The concept and application of antisense oligonucleotides.

Since the identification of the double-stranded DNA helix by Watson and Crick in 1953, the knowledge of nucleotide structure and function has been an important potential tool in the study and therapy of disease. There is recent clinical evidence that antisense oligonucleotides may be important therapeutic compounds in the clinical therapy of a range of diseases, including infection (viruses and bacteria), oncology, and inflammation. Our laboratory-based understanding of antisense oligonucleotide activity has provided a foundation for their use in several human diseases. Potentially relevant applications include inflammatory bowel disease therapy, psoriasis, transplantation, rheumatoid arthritis, cytomegalovirus retinitis, hepatitis C, and solid tumor therapy. Here we will outline these applications as well as our ongoing clinical trials for Crohn's disease.

Hepatic distribution of a phosphorothioate oligodeoxynucleotide within rodents following intravenous administration.

The pharmacokinetics of ISIS 1082, a 21-base heterosequence phosphorothioate oligodeoxynucleotide, were characterized within rodent whole liver, and cellular and subcellular compartments. Cross-species comparisons were performed using Sprague-Dawley rat and CD-1 mouse strains. Although whole liver oligonucleotide deposition and the proportion of drug found within parenchymal and nonparenchymal cells were similar between the two rodent species as a function of both time and dose, dramatic differences in subcellular pharmacokinetics were observed. Specifically, within murine hepatocyte nuclei, drug was observed at the 10 mg/kg dose, whereas in the rat nuclear-associated levels required the administration of 25 mg/kg. Under all experimental regimens, murine hepatic nuclear-associated drug concentrations were at least 2-fold higher than those found in rat liver cells. More detailed metabolic analysis was also performed using high performance liquid chromatography/electrospray-mass spectrometry (HPLC/ES-MS) and demonstrated that although the extent of metabolism was similar for rat and mouse, the pattern of n-1 metabolites varied as a function of both species and cell type. While rat and mouse hepatocytes and rat nonparenchymal cellular metabolites were predominantly products of 3'-exonuclease degradation, mouse nonparenchymal cells contained a majority of n-1 metabolites produced by 5'-exonucleolytic activity. Based upon these data, it would appear that subcellular oligonucleotide disposition and metabolism among rodent species are more divergent than whole organ pharmacokinetics might predict.

Investigating the structure of human RNase H1 by site-directed mutagenesis.

In this study we examine for the first time the roles of the various domains of human RNase H1 by site-directed mutagenesis. The carboxyl terminus of human RNase H1 is highly conserved with Escherichia coli RNase H1 and contains the amino acid residues of the putative catalytic site and basic substrate-binding domain of the E. coli RNase enzyme. The amino terminus of human RNase H1 contains a structure consistent with a double-strand RNA (dsRNA) binding motif that is separated from the conserved E. coli RNase H1 region by a 62-amino acid sequence. These studies showed that although the conserved amino acid residues of the putative catalytic site and basic substrate-binding domain are required for RNase H activity, deletion of either the catalytic site or the basic substrate-binding domain did not ablate binding to the heteroduplex substrate. Deletion of the region between the dsRNA-binding domain and the conserved E. coli RNase H1 domain resulted in a significant loss in the RNase H activity. Furthermore, the binding affinity of this deletion mutant for the heteroduplex substrate was approximately 2-fold tighter than the wild-type enzyme suggesting that this central 62-amino acid region does not contribute to the binding affinity of the enzyme for the substrate. The dsRNA-binding domain was not required for RNase H activity, as the dsRNA-deletion mutants exhibited catalytic rates approximately 2-fold faster than the rate observed for wild-type enzyme. Comparison of the dissociation constant of human RNase H1 and the dsRNA-deletion mutant for the heteroduplex substrate indicates that the deletion of this region resulted in a 5-fold loss in binding affinity. Finally, comparison of the cleavage patterns exhibited by the mutant proteins with the cleavage pattern for the wild-type enzyme indicates that the dsRNA-binding domain is responsible for the observed strong positional preference for cleavage exhibited by human RNase H1.

Human RNase III is a 160-kDa protein involved in preribosomal RNA processing.

