The Evolution of Antisense Oligonucleotide Chemistry-A Personal Journey

Over the last four decades, tremendous progress has been made in use of synthetic oligonucleotides as therapeutics. This has been possible largely by introducing chemical modifications to provide drug like properties to oligonucleotides. In this article I have summarized twists and turns on use of chemical modifications and their road to success and highlight areas of future directions.

Novel oligonucleotides containing two 3'-ends complementary to target mRNA show optimal gene-silencing activity

Oligonucleotides are being employed for gene-silencing activity by a variety of mechanisms, including antisense, ribozyme, and siRNA. In the present studies, we designed novel oligonucleotides complementary to targeted mRNAs and studied the effect of 3'-end exposure and oligonucleotide length on gene-silencing activity. We synthesized both oligoribonucleotides (RNAs) and oligodeoxynucleotides (DNAs) with phosphorothioate backbones, consisting of two identical segments complementary to the targeted mRNA attached through their 5'-ends, thereby containing two accessible 3'-ends; these compounds are referred to as gene-silencing oligonucleotides (GSOs). RNA and/or DNA GSOs targeted to MyD88, VEGF, and TLR9 mRNAs had more potent gene-silencing activity than did antisense phosphorothioate oligonucleotides (PS-oligos) in cell-based assays and in vivo. Of the different lengths of GSOs evaluated, 19-mer long RNA and DNA GSOs had the best gene-silencing activity both in vitro and in vivo. These results suggest that GSOs are novel agents for gene silencing that can be delivered systemically with broader applicability.

Antisense strategies for oncogene inactivation

Antisense oligonucleotides have been evaluated as antineoplastic agents in a series of clinical trials, with mixed results. However, phase III trials incorporating G3139, a phosphorothioate oligomer targeted to the initiation codon region of the bcl-2 mRNA, have recently been completed in advanced melanoma, myeloma, and chronic lymphocytic leukemia (CLL). This article discusses the mechanism of the antisense effect and its dependence on the cellular internalization of oligonucleotides and the activity of RNase H. It also describes the properties, specific and nonspecific, of phosphorothioate oligonucleotides, the predominant species in current clinical trials, and discusses pharmacokinetic data obtained from earlier phase I and II trials employing these molecules. While the application of antisense technology to the treatment of human cancer is conceptually straightforward, in practice there are many complicated, mechanistically based questions that must be considered.

Self-stabilized CpG DNAs optimally activate human B cells and plasmacytoid dendritic cells

We recently showed that 5'-terminal secondary structures in CpG DNA affect activity significantly more than those at the 3'-end [Biochem. Biophys. Res. Commun. 306 (2003) 948]. The need for an accessible 5'-end of CpG DNA for activity suggested that the receptor reads the DNA sequence from this end. In continuation of these studies, we have designed immunomodulatory oligonucleotides (IMOs), consisting of a nine-mer stimulatory domain, containing a CpG motif and a hairpin-loop structure at the 3'-end, referred to as self-stabilized CpG DNAs. We studied the ability of self-stabilized CpG DNAs to stimulate human B-cell proliferation and interferon-alpha (IFN-alpha) secretion in plasmacytoid dendritic cell (pDC) culture assays. Self-stabilized CpG DNAs activated human B cells and induced plasmacytoid dendritic cells to secrete high levels of IFN-alpha. While both stimulatory and secondary structures in CpG DNAs were required for pDC activation, CpG motifs were sufficient to activate B cells. Interestingly, CpG motifs were not required for activity in the hairpin duplex region. Further modifications of the hairpin duplex region with a mixture of oligodeoxynucleotides and oligo-2'-O-methylribonucleotides in a heteroduplex formation permitted activation of both human B cells and pDCs.

Secondary structures in CpG oligonucleotides affect immunostimulatory activity

Oligodeoxynucleotides containing CpG dinucleotides in specific sequence contexts activate the vertebrate immune system. Our previous studies showed that the 5(')-end of a CpG oligonucleotide should be accessible for receptor recognition and subsequent immune stimulation. Activity is abrogated if this end is blocked by joining two CpG oligos through 5(')-5(') linkage. It was not known whether a similar effect would arise from secondary structures at either end of a CpG oligo, such as hairpin loops or terminal dimers. In the present study we found that 5(')-terminal secondary structures affect activity significantly more than those at the 3(')-end. The need for an open 5(')-end suggests that the receptor responsible for immune stimulation reads the DNA sequence from this end. These results may also provide insights to place CpG motifs appropriately in DNA vaccines to induce additional Th1 type responses.

