Antisense phosphorothioate oligodeoxynucleotides: introductory concepts and possible molecular mechanisms of toxicity

Over the past 5 years or so, much attention has been given to the possible use of synthetic antisense oligonucleotide analogs as a new class of therapeutic agents that function by sequence-specific inhibition of genetic expression. The basic design concepts which underline this novel approach to drug discovery are briefly described herein, together with some of the chemical, biochemical, and pharmacological aspects of phosphorothioate oligodeoxynucleotides that are first-generation antisense compounds now under clinical investigation. Possible molecular mechanisms of toxicity for this class, and other structural types of antisense compounds are discussed with the hope of stimulating interest in future toxicological studies in this emerging area of drug development.

Reversal of methylmercaptopurine ribonucleoside cytotoxicity by purine ribonucleosides and adenine

6-Methylmercaptopurine ribonucleoside-5'-phosphate (MeSPuRMP), the sole metabolite of 6-methylmercaptopurine ribonucleoside (MeSPuRib), is a strong inhibitor of purine de novo synthesis, inducing depletion of intracellular purine nucleotides and subsequent cell death in several tumor cell lines. In this study prevention of MeSPuRib cytotoxicity by compounds of the purine salvage pathway was studied in Molt F4 human malignant T-lymphoblasts. Adenosine, adenine and inosine were able to prevent depletion of the adenine nucleotide pool when used in combination with 0.5 microM MeSPuRib, but had virtually no effect on depletion of guanine nucleotides. Nevertheless, these three purine compounds were able to reduce the cytotoxic effects induced by MeSPuRib. Addition of guanosine to cells treated with 0.5 microM MeSPuRib normalized the guanine nucleotide pool, but adenine nucleotides remained depleted. Under these conditions, inhibition of cell growth was significantly decreased. With the combination of guanosine and 10 microM MeSPuRib, cytotoxicity was increased compared to 10 microM MeSPuRib alone, associated with a depletion of adenine nucleotides to 9% of untreated cells. Since cell growth and cell viability of Molt F4 cells are less inhibited by MeSPuRib under conditions where adenine nucleotide depletion is prevented by purine compounds (and where the other nucleotides are depleted) we conclude that depletion of adenine nucleotides is an important factor in MeSPuRib cytotoxicity.

Uptake, intracellular distribution, and stability of oligodeoxynucleotide phosphorothioate by Schistosoma mansoni

The in vitro uptake, cellular distribution, efflux, stability, and toxicity levels of an oligodeoxynucleotide phosphorothioate (PS-oligonucleotide) have been studied in mature Schistosoma mansoni worms. The intracellular accumulation of 35S-labeled PS-oligonucleotide occurred roughly in proportion to the worm body mass over a wide concentration range, whether the worms were exposed singly or in mating pairs. Cellular uptake was dependent on the extracellular concentration. A minor fraction (13%) of the PS-oligonucleotide taken up by the worm accumulated in the surface tegumental coat. Most of the PS-oligonucleotide taken up localized in the cytosol (54%) and the nuclei-enriched (33%) fractions. In a time course study on adult worms in culture, oligonucleotide uptake was observed within the first 2 h and peaked at about 36 h. A decrease in the intracellular concentration of the PS-oligonucleotide was observed by 42 h. Analysis of the extracted oligonucleotides showed that PS-oligonucleotide was digested slowly. Efflux of the oligonucleotide was time and temperature dependent. Significant toxicity to the cultured worms did not occur until the PS-oligonucleotide concentration was over 8 mg/ml (1 mM).

