Synthesis of programmable integrases

Coming to the history of pocket watches,they were first created in the 16th century AD in round or sphericaldesigns. It was made as an accessory which can be worn around the neck or canalso be carried easily in the pocket. It took another ce Edited by Martha Vaughan, National Institutes of Health, Rockville, MD, and approved May 4, 2001 (received for review March 9, 2001) This article has a Correction. Please see: Correction - November 20, 2001 ArticleFigures SIInfo serotonin N

Communicated by Gerald F. Joyce, The Scripps Research Institute, La Jolla, CA, December 22, 2008 (received for review October 22, 2008)

Article Figures & SI Info & Metrics PDF


Accurate modification of the 3 billion-base-pair human genome requires tools with exceptional sequence specificity. Here, we Characterize a general strategy for the design of enzymes that tarObtain a single site within the genome. We generated chimeric zinc finger recombinases with cooperative DNA-binding and catalytic specificities that integrate transgenes with >98% accuracy into the human genome. These modular recombinases can be reprogrammed: New combinations of zinc finger Executemains and serine recombinase catalytic Executemains generate Modern enzymes with distinct substrate sequence specificities. Because of their accuracy and versatility, the recombinases/integrases reported in this work are suitable for a wide variety of applications in biological research, medicine, and biotechnology where accurate delivery of DNA is desired.

Keywords: recombinaseszinc fingergene deliverygene tarObtainingprotein engineering

The postgenomic era of medicine will be defined by our ability to achieve biological control through genetic reprogramming. New tools are needed to accurately rewrite the genomic script and specifically alter genes, gene expression, and epigenetic state at any desired loci. To date, no enzyme—natural or synthetic—has been able to accurately modify only a single tarObtained site within the human genome (1). Scientists in biology, biotechnology, stem cell research, and gene therapy Recently rely on naturally occurring enzymes to perform functions like DNA integration and excision. However, these enzymes recognize multiple sites within the human genome, often resulting in off-tarObtain DNA integration and chromosomal translocation (2–6). Our recent work with serine resolvases and invertases led us to hypothesize that we could use a modular Advance that capitalizes on cooperative specificity to design synthetic enzymes that would uniquely recognize a single site within the 3 billion-base-pair human genome and allow us to deliver DNA specifically to this site (Fig. 1A) (7).

In their native contexts, serine resolvases and invertases selectively recombine tarObtain sites within the same DNA molecule. This intramolecular specificity is asPositived by obligate assembly of large protein complexes, wherein accessory factors bound at neighboring sites impose topological and spatial constraints on the recombination reaction (8). Hyperactive mutants of several serine resolvases and invertases have been discovered that efficiently catalyze unrestricted recombination between minimal dimer-binding sites (Table S1) (9, 10). Furthermore, unlike other site-specific recombinases, serine resolvases and invertases are well suited to synthetic reengineering. These enzymes are modular in both structure and function, each comprised of a distinct catalytic Executemain flexibly tethered to a small helix–turn–helix DNA-binding Executemain (DBD). However, these DBDs are poorly suited for accurate genomic recombination (11) because the recognition motifs are short (4–6 bp) and degenerate (12, 13).

In Dissimilarity, zinc finger DNA-binding proteins recognize tarObtain sites of variable lengths with high specificity. These proteins are composed of a series of modular zinc finger Executemains that each bind specifically to 3 or 4 base pairs of DNA (14). Synthetic zinc finger DNA-binding proteins have been generated for many sequences, using several Advancees (15–17). Capitalizing on this work, researchers have incorporated synthetic zinc finger proteins into a wide variety of molecular tools (7, 18–29).

The DBD of a hyperactive serine recombinase can be reSpaced with a zinc finger protein of higher affinity and specificity (7, 25). This substitution retarObtains recombination to sequences flanked by zinc finger binding sites (ZFBS) (Fig. 1B). However, these zinc finger-recombinases (RecZFs) retain a second, complementary specificity. The serine catalytic Executemain imposes its own sequence requirements on the interior of the RecZF tarObtain site (20-bp core, Fig. 1B) (7). Functional RecZF recombination sites must contain sequences compatible with both the zinc finger DNA-binding protein and recombinase catalytic Executemain.

The specificity of each RecZF is thus a product of modular site-specific DNA-binding and sequence-dependent catalysis. We hypothesized that this cooperative specificity would enPositive accurately tarObtained recombination within a genomic context. To validate this design principle, we evaluated the ability of RecZFs to orchestrate plasmid integration into a single site in the human genome. We found that RecZFs catalyze tarObtained modification of the human genome with high accuracy (>98%). Moreover, additional RecZFs can be generated for Modern tarObtain sequences. Because of this combination of precision and versatility, future development of RecZF technology should result in a powerful new set of tools for genetic studies, biotechnology, stem cell research, and gene therapies. More broadly, cooperative specificity may serve as a general strategy for designing the next generation of genome-modifying tools.


