaging lines. Some have limited host 
ranges (13, 27) or produce only low titers of retroviral 
vectors (24). Both of the high-titer, wide-host-range packag- 
ing cell lines currently available are nearly identical in 
construction (3, 15) and rely on viral protein synthesis from 
a provirus almost identical with a helper virus, except that 
the signal for packaging of viral RNA has been deleted. 
Recently, however, low-level transmission of a retroviral 
vector lacking a packaging signal has been reported (12); 
thus, one might expect that the deleted viral genome in these 
retrovirus packaging cell lines would also be transmitted at 
low frequency. We demonstrate that transfer of the packag- 
ing function does indeed occur. In addition, it has previously 
been shown that certain retroviral vectors interact with these 
packaging cells to produce high levels of helper virus, 
presumably following recombination between the vector and 
packaging system (16). 
We tested several new designs for packaging cell lines and 
found that one of these, which contains several mutations in 
addition to deletion of the packaging signal, is not subject to 
the problems described above; we did not detect helper virus 
production or packaging function transfer using this line. 
• Corresponding author. 
even after introduction of vectors which caused helper virus 
production from previously described packaging lines. 
MATERIALS AND METHODS 
Cell culture. Cells were grown in Dulbecco modified Eagle 
medium with high glucose (4.5 g/liter) supplemented with 
10% calf serum (Psi-2 cells) or 10% fetal bovine serum (all 
other cell lines). Previously described cell lines included 
NIH 3T3 TK” (28), PA12 (15), Psi-2 (13), and 208F (19). The 
PA317 cell line generated by this study is available from the 
American Type Culture Collection (no. CRL 9078). Cells 
were free of Mycoplasma sp. as determined by using the 
DNA stain Hoechst 33258 (22). 
Virus assay. For assay of virus carrying selectable mark- 
ers, recipient cells were seeded at 5 x 10 5 per 60-mm dish on 
day 1. On day 2 the medium was changed to medium 
containing 4 p.g of polybrene per ml, and test virus samples 
were added. Unless otherwise indicated, virus was har- 
vested by exposing culture medium to confluent dishes of 
virus-producing cells for 16 h, removing the medium, and 
subjecting the medium to centrifugation at 3,000 x g for 5 
min to remove cells and debris. On day 3 the cells were split 
1:10 into selective medium; 10“ 7 M methotrexate for virus 
expressing a mutant dihydrofolate reductase (DHFR), 2 mg 
of G-418 (about 50% active) per ml for Neo virus, or HAT 
selective medium (30 pM hypoxanthine, 1 pM amethopterin, 
20 pM thymidine) for hypoxanthine-guanine phosphoribo- 
syltransferase (HPRT) virus. Colonies were stained and 
counted on day 9. Helper virus was measured by using the 
S*L - assay as previously described (15). 
Generation and testing of packaging cell lines. Clonal cell 
lines containing the packaging constructs were made by 
transfecting (4, 7) NIH 3T3 TK" cells, seeded the day before 
at 5 x 10 5 per 60-mm dish, with 10 pg of packaging construct 
DNA and 0.1 pg of the herpes simplex virus thymidine 
kinase gene carried as a BamHl fragment in pBR322 (2). The 
cells were grown in HAT selective medium and resultant 
TK* colonies wore isolated with cloning rings. Clones were 
screened for their ability to package a retroviral vector 
containing the selectable marker HPRT as follows. On day 
one, cells to be tested were seeded at 5 x 10 5 cells per 60-mm 
dish. On day 2, the cells were transfected with 10 pg of 
[ 368 ] 
Recombinant DNA Research, Volume 12 
