
==== Front
Nucleic Acids ResNucleic Acids ResnarNucleic Acids ResearchNucleic Acids Research0305-10481362-4962Oxford University Press 1743097110.1093/nar/gkm114Molecular BiologyR.KpnI, an HNH superfamily REase, exhibits differential discrimination at non-canonical sequences in the presence of Ca2+ and Mg2+ Saravanan Matheshwaran 1Vasu Kommireddy 1Kanakaraj Radhakrishnan 1Rao Desirazu N. 2Nagaraja Valakunja 13*1Department of Microbiology and Cell Biology, 2Department of Biochemistry, Indian Institute of Science, Bangalore 560 012 and 3Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore 560 064, India*To whom correspondence should be addressed +91-80-2360066891-80-23602697vraj@mcbl.iisc.ernet.in4 2007 11 4 2007 11 4 2007 35 8 2777 2786 14 1 2007 4 2 2007 8 2 2007 © 2007 The Author(s)2007This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.KpnI REase recognizes palindromic sequence, GGTAC↓C, and forms complex in the absence of divalent metal ions, but requires the ions for DNA cleavage. Unlike most other REases, R.KpnI shows promiscuous DNA cleavage in the presence of Mg2+. Surprisingly, Ca2+ suppresses the Mg2+-mediated promiscuous activity and induces high fidelity cleavage. To further analyze these unique features of the enzyme, we have carried out DNA binding and kinetic analysis. The metal ions which exhibit disparate pattern of DNA cleavage have no role in DNA recognition. The enzyme binds to both canonical and non-canonical DNA with comparable affinity irrespective of the metal ions used. Further, Ca2+-imparted exquisite specificity of the enzyme is at the level of DNA cleavage and not at the binding step. With the canonical oligonucleotides, the cleavage rate of the enzyme was comparable for both Mg2+- and Mn2+-mediated reactions and was about three times slower with Ca2+. The enzyme discriminates non-canonical sequences poorly from the canonical sequence in Mg2+-mediated reactions unlike any other Type II REases, accounting for the promiscuous behavior. R.KpnI, thus displays properties akin to that of typical Type II REases and also endonucleases with degenerate specificity in its DNA recognition and cleavage properties.
==== Body
INTRODUCTION
Type II REases are homodimers, which recognize short palindromic sequences of 4–8 bp in length and in the presence of Mg2+, cleave the DNA within the recognition site to generate 5′-phosphate and 3′-OH ends (1). Type II REases have been broadly classified into two groups, based on structural architecture and mode of DNA recognition (2,3). Further, EcoRI and many other enzymes bind to DNA preferentially at their recognition sites in the absence of divalent metal ions (4,5). In contrast, EcoRV, TaqI and MunI bind both specific and non-specific sequences with similar affinities in the absence of divalent metal ions. With the latter set of enzymes, binding to the canonical sequence is increased by about 1000-fold in the presence of divalent metal ions (6–8). Irrespective of this difference in binding characteristics, both the groups of enzymes need divalent metal ion, generally Mg2+, for catalysis.

Type II REases exhibit a very high degree of site specificity towards their recognition sequences over non-canonical sequences and non-specific sequences. However, many of them exhibit relaxed specificity under sub-optimal conditions (4,9). R.KpnI isolated from Klebsiella pneumoniae recognizes the palindromic DNA sequence 5′- GGTAC↓C –3′ and cleaves DNA generating four base 3′ overhangs (10,11). The enzyme exhibits promiscuous DNA cleavage in the presence of Mg2+ and attains exquisite sequence specificity in the presence of Ca2+ (12). In addition to imparting specific DNA cleavage, Ca2+ suppresses Mg2+- or Mn2+-mediated relaxed specificity of the enzyme (12).

Type II REases differ in their amino acid sequence from one another, and indeed from all other protein families (13). Recent structural and bioinformatic analysis of Type II REases revealed that they possess common structural fold around and comprising the active site (14). A vast majority of Type II REases utilize PD … D/EXK motif as a catalytic motif, wherein the carboxylic side chains of the acidic amino acids play a role in metal ion coordination and the bound water is involved in phosphodiester bond hydrolysis by acid–base mechanism (4). Unlike most other REases, R.KpnI utilizes HNH motif as a catalytic motif and follows the mechanism of ββα-Me finger nucleases, such as colicins E7, E9, Vvn and Serratia nuclease (15). These enzymes recognize and cleave degenerate or non-specific DNA sequences and employ a histidine residue to activate the attacking water molecule (16–18).