Ten years of antisense inhibition of brain G-protein-coupled receptor function

Antisense oligonucleotides (AOs) are widely used as tools for inhibiting gene expression in the mammalian central nervous system. Successful gene suppression has been reported for different targets such as neurotransmitter receptors, neuropeptides, ion channels, trophic factors, cytokines, transporters, and others. This illustrates their potential for studying the expression and function of a wide range of proteins. AOs may even find therapeutic applications and provide an attractive strategy for intervention in diseases of the central nervous system (CNS). However, a lack of effectiveness and/or specificity could be a major drawback for research or clinical applications. Here we provide a critical overview of the literature from the past decade on AOs for the study of G-protein-coupled receptors (GPCRs). The following aspects will be considered: mechanisms by which AOs exert their effects, types of animal model system used, detection of antisense action, effects of AO design and delivery characteristics, non-antisense effects and toxicological properties, controls used in antisense studies to assess specificity, and our results (failures and successes). Although the start codon of the mRNA is the most popular region (46%) to target by AOs, targeting the coding region of GPCRs is almost as common (41%). Moreover, AOs directed to the coding region of the GPCR mRNA induce the highest reductions in receptor levels. To resist degradation by nucleases, the modified phosphorothioate AO (S-AO) is the most widely used and effective oligonucleotide. However, the end-capped phosphorothioate AOs (ECS-AOs) are increasingly used due to possible toxic and non-specific effects of the S-AO. Other parameters affecting the activity of a GPCR-targeting AO are the length (mostly an 18-, 20- or 21-mer) and the GC-content (mostly varying from 30 to 80%). Interestingly, one-third of the AOs successfully targeting GPCRs possess a GC/AT ratio of 61-70%. AO-induced reductions in GPCR expression levels and function range typically from 21 to 40% and 41 to 50%, respectively. In contrast to many antisense reviews, we therefore conclude that the functional activity of a GPCR after AO treatment correlates mostly with the density of the target receptors (maximum factor 2). However, AOs are no simple tools for experimental use in vivo. Despite successful results in GPCR research, no general guidelines exist for designing a GPCR-targeting AO or, in general, for setting up a GPCR antisense experiment. It seems that the correct choice of a GPCR targeting AO can only be ascertained empirically. This disadvantage of antisense approaches results mostly from incomplete knowledge about the internalisation and mechanism of action of AOs. Together with non-specific effects of AOs and the difficulties of assessing target specificity, this makes the use of AOs a complex approach from which conclusions must be drawn with caution. Further antisense research has to be carried out to ensure the adequate use of AOs for studying GPCR function and to develop antisense as a valuable therapeutic modality.

Quantitative three-dimensional analysis and diffusion modeling of oligonucleotide concentrations after direct intraparenchymal brain infusion

We compared quantitative experimental results on the diffusion of 35S-labeled phosphorothioate oligonucleotide (PS-ODN) after intraparenchymal infusion in rat brain, with the distributions predicted by Fick's second law of diffusion. Fischer 344 rats underwent identical intracerebral infusions of 36S-PS-ODN. After 0, 5, 11, 23, and 47 h, groups of animals were sacrificed and sequential brain cryosections subjected to autoradiography. The resulting experimental data were compared to the predicted distributions, for estimation of the apparent free diffusion coefficient, D*. Volumes of distribution and total content of 36 S-PS-ODN in the parenchyma were also computed, to monitor loss of total material. The values for D* were within the expected range for the 21-mer PS-ODN used, but a progressive decrease in D* over time was noted. The model requires D* to remain constant and, thus, does not adequately explain the spread of 35S-PS-ODN following infusion. The progressive slowing of spread over time suggests that at later time points, 35S-PS-ODN may be fixed by tissue binding or cellular uptake in the brain. Loss of material via vascular and CSF clearance may also contribute to the lack of fit. Our results highlight issues to be addressed in the modeling and experimental design of the intraparenchymal infusion process.