A human RNase III gene encodes a protein of 160 kDa with multiple domains, a proline-rich, a serine- and arginine-rich, and an RNase III domain. The expressed purified RNase III domain cleaves double-strand RNA and does not cleave single-strand RNA. The gene is ubiquitously expressed in human tissues and cell lines, and the protein is localized in the nucleus of the cell. The levels of transcription and translation of the protein do not change during different phases of the cell cycle. However, a significant fraction of the protein in the nucleus is translocated to the nucleolus during the S phase of the cell cycle. That this human RNase III is involved in processing of pre-rRNA, but might cleave at sites different from those described for yeast RNase III, is shown by antisense inhibition of RNase III expression. Inhibition of human RNase III expression causes cell death, suggesting an essential role for human RNase III in the cell. The antisense inhibition technique used in this study provides an effective method for functional analysis of newly identified human genes.

Variations in mRNA content have no effect on the potency of antisense oligonucleotides.

A fundamental question with regard to antisense pharmacology is the extent to which RNA content or transcription rate or both affect the potency of antisense drugs. We have addressed this by controlling RNA content and transcription rate using either an exogenous gene expressed after transfection or an endogenous gene induced with a cytokine. We have demonstrated that in both A549 and HeLa cells, varying RNA copy numbers from <1 to >100 copies per cell has no effect on the potency of RNase H-active antisense drugs transfected into cells, nor did variation in transcription rate have an effect on potency. We demonstrate that this is because the number of oligonucleotide molecules per cell is vastly in excess of the RNA copy number. These data further suggest that a significant fraction of cell-associated antisense drug molecules may be unavailable to interact with the target RNA, an observation that is not surprising, as phosphorothioate oligonucleotides interact with many cellular proteins. We suggest that these data may extrapolate to in vivo results.

Potential roles of antisense technology in cancer chemotherapy.

Antisense technology may play a major role in cancer chemotherapy. It is clearly a tool of exceptional value in the functionalization of genes and their validation as potential targets for cancer chemotherapy. Additionally, there is now substantial evidence that antisense drugs are safe, and a growing body of data showing activity in animal models of human disease including cancer, and suggesting efficacy in patients with cancer. In this article, I review the progress in the technology, the anticancer antisense drugs in development and potential roles that antisense technology might play.

Highly efficient endonucleolytic cleavage of RNA by a Cys(2)His(2) zinc-finger peptide.

We have identified a 30-aa peptide that efficiently cleaves single-stranded RNA. The peptide sequence corresponds to a single zinc finger of the human male-associated ZFY protein; a transcription factor belonging to the Cys(2)His(2) family of zinc-finger proteins. RNA cleavage was observed only in the absence of zinc. Coordination with zinc resulted in complete loss of ribonuclease activity. The ribonuclease active structure was determined to be a homodimeric form of the peptide. Dimerization of the peptide occurred through a single intermolecular disulfide between two of the four cystines. The observed hydrolytic activity was single-stranded RNA-specific. Single-stranded DNA, double-stranded RNA and DNA, and 2'-methoxy-modified sequences were not degraded by the peptide. The peptide specifically cleaved pyrimidines within single-stranded RNA and the dinucleotide sequence 5'-pyr-A-3' was preferred. The RNA cleavage products consisted of a 3' phosphate and 5' hydroxyl. The initial rates of cleavage (V(0)) observed for the finger peptide were comparable to rates observed for human ribonucleases, and the catalytic rate (K(cat)) was comparable to rates observed for the group II intron rybozymes. The pH profile exhibited by the peptide is characteristic of general acid-base catalytic mechanisms observed with other ribonucleases. These observations raise interesting questions about the potential biological roles of zinc-finger proteins.

Determinants of aminoglycoside-binding specificity for rRNA by using mass spectrometry.

We have developed methods for studying the interactions between small molecules and RNA and have applied them to characterize the binding of three classes of aminoglycoside antibiotics to ribosomal RNA subdomains. High-resolution MS was used to quantitatively identify the noncovalent binding interactions between mixtures of aminoglycosides and multiple RNA targets simultaneously. Signal overlap among RNA targets was avoided by the addition of neutral mass tags that direct each RNA target to a unique region of the spectrum. In addition to determining binding affinities, the locations of the binding sites on the RNAs were identified from a protection pattern generated by fragmenting the aminoglycoside/RNA complex. Specific complexes were observed for the prokaryotic rRNA A-site subdomain with ribostamycin, paromomycin, and lividomycin, whereas apramycin preferentially formed a complex with the eukaryotic subdomain. We show that differences in binding between paromomycin and ribostamycin can be probed by using an MS-MS protection assay. We have introduced specific base substitutions in the RNA models and have measured their impact on binding affinity and selectivity. The binding of apramycin to the prokaryotic subdomain strongly depends on the identity of position 1408, as evidenced by the selective increase in affinity for an A1408G mutant. An A1409-G1491 mismatch pair in the prokaryotic subdomain enhanced the binding of tobramycin and bekanamycin. These observations demonstrate the power of MS-based methods to provide molecular insights into small molecule/RNA interactions useful in the design of selective new antimicrobial drugs.