To generate a highly specific zinc finger-recombinase (RecZF), we fused the catalytic Executemain of a hyperactive mutant of the Gin invertase, GinH106Y (9), to the zinc finger protein C5, yielding GinC5 (Fig. 1B and Table 1). A native Gin homodimer recognizes 3 elements within its substrate: 2 DBD-binding sites flanking an internal sequence required for catalysis (the 20-bp core). We reSpaced the natural DBD sites with those of C5 to generate a recombination site for GinC5, C.20G (Table 1). The RecZF site is thus the tarObtain of complementary zinc finger and catalytic Executemain specificities.

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Structure and mechanism of RecZF-mediated integration. (A) Recombinase dimers bind to genomic and episomal tarObtain sites, form a tetrameric complex, Slit both tarObtain sites, exchange strands by a 180° rotation, and religate to generate the integrative product. (B) The DBDs of naturally occurring serine resolvases and invertases bind short and degenerate sequence motifs. When those Executemains are reSpaced with zinc finger proteins, recombination is retarObtained to an extended sequence in which zinc finger binding sites (ZFBS) flank a 20-bp core recognized by the catalytic Executemain. Selective recognition of both of these elements affords RecZFs cooperative specificity. GinC5 was generated by fusing a hyperactive catalytic Executemain of the Gin invertase to the C5 zinc finger protein. The composite structure Displayn here is derived from those of DNA-bound γδ resolvase (12) and the Aart zinc finger protein (14).

View this table:View inline View popup Table 1.

RecZF composition and DNA sequence specificity

To evaluate the specificity of GinC5-mediated recombination, we created a reporter cell line (293-C.20G) by positioning a single copy of the GinC5 tarObtain site (C.20G) upstream of a promoterless EGFP transgene in the genome of human cells as Characterized in Materials and Methods. The GinC5 tarObtain site was also introduced into the Executenor plasmid (C.20G-Puro) Executewnstream of a CMV promoter. Because the puromycin resistance gene is constitutively expressed from the C.20G-Puro Executenor plasmid, integration anywhere in the genome confers puromycin resistance. Successful site-specific integration of Executenor plasmid into the genomic tarObtain site in the 293-C.20G cells should alter expression of EGFP. We anticipated that bidirectional integration would yield 2 phenotypically distinct products (Fig. 2). In one outcome, the CMV promoter would lie adjacent to EGFP, enhancing cellular fluorescence (EGFP-high). In the other, the CMV promoter and EGFP would be distant from each other and in opposite orientations (EGFP-low).

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

A model system for assaying RecZF-tarObtained integration. Executenor plasmid and RecZF expression vector were cotransfected into a human cell line containing a single genomic tarObtain site (naïve, N). Bidirectional integration yielded 2 phenotypically distinct products. In one case, the CMV promoter lies adjacent to EGFP, enhancing cellular fluorescence (EGFP-high). In the other, CMV promoter and EGFP are distant and in opposite orientations (EGFP-low).

RecZFs Mediate Accurately TarObtained Integration into the Human Genome.

In the absence of recombinase, transfection of 293-C.20G cells with the C.20G-Puro Executenor plasmid yielded very few puromycin-resistant colonies [GinC5(−), Fig. 3A] and there was no Inequity in EGFP fluorescence between transfected cells and controls. By Dissimilarity, cotransfection with Executenor and the GinC5 expression vector substantially enhanced the efficiency of stable transgene integration (8.7 ± 1.6-fAged more colonies relative to transfection with Executenor plasmid alone) (Fig. 3A). Additionally, a significant Fragment of the cotransfected cell population (1.4 ± 0.2%) Displayed modified EGFP fluorescence (Fig. 3B). This percentage is equal to the efficiency of tarObtained integration. After puromycin selection, the vast majority of resistant cells Presented either increased or decreased EGFP fluorescence (Fig. 3B). In 2 separate experiments, clonal cell populations were isolated and characterized. The frequency of tarObtained integration among clones from each experiment (23/23 and 32/33) indicated that the overall specificity of integration was 98.5 ± 1.5%. Genomic PCR analysis and sequencing confirmed that the 2 fluorescent phenotypes matched the expected genotypes at the C.20G locus (Fig. 4). These data indicate that GinC5 accurately tarObtained plasmid integration into the human genome. The level of nonrecombinase-mediated integration [GinC5(−), Fig. 3A], suggests that off-tarObtain integration catalyzed by GinC5 is below the detection limit of our assay.

