Vol. 6. 1986 
RETROVIRUS PACKAGING CELL LINES 
transfected into Psi-2 cells, helper virus production from 
clonal cell lines was not detected (6, 11). Psi-2 cells contain 
a packaging construct which is similar to that of PA12 cells 
except that the envelope regions are from murine viruses 
with different host ranges. Thus, one might predict that Psi-2 
cells would interact with N2 virus to generate helper virus in 
a fashion analogous to that of PA12 cells. We tested for 
production of helper virus from Psi-2 cells after introduction 
of N2 virus by using the XC helper virus assay (21). This 
assay is used for detection of Moloney murine leukemia 
virus, the virus on which the Psi-2 line is based. In addition 
to isolating clonal cell lines containing N2 virus, we analyzed 
cell populations containing many independent clones so that 
a rare event resulting in helper virus production might be 
detected. These populations of cells were made by transfec- 
tion of Psi-2 cells with pN2 or by infecting the cells with 
helper-free virus from the cell line PA317-N2 ell (Table 3). 
All of the populations of Psi-2 cells containing N2 virus 
secreted helper virus in addition to N2 virus (Table 4). In 
addition, three of the six clonal N2-transfected Psi-2 cell 
lines secreted helper virus in addition to N2 virus. The level 
of helper virus secretion was highly variable among the cells 
analyzed, which may reflect a relatively infrequent event 
which generates helper virus followed by slow spread of 
helper virus in the ceils. In contrast, we did not detect helper 
virus production from Psi-2 cells not containing vectors. 
Thus, even though it may be possible to isolate clones of N2 
virus-containing Psi-2 cells which do not secrete helper 
virus, production of helper virus is a frequent event. 
Analysis of packaging clones that produce low-titer virus. 
Some of the PA317 cell clones infected with selectable 
vectors produced low-titer virus, for example, PA317-SDHT 
c7 (Table 2) or PA317-N2 c3 (Table 3). The explanation for 
this might be that the integrated virus suffered mutations so 
that it could not efficiently replicate or that PA317 cells are 
heterogeneous in their ability to express virus-packaging 
functions. To answer this question, we infected several 
PA317-SDHT and PA317-N2 clones with AM-MLV helper 
virus, passaged the cells for over 2 weeks to allow the helper 
virus to spread, and assayed the clones for production of 
virus. After infection, all of the clones analyzed produced 
high vector titers (Tables 2 and 3), showing that the inte- 
grated SDHT and N2 vectors in these cells were not defec- 
tive. We therefore concluded that there is heterogeneity in 
the packaging ability of PA317 cells, possibly because of loss 
of the transfected pPAM3 DNA that is required for retrovi- 
rus vector packaging. 
Sensitive assay for helper virus production. We next at- 
tempted to increase the sensitivity of the S + L~ assay in an 
attempt to detect helper virus production from PA12 or 
PA317 cells. In many experiments performed in this labora- 
tory we have never seen helper virus production from PA12 
cells as measured by the S + L" assay. We have detected 
helper virus production from retrovirus vector-containing 
cells only in the case of N2 Neo virus and its derivatives. 
However, an attempt to increase the sensitivity of the assay 
might result in helper detection or detection of a rare event 
involving transfer of the packaging function. 
The S + L“ helper virus assay involves exposure of cat 
cells harboring a replication-defective transforming virus to 
test virus, followed by overlay of the cat cells with 
nontransformed rat NRK cells. The presence of helper virus 
in the test virus is indicated by focus formation in the 
otherwise flat NRK cells, owing to rescue of the transform- 
ing virus followed by infection and transformation of the 
NRK cells. To increase the sensitivity of the assay, we 
[ 372 ] 
TABLE 4. Virus production from Psi-2 cells containing 
N2 Neo virus* 
N2 vtni»- 
truixfected 
Pti-2 ccQ 
done* 
Neo virut titer 
(CFU/ml) 
Helper titer 
(PFU/mlp 
1 
2 x 10* 
<1 
2 
9 x 10 4 
<1 
3 
8 x 10 5 
UP 
4 
4 x 10 5 
>10 J 
3 
2 x 10* 
>10 J 
6 
3 x 10 5 
<1 
Population a 
>10" 
2 x 10 1 
Population b 
>10* 
8 
Population c 
>10* 
5 x HP 
Population d 
>10* 
>1(P 
* Cell populations a and b were derived by infection of Psi-2 cells with 
helper-free virus from PA317-N2 ell cells. Cell populations c and d were 
derived by transfection of Psi-2 cells with pN2 without cloning the resultant 
colonies. 
* Helper virus was measured by using the XC assay as previously described 
(17). 
repeatedly exposed cat cells at 12-h intervals to large quan- 
tities of medium harvested from cells to be tested for helper 
virus production and then cocultivated the cat cells with 
NRK cells. The assay was performed using PA12 and PA317 
cells and DHFR virus-producing PA12 and PA317 cells 
(PA12-SDHT c6 and PA317-SDHT c3 (Table 2]). No foci 
were induced by any of these cell lines; thus, we were unable 
to detect helper virus production by these lines (<0.1 helper 
virus per ml) (data not shown). In contrast, AM-MLV helper 
virus harvested from AM-MLV virus-producing NIH 3T3 
TK - cells induced 3 x 10 6 foci per ml of medium. In 
addition, critical observation revealed no differences be- 
tween the test plates and a mock-infected control dish; thus, 
there was no indication in this assay of transfer of the 
packaging function. Transfer of the packaging function into 
cat cells should have resulted in production of transforming 
virus, but the rate of production may have been too low to 
result in focus induction in NRK cells, or perhaps helper 
virus-mediated spread of transforming virus is required for 
focus formation. 
Detection of packaging function transfer. We designed the 
experiment depicted in Fig. 3 as a sensitive assay for 
detection of packaging function transfer. First, NIH 3T3 
(TK - ) cells harboring but not producing a Neo virus (NIH 
3T3-N2 cells) were exposed to medium harvested from 
various packaging cells. If the NIH 3T3-N2 cells beca m e 
infected by virus, resulting in transfer of the packaging 
function, they would begin to secrete Neo virus. We next 
mixed the infected NIH 3T3-N2 cells with rat HPRT" 
fibroblasts, cocultivated the cells for 4 days, and assayed the 
mixed populations for the presence of HPRT - , Neo virus- 
infected rat cells by exposing the cells to G-418 and 6- 
thioguanine and scoring resistant colonies. 6-Thioguanine 
kills HPRT- cells, so the NIH 3T3-N2 cells were killed, and 
G-418 kills rat cells unless they are infected with Neo virus. 
Colony formation was thus indicative of rescue of the Neo 
virus by virus produced by the packaging cells. The reason 
for the cocultivation step in this assay was that small 
amounts of virus produced by a few NIH 3T3-N2 cells were 
not detected in medium exposed to the cells (data not 
shown), presumably because the virus bound rapidly to cells 
in the dish which were not producing virus and thus had 
available viral receptors. Cocultivation of NIH 3T3-N2 cells 
Recombinant DNA Research, Volume 12 
