Antisense technology is based on the principle that antisense oligonucleotides (AO) can specifically alter gene expression and the pattern of splicing at the RNA level. The first application of such technology 25 years ago was for knockdown of targeted viral gene expression (Zamecnik et al., 1978), which was achieved by targeting mRNA with AO to block the viral gene translation or induce its degradation. Such strategies have now been developed as powerful tools for the functional analysis of targeted genes in the laboratory and are suggested to be of great therapeutic potential against diseases from viral infections to cancers. More recently, antisense technologies have been devised for interference with gene splicing (Figure 1). Such applications target pre-mRNA rather than mRNA and was first established in vitro for correction of altered exon splicing in β-thalassemia. Mutations in the β-globin gene activate a cryptic site in the pre-mRNA and as a result the intronic sequences are spliced into the mRNA. This causes disruption of normal protein synthesis. Kole and his colleagues designed specific AO to the cryptic splicing site and successfully blocked its recognition and restored the normal splicing pattern of the gene (Kole et al., 1993). The use of AO for selective exon skipping was quickly explored by several laboratories to correct gene mutations for the treatment of Duchenne muscular dystrophy (DMD).
DMD is a muscle degenerative disorder caused by nonsense or frame-shift mutations in the dystrophin gene. The disease affects 1 in every 3,500 live male births and is characterized by severe muscle wasting and weakness, which becomes clinically evident between the ages of 3 to 5 years. Affected individuals stop walking by 12 years of age and have a poor prognosis. A milder allelic form of the disease, Becker muscular dystrophy (BMD), is commonly caused by deletions or duplications in the same dystrophin gene, which, however, do not disrupt the reading frame, resulting in altered, but in-frame, and therefore translatable transcripts. BMD has a spectrum of phenotypes ranging from almost asymptomatic to the mild form of DMD with loss of ambulation in the late teens or early 20s.
The human dystrophin gene comprises 79 exons spanning more than 2.3 million base pairs. The muscle form of the dystrophin protein has 3,685 amino acids (427kd) and it can be divided into amino (N) terminal, rod, cysteine-rich and carboxy (C) terminal domains. The N terminal domain binds to the cytoplasmic actin filaments and the cysteine-rich domain to dystrophin-associated protein (DAP) complexes including dystroglycans, sarcoglycans and syntrophins, through which dystrophin connects itself to extracellular matrix components. The enormous size of the gene and the involvement in DMD of body-wide muscles including the heart and diaphragm, create huge hurdles for developing effective approaches of gene therapy and cell-based therapy.
However, DMD presents an unparalleled prospect for gene correction by alternative splicing. First, the rod domain of the dystrophin gene appears to be not critical for its functions. For instance, deletion of exons from 17-49 was associated with only a mild clinical phenotype (England et al., 1990). Similarly, an artificially constructed microdystrophin with deletion from exon 18 to 57 remains largely functional. Second, the majority of DMD mutations occur within this non-critical region of the dystrophin gene. Thus correction of frame-shift mutations in this region by skipping the mutated exon or other exons necessary for restoration of the reading frame will retain critical functions of the protein (Figure 1). Furthermore, dystrophin positive muscle fibres can be found in many DMD muscles. These so called “revertant fibres” appear to be the product of altered splicing that skips mutated exons and exons which disrupt the reading frame of the endogenous gene (Lu et al., 2000). Such observations suggest a high susceptibility of the dystrophin gene to alternative splicing. The presence of revertant dystrophin may also provide the immune tolerance to the dystrophin induced by AO.
In the last few years, experiments in several laboratories using myogenic cell cultures have demonstrated the principle that sequence-specific AO can induce targeted exon skipping to re-establish the reading frame of dystrophin mRNA (Mann et al., 2001). The effectiveness of such strategies for the treatment of dystrophinopathy has now been demonstrated in mdx mice, an animal model of DMD. mdx dystrophic mice have a nonsense point mutation in exon 23 of the dystrophin gene and, as a result, no dystrophin is expressed in the muscles apart from a small number of revertant fibres. AO delivered by intramuscular injection is able to effectively skip the mutated exon 23 and restore the reading frame. As a result, the treated muscle produces persistent dystrophin at normal levels in a large numbers of fibres resulting in functional improvement (Figure 2). Dystrophin induced by AO remains at detectable levels even 3 months after intramuscular injections. Furthermore, repeated administration enhances the dystrophin expression without eliciting immune responses (Lu et al., 2003).
