Gene, 45 (1986) 215-220
215
Elsevier GENE
1690
Expression
mRNA in H-vus transfected NIH/3T3 cells does not suppress the transformed
of anti-sense
phenotype (H-ras
oncogene;
RNA
duplex formation;
SV40 expression
plasmid;
DNA
rearrangements;
recombinant
DNA)
B. Salmons *, B. Gtoner **, R. Friis, D. Muellener and R. Jaggi Ludwig Institute for Cancer Research, Inselspital, 3010 Bern (Switzerland) (Received
February
(Revision
received
(Accepted
Tel. (031)-261241
lOth, 1986) June 26th, 1986)
July lst, 1986)
SUMMARY
We have attempted to reverse the transformed phenotype of cells expressing the H-ras oncogene. A plasmid in which the first exon of the H-ras oncogene was coupled to the SV40 early promoter in an anti-sense orientation was constructed. This construct was introduced into a clone of H-rus-transformed NIH/3T3 cells. Simultaneous expression of both the SV40 anti-sense construct and H-rus was observed. Anti-sense RNA was present in a lo-20-fold excess over sense H-ras RNA. Only a small fraction of the cytoplasmic RNA was present in a sense : anti-sense duplexed form. The expression of anti-sense H-ras RNA was not accompanied by a phenotypic reversion of transformed cells. The only phenotypic reversion we observed was accompanied by a loss of transfected H-rus sequences. The loss of transfected H-ras sequences occurs with a high frequency in cells supertransfected with the SV40 anti-sense construct.
INTRODUCTION
Cells in culture can be transformed by a number of viral and cellular oncogenes (reviewed by Bishop, 1985). Several mechanisms have been identified which result in the activation of proto-oncogenes to
* Present Medical
address: College
Department of Georgia,
of Cell and Molecular
Biology,
Augusta,
(U.S.A.)
GA
30912
Tel. (404)-828-3271. ** To whom
correspondence
and reprint
requests
should
be
addressed. Abbreviations: hygromycin
bp, base pair(s); B; kb, kilobase
polyacrylamide;
R, resistant;
ds, double
strand(ed);
or 1000 bp; nt, nucleotide(s); ss, single strand(ed);
HyB, PA,
tk, thymidine
kinase.
0378-I 119/86/$03.50
0 1986 Elsevier
Science Publishers
B.V. (Biomedical
oncogenes (Hayward et al., 1981; Reddy et al., 1982; Tabin et al., 1982; Taprowsky et al., 1983; Capon et al., 198313; Taub et al., 1982). If the identification of oncogenes involved in human tumours progresses and clinical applications are sought, strategies designed to inhibit oncogene expression will be of importance. Several recent reports suggest that the suppression of specific gene products can be achieved by expression of anti-sense mRNA (for reviews see Laporte, 1984; Weintraub et al., 1985). We have attempted to suppress the expression of the H-ras oncogene (Shih and Weinberg, 1982) using anti-sense H-ras transcripts. We were not able to reverse the transformed phenotype of cells transfected with this oncogene. Division)
216
EXPERIMENTAL
(a)
Anti-sense
AND
1985), with a plasmid
DISCUSSION
H-rus
transcripts
are
NIH/H-rus cells supertransfected anti-H-rus construct We have constructed
present
conferring
resistance
to the
antibiotic HyB (Bfochlinger and Diggelmann, 1984) into a clonal line of H-ras transformed NIH-3T3
in
with the SV40
cells (NIH/H-~+a$). Individual
HyBR cell clones were
isolated and grown into mass culture. To determine whether the SV40 anti-H-ras construct directs trans-
a plasmid cont~ning
the first
cription
of RNA with the opposite
polarity to H-vas
exon of the c-H-ras gene linked to the SV40 early promoter in an anti-sense orientation (SV40 anti-H-
transcripts (i.e., anti-sense RNA), total cytoplasmic RNA was prepared from four HyBR cell clones. The
vus; Fig, 1D). The SV40 anti-H-ras
RNA was size-separated by agarose gel electrophoresis (McMaster and Carmichael, 1977), trans-
co-transfected
plasmid
(Wigler et al., 1979; Salmons
was et al.,
C 123
4
5 -
origin
-
28s
+
18s
: ?. :
:
1,3 kb
e-O,7
kb
D Eco
Fig. 1. Antisense
Xbal
RI
H-ras transcripts
cytopfasmic
RNA from NIH/H-raf
respectively)
was glyoxylated
a 32P nick-translated
are present
Bamtil
in NIH/H-rus
for 3 min at SO”C, separated
18 S rRNA diagrammatic (nt 1644-1905
with the SV40 anti-H-ras
supertransfected
on a 1Y0 agarose
construct.
