Biochemistry 1998, 37, 6235-6239 6235

Up-Regulation of Luciferase Gene Expression with Antisense Oligonucleotides:

Implications and Applications in Functional Assay Development†
Shin-Hong Kang,‡ Moo-Jung Cho,*,‡ and Ryszard Kole§
School of Pharmacy, Department of Pharmacology, and Lineberger ComprehensiVe Cancer Center, UniVersity of
North Carolina, Chapel Hill, North Carolina 27599
ReceiVed February 5, 1998; ReVised Manuscript ReceiVed March 24, 1998

ABSTRACT: HeLa Tet-Off cells were transfected transiently as well as stably with a recombinant plasmid
(pLuc/705) carrying the luciferase gene interrupted by a mutated human β-globin intron 2 (IVS2-705).
The mutation in the intron causes aberrant splicing of luciferase pre-mRNA, preventing translation of
luciferase. However, treatment of the cells with a 2′-O-methyl-oligoribonucleotide targeted to the aberrant
splice sites induces correct splicing, restoring luciferase activity. The effects are sequence-specific, depend
on the concentration of the oligonucleotide, and can be modulated by the pretreatment of the cell line,
Luc/705, with tetracycline. Thus, the cell line provides, among others, a novel functional assay system
superior to other procedures that are based on protein down-regulation. In particular, the system would
be ideal in assessing the cellular delivery efficiency of antisense oligonucleotides.

Synthetic oligonucleotides have shown promising thera- influence of delivery-related issues on the overall antisense
peutic potentials as antigene or antisense agents (1-5). activity, because the antisense oligonucleotides may be
However, the development has been greatly hampered by targeted to sequence elements of different accessibility (14)
the uncertainties associated with the mechanism of action or delivered to the cells by different techniques (15). Here
(6), the nonspecific binding of certain oligomers to cellular we report on a novel cellular assay system in which the
proteins (7), the RNase H-mediated nonantisense mechanism activity of antisense oligonucleotides results in the up-
(8), and perhaps most importantly by the lack of means to regulation of the luciferase gene expression and therefore is
deliver the oligomers at the tissue/organ as well as (sub)- free of the drawbacks mentioned above.
cellular level (9-12).
The development of effective means to deliver antisense EXPERIMENTAL PROCEDURES
agents to the intracellular target has been slowed in part by Materials. LipofectAmine as well as the cell culture media
the lack of a simple, rapid, and reliable way to evaluate the used were from Life Technologies, Inc. (Gaithersburg, MD)
delivery efficiency. A typical cellular system for assessing unless specified otherwise. The phosphorothioate 2′-O-
antisense activity is based on the inhibition of protein methyl-oligoribonucleotides used in the present study were
synthesis. Such a system suffers from several drawbacks: gifts from Hybridon Inc. (Cambridge, MA) and used as
low signal-to-noise ratio, delayed response originating from received: ON-705, CCUCUUACCUCAGUUACA; ON-
the stability of protein translated from the targeted mRNA, 705M, CCUCUUACAUCAGUUACA; ON-119, UGAGAC-
and unknown contribution of cytotoxicity from a given UUCCACACUGAU; and ON-3′C, CAUUAUUGCCCUG-
strategy to global protein synthesis. Moreover, the assay AAAG were targeted to respective sequences in the intron
does not allow one to ascertain the intracellular site, nuclear 2 of human β-globin gene. As an additional control, a
versus cytoplasmic, of the antisense activity. Although the randomized oligonucleotide pool, that is, a preparation
latter issue can be addressed by confocal fluorescence synthesized as a library of all possible 18-mers, was used.
microscopy, this technique is far from being unequivocal HeLa Tet-Off cell line, luciferase plasmid pTRE-Luc and
(13). Likewise, it is frequently difficult to determine the hygromycin-resistant plasmid pTK-Hyg were purchased from
Clontech Laboratories, Inc. (Palo Alto, CA). Primers used
† Supported by USPHS Grant GM50768 (S.H.K. and M.J.C.) and in PCR were custom-synthesized in the Lineberger Com-
HL32352 (R.K.). prehensive Cancer Center, the University of North Carolina.
* To whom reprint requests should be directed. Tel: 919-966-1345.
Fax: 919-966-7778. E-mail: [email protected].
