|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAPID COMMUNICATIONS |
School of Biological Sciences, Nanyang Technological University, Singapore
Key Words. Site-specific recombination • Human embryonic stem cells • Cre recombinase •
integrase • Plasmid transfection • 
resolvase
Correspondence: Peter Dröge, Ph.D., Nanyang Technological University, School of Biological Sciences, 60 Nanyang Drive, 637551, Singapore. Telephone: 65-6316-2809; Fax: 65-6791-3856; e-mail: pdroge{at}ntu.edu.sg
| ABSTRACT |
|---|
|
|
|---|
integrase, phage P1 Cre recombinase, and mutant 
resolvase displayed distinct activities on episomal recombination substrates. Interestingly, cofactor-independent
integrase catalyzed the integrative pathway five times more efficiently than the excisive pathway. Such a degree of directionality in hESCs could be explored for sequential gene insertions into predetermined genomic sequences. We also report an improved, easy-to-use plasmid transfection system that employs silica microspheres and, in combination with SSR, could be applied to hESC genome engineering.
| INTRODUCTION |
|---|
|
|
|---|
It has been noted that plasmid transfection is rather inefficient with hESCs [5]. In addition, published data comparing different transfection strategies are scarce. This is especially true for reagent-based methods. We address this issue and directly compare transfection efficiencies obtained with different reagents. We also describe an efficient three-component plasmid transfection system for hESCs which employs silica microspheres [6].
Site-specific DNA recombination (SSR) systems derived from prokaryotic cells are valuable tools for various applications in eukaryotic cells. Notably, the phage P1 recombinase Cre is frequently used to splice out (delete) marker genes from genomes after gene targeting and for conditional mutagenesis in model organisms, particularly in the mouse [7]. Furthermore, the Cre, Flp,
C31, and
integrase system can achieve targeted insertion of foreign DNA into predetermined artificial or natural genome sequences [810]. The latter application is especially relevant to future gene therapy approaches with hESCs because it minimizes the risk of unwanted genome alterations due to random DNA insertions, which have been reported for viral vector-based strategies [11]. A main objective, therefore, is to develop SSR systems that can be used to safely modify hESC genomes for possible clinical applications.
As a first step toward this goal, we compared the enzymatic activities of three site-specific recombinases: wild-type Cre, a modified 
resolvase (
102NLS) derived from bacterial transposon 
bearing the two recombination-activating mutations E102Y and E124Q, and Int-h/218, which is a mutant phage
integrase (E174K/E218K) that functions in the absence of cofactors in mammalian cells [1217]. It has been shown that these enzymes faithfully catalyze DNA strand transfer on their respective target sequences inside mammalian cells without adding or deleting nucleotides to or from strands in the course of the reaction, respectively [12, 13, 15, 17]. We show here that these recombinases catalyzed DNA strand transfer reactions inside hESCs on plasmid substrates after cotransfection with the respective expression vector. Whereas Cre-mediated recombination is detectable in approximately 50% of transfected cells, the integrative recombination pathway catalyzed by Int-h/218 is observed in close to 20% of transfected cells. Interestingly, the excisive recombination pathway is significantly less active in hESCs than the integrative pathway. This finding will be explored to achieve controlled site-specific gene insertions into predetermined chromosomal sequences.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Transfection and Recombination Assays
hESCs were transfected 4 days after colony transfer. Approximately 2 x 105 cells in each well were transfected using Exgen 500 (Fermentas, Hanover, PA, http://www.fermentas.com), Lipofectamine 2000 (Invitrogen Corporation, Carlsbad, CA, http://www.invitrogen.com), Fugene 6 (Roche, Basel, Switzerland, http://www.roche.com), Effectene (Qiagen GmbH, Hilden, Germany, http://www.qiagen.com), or Effectene plus 0.15 µM silica microspheres (Polysciences, Inc., Warrington, PA, http://www.polysciences.com). Transfections using Exgen 500 and Lipofectamine 2000 were performed as described [5, 23]. For Fugene 6, hESC media without serum was mixed with either 30 µl (for 2 µg DNA) or 60 µl (for 4 µg DNA) of reagent to yield a total volume of 100 µl and incubated at room temperature (RT) for 5 minutes. The appropriate amount of DNA was added to the mixture, incubated for 45 minutes, and subsequently added drop-wise to each well. Transfection using Effectene was performed as recommended by the manufacturer at a 1:25 DNA:Effectene (µg:µl) ratio. The final mixture was incubated for 15 minutes at RT before application. Transfection with Effectene and silica microspheres was performed as above, but with the addition of 1 x 109 microspheres 5 minutes after the start of the final incubation period. The reagent/DNA mixtures were incubated with hESCs for 24 hours.