and Carmichael,
1977) and hybridized
in vitro from the bacteriophage
transcripts
are indicated
by arrows.
to
SP6 promoter
H-ras sense RNA, the RNA probe used in (C) recognises
and the 0.7-kb SV40 anti-H-ras
5 pg of total
cell clones I, 2, 3 and 7 (lanes 2, 3, 4 and 5,
gel (MeMaster
probe (A). RNA probes (ss) were synthesized
(Melton et al., 1984). The RNA probe used in (B) recognises RNA. The 1.3-kb H-ras transcripts
cells supertransfected
(lane I) and the SV40 anti-H-ras
6.6-kb H-ras-specific
Bgl II
SV40 anti-H-tax
Cross-hybridisation
with 28 and
is seen with all three probes. Autoradiographic exposure was 3 days (A), 1 day (B), and 2 hrs (C). Panel D shows a representation (not to scale) of the SV40 anti-H-ras construct. The 26t-bp SmaI-Xbal fragment of the H-ras gene in Capon
the XbuI and BamHI vector containing efficient expression
et al., 1983a) was converted
sites of pA8lG
(Paabo
the SV40 early promoter and stability
into a BarnHI-Xbal
et al., 1983) resulting as well as the second
of RNA (Namer
and Leder,
fragment
in transcripts intron
1979).
by linker ligation. with anti-sense
and polyadenylation
This fragment
orientation.
pA8lG
signal from the rabbit
was inserted
into
is an expression j?-globin gene for
217
A
B
C
D
123
123
1 2
12345
ferred to a nitrocellulose
‘b
t
3
,
filter and hybridized
to a
32P-labelled, nick-translated H-rus probe. Authentic 1.3-kb H-rus transcripts are detected in the parental NIH/H-rus cells and in the transfected cell clones 1, 3 and 7 (Fig. lA, lanes 1, 2, 4 and 5). This RNA is also detected probe
that
if the same filter is hybridized recognizes
lanes 1,2,4
H-ras
to an ss
transcripts
and 5). Hybridization
(Fig. lB,
with the ds H-rus
probe (Fig. 1A) or an ss anti-sense
specific rus probe
(Fig. 1C) reveals
in the anti-sense
a 0.7-kb
RNA
H-rus transfected cell clones 1, 2 and 7 (lanes 2, 3 and 5). Thus the SV40 anti-H-rus construct is transcribed into a 0.7-kb RNA in three of the four supertransfected cell clones. 147 133
(b) Anti-sense H-ras mRNA is expressed fold higher levels than H-rus mRNA
-b
at 10-20-
123 90
67
H -ras
anti-sense H-ras Fig. 2. RNase protection transcripts
mapping
of H-ras and SV40 anti-H-ras (A) 10 pg of total cytoplasmic
and duplex formation.
RNA from the NIH/H-ras anti-H-ras
parental
supertransfected
cells (lane 1) and the SV40
4 and 5, respectively)
was hybridized
32P-labelled
RNA
promoter
ss H-ras
Non-hybridized analysed
probe
and complementary
and RNase
for 16 h to a from
to the SV40 anti-H-r03
by separation
Protected
the SP6 transcript.
senting the protected
RNA: RNA duplexes
on a 6% PA-8
followed by autoradiography
by SV40 anti-H-ras
ss RNA probe is indicated.
