Making Antisense of Splicing

Antisense oligonucleotides have been used to correct cryptic splicing in primary transcripts encoded by mutated human genes. A well-studied group of mutations caused by cryptic splicing is found in patients with P-thalassemias. These are caused by mutations within intron 2 of bglobin that activate cryptic splice sites and lead to expression of abnormal bglobin mRNA and result in low levels of hemoglobin A (Hb A; Fig 1C). The use of these cryptic splice sites in erythroid cells in culture has been abrogated by blocking antisense oligonucleotides (Lacerra, Sierakowska et al. 2000; Vacek, Sazani et al. 2003). In a similar approach Hb A restoration was achieved using larger antisense RNA molecules synthesized in the targeted cells (Gorman, Suter et al. 1998). These RNAs were engineered to contain small nuclear structural features of RNAs (snRNAs), which presumably help in localizing the RNAs to the nucleus and may also enhance interactions with splicing factors (Gorman, Suter et al. 1998). The antisense oligonucleotides and RNAs exert their effect by binding to the cryptic splice sites and sterically blocking the binding of splicing factors (e.g., binding of U1 snRNP to the cryptic 5' splice site).

Antisense oligonucleotides directed at splice sites have also been used to alter the splicing of cystic fibrosis transmembrane conductance regulator (CFTR; Friedman, Kole et al. 1999), dystrophin (Pramono, Takeshima et al. 1996; Dunckley, Manoharan et al. 1998; Wilton, Lloyd et al. 1999; De Angelis, Sthandier et al. 2002), and cyclophilin transcripts (Liu, Asparuhova et al. 2004). Recently this technology has been applied to Hutchinson-Gilford Progeria Syndrome (HGPS), a disease that results in accelerated aging (Scaffidi and Misteli 2005). A common HSPG-causing mutation creates a cryptic 5' splice site in exon 11 of the gene-encoding lamin A/C (LMNA). Use of this cryptic site leads to formation of a dominant negative form of lamin A, 50 lamin A, which in turn results in profound cellular abnormalities (De Sandre-Giovannoli, Bernard et al. 2003; Eriksson, Brown et al. 2003). The molecular and cellular defects observed in fibroblasts of HSPG patients were reversed by blocking the use of the exon 11 cryptic 5' splice site with an antisense morpholino oligonucleotide (Scaffidi and Misteli 2005). Oligonucleotides targeting splice sites of alternatively used exons have been employed to alter splicing choices in Bcl-x (Taylor, Zhang et al. 1999), interleukin-5 receptor-alpha (Karras, Maier et al. 2001), and tau pre-mRNAs (Kalbfuss, Mabon et al. 2001; Fig 1C).

The studies described above target pre-mRNA sequences at or near the splice sites, however antisense oligonucleotides can be used to block splicing enhancers or silencers, which can be found in introns or exons quite far away from the splice sites (Takeshima, Wada et al. 2001; Bruno, Jin et al. 2004). Inhibition of splicing silencers (ESSs or ISSs) is of particular interest since it provides a way to activate otherwise repressed exons. This was achieved for the a exon of fibroblast growth factor receptor-1 (FGFR1) transcripts

(Bruno, Jin et al. 2004). The FGFR1 a exon is silenced by the action of two ISSs, and this silencing leads to the formation of an isoform of the receptor that predominates in glioblastomas. Bruno et al used antisense morpholino oligonucleotides to block these silencer elements in glioblastoma cells in culture promoting the inclusion of the a exon (Bruno, Jin et al. 2004).

While all of the work described above was carried out in tissue culture, proof of principle for splicing disruption by antisense oligonucleotides in animals had also been obtained. The first such system used transgenic mice harboring an EGFP reporter that was interrupted with the second intron from the Pglobin thalassemia-654 gene. This thalassemic mutation leads to cryptic splicing, interruption of the EGFP open reading frame, and low EGFP production (Sazani, Gemignani et al. 2002). Cryptic splicing was reduced using antisense oligonucleotides that blocked the splice sites and this resulted in increased levels of EGFP fluorescence (Sazani, Gemignani et al. 2002). This animal model has been very useful to study the effectiveness of different oligonucleotides.

Perhaps the most striking use of therapeutic oligonucleotides to alter splicing has been achieved in mouse models for Duchenne muscular dystrophy (DMD), which is an X-linked recessive disorder that leads to muscle wasting and weakness (OMIM # 310200). Severe DMD is usually associated with nonsense mutations in the dystrophin gene that lead to the complete absence of the protein, whereas milder allelic variants of DMD, such as Becker muscular dystrophy (BMD), are usually caused by internal truncations of the protein that retain partial function (Muntoni, Torelli et al. 2003). The milder phenotype of these truncations suggested that forcing the skipping of internal exons, which harbored nonsense codons, could be advantageous (Pramono, Takeshima et al. 1996; Dunckley, Manoharan et al. 1998; Wilton, Lloyd et al. 1999; reviewed in Aartsma-Rus, Janson et al. 2004). These potential therapies have been tested in mdx mice, which have a nonsense mutation in exon 23 in the dystrophin gene and are an animal model of human DMD (Sicinski, Geng et al. 1989). Dystrophin transcripts that include exon 23 encode a non-functional truncated dystrophin (and are also subject to nonsense mediated decay), whereas those that are missing exon 23 lead to the production of partially active dystrophin protein. Mdx mice were treated by intramuscular injection or transfection with 2'-OMe antisense oligonucleotides that blocked the splice sites of exon 23 to induce skipping of this exon (Mann, Honeyman et al. 2001; Lu, Mann et al. 2003; Wells, Fletcher et al. 2003). The oligonucleotide treated mdx mice show normal levels of dystrophin production in many muscle fibers and improved muscle function (Mann, Honeyman et al. 2001; Lu, Mann et al. 2003); however, the effects were restricted to a limited area and were short-lived. Recently two groups have attempted to deal with these issues. In one case a U7 snRNA was modified to include antisense sequences targeting the branchpoint sequence upstream of exon 23 and 5' splice site of this exon and was transduced effectively to a high number of muscle fibers by intramuscular or intra-arterial delivery of an adeno-associated viral (AAV)

vector (Goyenvalle, Vulin et al. 2004). This resulted in the skipping of exon 23 and a concomitant rescue of dystrophin expression and function (Goyenvalle, Vulin et al. 2004). This delivery method also led to sustained expression of the exon 23 dystrophin up to 13 weeks after injection of the AAV encoding the antisense U7 snRNA (Goyenvalle, Vulin et al. 2004). Similar recovery of expression and function in mdx mice was observed by repeated systemic intravenous administration of 2'- OMe oligonucleotides that block the 5' splice site of exon 23 (Lu, Rabinowitz et al. 2005). This study did not examine expression or function beyond two weeks after the last oligonucleotide injection, and thus the persistence of the effect could not be gleaned from the data shown. The authors examined the levels of dystrophin expression and noted significant differences among different muscles with almost complete lack of expression in the heart (Lu, Rabinowitz et al. 2005).

These impressive studies give cause for optimism as they demonstrate the potential for the antisense approach to alter splicing in a therapeutic manner. Although rigorous determination of specificity for these methods is still lacking the studies did not reveal any significant toxicity (Lu, Rabinowitz et al. 2005) and it is very likely that we will see clinical trials in the very near future. Given sequence variation between mouse and human DMD genes the mdx mice cannot be used to test the sequence dependence of antisense formulations. To overcome this problem a second mouse model, the hDMD mouse, was engineered to harbor a single full-length human wt DMD gene, and this mouse was used to show exon skipping mediated by 2'-OMe oligonucleotides (Bremmer-Bout, Aartsma-Rus et al. 2004). Obviously, diseases such as DMD are complex and correcting the skeletal muscle defects may only partially alleviate the plight of the patients, however, more sophisticated vehicles could eventually bring the therapeutics to most if not all the tissues affected.

0 0

Post a comment