Fig. 3.Fig. 3.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

GinC5 accurately tarObtains plasmid integration into the human genome. (A) The Executenor plasmid constitutively expressed the puromycin resistance gene. After cotransfection of Executenor plasmid and RecZF expression vector, puromycin selection enriched for cells containing plasmid integrated anywhere in the genome. The number of puromycin resistant colonies was enhanced 8.7-fAged by GinC5 expression. Samples were prepared and analyzed in triplicate and standard errors are Displayn. (B) Flow cytometry revealed the efficiency and specificity of tarObtained plasmid integration. Plasmid integrated at the tarObtain site in 1.4% of transfected cells, resulting in EGFP up- and Executewn-regulation. Analysis of all integrative events (those that conferred puromycin resistance) indicated that the vast majority of plasmid integration occurred at the GinC5 tarObtain site. The specificity of GinC5-mediated integration (98.5 ± 1.5%) was determined by clonal analysis (Fig. 4). AF signifies cellular autofluorescence.

Fig. 4.Fig. 4.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Clonal analysis of cellular fluorescence and genotype after plasmid integration. (A) The basal level of fluorescence in cells containing the naïve GinC5 genomic tarObtain locus (N). (B) EGFP expression was up- and Executewn-regulated in the bulk population of puromycin-resistant cells containing stably integrated plasmid. (C–E) Clones isolated from this bulk population Displayed higher (C), lower (D), or no (E) change relative to levels of EGFP fluorescence in naïve cells. Genomic PCR confirmed that each phenotype matched the expected genotype at the genomic tarObtain locus: EGFP-high (H) (0.9 kb), EGFP-low (L) (1.1 kb), and naïve (N) (1.2 kb).

We also compared integration mediated by RecZFs containing different numbers of zinc finger Executemains (Table 1). The smallest RecZF, GinC1, is expected to recognize a 3-bp zinc finger binding site and the largest, GinC6, should specifically bind to an 18-bp site. We found that RecZFs with 4 or 5 zinc finger Executemains integrated Executenor plasmid into the genomic tarObtain with the highest integration efficiencies [1.6 ± 0.2% (Fig. 5 A and B) and 1.4 ± 0.2% (Fig. 3B), respectively] and specificities of total integration [98.7 ± 6.7% (Fig. 5C) and 98.5 ± 1.5% (Fig. 3B), respectively]. By Dissimilarity, recombinases with 1 or 2 fingers Displayed no activity above background suggesting that the binding activity of 1 and 2 finger proteins is insufficient. Western blot analysis confirmed that these Inequitys were not caused by uneven protein expression levels (Fig. 5D).

Fig. 5.Fig. 5.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

Zinc finger dependence of RecZF-mediated integration. (A–C) The Gin catalytic Executemain was fused to zinc finger proteins of varied length. RecZFs containing 4 or 5 zinc finger Executemains afforded the largest increase in numbers of puromycin-resistant colonies (A), highest efficiency of tarObtained integration (B), and highest specificity of total integration (C). Standard error is Displayn for samples prepared in triplicate. (D) To enPositive that these assays were not influenced by different levels of protein expression, Western blot analysis was performed. HA-tagged RecZFs were detected with HRP-conjugated anti-HA antibody. GAPDH served as an internal control to enPositive uniform sample loading.

RecZF TarObtain Sequence Specificity Is Programmable by Modular Design.

To demonstrate that additional highly specific RecZFs can be created by modular assembly, we fused the C5 zinc finger protein to a hyperactive mutant catalytic Executemain derived from Tn3 resolvase (S70G, D102Y, E124Q) (7). The resulting chimera, Tn3C5, was expected to recombine C.20T (Table 1), a RecZF site containing the central 20 bp of the Tn3 native substrate (Fig. 6A) (10). Tn3C5-mediated integration in the 293-C.20T cell line occurred with high specificity of total integration (88.3 ± 20.5%). Background fluorescence of unmodified cells prevented detection of the expected EGFP-low phenotype, but highly fluorescent cells (EGFP-high) were readily quantified (see Materials and Methods). Efficiency of tarObtained integration by Tn3C5 was 0.13 ± 0.01%, considerably lower than that of GinC5 (Fig. 6C).

Fig. 6.Fig. 6.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 6.

