Modes of Action of Antisense Oligonucleotides
There are several main intracellular mechanisms of actions of antisense oligonucleotides to prevent the expression and translation of the target protein (Dias and Stein 2002, Fig. 1a-e). The first mechanism is based on the disruption of protein translation of target mRNA by base-pairing of single-stranded antisense oligodesoxynucleotides (► antisense DNA) to the respective complementary mRNA strand and physically/sterically obstructing the translation machinery, prevention of ribosomal complex formation, and translation of mRNA into the gene product (transla-tional arrest) (Fig. 1b). The second main mechanism is also based on the complementary binding between the antisense oligodesoxynucleotides (antisense DNA) and the target mRNA. This DNA/RNA hybrid can then be degraded by the enzyme RNase H, a mechanism which significantly enhances the efficacy of antisense potency (Fig. 1c). The RNase endonuclease specifically cleaves RNA in DNA/RNA heteroduplexes. Short-term effects of antisense oligodesoxynucleotides inhibiting neuronal responsiveness without altering gene protein content and availability have been described in neuroendocrine cells (Fig. 1d) (Neumann and Pittman 1998).
Also, ► ribozymes have antisense-like properties (Fig. 1d); those RNA molecules bind to and cleave their target mRNA.
A revolutionary development in antisense research is the finding that 20-25 nucleotide double-stranded RNAs, called siRNA, can efficiently block gene expression (Figs. 1e and 2) via the ► RNA interference (► RNAi) pathways. Any gene of interest with known sequence can be targeted by siRNAs that match the mRNA sequence.
A main problem for antisense technologies is the limitation of cellular uptake of these nucleic acids due to their hydro-philic nature. This is true for negatively-charged siRNA, or chemically modified antisense oligodesoxynucleotides with uncharged backbones such as methylphosphonates, phosphorothioates, or morpholino-derivatives.
For improved efficacy of antisense uptake, agents that enhance transmembrane permeation, i.e., cell-penetrating peptides (CPP) and drugs that target specific receptors, i.e., cell-targeting ligands (CTL) are discussed (Juliano et al. 2008). Also, antisense oligonucleotides tend to localize in lysosomes and endosomes where their antisense properties are poor. The use of vectors (liposomes, i.e., vesicular colloid vesicles) increases stability and cellular internalization. There are commercially available vectors (so-called eufectins), which are used in basic research.
In vivo, the cellular uptake of antisense oligonucleo-tides into the target tissue is further impaired by several biological barriers (capillary endothelium, extracellular matrix), degradation by serum and tissue nucleases (liver, spleen), and rapid excretion via the kidney; but if oligonucleotides are bound to plasma proteins, ultrafiltration is retarded (Juliano et al. 2008). In order to target the brain tissue, the blood-brain barrier prevents their uptake as polyanionic molecules, and antisense targeting the central nervous system (CNS) requires direct infusion.
If brain tissue is targeted, antisense oligonucleotides need to be directly infused either into the ventricular system or the target region to circumvent the ► blood-brain barrier; and they are relatively easily taken up by neurons. Prolonged action and slow delivery within the target region can be achieved by use of biotechnical modifications or polymer microparticle encapsulation with high biocom-patibility (Choleris et al. 2007).
Antisense Oligonucleotides. Fig. 1. Mechanisms of action of antisense oligonucleotides. (A) Regular transcription of a gene and translation of mRNA into the gene product. (B) Antisense oligonucleotides bind to the complementary strand of gene target mRNA and (B1), sterically block the translation process or (B2), the mRNA/oligonucleotide complex activates an RNase H cleaving the mRNA and preventing translation. (C) Short-term effects of antisense oligonucleotides on neuronal responsiveness via unknown mechanisms. (D) DNA and RNA enzymes (ribozymes) result in cleavage of mRNA. (E) siRNA activate intracellular mechanisms resulting in mRNA cleavage and translational arrest via the RNA interference (RNAi) pathway (see Fig. 2).
