Dual Role for Zn2+ in Maintaining Structural Integrity and Inducing DNA Sequence Specificity in a Promiscuous Endonuclease*

We describe two uncommon roles for Zn2+ in enzyme KpnI restriction endonuclease (REase). Among all of the REases studied, KpnI REase is unique in its DNA binding and cleavage characteristics. The enzyme is a poor discriminator of DNA sequences, cleaving DNA in a promiscuous manner in the presence of Mg2+. Unlike most Type II REases, the active site of the enzyme comprises an HNH motif, which can accommodate Mg2+, Mn2+, or Ca2+. Among these metal ions, Mg2+ and Mn2+ induce promiscuous cleavage by the enzyme, whereas Ca2+-bound enzyme exhibits site-specific cleavage. Examination of the sequence of the protein revealed the presence of a zinc finger CCCH motif rarely found in proteins of prokaryotic origin. The zinc binding motif tightly coordinates zinc to provide a rigid structural framework for the enzyme needed for its function. In addition to this structural scaffold, another atom of zinc binds to the active site to induce high fidelity cleavage and suppress the Mg2+- and Mn2+-mediated promiscuous behavior of the enzyme. This is the first demonstration of distinct structural and catalytic roles for zinc in an enzyme, suggesting the distinct origin of KpnI REase.

A large number of proteins have bound zinc ions, which contribute to protein stability and/or catalytic functions more widely than any other transition metal ions (1,2). A catalytic role for zinc was first shown in the case of carbonic anhydrase (3), and its structural role was first proposed and demonstrated for the transcription factor TFIIIA (4,5). Since then, the roles for Zn 2ϩ in numerous zinc-binding proteins have been identified and characterized. In many examples, the role of zinc ion is neither strictly structural nor catalytic, as in aminoacyl-tRNA synthetases, where zinc is involved in amino acid discrimination (6). Zinc binding motifs are structurally diverse and are present among proteins that perform a broad range of functions in various cellular processes. For instance, the motifs play a role in DNA recognition, transcription activation, protein folding and assembly, and protein-protein interactions (7). Zinc binding is observed in different groups of nucleases, I-PpoI, I-TevI, T4 endonuclease VII, DNA repair endonuclease IV, colicin E7, and S1 nuclease (8 -12). The binding of zinc is important for structural stability of I-PpoI, I-TevI, and T4 endonuclease VII and for catalysis in endonuclease IV and colicin E7. Bioinformatic analysis showed that McrA has a zinc binding fold, suggested to be needed for structural integrity (13). R.BslI contains two glucocorticoid receptor-like zinc (Cys 4 ) binding motifs, which are important for the protein-DNA and protein-protein interactions (14). In this paper, we describe two distinct roles for Zn 2ϩ in R.KpnI.
Type II REases 3 require Mg 2ϩ or a similar divalent metal ion to cleave DNA. Almost 3700 Type II restriction enzymes, representing more than 262 distinct specificities, are known to date (15). Most Type II REases belong to the PD . . . (D/E)XK superfamily (16). Recent structural and bioinformatics studies revealed that apart from the PD . . . (D/E)XK superfamily, few REases belong to other nuclease superfamilies, such as Nuc, HNH, and YIG-GIY, which are structurally unrelated to each other (17)(18)(19).
Sequence alignment and subsequent validation experiments showed that R.KpnI is the first member of the HNH superfamily (20). Although at first glance R.KpnI appeared to be a typical dimeric Type IIP REase recognizing and cleaving palindromic sequence GGTACC, it has several distinct features. The properties include prolific promiscuous activity in the presence of Mg 2ϩ which is further enhanced with Mn 2ϩ , efficient site specific high fidelity DNA cleavage when Ca 2ϩ is used instead of Mg 2ϩ , and suppression of the promiscuous cleavage activity in presence of Ca 2ϩ (21). Kinetic studies revealed that the Ca 2ϩmediated exquisite specificity is achieved at the step of DNA cleavage (22).
The alignment of McrA, T4 endonuclease VII, and R.KpnI is depicted in Fig. 1A. The former two enzymes have tetra-Cys Zn 2ϩ fingers, whereas R.KpnI has an unusual CCCH putative Zn 2ϩ finger. Here we describe the importance of the Zn 2ϩ finger motif in Zn 2ϩ coordination. Surprisingly, the bound Zn 2ϩ has more complex, multiple roles in R.KpnI function, in a manner distinct from any other restriction-modification system.

