Vel. 267, No. 18, Issue of June 25, pp. 12436-12439,1992 THE JOURNAL OF BIOLOGICAL CHEMISTRY

0 1992 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U. S A .

Specific Inhibition of Hepatitis B Viral Gene Expression in Vitro by Targeted Antisense Oligonucleotides*

(Received for publication, January 24, 1992) George Y. Wu* and Catherine H. Wu From the Division of Gastroenterology-Hepatology, Department of Medicine, Uniuersity of Connecticut School of Medicine, Furmington, Connecticut 06030

A 2 1-mer oligodeoxynucleotide complementary to the polyadenylation signal for human hepatitis B virus (HBV) was complexed to a soluble DNA-carrier system that is targetable to hepatocytes via asialoglycoprotein receptors present on those cells. A cell line, HepG2 (2.2.16) that possesses asialoglycoprotein receptors and is permanently transfected with hepatitis B virus (ayw subtype) was exposed to complexed antisense DNA or controls. In the presence of complexed anti- sense DNA, the concentration of hepatitis B surface antigen in medium was 80% lower than controls after 2 4 h. Furthermore, during the next 6 days, there was no significant increase in surface antigen concentra- tion in the presence of complexed antisense DNA. The inhibition could be effectively blocked by competition with an excess of free asialoglycoprotein. Total protein synthesis remained unchanged by exposure to com- plexed antisense sequences under identical conditions. In addition, HBV DNA in the medium and cell layers after 24-h exposure to complexed antisense sequences was 80% lower than in controls. The data indicate that antisense oligonucleotides complexed by a soluble DNA-carrier system can be targeted to cells via asial- oglycoprotein receptors resulting in specific inhibition of hepatitis B viral gene expression and replication.

We have previously shown that DNA in the form of bacte- rial plasmids can be targeted to hepatocytes resulting in foreign gene expression in vitro (1) and in vivo (2). This was accomplished through the use of a DNA-carrier system com- posed of: 1) a cell-targeting component consisting of a galac- tose-terminal (asialo-) glycoprotein capable of being recog- nized and internalized by specialized receptors on hepatocytes (3); and 2) a polycation, poly-L-lysine, which can bind DNA in a strong but non-covalent manner based on electrostatic attraction for the negative charges of DNA. Mixing of the carrier with DNA formed a complex that could be recognized by asialoglycoprotein receptors on hepatocytes. In these pre-

Grants DK-42182 (to G. Y. W.) and CA-01110 (to G. Y. W.), March * This work was supported in part by National Institutes of Health

of Dimes Grant 1-1256 (to G. Y. W.), and by a grant from TargeTech, Inc. (to C. H. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$To whom correspondence should be addressed Rm. AM-044, Dept. of Medicine, Division of Gastroenterology-Hepatology, Univer- sity of Connecticut School of Medicine, 263 Farmington Ave., Far- mington, CT 06030. Tel.: 203-679-3158; Fax: 203-679-3159.

vious studies, double-stranded DNA was used. However, in theory, the targetable DNA-carrier could also be used to deliver single-stranded DNA. This is of interest, because it has been shown, previously, that exposure of cultured cells to single-stranded DNA as antisense sequences can result in the inhibition of synthesis of a specific protein by hybridization of the antisense oligonucleotide to a specific complementary sequence present in mRNA (4).

MATERIALS AND METHODS

Cells and Cell Culture-Human hepatoma, HepG2 2.2.15 cells kindly provided by Dr. George Acs (Mt. Sinai School of Medicine, NY) and SK Hepl cells were grown in Dulbecco's modified Eagle's medium and 10% fetal calf serum as described previously (1).