Design and application of antisense oligonucleotides in cell culture, in vivo, and as therapeutic agents

1. Synthetic oligonucleotides can inhibit the expression of a gene in a sequence specific manner on the transcriptional and translational level. These molecules are usually referred to as antisense oligonucleotides. 2. Antisense mediated inhibition of gene expression is a valuable tool to analyze the function of a gene in vivo and can also be used for therapeutic gene suppression. 3. A number of factors such as the mode of action, specificity, chemistry, and pharmacology must be carefully considered for the design and successful application of antisense oligonucleotides. 4. Assay systems and controls must be chosen as to assure that the observed biological effects of antisense oligonucleotides do in fact reflect the result of a specific gene inhibition. 5. This article critically discusses these factors in view of the literature and our own experience with a wide range of cell types and animal models, targeting different genes. The emphasis is on the use of phosphorothioate oligodeoxynucleotides in cell cultures, in vivo, and as potential drugs.

Antisense phosphorothioate oligonucleotides: selective killing of the intracellular parasite Leishmania amazonensis

We targeted the mini-exon sequence, present at the 5' end of every mRNA of the protozoan parasite Leishmania amazonensis, by phosphorothioate oligonucleotides. A complementary 16-mer (16PS) was able to kill amastigotes--the intracellular stage of the parasite--in murine macrophages in culture. After 24 hr of incubation with 10 microM 16PS, about 30% infected macrophages were cured. The oligomer 16PS acted through antisense hybridization in a sequence-dependent way; no effect on parasites was observed with noncomplementary phosphorothioate oligonucleotides. The antisense oligonucleotide 16PS was a selective killer of the protozoans without any detrimental effect to the host macrophage. Using 16PS linked to a palmitate chain, which enabled it to complex with low density lipoproteins, improved the leishmanicidal efficiency on intracellular amastigotes, probably due to increased endocytosis. Phosphorothioate oligonucleotides complementary to the intron part of the mini-exon pre-RNA were also effective, suggesting that antisense oligomers could prevent trans-splicing in these parasites.

Antisense oligodeoxynucleotide phosphorothioate complementary to Gag mRNA blocks replication of human immunodeficiency virus type 1 in human peripheral blood cells

Gene-expression modulator 91 (GEM91) is a 25-nt antisense oligodeoxynucleotide phosphorothioate complementary to the Gag mRNA of human immunodeficiency virus type 1 (HIV-1). Cellular uptake and intracellular distribution of GEM91 within cells suggest that this oligomer is readily available for antisense activity. GEM91 inhibited HIV-1 replication in a dose-dependent and sequence-specific manner. In a comparative study, 2 microM GEM91 was as effective as 5 microM 3'-azido-3'-deoxythymidine in blocking virus replication during the 28-day treatment of an HIV-1-infected T-cell line. GEM91 also completely inhibited (> 99%) the growth of three different HIV-1 isolates in primary lymphocytes and prevented the cytopathic effect of the virus in primary CD4+ T cells. Similarly, treatment with GEM91 for 3 weeks of HIV-1/BaL-infected primary macrophages blocked virus replication. Based on GEM91 anti-HIV-activity, safety, and pharmacokinetic profile in animals, a clinical trial was started using this compound as an antisense oligonucleotide drug for the treatment of the acquired immunodeficiency syndrome.

The 2-5A system and HIV infection

2',5'-Oligoadenylates (2-5A) have an essential role in the establishment of the antiviral state of a cell exposed to virus infection. The key enzymes of the 2-5A system are the 2-5A forming 2',5'-oligoadenylate synthetase (2-5OAS), the activity of which depends on the presence of viral or cellular double-stranded RNA (dsRNA), and the 2-5A-activated ribonuclease (RNase L). Basic research in recent years has shown that the 2-5A system is a promising target for anti-HIV chemotherapy, particularly due to its interaction with double-stranded segments within HIV RNA. Two new strategies have been developed which yield a selective antiviral effect of 2-5A against HIV-1 infection: (1) development of 2-5A analogues displaying a dual mode of action (activation of RNase L and inhibition of HIV-1 RT) and (2) intracellular immunization of cells against HIV-1 infection by application of the HIV-1-LTR--2-5OAS hybrid gene. A further strategy is the inhibition of DNA topoisomerase I by longer 2-5A oligomers.

Pharmacokinetics of an antisense phosphorothioate oligodeoxynucleotide against rev from human immunodeficiency virus type 1 in the adult male rat following single injections and continuous infusion

An antisense phosphorothioate oligodeoxynucleotide 27-mer complementary to the rev gene mRNA of the human immunodeficiency virus (HIV-1) was administered to rats through intravenous injections and subcutaneous infusions in order to investigate the disposition of this compound. In addition, nonlethal toxic responses of the rat were evaluated. A biphasic plasma clearance with t1/2 alpha of 20-25 min and t1/2 beta of 27-41 hr was observed. Single doses ranging from 35 to 3257 micrograms were examined, and the plasma concentration and area under the plasma concentration-time curve (AUC) were found to be directly proportional to the dose. Continuous subcutaneous infusion of 50 mg over 28 days was also examined. The oligonucleotide is completely eliminated in the urine over 3 days. Electrophoretic analysis demonstrated that the excreted compound has the same mobility and UV-absorbance profile as the administered compound. Measurement of accumulation and distribution into tissues revealed unique tissue-specific rates and extent of oligonucleotide movement into and out of tissues. Results of the chronic infusion study suggest that uptake into tissue is not saturated, even after 28 days of infusion. Analysis of blood plasma from oligonucleotide-treated animals shows a possible transient elevation in levels of lactate dehydrogenase (LDH), alanine aminotransferase (ALT), and aspartate aminotransferase (AST), but not alkaline phosphatase (AP), gamma-glutamyltransferase (GT), and bilirubin. The data collectively support the potential utility of phosphorothioate oligonucleotides as therapeutic agents in vivo.

In vivo toxicological effects of rel A antisense phosphorothioates in CD-1 mice

To characterize the in vivo toxicity of phosphorothioate antisense oligonucleotides against rel A (p65 subunit of NF-kappa B transcription factor), forty-eight 6-week-old CD-1 mice were split into 4 groups (6/sex/group) receiving vehicle (phosphate-buffered saline) or doses of 50, 100, and 150 mg/kg of rel A antisense oligonucleotides intraperitoneally 3 times weekly for 2 weeks. Clinical signs of toxicity included weakness, and decreased motor activity and food consumption with body weight loss. Mortality occurred in 7 of 12 mice in the 150-mg/kg group and in 2 of 12 mice in the 100-mg/kg group, most of which died within the first 2 to 4 days of treatment. The remaining mice were necropsied on day 15. The major hematological finding was severe dose-dependent thrombocytopenia. The liver enzyme levels were mildly elevated in the serum of mid- and high-dose animals. At necropsy, increased spleen and liver weights were observed in treated mice, some of which also had mild pleural and/or peritoneal effusions. Histopathological examination revealed the likely cause of death to be acute renal failure due to renal cortical or tubular necrosis. Treatment-related changes were also found in the liver, spleen, bone marrow, and several other organs. In summary, the kidney, liver, and bone marrow (megakaryocytic lineage) were identified as the major target organs for toxicity with rel A antisense therapy.

Immune stimulation by an antisense oligomer complementary to the rev gene of HIV-1

Mice developed massive splenomegaly and polyclonal hypergammaglobulinemia within 2 days after intravenous injection of a phosphorothioate oligomer that is antisense to a portion of the rev region of the HIV-1 genome. Histologic examination of spleens from injected animals showed marked expansion of a uniform-appearing population of small lymphocytes and many mitoses. Spleen mononuclear cells (SMNCs) from injected animals showed approximately a 10-fold-increased uptake of [3H]thymidine and production of IgM and IgG. Flow cytometry analysis indicated that the responding cells were predominantly B-lymphocytes. The anti-rev oligomer also was mitogenic in vitro and stimulated immunoglobulin production by normal mouse SMNCs and human peripheral blood mononuclear cells. Similar immunologic effects were observed with an anti-rev 21-mer phosphorothioate, truncated at the 3' end, but not with a 20-mer human p53 antisense phosphorothioate or a 28-mer anti-rev phosphodiester. These observations are consistent with the possibility that DNA sequences homologous to the rev gene participate in the regulation of mammalian lymphocyte activation, proliferation and maturation.