Comparison of pharmacokinetics and tissue disposition of an antisense phosphorothioate oligonucleotide targeting human Ha-ras mRNA in mouse and monkey

The plasma pharmacokinetics and tissue disposition of ISIS 2503 were studied in mice following single and multiple bolus intravenous (iv) injections of 1-50 mg/kg, and in monkeys following single and multiple 2-h iv infusions of 1-10 mg/kg and bolus iv injections of 1 mg/kg of ISIS 2503. ISIS 2503 and its metabolites were measured in plasma, urine, and tissues using solid-phase extraction followed by capillary gel electrophoresis (CGE). In both species, the plasma clearance of ISIS 2503 was characterized by rapid distribution to tissues, and to a lesser extent, metabolism. The plasma clearance in mice was at least two-fold more rapid than in monkeys at equivalent doses. The plasma disposition (t1/2) increased with dose. The highest concentrations of oligonucleotide were consistently observed in the kidney and liver in both species. At equivalent doses, tissue concentrations in monkeys were much higher than tissue concentrations in mice. Urinary excretion of total oligonucleotide was a minor elimination pathway in both species at doses < 10 mg/kg. However, urinary excretion of total oligonucleotide in mice was increased to 12-29% as dose increased from 20 to 50 mg/kg.

Absorption of antisense oligonucleotides in rat intestine: effect of chemistry and length

An in situ single-pass perfusion model was used to assess the effect of chemical modification and length on permeability and absorption of various oligonucleotides in rat intestine. Phosphorothioate oligodeoxynucleotides (PS-ODN) were compared with oligoribonucleotides with 2'-methoxyethyl (MOE) or 2'-O-methyl (OMe) modifications. A 25-mer PS-OMe-modified oligonucleotide showed relatively poor permeability in this model, as did unmodified 20-mer PS-ODN (permeability coefficient [P(eff)] = 2-8 X 10(-6)cm/sec). Modifying some or all of the oligonucleotides with 2'-MOE groups on deoxyribose and 5'-methylation of the cytosines substantially increased intestinal permeability of oligonucleotides. Both partially and fully modified PS-MOE oligonucleotides showed a (2-4)-fold increase in permeability as compared with unmodified PS-ODN. The presence of a phosphodiester backbone in MOE-modified compounds led to further increases in intestinal permeability. PS-MOE composed of 6, 8, 10, 12, 14, 16, 18, 20, and 22 nucleotides were also examined. It was found that the permeability of these oligonucleotides increased linearly with decreasing length.

Importance of nucleotide sequence and chemical modifications of antisense oligonucleotides

The antisense approach is conceptually simple and elegant; to design an inhibitor of a specific mRNA, one needs only to know the sequence of the targeted mRNA and an appropriately modified complementary oligonucleotide. Of the many analogs of oligodeoxynucleotides explored as antisense agents, phosphorothioate analogs have been studied the most extensively. The use of phosphorothioate oligodeoxynucleotides as antisense agents in various studies have shown promising results. However, they have also indicated that quite often, biological effects observed could be solely or partly non-specific in nature. It is becoming clear that not all phosphorothioate oligodeoxynucleotides of varying length and base composition are the same, and important consideration should be given to maintain antisense mechanisms while identifying effective antisense oligonucleotides. In this review, I have summarized the progress made in my laboratory in understanding the specificity and mechanism of actions of phosphorothioate oligonucleotides and the rationale for designing second-generation mixed-backbone oligonucleotides.