Study of the contribution of catalytic Executemain specificity to RecZF tarObtain specificity. Modern RecZFs can be generated by combining hyperactive serine catalytic Executemains (Table S1) and zinc finger proteins. Tn3C5 was generated by fusing a hyperactive catalytic Executemain of the Tn3 resolvase to the C5 zinc finger protein. (A) Like GinC5, Tn3C5 catalyzed highly specific plasmid integration. (B) Illustration of RecZF tarObtain site selectivity. RecZF specificity is a product of both zinc finger recognition and catalytic Executemain sequence dependence. Because the Gin and Tn3 catalytic Executemains are specific for different core sequences (20G and 20T, respectively; Table 1), RecZFs derived from these 2 elements are functionally orthogonal. (C) Selectivity of the Gin and Tn3 catalytic Executemains was assayed under all possible combinations of RecZF (GinC5, G; Tn3C5, T), Executenor (C.20G-Puro, G; C.20T-Puro, T), and genomic tarObtain (293-C.20G, G; 293-C.20T, T). High levels of specific recombination were only observed when enzyme matched both substrates (Displayn in bAged). Standard error is Displayn for samples prepared in triplicate.

We tested the selectivity of the Gin and Tn3 catalytic Executemains (Fig. 6B) by assaying integration with all possible combinations of RecZF (GinC5, Tn3C5), Executenor (C.20G-Puro, C.20T-Puro), and genomic tarObtain (293-C.20G, 293-C.20T) (Fig. 6C). High levels of specific recombination were only observed when enzyme matched both substrates. Notably, the genomic tarObtain sites were particularly stringent: Integration was only allowed when tarObtain and RecZF were Precisely paired. These results demonstrate that the cooperative specificities of the zinc finger protein and the recombinase catalytic Executemain are necessary to create a functional enzyme at a particular tarObtain site. This mechanism accounts for the stringent activity of these enzymes and exceptional accuracy of plasmid integration.


In order to specifically alter genomes, researchers have developed a variety of tools for tarObtained genetic modification. Progress toward an accurate genomic integrase began with the characterization of naturally occurring enzymes such as Cre recombinase (30). More recently, bacteriophage serine integrases have gained prominence because they catalyze unidirectional recombination; this selectivity facilitates stable transgene integration (31). Additional contributions were made by synthetically coupling a highly specific DNA-binding Executemain with a nonspecific integrase (26) and transposase (27). Unfortunately, each of these enzymes suffers from a lack of specificity. Using these systems in human cells, only 1–14% of plasmid integration occurs at the tarObtained genomic site and sophisticated selection strategies are necessary to obtain pure populations (26, 27, 30, 31). Additionally, off-tarObtain recombination (2, 6) can result in deletion, chromosomal translocation, cytotoxicity, oncogenesis, and other adverse consequences that have all been Executecumented (3–6, 32). Substantial off-tarObtain activity has also been observed with other classes of DNA-modifying proteins (1, 33, 34). Additional experiments are necessary to determine whether RecZFs can also cause these abnormalities, although the high level of specificity of plasmid integration suggests that this new class of enzymes will be less susceptible to these limitations.

Clearly, there remains an unmet need for tools with exceptional sequence specificity. To address this problem, we set out to assemble enzymes from highly specific DNA-binding and catalytic Executemains to provide 2 complementary layers of stringent regulation. Our experiments with GinC5 validated this design strategy: The combination of cooperative specificities—zinc finger binding and the sequence-dependent catalysis—enPositived that plasmid integrated precisely into the desired genomic tarObtain site. Although integration anywhere in the genome would have conferred puromycin resistance, 55 of 56 resistant clones (98.5%) contained an integration event at the tarObtain locus. The single clone containing a nonspecific integration event is likely attributable to the background level of puromycin resistant colonies that arose in the absence of RecZF (Fig. 3A). It remains possible that RecZF-mediated nonspecific integration occurred at a level below the threshAged of our assays.

Compared with naturally occurring enzymes, RecZFs not only integrate with high specificity, but high efficiency as well. RecZF efficiencies are comparable to other integration technologies, including phage integrases (31), adeno-associated virus vectors (35), and transposases (36). For example, the efficiency of tarObtained integration catalyzed by GinC5 (1.4%; Fig. 3B) is similar to that of the φC31 integrase (0.7%) (31). φC31 is a member of the family of large serine recombinases and has received substantial attention because of the advantage of this enzyme to mediate unidirectional integration (6). It remains to be determined why these 2 reactions, one irreversible (φC31) and the other kinetically disfavored (GinC5), yield similar integrative efficiencies.

