Implications of cell cycle progression on functional sequence ... - Nature

Gene Therapy (2005) 12, 546–551 & 2005 Nature Publishing Group All rights reserved 0969-7128/05 $30.00 www.nature.com/gt

BRIEF COMMUNICATION

Implications of cell cycle progression on functional sequence correction by short single-stranded DNA oligonucleotides PA Olsen, M Randol and S Krauss Department for Cellular and Genetic Therapy, Institute for Microbiology, Rikshospitalet, Forskningsparken, Oslo, Norway

Oligonucleotide-based sequence alteration in living cells is a substantial methodological challenge in gene therapy. Here, we demonstrate that using corrective single-stranded oligonucleotides (ssODN), high and reproducible sequence correction rates can be obtained. CHO cell lines with chromosomally integrated multiple copy EGFP reporter genes routinely show rates of 4.5% targeted sequence correction after transfection with ssODN. We demonstrate that the cell cycle influences the rates of targeted sequence correction in vivo, with a peak in the early S phase during ssODN exposure. After cell division, the altered genomic sequence is predominantly passed to one daughter cell, indicating that targeted sequence alteration occurs after the

replication fork has passed over the targeted site. Although high initial correction rates can be obtained by this method, we show that a majority of the corrected cells arrest in the G2/M cell cycle phase, although 1–2% of the corrected cells form viable colonies. The G2/M arrest observed after targeted sequence correction can be partially released by caffeine, pentoxifylline or Go¨6976 exposure. Despite substantial remaining challenges, targeted sequence alteration based on ssODN increasingly promises to become a powerful tool for functional gene alterations. Gene Therapy (2005) 12, 546–551. doi:10.1038/sj.gt.3302437 Published online 27 January 2005

Keywords: in vivo repair; oligonucleotide; cell cycle; G2/M arrest

Various strategies have been developed to achieve therapeutic targeted gene repair, including small fragment homologous sequence1,2 ssODN,3–6 triple helixforming oligonucleotides containing a reactive group,7–9 bifunctional oligonucleotides based on a triple helixforming DNA recognition domain10,11 and branched oligonucleotides.12 Although a general proof of principle for oligonucleotide-based sequence alteration appears to be established, the paradigm for oligonucleotide-based genome alteration remain largely unknown.6 The central problem in studying this phenomenon are low rates of in vivo sequence correction rates in cell culture. Generally, observed sequence alteration rates on in vivo chromosomal templates in selected cell lines are in the range of 0.1%,6 while correction rates in untransformed primary cells are substantially lower.13–15 In the absence of a suitable selection system, such low rates allow only limited functional analysis, although Yoon and co-workers have recently suggested a selection system based on simultaneous targeting of two loci.16 A further problem is the variety of model systems and assays that were used to measure targeted sequence correction which lead to confusing and often conflicting results.6 Despite these obstacles, several consistent observations have been made: targeted sequence correction depends centrally on transcription, and ssODN that are complementary to Correspondence: Professor S Krauss, Institute of Microbiology, Section for Genetic Therapy, Rikshospitalet, Forskningsparken, 0349 Oslo, Norway Received 14 September 2004; accepted 23 November 2004; published online 27 January 2005

the nontranscribed strand show in most cases substantially higher correction rates than ssODN complementary to the transcribed strand in both episomal and chromosomal assays.17,18 The involvement of transcription-coupled repair has therefore been suggested. Furthermore, the absence of MSH2, a protein involved in mismatch repair (MMR), has been shown to increase targeted sequence alteration rates in mouse embryonic stem (ES) cells.13 The reason for this observation remains unclear, but a deficiency in the MMR machinery would increase the tolerance for a mismatch between ssODN and the target sequence, while a functional MMR could remove the mismatched base from the ssODN. Finally, in contrast to procaryotic systems,19 a model based on the integration of a ssODN into the lagging strand during replication could not be established in mammalian cells using episomal assays.20 To achieve high rates of ssODN-based sequence correction that would allow an improved tool for in vivo analysis; two CHO cell lines were created that contained multiple copies of a chromosomally integrated mutated EGFP (mEGFP-a and -b) reporter gene. Such a strategy was essential since the literature shows conflicting results between in vivo assays on chromosomal DNA, in vivo assays on episomal DNA and in vitro experiments carried out on cell extracts.6 As seen in Figure 1, CHOmEGFP-a and -b reporter cells showed significant rates of targeted sequence correction after ssODN exposure. After transfection of the CHO-mEGFP-a or CHOmEGFP-b reporter cells with ssODN (a2a and b2a), 4.5 and 0.35% cells, respectively, could routinely be obtained