Pseudo-cyclic oligonucleotides: in vitro and in vivo properties

We have designed and studied antisense oligodeoxynucleotides (oligonucleotides; oligos) which we call 'pseudo-cyclic oligonucleotides' (PCOs). PCOs contain two oligonucleotide segments attached through their 3'-3'- or 5'-5'-ends. One of the segments of the PCO is an antisense oligo complementary to a target mRNA, and the other is a short protective oligo that is 5-8 nucleotides long and complementary to the 3'- or 5'-end of the antisense oligo. As a result of complementarity between the antisense and protective oligo segments, PCOs form intramolecular pseudo-cyclic structures in the absence of the target RNA. The antisense oligo segment of PCOs used for the studies described here is complementary to an 18-nucleotide-long site on the mRNA of the protein kinase A regulatory subunit RIalpha (PKA-RIalpha). Thermal melting studies of PCOs in the absence and presence of the complementary RNA suggest that the pseudo-cyclic structures formed in the absence of the target RNA dissociate, bind to the target RNA, and form heteroduplexes. The results of RNase H cleavage assays suggest that PCOs bind to complementary RNA and activate RNase H in a manner similar to that of an 18-mer conventional antisense PS-oligo. In snake venom (a 3'-exonuclease) or spleen (a 5'-exonuclease) phosphodiesterase digestion studies, PCOs are more stable than conventional antisense oligos because of the presence of 3'-3'- or 5'-5'-linkages and the formation of intramolecular pseudo-cyclic structures. PCOs with a phosphorothioate antisense oligo segment inhibited cell growth of MDA-MB-468 and GEO cancer cell lines similar to that of the conventional antisense PS-oligo, suggesting efficient cellular uptake and target binding. The nuclease stability studies in mice suggest that PCOs have higher in vivo stability than antisense PS-oligos. The studies in mice showed similar pharmacokinetic and tissue distribution profiles for PCOs to those of antisense PS-oligos in general, but rapid elimination from selected tissues.

Bolus intravenous injection of phosphorothioate oligonucleotides causes hypotension by acting as alpha(1)-adrenergic receptor antagonists

Bolus intravenous injections of phosphorothioate oligonucleotides (PS-ODN) into primates cause profound hypotension, which has been attributed to complement activation, the biochemical pathway leading to acute inflammatory response. Because the hypotension was not accompanied by peripheral or pulmonary edema and epinephrine was not effective, but administration of 200 ml Ringer's lactate was effective, we examined the possibility that the 15-base PS-ODN interferes with sympathetic tone. We administered doses ranging from 3.3 to 10 mg/kg of a 15-base PS-ODN as a 30-60 s iv bolus into the right atrium of conscious Macaca mulatta. Blood pressure fell to 27 mm Hg following a 5.0 mg/kg dose, but no hypotension was observed after a 3.3 mg/kg dose; 10 mg/kg was lethal. Adrenergic receptor binding was evaluated in radioligand binding assays using rat cerebral cortex membranes with radiolabeled prazosin. The 15-base PS-ODN competes with prazosin for the alpha(1)-adrenergic receptor with an IC50 of 14 microM, which favors binding over serum albumin (K(d) = 37 to 48 microM). Admixing serum albumin with 5.0 mg/kg 15-base PS-ODN prior to injection prevented hypotension, suggesting that unbound PS-ODN interferes with sympathetic tone before binding to plasma proteins. Interactions of the 15-base PS-ODN with the alpha(1)-adrenergic receptor in vivo were confirmed by a decreased response to phenylephrine. Reducing the length from 15 to 9 or 5 bases abolished alpha(1)-adrenergic receptor binding in vitro and bolus infusion of 5.0 mg/kg of 9-base PS-ODN no longer produced hypotension. In conclusion, the 15-base PS-ODN shows cooperative binding to the alpha(1)-adrenergic receptor, which produces cardiovascular effects that are oligomer length, dose, and formulation dependent.

Biodistribution, stability, and antiviral efficacy of liposome-entrapped phosphorothioate antisense oligodeoxynucleotides in ducks for the treatment of chronic duck hepatitis B virus infection