EXPERIMENTAL PROCEDURES
Enzymes and DNA-T4 polynucleotide kinase, Pfu DNA polymerase, and DpnI were purchased from New England Biolabs. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. 1  Oligonucleotides (Sigma and Microsynth) were purified on 18% urea-polyacrylamide gel (33) and end-labeled with T4 polynucleotide kinase and [␥-32 P]ATP (6000 Ci/mmol). Mutagenesis, Expression, and Purification of Mutant Proteins-The model of R.KpnI was built using the structure of T4 endonuclease VII and other structurally characterized HNH superfamily nucleases. The detailed procedure has been described and discussed previously (20). Site-directed mutagenesis was performed by the megaprimer method (23). The mutations were confirmed by sequencing. The WT and mutants were expressed in Escherichia coli BL26 (F Ϫ omp T hsdSB (rB Ϫ mB Ϫ ) gal dcm ⌬lac (DE3) nin5 lac UV5-T7 gene 1) containing KpnI methyltransferase, and the cells were induced with 0.3 mM isopropyl-␤-D-thiogalactopyranoside as described previously (24). Cells were lysed by sonication in buffer A containing 10 mM potassium phosphate (pH 7.0), 1 mM EDTA, 7 mM 2-mercaptoethanol. The supernatant was subjected to 0 -50% ammonium sulfate fractionation. The samples were dialyzed against buffer A and purified by phosphocellulose and Hi-Trap heparin columns. The fractions containing the enzyme were pooled and dialyzed against buffer B (10 mM Tris-HCl (pH 7.4), 0.1 mM EDTA, 50 mM KCl, 5 mM 2-mercaptoethanol, and 50% glycerol). The concentration of the proteins was estimated by the method of Bradford (25).
Atomic Absorption Analysis-Purified R.KpnI (10 mg) was denatured and renatured in the presence or absence of 100 M ZnCl 2 or 5 mM EDTA and then dialyzed overnight against 20 mM Tris-HCl (pH 7.4), 150 mM NaCl at 4°C with buffer changes to eliminate excess metal ions or chelators. Chelex-100 resin (Sigma) was used to remove trace metal ions in all of the buffers. The samples were analyzed by atomic absorption spectroscopy. The dialyzed buffer after Chelex treatment was used as a blank, and the residual Zn 2ϩ background was subtracted from the measurement of protein samples.
Zn 2ϩ Blotting Assay-Purified R.KpnI and its mutants (0 -6 g) was slot-blotted onto nitrocellulose membrane presoaked in buffer (10 mM Tris, pH 7.5, 100 mM NaCl, 1 mM EDTA). Proper transfer was ascertained by Ponceau-S staining with transferred protein amounts estimated using Quantity One software. After transfer, the membrane was incubated at 37°C for 1 h in buffer containing 10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 10 mM MgCl 2 and subsequently washed three times (15 min each) in the same buffer. The membrane was next incubated in buffer containing 30 Ci of 65 ZnCl 2 (specific activity, 800 mCi/g; BARC, Mumbai) at room temperature for 1 h with gentle rocking. The unbound radioactivity was removed by washing the membrane three times with buffer (20 min each). The membrane was dried and exposed to a PhosphorImager screen.
Electrophoretic Mobility Shift Assay-Different concentrations of the WT (1-10 nM) and mutant R.KpnI (10 -100 nM) were incubated with the 0.2 pmol of end-labeled doublestranded oligonucleotide (20-mer) containing the R.KpnI recognition site in binding buffer (20 mM Tris-HCl (pH 7.4), 25 mM NaCl, and 5 mM 2-mercaptoethanol) for 15 min on ice. The free DNA and the enzyme-bound complexes were resolved by 8% native polyacrylamide gel electrophoresis in 1ϫ TBE buffer (89 mM Tris-HCl, 89 mM boric acid, and 1 mM EDTA), and then signals were detected by autoradiography.
In Vitro DNA Cleavage and Steady-state Kinetic Analysis-Purified R.KpnI and its mutants were incubated with 500 ng of plasmid DNA in buffer containing 10 mM Tris-HCl (pH 7.4), 5 mM 2-mercaptoethanol, 2 mM MgCl 2 , or 10 -100 M ZnCl 2 for 1 h at 37°C. The cleavage products were analyzed on 1% agarose gel. For kinetic analysis, the purified enzyme was dialyzed against 10 mM EDTA to remove any bound metal ions. Steady-state kinetic time courses with canonical DNA substrates were measured at DNA concentrations of 5-150-fold molar excess over dimeric enzyme (1 nM) in the presence of 100 M Zn 2ϩ . The kinetic parameters were determined as described (22).
Circular Dichroism-The wild type R.KpnI and its mutants harboring the C119A, C128A, C171A, and H174A mutations were analyzed by CD. The CD spectra were recorded at 25°C from 250 to 200 nm using a JASCO J-720 spectropolarimeter and a cuvette of path length 0.2 cm. The spectra were collected at scanning rate of 50 nm/min, and triplicate spectrum readings were collected per sample. All of the samples were base linecorrected before calculations. The buffer used was 50 mM Tris-HCl (pH 7.4), 75 mM NaCl, 1 mM 2-mercaptoethanol. The proteins were at a concentration of 0.2 g/l, and the molar ellipticity () was calculated using the equation, where obs is the observed ellipticity, M r is molecular weight, C is concentration (in mg/ml), l is the path length of the cuvette in centimeters, n refers to the number of residues, and deg is degrees. Thermal stability of the protein samples was assessed using CD by following changes in the spectrum with increasing temperature (25-75°C). A single wavelength (222 nm) was chosen to monitor the protein structure, and the signal at that wavelength is recorded continuously as the temperature is raised.
Tryptic Digestion of R.KpnI-Proteolytic digestions of different samples of R.KpnI and its mutants (1 mg/ml) were carried out in 50 mM Tris-HCl buffer, pH 8.5, and at 37°C using 1% trypsin. 5-l samples (5 g) were taken after various time periods, and trypsin was inactivated with buffer containing phenylmethylsulfonyl fluoride. Samples were analyzed by 15% SDS-PAGE.

R.KpnI Has Two Zn 2ϩ
Binding Sites-Sequence analysis and homology modeling of R.KpnI predicted the presence of an unusual CCCH zinc finger motif (20) different from other previously described commonly found zinc finger motifs (CCCC, CCHH, CCHC). The putative zinc-coordinating residues are shown in the model (Fig. 1B). The arrangement of cysteines and histidine (CCCH) in R.KpnI is rare among zinc finger proteins of prokaryotic origin. To estimate the bound Zn 2ϩ , we performed atomic absorption spectrometry (Table 1). Extensively dialyzed R.KpnI was found to bind 2 mol of Zn 2ϩ /mol of dimer. The dimeric nature of the enzyme has been established before Dual Role for Zn 2؉ in R.KpnI NOVEMBER 2, 2007 • VOLUME 282 • NUMBER 44 (24). Bound Zn 2ϩ was not replaceable with other metal ions, since even after exhaustive dialysis against a buffer that contained 10 mM MgCl 2 , 2 mol of Zn 2ϩ were still retained in the protein, indicating that the site was inert to exchange by Mg 2ϩ . However, when urea-denatured protein was renatured in presence of ZnCl 2 , the zinc content increased to 4 mol/mol of R.KpnI dimer. Dialysis of this preparation in the presence of MgCl 2 resulted in loss of 2 mol of Zn 2ϩ from R.KpnI, indicating the replacement of two of the four Zn 2ϩ ions by Mg 2ϩ . The other 2 mol of tightly bound zinc could not be replaced by Mg 2ϩ . These results show that R.KpnI monomer possesses two zinc-binding sites; one is replaceable with Mg 2ϩ , and another one is not ( Table 1).
Role of the Zn 2ϩ in Structural Integrity-To define the role of zinc atoms in R.KpnI, we compared the DNA binding and cleavage properties of native, zinc-demetalated (renatured in the absence of Zn 2ϩ ), and zinc-reconstituted enzymes. The zinc-demetalated R.KpnI (the apoenzyme with no Zn 2ϩ bound) did not bind DNA ( Fig. 2A). The enzyme had no DNA cleavage activity in the presence of 2 mM MgCl 2 . Similar experiments were carried out with native and zinc reconstituted R.KpnI (Fig.  2B). The zinc-reconstituted enzyme binds and cleaves the DNA in a promiscuous manner in Mg 2ϩ -catalyzed reactions, similar to the native enzyme. To investigate whether the loss of DNA binding and cleavage in Zn 2ϩ -demetalated R.KpnI is due to the structural alterations in the protein, we carried out CD analysis. In presence of Zn 2ϩ , the far-UV CD spectrum of R.KpnI has two negative maxima at 208 and 222 nm, which is a characteristic of helical conformation. The zinc-demetalated enzyme showed altered secondary structure compared with zinc-re-constituted enzyme, indicating the importance of Zn 2ϩ coordination to maintain the secondary structure of the R.KpnI (Fig. 3A). The stability of these proteins was monitored by CD thermal denaturation. Unfolding profiles were measured at 222 nm, from 30 to 75°C. The T m of zinc-reconstituted enzyme was increased by ϳ8°C over the zincdemetalated enzyme, indicating a role for Zn 2ϩ in stability of the enzyme (Fig. 3B). In accordance with CD spectroscopy and thermal melting experiments, proteolytic experiments also showed that the zinc-demetalated enzyme is more susceptible to trypsin cleavage than the native or zinc-reconstituted enzyme (Fig. 3C). We conclude that Zn 2ϩ is required for stabilization of the enzymatically active R.KpnI conformation.
The CCCH Motif Is Involved in Zn 2ϩ Coordination and Maintenance of the R.KpnI Structure-To establish that the zinc binding is through the unusual zinc finger motif shown in Fig. 1A, point mutations were generated in R.KpnI. The cysteines (Cys 119 , Cys 128 , and Cys 171 ) and histidine (His 174 ) of the putative motif were individually changed into alanine by sitedirected mutagenesis. The mutant proteins were analyzed for radioactive zinc binding using a zinc blotting assay. All of the alanine replacement mutants failed to bind radioactive zinc (Fig. 4A) in contrast to R.KpnI.
Zn 2ϩ has been shown to be essential for the folding and stability of many Zn 2ϩ finger proteins. Zinc blotting experiments indicate that Cys 119 , Cys 128 , Cys 171 , and His 174 are responsible for coordinating Zn 2ϩ in R.KpnI. We examined the effect of impairment in Zn 2ϩ coordination on the stability of the enzyme by monitoring the CD thermal melting curves of the alanine replacement mutant proteins. The normalized CD absorbance at 222 nm as a function of temperature for WT and mutants is shown in Fig. 4B. The mutant proteins showed decreased thermal stability compared with that of the WT R.KpnI, indicating that the mutations at the CCCH motif affect the folding of the enzyme. Further, the mutant enzymes showed increased protease susceptibility compared with R.KpnI, confirming the importance of the CCCH motif for the structural stability of the enzyme (Fig. 4C).
Effect of CCCH Zn 2ϩ Finger Mutations on DNA Cleavage and Binding-R.KpnI and its mutants were analyzed for the ability to cleave DNA. Under the assay conditions, wherein pUC18 DNA was completely cleaved by 1 nM WT enzyme, there was no cleavage product observed with all the four mutants even at a 100-fold excess of the enzyme (Fig. 4D). To address the cause for the loss of DNA cleavage observed with the mutants, we analyzed the DNA binding ability of the mutants by electrophoretic mobility shift assay. The mutants failed to bind the DNA containing R.KpnI recognition sequence (Fig. 4E). The mutants showed no detectable DNA binding and cleavage, due to the loss of the structure as observed in CD thermal melting  and proteolytic experiments. These results suggest that the loss of coordination with zinc affected the structural integrity of the protein, concomitantly affecting the activity of the enzyme. The loss of DNA binding and cleavage seen with single amino acid substitution in R.KpnI is a typical characteristic of Zn 2ϩ finger proteins where Zn 2ϩ has a structural role.
Specific DNA Cleavage and Suppression of Mg 2ϩ -and Mn 2ϩinduced Promiscuous Activity-The atomic absorption spectroscopy analysis of zinc-reconstituted R.KpnI showed that the enzyme binds 4 mol of zinc (Table 1). Among the 4 mol, only 2 mol can be readily replaced with Mg 2ϩ . This hints at the possibility of zinc ions binding to the active site to influence the enzymatic properties of R.KpnI in addition to the tight coordination at the CCCH motif. The additional 2 mol of Zn 2ϩ bound replacing the Mg 2ϩ may inhibit DNA cleavage. Surprisingly, the enzyme showed efficient DNA cleavage in the presence of 50 M Zn 2ϩ (Fig. 5A). To evaluate the role of Zn 2ϩ in the specificity of R.KpnI, we carried out DNA cleavage experiments at higher enzyme concentrations (50 -1000 nM) and in the presence of 100 M Zn 2ϩ . Even at such high concentrations of the enzyme, promiscuous cleavage is not detected, unlike in Mg 2ϩ -catalyzed reactions (Fig. 5B). In experiments using one of the noncanonical oligonucleotides (GtTACC) as a substrate in the presence of 2 mM Mg 2ϩ or 100 M Zn 2ϩ , the cleavage was observed only with Mg 2ϩ , indicating that the enzyme is highly specific in the presence of Zn 2ϩ (Fig. 5C). No detectable DNA cleavage was observed in the presence of Zn 2ϩ with any of the other noncanonical substrates.