Functional sequence correction by ssODN PA Olsen et al

547

Figure 1 Functional correction of mutated EGFP with ssODN. (a) Representative flow cytometry dot plots of CHO-mEGFP-a cells transfected with control ssODN (actr1a, left panel) or corrective ssODN (a2a, right panel) analyzed 24 h after transfection. In the dot plots, corrected cells (green fluorescence) and dead cells (propidium iodide stained, red fluorescence) are plotted against the X- and Y-axis, respectively. The sequences above the dot plot represent the genomic target sequence and the sequence of the ssODN (italic). Mismatches between ssODN and the genomic target are indicated by asterisks. (b) Detection of green cells appearing 24 h after transfection of the CHO-mEGFP-a cell line with ssODN a2a; fluorescence (EGFP, right panel) and phase contrast (PH, left panel) microscopy (scale bar represent 100 mm). (c) Sequencing of stable corrected green CHO-mEGFP-b cells obtained after sequence correction with ssODN b2a. Asterisks indicate positions of changed bases. The position of the targeted stop codon is underlined. ssODN were purchased from MWG Biotech and Eurogentec. All ssODN contained three phosphorothioate groups at the 30 and 50 ends. The reporter cell lines CHOmEGFP-a and CHO-mEGFP-b were prepared by stably transfecting CHO-k1 cells (ATCC CCL-61) with pmEGFP-a and pmEGFP-b, respectively. pmEGFP-a contains a Tyr 66 (TAC) to Ser 66 (TCC) missense mutation in the pEGFP-c1 plasmid (Clontech, BD Biosciences) that renders the EGFP protein nonfluorescent, and a silent Lys 45 (AAG to AAA) diagnostic mutation. pmEGFP-b was prepared by inserting the EGFP-coding sequence from pEGFP-c1 into the pNRTIS-21 vector.28 The pmEGFP-b construct contains a premature stop codon at Lys 41 (AAG) to stop 41 (TAG) that renders the EGFP protein nonfluorescent and the following silent point mutations: Gly 31 (GGC to GGA), Gly 33 (GGC to GGA), Gly 35 (GGC to GGA) and Gly 40 (GGC to GGA). The pmEGFP-a and pmEGFP-b constructs were stably transfected into CHO cells and individual clones were isolated after G418 selection (0.8 mg/ml). Clonal reporter cell lines were functionally selected by transfection with corrective ssODN (a2a and b2a) that contained a wild-type EGFP sequence. Quantitative Southern blot analysis established that CHO-mEGFP-a contained 60 copies of the reporter, while CHO-mEGFP-b contained 12 copies. For targeted sequence correction, cells were transfected in six-well plates (150 000 cells/well the day before transfection) with 4 mg ssODN and 8 ml Lipofectamine 2000 (Invitrogen)/well according to the manufacturer’s instructions. Cells were incubated for 1 h with the transfection mix, followed by 24 h incubation without transfection mix before analysis by flow cytometry (100 000 cells). For sequencing, the EGFP region was PCR amplified from genomic DNA from sorted (FACS DiVa cell sorter, BD Biosciences) converted CHO-mEGFP-b cells and cloned into a sequencing vector by the TOPO TA cloning kit (Invitrogen). Quantification of green converted cells and cell cycle analysis were performed on a Parterc PAS flow cytometer (Partec GmbH) equipped with a 488 nm argon laser and a HBO lamp. Image acquisition was carried out with a Zeiss Axiovert 200M microscopy equipped with a Zeiss Axiocam HR color camera.

