Reevaluation of the exact CAG repeat length in hereditary cerebellar ataxias using highly denaturing conditions and long PCR.

Hereditary cerebellar ataxias, including spinocerebellar ataxia type I (SCA1), dentato-rubro-pallidoluysian atrophy (DRPLA), and Machado-Joseph disease (MJD), have been associated with unstable CAG repeats. The length of the CAG repeat is a major factor in determining the age of onset of these diseases. In electrophoresis through acrylamide gels with formamide, the CAG repeat length following the polymerase chain reaction (PCR) coincides with the sequence-determined repeat length after subcloning. However, without formamide, PCR products with long CAG repeats appear 1-4 repeats shorter than when electrophoresed with formamide, and the repeat lengths are variable. In addition, the larger the CAG repeats are, the more difficult are the PCR reactions. A mixture containing thermostable Taq and Pwo DNA polymerases (so-called "long PCR") is much more sensitive than that with Taq polymerase alone in detecting- expanded CAG repeats. Therefore, highly denaturing conditions, especially formamide gel electrophoresis, and the "long PCR" protocol should be used to evaluate the exact CAG repeat length. We have used these principles to detect unstable CAG repeats. The normal ranges are 14-34 repeats for MJD, 6-31 repeats for DRPLA, and 21-32 repeats for SCA1.