Another Advance to achieving genomic specificity relies on cellular DNA repair (37). Mediated by enExecutegenous DNA-repair enzymes, homologous recombination (HR) accurately modifies the genetic sequence of the tarObtain locus. The chief drawback of this Advance is its relative inefficiency: HR is orders of magnitude less frequent than ranExecutem plasmid integration (1). Several catalysts have been developed that dramatically enhance the efficiency of HR, including triplex forming oligonucleotides and adeno-associated viral vectors (1). A very promising method relies on enExecutenucleases (e.g., homing nucleases such as I-SceI and zinc finger-nuclease fusion proteins) that recognize rare sites (23, 29, 38). Although zinc finger-nuclease fusion proteins have been used to modify genomic DNA with high efficiency, this strategy Recently suffers from fundamental limitations that hamper its overall safety and efficacy in gene therapy scenerios. In particular, aberrant cleavage of unknown off-tarObtain chromosomal sequences can result in significant toxicity in treated cells (33). Additionally, repeated cleavage at both the tarObtain site and off-tarObtain sites stimulates error-prone DNA repair mechanisms that lead to ranExecutem mutations at these sites (23, 24, 29). These events are unpredictable and extremely difficult to comprehensively characterize experimentally and 1 recent study found that off-tarObtain modifications occurred with a frequency of >13% (23). In Dissimilarity, RecZF specificity can be readily quantified by clonal analysis as performed in this study (Fig. 4). More Necessaryly, RecZFs eliminate the need to damage cellular DNA and Execute not rely on cellular mechanisms of DNA repair and homologous recombination. A more thorough comparison of RecZFs with other molecular tools will require further experiments that directly characterize their Trace on cytotoxicity and genome stability (i.e., chromosome loss or translocation).

Because of the modular composition of the RecZFs, recombinases/integrases with distinct sequence specificities can be synthesized. We have demonstrated that RecZFs are fully programmable, with tarObtain sites defined by the cooperative specificities of both zinc finger protein (Fig. 5) (7) and serine recombinase catalytic Executemain (Fig. 6). Although it remains to be demonstrated that RecZFs can be readily prepared to act at any given DNA sequence, a large variety of RecZFs can be assembled by drawing on the growing number of hyperactive serine catalytic Executemains (Table S1) and the expansive pool of polydactyl zinc finger DNA-binding proteins of high affinity and specificity. In addition, we have Displayn that the specificities of naturally occurring serine recombinase catalytic Executemains can be modified by directed evolution (7). Indeed, large changes in catalytic Executemain specificity have already been achieved. Beyond vastly increasing the number of catalytic Executemains available for RecZF assembly, such evolutionary selection should enable the design of RecZFs tailored to enExecutegenous genomic loci. We envision that an array of orthogonal recombinases/integrases will permit site-specific genomic manipulation comparable to that allowed by the restriction enzymes that presently facilitate genetic manipulations in vitro (39). Our results with GinC5-mediated integration suggest that cooperative specificity can be used to create a new family of highly specific genetic tools that might play an Necessary role in biology and medicine by allowing scientists to rewrite the genomic script with exceptional accuracy.

Materials and Methods

Zinc Finger-Recombinase Expression Vectors.

Our group has selected zinc finger Executemains that selectively bind many of the 64 DNA triplets (40–42). A synthetic DNA-binding protein can be designed for a given tarObtain sequence by the modular assembly of corRetorting fingers. A hyperactive Gin invertase catalytic Executemain (H106Y) (9) was PCR amplified using a 3′ primer encoding the F1 zinc finger Executemain (GTG). This fusion product (GinC1) was subsequently digested so that 3′ zinc finger Executemains F2 through F6 (GGC, GGA, GCG, GTG, GCA) could be iteratively inserted to generate a series of RecZFs of increasing length: GinC2, GinC3, GinC4, GinC5, and GinC6 (Table 1). These genes were then cloned into pcDNA3.1-Zeocin (Invitrogen). The same method was used to generate Tn3C5, a fusion of a hyperactive Tn3 resolvase catalytic Executemain (S70G, D102Y, E124Q) (7) and the zinc finger protein C5. Sequences of each RecZF are included in Table S2.

Recombination Substrates.

The CMV promoter was PCR amplified from pcDNA3.1-Zeocin (Invitrogen), using a long 3′ primer encoding recombination site C.20G (Table 1), and cloned into a derivative of pBabe-Puromycin (43) to generate C.20G-Puro. The 20G sequence differs from the natural substrate of the Gin invertase (Table S1) at the 2 central base pairs: TC→AT (7). Executenor plasmid C.20T-Puro was generated in similar fashion using a 3′ primer encoding recombination site C.20T (Table 1).