Although a number of REases have been characterized, kinetic analysis has been carried out only with few of the enzymes. Amongst these, RsrI, EcoRI, EcoRV, BamHI, MunI and PvuII, have been subjected to detailed kinetic studies (19–26). All these extensively studied enzymes belong to PD … D/EXK family, yet seem to have varied kinetic properties on canonical, non-canonical and non-specific sequences. Owing to its high degree of promiscuous activity, specific cleavage in the presence of Ca2+, suppression of Mg2+- and Mn2+-mediated promiscuity by Ca2+ and use of HNH motif instead of PD … D/EXK for catalysis (15), R.KpnI is markedly different from other Type II REases subjected to kinetic analysis so far. Here, we have investigated the DNA binding and catalytic behavior of R.KpnI and describe features underlying the metal ion dependence during specific and promiscuous DNA cleavage reactions.

MATERIALS AND METHODS
Enzymes and chemicals
Oligonucleotides used in the study are listed in Table 1. Oligonucleotides (Microsynth and Sigma) purified on 15% urea-polyacrylamide gel were 5′ end-labelled using T4 polynucleotide kinase and [γ-32P] ATP (6000 Ci/mmol) and further purified using G-50 spin column chromatography.
Table 1. Duplex DNA oligonucleotides used in this study

Duplex I	5′-ATTGCGTGGTACCCGCTCTT-3′ 3′-TAACGCACCATGGGCGAGAA-5′	
Duplex II	5′-ATTGCGTtGTACCCGCTCTT-3′ 3′-TAACGTaCATGGGGCGAGAA-5′	
Duplex III	5′-ATTGCGTGaTACCCGCTCTT-3′ 3′-TAACGTCtATGGGGCGAGAA-5′	
Duplex IV	5′-ATTGCGTGtTACCCGCTCTT-3′ 3′-TAACGCACaATGGGCGAGAA-5′	
Duplex V	5′-ATTGCGTGGaACCCGCTCTT-3′ 3′-TAACGCACCtTGGGCGAGAA-5′	
Duplex VI	5′-ATTGCGTGGTtCCCGCTCTT-3′ 3′-TAACGCACCAaGGGCGAGAA-5′	
Duplex VII	5′-ATTGCGTGGTAtCCGCTCTT-3′ 3′-TAACGCACCATaGGCGAGAA-5′	
Duplex VIII	5′-ATTGCGTGGTACtCGCTCTT-3′ 3′-TAACGCACCATGaGCGAGAA-5′	


T4 polynucleotide kinase (New England Biolabs), [γ 32P]-ATP, Hi-Trap Heparin column (Amersham Biochemicals), ampicillin, chloramphenicol, bovine serum albumin, polyethyleneimine, Commassie Brilliant Blue, and IPTG (Sigma), Phosphocellulose P11 (Whatman) were used in the study.

Expression and purification of R.KpnI
R.KpnI was expressed in E. coli BL26 [F−
omp T hsdSB 
gal dcm Δlac (DE3) nin5 lacUV5-T7 gene 1] cells in the presence of pACMK (plasmid-expressing kpnI methylase gene) (27). Escherichia coli BL26 cells harboring R.KpnI (pETRK) clone were grown at 37°C in LB medium containing 100 µg/ml ampicillin and 20 µg/ml chloramphenicol until the absorbance at 600 nm reached 0.6 and induced with 0.3 mM IPTG. The protein was purified to near homogeneity as described previously (27). Protein concentrations were estimated by the Bradford method using bovine serum albumin as the standard.

