HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kang, J. S. Articles by Kim, J.-S. Search for Related Content PubMed PubMed Citation Articles by Kang, J. S. Articles by Kim, J.-S. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 12, 8742-8748, March 24, 2000

Zinc Finger Proteins as Designer Transcription Factors^*

Jong Seok Kang and Jin-Soo Kim^§¶

From the  Samsung Biomedical Research Institute and ^§ Toolgen Incorporation, Sungkyunkwan University School of Medicine, 300 Chunchun-Dong, Changan-Ku, Suwon, Republic of Korea

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Recent progress in the design and selection of novel zinc finger proteins with desired DNA binding specificities now allows construction of tailor-made DNA-binding proteins that specifically recognize almost any predetermined DNA sequence. Such novel or "designer" zinc finger proteins with desired DNA binding specificities can serve as efficient transcription factors in various mammalian cell lines. In addition, they may be broadly useful in the regulation of endogenous genes in transgenic organisms and eventually in gene therapy applications. In this report, we use a series of transient and stable transfection experiments to demonstrate that the expression of a target gene can be controlled by changing the in vivo concentration of designer zinc finger proteins in a dose-dependent manner. We also report that designer zinc finger proteins can access their binding sites integrated into the genome and function as potent transcription factors. Our results suggest that designer zinc finger transcription factors that specifically recognize appropriate sites in the promoter of a target gene may have broad applications in the post-genomic era.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Many transcription factors in eukaryotes from yeast to plants and to mammals contain the C₂H₂ class of zinc finger domains as their DNA-binding modules. This class of zinc finger motifs is unique in that their DNA binding specificities are highly adaptable; unlike most other DNA-binding domains, dozens of zinc finger domains characterized thus far bind to highly diverse DNA sequences, with each zinc finger domain able to recognize distinctive DNA binding sites. Many groups have successfully exploited this adaptability of zinc finger domains in DNA recognition to isolate or design novel proteins with altered DNA binding specificities (1-7). Using phage display, zinc finger proteins that exhibit the desired DNA binding specificities can be selected for almost any predetermined DNA sequence.

Because many transcription factors consist of two modular units, that is, an effector (activator or repressor) domain and a DNA-binding domain, it should now be possible to construct novel sequence-specific transcription factors in which tailor-made zinc finger domains are combined with appropriate effector domains. However, the complexity of eukaryotic genomes, in particular the human genome, makes the design of gene-specific transcription factors a formidable challenge. Ideally, designer transcription factors should regulate the expression of one or a few target genes among the tens of thousands of genes in a typical eukaryotic genome. Use of a combinatorial approach (that is, the targeting of a gene of interest with two or more zinc finger transcription factors) may be an effective strategy for achieving gene specificity (8). A second approach would be to fuse two zinc finger proteins, each recognizing adjacent but nonoverlapping 9-base pair (bp)¹ binding sites to yield highly specific DNA-binding proteins (9-11). For example, Kim and Pabo (10) have recently demonstrated that designer proteins containing six zinc finger domains, in which two three-finger proteins are fused via improved linker segments, bind to DNA sequences of 18-20 bp with femtomolar dissociation constants and serve as extremely potent transcriptional repressors in vivo.

Such progress in the selection and design of novel zinc finger proteins with the desired DNA binding specificities should now allow construction of designer transcription factors that specifically regulate the expression of target genes. However, important issues must be considered prior to the application of engineered zinc finger proteins in transgenic organisms and eventually in human patients. For example, is the DNA recognition by the six-finger protein specific enough to selectively regulate the transcription of a target gene that exists among the tens of thousands of genes in a typical eukaryotic genome? Can the expression of a target gene be controlled by altering the concentration of zinc finger proteins in a dose-dependent manner? Can the engineered zinc finger proteins access binding sites in a chromatin complex? In this report, we address these questions by transiently and stably transfecting human cells with either a prototype three-finger peptide from Zif268 or a designer six-finger peptide constructed by fusing two three-finger peptides.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Construction-- Plasmids used in transient transfection studies were described previously (8, 10). Expression plasmids encoding the 268//NRE peptide under the control of the simian virus 40 (SV40) early promoter and the herpes simplex virus thymidine kinase (HSV-TK) promoter were constructed as follows: a DNA fragment encoding the 268//NRE peptide was obtained by polymerase chain reaction amplification using pCS-268//NRE (10) as a template and was subcloned into the NdeI and XbaI sites of pRL-SV40 and pRL-TK (Promega), respectively. Zinc finger expression plasmids used in stable transfection experiments were constructed by polymerase chain reaction amplification of DNA segments encoding the Zif268 peptide and the 268//NRE peptide from pCS-Zif268 and pCS-268//NRE, respectively (10). These DNA segments were inserted into the BamHI and KpnI sites of pIND (Invitrogen) to yield pIND-Zif268 and pIND-268//NRE. These expression plasmids were designed to produce zinc finger peptides with an S-Tag^TM and a nuclear localization signal from SV40 T antigen at the N terminus. Reporter plasmids used in stable transfection experiments were constructed as follows: DNA fragments that contained the luciferase reporter gene under the control of various synthetic promoters were generated by digesting pGL3-TATA/Inr, pGL3-TATA/Inr:N, pGL3-TATA/Inr:Z, and pGL3-TATA/Inr:N/Z (10) with MluI and XbaI and inserting the fragments into the MluI and XbaI sites of pcDNA3 (Invitrogen) to yield pc-TATA/Inr, pc-TATA/Inr:N, pc-TATA/Inr:Z, and pc-TATA/Inr:N/Z, respectively.