Preparation of Targetable Antisense DNA-A targetable, soluble DNA carrier was prepared by coupling asialoorosomucoid to poly-L- lysine (Sigma) (M, = 59,000) using l-ethyl-3-(3-dimethylaminopro- py1)carbodiimide (Pierce) as described previously (1) except that the conjugate was purified by cation exchange chromatography using a high pressure liquid chromatographic system (Rainin) employing an Aquapore C-300 column (Rainin) and stepwise elution with 0.1 M sodium acetate, pH 5.0, 2.5, 2.25, and 2.0. The second peak eluted from the column, as detected by UV absorption at 230 nm, was determined to be the optimal conjugate form and was used for all subsequent experiments. A 21-mer oligodeoxynucleotide, complemen- tary to a portion of the human hepatitis B virus (ayw subtype) (5) including the polyadenylation signal, corresponding to nucleotides 1903-1923 of the viral genome, was synthesized on an automated nucleotide synthesizer (Applied Biosystems) (6) using phosphoro- thioate linkages. As a control, a random 21-mer sequence was pre- pared in an identical fashion. The purity of oligonucleotides was determined by electrophoresis through 15% polyacrylamide gels stained with ethidium bromide. Antisense DNA was titrated with conjugate to form a soluble complex using an agarose gel retardation system as described previously (l), and a conjugate to DNA ratio of 1.6:l by weight (asialoorosomucoidDNA) was selected.

Assay for Receptor-mediated Uptake of Complened Antisense DNA-To evaluate uptake of oligonucleotides, antisense DNA was end-labeled with 'lP (7). DNA alone or in the form of a complex was added to the medium of HepG2 and SK Hepl cells to make 50 pM solutions with respect to added antisense DNA. Uptake was deter- mined as described previously for asialoglycoproteins (8). In brief, cells were incubated with ligands at 37 "C, and at regular time intervals, dishes were chilled to 4 "C, washed with cold 10 mM EDTA- phosphate-buffered saline, and the cell layers removed, and scintil- lation counted. To determine counts bound to the cell surface, iden- tical sets of cells were also incubated at 4 "C with ligands and, after washing as described above, the cell layers were removed, and adher- ent radioactivity was determined by scintillation counting. Uptake was calculated as the difference between total cell-associated counts at 37 "C and counts bound to the cell surface at 4 "C for each time point. All points were determined in triplicate, and results are shown as means & S.E. expressed as picomoles of DNA/106 cells.

Antisense DNA and Viral Gene Expression-To determine the effect of antisense DNA on viral gene expression, HepG2 2.2.15 cells were seeded 6 days preconfluence and incubated at 37 "C in medium containing antisense DNA alone, complexed antisense DNA, com- plexed random DNA, or medium alone. All media containing added DNA were initially 50 ,.LM with respect to DNA. At daily intervals, 50 pl of medium was sampled and assayed for hepatitis B surface antigen by an ELISA' (Abbott) method as described by the manufacturer, modified for quantitation using serially diluted standard surface antigen (Merck) which produced a linear response within the range of antigen levels found in the samples. Cell number was determined by microscopically counting cells stained with trypan blue. All points were determined in triplicate, and the results of four experiments are

' The abbreviations used are: ELISA, enzyme-linked immunosor- bent assay; HBV, hepatitis B virus.

12436

Targeted Antisense DNA against Hepatitis B Virus 12437

shown as means k S.E. expressed as pg/rnl/lO6 cells. Effect of Antisense DNA on Protein Secretion"HepG2 2.2.15 cells

were incubated with 50 FM complexed DNA as described in Fig. 2 except that 10 pCi of [35S]methionine (Amersham Corp.), specific activity 1000 Ci/mrnol, was added to label newly synthesized proteins. After 24 h, medium was moved, and cells were washed with phosphate- buffered saline and lysed with 1% sodium dodecyl sulfate which was subsequently removed with Triton X-100. Both media and cell lysates were treated with a specific rabbit anti-surface antigen antibody (DAKO) and precipitated with protein A-Sepharose (Sigma). Precip- itates were scintillation counted, and each point was assayed in triplicate. Total cell protein was determined by colorimetric assay (Bio-Rad). The results of three experiments are shown as means & S.E. expressed as cprn/mg of cell protein.

