Bioorganic & Medicinal Chemistry Letters 24 (2014) 813–816

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl

Sandwiched zinc-finger nucleases demonstrating higher homologous recombination rates than conventional zinc-finger nucleases in mammalian cells Tomoaki Mori, Koichi Mori, Takamasa Tobimatsu, Takashi Sera ⇑ Department of Applied Chemistry and Biotechnology, Graduate School of Natural Science and Technology, Okayama University, Tsushima-Naka, Kita-ku, Okayama 700-8530, Japan

a r t i c l e

i n f o

Article history: Received 2 August 2013 Revised 19 December 2013 Accepted 23 December 2013 Available online 3 January 2014 Keywords: Sandwiched zinc finger nuclease Single-chain FokI dimer Zinc finger protein Homologous recombination Genome editing

a b s t r a c t We previously reported that our sandwiched zinc-finger nucleases (ZFNs), in which a DNA cleavage domain is inserted between two artificial zinc-finger proteins, cleave their target DNA much more efficiently than conventional ZFNs in vitro. In the present study, we compared DNA cleaving efficiencies of a sandwiched ZFN with those of its corresponding conventional ZFN in mammalian cells. Using a plasmid-based single-strand annealing reporter assay in HEK293 cells, we confirmed that the sandwiched ZFN induced homologous recombination more efficiently than the conventional ZFN; reporter activation by the sandwiched ZFN was more than eight times that of the conventional one. Western blot analysis showed that the sandwiched ZFN was expressed less frequently than the conventional ZFN, indicating that the greater DNA-cleaving activity of the sandwiched ZFN was not due to higher expression of the sandwiched ZFN. Furthermore, an MTT assay demonstrated that the sandwiched ZFN did not have any significant cytotoxicity under the DNA-cleavage conditions. Thus, because our sandwiched ZFN cleaved more efficiently than its corresponding conventional ZFN in HEK293 cells as well as in vitro, sandwiched ZFNs are expected to serve as an effective molecular tool for genome editing in living cells. Ó 2013 Elsevier Ltd. All rights reserved.

Artificial endonucleases have been developed by protein-engineering techniques including fusion of a DNA-binding protein to a DNA-cleaving enzyme and alteration of target specificities of existing restriction enzymes.1–4 Among them, zinc-finger nucleases (ZFNs) have been demonstrated to efficiently cleave their target DNA in vitro and in vivo, allowing manipulation of genetic information.5,6 One concern about in vivo applications is that off-target cleavage may cause cytotoxicity because ZFNs can form a homodimer as well as a heterodimer via their FokI catalytic domains. To reduce off-target cleavage, two groups have engineered a FokI-dimer interface so that the engineered ZFN preferentially formed a heterodimer.7,8 However, because conventional ZFNs bind to both dsDNA substrates and cleavage products with equal affinities, they do not cleave target DNA with multiple turnovers differently from native restriction endonucleases. To solve this problem, we previously developed sandwiched ZFNs9–11 as novel ZFNs. The sandwiched ZFNs uniquely harbor a single-chain FokI dimer (designated scFokI12) sandwiched between two artificial zinc-finger proteins (AZPs13). Due to the unique molecular structure, our sandwiched ZFNs cleave target DNA with multiple turnovers like native restriction endonucleases. Because sandwiched ZFNs cleave the regions between two AZP-binding ⇑ Corresponding author. Tel./fax: +81 86 251 8194. E-mail address: [email protected] (T. Sera). 0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bmcl.2013.12.096

sites, dissociation constants of sandwiched ZFNs for DNA cleavage products are significantly increased. Therefore, sandwiched ZFNs are able to revisit and cleave their DNA substrates again, leading to multiple-turnover DNA cleavage (see Ref. 9 for more detailed explanation). Actually, sandwiched ZFNs demonstrated higher DNA-cleavage rates than conventional ZFNs in vitro.9–11 For example, a sandwiched ZFN cleaved a 100-fold excess of its target plasmid completely. However, under the same condition, multipleturnover cleavage by the corresponding conventional ZFN was not observed; only very faint DNA bands corresponding to a cleavage product were detected in the reactions with the conventional ZFN.10 In the present study, we examined whether our sandwiched ZFN cleaved target DNA more efficiently than its corresponding conventional ZFN in mammalian cells as well as in vitro. To this end, we used a plasmid-based single-strand annealing (SSA) reporter assay in HEK293 cells. Briefly, a gene encoding luciferase was divided into two segments, separated by a stop codon and a ZFN target site (Fig. 1A). Both segments contained a homologous 881-bp region in the direct repeat orientation. A ZFN-induced double-strand break between the segments allows efficient homologous recombination via SSA, resulting in the reconstitution of an active luciferase gene. The luciferase activity measured by an SSA reporter assay is therefore proportional to DNA-cleavage activity of a ZFN. In both the sandwiched and conventional ZFNs, two het-