Transfection efficiencies were determined with 2 µg of reporter plasmid pCMVssEGFP per well. Recombination assays were carried out with 2 µg of substrate vectors pCH-RLNRLE, p
IR for
integrative, or p
ER for
excisive recombination. Substrate vectors were cotransfected with 2 µg of respective recombinase vectors pPGKCre, pPGK
102NLS, or pCMVssInth/218 per well. pPGK or pCMV served as mock expression vectors. pCMVssEGFP was used as positive control for
integrative and excisive recombination, whereas plasmid pCH-RLE, the product vector that results from recombination on pCH-RLNRLE, served as positive control for Cre and 
resolvase reactions.
integrase recombination assays were performed using Fugene 6. Cre and 
resolvase assays were carried out with Effectene.
Immunostaining
Immunostaining of hESC colonies on Matrigel was performed using SSEA-1, SSEA-4, TRA-1-61, TRA-1-80, and OCT4 specific primary antibodies (Chemicon, Temecula, CA, http://www.chemicon.com). Antibodies were visualized with fluorescein iso-thiocyanate (FITC)labeled secondary antibodies (Chemicon). Alkaline phosphatase (AP) activity was detected according to the manufacturers protocol (Chemicon).
Flow Cytometry
Cells were washed 48 hours after transfection with phosphate-buffered saline (PBS) (Gibco, Invitrogen Corporation) and harvested through incubation with 0.05% trypsin/ethylenediaminetetraacetic acid (EDTA) for 5 minutes. Cells were collected, washed once with PBS, and resuspended in 200 µl of PBS for analysis in a FACSCalibur flow cytometer (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). A minimum of 20,000 total events were recorded per sample and analyzed using CELLQuest software (Becton, Dickinson and Company). All values were obtained after gating mock transfected (pCMV) hESCs to zero. Recombination efficiencies were calculated after subtracting mock control values from experimental values. The positive control values were taken as 100%.
Substrate and Expression Vectors
Plasmids pCMV, pCMVssInt-h/218, pCMVssEGFP, pCH-RLN-RLZ, p
IR, p
ER, pPGKCre, and pPGK
102NLS have been described [1316]. pCH-RLNRLE was constructed by inserting the coding region for enhanced green fluorescent protein (EGFP) into pCH-RLNRLZ. The EGFP gene was derived from pCMVssEGFP via polymerase chain reaction (PCR), with a short linker (GGSGG) replacing the EGFP start codon. The following primers were used in the PCR: L1EGFP-F (5'-CGGGGTACC-GGGTGGAAGCGGCGGTGTGAGCAAGGGCGAGGA-3') and L1EGFP-R (5'-CGCGGATCCGAGGCTAGAACTAGTGG-3'). Both the PCR product and vector pCH-RLNRLZ were cleaved with KpnI/BamHI. The EGFP insertion replaced a large portion of the LacZ gene and results in the expression of an N-terminal LacZ-Linker-EGFP fusion protein. The recombined product vector pCH-RLE served as positive control and was generated through transformation into 294-Cre cells. This Escherichia coli strain expresses Cre constitutively at a low level [24].
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
|
SSR in hESCs
We recently developed two new SSR systems for applications in eukaryotic cells and showed that mutant phage
integrases and the mutant 
resolvase 
102NLS, in particular, are recombination-proficient on episomal and genomic DNA substrates [1317]. However, because hESCs differ from other mammalian cells, including mouse ESCs, in features such as cell size, doubling time, and gene expression patterns [32, 33], we were interested in comparing the
and 
SSR system with the widely used Cre system inside hESCs. For this, we used substrate vectors that, when cotransfected with the respective recombinase expression vector, lead to EGFP expression as an easy read-out for recombination (Fig. 3
).