(n ¼ 30) that produced functional EGPF (Figures 1a, b and 2a, b). A flow cytometry analysis of CHO-mEGFP-a cells treated with a corrective ssODN (a2a, complementary to the nontranscribed strand) showed a distinct population of strongly green fluorescent cells (Figure 1a, right panel). ssODN allowed the introduction of sitespecific sequence alterations in the mEGFP target with high specificity. Control ssODN (actr1a) that did not contain a corrective mismatch showed no measurable background activity (Figure 1a, left panel). Furthermore, sequencing of green cells derived from the CHOmEGFP-b cell line after transfection with ssODN revealed that in all cases the observed changes in the

chromosomal template after ssODN exposure corresponded exactly to the sequence of the ssODN. Thus, using ssODN b2a that forms two mismatches when hybridized to the mEGFP-b reporter, both mismatched bases are altered in the chromosomal template. In no cases only one base, or alternative collateral bases changed (Figure 1c). Interestingly, it was observed that sequence correction by ssODN tolerates well more than one mismatch with the chromosomal target. The ssODN that contained two adjacent corrective mismatches (ssODN a2a and b2a; Figure 2a, b) led consistently to higher sequence correction rates when compared to ssODN containing only one mismatch (ssODN a1a, b1a; Gene Therapy


548

Figure 2 Correlation between targeted sequence correction and number of corrective mismatches. (a) Sequence correction rates of ssODN complementary to the nontranscribed strand in reporter cell line CHO-mEGFP-a. (b) Sequence correction rates of ssODN complementary to the nontranscribed strand in reporter cell line CHO-mEGFP-b. Error bars represent the standard deviation (s.d.) between three independent parallels. Asterisk indicates statistical significant difference relative to the a1a or b1a ssODN (Po0.03, paired Student’s t-test). Dots in the schematic drawings of the ssODN below the bars indicate the number and position of corrective mismatches. The sequence of the corrective ssODN were: actr1a 30 -tgg gac tgg agg ccg cac gtc acg aag-50 ; a1a 30 -tgg gac tgg atg ccg cac gtc acg aag50 ; a2a 30 -tgg gac tgg ata ccg cac gtc acg aag-50 ; a3a 30 -tgg gac tgc atg ccg cac gtc acg aag-50 ; a4a 30 -tgg gac tgc ata ccg cac gtc acg aag-50 ; bctr1a 30 tgg atg cct atc gac tgg gac ttc aag t-50 ; b1a 30 -tgg atg cct ttc gac tgg gac ttc aag t-50 ; b2a 30 -tgg atg ccg ttc gac tgg gac ttc aag t-50 ; b3a 30 -tgg atg ccg tgc gac tgg gac ttc aag t-50 ; b4a 30 -tgg atg ccg tgg gac tgg gac ttc aag t-50 (underlined bases indicates positions of mismatches).

Figure 2a, b), while more than three mismatches resulted in a reduction of the sequence correction frequencies (Figure 2b). CHO-mEGFP-a and CHO-mEGFP-b cell lines transfected with ssODN showed sequence correction only when they were transfected while they were dividing (Figure 3a, left panel). If the cells were confluent at the time of transfection, no sequence correction could be detected (not shown). Therefore, the correlation between ssODN-based sequence correction and cell cycle stages was tested. CHO cells can efficiently be arrested in the G1/S phase by mimosine, a drug that reversibly inhibits both replication initiation and strand elongation.21 After removing the drug, cells progress synchronically into the S and G2/M phase.22 As seen in Figure 3a (right panel), cells kept in the G1/S phase by continuous exposure to Gene Therapy

