Comparative analysis of antisense oligonucleotide sequences for targeted skipping of exon 51 during dystrophin pre-mRNA splicing in human muscle.

Duchenne muscular dystrophy (DMD) is caused by mutations in the dystrophin gene that result in the absence of functional protein. In the majority of cases these are out-of-frame deletions that disrupt the reading frame. Several attempts have been made to restore the dystrophin mRNA reading frame by modulation of pre-mRNA splicing with antisense oligonucleotides (AOs), demonstrating success in cultured cells, muscle explants, and animal models. We are preparing for a phase I/IIa clinical trial aimed at assessing the safety and effect of locally administered AOs designed to inhibit inclusion of exon 51 into the mature mRNA by the splicing machinery, a process known as exon skipping. Here, we describe a series of systematic experiments to validate the sequence and chemistry of the exon 51 AO reagent selected to go forward into the clinical trial planned in the United Kingdom. Eight specific AO sequences targeting exon 51 were tested in two different chemical forms and in three different preclinical models: cultured human muscle cells and explants (wild type and DMD), and local in vivo administration in transgenic mice harboring the entire human DMD locus. Data have been validated independently in the different model systems used, and the studies describe a rational collaborative path for the preclinical selection of AOs for evaluation in future clinical trials.

Recombinant adeno-associated viral (rAAV) vectors as therapeutic tools for Duchenne muscular dystrophy (DMD).

Duchenne muscular dystrophy (DMD) is a lethal genetic muscle disorder caused by recessive mutations in the dystrophin gene. The size of the gene (2.4 Mb) and mRNA (14 kb) in addition to immunogenicity problems and inefficient transduction of mature myofibres by currently available vector systems are formidable obstacles to the development of efficient gene therapy approaches. Adeno-associated viral (AAV) vectors overcome many of the problems associated with other vector systems (nonpathogenicity and minimal immunogenicity, extensive cell and tissue tropism) but accommodate limited transgene capacity (<5 kb). As a result of these observations, a number of laboratories worldwide have engineered a series of microdystrophin cDNAs based on genotype-phenotype relationship in Duchenne (DMD) and Becker (BMD) dystrophic patients, and transgenic studies in mdx mice. Recent progress in characterization of AAV serotypes from various species has demonstrated that alternative AAV serotypes are far more efficient in transducing muscle than the traditionally used AAV2. This article summarizes the current progress in the field of recombinant adeno-associated viral (rAAV) delivery for DMD, including optimization of recombinant AAV-microdystrophin vector systems/cassettes targeting the skeletal and cardiac musculature.

Screening for antisense modulation of dystrophin pre-mRNA splicing.

Most gene therapy approaches to genetic disorders aim to compensate loss-of-function by introducing recombinant cDNA-based minigenes into diseased tissues. The current report represents an ongoing series of studies designed to correct genetic mutations at the post-transcriptional level. This strategy modifies the binding of components of the spliceosome by high affinity hybridisation of small complementary (antisense) RNA oligonucleotides to specific pre-mRNA sequences. These, so-called 'splicomer' reagents are chemically modified to impart bio-stability, and are designed to cause skipping of mutant frame-shifting exon sequences leading to restoration of the reading frame and an internally deleted but partially functional gene product. For instance, Duchenne muscular dystrophy is generally caused by frame-shift mutations in the dystrophin gene, whereas in-frame deletions of up to 50% of the central portion of the gene cause Becker muscular dystrophy, a much milder myopathy, which in some cases can remain asymptomatic to old age. In the mdx mouse model of Duchenne muscular dystrophy, a mutation in exon 23 of the dystrophin gene creates a stop codon and leads to a dystrophin-deficient myopathy in striated muscle. In previous studies, we have demonstrated that forced skipping of this mutant exon by treatment of mdx muscle cells with splicomer oligonucleotides can generate in-frame dystrophin transcripts and restore dystrophin expression. Here, we report the results of an optimisation of splicomer sequence design by the use of both high-throughput arrays and biological screens. This has resulted in specific and, importantly, exclusive skipping of the targeted exon in greater than 60% of dystrophin mRNA, leading to the de novo synthesis and localisation of dystrophin protein in cultured mdx muscle cells.