This study investigated the feasibility of using liposomes to increase the hepatic delivery and antiviral efficacy of phosphorothioate antisense oligodeoxynucleotides (PS-ODN) for the in vivo treatment of hepatitis B virus (HBV) infection. Ducks infected with duck hepatitis B virus (DHBV) were used as the model. We studied the stability of an antisense PS-ODN in duck plasma, its integrity during the process of liposome entrapment, its in vivo biodistribution, plasma clearance, and excretion. In addition, the intrahepatic distribution of a labeled free and liposome-entrapped ODN was also investigated. The results of our studies show that: 1) phosphorothioate ODN remain stable during the process of liposome entrapment; 2) are stable in duck plasma for many hours; 3) are rapidly cleared from the plasma when injected intravenously; 4) intravenous injection of antisense ODNs entrapped within liposomes enhances delivery of the ODN to the liver; and 5) inhibit DHBV replication. Serum DHBV DNA levels fell rapidly, with a corresponding decrease in intrahepatic viral replicative intermediates at the end of the 5-day study period. Although inhibition of viral replication and a fall in the target protein was observed, a marked inhibition of viral replication was also observed with high doses of a random-sequence ODN. Thus, it is not certain that inhibition of viral replication was entirely through an antisense mechanism. Therefore, liposomes may be effective vehicles to improve the delivery of antisense oligonucleotides to the liver for the therapy of hepatotropic viruses.

Effect of aspirin on protein binding and tissue disposition of oligonucleotide phosphorothioate in rats

Pharmacokinetic studies of phosphorothioate oligodeoxynucleotides (PS-oligonucleotides) in animals show that following intravenous administration, PS-oligonucleotide clears out rapidly from the plasma and is distributed to majority of the organs. PS-oligonucleotides are bound to plasma proteins extensively. This study was aimed to determine the effect of aspirin, a commonly used drug, on pharmacokinetics of PS-oligonucleotides. In the present study, PS-oligonucleotide was administered to rats that had received aspirin by gavage. Pharmacokinetic study shows that if PS-oligonucleotide was administered following aspirin administration in rats, a) plasma pharmacokinetic parameters (t1/2alpha?, t1/2beta, AUC, etc.) had lower values, b) tissue disposition was different, and c) rate and route of elimination was affected in animals compared to rats receiving PS-oligonucleotide alone. This finding suggests that pharmacokinetics of PS-oligonucleotides can be affected with certain class of drugs, which may have direct impact on biological activity and safety.

Pharmacokinetics and metabolism of an oligodeoxynucleotide phosphorothioate (GEM91) in cynomolgus monkeys following intravenous infusion

The pharmacokinetics and metabolism of an antisense oligonucleotide phosphorothioate (GEM91) were studied in cynomolgus monkeys following intravenous infusion. [35S]-Labeled GEM91 was administered to 12 monkeys by means of a 2-hour intravenous infusion at a dose of 4 mg/kg. Plasma pharmacokinetic analysis revealed that the maximum plasma concentration was 41.7 microg equivalents/ml, which was achieved in 2.13 hours. The plasma elimination half-life was 55.8 hours based on radioactivity levels. Urinary excretion represented the major pathway of elimination, with 70% of the administered dose excreted in urine over 240 hours. The oligonucleotide was widely distributed to tissues. The highest concentrations were observed in the liver and kidney. Analysis of the extracted oligonucleotide following post-labeling with [32p] on polyacrylamide gel electrophoresis showed the presence of both intact and degraded oligonucleotide in plasma, kidney, liver, spleen, and lymph nodes. Based on the methods used for post-labeling (either 3'-end or 5'-end), different patterns of bands were observed on polyacrylamide gel electrophoresis, suggesting metabolic modification of the administered oligonucleotide.

Pharmacokinetics of oligonucleotides

The effectiveness of antisense oligonucleotides as therapeutic agents depends on their pharmacokinetics, tissue disposition, stability, elimination and safety profile. Pharmacokinetic data allow one to determine the frequency of administration and any potential toxicity associated with chronic administration. Phosphorothioate oligonucleotides degrade from the 3' end, the 5' end, and both the 3' and 5' ends in a time- and tissue-dependent manner. After intravenous administration in mice, rats and monkeys, phosphorothioate oligonucleotides are detected in plasma; they distribute rapidly and are retained in the majority of tissues. The major route of elimination is the urine. The pharmacokinetic profile is similar following subcutaneous, intradermal or intraperitoneal administration, but with lower maximum plasma concentrations. Phosphorothioate oligonucleotides have a short plasma half-life in humans. End-modified, mixed-backbone oligonucleotides (MBOs) contain nuclease-resistant 2'-O-alkylribonucleotides or methylphosphonate internucleotide linkages at both the 3' and 5' ends of phosphorothioate oligonucleotides. These end-modified MBOs have pharmacokinetic profiles similar to those of the parent phosphorothioate oligonucleotides, but they are significantly more stable in vivo and they can be administered orally. Centrally modified MBOs contain modified RNA or DNA in the centre of a phosphorothioate oligonucleotide. They show controlled degradation and elimination following administration in rats. The pharmacokinetics of antisense oligonucleotides depends on the sequence, the nature of the oligonucleotide linkages and the secondary structure.