Hygromycin was from Boeringer-Mannheim (Indianapolis,
‡ School of Pharmacy, University of North Carolina at Chapel Hill. IN).
§
Department of Pharmacology and Lineberger Comprehensive Methods. Plasmid and Cell Line Construction. The
Cancer Center, University of North Carolina at Chapel Hill.
1 Abbreviations: DMEM, Dulbecco’s modified eagle medium; FCS,
thalassemic β-globin intron IVS2-705 or its wild-type
fetal calf serum; IVS2-705, human β-globin intron 2 with a T-to-G equivalent IVS2-GL (16) was inserted into the luciferase
mutation at nucleotide 705; Luc/705, HeLa Tet-Off cells transfected cDNA sequence of the pTRE-Luc plasmid, between nucleo-
with pLuc/705; ON, oligonucleotide; PCR, polymerase chain reaction; tides 1368 and 1369, by PCR-based in vivo recombination
pLuc/705, recombinant luciferase plasmid interrupted by IVS2-705; (17) resulting in plasmids pLuc/705 or pLuc/GL, respectively
pLuc/GL, wild-type luciferase plasmid containing normal β-globin
intron 2; pTRE-Luc, wild-type luciferase plasmid from Clontech; RT- (Figure 1). The details of the construction are available on
PCR, reverse-transcription PCR. request. HeLa Tet-Off cells were transfected with the
S0006-2960(98)00300-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/14/1998
6236 Biochemistry, Vol. 37, No. 18, 1998 Accelerated Publications

FIGURE 1: Pre-mRNA of modified luciferase gene, pLuc/705.

Shaded boxes and solid thick lines represent sequences derived from
luciferase cDNA and human β-globin intron 2 carrying a T-to-G
mutation at nucleotide 705, respectively. Aberrant splice sites
activated by the mutation are indicated. Short bars under the intron
indicate phosphorothioate 2′-O-methyl-oligoribonucleotides targeted
to the aberrant splice sites (ON-705, ON-3′C), an upstream sequence
(ON-119), or with a random sequence (ON-Ran). ON-705M
hybridizes to its target sequence with a single mismatch. ON-705
and ON-3′C prevent aberrant (dashed line) and restore correct (solid
thin line) splicing.

recombinant plasmids using LipofectAmine as a carrier

following the supplier’s instructions. Stable transfectants
were obtained by cotransfecting the cells with the hygro-
mycin-resistant plasmid and one of the above plasmids or
the original pLuc plasmid, followed by selection in Dulbec-
co’s modified eagle medium (DMEM) containing hygromy-
cin (200 µg/mL) and 10% fetal calf serum (FCS).
Oligonucleotide Treatment. The cells were treated with
a complex of LipofectAmine and oligonucleotides. The FIGURE 2: Luciferase activity in HeLa cells transiently transfected
complex was prepared as follows. A 0.1-mL aliquot of an with plasmids pTRE-Luc, pLuc/705, and pLuc/GL (A), and in
oligonucleotide at a given concentration in Opti-MEM was transiently transfected HeLa Luc/705 cells treated with oligonucleo-
tides ON-705 (black columns), ON-119 (shaded), and ON-Ran
mixed with 0.1 mL of Opti-MEM containing 4 µL of (white) (B). In this and subsequent figures, the activity of luciferase
LipofectAmine. After being briefly mixed, the preparation was normalized in total cellular protein and is presented in relative
was left at room temperature for 15 min, followed by dilution luminescence units (RLU)/µg of protein. Vertical bars in A indicate
to 1.0 mL with Opti-MEM. The cells were incubated for 5 standard deviations (n ) 3).
h. Subsequently the medium was replaced with DMEM
transfected with the vectors carrying either intact luciferase
containing 10% FCS, and an additional 18 h later, the protein
cDNA (pTRE-Luc) or the one with the coding sequence
and the RNA were collected for luciferase (18) and RT-PCR
interrupted by the wild-type β-globin intron (pLuc/GL)
(19) assays, respectively. Forward and reverse primers
exhibited 20-25-fold higher activity (Figure 2A). The latter
(TTGATATGTGGATTTCGAGTCGTC and TGTCAAT-
result shows that the wild-type β-globin intron can be spliced
CAGAGTGCTTTTGGCG, respectively) hybridized to the
out even though it is located in an unnatural sequence
luciferase sequences flanking the intron.
context. The difference in the levels of luciferase expression
RESULTS in cells transfected with pTRE-Luc and pLuc/GL is most
likely due to differences in the transfection efficiency.