RNA
complementary
transcribed
to the H-ras
were analysed
as described
senting the protected
gel
repreand the
(B) 10 pg of total cyto-
(lane 1) and SV40 anti-H-rus
1 and 7 (lanes 2 and 3) was hybridized
ss anti-H-ras
were
M urea sequencing
for 3 h. The 350-nt fragment
region shared
plasmic RNA from NIH/H-rascells clones
at 45°C
transcribed
(ss) RNA was digested with RNase A (40 pg/ml)
Tl (2 pg/ml).
32P-labelled
1, 2, 3 and 7 (lanes 2, 3,
cell clones
to a 32P-labelled
from the SP6 promoter, transcript.
RNA.
and
RNA duplexes
in (A). The 133-nt fragment
repre-
region shared by H-rns and the 32P-labelled
ss RNA probe is indicated.
The X-ray film was exposed
to the
gel for seven days. (C) and (D) 10 pg of total cytoplasmic
RNA
from NIH/H-ras
(lane 1) and SV40 anti-H-rtrs
(lanes 2 and 3) was either digested plasmic
RNA: RNA duplex structures
at 45’C
and then digested
to SV40 anti-H-rus
duplexes
by both procedures
formed
to isolate cyto-
(C) or hybridized
to a ‘ZP-labelled
transcripts
detected in clones 1,2 and 7 (Fig. 2A, lanes 2, 3 and 5; 350 nt fragment). Expression of the H-rus (sense) transcripts was determined using radioactively labelled ss RNA spanning the first exon of H-rus. H-rus transcripts share 133 nt with the complementary SP6 RNA probe. Parental NIH/H-rus cells, and the clones 1 and 7, express similar levels of H-rus mRNA (Fig. 3B, lanes 1,2 and 3; 133-nt fragment). Similar levels of ~21’“” expression in NIH/H-rus cells and clones 1 and 7 are observed by Western blotting analysis (Table I). The intensity of the signal
clones 1 and 7
with RNase to hybridize
H-ras transcripts rehybridized
with RNase
Correctly initiated SV40 anti-H-rus mRNA was measured using the RNase protection technique as outlined by Melton et al. (1984). Briefly, radioactively labelled RNA complementary to SV40 antiH-rus RNA was transcribed in vitro from the SP6 bacteriophage promoter. The radioactive RNA probe was hybridized to total RNA isolated from the parental NIH/H-rus and the SV40 anti-H-rus supertransfected cell clones. The RNA was then treated with ss-specific RNase. RNA: RNA duplexes protected from digestion were visualized after gel separation by autoradiography. SV40 anti-H-rus RNA shares 350 nt with the complementary SP6 RNA probe. Correctly initiated anti-H-rus transcripts are
for 16 h
all available
(D). RNA: RNA
were then denatured
ss RNA complementary
and to the
H-ras transcript.
RNA:probe
RNA duplexes were then analysed
as described
in (A). The 133.nt fragment
of homology
shared
indicated labelled
representing
by H-ras and SV40 anti-H-ms
by an arrow. HpaII-digested and served as marker
pBR322
fragments
in bp). The X-ray film was exposed
the region
transcripts
is
DNA was end-
(see left margin;
sizes
to the gel for seven days.
218
TABLE I Summary
of biological
Cell
properties
of the transfected
Morphology
a
clone
cell clones
Soft agar
Tumours
growth b
nude mice”
H-ras
in
fragments
(%I
d
sv40
Human’
anti-H-ras
p2 1 ra5
fragment
(kb)
*
(kb) 51
1
T
2
NT
+ _
23
1.4
2.8
+ _
3
T
13
+
6.6
1.4 -
7
T
57
+
6.6, 15
1.4
+
NIH/H-ras
T
52
NT
6.6, 15, 23 -
-
NIH/3T3
+ _
+ _
a The morphology
0
0
of cells grown
in monolayer
culture
is indicated.