Properties of cloned and expressed human RNase H1.

We have characterized cloned His-tag human RNase H1. The activity of the enzyme exhibited a bell-shaped response to divalent cations and pH. The optimum conditions for catalysis consisted of 1 mM Mg(2+) and pH 7-8. In the presence of Mg(2+), Mn(2+) was inhibitory. Human RNase H1 shares many enzymatic properties with Escherichia coli RNase H1. The human enzyme cleaves RNA in a DNA-RNA duplex resulting in products with 5'-phosphate and 3'-hydroxy termini, can cleave overhanging single strand RNA adjacent to a DNA-RNA duplex, and is unable to cleave substrates in which either the RNA or DNA strand has 2' modifications at the cleavage site. Human RNase H1 binds selectively to "A-form"-type duplexes with approximately 10-20-fold greater affinity than that observed for E. coli RNase H1. The human enzyme displays a greater initial rate of cleavage of a heteroduplex-containing RNA-phosphorothioate DNA than an RNA-DNA duplex. Unlike the E. coli enzyme, human RNase H1 displays a strong positional preference for cleavage, i.e. it cleaves between 8 and 12 nucleotides from the 5'-RNA-3'-DNA terminus of the duplex. Within the preferred cleavage site, the enzyme displays modest sequence preference with GU being a preferred dinucleotide. The enzyme is inhibited by single-strand phosphorothioate oligonucleotides and displays no evidence of processivity. The minimum RNA-DNA duplex length that supports cleavage is 6 base pairs, and the minimum RNA-DNA "gap size" that supports cleavage is 5 base pairs.

Multiplexed screening of neutral mass-tagged RNA targets against ligand libraries with electrospray ionization FTICR MS: a paradigm for high-throughput affinity screening.

We demonstrate that binding of mixtures of aminoglycosides can be measured simultaneously against multiple RNA targets of identical length and similar (or identical) molecular weight. Addition of a neutral mass tag to one of the RNA targets shifts the detected peaks to a higher mass/charge ratio, where complexes with small molecules can be identified unambiguously. An appropriately placed neutral mass tag does not alter RNA--ligand binding. The utility of this strategy is demonstrated with model RNAs corresponding to the decoding region of the prokaryotic and eukaryotic rRNAs and a mixture of five aminoglycosides. Complexes are observed between the aminoglycoside library and the prokaryotic rRNA model, while no aminoglycoside was observed to bind to the mass-tagged eukaryotic rRNA model. The differential binding data is consistent with the eukaryotic A-site rRNA having a different conformation compared with the prokaryotic A-site that prevents entry and binding of neomycin-class aminoglycosides. Mass spectrometric analysis of neutral mass-tagged macromolecular targets represents a new high-throughput screening paradigm in which the interaction of multiple targets against a collection of small molecules can be evaluated in parallel.

Molecular mechanisms of action of antisense drugs.

Given the progress reported during the past decade, a wide range of chemical modifications may be incorporated into potential antisense drugs. These modifications may influence all the properties of these molecules, including mechanism of action. DNA-like antisense drugs have been shown to serve as substrates when bound to target RNAs for RNase Hs. These enzymes cleave the RNA in RNA/DNA duplexes and now the human enzymes have been cloned and characterized. A number of mechanisms other than RNase H have also been reported for non-DNA-like antisense drugs. For example, activation of splicing, inhibition of 5'-cap formation, translation arrest and activation of double strand RNases have all been shown to be potential mechanisms. Thus, there is a growing repertoire of potential mechanisms of action from which to choose, and a range of modified oligonucleotides to match to the desired mechanism. Further, we are beginning to understand the various mechanisms in more detail. These insights, coupled with the ability to rapidly evaluate activities of antisense drugs under well-controlled rapid throughput systems, suggest that we will make more rapid progress in identifying new mechanisms, developing detailed understanding of each mechanism and creating oligonucleotides that better predict what sites in an RNA are most amenable to antisense drugs of various chemical classes.