mimosine before and after transfection with ssODN displayed no sequence correction. If cells were released from the G1/S arrest and transfected with ssODN (b2a) in intervals, sequence correction rates changed depending on the stage of the cell cycle. Correction frequencies increased 10-fold if the cells were transfected immediately after release from the G1/S arrest (Figure 3b, T0). When the cells were transfected after they had progressed to the late S phase and G2/M phase, correction frequencies dropped to 0.57 and 0.32% (Figure 3b, T8 and T12, respectively). Cells arrested in the G2/M phase, by the microtubule-depolymerizing agent nocodazole, displayed correction rates similar to cells that had progressed to the S or G2/M phase (Figure 3b, right panel). Control experiments with fluorescently labeled ssODN showed that uptake of ssODN did not alter significantly throughout cell cycle progression (not shown). This demonstrates that the observed alterations are not due to differential uptake of the ssODN but rather reflect cell cycle-specific changes in the capability of undergoing sequence alteration. The preferred stage for targeted sequence correction is thus cells in the S phase, but correction can also take place in the G2/M phase. No measurable correction was taking place when cells were arrested in the G1 phase. Ferrara and Kmiec23 have recently reported increased ssODN-mediated sequence correction rates in the late S phase/early G2 phase. In their system, DLD-1 cells were pretreated with camptothecin, a drug that induces DNA strand breaks with a subsequent shift of the cells towards the late S/early G2 cell cycle phase before ssODN transfection. To further confirm the importance of the cell cycle in targeted sequence correction, the fate of corrected cells was analyzed. If targeted sequence correction can occur in the G1 phase, one would expect that the alteration would be passed on to both daughter cells after replication and cell division. However, if the repair process occurred during or after replication, it would be expected that only one daughter cell would contain the sequence-corrected template. We therefore followed the intensity of EGFP fluorescence in corrected cells after cell division. Figure 3c and d shows that from 80% of the corrected cells only one daughter cell maintained strong EGFP staining, while the other daughter cell displayed decreased fluorescence intensity. In CHO cells stably expressing wild-type EGFP, the fluorescence intensity did not differ between daughter cells after cell division (Figure 3d, filled bars). The reduction of fluorescence in one of the daughter cells again supports a scenario where targeted sequence alteration occurs predominantly in the S or G2/M phase. Stable colonies harboring the nucleotide change introduced by ssODN could be routinely established. Nevertheless, the number of colonies that emerged did not correspond to the initial number of green corrected cells. Over a period of 4 days, the number of green cells dropped by 40–60% every 24 h, independent on the analyzed reporter cell line (Figure 4a). From the total amount of green cells detected 24 h after ssODN transfection, only 1–2% was capable of forming viable colonies. The majority of the initially green cells did not divide, but started to round up and eventually detached from the surface, although they did not loose their green fluorescence (Figure 4b). To determine at which stage of the cell cycle the majority of converted cells remained,