T, transformed
+
-
morphology:
NT, non-transformed,
normal
morphology. b 5 x lo4 cells were plated into medium containing 20 days and expressed ’ Tumourigenicity
as a percentage
of the cell clones was determined
2 weeks and are indicated d The presence by Southern e Expression antiserum,
0.35 % agar. The presence
of the total number
by a plus-symbol.
of H-ras-specific
blot analysis
BarnHI
by subcutaneous
Mice without
DNA fragments
and hybridization
to H-ras-
of human p21 was determined
by Western
generously
supplied
tumours
injection
of
1 x
were observed
was determined
lo6 cells into nude mice. Tumours
and the SV40 anti-H-ras-specific nick-translated
blotting
of cell-membrane
by Dr. R. Sweet, followed
colonies
for an eight-week
or SV40-specific analysis
by immunovisualization
from the RNase-resistant fragment indicative of the anti-H-ras transcripts (Fig. 2A, 350~nt fragment) was compared with that from H-ras transcripts (Fig. 2B, 133-nt fragment). Differences in size of the protected fragments and exposure time of the autoradiograms were taken into consideration. A IO-20fold greater level of expression of SV40 anti-H-ras transcripts was detected. (c) A small fraction of sense H-ras transcripts is duplexed with anti-sense
of anchorage-independent
after
of cells plated.
H-ras transcripts
One mechanism by which anti-sense mRNA might inhibit sense RNA expression is by the formation of RNA : RNA duplexes. These ds molecules can inhibit translation (Paterson et al., 1977). The SV40 anti-H-ras RNA shares 133 nt of sequence complementary to H-W mRNA. RNA was isolated from NIH/H-rus cells and the SV40 anti-H-rus supertransfected cell clones 1 and 7, and digested with RNase. If the two RNAs are duplexed intracellularly, a fragment of 133 nt should be protected from the ss-specilic digestion of RNase. RNaseresistant RNA was detected by denaturation of the duplex and hybridization to a radioactively labelled probe covering the complementary region. After a
EcoRI-BglII
appeared
after
period. DNA fragment
were determined
DNA probes. extracts
using an anti-human
with ‘251-labelled
protein
p2 l’““-specific
A.
second RNase digestion, the protected labelled probe fragment indicative of the in vivo duplexed region was visualized by gel separation and autoradiography. A weak 133-nt signal is detected using RNA isolated from cell clones 1 and 7 (Fig. 2C, lanes 2 and 3), indicating that H-rus and SV40 antiH-W transcripts are present in these cell clones in a duplexed form. The same experiment was carried out after in vitro incubation of total RNA under hybridization conditions. This allows the formation of all possible H-rus : SV40 anti-H-rus duplex molecules. These conditions result in an increased 133-nt signal (Fig. 2D, lanes 2 and 3). No signal is detected when RNA from NIH/H-rus cells is tested in this assay (Fig. 2, C and D, lanes 1). These results indicate that, under normal cytoplasmic conditions, only a small fraction of H-ru.s RNA is duplexed with SV40 anti-H-rus transcripts. (d) The transfected rearranged in SV40
H-rus gene is frequently anti-H-rus supertransfected
cells
The parental NIH/H-rus cell line was obtained by transfection of aplasmid containing a 6.6-kb BumHI DNA fragment encoding the entire H-rus gene
219
