Digestion of Terminal Restriction Endonuclease Recognition Sites on PCR Products

One of the common methods for cloning polymerase chain reaction (PCR) products is overhanging-end cloning (also known as sticky-end or directional cloning). Frequently, it is not possible to use restriction enzyme sites already present in the amplified product, and primers that encode recognition sites of restriction endonucleases in addition to the specific sequence have to be designed. After amplification of the target sequence with these primers, the PCR products are purified, digested with restriction enzymes and cloned into vectors treated with the same enzymes. However, it has been found that many restriction enzymes fail to cleave at the end of PCR fragments (1–5). To circumvent this problem, the addition of at least three more nucleotides at the end of a restriction site (1) was suggested. In contrast to this, it was claimed in one study that even 3–4 extra bases might be insufficient for reliable cutting (3). Some publications (4,5) and commercial catalogues contain lists of base pair extensions required for efficient restriction endonuclease cleavage near DNA termini. One major disadvantage of these lists is that they are all based on experiments in which a plasmid vector has been subsequently digested with two enzymes specific for neighboring cleavage sites. Because the distance of the two restriction sites in a plasmid vector is predetermined, these lists give no information about the minimal number of base pairs really necessary for the restriction enzyme to successfully attach to its cleavage site; furthermore, the experiments have not been performed with PCR products. We per-