1 2 3 4 ft ft t ft ft 
Fig. 4. Neo^-coded phosphotransferase activ- 
ity in extracts from spleen foci. Spleen foci 
were assayed for neomycin phosphotransfer- 
ase activity t/V). l.ysates were subjected to 
electrophoresis on a nondenaturing polyacryl- 
amide gel: the gel was then overlayed with 
agarose containing kanamycin at 25 (ig/ml and 
2 nM |y-' : P|ATP (>5000 Ci/mmol). Subse- 
quently. the gel was blotted with Whatman 
PHI paper. The arrow indicates the position of 
neo*-coded phosphotransferase activity. 
(Lane I) Lysate of I x 10' F-5B cells. (Lane 
2) Lysate of approximately I x 10' uninfect- 
ed spleen focus cells. (Lane 3) Lysate of K 
pooled foci from one mouse, approximately 
5 * 10-’ cells after bone marrow infection with 
F-5B cells as described (legend to Table I). 
(Lanes 4 to 9) Lysate of six individual foci 
(approximately 0.5 x |0' to 2 x |(F cells 
each) after bone marrow infection with F-5B 
cells. 
In (he above experiments, bone mar- 
row cells were infected by co-cultivating 
them with virus-producing cells. To de- 
termine whether direct contact with vi- 
rus-producing cells was required, the fol- 
lowing infection protocols were per- 
formed: (i) infection by co-cultivation 
with F-5B cells: (ii) infection by co- 
cultivation with F-5B cells that had been 
washed three times with phosphate-buff- 
ered saline and then had 10 ml of fresh 
medium added just before addition of 
bone marrow cells (to reduce the starting 
virus titer to near zero): (iii) infection 
using cell-free medium derived from (ii): 
and (iv) infection as in (iii) but using 
frozen medium from F-5B cells rather 
than fresh. The results demonstrate that 
supernatant alone can infect bone mar- 
row cells but at about one-third the effi- 
ciency obtained when F-5B cells are 
present (Table 2). The lower efficiency 
could simply be due to the lack of fresh 
viral particles being generated over the 
24 hour co-cultivation period. 
Although N2 virus was known to be 
stable when used to infect tissue culture 
cells. Southern blots were performed 
with DNA obtained from individual foci 
to test whether rearrangements or dele- 
tions of the proviral sequences occur 
during the proliferation and differentia- 
tion of the infected hematopoietic stem 
cells in vivo (Fig. 3. A and B). No 
evidence of gross rearrangement was de- 
tected with several restriction endonu- 
cleases. With Sac I. which digests N2 
within both LTR's. releasing a fragment 
2H IM-C'KMHI-R IVX< 
nearly unit length, a band of expected 
size (3.2 kb) (Fig. I) was always found. 
When the enzyme Xho I was used, 
which cleaves N2 at only one internal 
site, several bands with different intensi- 
ties were often detected on the blots 
when infection took place at a high viral 
titer. This indicated that some stem cells 
had been infected several limes. DNA 
was also digested with Pst I generating 
the expected 0.9-kb fragment. 
Southern blot analysis provides only 
coarse evidence for the absence of se- 
quence alterations. However, the pres- 
ence of transcripts and phosphotransfer- 
ase enzyme activity would provide evi- 
dence that the transcriptional machinery 
of the provirus remained functional. To 
this end. mice were lethally irradiated 
and given a portion of the cells from a 
DNA-positive spleen focus. Spleens 
were removed from these secondary 
mice and RNA was prepared (#) and 
analyzed. A neo K -containing fragment of 
the appropriate size was detected on Tl 
ribonucleasc gels (9). 
The neo K 'genc product was assayed 
by its phosphotransferase activity in ex- 
tracts from individual and pooled foci. 
The majority of foci tested demonstrated 
expression of the neo K gene. All six of 
the individual foci (Fig. 4. lanes 4 to 9) 
were positive but to different extents. 
Table 2. Bone marrow infection with cell-free 
virus-containing medium. Bone marrow cells 
were isolated as described in Table I. Single 
cell suspensions were cultivated for 24 hours. 