(Tabin et al., 1982) into NIH/3T3 cells. Digestion of genomic DNA from NIH/H-ras cells with BamHI, followed by separation on a 0.8% agarose gel, transfer to nitrocellulose and hybridization to a nicktranslated H-ras-specific probe, reveals three hybridizing bands of 6.6,15 and 23 kb (Table I). The two larger fragments probably contain H-ras copies which have lost one of the flanking BamHI sites during transfection and integration. The analysis of the three SV40 anti-H-ras supertransfected NIH/Hras clones reveals that these sublines have lost or rearranged some of these H-ras-specific BamHI fragments. SV40 anti-H-ras supertransfected cell clone 1 contains only the 23-kb fragment (Table I). This fragment contains a functional H-ras gene copy since this clone grows in soft agar and forms tumours upon injection into nude mice (Table I). The nontumourigenic clone 2 does not express H-ras mRNA (Fig. 1, A and B, lanes 3). This clone has lost all the parental H-ras-specific fragments and a novel 2.8-kb fragment is detected (Table I). Clone 3 is tumourigenie and has only retained the 6.6-kb fragment of the parental NIH/H-ras cells (Table I). Both the 6.6and 15-kb fragments are retained in the tumourigenic clone 7 (Table I). The presence of the SV40 anti-Hras construct in the supertransfected cell lines can be determined by restriction enzyme digestion with EcoRI + BglII, followed by hybridization to an SV40-specific probe (Fig. 1D). The indicative 1.4-kb EcoRI-BglII fragment is present in the DNA of cell clones 1,2 and 7. It is not present in clone 3 or in the parental NIH/H-ras cells (Table I). The only clone that shows a phenotypic reversion (clone 2) has lost functional copies of the H-ras gene. We have not been able to reverse the transformed phenotype of H-ras-transfected cells by simultaneous expression of H-ras and anti-sense H-ras RNA. We have obtained similar results with a construct in which the SV40 early promoter is linked, in the anti-sense orientation, to the v-mos gene (Van Beveren et al., 198 1) and transfected into cells transformed by the v-mos oncogene (not shown). Recently Kim and Wold (1985) have demonstrated that even a 300-fold excess of anti-sense tk RNA over sense tk RNA reduces, but does not completely abolish, tk enzyme activity. N. Hasan, G. Somasekhar and W. Szybalski (personal communication) also have found that a large excess of both galK and ,?-phage gene-N anti-mRNAs have little effect on the expres-
sion of the corresponding sense genes. A complete suppression of H-ras expression is probably required to reverse the transformed phenotype of NIH/H-ras cells. Sufficiently high levels of anti-sense H-ras RNA cannot be obtained in supertransfected cells even if the anti-sense construct is transcriptionally regulated by the strong SV40 promoter. The extremely high levels of anti-sense RNA necessary for absolute gene suppression will probably limit its applications.
ACKNOWLEDGEMENTS
We thank Dr. S. Kozma (Ludwig Institute, Bern, Switzerland) for NIH/H-ras cells, Dr. R. Sweet (Smith, Kline and French Laboratories, Swedeland, PA, U.S.A.) for anti-human ~21’“” antiserum, and Dr. W. Schaffner (University ofZurich, Switzerland) for plasmid pA8 1G. We are grateful for the excellent technical assistance of Ms. S. Saurer and A. Schlafli. We would also like to thank Dr. W.H. Glinzburg for critical comments on the manuscript, Ms. C. Wiedmer and M.T. DiabatC for editorial assistance and J. Grtinig and P. Wegmliller for art work.
REFERENCES Bishop,
J.M.: Viral oncogenes.
Blochlinger,
Cell 42 (1985) 23-38.
K. and Diggelmann,
transferase
as a selectable
ments with higher eucaryotic 2929-293 Capon,
Goeddel,
bladder
D.J.,
Edaman,
carcinoma
Seeburg,
oncogene
Seeburg,
sequence
P.H. and of the T24
and its normal
homo-
33-37.
P.H.,
McGrath,
J.P.,
A.D. and Goeddel,
of Ki-ras gene in human Hamer,
A.D.,
nucleotide
302 (1983a)
U., Levinson,
different
experi-
cells. Mol. Cell. Biol. 4 (1984)
E.Y., Levinson,
D.V.: Complete
logue. Nature Capon,
B phospho-
for DNA transfer
1.
D.J., Chen,
human
H.: Hygromycin
marker
Hayflick,
colon and lung carcinomas
point mutations.
Nature
J.S.,
D.V.: Activation by two
304 (1983b) 507-513.
D.H. and Leder, P.: Splicing and the formation
of stable
RNA. Cell 18, 1299-1302. Hayward,
W.S.,
cellular lymphoid
Neel,
oncogenes leukosis.