Stable Transfection Experiments-- To establish cell lines that express zinc finger proteins under the control of an ecdysone-responsive promoter, pIND-Zif268 or pIND-268//NRE was transfected into human cell line EcR-293 (Invitrogen), in which modified subunits of the ecdysone receptor (VgEcR and RXR) are constitutively expressed. Cells resistant to G418 were selected in medium containing G418 at 0.80 mg/ml and tested for expression of zinc finger proteins using the S-Tag Rapid Assay^TM kit (Novagen) (12, 13). To establish cell lines that contain reporter genes integrated into the genome of the cell, EcR-293 cells were stably transfected with reporter plasmid pc-TATA/Inr, pc-TATA/Inr:N, pc-TATA/Inr:Z, or pc-TATA/Inr:N/Z. G418-resistant cells were selected in medium containing G418 at 0.80 mg/ml.

Transient Transfection Experiments-- 293 cells were transiently transfected by calcium phosphate precipitation as described (14). Transfection experiments typically used cells at 10-30% confluency in monolayer cultures (60-mm diameter plates), and the following plasmids were added: (i) 0.5 µg of either an empty expression plasmid (pCS) or an expression plasmid encoding the zinc finger proteins; (ii) 0.5 µg of a reporter plasmid encoding firefly luciferase; (iii) 2.5 µg of an activator plasmid encoding GAL4-VP16; (iv) 0.25 µg of an internal control plasmid encoding Renilla luciferase (pRL-SV40; Promega); and (v) 11.25 µg of a carrier plasmid (pUC19 or pBluescript II KS⁺). Stably transfected cell lines in which zinc finger proteins were expressed under the control of the ecdysone-responsive promoter were transiently transfected in 6-well plates with (i) 0.25 µg of a reporter plasmid encoding firefly luciferase; (ii) 1.25 µg of an activator plasmid encoding GAL4-VP16; (iii) 0.125 µg of an internal control plasmid encoding Renilla luciferase (pRL-CMV); and (iv) 5.625 µg of carrier plasmid (pUC19 or pBluescript II KS⁺). Ponasterone A was added to the culture media at 5 µM to induce the expression of zinc finger proteins. Luciferase activities in the transfected cells were measured using the Dual-Luciferase^TM Reporter Assay System (Promega).

Gel Shift Assays-- Nuclear extracts were prepared from stably transfected cell lines as described (15). Binding reactions contained radioactive DNA probe (~1 nM) and nuclear extracts (5 µg of total protein) in 20 µl of binding buffer (20 mM Tris-HCl, pH 7.7, 120 mM NaCl, 5 mM MgCl₂, 20 µM ZnSO₄, 10% (v/v) glycerol, 5 mM dithiothreitol, and 0.1 mg/ml of poly(dI-dC)). The DNA binding reactions were incubated at room temperature for 30 min and subjected to electrophoresis on 5% polyacrylamide gels. The DNA sequences of the binding sites were: Z site, 5'-CCGGGTCGTGCGTGGGCGGTACCG-3', and N/Z site, 5'-TCTGCAAGGGTTCAGGCGTGGGCGGTAAG-3' (in each case, the 9- or 19-bp recognition sequence is underlined).

Southern Blot Analyses-- Genomic DNA (10 µg) isolated from stably transfected cell lines was digested with restriction enzymes, subjected to electrophoresis on a 0.8% agarose gel, and blotted onto a nitrocellulose membrane. The membrane was probed with a radiolabeled luciferase DNA segment isolated from pGL3-TATA/Inr. The hybridization reaction was performed overnight at 42 °C in reaction buffer (6× SSC, 5× Denhardt's solution, 0.1% SDS, 100 µg/ml salmon sperm DNA, 0.5% dextran sulfate, and 50% formamide), and the membrane was washed with buffer (0.2× SSC and 0.1% SDS) at 65 °C.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Six-finger Protein-DNA Recognition in Vivo-- We first studied the DNA binding specificity of the six-finger protein 268//NRE in human cells using transient transfection assays. The 268//NRE peptide was constructed by linking the three-finger peptide from Zif268, which recognizes the site 5'-GCGTGGGCG-3', and the three-finger "NRE" peptide (a variant of the Zif268 peptide selected previously by phage display (7)), which binds specifically to part of a nuclear hormone response element (5'-AAGGGTTCA-3') (10). Previous studies have demonstrated that zinc finger proteins efficiently repress transcription of a reporter gene when the proteins are bound to sites near the transcription start point (8, 10). In this study, we used reporter constructs in which one of various forms of a 19-bp binding site for the 268//NRE peptide was incorporated near the transcription start point (Fig. 1A). The reporter constructs also contained five GAL4 binding sites upstream of the promoter of the reporter gene. Each of these reporter plasmids was cotransfected with a plasmid encoding the 268//NRE peptide and a plasmid encoding GAL4-VP16, and the activity of the reporter gene (that is, firefly luciferase) was measured from cell extracts after 2 days of incubation.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Schematic representations of zinc finger proteins and reporter constructs used in transfection studies. A, promoters of luciferase reporter genes. The nucleotide sequences of the promoters of the luciferase reporter gene are aligned to show the location of the zinc finger binding sites (5'-AAGGGTTCA-3' and 5'-GCGTGGGCG-3'), and the positions are numbered with respect to the transcription start point (+1). Identical nucleotides are shown as dots, and mutated nucleotides are underlined. B, zinc finger proteins and their binding sites. The cocrystal structure of the Zif268-DNA complex (28) revealed that each of the three finger domains of the Zif268 peptide primarily contact a three-base subsite as depicted in the figure. The three finger domains of the NRE peptide also may contact DNA in a similar manner. Each finger domain is represented by a circle.