Electrophoretic mobility shift assay
Different concentrations of R.KpnI or its mutant H149A (0.5–256 nM) were incubated with 1 nM of end-labeled double-stranded oligonucleotides containing canonical (duplex I) and non-canonical sites (duplex II to VIII) in the binding buffer [20 mM Tris-HCl (pH 7.4), 35 mM NaCl and 5 mM 2-mercaptoethanol] on ice for 15 min. The gels were electrophoresed at 4°C and 100 V for 2 h. The amounts of DNA in free and bound forms were quantitated using Phosphorimager Image Gauge software Version 2.54. The assays were repeated thrice and the dissociation constants were determined by fitting these data to a standard hyperbolic curve using Graphpad Prism software Version 4. For EMSA in the presence of metal ions, the binding buffer contained 20 mM Tris-HCl (pH 7.4), 5 mM 2-mercaptoethanol and either 5 mM CaCl2 or MgCl2. The gel running buffer (pH 8.2) contained 89 mM Tris-boric acid and either 2 mM CaCl2 or MgCl2. All other reaction conditions for electrophoresis and analysis were identical to those performed in the absence of divalent metal ions.

Steady-state kinetic assays
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 and non-canonical DNA substrates were measured at DNA concentrations of 5–150-fold molar excess over dimeric enzyme (1 nM). The steady-state reactions were carried out at 37°C in an assay buffer containing 10 mM Tris-HCl, pH 7.4, 35 mM NaCl, 5 mM 2-mercaptoethanol and 5 mM Mg2+ or Mn2+ or Ca2+. Aliquots were taken out at 10 different time points for the determination of initial velocity. The reactions were terminated by adding an equal volume of stop dye containing urea (4 M), 2 mM EDTA, 2 μl formamide, 1 mM NaOH, 0.05% xylene cyanol, 0.05% bromophenol blue. The samples were analyzed on 15% denaturing 8 M urea–polyacrylamide gels. Electrophoresis was carried out at 250 V with 1× Tris-boric acid-EDTA as the running buffer and cleavage products were quantified by phosphorimaging analysis. Steady-state kinetic parameters were determined by fitting the change in the velocity with substrate concentration to the double reciprocal (1/v versus 1/[S]) Lineweaver–Burk plot using Graphpad Prism 4. The turnover number (kcat) was calculated as the ratio of Vmax to the enzyme concentration used (28).

Preincubation studies
Preincubation studies were carried out by initiating the reactions with metal ion (5 mM) or oligonucleotide substrates (2.5 nM) or R.KpnI (0.5 nM) at 15°C. Samples were removed after fixed time intervals and analyzed as described for cleavage assays.

Intrinsic fluorescence
Fluorescence emission spectra were collected on a Jobin-Yvon fluorometer FluoroMax 3 (HORIBA, Jobin-Yvon/Spex Division, Longjumeau, France), thermostated at 25°C. EDTA-treated wt R.KpnI or H149A (2 µM concentration each) in a buffer containing 20 mM Tris-HCl pH (7.4) and 5 mM 2-mercaptoethanol were incubated with different concentrations of metal ions for 15 min at 25°C. All samples were incubated until the equilibrium was established under a particular set of conditions before measuring the fluorescence intensities. All fluorescence emission spectra and fluorescence intensities from titrations were corrected for protein tryptophan fluorescence by subtraction of control spectra and control titrations. Fluorescence intensities were plotted against the metal ion concentrations, and the data were analyzed according to the modified Stern–Volmer equation (29).
  
where F0 and F are fluorescent intensities in the absence and presence of cofactor respectively, Q is the quencher (metal ion) concentration; fa is the fractional number of fluorophores, KQ is the quenching constant. The dissociation constants were calculated from the modified Stern–Volmer plot (a plot of F0/(F0 − F) versus 1/[Q]), where KQ = 1/Kd (29).