Inhibition of atherosclerosis in apolipoprotein-E-deficient mice following muscle transduction with adeno-associated virus vectors encoding human apolipoprotein-E.

Apolipoprotein E (apoE) is a multifunctional plasma glycoprotein involved in lipoprotein metabolism and a range of cell signalling phenomena. ApoE-deficient (apoE(-/-)) mice exhibit severe hypercholesterolaemia and are an excellent model of human atherosclerosis. ApoE somatic gene transfer and bone marrow transplantation in apoE(-/-) mice results in reversal of hypercholesterolaemia, inhibition of atherogenesis and regression of atherosclerotic plaque density. Replication defective adeno-associated virus vectors (rAAVs) are an attractive system currently in clinical trial for muscle-based heterologous gene therapy to express secreted recombinant plasma proteins. Here we have applied rAAV transduction of skeletal muscle to express wild-type (epsilon3) and a defective receptor-binding mutant (epsilon2) human apoE transgene in apoE(-/-) mice. In treated animals, apoE mRNA was present in transduced muscles and, although plasma levels of recombinant apoE fell below the detection levels of our ELISA (ie <10 ng/ml), circulating antibodies to human apoE and rAAV were induced. Up to 3 months after a single administration of rAAV/apoE3, a significant reduction in atherosclerotic plaque density in aortas of treated animals was observed (approximately 30%), indicating that low-level rAAV-mediated apoE3 expression from skeletal muscle can retard atherosclerotic progression in this well-defined genetic model.

Oligonucleotide-based gene correction strategies: applications to neuromuscular and cardiovascular diseases.

Gene augmentation is an attractive and viable approach in treatment of inherited diseases, despite its limitations, ie, the amount of viral genome in replication-defective viral vectors is often too small for larger copy DNAs to be inserted. In addition, most viral vectors, when assessed in vivo, can elicit a host immune response, especially in cases of introduction of a gene copy where protein is completely absent, may have potential mutagenic effect on the host genome, or may be down-regulated. Therefore, alternative therapeutic approaches are being investigated, such as chimeraplasty, in which a mutated allele that already exists in an affected individual can be corrected. Although the only gene defects that can be corrected by chimeraplasty are point mutations, and the correction frequencies are variable, it has been observed that intracellular delivery of oligonucleotides is likely to be more efficient than that of plasmid DNA or viral vectors, because corrected genes are expressed from their autologous promoters, ensuring thus correct spatial and temporal expression, and host immune response is not elicited. Another strategy in the therapy of inherited diseases, such as Duchenne muscular dystrophy, is the application of antisense RNA oligonucleotides, or splicomers, to the exclusion of exons whose mutation leads to production of deficient essential proteins. Here we report the recent progress made and problems encountered in each of these fields, and discuss the potential of nucleotide-based gene correction strategies in treatment of neuromuscular and cardiovascular diseases.

Gene correction of the apolipoprotein (Apo) E2 phenotype to wild-type ApoE3 by in situ chimeraplasty.

Apolipoprotein (apo) E is a polymorphic plasma protein, synthesized mainly by liver. Here, we evaluate whether synthetic DNA-RNA oligonucleotides (chimeraplasts) can convert a dysfunctional isoform, apoE2 (C --> T, R158C), which causes Type III hyperlipidemia and premature atherosclerosis, into apoE3. First, we treated recombinant Chinese hamster ovary cells stably secreting apoE2 with a 68-mer apoE2 to apoE3 chimeraplast. About one-third of apoE2 was converted to apoE3, and the repair was stable through 12 passages. Subcloning treated cells produced both apoE2 and apoE3 clones. Direct sequencing and reverse transcription polymerase chain reaction confirmed the genotype, whereas phenotypic change was verified by isoelectric focusing and immunoblotting of secreted proteins. Second, we established that the APOE2 gene can be targeted both in vivo, using transgenic mice overexpressing human apoE2, and in chromosomal context, using cultured lymphocytes from a patient homozygous for the epsilon2 allele. We conclude that chimeraplasty has the potential to convert the apoE2 mutation in patients with Type III hyperlipidemia to apoE3.