InactiVation of Luciferase Gene by Insertion of IVS2-705. Restoration of Luciferase ActiVity by ON-705, an Antisense
A single-point mutation at nucleotide 705 in the intron 2 of Oligonucleotide Targeted to the Aberrant Splice Site. The
human β-globin gene, referred to as IVS2-705 in this report, cells transiently transfected with pLuc/705 plasmid were
creates an additional 5′ splice site and activates a cryptic 3′ subsequently treated with a complex of LipofectAmine and
splice site 126 nucleotides upstream. This and similar the 18-mer 2′-O-methyl-phosphorothioate oligoribonucleotide
mutations lead to aberrant splicing which results in the (ON-705). The latter is targeted to the aberrant 5′ splice
retention of the intron fragment with a stop codon in the site created by the 705 mutation (16, 21). This treatment
coding sequence of the β-globin mRNA and concomitant resulted in a dose-dependent restoration of luciferase activity
inhibition of the translation of the full-length β-globin (Figure 2B), which increased approximately 50-fold over the
protein. This mechanism is responsible for a large number background at the optimum concentration of the oligonucleo-
of cases of a genetic blood disorder, thalassemia (20). tide (0.3 µM). A decline seen at 0.5 µM is due to the
Insertion of the IVS2-705 human globin intron into the improper charge ratio of the cationic lipid and the negatively
cDNA of luciferase plasmid pLuc/705 resulted in an aber- charged nucleic acid in the complex (22). Note that a
rantly spliced intron and a generation of the stop codon constant amount of LipofectAmine was used in this and
interfering with the translation of luciferase (Figure 1). Thus, subsequent experiments (see Methods section).
HeLa cells transiently transfected with the pLuc/705 exhib- The effect was sequence-specific; neither an oligonucleo-
ited only background levels of luciferase activity, while those tide targeted to a sequence adjacent to nucleotide 119 (ON-
Accelerated Publications Biochemistry, Vol. 37, No. 18, 1998 6237

FIGURE 3: Restoration of splicing by antisense oligonucleotides in the stably transfected HeLa Luc/705 cell line. (A) Luciferase activity.
The oligonucleotides used were: ON-705 ([), ON-705M (9), ON-119 (2), and ON-Ran (×). (B) RT-PCR. Total cellular RNA was
subjected to RT-PCR. The 268- and 142-nucleotide bands correspond to the aberrant and correct mRNA, respectively: lanes 1 and 2,
control cells expressing the luciferase cDNA (Luc) and the pLuc/GL construct (GL) in which the luciferase cDNA is interrupted by the
wild-type β-globin intron, respectively; lanes 3-8, Luc/705 cells treated with ON-705 at varying concentrations; and lanes 9 and 10,
Luc/705 cells treated with ON-119 and ON-Ran, respectively. In (C) and (D), the luciferase activity and RT-PCR assays are presented,
respectively, for the Luc/705 cells treated with ON-3′C. The latter is targeted to the 3′ splice site activated by the IVS2-705 mutation. All
designations are the same as above.

119), which is irrelevant for splicing, nor the one with a To obtain direct evidence that the generation of luciferase
random sequence (ON-Ran) showed any effect on luciferase in Luc/705 cells was due to the correction of aberrant splicing
activity (Figure 2B). Furthermore, the luciferase level in of the modified luciferase pre-mRNA, we have subjected
cells transfected with pLuc/GL in which the intron is the total RNA from the cells treated with the oligonucleotides
correctly spliced was not affected by the treatment with to quantitative reverse transcription PCR (RT-PCR) (19, 25).