The high efficiency of AO induced exon skipping in vivo in our study is the result of several improvements to the antisense approach. Factors known to be especially important for effective antisense-induced exon skipping are: 1) optimal sequence of AO for the targeted exon; 2) stability of AO, which is dependent on the type of chemical modifications; and 3) vector systems which determine delivery efficiency of AO into targeted cells. Use of an effective AO specific to the targeted exon is crucial for the success of antisense therapy. There is, however, no satisfactory method to design and select such an AO, although several methods have been described. For example, nucleic acid array can be used to find out the most effective binding site for targeting RNA, but the structure and accessibility of RNA in a living cell are likely to be different from those of purified RNA or in vitro transcribed RNA. The most reliable approach of selecting effective AO so far is still the screening of a large number of candidate AO in a relevant cell culture system. For example, the effective AO used in our study for the skipping of mutated exon 23 in the dystrophin gene was discovered through such tedious in vitro testing and selections.
Oligo DNA or RNA without chemical modification is unstable and is quickly degraded by DNase or RNase H. This becomes even more problematic when AO is injected into either the tissue or bloodstream, where high levels of such enzyme activity would be expected and be difficult to suppress. One of the major challenges for antisense therapy is therefore to stabilize AOs in biological systems. One most promising solution is chemical modification, which has been a focus in the development of antisense strategies, particularly for the antisense induced gene knockdown. There have been a large number of chemically modified nucleotides described, each having its own unique properties. In our recent study, the use of 2′-O-methyl phosphorothioated antisense ribonucleotide (2OMeAO) provides a significant improvement in the induction of exon skipping over the use of unmodified or phosphorothioated antisense nucleotides which did not induce detectable levels of dystrophin expression. 2OMeAO belongs to the second generation of chemically modified AO with improved target binding affinity and a higher resistance to RNase H degradation. The majority of other chemically modified AOs have yet to be examined in exon skipping models and some of them might be more effective than 2OMeAO. The potential of each class of AO may be realized only when they are used in combination with different delivery systems and also depend on the types of target cells or tissues.
A major factor determining the effectiveness of exon skipping in vivo is the efficiency of the delivery system. In our recent study, expression of near-normal levels of dystrophin was reliably achieved only when 2OMeAO was delivered with the aid of the non-ionic tri-block co-polymer F127, listed in the pharmacopoeia as an “(inactive) excipient” and widely used for drug delivery. We have previously found that this compound is the most effective block co-polymer for improving plasmid delivery into muscle. F127 is the least toxic among the block co-polymers tested and causes no microscopic damage outside the needle track on intramuscular injection and, for delivery of small nucleic acids into muscle it appears better suited than any other known reagents or delivery systems, such as electroporation. Although not as efficient as recombinant AAV and AV vectors, and marginally less efficient than electroporation for large plasmids, F127 shows a similar efficiency for AO delivery to that achieved by electroporation (Wells et al., 2003). The reagent is cheap, the preparation is simple, and the results are highly reproducible. Furthermore, the known chemical composition of block co-polymers permits a systematic investigation of their mode of action as a basis for rational design of improved variants that might achieve vascular delivery. Development of such low-toxicity compounds for intracellular delivery of nucleic acids has so far been limited, but could prove to be more rewarding than the development of other delivery systems such as viral vectors and electroporation.
In summary, exon skipping induced by antisense oligonucleotide has now established the realistic practicality of an approach that is applicable, in principle, to a majority of cases of DMD as well as other diseases in which exon-splicing might be used to circumvent nonsense and frame-shift mutations. Antisense therapy is limited in its duration of maintaining efficacy and would thus have to be administered at regular intervals for diseases such as DMD. However, this feature can also be advantageous for clinical trials since it permits withdrawal in case of any potential adverse effects and further optimizations can be introduced at any stage. Clinical trials of antisense therapy for DMD by local muscle injection will probably be conducted within the next 2 years. Further improvements in the chemical modification of AO, the efficiency of delivery systems and optimization of the AO sequence will be the key factors for achieving therapeutic effect in patients.
Restoration of correct splicing in thalassemic pre-mRNA by antisense oligonucleotides.
Dominski Z and Kole R.
Department of Pharmacology, University of North Carolina, Chapel Hill, NC, USA.
PNAS 90:8673-8677, 1993.