In vivo analysis of functional regions within yeast Rap1p.

We have analyzed the in vivo importance of different regions of Rap1p, a yeast transcriptional regulator and telomere binding protein. A yeast strain (SCR101) containing a regulatable RAP1 gene was used to test functional complementation by a range of Rap1p derivatives. These experiments demonstrated that the C terminus of the protein, containing the putative transcriptional activation domain and the regions involved in silencing and telomere function, is not absolutely essential for cell growth, a result confirmed by sporulation of a diploid strain containing a C terminal deletion derivative of RAP1. Northern analysis with cells that expressed Rap1p lacking the transcriptional activation domain revealed that this region is important for the expression of only a subset of Rap1p-activated genes. The one essential region within Rap1p is the DNA binding domain. We have investigated the possibility that this region has additional functions. It contains two Myb-like subdomains separated by a linker region. Individual point mutations in the linker region had no effect on Rap1p function, although deletion of the region abolished cell growth. The second Myb-like subdomain contains a large unstructured loop of unknown function. Domain swap experiments with combinations of elements from DNA binding domains of Rap1p homologues from different yeasts revealed that major changes can be made to the amino acid composition of this region without affecting Rap1p function.

Prospective comparison of unenhanced spiral computed tomography and intravenous urogram in the evaluation of acute flank pain.

OBJECTIVES: To prospectively compare the diagnostic ability of unenhanced spiral computed tomography (NCCT) and intravenous urogram (IVU) in the evaluation of adults with acute flank pain. METHODS: After giving informed consent, 106 adult patients with acute flank pain suspected of having urolithiasis underwent NCCT followed by IVU. Subsequent follow-up was scheduled within 72 hours in the Urology Clinic. Each NCCT was read by a single radiologist who was unaware of clinical history and IVU results. Each IVU was read by a different radiologist who was unaware of clinical history and NCCT results. Sensitivity, specificity, and positive and negative predictive values were determined for NCCT and IVU. RESULTS: The diagnosis of ureterolithiasis was defined as unequivocal evidence of urolithiasis on either NCCT or IVP. Seventy-five of 106 patients evaluated were diagnosed with ureterolithiasis. Clinical follow-up was available in 74 (98%) stone patients and in 31 (100%) of 31 non-stone patients. In 72 of the 75 patients diagnosed with ureteral calculi, the NCCT made the diagnosis. IVU made the diagnosis in 65 of the 75 patients. Of the 31 patients without ureterolithiasis, the NCCT was negative in all cases. IVU was negative in 29 of the 31 cases. Unenhanced spiral CT was 96% sensitive and 100% specific (P <0.001). IVU was 87% sensitive and 94% specific (P <0.001). Compared with IVU, using the log odds ratio and Fisher's exact test, NCCT was significantly better able to predict the presence of urolithiasis (P=0.015). CONCLUSIONS: NCCT accurately diagnoses ureterolithiasis in patients presenting with acute flank pain. NCCT is significantly better than IVU in determining the presence of urolithiasis.

lambda Rap protein is a structure-specific endonuclease involved in phage recombination.

Bacteriophage lambda encodes a number of genes involved in the recombinational repair of DNA double-strand breaks. The product of one of these genes, rap, has been purified. Truncated Rap proteins that copurify with the full-length form are derived, at least in part, from a rho-dependent transcription terminator located within its coding sequence. Full-length and certain truncated Rap polypeptides bind preferentially to branched DNA substrates, including synthetic Holliday junctions and D-loops. In the presence of manganese ions, Rap acts as an endonuclease that cleaves at the branch point of Holliday and D-loop substrates. It shows no obvious sequence preference or symmetry of cleavage on a Holliday junction. The biochemical analysis of Rap gives an insight into how recombinants could be generated by the nicking of a D-loop without the formation of a classical Holliday junction.