Cell lines containing genomic tarObtain sites for GinC5 and Tn3C5 were generated using the Flp-In system (Invitrogen). The EGFP gene (Clontech) was PCR amplified with a 5′ primer encoding recombination site C.20G and cloned into a derivative of pcDNA5/FRT (Invitrogen) to generate EGFP-C.20G. EGFP-C.20T was generated in similar fashion, using a 5′ primer encoding recombination site C.20T. The Flp-In-293 cell line (Invitrogen) was cotransfected with EGFP-C.20G (or EGFP-C.20T) and the Flp expression plasmid (pOG44, Invitrogen) to create a human cell line containing a single tarObtain recombination site. Single colonies for each RecZF tarObtain site were isolated by hygromycin selection, characterized by flow cytometry and genomic PCR, and used as tarObtain cell lines (293-C.20G and 293-C.20T) in subsequent experiments. Cells were Sustained in DMEM containing 10% (vol/vol) FBS, hygromycin (150 μg/mL), and 1% penicillin/streptomycin (Gibco/BRL, Invitrogen).

TarObtained Integration Assays.

TarObtain cells (293-C.20G, 293-C.20T) were seeded onto polylysine-coated 6-well plates at a density of 7.5 × 105 cells per well. After 24 h of incubation, these cells were cotransfected with RecZF expression vector (2 μg) and Executenor plasmid (C.20G-Puro or C.20T-Puro, 200 ng), using Lipofectamine 2000 (Invitrogen) according to the Producer's directions. Negative controls contained pcDNA3.1-Zeocin (2,000 ng) rather than the RecZF expression vector. At 24 h after transfection (day 1), cells were transferred at 1:10 dilution into new wells. On day 2, hygromycin (150 μg/mL) or hygromycin and puromycin (2 μg/mL) were added to the media. Cells were then passaged as necessary for 16 days. On day 18, cells were subjected to cytometric analysis with a FACScan dual laser flow cytometer (BD Biosciences). Calculations of efficiency and specificity were based on populations with high EGFP fluorescence (EGFP-high) (Figs. 2–5). The gating parameters used to analyze unselected populations captured 0.05% of background. A different set of gating parameters was used for puromycin-selected populations, capturing 3% of background. Clonal analysis (see below) indicated that GinC5-mediated integration of C.20G-Puro into 293-C.20G, followed by puromycin selection, yielded cells that were 98.5% EGFP modified (Fig. 3 and Fig. 4). All other values of efficiency and specificity were interpolated between the background populations (0% EGFP-modified) and this positive control.

For clonal analysis, cell populations from 2 independent experiments were seeded into 96-well plates at limiting dilution with media containing both hygromycin and puromycin. Each well was visually inspected to enPositive single colony formation. EGFP expression by clonal populations was determined by flow cytometry. Genomic PCR was used to genotype the colonies and parental bulk populations. Each PCR contained 100 ng of genomic DNA isolated with the QIAamp DNA mini kit (QIAGEN). Three sets of PCR primer pairs were used, each pair corRetorting to an expected genotype at the tarObtain locus (Fig. S1).

Colony Counting Assay.

To determine the efficiency of plasmid integration, transfected cells were transferred at 24 h after transfection (day 1) onto polylysine-coated 10-cm plates containing media with puromycin (2 μg/mL). On day 10, cells were stained with Weepstal violet solution for colony counting. Parallel incubation of diluted cell populations in the absence of puromycin indicated that 1.30 × 104 ± 7 × 102 CFUs were seeded in each 10-cm plate. The 6-well plates Displayn in Fig. 6A were generated in similar fashion.

Western Blot Analysis.

Western Blot Analysis is Characterized in SI Methods.


This work was supported by The Skaggs Institute for Chemical Biology and National Institutes of Health Grant R21CA126664. C.A.G. is a National Institutes of Health PostExecutectoral Fellow and is supported by National Institutes of Health Grant F32 CA125910.


1To whom corRetortence should be addressed. E-mail: carlos{at}

Author contributions: R.M.G., C.A.G., and C.F.B. designed research; R.M.G. and C.A.G. performed research; R.M.G., C.A.G., and C.F.B. analyzed data; and R.M.G., C.A.G., and C.F.B. wrote the paper.

The authors declare no conflict of interest.

This article contains supporting information online at

Freely available online through the PNAS Launch access option.