In order to assess the effect of complexed antisense DNA on total protein synthesis, cells treated with complexed DNA and labeled with ['"SS]methionine as described above were separated from media. Total protein was precipitated with 10% trichloroacetic acid and counted. The results of triplicate assays of three experiments are shown as means & S.E. expressed as cpm/mg of cell protein.

Effect of Antisense DNA on Production of Viral DNA-Cells were incubated with 50 PM antisense DNA alone or in the form of com- plexes. After 24 h, medium was removed, and HBV DNA was ex- tracted from the medium according to the method of Sells et al. (9), and Eom the cell layer according to the method of Hirt (10). Total cell protein was determined by colorimetric assay (Bio-Rad). DNA extracted from equal volumes (40 ml) of medium and from approxi- mately equal numbers of cells (10') were applied on an agarose gel. HBV DNA was identified by Southern blot using an EcoRI-BglII fragment of the HBV genome (nucleotide 0-1982) as a probe labeled with "P and exposed to x-ray film as described previously (11). Relative quantitation was achieved by densitometry and confirmed by scintillation counting of corresponding bands.

RESULTS AND DISCUSSION

The first objective was to determine whether a single- stranded DNA oligonucleotide could be bound by an asialo- glycoprotein-based carrier and whether it could be specifically targeted to asialoglycoprotein receptor-bearing cells.

Fig. 1 shows that in SK Hepl (asialoglycoprotein receptor (-) cells), the average uptake of antisense DNA alone was less than 0.05 pmol/h/million cells over 4 h of exposure. Incubation with DNA in the form of a complex did not improve the uptake in these cells. Similarly, uptake of anti- sense DNA alone by HepG2 (asialoglycoprotein receptor (+)) cells was as low as observed in receptor (-) cells over the course of the 4 h. However, in the receptor (+) cells, complexed antisense DNA was taken up nearly linearly at an average rate 12 times faster than antisense DNA alone through the 4 h of incubation. Accumulation of complexed antisense was also 12 times greater than antisense DNA alone after 4 h of exposure. The uptake of complexed DNA was virtually com- pletely blocked by administration of a large molar excess of free carrier protein, asialoorosomucoid, to compete for recep- tor binding. This confirmed the involvement of asialoglyco- protein receptors in the differential delivery of the antisense DNA.

To determine whether the targeted antisense DNA was functional, effects on hepatitis B viral gene expression were evaluated. Fig. 2 shows that in untreated control cells, hepa- titis B viral surface antigen steadily increased in concentra- tion in their media, rising from 1 Fg/ml/106 cells on the 1st day to 5.5 pg/rn1/1O6 cells by the 7th day. Exposure of cells to antisense DNA alone had no significant effect until the 3rd day at which time surface antigen concentration was 30% lower than untreated controls. Nevertheless, surface antigen concentrations in the presence of antisense DNA alone con- tinued to rise steadily throughout the 7 days of exposure. However, treatment with complexed antisense DNA resulted in an 80% inhibition after the 1st day and 95% inhibition by the 7th day compared with untreated controls. There was no

I /

2 3 ./ 0 2

Time [houral Time ihounl

FIG. 1. Uptake of complexed antisense DNA by HepG2 and SK Hepl cells. A targetable, soluble DNA carrier was prepared by coupling asialoorosomucoid to poly-L-lysine using a water-soluble carbodiimide as described previously (1) and purified as described under "Materials and Methods." The conjugate was complexed to a 21-mer oligodeoxynucleotide, complementary to a portion of the human hepatitis B virus including the polyadenylation signal, and labeled with '"P. DNA alone, or in the form of a complex, was added to medium to make it 50 PM with respect to added antisense DNA. Uptake was determined as described previously for asialoglycopro- teins (8). Cells were incubated at 37 "C, and at regular time intervals, dishes were chilled, washed with 10 mM EDTA-phosphate-buffered saline. The cell layer was removed, mixed with scintillant, and counted on a scintillation counter. To determine counts bound to the cell surface, identical sets of cells were also incubated at 4 "C with ligands and, after washing as described above, the cell layers were removed, and radioactivity was determined by scintillation counting. Uptake was calculated as the difference between cell-associated counts at 37 "C and counts bound to the cell surface at 4 "C for each time point. All points were determined in triplicate, and results are shown as means k S.E. expressed as picomoles of DNA/106 cells. 0, antisense DNA alone; +, complexed antisense DNA A, complexed antisense DNA plus a 100-fold molar excess of asialoorosomucoid.

