PCR amplification of long DNA fragments.

The polymerase chain reaction (PCR) has recently evolved as a standard laboratory technique, popular in all areas of molecular biology research. However, the technique still has two limitations: the relatively low fidelity of Taq polymerase when compared with other polymerases (1), and its inability to efficiently amplify fragments higher than 3 kbp (2, 3). Although these two issues are irrelevant in most PCR applications, they are limiting factors in some cases such as the amplification of large constructs for in vitro mutagenesis and the amplification of eukaryotic genomic DNA segments containing introns of unknown length. We have developed PCR conditions allowing the efficient amplification of DNA segments with 6 kbp (see fig. 1 legend). We were interested in the amplification of several segments (ranging 1 to 6 kbp) of the Cyllla CAT construct (4), in order to develop a PCR based method for the in vitro mutagenesis of the upstream regulatory domain of the Cyllla cytoskeletal sea urchin gene. We found the absence of KC1 to be optimal for the amplification of DNA molecules in the range of 3 6 kbp: bands of these sizes were faint or undetectable in ethidium bromide stained agarose gels after 30 PCR cycles (using 1 — 100 ng of template and buffers PEC and I in Table 1), yet several micrograms of 3 —6 kbp molecules were obtained after 30 cycles (using 10 ng of template and buffer II). Given the large temperature dependence of pH in buffers made with Tris (5), we tested a Tricine buffer (III in Table 1) without KC1. As shown in Figure 1, using buffer III it is possible to obtain several micrograms of a 6.2 kbp PCR product even after only 10 PCR cycles. Template amounts used in these 10 cycle experiments were 100 ng to 1 fig, with only small differences in product yield.