All cultures of bone marrow contained IL-3 
(20 U/ml). Polybrene (4 ng/ml). penicillin ( 100 
U/ml). and streptomycin (100 p.g/ml). The 
cells were cultivated for 24 hours, recovered, 
and injected into lethally irradiated mice. 
Spleen foci were recovered after 10 days for 
dot blot analysis (Table I). 
Medium 
Foci 
DNA 
positive 
(No.) 
DNA 
positive 
condition 
analyzed 
foci 
m 
Willi cells 
24 hour* 
14 
II 
79 
Fresht 
12 
9 
75 
No cells 
24 hour! 
10 
3 
30 
Frozen 24 hour§ 
12 
1 
8 
‘Co-cultivation with F-5B cells producing N2 at a 
titer of ! x ||f cfu/ml 24 hours after a medium 
change. 1F-5B cells with fresh medium. The 
cells were washed three times with PBS and fresh 
medium was added just prior to the addition of bone 
marrow cells. tThc bone marrow cells were plat- 
ed into cell-free medium removed from the confluent 
F-5B cells in B. The F-5B cells had been growing in 
this medium for 24 hours producing a titer of 
2.J x 10*. Before the bone marrow cells were add- 
ed. the medium was centrifuged to remove cells and 
9 ml of supernatant was placed in a separate tissue 
culture dish. SThe bone marrow cells were plat- 
ed into cell-free F-5B medium that had been stored 
for several weeks in liquid nitrogen after removal 
from confluent F-5B cells. F-5B cells had been 
growing in this medium for 24 hours and had a titer 
of 1.6 » IlF cfu ml 
t 2 3 4 $ • r 
Fig. 5. Nco M -codcd phosphotransferase activ- 
ity in the bUntd and bone marrow of four long- 
term reconstituted mice. (Lane I) Lysate of 
I x lo* 1 whole blood cells from mouse A. 
(Lane 2) Lysate of I x l(T bone marrow cells 
from mouse A. (Lane 3) Lysate of I x 10** 
whole blood cells from mouse B. (Lane 4) 
Lysate from I x 10* bone marrow cells from 
mouse B. (Lane 5) Lysate from I x |0* whole 
blood cells from mouse C. <l.ane ft) Lysate 
from l x |0* bone marrow cells from mouse 
C. (Lane 7) I x 10* bone marrow cells from 
mouse D. The positive band in lane 7 is much 
clearer on longer exposures. The arrow indi- 
cates the position of neo M -coded phospho- 
transferase activity. The dark slower migrat- 
ing band in each lane represents a phospho- 
transferase activity seen to various degrees in 
all tissues studied, and is unrelated to the 
presence or absence of the neo* gene. 
The variation in activity among different 
foci might be partially due to varying 
sizes of the foci since cell numbers were 
not exactly equalized. More likely, how- 
ever. is that the neo K gene may be ex- 
pressed at different levels in different 
foci, either because of multiple single- 
copy insertions of the N2 vector or for 
some other reason (such as a position 
effect due to the random chromosomal 
integration of each proviral sequence). 
Studies arc under way to evaluate these 
possibilities. 
To determine the long-term structural 
stability of N2 proviral sequences, total- 
ly reconstituted animals were obtained 
by injecting irradiated mice with 5 x I0 6 
infected bone marrow cells and letting 
those cells repopulate the animal over a 
period of several months. Spleens, bone 
marrow, and blood were recovered and 
analyzed by Southern blot or phospho- 
transferase assays (or both). Southern 
blots demonstrated that the N2 se- 
quences remained intact even after 4 
months (Fig. 30. However, the propor- 
tion of total hematopoietic cells carrying 
N2 may have decreased since band in- 
tensity in spleen DNA appeared to be 
reduced when compared with an equal 
amount of DNA isolated from spleen foci 
(compare Fig. 3. A and B. with C). • 
Of major interest was whether the 
blood and bone marrow of long-term 
reconstituted animals express the neo R 
gene. Phosphotranferase was detected in 
the bone marrow of three of the four 
animals tested (Fig. 5. lanes 2. 6. and 7). 
In one animal (lane 5). a strong neo K - 
nv7 
[238] 
Recombinant DNA Research, Volume 12 