814

T. Mori et al. / Bioorg. Med. Chem. Lett. 24 (2014) 813–816

A

(GGGGS)3

sandwiched ZFN

pSSA-1:

GGTCGGGACC ATATGT GTTGCGGGAT CCAGCCCTGG TATACA CAAGGCCCTA

conventional ZFN pair

pSSA-2:

GGTCCCGACC ATATGT GTTGCGGGAT CCAGGGCTGG TATACA CAAGGCCCTA

PCMV

LH RH inactive luciferase

DSB PCMV

SSA PCMV

PCMV active luciferase

Relative lucifease activity (fold)

B 10 8 6 4 2 0

0.12 0.06 0.03

0

0.12 0.06 0.03

0

sandwiched (fmol)

conventional (fmol)

0.12 0.06 0.03

0.12 0.06 0.03

Relative lucifease activity (fold)

C 10 8 6 4

different DNA strand. Accordingly, the pSSA-1 reporter plasmid was used for the sandwiched ZFN and the pSSA-2 one for the conventional ZFN, as shown in Figure 1A. HEK293 cells were co-transfected with a ZFN expression plasmid and its SSA reporter plasmid. Luciferase activity was measured at three and six days post-transfection. As shown in Figure 1B, three days after transfection, the sandwiched ZFN activated a luciferase gene much more effectively than the conventional ZFN at all equimolar amounts of transfected ZFN-expression plasmids. For example, in the transfection with 0.12 fmol of a ZFN-expression plasmid, the sandwiched ZFN activated a luciferase gene 4.5 times, while the conventional ZFN activated a luciferase gene only 1.1 times. Furthermore, six days after transfection, the sandwiched ZFN showed greater DNA-cleavage rates than that of the conventional ZFN in HEK293 cells as well (Fig. 1C). The sandwiched ZFN enhanced luciferase gene expression 8.8 times when 0.12 fmol of the expression plasmid was transfected. In contrast, the conventional ZFN enhanced the gene expression only 1.2 times under the same conditions. To further confirm that sandwiched ZFNs have greater abilities of homologous recombination than conventional ZFNs, we additionally constructed two different sandwiched ZFNs (designated ‘sandwiched ZFN(2)’ and ‘sandwiched ZFN(3)’), which targeted 50 -GGGGAGCAGGATATGTTAGGGAGCCC-30 and 50 GGAGAAGGACATATGTGTGGGCTTGT-30 harboring unique AZPbinding sites (underlined), respectively, and their corresponding conventional ZFNs (see Supplementary Fig. S1). We then performed SSA reporter assays as described above. As shown in Supplementary Figure S1, these two sandwiched ZFNs demonstrated higher recombination rates than conventional ZFNs as well. Western blot analysis of sandwiched and conventional ZFNs revealed that the greater DNA cleavage by the sandwiched ZFN was not caused by higher expression of the sandwiched ZFN in HEK293 cells. Because conventional ZFNs must form a dimer for DNA cleavage at one site, the ratio of signal intensity of the sandwiched ZFN observed in Western blot analysis to that of the conventional ZFN is 1 to 2 when an equal amount of sandwiched ZFN is expressed. As shown in Figure 2, the band intensity of sandwiched ZFN is lower than half of that of the conventional ZFN at both three and six days after transfection. Namely, the expression level of the sandwiched ZFN was lower than that of the conventional ZFN, indicating that the greater DNA-cleaving activity of the sandwiched ZFN was not due to higher expression of the sandwiched ZFN. We also obtained the same or similar results regarding sandwiched ZFN(2) and sandwiched ZFN(3) (see Supplementary Fig. S2).