varying concentrations of the oligonucleotides (data not The results show that ON-705 indeed induced correction of
shown). Taken together, these results indicate that the effects pre-mRNA splicing. In the untreated cells the only detect-
of ON-705 were due solely to its ability to restore correct able RT-PCR product was 268 nucleotides long. It is the
splicing of the IVS2-705 intron and exclude the possibility aberrantly spliced luciferase mRNA that retained the 126-
of nonspecific activation of the luciferase gene expression nucleotide fragment of the intron (Figure 3B, lane 3). This
(see also below). They clearly show that the concept of an result indicates that, despite the presence of normal 3′ and
assay system based on protein up-regulation by oligonucleo- 5′ splice sites at the ends of the inserted intron, splicing
tides works well, with a high signal-to-noise ratio. Note that follows exclusively the aberrant pathway, leading to little
2′-O-methyl-phosphorothioate oligoribonucleotides do not or no correctly spliced luciferase mRNA. However, a
promote cleavage of the hybridized RNA by cellular RNase significant level of the 142-nucleotide product representing
H (23). This is important because the degradation of the correct splicing was already generated at the lowest con-
RNA would have led to the removal of the splicing substrate centration of the oligonucleotide tested, 0.02 µM (lane 4).
(24). Analysis of the data, by densitometry of the autoradiogram
Restoration of Luciferase ActiVity and Correction of Pre- and calculation of the ratio of the corrected product to the
mRNA Splicing in Stably Transfected Cell Line. The sum of the corrected and aberrant products, indicated that
experimental system was further improved by generating a the correction of splicing increased in a dose dependent
HeLa cell line stably transfected with the pLuc/705 construct. manner (lanes 5-7) and reached close to 80% at 0.4 µM
Figure 3A shows that the treatment of the Luc/705 cells with oligonucleotide. In agreement with the results obtained by
ON-705 generates luciferase activity in a dose-dependent and luminometry, 50% correction occurred at approximately 0.1
sequence-specific manner. Control oligonucleotides, ON- µM oligonucleotide (Figure 3A).
119 and ON-Ran, were inactive while a single mismatch in As expected, the control cell lines, constitutively express-
ON-705M resulted in an approximately 70% decrease in the ing either the luciferase mRNA transcribed from an intronless
antisense effectiveness of ON-705. construct (pTRE-Luc, lane 1) or the one with a wild-type
6238 Biochemistry, Vol. 37, No. 18, 1998 Accelerated Publications
indicated the inadequacy of such an assay system (7). Even
an assay based on down-regulation of a given protein,
although more specific, suffers from several drawbacks
described in the Introduction.
In contrast to previous approaches (28), the oligonucleotide
ON-705 corrects the aberrant splicing of the mutated intron
which interrupts the coding sequence of a luciferase gene.
The positive readout thus obtained effectively eliminates the
flaws associated with a functional assay based on protein
down-regulation.
Analysis of a kinetic scheme involving two compartments,
one for the mRNA pool and the other for the protein pool
(29, 30), can demonstrate that the luciferase activity at a
steady state is proportional to the extracellular concentration
FIGURE 4: Restoration of luciferase activity in Luc/705 cells that of ON-705 with rate-limiting cellular entry of ON-705.
had been incubated for 48 h with tetracycline at indicated Assumptions made in deriving this conclusion include the
concentrations prior to subsequent treatment with 0-0.2 µM ON-
705. following: the gene is transcribed following a zero-order
process (31); hybridization occurs rapidly compared with
β-globin intron (lane 2), showed only correctly spliced other kinetic processes involved; and the extracellular
mRNA. In another series of control experiments, irrelevant concentration of ON-705 remains constant during the 5-h
oligonucleotides, ON-119 and ON-Ran, were found to be incubation of the cells with the oligonucleotide. Since
completely ineffective in inducing the cells to correctly splice splicing occurs exclusively in the nucleus, the luciferase
luciferase mRNA (lanes 9 and 10, respectively). Interest- activity observed reflects only the nuclear concentration of
ingly, an antisense oligonucleotide targeted to the 3′ cryptic the oligonucleotide. The nuclear envelope embedded with
splice site (ON-3′ C), which is activated by the 705 mutation, pores of 70 nm (32) is, however, expected to allow free
also resulted in a dose-dependent increase in the luciferase passage of oligonucleotides of 15-20-mer in size. There-
activity (Figure 3C) and concomitant correction of the fore, the results obtained will be inclusive in assessing both
splicing (Figure 3D). However, this oligonucleotide is less cytoplasmic and nuclear delivery of an oligonucelotide. In
effective and even at 0.4 µM generates no more than 20% summary, the present system provides a means of testing
of the corrected product. This may reflect its shorter length and an objective comparison of various cellular delivery
(17-mer vs 18-mer for ON-705) and/or decreased accessibil- strategies as well as other parameters such as the influence
ity of the 3′ splice site (19) which interacts with a large of the backbone structure of oligonucleotides.