We found that the 268//NRE peptide could efficiently discriminate among closely related binding sequences. Consistent with previous transfection experiments (10), the 268//NRE peptide, when expressed under the control of the cytomegalovirus (CMV) promoter, gave 137-fold (that is, 99.3%) repression of VP16-activated transcription from a promoter that contained the 19-bp zinc finger binding site N/Z (5'-AAGGGTTCAGGCGTGGGCG-3') (Fig. 2A). In contrast, the 268//NRE peptide gave much weaker 1.4- and 16-fold (30 and 94%, respectively) repression of transcription form a promoter that contained the 9-bp partial binding site N (5'-AAGGGTTCA-3') and Z (5'-GCGTGGGCG-3'), respectively. When we used a promoter that contained the mutated N/mZ1 site (5'-AAGGGTTCAGGCGTGGGCC-3'), in which a single G base at the 3' end of the N/Z site was substituted by a C base (underlined), the 268//NRE peptide still gave very strong 123-fold repression (Fig. 2A). In contrast, the six-finger peptide gave only 7.0-fold repression from a promoter that contained the N/mZ2 site (5'-AAGGGTTCAGGCGTGGCCC-3'), in which two G bases at the 3' end of the 19-bp recognition sequence were mutated. (The three-finger Zif268 peptide also gave slightly weaker repression with the mZ1 site than it did with the optimal Z site and gave no repression with the mZ2 site at the promoter; data not shown.) Similarly, the 268//NRE peptide gave only 8.3-fold repression of transcription from a promoter that contained the mN/Z site (5'-TCTGGTTCAGGCGTGGGCG-3'), in which the three bases at the 5' end of the 19-bp site were mutated (Fig. 2A). As shown in Fig. 1B, the three bases at the 5' end (AAG) and the 3' end (GCG) of the N/Z site are recognized by finger 6 (the zinc finger domain at the C terminus) and finger 1 (the zinc finger domain at the N terminus), respectively, of the 268//NRE peptide. Our results indicate that both finger 1 and finger 6 of the 268//NRE peptide contribute significantly to the DNA-binding affinity of the six-finger peptide for the N/Z site in vivo by making base-specific interactions. Previous studies have shown that either the C-terminal or N-terminal zinc finger domains of three-finger proteins contribute less to the stability of the zinc finger protein-DNA complex than do those in the middle (16). Our results imply that each of the six finger domains of the 268//NRE peptide contributes significantly to the affinity and specificity of the six-finger protein for the 19-bp recognition sequence and suggest that the DNA recognition by the six-finger protein is highly specific in vivo.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Transcriptional repression by the six-finger 268//NRE protein under the control of various promoters. Three promoters were used in this study, the CMV immediate early promoter (A), the SV40 early promoter (B), and the HSV-TK promoter (C). Human 293 cells were transiently transfected with an activator plasmid encoding GAL4-VP16, an internal control plasmid encoding Renilla luciferase, a reporter plasmid encoding firefly luciferase, and an effector plasmid encoding the 268//NRE protein under the control of the various promoters. Renilla and firefly luciferase activities were measured 48 h after transfection. The firefly luciferase activities were normalized with respect to the Renilla luciferase activities to correct for transfection efficiency. Repression levels (fold repression) were obtained by dividing the normalized luciferase activities from the cells transfected with an empty effector plasmid by those from the cells transfected with the effector plasmid encoding the 268//NRE peptide. The data represent an average of three independent experiments, and the S.E. is shown.