B.G. and Astrin, by promoter Nature
in cells expressing
Cell 42 (1985) 129-138.
of
in ALV-induced
290 (1981) 475-480.
Kim, S.K. and Wold, B.J.: Stable reduction activity
S.M.: Activation
insertion
of thymidine
kinase
high levels of antisense
RNA.
220
Laporte,
D.C.:
control
Anti-sense
RNA:
of gene expression.
McMaster,
G.K. and Carmichael,
double stranded
A new mechanism
Trends
Biochem.
G.G.: Analysis
of single and
nucleic acids on polyacrylamide
gels by using glyoxal and acridine
orange.
for the
9 (1984) 463. and agarose
Proc. Natl. Acad.
Sci. USA 74 (1977) 4835-4838. Melton,
D.A., Krieg, P.A., Rebagliati,
M.R., Maniatis,
K. and Green, M.R.: Efficient in vitro synthesis active RNA and RNA hybridization containing
a bacteriophage
ofbiologically
probes
SP6 promoter.
T., Zinn,
from plasmids
Nucl. Acids Res.
12 (1984) 7035-7056. Pllbo,
S., Weber,
viral Paterson,
protein
vector.
transplantation synthesised
antigens
from
and a
a mammalian
by DNA.
Proc.
Natl.
mRNA
gene identiti-
hybrid
Acad.
arrested
Sci. USA
cell
74 (1977)
4370-4374. Reddy,
mutation
forming Salmons,
is responsible
properties
oncogene.
Nature
B., Groner,
Production fection Virology
R.K., Santos,
E. and Barbacid,
for the acquisition
by the T24 human
bladder
M.: A of trans-
carcinoma
300 (1982) 149-152. B. Calberg-Bacq,
Tabin,
C.J., Bradley,
Papageorge,
of mouse
of a recombinant
mammary proviral
144 (1985) 101-I 14.
tumour
H.:
virus upon trans-
DNA into cultured
Isolation
bladder
of a transforming
carcinoma
S.M., Bargmann,
A.C., Scolnick,
cell line. Cell 29
E.H.: Mechanism
Nature
300 (1982) 143-149.
Taprowsky,
E., Shimizu,
Structure
CL,
of activation K., Goldfarb,
and activation
Weinberg,
R.A.,
E.M., Char, R., Lowy, D.R. and
Chang,
of a human oncogene. M. and
Wigler,
ofthe human N-rasgene.
M.:
Cell (1983)
581-586. I., Morton,
S., Aaronson,
C., Lenoir, G., Swan, D., Tronick,
S. and Leder,
P.: Translocation
gene into the immunoglobulin Burkitt
lymphoma
Van Beveren,
of the c-mJ,c
heavy chain locus in human
and murine
plasmacytoma
cells. Proc.
Sci. USA 79 (1982) 7837-7841. C., Van Straten,
I.M.: Nucleotide
sequence
F., Gallenshaw,
J.A. and Verma,
of the genome of a murine sarcoma
virus. Cell 27 (1981) 97-108. H., Izant, J.G. and Harland,
as a molecular
tool for genetic
R.M.: Anti-sense
analysis.
Trends
RNA
Genet.
1
(1985) 22-25. Wigler, M., Sweet, R., Sim, G.K., Weld, B., Pellicer, A., Lacy, E., Maniatis,
C-M. and Ponta,
R.A.:
(1982) 161-169.
Weintraub,
E.P., Reynolds,
point
Weinberg,
from a human
Nat]. Acad.
B. and Kuff, E.: Structural
and mapping
free translation.
W. and Peterson,
Cell 33 (1983) 445-453.
B., Roberts,
cation
O., Schaffner,
between
membrane
expression
C. and
sequence
Taub, R., Kirsch,
F., Kaempe,
P.A.: Association
Shih,
mammalian eukaryotes.
T., Silverstein, cells
with
S. and Axel, R.: Transformation genes
Cell 16 (1979) 777-785.
cells. Communicated
by Z. HradeEnC.
from
prokaryotes
of and