In the qualitative experiments described above, Zn 2ϩ -mediated enzyme activity appeared to be comparable with the activity in the presence of other metal ions. We resorted to kinetic analysis to obtain quantitative information about Zn 2ϩ DNA cleavage. Kinetic analysis in the presence of Zn 2ϩ revealed the turnover number (k cat ) of the enzyme to be 2.12 min Ϫ1 , which is comparable with that of Ca 2ϩ (2.20 min Ϫ1 ), showing that Zn 2ϩ -mediated DNA cleavage is as efficient as Ca 2ϩ -dependent cleavage (  (Fig. 5D)

DISCUSSION
Zinc finger proteins are involved in fundamental cellular processes (viz. replication, transcription, repair, translation, and programmed cell death) (7). Zinc finger motifs have also been discovered and implicated in maintenance of the structural architecture in a number of nucleases (9,14,26). We demonstrate that a sequence motif ( 119 CX 8 CX 42 CX 2 H 174 ) found in R.KpnI is a zinc binding motif. Based on the conserved arrangements of cysteines and/or histidines, several classes of zinc fin-  Table 1) as described under "Experimental Procedures." ger families (CCHH, CCCC, CCHC, and CHCC), have been characterized and shown to be involved in interactions with DNA, RNA, or other proteins (27). CCCH-type zinc fingers were identified in a number of RNA-binding proteins of eukaryotic origin (28) and also found in Mcm10 protein, which is essential for the formation of active homocomplex (29). The arrangement of three cysteines and a histidine in the CCCH zinc finger found in R.KpnI is rare in prokaryotic proteins. The CCCH zinc fingers were identified in replication protein A homologues in different lineages of Euryarcheaota (30), RPA41 from Pyrococcus furiosus (31) and 50 S ribosomal protein L36 from Thermus thermophilus (32).
R.KpnI thus is a new member of the CCCH Zn 2ϩ finger family and is the first REase to have this motif. The Zn 2ϩ fingers in other members having this motif listed above have varied roles in protein-protein interactions, protein-nucleic acid interactions, structural integrity, and folding. From the results presented in Fig. 2, it is clear that tightly bound Zn 2ϩ has a structural role, since it supports Mg 2ϩ -mediated promiscuous cleavage. The second Zn 2ϩ atom, which is loosely bound to the active site, imparts catalytic function. A peculiar characteristic of R.KpnI is its highly promiscuous behavior in the presence of Mg 2ϩ not seen with any other REase (21). The Mg 2ϩ (and Mn 2ϩ )-mediated promiscuous cleavage by R.KpnI is completely suppressed by Zn 2ϩ meanwhile, inducing the high fidelity cleavage. The architectural plasticity of the R.KpnI active site allows the binding of Mg 2ϩ , Mn 2ϩ , Ca 2ϩ , or Zn 2ϩ , which have different coordination chemistry and geometry to induce promiscuous or specific cleavage. Thus, in R.KpnI, Zn 2ϩ has both structural and catalytic roles, together not found in any enzyme so far. The tightly bound Zn 2ϩ at the CCCH motif imparts structural integrity for the enzyme, whereas the readily exchangeable Zn 2ϩ at the active site induces high specificity cleavage.
Finally, Zn 2ϩ finger motifs as such appear to be extremely rare in nucleases of prokaryotic origin. Although the HNH motif is commonly found in diverse classes of nucleases, the zinc finger motifs are found only in McrA, I-PpoI, and T4 endonuclease VII belonging to the superfamily. Although the Cys 4 zinc finger of T4 endonuclease VII has a structural role, the function of similar zinc finger in McrA is not known. The two Zn 2ϩ fingers (CCCH and CCHC) in I-PpoI are also important for structural stabilization of the protein core. The catalytic and structural role for Zn 2ϩ in R.KpnI hints at its distant origin and possibly additional yet unknown function in Klebsiella pneumoniae.