Zinc Finger Proteins Expressed under the Control of Various Promoters-- To test the effect of changes in concentration of zinc finger proteins in vivo on transcriptional repression, we expressed the 268//NRE peptide under the control of various promoters and performed transfection experiments. Three promoters used in this study were the CMV immediate early promoter, the SV40 early promoter, and the HSV-TK promoter. Control transfection experiments in 293 cells using plasmids encoding Renilla luciferase under the control of each of these promoters revealed that in 293 cells the CMV promoter was about 30-fold stronger than the SV40 promoter (that is, the luciferase activity observed with the CMV promoter was about 30 times higher than that observed with the SV40 promoter) and about 1,000-fold stronger than the HSV-TK promoter (data not shown). As expected, the 268//NRE peptide expressed under the control of the SV40 and HSV-TK promoters repressed transcription of the reporter gene to a lesser degree than did the 268//NRE peptide expressed from the CMV promoter (Fig. 2, B and C). When expressed under the control of the SV40 and HSV-TK promoters, the six-finger peptide gave 25- and 2.8-fold repression, respectively, with the N/Z site. The 268//NRE peptide again gave strongest repression with the N/Z site, somewhat reduced repression with the N/mZ1 site, and significantly reduced repression with the other sites (Fig. 2). (The Zif268 peptide also showed a similar repression pattern with these sites when expressed under the control of the three promoters; data not shown.) These results indicate that the repression level of target gene expression is highly dependent on the cellular concentration of zinc finger transcription factors.

Inducible Expression of Zinc Finger Proteins-- We next tested whether expression of the reporter gene could be regulated by controlling the concentration of zinc finger proteins in vivo with a small molecule in a dose-dependent manner. To this end, we adopted an ecdysone-inducible expression system in which the concentration of a protein of interest whose expression is under the control of an ecdysone-responsive promoter could be regulated by adjusting the amount of the insect hormone ecdysone (or its analog ponasterone A) in the culture media (17). We constructed effector plasmids in which the genes encoding either the Zif268 peptide or the 268//NRE peptide were placed under the control of a promoter containing ecdysone-responsive elements. We then stably transfected each of these plasmids into the human cell line EcR-293, in which modified ecdysone receptors are constitutively expressed. Because the zinc finger peptides were fused with an S-Tag^TM, we screened G418-resistant clones by assaying ribonuclease activity after activation of the tagged proteins with S-protein (12, 13). The isolated cell lines expressed either the Zif268 peptide (cell line c268) or the 268//NRE peptide (cell line c268//NRE) only in the presence of ponasterone A. Gel shift assays confirmed that the expression of zinc finger peptides was induced by ponasterone A in these cell lines (data not shown).

Using cell lines c268 and c268//NRE, we sought to determine whether expression of the reporter gene could be regulated by ponasterone A in a dose-dependent manner. In this series of experiments, each of these cell lines was transiently transfected with the GAL4-VP16 activator plasmid and a reporter plasmid that contained either the Z site or the N/Z site in the promoter. Varying amounts of ponasterone A were added to the culture media of the transfected cells. A gradual increase in the level of repression of activated transcription was observed with increasing concentrations of ponasterone A (up to 20 µM in the culture media) (Fig. 3A). When ponasterone A was added to a final concentration of 20 µM, the Zif268 peptide expressed in the c268 cell line gave 8.0-fold repression of activated transcription from the promoter that contained the Z site. Under these same conditions, the 268//NRE peptide expressed in the c268//NRE cell line gave 24-fold repression of activated transcription from the promoter that contained the N/Z site. In contrast, no significant repression was observed in these cell lines in the presence of ponasterone A from a control promoter that contained no zinc finger binding sites (Fig. 3B), indicating that repression of target gene expression in these cell lines was binding site-specific. The repression levels observed with these stably transfected cell lines even at the highest concentration of ponasterone A tested (20 µM) was not as high as those observed with transiently transfected cells in which zinc finger proteins were expressed under the control of the CMV promoter. This is consistent with our observation that the cellular concentrations of zinc finger proteins (measured using the S-Tag^TM assay kit) whose expression was induced by ponasterone A in these cell lines were significantly lower than those of zinc finger proteins expressed under the control of the CMV promoter in transiently transfected 293 cells (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Regulation of target gene expression by ecdysone-inducible expression of zinc finger proteins. A, dose-dependent repression of gene expression by zinc finger proteins. The c268 and c268//NRE cell lines were transiently transfected with the indicated reporter plasmids encoding firefly luciferase, an internal control plasmid encoding Renilla luciferase, and an activator plasmid encoding GAL4-VP16. Varying amounts of ponasterone A were added to the culture media, and luciferase activities were measured 48 h later. Fold repression was calculated by dividing the normalized luciferase activities from cells in the absence of ponasterone A by those from cells in the presence of ponasterone A. B, site-specific repression of gene expression by zinc finger proteins expressed under the control of an ecdysone-responsive promoter. The c268 and c268//NRE cell lines were transiently transfected with the indicated reporter plasmids encoding firefly luciferase, an internal control plasmid encoding Renilla luciferase, and an activator plasmid encoding GAL4-VP16. Ponasterone A (5 µM) was added to the culture media to induce the expression of zinc finger proteins. Fold repression was calculated as described in A. The data represent an average of three independent experiments, and the S.E. is shown.