Summary: The authors used antisense 2′-O-methyl ribo-oligonucleotides to target specific sequence elements in mutated human β-globin pre-mRNAs to restore correct splicing of these RNAs in vitro. Some mutations of the b-globin gene cause aberrant splicing and were identified as the underlying causes of β-thalassemia. Correct splicing was restored by oligonucleotides targeted against the aberrant 5′ splice sites created by the mutations in the second intron or against a cryptic 3′ splice site located upstream and activated in the mutated gene. The experiments represent the first attempt to restore the function of a defective gene by alternation in splicing.
Very mild muscular dystrophy associated with deletion of 40% dystrophin.
England SB, Nicholson LV, Johnson MA, Forrest SM, Love DR, Zubrzycka-Gaarn EE, Bulman DE, Harris JB, Davies KE.
Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford, UK.
Nature 343:180-182, 1990.
Summary: Authors reported large deletional, but in-frame mutations in dystrophin gene in BMD with mild clinical phenotypes.
Massive idiosyncratic exon skipping corrects the nonsense mutation in dystrophic mouse muscle and produces functional revertant fibers by clonal expansion.
Lu QL, Morris GE, Wilton SD, Ly T, Artem’yeva OV, Strong P, Partridge TA.
Muscle Cell Biology, Medical Research Council Clinical Science Center, Hammersmith Hospital, London, UK.
J Cell Biol 148:985-996, 2000.
Summary: The authors show that reversion of dystrophin expression in mdx mice muscle involves unprecedented massive loss of up to 30 exons and exon skipping in two non-contiguous regions. Genomic deletion from the region commonly skipped was not detected. Several alternatively processed transcripts that could account for some of the revertant dystrophins were detected. The authors suggested that revertant dystrophins are functional and aberrant splicing is the most likely mechanism for the restoration of dystrophin.
Functional amounts of dystrophin produced by skipping the mutated exon in the mdx dystrophic mouse.
Lu QL, Mann CJ, Lou F, Bou-Gharios G, Morris GE, Xue SA, Fletcher S, Partridge TA, Wilton SD.
Muscle Cell Biology, MRC Clinical Science Centre, Hammersmith Hospital, London, UK.
Nat Med 9:1009-1015, 2003.
Summary: The authors reported antisense therapy in vivo in the mdx dystrophic mouse for skipping a nonsense mutation in exon 23 by combining a potent transfection protocol with a 2′-O-methyl phosphorothioated antisense oligoribonucleotide (2OMeAO). The treated mice show persistent production of dystrophin at normal levels in large numbers of muscle fibers and functional improvement of the treated muscle. Data established the realistic practicality of an approach that is applicable, in principle, to a majority of cases of severe dystrophinopathy.
Antisense-induced exon skipping and synthesis of dystrophin in the mdx mouse.
Mann CJ, Honeyman K, Cheng AJ, Ly T, Lloyd F, Fletcher S, Morgan JE, Partridge TA, Wilton SD. Australian Neuromuscular Research Institute, Centre for Neuromuscular and Neurological Disorders, University of Western Australia, Perth, Western Australia, Australia.
PNAS 98:42-47, 2001.
Summary: The authors reported the use of 2′-O-methyl antisense oligoribonucleotides to block motifs involved in normal dystrophin pre-mRNA splicing in cell culture and the mdx mouse model. Skipping exon 23 in the dystrophin gene was first optimized in vitro in transfected H-2K(b)-tsA58 mdx myoblasts and then induced in vivo. mRNA with exon 23 skipped was induced at high levels in vitro and production of dystrophin protein was detectable in vivo.
Enhanced in vivo delivery of antisense oligonucleotides to restore dystrophin expression in adult mdx mouse muscle.
Wells KE, Fletcher S, Mann CJ, Wilton SD, Wells DJ.
Department of Neuromuscular Diseases, Imperial College London, Charing Cross Hospital, London, UK.
FEBS Lett 552(2-3):145-149, Sep. 25, 2003.
Summary: The Authors described the use of hyaluronidase-enhanced electrotransfer to deliver uncomplexed 2′-O-methyl modified phosphorothioate AO to adult dystrophic mouse muscle. They reported dystrophin expression in 20-30% of fibres in tibialis anterior muscle after a single AO injection.
Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide.
Zamecnik PC and Stephenson ML.
PNAS 75:280-284, 1978.
Summary: The authors added tridecamer d(A-A-T-G-G-T-A-A-A-A-T-G-G), which is complementary to 13 nucleotides of the 3′- and 5′-reiterated terminal sequences of Rous sarcoma virus 35S RNA, to chick embryo fibroblast tissue cultures infected with Rous sarcoma virus, and reported inhibition of virus production.
[Discovery Medicine, 3(19):39-44, 2003]