The most common strategy for siRNA in vivo delivery is to construct expression vectors driven by a constitutive-ly active RNA polymerase III (Pol III) promoter (e.g., the U6 promoter), to drive transcription of small hairpin RNA (shRNA). shRNA are sequences of RNA that make tight hairpin turns upon intramolecular Watson-Crick base-pairing (Fig. 2b), with perfect duplex RNA of 18-25 nucleotides in length. Cellular enzymes, most likely involving Dicer, process shRNA into siRNA molecules that are capable of performing gene silencing via the RNAi pathway. Although plasmid vectors are effective at delivering shRNA and consequently, siRNA to cultured cells and for the generation of transgenic plants or mice, there are limitations for their use, e.g., in somatic cells of adult brain tissue. To overcome these limitations, viral vectors for the delivery of siRNA have been developed, including lentiviral vectors derived from human immunodeficiency virus-1 (HIV-1) and adenoviral and adeno-associated viral vectors that are used to deliver siRNA into selected brain regions of rodents (Kuhn et al. 2007).
Chemical Modifications of Antisense Oligonucleotides
Antisense oligonucleotides are rapidly degenerated by intracellular enzymatic activity. In order to increase efficiency, the cellular uptake and intracellular stability can be
Antisense Oligonucleotides. Fig. 2. Specific knockdown of a target mRNA by siRNA via the RNAi pathway. (a) Molecular hallmarks of an siRNA molecule are 18- to 23-nucleotide duplexed RNA region with 2-nt unphosphorylated 3' overhangs and 5' phosphorylated ends. (b) Mechanism of inhibition of gene expression by siRNA. Long siRNA or shRNA are cleaved by the RNase III family member Dicer to produce siRNA molecules. Alternatively, siRNA can be chemically synthesized and transfected into cells. These siRNAs are then unwound by RISC and incorporated as single-stranded antisense RNA to guide RISC to mRNA transcripts of complementary sequence. This leads to selective endonucleolytic cleavage of the matching target mRNA with subsequent elimination by further cellular RNases.
improved by ► chemical modifications of antisense oligonucleotides (Dias and Stein 2002). The most common chemical modifications employed are replacing an oxygen group of the phosphate-diester backbone with either a methyl group (methyl phosphonate oligonucleotide) or a sulfur group (phosphorothioate oligonucleotide) which have been introduced into clinical therapeutic trials. Increased specificity and efficacy of phosphorothioates are achieved with chimeric oligonucleotides in which the RNase H-competent segment (the phosphorothioate moiety) is bounded on both termini by a high-affinity region of modified RNA. Other chemical modifications include 2'-OH modifications, locked nucleic acids, peptide nucleic acids, and morpholino compounds or hexitol nucleic acids. Although high RNA affinity and high stability have been reported, they do not support RNase H activation. Nevertheless, they can exert their antisense activity via translational arrest or modulation of ► alternative splicing.
Antisense oligonucleotide technologies including oligode-soxynucleotides and siRNA are used in psychopharmacology to sequence-specifically, transiently, and locally downregulate the expression of target genes (neuro-peptides and their receptors, neurotransmitter receptor subtypes and subunits, steroid receptors, neuronal enzymes, transcription factors), identified by microarray and proteomic approaches or by human genetic studies in vivo in addition or alternatively to established pharmacological means (selective ► agonists and ► antagonists, generation of knockout mice) (Hoyer et al. 2006; Landgraf et al. 1997; McCarthy 1998) for detailed methodological description (Fendt et al. 2008).
Functional consequences of antisense-induced down-regulation of neuronal gene products in the brain can be monitored in a behavioral, neuroendocrine, or neuronal context. Antisense oligonucleotides need to be directly infused into the brain ventricles or into a selected brain target region over several days using an osmotic minipump or repeated infusions. They are relatively easily taken up by neurons, have a relatively high stability in neurons and in CSF - likely due to negligible cell proliferation within the brain and lack of exo- and endonucleases within the CSF (McCarthy 1998). Nevertheless, in order to increase central efficiency, mainly phosphodiester and phosphorothioated antisense oligonucleotides are used in psychopharmacology, but unspecific effects were repeatedly described making respective controls essential.
Both acute effects on neuronal functions after single infusion as well as long-term effects after repeated or chronic administration have been described. Often, short-term, antisense-induced blockade of neuronal activation without altering the availability of the gene product (Figs. 1 and 2) is likely to contribute to the observed behavioral, neuroendocrine, and neuronal effects of antisense oligonucleotides, especially if neuropeptides are targeted and stored in rather large amounts (Neumann and Pittman 1998).