The inducible zinc finger expression system also may allow reversible regulation of target gene expression. To test this possibility, we first treated the c268 and c268//NRE cell lines with 5 µM ponasterone A for 24 h. These cells were then transiently transfected with reporter plasmids that contained either the Z site or the N/Z site in the promoter, and ponasterone A was removed from the culture media at various time points. Luciferase activity was measured 48 h after transient transfection and was compared with control luciferase levels in cells not treated with ponasterone A before and after parallel transient transfections. Luciferase activity in the c268 cell line when ponasterone A was removed after the initial 24-h treatment (Fig. 4A, 0 h) was nearly equivalent (87%) to that in cells not treated with ponasterone A, which was set to 100%. However, luciferase activity in the c268//NRE cell line when ponasterone A was removed after the initial 24-h treatment (Fig. 4A, 0 h) was only 17% of that in the cells not treated with ponasterone A. Gel shift experiments with nuclear extracts prepared from the c268 and c268//NRE cells showed that the Zif268 peptide in the c268 cell line was almost undetectable at 48 h after depleting ponasterone A from the culture media (Fig. 4B). In contrast, a significant amount of the 268//NRE peptide in the c268//NRE cell line was detected even at 84 h after depleting ponasterone A. This suggests that the Zif268 peptide has a shorter half-life in cells than does the 268//NRE peptide. (An alternative explanation is that the six-finger 268//NRE peptide is simply more potent in DNA binding than the three-finger Zif268 peptide. Previous studies have shown that the 268//NRE peptide binds to the 19-bp recognition sequence about 6,000-fold more tightly than the Zif268 peptide (10).) Our results suggest that it may be possible to reversibly regulate target gene expression by adopting an inducible system to express zinc finger proteins. However, the long half-life (or high affinity for binding sites) of some zinc finger proteins in cells may not allow rapid on-off control of target gene expression.

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Reversible regulation of target gene expression using the ecdysone-inducible zinc finger expression system. A, reversible regulation of target gene expression. The c268 and c268//NRE cell lines were pretreated with ponasterone A (5 µM) for 24 h and then transiently transfected with an internal control plasmid encoding Renilla luciferase, an activator plasmid encoding GAL4-VP16, and a reporter plasmid containing either the Z site or the N/Z site in the promoter. At various time points after transient transfection (0, 12, 24, 36, and 48 h), ponasterone A was depleted from the culture media. The firefly luciferase activities were measured 48 h after transient transfection and normalized with respect to the Renilla luciferase activities. As a control, cells were not treated with ponasterone A before and after transient transfection, and the normalized luciferase activity from these cells measured 48 h after transfection was set to 100%. B, gel shift assays. The c268 and c268//NRE cell lines were treated with ponasterone A (5 µM) for 24 h, and then ponasterone A was depleted from the culture media. Nuclear extracts were prepared at the indicated time points (lanes 2-5) after depleting ponasterone A from the media. Nuclear extracts also were prepared from cells not treated with ponasterone A (lane 1). The DNA binding reactions contained nuclear extract (5 µg of total protein), a 32P-labeled zinc finger binding site (~ 1 nM) and poly(dI-dC) (0.1 mg/ml). The DNA binding reactions were incubated at room temperature for 30 min and then subjected to electrophoresis on 5% polyacrylamide gels.

Repression of Reporter Genes Integrated into the Genome-- We also used stable transfection experiments to investigate whether zinc finger proteins could bind to target sites that are packaged in a chromatin complex. Although several previous studies have focused on the potential use of zinc finger proteins as transcription factors in various mammalian cell lines (8-10, 18-20), most of these studies have involved transient transfection experiments, where the target genes are not integrated into the genome. Because transiently transfected DNA may not form tight chromatin complexes with histones (21), we decided to test whether engineered zinc finger proteins can access binding sites that are integrated into the genome and serve as efficient transcription factors. There also was a concern that zinc finger proteins selected in vitro via phage display (such as the "NRE" moiety of the 268//NRE peptide) might not serve as efficient transcription factors, because phage display procedures utilize naked DNA-binding sites not in a chromatin complex. Therefore, we stably transfected EcR-293 cells with luciferase reporter plasmids that contained the various zinc finger binding sites in their promoters and obtained G418-resistant clones. From those drug-resistant clones, we isolated cell lines that showed high luciferase activity when transiently transfected with the activator plasmid encoding GAL4-VP16 but low activity in the absence of the activator plasmid.

We observed sequence-specific repression of the reporter gene integrated into the cellular genome by zinc finger proteins in these cell lines. Thus, when the plasmid encoding the Zif268 peptide was transiently transfected into the cell lines in which the promoter of the integrated reporter gene contained either the Zif268 site (designated as cTATA/Inr:Z) or the N/Z site (designated as cTATA/Inr:N/Z-1, 2, and 3), 6.2-16-fold repression of VP16-activated transcription was observed (Fig. 5). (Schematic representations of reporter plasmids that were used to construct these cell lines are shown in Fig. 1A.) In contrast, the Zif268 peptide gave no significant repression in the cTATA/Inr and cTATA/Inr:N cell lines, in which the promoter of the integrated reporter gene contained no Zif268 binding site. The 268//NRE peptide gave 77-92-fold repression in the cTATA/Inr:N/Z-1, 2, and 3 cell lines, in which the promoter of the integrated reporter gene contained the 19-bp N/Z site. (All of these clones were constructed by stable transfection of reporter plasmid pc-TATA/Inr:N/Z (Fig. 1A) into EcR-293 cells.) The 268//NRE peptide also gave 33-fold repression in the cTATA/Inr:Z cell line, in which the promoter contained the Zif268 site. In contrast, the six-finger protein gave no significant repression in cell lines cTATA/Inr and cTATA/Inr:N.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Transcriptional repression of chromosomally integrated reporter genes by zinc finger proteins. Various reporter plasmids encoding firefly luciferase were stably transfected into EcR-293 cells. The stably transfected cell lines were then transiently transfected with effector plasmids encoding zinc finger proteins (or the empty effector control plasmid), an activator plasmid encoding Gal4-VP16, and an internal control plasmid encoding Renilla luciferase. Renilla and firefly luciferase activities were measured 48 h after transfection. The firefly luciferase activities were normalized with respect to the Renilla luciferase activities to correct for transfection efficiency. Repression levels (fold repression) were obtained by dividing the normalized luciferase activities from the cells transfected with the empty effector plasmid by those from the cells transfected with the effector plasmids encoding the zinc finger peptides. The data represent an average of three independent experiments, and the S.E. is shown.

