The fission yeast Schizosaccharomyces pombe is an attractive model organism for studying eukaryotic biology due to the ease of genetic and molecular manipulations (3). In recent years, several selectable markers and promoters have been developed to enhance our ability to manipulate gene expression in this yeast. Recently, Invitrogen (Carlsbad, CA, USA) has developed the SpECTRA™ S. pombe Expression System that consists of three TOPO® cloning vectors that allow for rapid cloning of PCR products, with the option to epitope-tag the gene product (adding a carboxy-terminal V5-6his tag). Transcription of cloned open reading frames is under the control of the thiamine-repressible promoters nmt1, nmt41, and nmt81. (Promoters nmt41 and nmt81 are TATA box mutant derivatives of nmt1 that allow for different levels of transcription.) However, as all three plasmids utilize the same selectable marker (simian virus 40-driven Saccharomyces cerevisiae LEU2, which complements an S. pombe leu1 mutant allele), use of these vectors limits the choice of host strains and the possibility to transform strains with more than one plasmid construct. To enhance the use of the SpECTRA vectors by allowing for the transformation of Leu+ strains or for the introduction of two distinct plasmids into a host strain, we have developed a protocol that facilitates the exchange of the LEU2 marker for another selectable marker. This protocol takes advantage of the robust homologous recombination machinery found in yeasts, which has been used in the past to rescue chromosomal mutations onto plasmids by gap repair (8) or to screen mutant alleles created by PCR by co-transformation of the PCR product with a linearized plasmid (5,6). In both situations, the plasmid DNA is linearized before transformation and is repaired and recircularized by DNA that bears homology to the ends of the linearized plasmid. Using a protocol described here, we can readily replace the LEU2 -selectable marker in a SpECTRA vector with either the S. pombe ura4 or his3 genes. This technique could be used with other selectable markers as well. The general procedure is as follows. Marker replacement is carried out by co-transforming S. pombe cells with a SpECTRA-derived plasmid carrying an insert of interest that has been digested with restriction enzymes that cut within the LEU2 gene (Figure 1) along with a PCR product that carries a different selectable marker. The PCR product is flanked by sequences identical to sequences flanking LEU2 in the SpECTRA vectors (Table 1). Transformants are pooled and subjected to “smash and grab” (4) to rescue autonomous plasmids back into E. coli. Plasmids from these E. coli transformants are then screened to identify the desired construct. Figure 1 Schematic diagram of SpECTRA plasmid Table 1 Sequence of Marker Swap PCR Oligonucleotides (5′→3′) The oligonucleotides used in this study were designed as follows. The 5′-ends of the forward oligonucleotides contain 53 nucleotides of sequence identical to a region immediately 5′ to the unique NcoI restriction site in the SpECTRA vectors, while the 5′-ends of the reverse oligonucleotides contain 56 nucleotides of sequence identical to the complement strand immediately 3′ to the unique NsiI site in these vectors (Figure 1 and Table 1). The 3′-ends of each oligonucleotide set contain 21–24 nucleotides of sequence able to prime a PCR that could amplify either the S. pombe ura4- or the his3- selectable marker (for his3, the oligonucleotides were designed specifically for plasmid pAFl; see Reference 7). PCR was carried out using the FailSafe™ PCR System (EPICENTRE, Madison, WI, USA), although the specific system used for PCR is most likely not critical to the success of the procedure. The PCR products were 1528 bp and 2150 bp for the ura4 and his3 reactions, respectively (Figure 1). For the marker exchange to occur during yeast transformation, the plasmid must be linearized or gapped by restriction digestion within the LEU2 region. The PCR product that will be used to repair the gap possesses ends that are homologous to the vector immediately adjacent to unique NcoI and NsiI sites in the plasmid (Figure 1). However, one cannot use restriction enzymes that also cut within the DNA that was cloned into the SpECTRA vector before carrying out the marker swap. As indicated in Figure 1, in addition to NcoI and NsiI, StuI, BstEII, BstBI, BtgI (DsaI), XcmI, BspMI, and AgeI all cut the vector specifically within this region. (Complete sequence and restriction map information for these vectors can be found at http://www.invitrogen.com:80/content/vectors/pnmt1topo_seq.txt and http://www.invitrogen.com:80/content/vectors/pnmt1topo_rest.htm.) We have successfully swapped the selectable marker in a pNMT41 plasmid derivative that had been gapped either with NcoI and NsiI digestion or with NcoI and XcmI digestion (data not shown). As XcmI cuts approximately 0.9 kb from the region of homology present within the PCR product, we infer that any of the enzymes mentioned above can be used to create a gapped vector suitable for marker exchange. The gap repair transformation was carried out using approximately 2 μg digested plasmid DNA (10 μL volume) along with 2 μg PCR product (10 μL volume) to co-transform either strain FWP16 (h+ ura4-D18) to Ura+ or strain JSP303 (h+leu1-32 his3-D1) to His+, plating transformants onto PM-uracil or PM-histidine (9) medium, respectively. The transformation was performed using the protocol of Bahler et al. (2), which enhances homologous recombination events. In a control transformation, only the PCR product was added to competent cells to determine the level of colony formation due to nonhomologous integration of the PCR product or to uptake and unstable maintenance of the PCR product. Transformation with PCR product alone gave a background of several hundred colonies, while co-transformation of the PCR product with the gapped plasmid resulted in greater than 1000 colonies after four days at 30°C. The background level of His+ or Ura+ colonies generated by events other than plasmid gap repair does not pose a problem for the identification of the desired plasmids since these colonies do not contain DNA that can transform E. coli to ampicillin-resistance. Gap-repaired plasmids were rescued from S. pombe transformants into E. coli using a modified version of the “smash and grab” method (4), as described in the protocol in Table 2. Following this procedure, we typically observe 10–50 E. coli transformants. Restriction mapping of candidate plasmids demonstrated that 80%–100% of E. coli transformants carry the desired plasmid. EcoRV can be used to identify ura4-containing plasmids, as this enzyme cuts once within ura4 while not cutting within LEU2 (Figure 1). Additionally, HindIII cuts once within LEU2 while not cutting within ura4. EcoRI can be used to identify his3-containing plasmids, as this enzyme generates a 0.55-kb fragment internal to his3 (Figure 1). Finally, we have shown that the purified plasmids can transform ura4- or his3- mutant strains to Ura+ or His+, respectively (data not shown). Table 2 Rescue Protocol for Gap-Repaired Plasmids Using the protocol described here, we have been able to convert the LEU2- marked SpECTRA plasmids quickly and efficiently to either ura4- or his3- marked plasmids. This strategy can be easily adapted for the incorporation of other selectable markers by modifying the gene-specific sequence (3′-end, underlined in Table 1) of the PCR oligonucleotides combined with an appropriate template DNA. This general technique could also be used for marker exchange in other S. pombe or S. cerevisiae shuttle vectors in situations where a lack of appropriate restriction sites precludes standard subcloning procedures. The ease with which plasmid markers can be exchanged by gap repair transformation greatly enhances the usefulness of the SpECTRA cloning vectors by allowing for transformation of Leu+ host strains or for the co-transformation of host strains with SpECTRA-derived constructs that express two different proteins.