Full Text (PDF)
For cellular and animal transgenesis, FLP- and Cre-recombinase gene capture systems are highly effective to provide stable integration of a donor plasmid carrying the transgene cassette of interest into an engineered genomic locus in a given cell line. However, in many protocols, the entire plasmid bacterial backbone is integrated along with the transgene cassette. Here, we present a very simple yet highly efficient method for excluding plasmid bacterial backbone integration. The transgene cassette, including a single FLP recognition target site, is specifically amplified by PCR, and the resulting DNA ligated into minicircles can serve as donor DNA in FLP-mediated recombination. Interestingly, the elimination of the bacterial backbone increased expression of the inserted transgene. The presented method is simple and efficient for generating transgene cassette insertions devoid of the bacterial backbone.
Site-specific recombinases represent important tools in eukaryotic transgenesis. Recombinases like FLP and Cre have been widely used to direct transgene insertion to constitutively active chromosomal locations and to eliminate transgene concatemerization (1,2,3). By sequential transgenesis, cells of interest are first engineered to contain a ‘docking’ site containing recombinase recognition sequences and a stably expressed marker gene. Such cells are used subsequently for transgene cassette insertion through site-specific recombination. In many procedures, recombination-mediated transgenesis will result in co-insertion of the plasmid bacterial backbone (BB) which can influence transgene expression and constitutes an undesired source of bacterial DNA and resistance genes for the recipient cells (4). Recently, recombinase-mediated cassette exchange was developed by relying on tandem-incompatible recombination target sites for transgene insertion (5). One advantage of recombinase-mediated cassette exchange is the possibility to exclude co-insertion of the BB. Accordingly, BB insertion into the recipient cell genome will depend on additional random co-integrations of the transgene vector. Random integration is unfavored compared with recombinase-mediated integration, but random co-integration of BB sequences is observed if cells are prone to random integrations (6). A limitation of the use of recombinase-mediated cassette exchange for BB removal is the requirement for tandem recombination sites, which excludes widely used transgenesis vectors, cell lines, and animal models possessing only a single recombination site. Generation of donor minicircles in bacteria represents an alternative strategy for production of BB-deficient DNA substrates for transgenesis (7,8). This approach is complicated by the requirements for specific bacterial strains and vector designs. We sought a simple, time-saving, and general alternative method to avoid integrating the BB. Here, we present a very simple yet highly efficient method in which a transgene expression cassette without BB is amplified by PCR and the resulting DNA is ligated into minicircles to serve as donor DNA in FLP-mediated recombination.
The transgene donor plasmid used in our analysis is a modified version of pcDNA5/FRT (Invitrogen, Carlsbad, CA, USA) (FRT: FLP recognition target) in which the red fluorescent protein (DsRed) gene was inserted. This vector, designated pcDNA5/FRT-red (Figure 1A), also contains the Hygromycin resistance gene that lacks a start methionine (Invitrogen). To generate recombinase substrate DNA without BB, the DsRed/Hygromycin transgene cassette was amplified by PCR (see Supplementary Material for details) using primers with terminal BglII sites recognizing sequences immediately upstream of the CMV promoter and downstream of the Hygromycin resistance gene polyadenylation signal, respectively (primer locations indicated in Figure 1A). The PCR sample was digested with DpnI to remove template DNA. For generation of minicircles, DNA was cleaved with BglII and used in a T4 DNA ligation reaction (New England BioLabs, Ipswich, MA, USA) favoring intramolecular ligation. The minicircle DNA is schematically illustrated in Figure 1B.
For site-directed transgenesis, we used different cell lines. The FLP-In T-Rex HEK293 and BHK cell lines have insertion of a single FRT site linked to the Zeocin resistance gene (Invitrogen). Upstream from the FRT site is a methionine codon conferring the start codon to the Hygromycin resistance gene of the donor plasmid or minicircle after FLP-mediated recombination. Additionally, we have generated a HEK293-derived cell line, HEK293–6F with 6 insertions of the Sleeping Beauty transposon SBT/SV40-FGIP (Figure 1C). SBT/SV40 -FGIP includes a FRT-EGFP fusion gene driven by the SV40 promoter (9) and a methionine codon located upstream from the FRT site conferring the start codon to the Hygromycin resistance gene of the donor plasmid or minicircle after FLP-mediated recombination.
For the characterization of transgenesis using minicircle donor DNA, we initially used the HEK 293–6F cell line since the presence of multiple FRT sites will allow simultaneous examination of several transgenesis events in each cell clone. For FLP-mediated transgenesis, 0.5 µg cDNA5/FRT-red (or a similar amount of minicircle DNA) was transfected into HEK 293 – 6F cells. The DNA was co-transfected with 5.5 µg either FLP-recombinase expression vector pOG4 4 (Invitrogen) or pUC19 as a negative control. The next day, cells were diluted 10× and after 2 days, Hygromycin was added. Resistant colonies were formed only from cells co-transfected with pOG4 4 and pcDNA5/FRT-red or minicircle DNA, verifying that transgenesis was FLP-dependent. The number of colonies was similar for the two DNA donors (30–55 colonies per 105 transfected HEK 293–6F cells). Genomic DNA was purified from colonies independent of the fluorescence status to ensure nonbiased analysis of targeted integration events. Recombination into the FRT site was examined by PCR screening of genomic DNA using primers recognizing the SV40 promoter and the Hygromycin resistance gene (primer combination 3 and 5 indicated in Figure 1D). A 300-bp fragment was expected after recombination into the FRT site and indeed, such a PCR product was observed from all examined genomic DNA samples (Figure 2A). Sequencing confirmed precise recombination (data not shown). This clearly supports that minicircle DNA devoid of BB can be used for FLP-mediated transgenesis. We addressed BB deletion by PCR using primers located in the SV40 promoter and upstream from the CMV promoter (primer combination 3 and 4 in Figure 1D). In all six examined DNA samples from minicircle transgenesis, a 1.7-Kb fragment was detected which represents the expected size for a BglII-site ligation and, accordingly, an absence of the BB (Figure 2B). The lack of BB was confirmed by sequencing (data not shown). Southern blot analyses showed DNA rearrangements in clone #2 from the minicircle donor (Supplementary Figure S1A). Clone #3 from the minicircle donor indicated existence of transgene concatemerization (Supplementary Figure S1A).