Plasmodium falciparum: effect of chemical structure on efficacy and specificity of antisense oligonucleotides against malaria in vitro

Antisense oligodeoxynucleotides (AS ODNs) have shown promise both as potential anti-malarial chemotherapeutic agents and as a means for identifying genes critical for parasite survival. Because conventional ODNs containing phosphodiester (PO) groups are subject to rapid nuclease degradation, ODNs with phosphorothioate (PS) groups are commonly used. However, at high concentration, these lose target specificity, and in some animal models, they become toxic. We compared a variety of chemical modifications (PO, PS, PO-PS hybrids, 2'-O-methyl-2'-deoxy chimeras) and structural modifications (sequence alterations favoring self-stabilizing loop formation) for their ability to inhibit Plasmodium falciparum malaria cultured in vitro. All modifications were done using an AS ODN sequence targeted against dihydrofolate reductase thymidylate synthase (DHFR). Inhibition by PO-PS hybrids containing as few as three PS groups at the 3'- and 5'-ends did not differ significantly from that obtained using compounds containing all-PS groups. Similarly, inhibition by PS chimeric compounds containing 2'-O-methyl modifications did not differ significantly from that of conventional PS compounds. In contrast, while inhibition by PO-PS hybrid chimeras did not differ significantly from that of all-PS compounds at low concentrations, at 1 microM they inhibited parasite growth 25% less (P < 0.001) than all-compounds or PS 2'-O-methyl-2'-deoxy chimeras. Extension of the nucleotide sequence to increase stem-loop formation yielded two compounds which inhibited parasite growth about 20% more than unmodified compounds, though this difference was not significant. Furthermore, most of this increase appears to correlate with the greater number of PS groups associated with the increased ODN length. We conclude that limiting the number of PS groups and inclusion of PO 2'-O-methyl groups may yield compounds with high antisense activity but low non-sequence-dependent effects. Such compounds are currently being tested in vivo.

Toxicologic effects of an oligodeoxynucleotide phosphorothioate and its analogs following intravenous administration in rats

The aim of the present study is to evaluate the in vivo toxicologic effects of a phosphorothioate oligodeoxynucleotide (PS oligo) and three of its analogs [PS oligo containing four methylphosphonate linkages at the 3' and 5'-ends (MBO 1), PS oligo containing four 2'-O-methylribonucleosides at both the 3'- and 5'-ends (MBO 2), and PS oligo containing an 8 bp loop region at the 3'-end (self-stabilized oligo)]. Oligodeoxynucleotides were administrated intravenously to male and female rats at doses of 3, 10, and 30 mg/kg/day for 14 days. Rats were killed on day 15, blood samples were collected for hematology and clinical chemistry determinations, and tissues, including lymph nodes, spleens, livers, and kidneys, were subjected to pathologic examinations. The toxicity profiles of the four oligodeoxynucleotides were very similar, but differed in magnitude. In terms of the severity of the abnormalities caused by the oligodeoxynucleotides, the order was MBO 2 > PS oligo > self-stabilized oligo > MBO 1. Alterations in hematology parameters included thrombocytopenia, anemia, and neutropenia. Abnormalities in clinical chemistry parameters observed with PS oligo or MBO 2 were dose-dependent elevation of liver transaminases and reduction of the levels of alkaline phosphatase, albumin, and total protein. In addition, MBO 2 caused elevation of the total bilirubin level in male rats at the 30 mg/kg dose. No major alterations in hematology or clinical chemistry were observed in rats receiving MBO 1 or self-stabilized oligo. Dose-dependent enlargements of spleen, liver, and kidney were observed, especially in rats receiving PS oligo and MBO 2. Pathologic studies showed a generalized hyperplasia of the reticuloendothelial (RE) system in the tissues examined. Alterations in the spleen were mainly RE cell hyperplasia and hematopoietic cell proliferation. In addition to RE cell hyperplasia, lymph nodes showed necrosis, hepatocytes showed cytologic alterations and necrosis, and kidneys showed renal tubule regeneration. The severity of pathologic changes observed was oligodeoxynucleotide dependent, in the order of MBO 2 > PS oligo > self-stabilized oligo > MBO 1.

