The goal of this work was to establish a system that facilitates the cloning and then allows a flexible shuttling of the corresponding sequences and/or expression cassettes into the appropriate vector systems. To reach this aim we constructed a set of vectors that take advantage of a Cre-dependent recombinase system .
First we constructed the donor plasmid. We selected a pCR-TOPO vector as backbone and removed the ampicillin resistance (ampR) gene by BspHI digestion and subsequent religation of the plasmid. In a second step a 1.8 kb DNA cassette was inserted by using EcoRI sites. This artificial cassette (K42) has a modular structure and contains a histone promoter (H4-1), a signal peptide (encoding the first 39 aa of P. falciparum surface protein MSP1) fused to the EYFP reporter protein and a histidine stretch (6xH) as well as the beta tubulin terminator (BTU2) of T. thermophila. All of these DNA modules can be easily changed by using unique restriction sites. The whole cassette is flanked by loxP sites on the 5' and 3' ends (see Additional file 1; for basic donor plasmid see figure 5).
A chloramphenicol resistance (CmR) was inserted between the loxP site and the BTU2 terminator to reduce background clones, because this CmR is only translated in E. coli if a correct site-specific recombination between acceptor and donor plasmid has been occurred. However, due to the modular architecture of the pDL-plasmids also other resistance markers can be used for this purpose (e.g. tetracycline, zeocin etc.). In a next step, the sacB marker gene was inserted into the intermediate vector. The sacB gene product metabolises sucrose into levansucrose, a toxic substance for E. coli cells. The parallel usage of the Cm resistance (selection) and the sacB (counter-selection) gene strongly inhibits the presence of non-recombinant clones. Finally, we replaced the Enhanced Yellow Fluorescence Protein (EYFP) cDNA by the MSP119 cDNA from P. falciparum or the blasticidin resistance gene (bsdR) to demonstrate that the whole system facilitates cloning and expressing foreign genes like previously shown for other host systems e.g. arabidopsis.
Recently, we described the bifunctional dihydrofolate reductase and thymidylate synthase (DHFR-TS) of T. thermophila. Both enzyme activities play a crucial role in DNA synthesis. The loss of these essential activities can be used as an auxotrophic marker in T. thermophila. We developed a vector system that combines the knock out of the endogenous DHFR-TS gene with the knock in of an expression cassette that encodes a foreign gene (pKOI) . Appropriate acceptor vectors for the T. thermophila system were created by cloning the loxP-promoter site (loxprom) into this pKOI vector backbone as well as into a previously described rDNA based episomal plasmid (pH4T2). The new vectors were named into pKOIX (pKOI backbone) and pAX (pH4T2 backbone). A scheme of the acceptor vector structure is given in figure 1.
The final expression vectors pAX-MSP119/pAX-bsdR and pKOIX-MSP119/pKOIX-bsdR were generated via the novel recombinase mechanism. In general, a mixture of 100 ng of donor and 100 ng of acceptor plasmid yielded 30 to 50 initial positive clones that were able to grow on LB-agar plates supplemented with both chloramphenicol and ampicillin.
We picked six clones of the MSP119 recombinase reactions and analyzed them by diagnostic PCR and restriction analysis to test the efficiency of the novel recombinase approach. All of the analyzed clones were positive (6/6). This finding was independent of the used acceptor plasmid. The recombinant pAX- as well as pKOIX-plasmids carry complete expression cassettes and an all complete plasmid backbone. The results are shown in figure 2. The left column illustrates the results using the pAX and right column the pKOIX backbone. Positive clones were analyzed by restriction analysis (XhoI and SacI) and diagnostic PCR (360 bp fragment in positive clones), verifying a correct recombinase event. In general nearly all clones (80–100%) obtained were positives and have a complete backbone. We obtained such quotes in all performed Cre-recombinase reactions (data not shown).
We compared the recombinase efficiency of generating recombinant expression plasmids standard cloning techniques. The pDL-MSP119 (see Additional file 2) was digested with NotI and SacI and the corresponding insert (H4-1-MSP119-BTU2) was ligated into the pre-cut pH4T2 vector to obtain an rDNA-based MSP119 expression plasmid. An analogous approach was done with the pKOI plasmid. Eight clones were randomly picked and analyzed by restriction analysis. All of them (8/8) were negative and most of the pH4T2 backbones were degraded or fragmented during the ligation, transformation selection and propagation procedure. The supplementary figure 1 illustrates a typical result of such an approach. In most cases only 2–3% (one of 30 to 50 clones) carries the complete expression cassettes in a complete plasmid.
Extracts of cells that were transformed with the pKOIX-MSP119 plasmid were analyzed for expression of a 19 kDa protein fragment of the MSP1 protein. We used cell extracts of four independent stably transformed strains and compared them to the non-transformed 1868/7 wildtype strain. In all tested cell extracts the recombinant 19 kDa fragment of MSP1 (MSP119) could be detected by the specific monoclonal antibodies (mAb2.2, mAb7.5, mAb12.8 and mAb12.10 were kindly provided by Prof. McBride, Edinburgh, UK) [26,27]. In the wildtype negative control no signal could be found. The results are summarized in figure 3. In previous expression experiments we could also demonstrate that the rDNA plasmid is capable of expressing the MSP1 C-terminus (data not shown).
We attempted to confirm the Cre-recombinase dependent cloning by using a second independent expression module. Therefore pAX and pKOIX constructs that carry the bsdR expression cassette (pAX-bsdR, pKOIX-bsdR) were transformed into conjugating and vegetative T. thermophila wildtype strains according to protocols previously described. The transformants (clone1- clone10) were cultivated in SPP-medium supplemented with thymidine and increasing concentrations of paromomycin (figure 4B) to ensure a stable propagation of the clones. The same clones were cultivated in SPP-medium without antibiotics (figure 4A). In a second step we performed a blasticidin growing assay and switched the antibiotic from paromomycin to blasticidin (figure 4C) or applied both antibiotics in parallel. In all experiments the wildtype (WT) that did not contain a resistance gene died within 2–5 days. As expected the mock transformant (MT) that only carried the neo2 cassette (resistance against paromomycin) died when blasticidin was added to the SPP-medium (bsd antibiotic control).
Interestingly, we observed that the pKOIX-bsdR transformants were more stable when compared to the pAX clones (figure 4D). Eight out of the ten analyzed independent pKOIX-bsdR clones are resistant against both antibiotics. In contrast to that only ~30% of the pAX clones (3/9) displayed both resistances. This is probably due to recombination events between the rDNA plasmids backbones and/or due to the very similar architecture of the neo2 and bsdR resistance cassettes (see figure 1). In summary these results illustrate that the Cre-dependent modular donor plasmids in combination with the recently described knock out/knock in concept (pKOIX) provides an easy and sustainable system to establish a resistance testing tool.