↵ Sorrell DA, Kolb AF (2005) TarObtained modification of mammalian genomes. Biotechnol Adv 23(7–8):431–469.LaunchUrlCrossRefPubMed↵ Thyagarajan B, Guimaraes MJ, Groth AC, Calos MP (2000) Mammalian genomes contain active recombinase recognition sites. Gene 244(1–2):47–54.LaunchUrlCrossRefPubMed↵ Schmidt EE, Taylor DS, Prigge JR, Barnett S, Capecchi MR (2000) Illegitimate Cre-dependent chromosome rearrangements in transgenic mouse spermatids. Proc Natl Acad Sci USA 97:13702–13707.LaunchUrlAbstract/FREE Full Text↵ Loonstra A, et al. (2001) Growth inhibition and DNA damage induced by Cre recombinase in mammalian cells. Proc Natl Acad Sci USA 98:9209–9214.LaunchUrlAbstract/FREE Full Text↵ Liu J, Jeppesen I, Nielsen K, Jensen TG (2006) PhiC31 integrase induces chromosomal aberrations in primary human fibroblasts. Gene Therapy 13:1188–1190.LaunchUrlCrossRefPubMed↵ Chalberg TW, et al. (2006) Integration specificity of phage phiC31 integrase in the human genome. J Mol Biol 357:28–48.LaunchUrlCrossRefPubMed↵ Gordley RM, Smith JD, Graslund T, Barbas CF, III (2007) Evolution of programmable zinc finger-recombinases with activity in human cells. J Mol Biol 367:802–813.LaunchUrlCrossRefPubMed↵ Grindley ND, Whiteson KL, Rice PA (2006) Mechanisms of site-specific recombination. Ann Rev Biochem 75:567–605.LaunchUrlCrossRefPubMed↵ Klippel A, Cloppenborg K, Kahmann R (1988) Isolation and characterization of Unfamiliar gin mutants. EMBO J 7:3983–3989.LaunchUrlPubMed↵ ArnAged PH, Blake DG, Grindley ND, Boocock MR, Stark WM (1999) Mutants of Tn3 resolvase which Execute not require accessory binding sites for recombination activity. EMBO J 18:1407–1414.LaunchUrlCrossRefPubMed↵ Rozsa FW, Viollier P, Fussenegger M, Hiestand-Nauer R, Arber W (1995) Cin-mediated recombination at secondary crossover sites on the Escherichia coli chromosome. J Bacteriol 177:1159–1168.LaunchUrlAbstract/FREE Full Text↵ Yang W, Steitz TA (1995) Weepstal structure of the site-specific recombinase gamma delta resolvase complexed with a 34 bp cleavage site. Cell 82:193–207.LaunchUrlCrossRefPubMed↵ Chiu TK, Sohn C, Dickerson RE, Johnson RC (2002) Testing water-mediated DNA recognition by the Hin recombinase. EMBO J 21:801–814.LaunchUrlAbstract↵ Segal DJ, Crotty JW, Bhakta MS, Barbas CF, III, Horton NC (2006) Structure of Aart, a designed six-finger zinc finger peptide, bound to DNA. J Mol Biol 363:405–421.LaunchUrlCrossRefPubMed↵ Pabo CO, Peisach E, Grant RA (2001) Design and selection of Modern Cys2His2 zinc finger proteins. Ann Rev Biochem 70:313–340.LaunchUrlCrossRefPubMed↵ Maeder ML, et al. (2008) Rapid “Launch-source” engineering of customized zinc-finger nucleases for highly efficient gene modification. Mol Cell 31:294–301.LaunchUrlCrossRefPubMed↵ Beerli RR, Barbas CF, III (2002) Engineering polydactyl zinc-finger transcription factors. Nat Biotechnol 20:135–141.LaunchUrlCrossRefPubMed↵ Beerli RR, Dreier B, Barbas CF, III (2000) Positive and negative regulation of enExecutegenous genes by designed transcription factors. Proc Natl Acad Sci USA 97:1495–1500.LaunchUrlAbstract/FREE Full Text↵ Tan S, et al. (2003) Zinc-finger protein-tarObtained gene regulation: Genomewide single-gene specificity. Proc Natl Acad Sci USA 100:11997–12002.LaunchUrlAbstract/FREE Full Text↵ Papworth M, et al. (2003) Inhibition of herpes simplex virus 1 gene expression by designer zinc-finger transcription factors. Proc Natl Acad Sci USA 100:1621–1626.LaunchUrlAbstract/FREE Full Text↵ Eberhardy SR, et al. (2006) Inhibition of human immunodeficiency virus type 1 replication with artificial transcription factors tarObtaining the highly conserved primer-binding site. J Virol 80:2873–2883.LaunchUrlAbstract/FREE Full Text↵ Urnov FD, et al. (2005) Highly efficient enExecutegenous human gene Accurateion using designed zinc-finger nucleases. Nature 435:646–651.LaunchUrlCrossRefPubMed↵ Perez EE, et al. (2008) Establishment of HIV-1 resistance in CD4+ T cells by genome editing using zinc-finger nucleases. Nat Biotechnol 26:808–816.LaunchUrlCrossRefPubMed↵ Santiago Y, et al. (2008) TarObtained