RESULTS AND DISCUSSION
Steady-state kinetic analysis of R.KpnI in the presence of Mg2+, Mn2+and Ca2+
Previously, we demonstrated that Ca2+, in addition to Mg2+ and Mn2+, supported the R.KpnI endonuclease cleavage activity (12). Unlike many other REases, the enzyme showed high promiscuous activity in the presence of Mg2+ and this behavior was compleletely suppressed with Ca2+ resulting in high specificity cleavage (12). For determining the kinetic parameters of R.KpnI under steady-state reaction conditions, end- labeled 20-mer duplex I containing the canonical site (GGTACC) was used as a substrate. Steady-state kinetic reactions were carried out under conditions of 5–150-fold molar excess of substrates. In a plot of product formed versus time shown in Figure 1A, all time courses displayed biphasic kinetics. The steady-state kinetics shown in Figure 1A is consistent with a product burst. The burst amplitude is proportional to the enzyme concentration (Figure 1B). These data would imply that the enzyme performs the first turnover fast and the slow phase corresponds to the product release. Thus, the kcat values most likely reflect the product release rates. Analysis of this data in the steady- state phase showed that the enzyme obeys the Michaelis–Menten kinetics (Figure 1C). Kinetic constants KM and Vmax were obtained from Lineweaver–Burk plot (Figure 1D). Amongst all REases subjected to steady-state kinetic analysis, R.KpnI has higher kcat in the presence of Mn2+ compared to that of Mg2+. The specificity constant (kcat/KM) for the enzyme did not vary significantly in the presence of Mg2+ and Mn2+; the values were 3-fold less when Ca2+ was used as a cofactor (Table 2). Under these steady- state conditions, the turnover number of the enzyme is comparable with Mg2+ and Mn2+ as cofactors, and only 2-fold lower for Ca2+. These features are unique to R.KpnI. In contrast, most other Type II REases exhibit no cleavage with Ca2+ and reduced DNA cleavage at the canonical site with Mn2+ (30). For instance, EcoRV has 15-fold higher kcat/KM value for canonical duplex with Mg2+ when compared to that of Mn2+ (22). MunI and EcoRI also showed reduced DNA cleavage with Mn2+ as cofactor (8,31).
Table 2. Kinetic constants of R.KpnI for canonical and non-canonical substrates

DNA Substrates	KM (M)	kcat(s−1)	kcat/KM(M−1 s−1)	Discrimination factor*	
Duplex I (-GGTACC-)					
    Mg2+ as cofactor	22 (±2.6) × 10−9	0.072 ± 0.002	3.2 (±0.3) × 106		
    Mn2+ as cofactor	21 (±3.4) × 10−9	0.077 ± 0.003	3.6 (±0.4) × 106		
    Ca2+ as cofactor	36 (±2.8) × 10−9	0.037 ± 0.003	1.0 (±0.1) × 106		
Duplex II (-tGTACC)					
    Mg2+ as cofactor	221 (±10.6) × 10−9	0.00029 ± 0.0001	1.3 (±0.06) × 103	2500	
    Mn2+ as cofactor	198 (±8.4) × 10−9	0.00081 ± 0.0001	4.1 (±0.50) × 103	900	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex III (-GaTACC)					
    Mg2+ as cofactor	69 (±9.3) × 10−9	0.0026 ± 0.0002	3.76 (±0.2) × 104	90	
    Mn2+ as cofactor	112 (±7.4) × 10−9	0.013 ± 0.002	1.16 (±0.1) × 105	30	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex IV (-GtTACC)					
    Mg2+ as cofactor	77 (±1.6) × 10−9	0.0028 ± 0.0001	3.6 (±0.02) × 104	90	
    Mn2+ as cofactor	123 (±2.7) × 10−9	0. 012 ± 0.0016	1.0 (±0.06) × 105	40	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex V (-GGaACC)					
    Mg2+ as cofactor	210 (±4.3) × 10−9	0.00068 ± 0.00004	3.2 (±0.2) × 103	1000	
    Mn2+ as cofactor	168 (±8.4) × 10−9	0.00032 ± 0.00002	1.9 (±0.1) × 104	200	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex VI (-GGTtCC)					
    Mg2+ as cofactor	164 (±3.1) × 10−9	0.0027 ± 0.0005	1.64 (±0.5) × 104	200	
    Mn2+ as cofactor	103 (±4.1) × 10−9	0.013 ± 0.003	1.26 (±0.3) × 105	30	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex VII (-GGTAtC)					
    Mg2+ as cofactor	174 (±2.8) × 10−9	0.0010 ± 0.0001	5.7 (±0.9) × 103	550	
    Mn2+ as cofactor	135 (±3.2) × 10−9	0.0034 ± 0.00012	2.56 (±0.1) × 104	140	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
Duplex VIII (-GGTACt)					
    Mg2+ as cofactor	199 (±3.0) × 10−9	0.00091 ± 0.00004	4.6 (±0.2) × 103	700	
    Mn2+ as cofactor	156 (±2.6) × 10−9	0.0043 ± 0.00014	2.7 (±0.2) × 104	140	
    Ca2+ as cofactor	ndc	ndc	ndc	–	
*Discrimination factor = [kCat/KM(canonical substrate)]/[kCat/KM(non-canonical substrate)].

ndc: no detectable cleavage.