|

resolvase (taken as 100% recombination efficiency). We found that Cre recombined pCH-RLNRLE in more than half of the transfected cells (Fig. 4
resolvase. The latter finding is in agreement with previous results which showed that 
resolvase is almost as efficient as Cre in murine cells, but significantly less so in human cells [34]. We want to emphasize that we are unable to determine recombination efficiencies at plasmid level because the copy number of internalized substrate DNA molecules accessible for the recombinase is unknown. This number varies per cell, and it is reasonable to assume that a single recombined copy will suffice to produce an EGFP-positive cell. Hence, the actual percentage of recombined substrate molecules could be lower than the percentage of EGFP-expressing cells normalized to the positive control.
|
Int-h/218 recombined the substrate for integrative recombination, p
IR, in 20% of transfected cells (Fig. 4
ER, the substrate for excisive recombination. This deviates from our previous findings with HeLa cells in which this difference is less than twofold [15]. Because integrative recombination between attachment sites attB and attP generates sites attL and attR, which are then substrates for excisive recombination, our results imply that cofactor-independent
integrases display a significant degree of directionality inside hESCs. The molecular basis for this pronounced directionality in hESCs is not understood at the moment and deserves further investigation.
Our finding that the Cre recombinase is able to recombine episomal substrates in a significant fraction of cotransfected cells indicates that this recombinase is a good candidate for future hESC genome manipulations such as removal of marker genes. However, in our direct comparison with Cre, the resolvase mutant 
102NLS is significantly less active in hESCs than in CHO cells and, therefore, may be useful only for specific applications that do not require high recombination efficiencies. It is not clear at present what causes the reduced activity of 
102NLS in human cells. We can exclude differences in nuclear localization because both Cre and 
102NLS contain functional nuclear localization signals [16, 35]. One possible factor to consider may be protein modification of 
resolvase in human cells.
The
integrative pathway catalyzed by Int-h/218 is quite efficient in hESCs. Furthermore, because of the directionality that
Int-h/218 exhibited in hESCs, in particular, this SSR system could prove very valuable for targeted gene insertions into genomic attachment sites. These sites could be artificially introduced or occur naturally in the hESC genome. Examples for
integrasemediated recombination genome insertions have been reported with human Burkitts lymphoma cells and mammalian artificial chromosomes [9, 10].
Currently, we are evaluating experimentally more than 1,000 different human genome sequence tracts as potential target sites that are homologues to the 21-bp attB site. These tracts contain inverted consensus
integrase recognition sites separated by a 7-bplong spacer. Nearly all of these genomic sequences are single copy because of a unique spacer. Based on the high fidelity of the
system in mammalian cells (Dröge, unpublished results), it should be possible to specifically direct the insertion of vectors into a few selected sites that may be more accessible in hESC chromatin. Because a strategy of gene insertion employing the pair attB/attP, or derivatives thereof, will generate attL/attR sites in the genome, integration could be stable even in the continued presence of the recombinase. This should allow for sequential targeting events into different genomic loci.
| ACKNOWLEDGMENTS |
|---|
|
|
|---|
| REFERENCES |
|---|
|
|
|---|
integrase mutants. J Mol Biol 2000;296:11751181.[CrossRef][Medline]
integrases: implications for synaptic complex formation and the reactivity of episomal DNA segments. J Mol Biol 2002;319:305314.[CrossRef][Medline]
integrase. Genesis 2002;32:203208.[CrossRef][Medline]
resolvase mutants: implications for the topology of episomal DNA. FEBS Lett 2000;471:147150.[CrossRef][Medline]
resolvase in eukaryotischen Zellen. Dissertation, University of Cologne, Germany, 2001.This article has been cited by other articles:
![]() |
S. M. Tan, S. T. Wang, H. Hentze, and P. Droge A UTF1-based selection system for stable homogeneously pluripotent human embryonic stem cell cultures Nucleic Acids Res., September 25, 2007; 35(18): e118 - e118. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Haupt, F. Edenhofer, M. Peitz, A. Leinhaas, and O. Brustle Stage-Specific Conditional Mutagenesis in Mouse Embryonic Stem Cell-Derived Neural Cells and Postmitotic Neurons by Direct Delivery of Biologically Active Cre Recombinase Stem Cells, January 1, 2007; 25(1): 181 - 188. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Bolusani, C.-H. Ma, A. Paek, J. H. Konieczka, M. Jayaram, and Y. Voziyanov Evolution of variants of yeast site-specific recombinase Flp that utilize native genomic sequences as recombination target sites Nucleic Acids Res., October 6, 2006; 34(18): 5259 - 5269. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
