Interruption of perfect CAG repeats by CAA triplets improves the stability of glutamine-encoding repeat sequences.

Several disorders are caused by the genomic expansion of a trinucleotide repeat beyond a specific critical size in the coding region of genes. The largest group of these diseases, which includes Huntington’s disease and the spinocerebellar ataxias of type 1, 2, 3, 6, and 7, are caused by expansion of a CAG repeat resulting in proteins with an expanded polyglutamine stretch (14). A correlation between the lengths of the repeat and the age of onset of the diseases has been found. However, functional studies and the generation of cell and animal models expressing very long repeats have been hampered by the significant instability of long perfect CAG repeats in E. coli (2,9,10,13) and the constraints on the size of homogeneous CAG repeats that can be efficiently and accurately produced using PCR (1,13). Several PCR-based methods for cloning long repeats and introducing these repeats in target genes have, nevertheless, been described, but these methods have not resulted in the generation of stable clones with very long repeats (5,6,11). We now report on a strategy that exploits both the redundancy of the genetic code and the special properties of type II restriction enzymes, which cleave at an exact distance outside their recognition site to generate constructs expressing long homopolymeric amino acid repeats. The method described here is fast, easy to perform, and results in stable constructs encompassing very long repeats. First, we tested whether the stability problems of homogeneous CAG repeats could be solved by the interruption of a perfect repeat at the DNA level by another triplet encoding the same amino acid (i.e., for polyglutamine interruption of a CAG repeat with CAA). Mixed CAG/CAA repeats were constructed using two complementary synthetic DNA oligonucleotides 5′-GCAGCAACAGCAGCAGCAACA-3′ and 5′-TGCTGTTGCTGCTGCTGTTGC-3′. Oligonucleotides were phosphorylated with T4 polynucleotide kinase (Amersham Biosciences, Piscataway, NJ, USA), hybridized and ligated using T4 DNA ligase (Roche Applied Science, Indianapolis, IN, USA). Ligation yields a 21-bp repeat with the basic sequence [(CAG)-CAA-(CAG)-CAA]n. The ligated fragments were treated with AmpliTaq® DNA polymerase (Applied Biosystems, Foster City, CA, USA) to create an A extension, followed by electrophoresis on a 2% agarose gel, and the isolation of fragments in the 300–400 bp size range using the Nucleotrap extraction kit (Macherey-Nagel, Düren, Germany). The purified fragments were subcloned in the pGEM®-T Easy vector containing T extensions (Promega, Madison, WI, USA) and transformed to STBL-2 cells (Invitrogen, Carlsbad, CA, USA). The insert lengths of the clones obtained were determined with PCR (buffer: 16.6 mM [NH4]SO4, 67 mM Tris-HCl, pH 8.8, 10 mM β-mercaptoethanol, 6.7 μM EDTA, and 170 μg/mL BSA) using M13-primers, and both strands of the three largest inserts were sequenced with the T7-primer and the M13-reverse primer, respectively. One clone, designated p98Gln, contained a mixed repeat encoding a perfect Gln-repeat of 98 residues. The two other clones had one or more variations that caused an interruption of the encoded Gln-repeat sequence. To test the stability of the insert in E. coli, p98Gln was transformed to a bacterial strain often used in cloning experiments, XL-1 Blue® (Stratagene, La Jolla, CA, USA). Plasmid DNA was isolated from independent colonies, and the insert lengths were determined with restriction enzyme digestion. Nine out of 10 contained an insert of the appropriate length; one clone contained a deletion of about 20 bp (probably one 21-mer repeat unit). To determine the insert lengths of the first nine clones more acBenchmarks