Various neuropeptides and their receptors have been targeted using antisense oligonucleotides in the absence of highly selective receptor antagonists (Hoyer et al. 2006; Landgraf et al. 1997; McCarthy 1998), including the receptors for neuropeptide Y1, corticotropin-releasing hormone (CRH), oxytocin and its receptor, vasopressin
Antisense Oligonucleotides. Fig. 3. Antisense oligodesoxynucleotides targeting the brain prolactin receptor (PRL-R-AS, icv, 6 days) reduced PRL-R density in the choroid plexus, increased anxiety-related behavior on the plus-maze, and elevated the acute stress-induced rise in plasma ACTH in virgin rats. (Data from Torner et al. 2001.)
and its Via receptor, or the long form of prolactin receptors (Fig. 3).
Antisense oligonucleotides have also been used to target the expression of other neurotransmitter and steroid receptors including the receptors for 5-HT2A, dopamine (D2), progesterone, as well as NMDA receptor subunits and metabotropic glutamate receptor subtypes. Also, immediate early genes (c-fos, c-jun) and enzymes essential for neurotransmitter synthesis (GABA, 5-HT) have been successfully targeted in order to reveal local effects and behavioral and other functional consequences.
To exclude sequence-independent, unspecific effects of mostly chemically modified antisense oligonucleotides in vivo and in vitro, appropriate control groups are essential, i.e., the use of oligonucleotides with identical chemical modification and length which do not match a particular gene. Possible controls include a sense-strand sequence from the same site of mRNA, a scrambled sequence of the nucleotides used in the antisense oligo-nucleotides in random order, a nonsense oligonucleotide of same length but consisting of a different nucleotide composition without complementarity to any known mRNA sequence.
siRNA in Psychopharmacology
Among the earliest studies of siRNA in the brain were the attempts to perform icv infusion of synthetic siRNA using osmotic minipumps in rodents and concomitant behavioral analysis (Hoyer et al. 2006). To establish this strategy, enhanced green fluorescent protein overexpressed in a transgenic mouse line was successfully targeted in several brain regions. Also, targeting the dopamine transporter
(DAT) resulted in reduced DAT mRNA and protein levels in substantia nigra, ventral tegmental areas (VTAs), and nucleus accumbens (Hoyer et al. 2006) accompanied by behavioral changes. Additional proof of concept for siRNA in rodent models of neuropsychiatric disease was provided by siRNA-induced reduction in serotonin transporter (SERT) mRNA in the raphe nucleus after 2 weeks of icv siRNA infusion with antidepressant-like behavioral consequences.
In addition, targeting the ► metabotropic glutamate receptor subtype 7 (mGluR7) by selective siRNA resulted in robust emotional and stress-related changes despite limited local mRNA knockdown (25%) (Fendt et al. 2008).
The use of viral vectors for RNAi in the mammalian brain is particularly suited for local gene knockdown even in small brain nuclei (Hoyer et al. 2006).
The major advantages of siRNA over classical antisense oligonucleotides and ribozymes appear to be threefold (Zamecnik and Stephenson 1978): siRNA is generally more potent, efficient, and selective because it utilizes natural and efficient cellular machinery, i.e., the RNAi pathway providing assistance at multiple steps for the actions of siRNA, while antisense oligonucleotides and ribozymes hybridize to their targets without any assistance (Dias and Stein 2002). siRNA approaches always result in efficient endonucleolytic cleavage and subsequent elimination of the target mRNA, whereas antisense oligonucleotides not recognized by RNase H result in translational block only via steric hindrance (Neumann and Pittman 1998). Antisense oligonucleotides require relatively high concentrations to target their matching mRNA because at least one oligonucleotide per target RNA is required for translational blockade or RNase H degradation resulting in toxic side effects. In contrast, one guide strand of siRNA incorporated in RISC (RNA-induced silencing complex) can sequentially bind to and eliminate several mRNA transcripts (multiple turnover).
Limitations of siRNA approaches include (Zamecnik and Stephenson 1978) their capacity to induce cellular virus defense mechanisms, mainly interferon gene induction and cellular arrest and (Dias and Stein 2002) possible unspecific off-target effects on closely related sequences making careful design of the siRNA sequence and use of different siRNA molecules targeting the same mRNA transcript essential.
Was this article helpful?