Chromosome walking by cloning of distinct PCR fragments.

For functional characterization of genes, it is often interesting to identify additional upor downstream sequences in a fast and inexpensive way. Constructing a phage λ or cosmid library is time consuming, and subcloning of the corresponding fragments has to be performed to isolate DNA that can be sequenced (3). Alternative methods such as inverse PCR (2) or onesided PCR (4) have been developed, but it is sometimes a tedious effort to sequence the PCR product. Small amounts of byproduct, sometimes invisible by ethidium bromide staining after agarose electrophoresis, are sufficient to generate a second sequence that makes the determination of the primary sequence impossible. Some sequencing facilities do not accept long PCR products and rely on sequencing of vector constructs. We have developed a strategy in which the PCR product is cloned in a straightforward way and sequencing of already known vector DNA is reduced to a minimum. The work was carried out with strains of the thermophilic Gram-positive bacterium Aneurinibacillus thermophilus, a member of the Bacillaceae (1). However, since the method is based on cloning of PCR products, it is compatible for other organisms as well. Chromosomal DNA of the corresponding organism was digested with restriction enzymes that cut the multiple cloning site (MCS) of a favorite cloning vector (we used pBCKS from Stratagene, La Jolla, CA, USA). When using an inverse PCR strategy for sequencing, it is necessary to have a restriction site near the sequence where the PCR primers are set. This is very restrictive, especially if only a small amount of sequence information is available (e.g., after cloning DNA by means of a degenerate probe). In our strategy, it is possible to use any restriction site, whether or not this site is on the known sequence. The only prerequisite is a cloning vector with this specific recognition sequence and an additional restriction site for a blunt end-generating endonuclease. The digested chromosomal DNA was ligated into the linearized and dephosphorylated vector. A forward primer was designed at the 3′ end of the known sequence (Figure 1). In contrast to inverse PCR, only one specific primer is sufficient. Using the ligation reaction products as templates (approximately 20 ng DNA), PCRs are carried out with the 3′ primer and a primer flanking the MCS (here we used M13 universal forward primer). Expand High Fidelity System (Roche Molecular Biochemicals, Penzberg, Germany) was used for the amplification reaction. This kit consists of both the Taq DNA and the Pwo DNA polymerases. Proofreading Pwo DNA polymerase ensures higher fidelity and the generation of blunt end reaction products. The length of the PCR product depends on the position of the restriction sites on the unknown sequence. In a screening experiment, the approximate length of the amplification product using libraries generated after restriction with different enzymes can be determined. In the case of pBCKS, the following enzymes were used: BamHI, ClaI, EcoRI, HindIII, KpnI, PstI, SalI, SstI, XbaI, and XhoI (Life Technologies, Eggenstein, Germany). The reaction with the PCR product of appropriate length was optimized so that only one specific band was detected on the agarose gel. The amplification product was digested with the corresponding restriction enzyme, resulting in two fragments, one of which contained only unknown sequence. The smaller fragment, consisting of vector sequence was removed by preparative agarose gel electrophoresis. The purified DNA fragment now possesses one distinct sticky end on one side and a blunt end on the other. The sticky end is phosphorylated, whereas the blunt end carries no phosphate residue, due to the PCR. The fragment was cloned into the vector cut by the selected sticky end and blunt end endonuclease, respectively. To reduce religation of the cloning vector due to incomplete digestion, the plasmid was first cut with the corresponding enzyme, the ends were dephosphorylated, and, after removal of the calf intestinal alkaline phosphatase, a digestion with the blunt endgenerating enzyme was carried out. In the final step, the fragment was ligated into the vector (Figure 1), and the corresponding plasmid was transformed into the E. coli strain TG1 (Stratagene) by electroporation. Selection of positive clones can be easily done by in situ PCR using the established PCR protocol. For in situ PCR, colonies were picked, and cells were suspended in 10μL aliquots of sterile water. One microBenchmarks