Rap1p is a negative regulator of the RAP1 gene.

The promoter of the RAP1 gene contains four potential binding sites for Rap1p, located between the UAS and the RNA initiation site. We have confirmed that three of these sites are recognised by Rap1p in vitro. Different combinations of the three sites were then mutated to abolish Rap1p binding, and the effect of these mutations on promoter activity was determined. When all three Rap1p sites were mutated, the activity of the promoter increased by about 130%, indicating that at least one of the sites is a negative element. Analysis of promoters with different combinations of the mutant sites revealed that the 5'-most site (A) is the principal target for repression. To test the involvement of Rap1p in controlling RAP1 expression, we have measured transcription of the chromosomal RAP1 gene in a RAP1 wild-type strain and two strains containing rap1ts mutations. At a semi-permissive temperature, the RAP1 promoter was more active in the rap1ts strains than in the RAP1 wild-type strain, suggesting that expression of the chromosomal RAP1 gene is greater when the activity of Rap1p in the cell is compromised. The activities of the wild-type promoter, and the promoter with mutations in the three Rap1p-binding sites, were compared in sir1, sir2, sir3 and sir4 mutant strains. In each case, the mutated promoter was significantly more active than the wild-type promoter, implying that the repression mechanism is not dependent on any one of the SIR gene products.

The multifunctional transcription factors Abf1p, Rap1p and Reb1p are required for full transcriptional activation of the chromosomal PGK gene in Saccharomyces cerevisiae.

We have identified two new transcription factor binding sites upstream of the previously defined UAS within the phosphoglycerate kinase (PGK) gene promoter in Saccharomyces cerevisiae. These sites are bound in vitro by the multifunctional factors Cpf1p and Reb1p. We have generated targeted deletions of Rap1p, Abf1p and Reb1p binding sites in the promoter of the chromosomal copy of the PGK gene. Northern blot analysis confirmed that most PGK promoter activity is mediated through the Rap1p binding site. However, significant effects are also mediated through both the Reb1p and Abf1p sites. In contrast, when the promoter is present on a high-copy-number plasmid, both the Abf1p and Reb1p sites play no role in transcriptional activation. The role of Cpf1p was examined using a cpf1 null strain. Cpf1p was found to have little if any, effect on activation of either the chromosomal or plasmid-borne PGK gene.

A Reb1p-binding site is required for efficient activation of the yeast RAP1 gene, but multiple binding sites for Rap1p are not essential.

The Saccharomyces cerevisiae RAP1 protein (Rap1p) is a key multifunctional transcription factor. Using gel retardation analysis, four binding sites for Rap1p have been identified within the promoter of the RAP1 gene. These sites are located downstream of a binding site for the transcription factor Reb1p. The Reb1p site and an associated AT-rich region are important for transcriptional activation, but deletion of three of the Rap1p-binding sites had little effect on promoter activity. The activity of the RAP1 promoter has been analysed in a yeast strain (YDS410) that contains a temperature-sensitive mutation in the RAP1 gene. This mutation renders the DNA-binding activity of Rap1p temperature dependent. When YDS410 was grown at a semi-permissive temperature (30 degrees C), the activity of the RAP1 promoter increased by approximately 170%, compared with the same strain grown at the permissive temperature (25 degrees C). A RAP1 promoter in which three of the four Rap1p-binding sites had been deleted, showed only a small increase in activity in the same experiment. These data confirm that Rap1p is not required for activation of the RAP1 gene, and suggest a role for Rap1p in negative autoregulation.

Use of a selection technique to identify the diversity of binding sites for the yeast RAP1 transcription factor.