Depletion of protein kinase C-alpha by antisense oligonucleotides alters beta-adrenergic function and reverses the phorbol ester-induced reduction of isoproterenol-induced adenosine 3'-5'-cyclic monophosphate accumulation in murine Swiss 3T3 fibroblasts.

Beta-adrenergic agonists are well known to increase the activity of adenylate cyclase, yielding increases of the intracellular concentration of cAMP. It has been reported that activation of protein kinase C (PKC) by phorbol esters reduces the amplitude of isoproterenol-induced cAMP production in a 3T3-L1 cell line. In this study, we investigated whether PKC-alpha is involved in this process in murine Swiss 3T3 fibroblasts. A 20-mer phosphorothioate oligonucleotide designed to hybridize to the AUG initiation codon of the murine PKC-alpha mRNA, which contains 2'-O-methoxyethyl modifications incorporated into the 5' and 3' segments of the oligonucleotide, was used to assess the putative role of PKC-alpha in the beta-adrenergic receptor regulation. ISIS 14012 reduced PKC-alpha mRNA for over 72 hr after the initial treatment and the reduction was concentration dependent, whereas the mismatch control, ISIS 13818, had no effect. This depletion was found to be selective; ISIS 14012 had no effect on the mRNA expression of PKC-delta and PKC-zeta. ISIS 14012 reduced in a time and concentration-dependent fashion the levels of immunoreactive PKC-alpha protein by over 85% at 72 hr after treatment. Depletion of PKC-alpha inhibited the effect of isoproterenol-induced cAMP production by phorbol dibutyrate (PdBu). This finding is corroborated by the use of a nonspecific inhibitor of PKC, GF-109203x, which also prevented the effect of PdBu. Depletion of PKC-alpha by ISIS 14012 potentiated isoproterenol-induced cAMP production in cells untreated with PdBu. However, neither depletion of PKC-alpha nor PKC activation by a phorbol ester altered beta-adrenergic receptor affinity and density. PKC activation by PdBu did not alter forskolin-induced cAMP levels, but enhanced cAMP production by cholera toxin. PKC-alpha inhibition by ISIS 14012 had no effect on either cholera toxin-induced increases in cAMP or the acute effects of phorbol esters on cholera toxin in induction of cAMP. Thus, PKC-alpha appears to be involved in the regulation of beta-adrenergic receptor coupling to adenylate cyclase, possibly by phosphorylating the Gs protein, but other PKC isotypes must be involved in the effects observed when cells are treated with cholera toxin.

In vivo distribution and metabolism of a phosphorothioate oligonucleotide within rat liver after intravenous administration.

In the rat, the liver represents a major site of phosphorothioate oligodeoxynucleotide deposition after i.v. administration. For this reason, we examined the intracellular fate of ISIS 1082, a 21-base heterosequence phosphorothioate oligodeoxynucleotide, isolated from parenchymal and nonparenchymal cell types after systemic dosing using established perfusion and separation techniques followed by CGE. Isolated cells were further fractionated into nuclear, cytosolic and membrane constituents to assess the intracellular localization, distribution and metabolic profiles as a function of time and dose. After a 10-mg/kg i.v. bolus, intracellular drug levels where maximal after 8 hr and diminished significantly thereafter, suggesting an active efflux mechanism or metabolism. Nonparenchymal (i.e., Kupffer and endothelial) cells contained approximately 80% of the total organ cellular dose, and this was equivalently distributed between the two cell types, while the remaining 20% was associated with hepatocytes. Nonparenchymal cells contained abundant nuclear, cytosolic and membrane drug levels over a wide dose range. In contrast, at doses of less than 25 mg/kg, hepatocytes contained significantly less drug with no detectable nuclear-association. Doses at or above 25 mg/kg appeared to saturate nonparenchymal cell types, whereas hepatocytes continued to accumulate drug in all cellular compartments, including the nucleus. Our results suggest that although pharmacokinetic parameters vary as a function of hepatic cell type, significant intracellular delivery can be readily achieved in the liver after systemic administration.