number of splicing factors (26). The luciferase assay is extremely sensitive; for example,
Modulation of Luciferase Gene Expression by Tetracycline approximately 1000-fold more sensitive than the chloram-
and Effects of ON-705. The HeLa cells used in the present phenicol acetyltransferase assay (33). In Tet-Off HeLa cells,
study harbor tetracycline-controlled transactivators while the the luciferase expression is further amplified and tightly
luciferase gene promoter contains a tetracycline-responsive regulated compared with the cells transfected with other
element. In this system, commonly referred to as Tet-Off, expression vectors (34). Since the Luc/705 cell line offers
the transcription of the luciferase gene is inhibited in the an additional 3′ cryptic splice target site that results in a
presence of tetracycline (27). Thus, the expression of positive readout, two different sequences can be tested
luciferase in the HeLa Luc/705 cell line can be controlled at simultaneously with one serving as the “internal standard or
the level of transcription and/or the RNA processing by the reference.” In theory, this approach should allow separate
combined activity of tetracycline and the ON-705 oligo- assessment of the relative contributions of uptake versus
nucleotide. Indeed, when the cell line was first incubated hybridization to the target.
for 48 h with 0-100 ng/mL of tetracycline and subsequently The fact that our assay is not applicable to antisense
exposed to 0-0.2 µM ON-705/LipofectAmine complex, the oligonucleotides which promote cleavage of the target RNA
luciferase activity was found to depend on the concentration by RNase H is not a serious limitation. None of the most
of both tetracycline and ON-705 (Figure 4). At 0.2 µM ON- promising 2′-O-alkyl-oligomers, so-called morpholino-oli-
705, the luciferase activity decreased approximately 14-fold gomers (35), methylphosphonates, and the peptide nucleic
as tetracycline concentration was increased from 0 to 100 acids (4) activate RNase H and therefore can be tested in
ng/mL. Conversely, at a given concentration of tetracycline, our system. Finally, but not in the least, in the Tet-Off
the luciferase activity increased in proportion to the oligo- system gene expression is reduced in proportion to the
nucleotide concentration. subcytotoxic concentration of tetracycline and shows the
highest level in the absence of the antibiotic. The ability to
DISCUSSION regulate the pre-mRNA level by the antibiotic provides an
The large body of work on antisense oligonucleotides opportunity to investigate how the concentration of the target
made apparent the importance of the functional assay used affects the requirements of the antisense activity, especially
for assessing their activity. Although the inhibition of cell in terms of the kinetics of the cellular delivery of the
growth was frequently equated with evidence for the down- oligonucleotides. The same subject has been already ad-
regulation of the target gene, the findings that oligonucleo- dressed in the literature for an oligonucleotide which can
tides may nonspecifically bind to proteins and therefore affect interfere with the luciferase expression via the activation of
the cell function by unpredictable mechanisms clearly RNase H (36).
Accelerated Publications Biochemistry, Vol. 37, No. 18, 1998 6239
Clearly the HeLa Luc/705 cell line described in the present 16. Sierakowska, H., Montague, M., Agrawal, S., and Kole, R.
study represents a versatile assay system that allows easy, (1997) Nucleosides Nucleotides 16, 1173-1182.
rapid, and quantitative analysis of the transport involved in 17. Jones, D. H. (1994) PCR Methods Appl. 3, S141-S148.
18. Brasier, A. R., Tate, J. E., and Habner, J. F. (1989) BioTech-
the action of antisense oligonucleotides. The Tet-Off system, niques 7, 1116-1122.
in combination with the pLuc/705 construct described here, 19. Sierakowska, H., Sambade, M. J., Agrawal, S., and Kole, R.
can be easily established in different cell lines (34) extending (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 12840-12844.
the range of the assay to a variety of cell types having 20. Schwartz, E., and Benz, E. J. (1991) in The Thalassemia
Syndrome (Hoffman, R., Benz, E. J., Shattil, S. J., Furie, B.,
different uptake or endocytic characteristics. and Cohen, H. J., Eds.) pp 368-392, Churchill Livingstone
Press, New York.