2 0

1 0

sandwiched (fmol)

0

conventional (fmol)

Figure 1. DNA cleavage rates of sandwiched and conventional ZFNs in HEK293 cells by SSA assay. (A) Plasmid-based SSA assay. Plasmids used for the SSA assays and SSA reaction mechanism are illustrated. Single-strand resection of the DNA ends allows the left (LH) and right (RH) homology regions of the split luciferase gene to anneal and recombine to form an active luciferase gene. The sense strands of DNA sequences recognized by each ZFN are shown in color. Relative luciferase activities three (B) or six (C) days after transfection of sandwiched or conventional ZFNexpression plasmid. Amounts of sandwiched or conventional ZFN-expression plasmid used are indicated below each lane. The data represent an average of three independent experiments, and the SD is shown.

erogeneous AZPs, which recognize two 10-bp targets of 50 GGTCGGGACC-30 and 50 -GTTGCGGGAT-30 , respectively, were used as the DNA-binding domains. However, the target sites of these ZFNs are different (Fig. 1A): two AZP-binding sites in the target of the sandwiched ZFN reside in the same DNA strand, but each AZP-binding site in that of the conventional ZFN resides in a

75 kDa 55 kDa anti-T7 35 kDa

2

3

4

5

6

*

sandwiched ZFN

conventional ZFN pair

anti-GAPDH

Figure 2. Immunoblots of ZFN derivatives in cell lysates. The molecular weights of sandwiched and conventional ZFNs are 75.8 and 39.2 kDa, respectively. Lane 1, sample from cells (3 days post-transfection) transfected with pcDNA3.1 as a control; lane 2, sample from cells (3 days post-transfection) transfected with sandwiched ZFN; lane 3, sample from cells (3 days post-transfection) transfected with conventional ZFN; lane 4, sample from cells (6 days post-transfection) transfected with pcDNA3.1 as a control; lane 5, sample from cells (6 days posttransfection) transfected with sandwiched ZFN; lane 6, sample from cells (6 days post-transfection) transfected with conventional ZFN. The antibody used for each panel is indicated on the left of each panel. The additional bands indicated with an asterisk in the upper panel are artifacts caused by cross-reactivity of an anti-T7 antibody.

T. Mori et al. / Bioorg. Med. Chem. Lett. 24 (2014) 813–816

One potential drawback of sandwiched ZFNs is that their greater DNA-cleavage rates may cause cytotoxicity. Therefore, we investigated the cytotoxicity of the sandwiched ZFN in a modified MTT assay with a highly water-soluble disulfonated tetrazolium salt,14 which is commercially available as a Cell Counting Kit-8 from Dojindo. In this assay, the cytotoxicity of a drug or molecule of interest is based on absorbance at 450 nm of the corresponding water-soluble formazan dye produced by cellular dehydrogenases. As shown in Figure 3A, the cell viability at three days post-transfection with the sandwiched ZFN-expression plasmid (0.12 fmol), which enhanced luciferase gene expression 4.5 times, was 100% ± 4%. Moreover, the cell viability at six days after transfection with the expression plasmid (0.12 fmol), in which 8.8 times activation of a luciferase gene was observed, was 100% ± 4% (Fig. 3B). Thus, under the experimental conditions, the sandwiched ZFN did not significantly reduces cell viability. In this study, we demonstrated that our sandwiched ZFN cleaved its target plasmid and induced homologous recombination more efficiently than its corresponding conventional ZFN in HEK293 cells. Therefore, the sandwiched ZFN might have more chances to cleave non-specific genomic sites, leading to severe cytotoxicity. However, we did not observe any significant cytotoxicity due to the sandwiched ZFN in a modified MTT assay. Although we are interested in investigating the genome-editing properties of sandwiched ZFNs as the next step, we have to examine the off-target effects carefully. The scFokI moiety12 used as the DNA-cleavage domain can be sandwiched with any DNA-binding domain including transcription

A Cell viability (%)

100 80 60 40 20 0

0.12 0.06 0.03

0

sandwiched (fmol)

0.12 0.06 0.03

0

conventional (fmol)

B Cell viability (%)