It is possible that many copies of the luciferase reporter DNA may have integrated at a single site in the genome. If this happens, then the reporter gene may not have been packaged as a tight chromatin complex. This loose structure might help zinc finger proteins to access their binding sites more efficiently and thus to serve as strong transcriptional repressors in our transfection experiments. This hypothesis prompted us to perform Southern blot analyses to characterize the integrated reporter constructs. Genomic DNA isolated from the cell lines in which the luciferase reporter plasmids were stably integrated was digested with either HindIII (which has a single recognition site in the integrated plasmids) or NdeI (which has no recognition site in the plasmids) or both and was subjected to electrophoresis on agarose gels. We estimated the copy number of the integrated plasmids from the size, intensity, and number of bands in the Southern blot. Only one or a few copies of the reporter plasmids were integrated into the genome in most of the cell lines. For example, genomic DNA isolated from the cTATA/Inr:N/Z-3 cell line showed a single hybridizing fragment of 6-10 kilobases (the size of the integrated plasmid was 7.3 kilobases), when digested with either HindIII or NdeI or both (Fig. 6, lanes 19-21). This suggests that a single copy of the plasmid was integrated into the genome at a single site in the cTATA/Inr:N/Z-3 cell line. In contrast, the large size and strong intensity of the fragments observed in digested DNA from the cTATA/Inr:N/Z-2 cell line suggest that many copies of the reporter plasmid were integrated into the genome (Fig. 6, lanes 16-18). The different band patterns observed with digested genomic DNA from the cTATA/Inr:N/Z-1, 2, and 3 cell lines indicate that these cell lines are indeed independent clones. We note that despite the differences in copy numbers and site of integration of the reporter plasmid, the 268//NRE peptide gave comparable levels of repression of activated transcription in these cell lines (Fig. 5). Taken together, these results indicate that engineered zinc finger proteins can access binding sites in the context of the genome and serve as potent transcription factors.

View larger version (64K): [in this window] [in a new window]   Fig. 6.   Southern blot analyses. Genomic DNA (10 µg) isolated from the indicated cell lines was digested with HindIII (indicated as H; lanes 1, 4, 7, 10, 13, 16, and 19), NdeI (indicated as N; lanes 2, 5, 8, 11, 14, 17, and 20), or HindIII + NdeI (indicated as H+N; lanes 3, 6, 9, 12, 15, 18, and 21), subjected to electrophoresis on a 0.8% agarose gel, and then blotted onto a nitrocellulose membrane. The membrane was hybridized with a 32P-labeled luciferase DNA probe from pGL3-TATA/Inr. The sizes and positions of molecular size markers in kilobases (kb) are indicated at the left.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Artificial regulation of gene expression using "designer" transcription factors may have broad applications in the post-genomic era. Genome projects for various organisms will eventually reveal the DNA sequences of tens of thousands of novel genes whose functions are yet to be determined. Experiments that employ DNA microarrays and differential display polymerase chain reaction also will identify candidate genes whose expression levels may contribute to the phenotypic differences among the cell types used in the analyses. To study the function(s) of a gene of interest or to validate the biological role(s) of a candidate gene, biologists often seek to up- and down-regulate the expression of the gene in a living organism or in cell culture and monitor the biological consequences of the altered expression. Zinc finger proteins appear to be well suited for these types of experiments, because a single zinc finger protein can be used for both up- and down-regulation of a gene of interest by fusing the zinc finger DNA-binding domain with appropriate effector domains or by targeting the zinc finger domain to critical sites within the promoter of the gene of interest. To utilize zinc finger domains in such experiments, two three-finger proteins are selected or designed to bind to two nonoverlapping 9-10-bp target sites that are adjacent to each other. The two three-finger proteins are then connected via an appropriate linker to yield a six-finger protein that recognizes the extended 18-20-bp target site. The resulting zinc finger protein may serve as a highly specific transcription factor in cell culture or in transgenic organisms, and this approach may eventually be extended to applications in gene therapy.