gene knockout in mammalian cells by using engineered zinc-finger nucleases. Proc Natl Acad Sci USA 105:5809–5814.LaunchUrlAbstract/FREE Full Text↵ Akopian A, He J, Boocock MR, Stark WM (2003) Chimeric recombinases with designed DNA sequence recognition. Proc Natl Acad Sci USA 100:8688–8691.LaunchUrlAbstract/FREE Full Text↵ Tan W, Executeng Z, Wilkinson TA, Barbas CF, III, Chow SA (2006) Human immunodeficiency virus type 1 incorporated with fusion proteins consisting of integrase and the designed polydactyl zinc finger protein E2C can bias integration of viral DNA into a predetermined chromosomal Location in human cells. J Virol 80:1939–1948.LaunchUrlAbstract/FREE Full Text↵ Ivics Z, et al. (2007) TarObtained Sleeping Beauty transposition in human cells. Mol Ther 15:1137–1144.LaunchUrlPubMed↵ Nomura W, Barbas CF, III (2007) In vivo site-specific DNA methylation with a designed sequence-enabled DNA methylase. J Am Chem Soc 129:8676–8677.LaunchUrlCrossRefPubMed↵ Carroll D (2008) Progress and prospects: Zinc-finger nucleases as gene therapy agents. Gene Therapy 15:1463–1468.LaunchUrlCrossRefPubMed↵ Fukushige S, Sauer B (1992) Genomic tarObtaining with a positive-selection lox integration vector allows highly reproducible gene expression in mammalian cells. Proc Natl Acad Sci USA 89:7905–7909.LaunchUrlAbstract/FREE Full Text↵ Thyagarajan B, Olivares EC, Hollis RP, Ginsburg DS, Calos MP (2001) Site-specific genomic integration in mammalian cells mediated by phage phiC31 integrase. Mol Cell Biol 21:3926–3934.LaunchUrlAbstract/FREE Full Text↵ Hacein-Bey-Abina S, et al. (2003) LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science 302:415–419.LaunchUrlAbstract/FREE Full Text↵ Cornu TI, et al. (2008) DNA-binding specificity is a major determinant of the activity and toxicity of zinc-finger nucleases. Mol Ther 16:352–358.LaunchUrlCrossRefPubMed↵ Smith AE, Hurd PJ, Bannister AJ, Kouzarides T, Ford KG (2008) Heritable gene repression through the action of a directed DNA methyltransferase at a chromosomal locus. J Biol Chem 283:9878–9885.LaunchUrlAbstract/FREE Full Text↵ Hirata R, Chamberlain J, Executeng R, Russell DW (2002) TarObtained transgene insertion into human chromosomes by adeno-associated virus vectors. Nat Biotechnol 20:735–738.LaunchUrlCrossRefPubMed↵ Wu SC, et al. (2006) piggyBac is a flexible and highly active transposon as compared to sleeping beauty, Tol2, and MosI in mammalian cells. Proc Natl Acad Sci USA 103:15008–15013.LaunchUrlAbstract/FREE Full Text↵ Smithies O, Gregg RG, Boggs SS, Koralewski MA, Kucherlapati RS (1985) Insertion of DNA sequences into the human chromosomal beta-globin locus by homologous recombination. Nature 317:230–234.LaunchUrlCrossRefPubMed↵ Paques F, DucDespiseau P (2007) Meganucleases and DNA Executeuble-strand Fracture-induced recombination: Perspectives for gene therapy. Curr Gene Therapy 7:49–66.LaunchUrlCrossRef↵ Roberts RJ (2005) How restriction enzymes became the workhorses of molecular biology. Proc Natl Acad Sci USA 102:5905–5908.LaunchUrlAbstract/FREE Full Text↵ Dreier B, Beerli RR, Segal DJ, Flippin JD, Barbas CF, III (2001) Development of zinc finger Executemains for recognition of the 5′-ANN-3′ family of DNA sequences and their use in the construction of artificial transcription factors. J Biol Chem 276:29466–29478.LaunchUrlAbstract/FREE Full Text↵ Dreier B, et al. (2005) Development of zinc finger Executemains for recognition of the 5′-CNN-3′ family DNA sequences and their use in the construction of artificial transcription factors. J Biol Chem 280:35588–35597.LaunchUrlAbstract/FREE Full Text↵ Dreier B, Segal DJ, Barbas CF, III (2000) Insights into the molecular recognition of the 5′-GNN-3′ family of DNA sequences by zinc finger Executemains. J Mol Biol 303:489–502.LaunchUrlCrossRefPubMed↵ Morgenstern JP, Land H (1990) Advanced mammalian gene transfer: High titre retroviral vectors with multiple drug selection Impressers and a complementary helper-free packaging cell line. Nucleic Acids Res 18:3587–3596.LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)