ACKNOWLEDGMENT 21. Dominski, Z., and Kole, R. (1993) Proc. Natl. Acad. Sci. U.S.A.
The authors thank Dr. Sudhir Agrawal from Hybridon, 90, 8673-8677.
22. Hawley-Nelson, P., Ciccarone, V., Gebeyehu, G., Jessee, J.,
Inc. for the oligonucleotides used in the present study and and Felgner, P. L. (1993) Focus 15, 73-79.
Liz Smith for technical assistance. 23. Sproat, B. S., Lamond, A. I., Beijer, B., Neuner, P., and Ryder,
U. (1989) Nucleic Acids Res. 17, 3373-3386.
REFERENCES 24. Furdon, P. J., Dominski, Z., and Kole, R. (1989) Nucleic Acids
Res. 17, 9193-9204.
1. Agrawal, S. (1996) TIBTECH 14, 376-387. 25. Chen, I. T., and Chasin, L. A. (1993) Mol. Cell. Biol. 13, 289-
2. Crooke, S. T., and Bennett, C. F. (1996) Annu. ReV. Phar- 300.
macol. Toxicol. 36, 107-129. 26. Moore, M. J., Query, C. C., and Sharp, P. A. (1993) in The
3. Maher, L. J. (1996) Cancer InVest. 14, 66-82. RNA World (Gesteland, R. F., and Atkins, J. F., Eds.) pp 303-
4. Nielsen, P. E. (1996) Perspect. Drug DiscoVery Des. 4, 76- 357, Cold Spring Harbor Laboratory Press, New York.
84. 27. Gossen, M., and Bujard, H. (1992) Curr. Opin. Cell Biol. 89,
5. Stein, C. A., and Narayanan, R. (1996) Perspect. Drug 5547-5551.
DiscoVery Des. 4, 41-50. 28. Hodges, D., and Crooke, S. T. (1995) Mol. Pharmacol. 48,
6. Crooke, S. T. (1996) Antisense Nucleic Acid Drug DeV. 6, 905-918.
145-147. 29. Hargrove, J. L., and Schmidt, F. H. (1989) FASEB J. 3, 2360-
7. Stein, C. A. (1996) TIBTECH 14, 147-149. 2370.
8. Branch, A. D. (1996) Hepatology 24, 1517-1529. 30. Ramanathan, M., MacGregor, R. D., and Hunt, C. A. (1993)
9. Vlassov, V. V., Balakireva, L. A., and Yakubov, L. A. (1994) Antisense Res. DeV. 3, 3-18.
Biochim. Biophys. Acta 1197, 95-108. 31. Ross, J. (1995) Microbiol. ReV. 59, 423-450.
10. Agrawal, S., Temsamani, J., Galbraith, W., and Tang, J. (1995) 32. Goldberg, M. W., and Allen, T. D. (1995) Curr. Opin. Cell
Clin. Pharmacokinet. 28, 7-16. Biol. 7, 301-309.
11. Srinivasan, S. K., and Iversen, P. (1995) J. Clin. Lab. Anal. 33. De Wet, J. R., Wood, K. V., DeLuca, M., Helinski, D. R.,
9, 129-137. and Subramani, S. (1987) Mol. Cell. Biol. 7, 725-737.
12. Cho, M. J., and Juliano, R. (1996) TIBTECH 14, 153-158. 34. Yin, D. X., Zhu, L., and Schimke, R. T. (1996) Anal. Biochem.
13. Politz, J. C., Taneja, K. L., and Singer, R. H. (1995) Nucleic 235, 195-201.
Acids Res. 23, 4946-4953. 35. Summerton, J., and Weller, D. (1997) Antisense Nucleic Acid
14. Ho, S. P., Bao, Y., Lesher, T., Malhotra, R., Ma, L. Y., Drug DeV. 7, 187-195.
Fluharty, S. J., and Sakai, R. R. (1998) Nature Biotechnol. 36. Flanagan, W. M., Kothavale, A., and Wagner, R. W. (1996)
16, 59-63. Nucleic Acids Res. 24, 2936-2941.
15. Gewirtz, A. M., Stein, C. A., and Glazer, P. M. (1996) Proc.
Natl. Acad. Sci. U.S.A. 93, 3161-3163. BI980300H