Figure 1. Steady-state kinetic analysis of R.KpnI with its canonical substrate in the presence of Mg2+. (A) R.KpnI (1 nM) was incubated with increasing concentrations (5–150 nM) of canonical duplex I in the reaction buffer. Reactions were initiated by the addition of 5 mM of Mg2+ as described in the Materials and Methods section. The reactions were stopped by adding a stop dye containing 10 mM EDTA. The samples were analyzed on a 15% denaturing 8 M urea-polyacrylamide gel. After electrophoresis, the gels were dried and visualized by phosphorimager and the intensity of the bands was quantified. (B) Rate of DNA cleavage with different concentrations of R.KpnI. 150 nM of duplex I was incubated with increasing concentrations of R.KpnI (0.5–10 nM) in the reaction buffer. The conditions of DNA cleavage reaction are described in Figure 1A. (C) A representative Michaelis–Menten plot. The graphs were obtained by plotting the rates of DNA cleavage at different concentrations of oligonucleotide substrate using Graphpad Prism software Version 4. (D) A representative Lineweaver–Burk plot from which kinetic constants were obtained.



Steady-state kinetic analysis at non-canonical sites
Next, we tested the DNA cleavage of non-canonical substrates in the presence of different metal ions. For carrying out kinetic studies with DNA containing the promiscuous sites, 20-mer double-stranded oligonucleotides (duplex II to VIII containing substitutions at each position of the hexanucleotide recognition sequence, Table 1) were chosen as substrates. DNA cleavage was not detectable with any of the non-canonical sites when reactions were carried out with Ca2+ as a cofactor even at high concentrations of the enzyme and prolonged incubation periods. However, in the presence of Mg2+ and Mn2+, all the non-canonical substrates were cleaved, although the extent of cleavage was varied. The sequences GtTACC, GaTACC, GGTtCC and GGTAtC were cleaved better than the sequences tGTACC, GGaACC and GGTACt. The kinetic parameters of the detailed analysis are summarized in Table 2. With Mg2+ and Mn2+ as cofactors, R.KpnI discriminates the non-canonical sites by 90–2500 times and 30–900 times respectively (Table 2). In contrast, all other REases studied so far exhibit a larger degree of discrimination in the presence of Mg2+, when recognizing the non-canonical sequences. For instance, with its natural cofactor Mg2+, and under optimal reaction conditions, EcoRI and EcoRV cleave their respective recognition sites 105 and 106 times faster than sites differing by one base pair (20,32). In Mn2+ -catalyzed reactions, however, the discrimination of EcoRV, EcoRI and TaqI endonucleases against non-canonical sequences was greatly reduced (22,33,34). From the kinetic constants obtained for R.KpnI, it is evident that R.KpnI exhibits very low degree of discrimination during Mg2+-mediated cleavage, accounting for its highly promiscuous behavior (Table 2).