549

Figure 3 Gene correction and cell cycle. (a) Cell cycle distribution and correction rates in exponentially growing (left panel) and mimosine G1/S cell cyclearrested (right panel) CHO-mEGFP-b cells. The bars next to the DNA histograms represent gene correction frequencies obtained from CHO-mEGFP-b cells transfected with ssODN b2a when the cells were at the indicated cell cycle phases. (b) Mimosine-arrested CHO-mEGFP-b cells that were released from the mimosine block and transfected with b2a ssODN at 0 h (T0), 8 h (T8) and 12 h (T12) after mimosine removal (left panels). CHO-mEGFP-b cells arrested in the G2/M phase by nocodazole and transfected with the corrective b2a ssODN (right panel). (c) Representative fluorescent images of caffeinetreated dividing CHO-mEGFP-a cells 24 h after transfection with ssODN a2a. Note the difference in fluorescence intensity between the two daughter cells (scale bar represent 100 mm). (d) Quantification of fluorescence intensity in dividing pairs of stable CHO-EGFP wild-type (wt) cells (closed bars) and sequence-corrected CHO-mEGFP-a (open bars) cells. The fluorescence intensity of each cell in 34 pairs of dividing cells from each cell line was measured and compared (MCID analysis software, Imaging Research Inc.). Error bars represent s.d. between the fluorescence intensity ratios of the cell pairs in the respective group. A ratio of 1.0 means no difference in fluorescence intensity of the cells in a dividing pair (as observed in dividing wt cells). Cells were arrested at the G1/S border by the procedure described.21 Briefly, the cells were grown for 48 h in media containing 0.2% FBS, then incubated for 24 h in growth media containing 5% FBS and 100 mM mimosine (Sigma). The cells were transfected (1 h) in media containing mimosine. After one wash with PBS, the cells were either kept in the G1 phase by continued incubation with mimosine or allowed to synchronically progress into the cell cycle by removal of mimosine. For cell cycle analysis, the cells were ethanol fixed, DAPI stained and analyzed for DNA content by flow cytometry. Cells were arrested at the G2/M phase by overnight incubation with nocodazole (0.5 mg/ml) (Sigma-Aldrich). For caffeine (Sigma-Aldrich) treatment, the cells were incubated in 2 mM caffeine 2 h before and 3 h after ssODN transfection.

the cell cycle distribution was analyzed by flow cytometry. When CHO-mEGFP-a cells were exposed to a2a ssODN (490% transfection efficiency as determined by fluorescently labeled control ssODN), transfected cells displayed a normal cell cycle distribution (Figure 4c, left panel), indicating that the presence of ssODN per se did not have any measurable affects on cell cycle distribution. In contrast, when green converted cells were

analyzed separately, a clear accumulation of these cells in the G2/M phase was apparent (Figure 4c, right panel). This indicates that ssODN, upon altering a chromosomal sequence, are likely to trigger damage response pathways that the majority of cells are unable to overcome. The G2/M arrest is not absolute, and consistently a few cells were capable of forming viable colonies with stable sequence alteration, indicating that a small number of Gene Therapy


550

Figure 4 Sequence correction and G2/M cell cycle arrest. (a) CHOmEGFP-a cells were transfected with a2a ssODN and the percentage of green cells was quantified after 24, 48, 72 and 96 h. Error bars indicate s.d. between three independent parallels. (b) Sequential fluorescent images of a representative corrected CHO-mEGFP-b cell after transfection (30, 44, 72, 96, 114 and 120 h) with ssODN b2a without caffeine treatment (scale bar represent 50 mm). (c) Cell cycle distribution of uncorrected and corrected CHO-mEGFP-b cells 24 h after transfection with ssODN b2a. The cell cycle distribution was visualized by flow cytometry after incubation of the cells with the live cell penetrating dye Hoechst 33342 (Molecular Probes). Cell cycle distribution of all the transfected cells (whole population, left panel) and only the corrected cells (green population, right panel) are shown. A clear accumulation of cells in the G2/M phase is evident in the corrected cells.

cells either escapes from the arrest, or does not enter the G2/M arrest after targeted sequence alteration. However, most corrected cells remain in the G2/M arrest. It is known that cells can respond to DNA damages that arise during or after replication by the activation of the ATM/ ATR kinase-dependent G2/M checkpoint.24 We therefore tested the impact of caffeine and pentoxifylline that target ATM/ATR, and of Go¨6976 that inhibits with high specificity the ATM/ATR downstream mediators Chk1 and/or Chk 2.25 Compared to caffeine and pentoxyfylline, Go¨6976 has a greater selectivity in abrogating DNA damage-induced cell cycle arrest.25 If an ATM/ATR-dependent G2/M arrest was responsible for the failure of most green corrected cells to progress through mitosis, exposure of the cells to either caffeine, pentoxifylline or Go¨6976 would be expected to release cells from a G2/M block.25,26 As seen in Figure 5a and b, when CHO-mEGFP-b cells were transfected with ssODN b2a, 5% of the green corrected cells had undergone cell division 24 h after transfection. However, if the Gene Therapy