Our studies suggest that the zinc finger approach has many useful properties that make this system ideal for broad applications in artificial gene regulation: (i) Our results with the six-finger protein (268//NRE) suggest that the zinc finger approach is highly specific. In our transfection studies, the 268//NRE peptide discriminated very efficiently among closely related sequences. Six-finger proteins that recognize target sequences of 18-20 bp may serve as "genome-specific" transcription factors, regulating only a limited number of genes in a complex genome. (ii) Our stable transfection studies suggest that engineered zinc finger proteins could access DNA-binding sites in a chromatin complex and serve as efficient transcription factors, irrespective of the locations of such binding sites in the genome. (iii) Conditional regulation of a target gene can be achieved by subjecting the expression of zinc finger transcription factors to inducible expression systems. By adopting an inducible expression system, we were able to regulate the cellular concentration of zinc finger proteins, and this allowed us to control the expression of a reporter gene in a dose-dependent manner.

Many inducible gene expression systems using small molecules such as ecdysone (17), tetracycline (22), rapamycin (23), and RU486 (24) have been developed to conditionally regulate the expression of a gene of interest. In these systems, the promoter of the target gene must be manipulated to contain the appropriate binding sites for artificial activators. Thus, these systems cannot be used to achieve the conditional regulation of endogenous genes in organisms other than mouse, in which gene targeting via homologous recombination is possible. Our results suggest that conditional regulation of endogenous genes in cell culture and in most organisms can be achieved by combining an inducible expression system with our zinc finger approach. The zinc finger approach will circumvent the necessity of manipulating endogenous promoters, because novel zinc finger peptides can be selected or designed to specifically recognize endogenous DNA sites.

The zinc finger approach may provide several important advantages over antisense or ribozyme approaches, which are broadly used to regulate endogenous gene expression: (i) Predicting effective binding sites for zinc finger proteins is straightforward. Previous transfection studies have shown that zinc finger proteins serve as efficient transcriptional repressors when they bind to critical sites near the TATA box or the transcription start point (8). In contrast, it is much more difficult to choose effective target sites for antisense or ribozyme molecules, and the target sites in mRNA usually are determined by trial and error (25). (ii) The zinc finger approach may give much stronger repression than do the antisense or ribozyme approaches. Our study shows that a six-finger protein gives >99% repression of activated transcription in vivo. Thus it appears that blocking the transcription of a single copy of a target gene via the zinc finger approach is more efficient than inhibiting the translation of many copies of mRNA via the antisense or ribozyme approaches, which seldom give >90% repression of target gene expression. (iii) Targeting two or more binding sites in the promoter of a gene of interest with two or more zinc finger proteins can give stronger repression than targeting a single site with a single zinc finger protein (8). In contrast, targeting multiple sites in an mRNA using antisense or ribozyme molecules seldom gives synergistic or additive effects. The cumulative effect of multiple zinc finger proteins on transcriptional repression may be useful not only to shut down the expression of a target gene but also to reduce the possibility of selecting resistant clones. In principle, when applied to the suppression of oncogenes or viral genes, all three approaches (zinc finger, ribozyme, and antisense) may suffer from potential problems associated with resistant clones in which point mutations have been introduced at the binding site. When two or more zinc finger proteins bind to distinctive sites, however, it would be much more rare for a mutant clone to arise, because it would require that mutations be introduced into all of the binding sites simultaneously in the same clone. (iv) The zinc finger approach is more flexible in that the zinc finger DNA-binding domains can be linked to domains from other proteins such as basal transcription factors (26), transcriptional repression domains (9, 20), and transcriptional activation domains (9, 19, 20). Thus, the zinc finger approach can be used both to down- and up-regulate target gene expression.

The fusion of transcriptional repression domains such as the Krüppel-associated box (KRAB) domain to zinc finger proteins with the desired DNA binding specificities appears to be a powerful method for repression of target gene expression (9, 20). It has been demonstrated that the KRAB domain, when fused to heterologous DNA-binding proteins, represses the expression of a reporter gene, even when bound to sites a few kilobase pairs upstream from the promoter of the gene (27). This property of repression at a distance may be advantageous in many cases, because the zinc finger proteins that are fused to the KRAB domain can be selected or designed to bind target sites anywhere within a few kilobase pairs from the promoter of a target gene. In contrast, if isolated zinc finger proteins (that is, without KRAB domain fusion) are used as transcriptional repressors, the binding sites must be located within a ~70-bp region around the TATA box and the transcription start point (8). This restriction may impose some limitations on the choice of appropriate DNA sites.