significant increase in surface antigen concentration after the first 24 h. Complexed random DNA of the same size had no effect on antigen concentrations a t any time point under identical conditions.

The lack of accumulation of hepatitis B surface antigen in the medium of cells treated with complexed antisense DNA could have been due to a block in protein secretion. To examine this possibility, the synthesis of new surface antigen was measured in the media and cell layers. Table I, A, shows that radiolabeled immunoprecipitable surface antigen in both the medium and cells after treatment with complexed anti- sense DNA for 24 h were decreased to an equal extent (80%) compared with untreated cells. There was no significant in- tracellular accumulation of newly synthesized antigen that would have been expected if a block in protein secretion had occurred.

Table I, B, shows that neither complexed antisense DNA nor random DNA had a significant effect on total newly synthesized protein in the cell layer. A small, 2% decrease was detected in newly synthesized secreted protein in cells exposed to complexed antisense DNA. This likely reflects the contribution of the inhibition of viral surface antigen synthe- sis noted in Table I, A. The data, overall, indicate that the observed inhibition of hepatitis B surface antigen synthesis by complexed antisense DNA was specific and could not have been due to a generalized inhibition of total protein synthesis.

Finally, in order to determine whether viral replication was affected, DNA was extracted from the medium, and cell layers were exposed to oligonucleotides for 24 h and analyzed by Southern blots. Fig. 3, lanes I and 5, shows that untreated cells produced bands at positions expected for relaxed circular and single-stranded linear viral replicative DNA forms. Other minor bands are present, a t 2.3 kilobases for example, as described previously for this cell line (9). Lanes 2 and 6 show

12438 7

6

Targeted Antisense DNA against Hepatitis B Virus

kb - untreated control * fmtlsense DNA Alonb - Complexed Random DNA - - - -. - - - Complexed Antisense DNA

T

Y """""." """.""""."""".""""I-"""-~ ..- , 1 2 3 4 5 6 7

Time (days)

FIG. 2. Effect of complexed antisense DNA on hepatitis B virus surface antigen concentration in culture medium. HepG2 2.2.15 cells were incubated at 37 "C in medium containing antisense DNA alone, complexed antisense DNA, complexed random DNA, or medium alone. All media containing added DNA were initially 50 pM with respect to DNA. A t daily intervals, medium was sampled and assayed for the presence of hepatitis B surface antigen by an ELISA (Abbott) method as described by the manufacturer, modified as described under "Materials and Methods." Cell number was deter- mined by microscopically counting cells stained with trypan blue. All points were determined in triplicate, and the results of four experi- ments are shown as means f S.E. expressed as pg/ml/106 cells.

TABLE I Effect of antisense DNA on protein synthesis in HepG2 2.2.15 cells

Treatment Cell layer' Cell medium" A. Immunoprecipitable hepatitis B surface antigenb

Untreated control 56,100 f 2,321 114,500 f 2,442 Complexed antisense DNA 10,200 k 1,009 15,300 2 890 Comdexed random DNA 52.500 f 4.534 122.220 f 5,742

B. Total TCA-precipitable radioactivity'

Untreated control 184,498 2 2,258 712,498 f 5,435 Complexed antisense DNA 188,844 f 6,240 684,302 f 9,678 Complexed random DNA 183,591 f 5,444 706,240 f 7,544

cpm/mg cell protein. I, After 24 h of incubation. ' TCA, trichloroacetic acid.

that treatment of cells with complexed antisense DNA de- creased the amount of all viral DNA forms in the medium by approximately 80% compared with untreated cells (lanes 1 and 5 ) . Complexed random DNA, lanes 3 and 7, had no detectable effect on the levels of HBV DNA under identical conditions. Antisense DNA alone, lanes 4 and 8, decreased HBV DNA by approximately 30% relative to untreated con- trols. The number of viable cells as determined by trypan blue exclusion was not affected by treatment with any form of DNA (data not shown).

It has been shown previously that many cell types are capable of taking up free oligonucleotides (12). Loke et dl. (12) showed that the rates of uptake were inversely propor-

4.0 - w

2.0 -

1.3 -

FIG. 3. Southern blots of DNA extracted from medium and cells after 24 h of exposure to antisense DNA. Cells were incubated as described in Fig. 2 with 50 PM antisense DNA alone or in the form of complexes as described in Fig. 2. After 24 h, medium was removed and DNA was extracted from the medium (9) and from the cell layer (10). Total cell protein was determined by colorimetric assay (Bio-Rad). DNA extracted from equal volumes of medium and from approximately equal numbers of cells were applied on an agarose gel. HBV DNA was identified by Southern blot using an EcoRI-BglII fragment of the HBV genome as a probe labeled with '*P and exposed to x-ray film (11). Relative quantitation was achieved by densitometry and confirmed by scintillation counting of corresponding bands, normalized to equal volume or cell number, for media and cell layers, respectively. Duplicate blots were performed, a representative of which is shown above. Lanes 1-4, cell lysates; lanes 5-8, media; lanes 1 and 5, untreated controls; lanes 2 and 6, treated with complexed antisense DNA; lunes 3 and 7, treated with complexed random DNA; and lanes 4 and 8, treated with antisense DNA alone. Expected positions for relaxed circular (RC), and single-stranded (SS) forms are indicated on the right.

tional to the size of the oligonucleotide. However, in general, the longer the DNA sequence, the greater the specificity for target mRNA molecules. These conflicting properties illus- trate two common problems with the current use of antisense oligonucleotides: inefficient uptake and lack of cell specificity. In order to improve uptake of antisense oligonucleotides, Lemaitre et al. (13) covalently coupled an oligonucleotide to polylysine and obtained antiviral effects at severalfold lower concentrations than could be obtained with free DNA. How- ever, the delivery was not cell-specific. Our uptake data indi- cate that not only can transport of oligonucleotides into cells be greatly enhanced, but the uptake can also be directed to specific cells mediated by an asialoglycoprotein-based DNA- carrier system.

Because of the specificity of DNA hybridization with mRNA to form hybrids implicated in antisense-mediated inhibition of translation, antisense oligonucleotides have been used successfully to study normal gene expression in vitro (14). For similar reasons, antisense oligonucleotides have also been examined previously for antiviral effects (15, 16). For example, Agrawal et al. (16) administered infectious virus (HIV) together with antisense oligonucleotides in a non- targeted manner to cell media. Specific inhibition of viral replication was demonstrated. Similarly, Lemaitre et dl. (13) studied a model of acute viral infection in which antisense was preadministered to cells with substantial specific antiviral effects. Our experiments differ from these previous studies in that our cells had a pre-existing, stable viral infection with

Targeted Antisense DNA against Hepatitis B Virus 12439

viral production maintained by an integrated viral genome. Our data indicate that although a stable infection existed, delivery of antisense oligonucleotides can dramatically inhibit viral gene expression in a specific manner. However, in uiuo, persistent production of hepatitis B virus is usually due to the presence of unintegrated viral DNA (17). Integration of the viral genome into that of the host is usually associated with a cessation of production of complete viral particles (18). Whether targeted antisense delivery can be effective in the presence of an infection generated by unintegrated viral DNA remains to be seen. However, asialoglycoprotein uptake in hepatitis virus-infected HepGP cells was found to be not substantially different from non-infected HepG2 cells (data not shown), indicating that infection by the virus did not alter the receptor activity in these cells. This suggests that targeted delivery of antisense oligonucleotides may be generally appli- cable to naturally infected hepatocytes, which are otherwise normal, via asialoglycoprotein receptors.

Finally, it should be noted that the oligonucleotides used in this current work were linked together by phosphorothioate bonds. These linkages are less susceptible to nuclease degra- daticl than normal phosphodiester bonds. However, a variety of other synthetic strategies have been developed to confer nuclease resistance to antisense oligonucleotides (19). Forms that retain polyanionic character may also be deliverable by a receptor-mediated carrier system to provide enhanced and prolonged efficacy in a targeted manner.

Acknowledgments-We thank Dr. George Acs for kindly providing us with the HepG2 2.2.15 cells and Brenda Kawecki for assistance in the preparation of the manuscript.

REFERENCES 1. Wu, G. Y., and Wu, C. H. (1987) J. Biol. Chem. 262,4429-4432 2. Wu, G. Y., and Wu, C. H. (1988) J. Biol. Chern. 263, 14621-

14624

3. Ashwell, G., and Morell, A. G. (1974) Adu. Enzymol. Relat. Areas

4. Stein, C. A., and Cohen, J. S. (1988) Cancer Res. 48, 2659-2668 5. Hirschman, S. Z., Price, P., Garfinkel, E., Christman, J., and Acs,

G. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,5507-5511 6. Matsukura, M., Shinozuka, K., Zon, G., Mitsuya, H., Reitz, M.,

Cohen, J. S., and Broder, S. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 7705-7710

7. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laborat~ory Manual, 2nd Ed., Vol. 2, p. 11.31, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

8. Schwartz, A. L., Fridovich, S. E., Knowles, B. B., and Lodish, H. F. (1981) J. Biol. Chem. 256,8878-8881

9. Sells, M. A,, Chen, M. L., and Acs, G. (1987) PFOC. Natl. Acad. Sci. U. S. A . 84, 1005-1009

Mol. Biol. 41, 99-128

10. Hirt, B. (1967) J. Mol. Biol. 26, 365-371 11. Sambrook. J.. Fritsch. E. F.. and Maniatis. T. (1989) Molecular

Cloning:'A Laboratory Manual, 2nd Ed., Vol. 2, pp. 10.14-10.15, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

2. Loke, S. L., Stein, C. A., Zhang, X. H., Mori, K., Nakanishi, M., Subasinghe, C., Cohen, J. S., and Neckers, L. M. (1989) Proc.

3. Lemaitre, M., Bayard, B., and Lebleu, B. (1987) Proc. Natl. Acad.

4. Bevilaqua, A., Erikson, A. P., and Hieber, V. (1988) Proc. Natl.

Natl. Acad. Sci. U. S. A . 86, 3474-3478

Sci. U. S. A . 84, 648-652

1

1

1 Acadl Sci. U. S. A . 85, 831-835

15. Goodchild, J., Agrawal, S., Civeira, M. P., Sarin, P. S., Sun, D., and Zamecnik, P. C. (1988) PFOC. Natl. Acad. Sci. U. S. A . 85,

16. Agrawal, S., Ikeuchi, T., Sun, D., Sarin, P. S., Konopka, A., Maizel, J., and Zamecnik, P. C. (1989) PFOC. Natl. Acad. Sci.

17. Shafritz, D., Shouval, D., Sherman, H. I., Hadziyannis, S. J., and

18. Ganem, D. (1982) Reu. Infect. Dis. 4, 1026-1047 19. Miller, P. S., Agris, C. H., Aurelian, L., Blake, K. R., Murakami,

A., Reddy, M. P., Spitz, A. A., and Ts'o, P. 0. P. (1985) Nucleosides & Nucleotides 6, 769-776

5507-5511

U. S. A. 86, 7790-7794

Kew, M. C. (1981) N . Engl. J. Med. 305, 1067-1073