Basis of discrimination
A high degree of sequence specificity is achieved by Type II REases as a result of their discrimination of the non-canonical sequences from the canonical sequence either during binding or cleavage or at both the steps (4,20,35). The different metal ions may have a varied influence on the enzyme binding to canonical and non-canonical DNA sequences. Alternatively, the differences in the metal ion binding to the enzyme active site could account for the property. The results presented in Figure 2 demonstrate that all the three metal ions use the same active site characterized earlier (15). A mutant R.KpnI, in which invariant H149 of the HNH motif is replaced by alanine, did not show any cleavage irrespective of the metal ions used, indicating that the same catalytic scheme is employed in all the cases. Similar metal-ion-binding pattern by the wt and H149A mutant suggests that H149 is not implicated in metal ion binding. Fluorescence spectroscopy was used to evaluate binding of the metal ions to the enzyme, taking advantage of the presence of three tryptophans in the protein. The change in tryptophan fluorescence intensity was measured as a function of the metal ion concentration. The results presented in Figure 3A show representative spectra with Mg2+ as cofactor. With increasing concentrations of the metal ion, fluorescence quenching was observed. Mn2+ and Ca2+ also showed similar concentration-dependent fluorescence quenching. However, monovalent cations such as Na+ did not influence the fluorescence spectra of the enzyme (data not shown). A modified Stern–Volmer plot was used to analyze the quenching (Figure 3B). The linearity of the plot suggests that Mg2+ binding is a dominant fluorescence-quenching phenomenon over the range of concentrations used (0.1–10 mM). The KD values calculated from modified Stern–Volmer relationship are 550 µM for Mg2+, 200 µM for Mn2+ and 700 µM for Ca2+. The data presented in Figure 3C and D show comparable binding of all the three metal ions to both wt and active site mutant H149A. These results indicate that the metal binding to the enzyme by itself may not account for the lower discrimination exhibited by the enzyme.
Figure 2. DNA cleavage activity of H149A mutant R.KpnI in the presence of different metal ions. End-labeled duplex I was incubated with R.KpnI (1 nM) and H149A (10, 100 nM) in the presence of 5 mM of Mg2+, Mn2+ and Ca2+ as indicated. In all the cases, the reactions were incubated at 37°C for 30 min, and the products were analyzed on 12% urea–polyacrylamide gel.


Figure 3. Fluorescence emission spectra of wt and mutant R.KpnI. Fluorescence emission spectra were recorded as described under the Materials and Methods section. (A) Background-corrected fluorescence emission spectra of R.KpnI (20 μM) with increasing amounts of MgCl2 (0–10 000 μM) in binding buffer (50 mM Tris-HCl pH 7.4 and 50 mM NaCl,) (B) Modified Stern–Volmer plot for Mg2+ quenching. Fluorescence changes were calculated from the maximal intensity of enzyme alone (before addition of cofactor), F0, and the maximal intensity after the addition of Mg2+, F (0.1–10 mM). (C) wt R.KpnI and (D) mutant H149A in the presence of 1 mM Mg2+, Mn2+ or Ca2+ respectively.



In order to understand the role of divalent metal ions in DNA binding, we investigated the binding affinity of R.KpnI using the wt and H149A mutant. Initially, we carried out the DNA-binding analysis in the absence of any metal ion cofactors. Both the wt and the H149A mutant enzymes bound with similar affinity to the canonical site (duplex I) as well as non-canonical sites (duplex II to VIII) (Table 3). Next, we investigated the DNA-binding properties of the mutant H149A in the presence of Mg2+ or Mn2+ or Ca2+ using the canonical and non-canonical sequences. The binding affinity of the enzyme to the DNA was comparable with all the three metal ions. More importantly, the metal ions did not enhance the binding of the enzyme to the recognition sequence (Table 3).
Table 3. DNA-binding affinities of wt R.KpnI and mutant H149A REases

Enzyme	Kd (nM)			
R.KpnI (absence of metal ions)				
    Duplex I (-GGTACC-)	9 ± 2.3			
    Duplex II (-tGGACC-)	8 ± 2.4			
    Duplex III (-GaTACC-)	13 ± 1.7			
    Duplex IV (-GtTACC-)	12 ± 0.7			
    Duplex V (-GGaACC-)	16 ± 0.3			
    Duplex VI (-GGTtCC-)	16 ± 2.1			
    Duplex VII (-GGTAtC-)	29 ± 3.2			
    Duplex VIII (-GGTACt-)	28 ± 0.7			
H149A (absence of metal ions)				
    Duplex I (-GGTACC-)	8 ± 1.6			
    Duplex II (-tGGACC-)	18 ± 1.4			
    Duplex III (-GaTACC-)	14 ± 0.7			
    Duplex IV(-GtTACC-)	13 ± 0.4			
    Duplex V (-GGaACC-)	28 ± 1.8			
    Duplex VI (-GGTtCC-)	23 ± 0.7			
    Duplex VII (-GGTAtC-)	19 ± 3.2			
    Duplex VIII (-GGTACt-)	26 ± 2.7			
H149A (presence of metal ions)	Mg2+	Mn2+	Ca2+	
    Duplex I (-GGTACC-)	8 ± 2.4	9 ± 1.4	7 ± 1.3	
    Duplex II (-tGGTACC-)	11 ± 1.4	12 ± 3.1	21 ± 4.3	
    Duplex III (-GaTACC-)	12 ± 1.7	12 ± 2.1	21 ± 4.3	
    Duplex IV (-GtTACC-)	11 ± 1.6	12 ± 0.8	11 ± 3.1	
    Duplex V (-GGaACC-)	21 ± 6.4	28 ± 3.3	19 ± 1.9	
    Duplex VI (-GGTtCC-)	19 ± 3.9	17 ± 1.7	10 ± 2.6	
    Duplex VII (-GGTAtC-)	21 ± 4.4	27 ± 3.4	25 ± 3.4	
    Duplex VIII (-GGaACt-)	19 ± 1.6	17 ± 3.4	15 ± 2.8	