Pharmacokinetics, metabolism, and elimination of a 20-mer phosphorothioate oligodeoxynucleotide (CGP 69846A) after intravenous and subcutaneous administration

The pharmacokinetics, tissue distribution and metabolism of CGP 69846A, a 20-mer phosphorothioate oligodeoxynucleotide targeted against the 3'-untranslated region of human c-raf-1 kinase mRNA, were investigated in vivo in rats after intravenous and subcutaneous administration. Intravenous disposition studies with [3H]CGP 69846A were supported with analysis by capillary gel electrophoresis and electrospray mass spectrometry. In combination, these techniques provide a detailed account of the pharmacokinetic and metabolic profile for this compound. The elimination of CGP 69846A after a single intravenous dose was studied over extended periods in mice using whole-body autoradiography and capillary gel electrophoresis. Subcutaneous administration to rats resulted in a significant bioavailability with peak plasma levels 4.5-fold lower than after intravenous dosing. This dose route resulted in low interanimal variability and only slightly greater metabolism of the oligonucleotide compared to the intravenous administration.

Central overexpression of the TRH precursor gene induces hypertension in rats: antisense reversal

Extrahypothalamic TRH participates in cardiovascular regulation and spontaneous hypertension of the rat. To investigate whether an increase in central TRH activity produces hypertension we studied the effect of the preTRH overproduction induced by I.C.V. transfection with a naked eukaryotic expression plasmid vector which encodes preTRH (pCMV-TRH). Northern blot analysis and RT-PCR showed that pCMV-TRH was transcribed in vitro and in vivo. At 24, 48, and 72 hours, pCMV-TRH (100 microg) in a significant and dose-dependent manner increased 37%, 84%, and 49%, respectively, the diencephalic TRH content and SABP (42+/-3, 50+/-2, and 22+/-2 mm Hg, respectively) with respect to the vector without the preTRH cDNA insert (V[TRH(-)]) as measured by RIA and the plethysmographic method, respectively, in awake animals. In addition, using immunohistochemistry we found that the increase of TRH was produced in circumventricular areas where the tripeptide is normally located. To further analyze the specificity of these effects we studied the actions of 23-mer sense (S), antisense (AS), and 3'self-stabilized sense (Ss) and antisense (ASs) phosphorothioate oligonucleotides against the initiation codon region. Only ASs inhibited the increase of TRH content and SABP induced by pCMV-TRH treatment. In addition, pCMV-TRH-induced hypertension seems not to be mediated by central Ang II or serum TSH. To summarize, central TRH overproduction in periventricular areas induced by I.C.V. transfection produces hypertension in rats which is reversed by specific antisense treatment. This model may help in testing effective antisense oligodeoxynucleotides against other candidate genes.

In vivo metabolic profile of a phosphorothioate oligodeoxyribonucleotide

Antisense phosphorothioate oligodeoxyribonucleotides (PS oligonucleotides) have the ability to inhibit individual gene expression in the potential treatment of cancer and viral diseases. Following administration in vivo, PS oligonucleotides are rapidly cleared from the plasma and distributed to various organs. However, the manner in which administered oligonucleotides are metabolized in plasma and tissues is poorly understood. In this study, a 25-mer PS oligonucleotide (GEM91) complementary to the gag gene mRNA of the human immunodeficiency virus (HIV-1) was administered to mice through intravenous injections to investigate its metabolism. The PS oligonucleotide was extracted from plasma at 1 hour postadministration and from kidney and liver at 24 hours postadministration. After extraction, the PS oligonucleotide and its metabolites were tailed with dA and annealed to a dT-tailed plasmid. The recombinant plasmid was ligated and used to transform competent bacteria. The region of interest containing the PS oligonucleotide was then sequenced. Our results show that degradation of the PS oligonucleotide in plasma was primarily from the 3'-end. However, in kidney and liver, degradation was primarily from the 3'-end, but a large proportion of the PS oligonucleotide was degraded from the 5'-end as well. We also studied the metabolism of PS oligonucleotide in plasma after 2-hour intravenous infusion in HIV-infected patients. The degradation of the PS oligonucleotide in plasma was primarily from the 3'-end. This study is important in understanding the metabolism of antisense PS oligonucleotide in vivo in general but also provides guidance for designing second-generation antisense oligonucleotides with improved stability and safety profile.