Figure 5 Release from the induced G2/M cell cycle arrest of corrected cells by caffeine, pentoxifylline and Go¨6976. (a) Quantification of percentage of dividing CHO-mEGFP-b cells after transfection with b2a ssODN. Quantification was performed on untreated cells and on cells incubated with caffeine, pentoxifylline or Go¨6976. Results from a representative experiment where at least 500 cells/cell pairs were counted for each conditionare shown. (b) Fluorescent images of CHO-mEGFP-b cells 24 h after transfection with ssODN a2a. The images are of untreated cells and cells treated with caffeine, pentoxifylline and Go¨6976, dividing cell pairs are indicated with an arrowhead (scale bar represent 100 mm). For drug treatment, the cells were incubated with 3 mM caffeine, 3 mM pentoxifylline (Sigma-Aldrich) or 150 nM Go¨6976 (Calbiochem) for 3 h before ssODN transfection followed by continued exposure to the drugs over night. The presence of the drugs during transfection did not influence on transfection efficiency as determent by uptake of fluorescently labeled control ssODN (not shown).

cells received an overnight exposure of 3 mM caffeine, or 3 mM pentoxifylline, 15 and 12% of the corrected cells, respectively, entered visible division (Figure 5a, b). The strongest effect on the G2/M arrest was seen after treatment with 150 nM of the Chk1 inhibitor Go¨6976. When cells were treated overnight with Go¨6976, 30% fluorescent dividing cells were observed (Figure 5a, b). Treatment of cells with caffeine, pentoxifylline or Go¨6976 did not give rise to increased number of corrected cells forming viable colonies (not shown). This was not surprising since it is known that abrogation of a G2 cell cycle arrest (preferentially in p53-defective cells) by these drugs drives the cells through a lethal mitosis.27 The effective G2/M block and low rate of cells that processed into normal cell division after ssODN


transfection indicates the presence of effective defense mechanisms against ssODN-based genome alterations in CHO cells, and possibly in other cell types. A substantial challenge remains to understand in more detail the mechanisms that enable, or counteract targeted sequence alteration. Improvements may be achieved by either altering the design of ssODN, by stimulating the pathways that allow ssODN-based targeted sequence alteration, or by interfering with the cellular defense mechanisms that can inhibit cell cycle progression of corrected cells. The elucidation of the functional basis for targeted sequence correction that is now underway starts to provide a basis for systematic improvements of the technology.

Acknowledgements This work has been funded by the EU Grant QLK3-CT2000-00634, and by a grant of the research council of Norway (NFR advanced research program). We thank James Booth and Solene Geraudie for reading through the manuscript and helpful discussions. Finally, we specially thank Dr Rolf Seljelid for valuable support.

References 1 Kunzelmann K et al. Gene targeting of CFTR DNA in CF epithelial cells. Gene Therapy 1996; 3: 859–867. 2 Sangiuolo F et al. In vitro correction of cystic fibrosis epithelial cell lines by small fragment homologous replacement (SFHR) technique. BMC Med Genet 2002; 3: 8. 3 Campbell CR et al. Homologous recombination involving small single-stranded oligonucleotides in human cells. New Biol 1989; 1: 223–227. 4 Igoucheva O, Alexeev V, Yoon K. Targeted gene correction by small single-stranded oligonucleotides in mammalian cells. Gene Therapy 2001; 8: 391–399. 5 Moerschell RP, Tsunasawa S, Sherman F. Transformation of yeast with synthetic oligonucleotides. Proc Natl Acad Sci USA 1988; 85: 524–528. 6 Igoucheva O, Alexeev V, Yoon K. Oligonucleotide-directed mutagenesis and targeted gene correction: a mechanistic point of view. Curr Mol Med 2004; 4: 445–463. 7 Faruqi AF et al. Recombination induced by triple-helix-targeted DNA damage in mammalian cells. Mol Cell Biol 1996; 16: 6820–6828. 8 Nagatsugi F, Sasaki S, Miller PS, Seidman MM. Site-specific mutagenesis by triple helix-forming oligonucleotides containing a reactive nucleoside analog. Nucleic Acids Res 2003; 31: e31. 9 Seidman MM, Glazer PM. The potential for gene repair via triple helix formation. J Clin Invest 2003; 112: 487–494.

