Multivariate analysis of factors influencing repeat expansion detection.

Repeat expansion detection (RED) is a powerful tool for detection of expanded repeat sequences in the genome. In RED, DNA serves as a template for a repeat-specific oligonucleotide. A thermostable ligase is used to ligate oligonucleotides that have annealed at adjacent positions, creating multimers in a thermal cycling procedure. The products are visualized after gel electrophoresis, transfered to a membrane and subsequently hybridized. Multiple linear regression (MLR) and partial least square (PLS) techniques were used to reveal the most influential factors in the amplification reaction and to identify possible interacting factors. Ligation temperature proved to be the most important factor in the reaction: Temperatures far below the melting point of the oligonucleotide increased the yield considerably. Higher cycle number resulted in a continuous rise in intensity, indicating that the ligase remained active even after 700 cycles or 12 hr of cycling. In addition, the concentration of ligase was found to be important. Using optimal parameters, a 5.5- and 3.2-fold increase in the yield of 180- and 360-nucleotide products respectively was obtained. The improved sensitivity makes the method more robust and facilitates detection of repeat expansions. This improvement may be particularly useful in development of RED for diagnostic purposes as well as for nonradioactive detection of RED products. Based on these results, a new protocol for the RED method was developed taking into account the risk of introducing artifacts with increased enzyme concentrations and lowered annealing temperatures.