100 80 60 40 20 0

815

activator-like effectors15,16 (TALEs). Although the DNA-binding affinities of TALEs seem to be generally lower than those of AZPs,17,18 TALEs are able to bind to less G-rich targets,15–18 which are not generally good targets for AZPs. TALE nucleases (TALENs) have also demonstrated effective genome editing (reviewed in Ref. 19). Therefore, fusion of the scFokI cleavage domain to TALEs (i.e., sandwiched TALENs) may be useful for targeting more diverse DNA sequences. Another potential advantage of sandwiched ZFNs (or sandwiched TALENs) may be the more efficient biallelic disruption of endogenous target genes due to the greater DNA-cleavage rates. In previous studies, conventional ZFNs caused biallelic disruption at approximately one-eighteenth to one-third of monoallelic disruption rates.6,20 Greater DNA-cleavage rates of sandwiched ZFNs may increase the rates of biallelic disruption or modification compared with conventional ZFNs. Recently, the clustered regulatory interspaced short palindromic repeat (CRISPR)/Cas system, which was discovered as an adaptive immune system against invading genetic elements such as viruses and plasmids in bacteria and archaea, has emerged as a new RNA-guided DNA endonuclease (reviewed in Ref. 21). The system requires only short CRISPR RNA guided to a DNA target and the Cas9 protein to efficiently cleave a target DNA, and it demonstrates not only higher genome-editing rates22,23 but also more efficient biallelic disruption24 due to greater DNA-cleavage rates than conventional ZFNs and TALE nucleases. Therefore, we are also interested in comparing the DNA-cleavage rates and genome editing efficiencies of our sandwiched ZFNs with those of CRISPR-Cas systems. In the present study, we used a flexible (GGGGS)3 linker2 between AZP and FokI moiety because our sandwiched ZFNs harboring the 15-amino acid (aa) peptide linker previously demonstrated enough high DNA-cleaving rates in vitro.10,11 Actually, the length of the peptide linker influences cleaving rates of ZFN derivatives.25,26 For example, 4- and 8-aa linkers induced the highest recombination rates among 0–20-aa linkers at 6-bp spacers between two AZP-binding sites.26 However, ZFNs harboring longer peptide linkers (e.g., 16- or 18-aa linker) were shown to be still active in mammalian cells.25,26 Therefore, low homologous recombination rates of conventional ZFNs in the present study seem to be due to mainly their low DNA-cleaving rates caused by transfection of relatively small amounts of expression plasmids, but not to the linker length. Transfection of a higher amount (0.24 fmol) of conventional ZFNexpression plasmids increased recombination rates from 1.1 times (transfection of the 0.12 fmol plasmid, Fig. 1B) to 1.7 times at three days post-transfection (data not shown). We note again that sandwiched ZFNs used in the present study harbor the same 15-amino acid peptide linker. In summary, our sandwiched ZFN demonstrated higher DNAcleavage rates than those of a corresponding conventional ZFN in living mammalian cells without cytotoxicity. Thus, sandwiched ZFNs are expected to serve as an effective molecular tool for genome editing in living mammalian cells. Supplementary data

0.12 0.06 0.03

0

sandwiched (fmol)

0.12 0.06 0.03

0

conventional (fmol)

Figure 3. Cytotoxicity of sandwiched and conventional ZFNs. Three (A) or six (B) days after transfection, the amounts of a water-soluble formazan dye generated by cellular enzymes were measured as absorbance at 450 nm (OD450), and then cell viabilities (%) were calculated from the ratio of OD450 in the presence of ZFN to OD450 in the absence of ZFN as the control (100%). Amounts of sandwiched or conventional ZFN-expression plasmid used are indicated below each lane. The data represent an average of four independent experiments, and the SD is shown.

Supplementary data (materials and methods; Fig. S1; Fig. S2) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.12.096. References and notes 1. 2. 3. 4.

Dehua, P.; Schultz, P. G. J. Am. Chem. Soc. 1990, 112, 4579. Kim, Y.-G.; Cha, J.; Chandrasegaran, S. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 1156. Wu, J.; Kandavelou, K.; Chandrasegaran, S. Cell. Mol. Life Sci. 2007, 64, 2933. Pâques, F.; Duchateau, P. Curr. Gene Ther. 2007, 7, 49.