551 10 Chan PP et al. Targeted correction of an episomal gene in mammalian cells by a short DNA fragment tethered to a triplex-forming oligonucleotide. J Biol Chem 1999; 274: 11541–11548. 11 Datta HJ et al. Triplex-induced recombination in human cell-free extracts. Dependence on XPA and HsRad51. J Biol Chem 2001; 276: 18018–18023. 12 Olsen PA, McKeen C, Krauss S. Branched oligonucleotides induce in vivo gene conversion of a mutated EGFP reporter. Gene Therapy 2003; 10: 1830–1840. 13 Dekker M, Brouwers C, Te RH. Targeted gene modification in mismatch-repair-deficient embryonic stem cells by singlestranded DNA oligonucleotides. Nucleic Acids Res 2003; 31: e27. 14 Nickerson HD, Colledge WH. A comparison of gene repair strategies in cell culture using a lacZ reporter system. Gene Therapy 2003; 10: 1584–1591. 15 Nickerson HD, Colledge WH. A LacZ-based transgenic mouse for detection of somatic gene repair events in vivo. Gene Therapy 2004; 11: 1351–1357. 16 Alexeev V, Igoucheva O, Yoon K. Simultaneous targeted alteration of the tyrosinase and c-kit genes by single-stranded oligonucleotides. Gene Therapy 2002; 9: 1667–1675. 17 Igoucheva O, Alexeev V, Pryce M, Yoon K. Transcription affects formation and processing of intermediates in oligonucleotidemediated gene alteration. Nucleic Acids Res 2003; 31: 2659–2670. 18 Liu L et al. Strand bias in targeted gene repair is influenced by transcriptional activity. Mol Cell Biol 2002; 22: 3852–3863. 19 Li XT et al. Identification of factors influencing strand bias in oligonucleotide-mediated recombination in Escherichia coli. Nucleic Acids Res 2003; 31: 6674–6687. 20 Brachman EE, Kmiec EB. DNA replication and transcription direct a DNA strand bias in the process of targeted gene repair in mammalian cells. J Cell Sci 2004; 117: 3867–3874. 21 Kalejta RF, Hamlin JL. The dual effect of mimosine on DNA replication. Exp Cell Res 1997; 231: 173–183. 22 Orren DK, Petersen LN, Bohr VA. A UV-responsive G2 checkpoint in rodent cells. Mol Cell Biol 1995; 15: 3722–3730. 23 Ferrara L, Kmiec EB. Camptothecin enhances the frequency of oligonucleotide-directed gene repair in mammalian cells by inducing DNA damage and activating homologous recombination. Nucleic Acids Res 2004; 32: 5239–5248. 24 Abraham RT. Cell cycle checkpoint signaling through the ATM and ATR kinases. Genes Dev 2001; 15: 2177–2196. 25 Kawabe T. G2 checkpoint abrogators as anticancer drugs. Mol Cancer Ther 2004; 3: 513–519. 26 Anderson HJ, Andersen RJ, Roberge M. Inhibitors of the G2 DNA damage checkpoint and their potential for cancer therapy. Prog Cell Cycle Res 2003; 5: 423–430. 27 Eastman A. Cell cycle checkpoints and their impact on anticancer therapeutic strategies. J Cell Biochem 2004; 91: 223–231. 28 Tenev T et al. Perinuclear localization of the protein-tyrosine phosphatase SHP-1 and inhibition of epidermal growth factorstimulated STAT1/3 activation in A431 cells. Eur J Cell Biol 2000; 79: 261–271.

Gene Therapy