We have used the technique known as selected and amplified binding (SAAB) to isolate binding sites for the yeast transcription factor RAP1 from a degenerate pool of oligonucleotides. A total of 47 sequences were isolated, of which two were shown to be contaminating non-RAP1 binding sites. After excluding these two sequences the remainder of the sequences were used to derive a new consensus binding site for RAP1. The new consensus 5' A/G T A/G C A C C C A N N C C/A C C 3' is a significant extension of the existing consensus (4). It is longer by two base pairs at the 5' end and is significantly more constrained at the 3' end. An analysis of the combinations of mis-matches in individual SAAB sequences, compared to the consensus RAP1 binding site, has allowed us to analyse the structure of the RAP1 binding site in some detail. The binding site can be sub-divided into three regions; a core binding site, a 5' flanking region and a 3' flanking region. The core binding site, consisting of the sequence 5'CACCCA3', is critical for recognition by RAP1. The less conserved flanking regions are not as important. Interactions between RAP1 and these regions probably stabilise the interaction between RAP1 and the core binding site. Each of the sequences isolated in the SAAB analysis was used to search release 78 of the EMBL+GenBank DNA data base. The searches identified 102 potential binding sites for RAP1 within promoters of yeast genes.

Alternative splicing of a human alpha-tropomyosin muscle-specific exon: identification of determining sequences.

The human alpha-tropomyosin gene hTMnm has two mutually exclusive versions of exon 5 (NM and SK), one of which is expressed specifically in skeletal muscle (exon SK). A minigene construct expresses only the nonmuscle (NM) isoform when transfected into COS-1 cells and both forms when transfected into myoblasts. Twenty-four mutants were produced to determine why the SK exon is not expressed in COS cells. The results showed that exons NM and SK are not in competition for splicing to the flanking exons and that there is no intrinsic barrier to splicing between the exons. Instead, exon SK is skipped whenever there are flanking introns. Splicing of exon SK was induced when the branch site sequence 70 nucleotides upstream of the exon was mutated to resemble the consensus and when the extremities of the exon itself were changed to the corresponding NM sequence. Precise swaps of the NM and SK exon sequences showed that the exon sequence effect was dominant to that of intron sequences. The mechanism of regulation appears to be unlike that of other tropomyosin genes. We propose that exclusion of exon SK arises because its 3' splicing signals are weak and are prevented by an exon-specific repressor from competing for splice site recognition.

Mutually exclusive splicing of calcium-binding domain exons in chick alpha-actinin.

We have determined the complete sequence of chick brain alpha-actinin (892 amino acids; 107,644 Da). The sequence differs from that of smooth muscle alpha-actinin only in the region of the first EF-hand calcium-binding motif, where 27 residues in brain alpha-actinin are replaced by just 22 residues in the smooth muscle isoform. This probably accounts for the different calcium sensitivities of the two isoforms with respect to actin binding. Analysis of the gene structure showed that this region of sequence divergence is encoded by two separate exons whose incorporation is mutually exclusive. We have determined the proportion of the two transcripts in various tissues and cell lines using poly(A)+ RNA and a quantitative assay based on the polymerase chain reaction. MRC-5 fibroblasts and HeLa cells express mRNAs encoding both isoforms, whereas Namalwa lymphoblastoid cells, which lack actin stress fibers, express only the non-muscle mRNA. Both isoforms of alpha-actinin became incorporated into stress fibers and cell-matrix junctions when full-length chick alpha-actinin cDNAs were expressed in monkey COS cells. The levels of chick alpha-actinin mRNAs were found to be serum-inducible, suggesting that alpha-actinin may be an early response gene.

Effects of RNA secondary structure on alternative splicing of pre-mRNA: is folding limited to a region behind the transcribing RNA polymerase?

The use in vivo of an alternative 5' splice site sequestered within a short stem of potential RNA secondary structure is determined by the length of the loop. Above a threshold length of loop, the alternative site is used despite the potential structure. In contrast, the alternative site is used very little or not at all during splicing in vitro with all lengths of loop that we have tested. A model is proposed which suggests that pre-mRNA is free to fold only within a limited period after transcription.