816

T. Mori et al. / Bioorg. Med. Chem. Lett. 24 (2014) 813–816

5. Urnov, F. D.; Miller, J. C.; Lee, Y.-L.; Beausejour, C. M.; Rock, J. M.; Augustus, S.; Jamieson, A. C.; Porteus, M. H.; Gregory, P. D.; Holmes, M. C. Nature 2005, 435, 646. 6. Santiago, Y.; Chan, E.; Liu, P.-Q.; Orlando, S.; Zhang, L.; Urnov, F. D.; Holmes, M. C.; Guschin, D.; Waite, A.; Miller, J. C.; Rebar, E. J.; Gregory, P. D.; Klug, A.; Collingwood, T. N. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 5809. 7. Miller, J. C.; Holmes, M. C.; Wang, J.; Guschin, D. Y.; Lee, Y.-L.; Rupniewski, I.; Beausejour, C. M.; Waite, A. J.; Wang, N. S.; Kim, K. A.; Gregory, P. D.; Pabo, C. O.; Rebar, E. J. Nat. Biotechnol. 2007, 25, 778. 8. Szczepek, M.; Brondani, V.; Büchel, J.; Serrano, L.; Segal, D. J.; Cathomen, T. Nat. Biotechnol. 2007, 25, 786. 9. Mineta, Y.; Okamoto, T.; Takenaka, K.; Doi, N.; Aoyama, Y.; Sera, T. Biochemistry 2008, 47, 12257. 10. Mori, T.; Kagatsume, I.; Shinomiya, K.; Aoyama, Y.; Sera, T. Biochem. Biophys. Res. Commun. 2009, 390, 694. 11. Shinomiya, K.; Mori, T.; Aoyama, Y.; Sera, T. Biochem. Biophys. Res. Commun. 2011, 414, 733. 12. Mino, T.; Aoyama, Y.; Sera, T. J. Biotechnol. 2009, 140, 156. 13. Sera, T.; Uranga, C. Biochemistry 2002, 41, 7074. 14. Ishiyama, M.; Miyazono, Y.; Sasamoto, K.; Ohkura, Y.; Ueno, K. Talanta 1997, 44, 1299. 15. Moscou, M. J.; Bogdanove, A. J. Science 2009, 326, 1501. 16. Boch, J.; Scholze, H.; Schornack, S.; Landgraf, A.; Hahn, S.; Kay, S.; Lahaye, T.; Nickstadt, A.; Bonas, U. Science 2009, 326, 1509.

17. Christian, M. L.; Demorest, Z. L.; Starker, C. G.; Osborn, M. J.; Nyquist, M. D.; Zhang, Y.; Carlson, D. F.; Bradley, P.; Bogdanove, A. J.; Voytas, D. F. PLoS ONE 2012, 7, e45383. 18. Meckler, J. F.; Bhakta, M. S.; Kim, M.-S.; Ovadia, R.; Habrian, C. H.; Zykovich, A.; Yu, A.; Lockwood, S. H.; Morbitzer, R.; Elsäesser, J.; Lahaye, T.; Segal, D. J.; Baldwin, E. P. Nucleic Acids Res. 2013, 41, 4118. 19. Mussolino, C.; Cathomen, T. Curr. Opin. Biotechnol. 2012, 23, 644. 20. Geurts, A. M.; Cost, G. J.; Freyvert, Y.; Zeitler, B.; Miller, J. C.; Choi, V. M.; Jenkins, S. S.; Wood, A.; Cui, X.; Meng, X.; Vincent, A.; Lam, S.; Michalkiewicz, M.; Schilling, R.; Foeckler, J.; Kalloway, S.; Weiler, H.; Ménoret, S.; Anegon, I.; Davis, G. D.; Zhang, L.; Rebar, E. J.; Gregory, P. D.; Urnov, F. D.; Jacob, H. J.; Buelow, R. Science 2009, 325, 433. 21. Bhaya, D.; Davison, M.; Barrangou, R. Annu. Rev. Genet. 2011, 45, 273. 22. Cong, L.; Ran, F. A.; Cox, D.; Lin, S.; Barretto, R.; Habib, N.; Hsu, P. D.; Wu, X.; Jiang, W.; Marraffini, L. A.; Zhang, F. Science 2013, 339, 819. 23. Mali, P.; Yang, L.; Esvelt, K. M.; Aach, J.; Guell, M.; DiCarlo, J. E.; Norville, J. E.; Church, G. M. Science 2013, 339, 823. 24. Wang, H.; Yang, H.; Shivalila, C. S.; Dawlaty, M. M.; Cheng, A. W.; Zhang, F.; Jaenisch, R. Cell 2013, 153, 910. 25. Bibikova, M.; Carroll, D.; Segal, D. J.; Trautman, J. K.; Smith, J.; Kim, Y.-G.; Chandrasegaran, S. Mol. Cell. Biol. 2001, 21, 289. 26. Händel, E.-M.; Alwin, S.; Cathomen, T. Mol. Ther. 2009, 17, 104.

Sandwiched zinc-finger nucleases demonstrating higher homologous recombination rates than conventional zinc-finger nucleases in mammalian cells.

We previously reported that our sandwiched zinc-finger nucleases (ZFNs), in which a DNA cleavage domain is inserted between two artificial zinc-finger...
353KB Sizes 0 Downloads 0 Views