PCR-RAPD has been widely used for genetic analysis in several organisms. However, due to complex interactions among the components of the PCR reaction it is unlikely that a single amplification condition can be suitable for all situations. In order to determine the optimum conditions for using PCR-RAPD in taxonomical analyses of the genus Atta, the following factors were tested: concentrations of MgCl2, DNA, BSA, cycling programs and methods of DNA extraction, using Fractional Factorial Design and Central Composite Design. DNA extraction methods had little influence on PCR-RAPD reactions, while the cycling programs showed the largest effect. The MgCl2 concentration was less important than the amount of DNA used in the reaction. Using cycling program P2, a significant improvement in the number of DNA fragments was achieved when MgCl2 concentration was increased to 3.0 mM and the BSA and DNA concentrations were reduced from 0.1% to 0.05% and from 2 to 1 ng/mul, respectively. The optimum conditions for PCR-RAPD with Atta specimens obtained in this study were: 25 mul reaction with 10 mM Tris-HCl (pH 9.0), 3.0 mM MgCl2, 50 mM KCl, 100 muM of each dNTP, 0.2 muM of primer, 1.0 U of Taq polymerase, 25 ng template DNA (extracted by Cheung's method) and BSA 0.05%, using the amplification program which consisted of 3 min at 94ºC, 3 min at 35ºC, followed by 40 cycles of 1 min at 94ºC, 1 min at 36ºC and 2 min at 72ºC, and a 5 m final extension at 72ºC.
RAPD-PCR; optimization; DNA extraction; leaf-cutting ants