Effects of synthetic oligonucleotides on human complement and coagulation

Oligodeoxynucleotide phosphorothioates (PS-oligos) are being studied as novel therapeutic agents based on their ability to inhibit gene expression. Preclinical studies produced unanticipated complement and coagulation effects in monkeys receiving high-dose PS-oligo. In the present in vitro studies, PS-oligo inhibited normal human blood clotting as well as subsequent assays for prothrombin fragment PF(1+2) and hemolytic complement. PS-oligo treatment of normal donor plasma produced concentration-dependent prolongations of clotting times, with the activated partial thromboplastin time more sensitive than prothrombin time or thrombin clotting time. PS-oligo treatment of normal donor serum similarly reduced hemolytic complement activity in a concentration-dependent manner. Reduced hemolysis correlated with increased levels of complement fragment C4d. The anti-heparin drug protamine sulfate inhibited in vitro effects of PS-oligo in both complement and coagulation assays, suggesting that charged residues in internucleotide linkages of PS-oligo mediated the observed activities. Therefore, oligonucleotides with varying internucleotide linkages, nucleotide sequence, or secondary structure were compared. Both complement and coagulation effects appeared to be independent of nucleotide sequence but were strongly related to the nature of internucleotide linkages. Several of these modified oligonucleotides have been shown previously to retain potent antisense activity and thus may represent viable alternatives for antisense therapeutics.

A nonradioisotope approach to study the in vivo metabolism of phosphorothioate oligonucleotides

A 25-mer phosphorothioate oligodeoxynucleotide (GEM 91) complementary to the gag gene mRNA of HIV-1 virus was administered intravenously (i.v.) at a dose of 10 mg/kg/day for 8 weeks or 25 mg/kg single dose subcutaneously (SC) to adult Rhesus monkeys. No radioactive markers were used. A capillary gel electrophoresis (CGE) method with UV detection was used to determine the concentration of GEM 91 in plasma and the metabolite profile. The metabolite profile was virtually the same following a single dose of either 10 mg/kg i.v. or 25 mg/kg SC. A different metabolite profile was observed after 4 or 8 weeks of multiple i.v. doses of 10 mg/kg/day. The extract was subjected to matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOFMS) for positive identification. Mass spectrometry confirmed the major metabolic pathway in vivo to be via 3'-end exonuclease activity. The extract was then subjected to a hybridization-assisted ligation reaction in which only 5'-end intact metabolites were labeled. Analysis by CGE with laser-induced fluorescence (LIF) detection allowed each of these metabolites to be quantified with a limit of detection of 1 ppb (ng/ml). MALDI-TOFMS identified components digested from both ends of the DNA. This study demonstrates that the combination of quantitative CGE-LIF and MALDI-TOFMS yields a powerful and unique approach to study the metabolism of phosphorothioate oligonucleotides.

Novel enzymatic and immunological responses to oligonucleotides

Oligonucleotide phosphorothioates (PS-oligos) are being studied as antisense agents for viral infection and cancer. In preclinical studies, PS-oligos produced dose-dependent changes in heart rate and blood pressure and significantly reduced serum hemolytic complement, which could be avoided by slowing infusion rates. Here, in vitro PS-oligo treatment of either human, rhesus monkey or guinea pig serum reduced hemolytic complement and further inhibited in vitro coagulation when added to whole blood or citrated plasma. These effects were dependent upon both oligonucleotide dose and structure. Oligonucleotides having identical sequences but containing methylphosphonates (Chimeric), 2'-O-methyl ribonucleosides (Hybrid) or 3' hairpin loop (Self-stabilized) had altered effects on complement and coagulation in vitro.