Further confirmation of the role of adenyl cyclase and of cAMP-dependent protein kinase in primary afferent hyperalgesia

Recent evidence has suggested that cAMP plays a role as a second messenger in the decrease in nociceptive threshold (or hyperalgesia) produced by agents acting on primary afferent terminals. In support of this hypothesis we report that intradermal injection of a direct activator of adenyl cyclase, forskolin, produces a dose-dependent hyperalgesia in the rat. The duration of this hyperalgesia was prolonged by the phosphodiesterase inhibitors, isobutylmethylxanthine and rolipram. Forskolin hyperalgesia was antagonized by the Rp isomer of cyclic adenosine-3'5'-monophosphothioate, an analog of cAMP that prevents the phosphorylation of the cAMP protein kinase. The Rp isomer of cyclic adenosine-3'5'-monophosphothioate also inhibited the hyperalgesia induced by a membrane-permeable analogue of cAMP, 8-bromocyclic adenosine monophosphate, as well as the hyperalgesia induced by agents that are presumed to act directly on primary afferent nociceptors: prostaglandin E2, prostaglandin I2, (8R,15S)-dihydroxyicosa(5E-9,11,13Z)tetraenoic acid; and the adenosine A2-agonist 2-phenylaminoadenosine. Although the cAMP second messenger system contributes to primary afferent hyperalgesia, we found no evidence for a contribution of protein kinase C. Thus, hyperalgesia induced by prostaglandin E2, prostacyclin (prostaglandin I2), (8R,15S)-dihydroxyicosa(5E-9,11,13Z)tetraenoic acid, the adenosine A2-agonist 2-phenylaminoadenosine, 8-bromocyclic adenosine monophosphate and the direct activator of adenyl cyclase, forskolin, were not significantly attenuated by the selective inhibition of protein kinase C by the 19-31 fragment of protein kinase C. Two other inhibitors of protein kinase C, sphingosine and staurosporine, also failed to attenuate prostaglandin E2-induced hyperalgesia.

The stability, toxicity and effectiveness of unmodified and phosphorothioate antisense oligodeoxynucleotides in Xenopus oocytes and embryos

The properties of antisense phosphorothioate and unmodified oligodeoxynucleotides have been studied in Xenopus oocytes and embryos. We find that phosphorothioates, like unmodified oligodeoxynucleotides, can degrade Vg1 mRNA in oocytes via an endogenous RNase H-like activity. In oocytes, phosphorothioate oligodeoxynucleotides are more stable than unmodified oligodeoxynucleotides and are more effective in degrading Vg1 mRNA. In embryos, neither unmodified nor phosphorothioate deoxyoligonucleotides were effective in degrading Vg1 message at sub-toxic doses.

Combination therapy of Schistosoma japonicum by tubercidin and nitrobenzylthioinosine 5'-monophosphate

Coadministration of nitrobenzylthioinosine 5'-monophosphate (NBMPR-P) with high doses of tubercidin by i.p. injection into Schistosoma japonicum infected mice beginning 5 weeks post-infection was highly toxic to the parasite but not the hose. Combination therapy resulted in a striking reduction in the number of worms, and the few worms that could be found were stunted. Combination therapy also caused a drastic reduction in the number of eggs in the livers (from 86,500 to 2,800 eggs/liver) and intestines (from 2,200 to 74 eggs/cm2), and 95% of eggs that were found were dead, indicating the termination of oviposition. Mice receiving the combination of tubercidin plus NBMPR-P appeared healthy and had normal size livers and spleens. These results demonstrate that by combining NBMPR-P with tubercidin high selective toxicity against S. japonicum can be achieved, as was shown previously with S. mansoni.