In terms of specificity, however, the targeting of critical sites around the transcription start point with isolated zinc finger proteins may be more desirable. It has been estimated that there are thousands or more adventitious binding sites in the human genome for any three-finger protein that recognizes a 9-10-bp DNA sequence. Gel shift experiments with nonspecific competitor DNA revealed that the three-finger Zif268 peptide specifically recognizes a 7-8-bp site (7, 10), although its binding site covers 9-10 bp in the Zif268-DNA cocrystal structure (28). This finding suggests that there might be about 100,000 sites in the human genome where the Zif268 peptide could bind. When isolated zinc finger proteins bind to only a few bp upstream of the TATA box or ~50 bp downstream of the transcription start site, they exert no repression at all (8, 26). Thus, isolated zinc finger proteins that bind to most of the numerous potential binding sites in the genome would not affect the gene expression. (One isolated report by Choo et al. (18) showed that a designer zinc finger protein gave effective repression when it bound to a site far downstream of the transcription start point. However, this DNA binding event would need to block transcriptional elongation rather than transcriptional initiation. We believe that it is an exception rather than a rule that zinc finger proteins can block transcriptional elongation and not transcriptional initiation. We found that even the six-finger 268//NRE peptide that bound to the 18-20-bp DNA sites with femtomolar dissociation constants could not block transcriptional elongation in our transfection experiments, when the 268//NRE peptide was targeted to a site more than 50 bp downstream of the transcription start point (data not shown). In dramatic contrast, the six-finger peptide gave highly efficient (137-fold) repression of VP16-activated transcription, when it bound to a site near the transcription start point (Fig. 2A), where the transcription initiation complex is formed.)

Zinc finger-KRAB fusion proteins, however, are likely to affect the expression of many genes other than the intended target gene, because the KRAB domain has the ability to function at a distance. This feature of the KRAB domain may cause serious problems in the application of the zinc finger approach in transgenic organisms or in gene therapy. The zinc finger-KRAB fusion approach also may suffer from problems associated with the quenching effect of the repression domain. It seems possible that the KRAB domain of overexpressed zinc finger-KRAB fusion protein in cells may perturb the expression patterns of many genes indirectly by sequestering essential cellular transcription factors. In contrast, it is much less likely that an isolated zinc finger protein would show such pleiotropic effects on gene expression.

	FOOTNOTES

^* This work was supported by Grant B98018 from the Samsung Biomedical Research Institute.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed. E-mail: jinsookim@netsgo.com.

	ABBREVIATIONS

The abbreviations used are: bp, base pair(s); SV40, simian virus 40; HSV-TK, herpes simplex virus thymidine kinase; CMV, cytomegalovirus; KRAB, Krüppel-associated box.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				C.-O. Yun, H.-C. Shin, T.-D. Kim, W.-H. Yoon, Y.-A Kang, H.-S. Kwon, S. K. Kim, and J.-S. Kim Transduction of artificial transcriptional regulatory proteins into human cells Nucleic Acids Res., September 1, 2008; 36(16): e103 - e103. [Abstract] [Full Text] [PDF]

				W. Tan, Z. Dong, T. A. Wilkinson, C. F. Barbas III, and S. A. Chow 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 Region in Human Cells J. Virol., February 15, 2006; 80(4): 1939 - 1948. [Abstract] [Full Text] [PDF]

				W. Tan, K. Zhu, D. J. Segal, C. F. Barbas III, and S. A. Chow Fusion Proteins Consisting of Human Immunodeficiency Virus Type 1 Integrase and the Designed Polydactyl Zinc Finger Protein E2C Direct Integration of Viral DNA into Specific Sites J. Virol., February 1, 2004; 78(3): 1301 - 1313. [Abstract] [Full Text] [PDF]

				T. G. Uil, H. J. Haisma, and M. G. Rots Therapeutic modulation of endogenous gene function by agents with designed DNA-sequence specificities Nucleic Acids Res., November 1, 2003; 31(21): 6064 - 6078. [Abstract] [Full Text] [PDF]

				D. Xu, D. Falke, and R. L. Juliano P53-Dependent Cell-Killing by Selective Repression of Thymidine Kinase and Reduced Prodrug Activation Mol. Pharmacol., August 1, 2003; 64(2): 289 - 297. [Abstract] [Full Text] [PDF]

				X. Guan, J. Stege, M. Kim, Z. Dahmani, N. Fan, P. Heifetz, C. F. Barbas III, and S. P. Briggs Heritable endogenous gene regulation in plants with designed polydactyl zinc finger transcription factors PNAS, October 1, 2002; 99(20): 13296 - 13301. [Abstract] [Full Text] [PDF]

				D. Xu, D. Ye, M. Fisher, and R. L. Juliano Selective Inhibition of P-glycoprotein Expression in Multidrug-Resistant Tumor Cells by a Designed Transcriptional Regulator J. Pharmacol. Exp. Ther., September 1, 2002; 302(3): 963 - 971. [Abstract] [Full Text] [PDF]

				R.L. Juliano, A. Astriab-Fisher, and D. Falke Macromolecular Therapeutics: Emerging Strategies for Drug Discovery in the Postgenome Era Mol. Interv., April 1, 2001; 1(1): 40 - 53. [Abstract] [Full Text]

				M. Moore, Y. Choo, and A. Klug From the Cover: Design of polyzinc finger peptides with structured linkers PNAS, February 13, 2001; 98(4): 1432 - 1436. [Abstract] [Full Text] [PDF]

				M. E. Laurance, D. B. Starr, E. F. Michelotti, E. Cheung, C. Gonzalez, A. W. Tam, J. Deikman, C. A. Edwards, and A. J. Bardwell Specific down-regulation of an engineered human cyclin D1 promoter by a novel DNA-binding ligand in intact cells Nucleic Acids Res., February 1, 2001; 29(3): 652 - 661. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal