- Email: [email protected]
Antisense phosphorothioates as antivirals against human immunodeficiency virus (HIV) and hepatitis B virus (HBV)
Toxicology Letters ELSEVIER
Toxicology Letters 82/83 (1995) 435-438
Antisense phosphorothioates as antivirals against human immunodeficiency virus (HIV) and hepatitis B virus (HBV) Makoto Matsukura”‘“,
Kazuo Koikeb, Gerald Zen”
“Department of Chdd Development, Kumamoto University School of Medicine, l-l-l Honjo, Kumamoto, blst Department of Internal Medicine, Tokyo Unwersity Faculty of Medicine, Tokyo, Japan ‘Lynx Therapeutics. USA
860 Japan
Abstract For the past 10 years we have focused on the development of antisense treatments against viral infections such as human immunodeficiency virus (HIV) and hepatitis B virus (HBV). We have demonstrated that phosphorothioate oligomers have in vitro anti-HIV activities and in vivo anti-HBV activity. In vitro anti-HIV activities of phosphorothioate oligodeoxynucleotides (S-oligo) were classified into sequence non-specific (non-antisense) and sequence specific (antisense). We found that toxicity of S-oligo could be also classified into the same 2 categories. Since some of non-specific toxicity is dependent on experimental conditions, in each experiment, we should pay attention to minimize non-specific toxicity of the oligomers.
Keywords: Antisense
DNA; Antivirals;
HIV, HBV
1. Materials and methods 1.1. Human
immunodeficiency
virus (HIV)
In vitro assay systems we used were (1) cytopathic effect inhibition assay and (2) viral gene expression inhibition assay. In the first assay we used the HIV-uninfected cell line, ATH8, which is transformed and immortalized with the infection of human T-lymphotropic Virus Type I (HTLV-I). ATH8 cells are very sensitive to the cytopathic effect of HIV and most of cells die within a week after exposure to HIV In the
* Corresponding
author.
second assay, we used human T cell line, H9, which is sensitive to the infection of HIV but can survive and become chronically HIV-infected H9 cells.
1.1.1. Cytopathic
effect inhibition
assay
ATH8 cells were centrifuged, pelleted and exposed to HIV,,,, for 1 h (500 virions per cell). Complete medium, RPM11640 containing 15% fetal calf serum (FCS) and recombinant interleukin-2 (20 units per ml) were used with various concentration of the oligomers added. The number of viable cells was counted in a hemocytometer using the trypan blue dye-exclusion method on day 7 following exposure to the virus.
0378-4274/95/$09.50 @ 1995 Elsevier Science Ireland Ltd. All rights reserved SSDl 0378-4274(95)03574-S
436
M. Matsukura et al. I Toxicology Letters 82183 (1995) 435-438
1.1.2. Viral gene expression
inhibition assay
To test whether an oligomer has antisense activity on HIV viral expression, we used chronically HIV-l-infected H9 cells (H9/111,). H9/111, cells were washed extensively to remove the previously produced virions from the medium. After washing, 1250 H9/111, cells per well in a 96-well culture plate were cultured in the presence or absence of oligomers in 200 ,ul of the complete medium (RPM11640 + 15% FCS). Culture supernatants were collected on day 5 and assayed for p24 gag protein by using RIA or ELISA. The cells were pulsed with [3H]thymidine and harvested next day to assess cytotoxicities of oligomers. 1.1.3. Reverse transcriptase (RT) assay We reported previously the RT assay [l] used and describe it only briefly. Oligo(dT)-poly (rA) was used as primer-template in the presence of [3H]thymidine triphosphate. As a control, oligo(dT)-poly(dA) was used. Phosphorothioate oligodeoxynucleotides (S-oligo) or n-oligo at various concentrations was added to the assay system. 1.1.4. Oligodeoxynucleotides used We employed the initiation site of rev as the target site of HIV gene with reasons such as conservativeness of the sequence and importance in the viral replication. As controls, S-oligo with sense and random sequence and n-oligo with antisense sequence were tested. 1.2. Hepatitis B virus (HBV) HBV has been know to be a very difficult virus for in vitro infection studies. We therefore employed an animal model using HBx transgenic mice. In HBx transgenic mice, histological precancerous changes, including vacuolation and altered foci in the liver, were followed by the development of hepatocellular carcinoma at 1 year of age or older. Because of the insertion of the HBV transactivator gene, HBx, these precancerous changes occurred.
1.2.1. Transgenic mice
Production of HBx gene transgenic mice was reported previously [2]. We used male mice from the H9 strain of HBx transgenic mouse which was homozygous for HBx gene. Mice were maintained and cared for in accordance with the guidelines established by the National Institutes of Health. 1.2.2. Oligodeoxynucleotides used Two sets of sense and antisense S-oligos were designed from the standpoint of base composition (rich in CC) and/or secondary stucture. One set, S-l (sense) and AS-1 (antisense), was aimed to the GC-rich region of HBx gene and a second set, S-2 and AS-2, was designed to cover the initiation site. Both oligomers were 27-mer and synthesized in the amount of hundreds of milligrams. Sequences of the oligomers will be published elsewhere [3]. 1.2.3. RNA extraction, RT-polymerase chain reaction (PCR) and histological studies of tissue RNA was extracted from liver tissues as described by Chomczynski and Sacchi [4]. RNA was analyzed by RT-PCR using a set of primers [3]. Tissue sections fixed in 10% neutral-buffered formalin were used for hematoxylin and eosin staining.
2. Results 2.1. Anti-HIV activities
We found that S-oligo could exhibit 2 different anti-HIV activities. One is sequence-non-specific and another is sequence-specific anti-HIV activity. Sequence-non-specific activity was found in the cytopathic effect inhibition assay where even homo-oligomer S-oligo could inhibit de novo infection (no HIV DNA synthesis) possibly with multiple mechanisms including inhibition of HIV RT. On the other hand, the result obtained from the viral gene expression assay was sequencespecific anti-HIV activity, namely antisense activity. In the enzyme assay for HIV RT, S-oligo
M. Matsukura
et al. I Toxicology
showed much more potent inhibition compared to n-oligo. Such inhibition, however, could be reversed by addition of bovine serum albumin (BSA) in the assay solution suggesting that the inhibition of RT activity by S-oligo is a result of non-specific protein binding. Similarly, increase of percentage of FCS in the complete medium eliminated the cytotoxicity of S-oligo whereas a higher percentage than 30% of FCS is somehow toxic and, under such conditions, FCS did not compete with the cytotoxicity of S-oligo (data not shown). In the viral gene expression assay, n-oligo showed a profound decrease of [“Hlthymidine uptake without decrease of viable cell number. This pseudo-cytotoxicity was derived from competition of non-radioactive thymidine from n-oligo degraded by nucleases in the FCS. Soligo is known to be nuclease-resistant, but could be degraded very slowly in culture media and also in vivo. One should take account of not only the oligomers but also the metabolites for toxicity* 2.2. Anti-HBV activity The data from in vivo experiments using HBx transgenic mice will be published elsewhere. Therefore, we describe the results here briefly without figures and tables. 2.2.1. Down-regulation of HBx gene expression As an initial trial, we injected 1 of 4 S-oligos i.p. into the mice for 7 consecutive days (1 mg/ day) at the age of 3 month old. On the last day of injection, mice were sacrificed and gene expression of HBx in liver was examined by means of the RT-PCR. We found a significant decrease of HBx gene expression in AS-Ztreated mice whereas other oligomers did not cause a decrease (data not shown). 22.2. Long-term treatment with AS-2 We started the treatments before the mice were 3 months old. At 1 week old, even 0.2 mg/day injection 3 times a week was found to be highly toxic to mice, which resulted in death at 4 weeks old. Then we delayed the initiation of the
Letters 82183 (199-Y) 435-438
437
treatment to 2 weeks old and reduced the dosage. With this protocol, the treatment was continued until the mice were 10 weeks old. Significant reduction of the expression and histologically milder changes were found in mice treated with AS-2 but not in mice treated with S-2 or PBS. In a series of treatments, cell infiltration into liver was found in mice treated with AS-1 and S-l suggesting that toxic reaction could be sequence specific or base composition of oligomers (GC-rich) (data not shown).
3. Discussion Toxicological consideration of oligomers could not be obtained without knowing more about the characteristics of the experiments. We should compare experiments with each other in terms of their advantages and disadvantages. In vitro assay systems which we used have advantages over in vivo experiments in the following ways: (1) a relatively small amount of oligomer is required; (2) because of relative ease, many samples including several controls can be tested simultaneously; (3) because of relatively defined and short-term assay conditions, experiments can be repeatable in a short time and the data obtained can be reproduced. There are also significant disadvantages as follows: (1) in vitro assay systems have artificial conditions such as use of particular immortalized cell lines, artificially conditioned medium containing FCS, relatively rapid cell growth (possibly different sensitivity to cytotoxicity of compounds) and/or lack of heterogenous cell populations; (2) a larger or smaller amount of viral production in comparison to in vivo situations; (3) lack of pharmacokinetics and tissue distribution. Since in vivo HIV experiments using animal models are neither definitive nor easy, we have been performing in vitro assays for detection of antisense activity. We have found at least several mechanisms of in vitro anti-HIV activities, which were initially confusing and made the antisense strategy perhaps less appealing to some of the researchers. Several mechanisms in anti-HIV activity might be also true in toxicity of oligo-
438
M. Matsukura
et al. I Toxicology
mers (at least 2: sequence-non-specific and sequence-specific toxicity). In monocytes, S-oligo could introduce superoxide (unpublished result), which could be partially reduced by indomethacin. S-oligo could also introduce the viral expression of HIV in dormant HIV-infected cells (unpublished data). Although suitable in vitro assays should be generally employed for antisense studies at the start, in vivo experiments using animal models could sometimes be better in sequence specificity and interpretability of data including toxicity. If enough oligomer is available, one might want to perform in vivo experiments without in vitro assays.
References [l] Mitsuya, H., Jarrett, F.D.M., Sarngadharan,
R.F., Matsukura, M., Veronese, M.G., Johns, D.G., Reitz, M. and
Letters 82183 (1995) 435-438 Broder, S. (1987) Long-term inhibition of HTLV-IIIILAV (human immunodeficiency virus) DNA synthesis and RNA expression in T-cells protected by 2’,3’-dideoxynucloesides in vitro. Proc. Nat]. Acad. Sci. USA 84, 20332037. 121Kim, C.-M., Koike, K., Saito, I., Miyamura, T. and Jay, G. (1991) HBx gene of hepatitis B virus induced liver cancer in transgenic mice. Nature 351, 317-320. M., Kurokawa, K. and Koike. K. [31Moriya, K., Matsukura, In vivo inhibition of hepatitis B virus gene expression by antisense phosphorothioate oligodeoxynucleotides. (Submitted.) P. and Sacchi, N. (1987) Single-step meth[41Chomczynski, od of RNA isolation by acid guanidinium thiocyanatephenol-chloroform extraction. Anal. Biochem. 162. 156159.
Toxicology Letters 82/83 (1995) 435-438
Antisense phosphorothioates as antivirals against human immunodeficiency virus (HIV) and hepatitis B virus (HBV) Makoto Matsukura”‘“,
Kazuo Koikeb, Gerald Zen”
“Department of Chdd Development, Kumamoto University School of Medicine, l-l-l Honjo, Kumamoto, blst Department of Internal Medicine, Tokyo Unwersity Faculty of Medicine, Tokyo, Japan ‘Lynx Therapeutics. USA
860 Japan
Abstract For the past 10 years we have focused on the development of antisense treatments against viral infections such as human immunodeficiency virus (HIV) and hepatitis B virus (HBV). We have demonstrated that phosphorothioate oligomers have in vitro anti-HIV activities and in vivo anti-HBV activity. In vitro anti-HIV activities of phosphorothioate oligodeoxynucleotides (S-oligo) were classified into sequence non-specific (non-antisense) and sequence specific (antisense). We found that toxicity of S-oligo could be also classified into the same 2 categories. Since some of non-specific toxicity is dependent on experimental conditions, in each experiment, we should pay attention to minimize non-specific toxicity of the oligomers.
Keywords: Antisense
DNA; Antivirals;
HIV, HBV
1. Materials and methods 1.1. Human
immunodeficiency
virus (HIV)
In vitro assay systems we used were (1) cytopathic effect inhibition assay and (2) viral gene expression inhibition assay. In the first assay we used the HIV-uninfected cell line, ATH8, which is transformed and immortalized with the infection of human T-lymphotropic Virus Type I (HTLV-I). ATH8 cells are very sensitive to the cytopathic effect of HIV and most of cells die within a week after exposure to HIV In the
* Corresponding
author.
second assay, we used human T cell line, H9, which is sensitive to the infection of HIV but can survive and become chronically HIV-infected H9 cells.
1.1.1. Cytopathic
effect inhibition
assay
ATH8 cells were centrifuged, pelleted and exposed to HIV,,,, for 1 h (500 virions per cell). Complete medium, RPM11640 containing 15% fetal calf serum (FCS) and recombinant interleukin-2 (20 units per ml) were used with various concentration of the oligomers added. The number of viable cells was counted in a hemocytometer using the trypan blue dye-exclusion method on day 7 following exposure to the virus.
0378-4274/95/$09.50 @ 1995 Elsevier Science Ireland Ltd. All rights reserved SSDl 0378-4274(95)03574-S
436
M. Matsukura et al. I Toxicology Letters 82183 (1995) 435-438
1.1.2. Viral gene expression
inhibition assay
To test whether an oligomer has antisense activity on HIV viral expression, we used chronically HIV-l-infected H9 cells (H9/111,). H9/111, cells were washed extensively to remove the previously produced virions from the medium. After washing, 1250 H9/111, cells per well in a 96-well culture plate were cultured in the presence or absence of oligomers in 200 ,ul of the complete medium (RPM11640 + 15% FCS). Culture supernatants were collected on day 5 and assayed for p24 gag protein by using RIA or ELISA. The cells were pulsed with [3H]thymidine and harvested next day to assess cytotoxicities of oligomers. 1.1.3. Reverse transcriptase (RT) assay We reported previously the RT assay [l] used and describe it only briefly. Oligo(dT)-poly (rA) was used as primer-template in the presence of [3H]thymidine triphosphate. As a control, oligo(dT)-poly(dA) was used. Phosphorothioate oligodeoxynucleotides (S-oligo) or n-oligo at various concentrations was added to the assay system. 1.1.4. Oligodeoxynucleotides used We employed the initiation site of rev as the target site of HIV gene with reasons such as conservativeness of the sequence and importance in the viral replication. As controls, S-oligo with sense and random sequence and n-oligo with antisense sequence were tested. 1.2. Hepatitis B virus (HBV) HBV has been know to be a very difficult virus for in vitro infection studies. We therefore employed an animal model using HBx transgenic mice. In HBx transgenic mice, histological precancerous changes, including vacuolation and altered foci in the liver, were followed by the development of hepatocellular carcinoma at 1 year of age or older. Because of the insertion of the HBV transactivator gene, HBx, these precancerous changes occurred.
1.2.1. Transgenic mice
Production of HBx gene transgenic mice was reported previously [2]. We used male mice from the H9 strain of HBx transgenic mouse which was homozygous for HBx gene. Mice were maintained and cared for in accordance with the guidelines established by the National Institutes of Health. 1.2.2. Oligodeoxynucleotides used Two sets of sense and antisense S-oligos were designed from the standpoint of base composition (rich in CC) and/or secondary stucture. One set, S-l (sense) and AS-1 (antisense), was aimed to the GC-rich region of HBx gene and a second set, S-2 and AS-2, was designed to cover the initiation site. Both oligomers were 27-mer and synthesized in the amount of hundreds of milligrams. Sequences of the oligomers will be published elsewhere [3]. 1.2.3. RNA extraction, RT-polymerase chain reaction (PCR) and histological studies of tissue RNA was extracted from liver tissues as described by Chomczynski and Sacchi [4]. RNA was analyzed by RT-PCR using a set of primers [3]. Tissue sections fixed in 10% neutral-buffered formalin were used for hematoxylin and eosin staining.
2. Results 2.1. Anti-HIV activities
We found that S-oligo could exhibit 2 different anti-HIV activities. One is sequence-non-specific and another is sequence-specific anti-HIV activity. Sequence-non-specific activity was found in the cytopathic effect inhibition assay where even homo-oligomer S-oligo could inhibit de novo infection (no HIV DNA synthesis) possibly with multiple mechanisms including inhibition of HIV RT. On the other hand, the result obtained from the viral gene expression assay was sequencespecific anti-HIV activity, namely antisense activity. In the enzyme assay for HIV RT, S-oligo
M. Matsukura
et al. I Toxicology
showed much more potent inhibition compared to n-oligo. Such inhibition, however, could be reversed by addition of bovine serum albumin (BSA) in the assay solution suggesting that the inhibition of RT activity by S-oligo is a result of non-specific protein binding. Similarly, increase of percentage of FCS in the complete medium eliminated the cytotoxicity of S-oligo whereas a higher percentage than 30% of FCS is somehow toxic and, under such conditions, FCS did not compete with the cytotoxicity of S-oligo (data not shown). In the viral gene expression assay, n-oligo showed a profound decrease of [“Hlthymidine uptake without decrease of viable cell number. This pseudo-cytotoxicity was derived from competition of non-radioactive thymidine from n-oligo degraded by nucleases in the FCS. Soligo is known to be nuclease-resistant, but could be degraded very slowly in culture media and also in vivo. One should take account of not only the oligomers but also the metabolites for toxicity* 2.2. Anti-HBV activity The data from in vivo experiments using HBx transgenic mice will be published elsewhere. Therefore, we describe the results here briefly without figures and tables. 2.2.1. Down-regulation of HBx gene expression As an initial trial, we injected 1 of 4 S-oligos i.p. into the mice for 7 consecutive days (1 mg/ day) at the age of 3 month old. On the last day of injection, mice were sacrificed and gene expression of HBx in liver was examined by means of the RT-PCR. We found a significant decrease of HBx gene expression in AS-Ztreated mice whereas other oligomers did not cause a decrease (data not shown). 22.2. Long-term treatment with AS-2 We started the treatments before the mice were 3 months old. At 1 week old, even 0.2 mg/day injection 3 times a week was found to be highly toxic to mice, which resulted in death at 4 weeks old. Then we delayed the initiation of the
Letters 82183 (199-Y) 435-438
437
treatment to 2 weeks old and reduced the dosage. With this protocol, the treatment was continued until the mice were 10 weeks old. Significant reduction of the expression and histologically milder changes were found in mice treated with AS-2 but not in mice treated with S-2 or PBS. In a series of treatments, cell infiltration into liver was found in mice treated with AS-1 and S-l suggesting that toxic reaction could be sequence specific or base composition of oligomers (GC-rich) (data not shown).
3. Discussion Toxicological consideration of oligomers could not be obtained without knowing more about the characteristics of the experiments. We should compare experiments with each other in terms of their advantages and disadvantages. In vitro assay systems which we used have advantages over in vivo experiments in the following ways: (1) a relatively small amount of oligomer is required; (2) because of relative ease, many samples including several controls can be tested simultaneously; (3) because of relatively defined and short-term assay conditions, experiments can be repeatable in a short time and the data obtained can be reproduced. There are also significant disadvantages as follows: (1) in vitro assay systems have artificial conditions such as use of particular immortalized cell lines, artificially conditioned medium containing FCS, relatively rapid cell growth (possibly different sensitivity to cytotoxicity of compounds) and/or lack of heterogenous cell populations; (2) a larger or smaller amount of viral production in comparison to in vivo situations; (3) lack of pharmacokinetics and tissue distribution. Since in vivo HIV experiments using animal models are neither definitive nor easy, we have been performing in vitro assays for detection of antisense activity. We have found at least several mechanisms of in vitro anti-HIV activities, which were initially confusing and made the antisense strategy perhaps less appealing to some of the researchers. Several mechanisms in anti-HIV activity might be also true in toxicity of oligo-
438
M. Matsukura
et al. I Toxicology
mers (at least 2: sequence-non-specific and sequence-specific toxicity). In monocytes, S-oligo could introduce superoxide (unpublished result), which could be partially reduced by indomethacin. S-oligo could also introduce the viral expression of HIV in dormant HIV-infected cells (unpublished data). Although suitable in vitro assays should be generally employed for antisense studies at the start, in vivo experiments using animal models could sometimes be better in sequence specificity and interpretability of data including toxicity. If enough oligomer is available, one might want to perform in vivo experiments without in vitro assays.
References [l] Mitsuya, H., Jarrett, F.D.M., Sarngadharan,
R.F., Matsukura, M., Veronese, M.G., Johns, D.G., Reitz, M. and
Letters 82183 (1995) 435-438 Broder, S. (1987) Long-term inhibition of HTLV-IIIILAV (human immunodeficiency virus) DNA synthesis and RNA expression in T-cells protected by 2’,3’-dideoxynucloesides in vitro. Proc. Nat]. Acad. Sci. USA 84, 20332037. 121Kim, C.-M., Koike, K., Saito, I., Miyamura, T. and Jay, G. (1991) HBx gene of hepatitis B virus induced liver cancer in transgenic mice. Nature 351, 317-320. M., Kurokawa, K. and Koike. K. [31Moriya, K., Matsukura, In vivo inhibition of hepatitis B virus gene expression by antisense phosphorothioate oligodeoxynucleotides. (Submitted.) P. and Sacchi, N. (1987) Single-step meth[41Chomczynski, od of RNA isolation by acid guanidinium thiocyanatephenol-chloroform extraction. Anal. Biochem. 162. 156159.