The difference in the ability of the enzyme to cleave non-canonical sequences in Mg2+- versus Ca2+-mediated reactions allowed us to explore the mechanism of Ca2+-specific discrimination of non-canonical sequence for DNA cleavage. From the DNA-binding studies presented in Figure 4A and B, it is evident that the enzyme binds to the non-canonical sites with comparable affinity to that of canonical sequence both in the presence and absence of Ca2+(Table 3). This was rather surprising as these sequences are refractile to Ca2+-mediated DNA cleavage (Table 2). The mutant H149A exhibits similar affinity with both canonical and non-canonical substrates in the presence of Ca2+ (Figure 4C, D and E). From these studies, it is clear that a Ca2+-mediated higher degree of discrimination by the enzyme is exerted only at the step of DNA cleavage and not at the step of binding.
Figure 4. Electrophoretic mobility shift analyses for the determination of the binding affinity of the wt R.KpnI and mutant H149A in the presence of Ca2+. (A) and (B) Gel shift showing the native gel (inset) with duplex II and III. The wt R.KpnI was titrated over the concentration range from 0 to 256 nM, and the concentration of the DNA was 25 pM. The intensity of the shifted DNA band are expressed as% bound. (C), (D) and (E) are the gel shifts showing the native gels (inset) with duplex I, II and III respectively using mutant H149A. The average values of the three different experiments were plotted in the graph.



Preincubation studies
While the above experiments revealed the basis for differential discrimination imposed by the metal ions for DNA cleavage, it is not yet clear how these differences are brought about by the enzyme using the same active site. A possible reason could be that the enzyme follows altered reaction path(s) in the presence of different metal ions. The order of binding to substrate could be varying with each metal ion. In order to understand the mechanistic details and the order of events involved in DNA cleavage reaction, we have carried out preincubation studies. DNA cleavage was initiated either by the addition of R.KpnI to preincubated DNA and Mg2+ or addition of DNA to R.KpnI preincubated with Mg2+ or by the addition of Mg2+ to preincubated DNA-R.KpnI complexes. Similar experiments were carried out with Mn2+ and Ca2+ as cofactors instead of Mg2+. The order of preincubation had significant influence on the initial rate of the product formation. Most of the DNA was cleaved within the initial time point in the case of metal-ion-initiated reactions (Figure 5). On the other hand, as can be seen from Figure 5A, at every time point between 30 and 120 s, 2-fold lower rates were observed with the reactions initiated with DNA or the enzyme, implying that the pre-formed enzyme–DNA complex could be catalytically more efficient than the pre-formed enzyme–metal ion complex. However, after 10 min, the extent of cleavage was comparable in all the three cases. From these results, it appears that R.KpnI follows a rather random order in substrate and cofactor binding, with a preference for binding the DNA substrate first.
Figure 5. Preincubation studies of R.KpnI by using canonical substrate. Preincubation studies were carried out by initiating the reaction either by the addition of metal ions or DNA or enzyme. (A), (B) and (C) are graphical representations of preincubation experiments results carried out in the presence of Mg2+, Mn2+ and Ca2+ respectively.



A hallmark characteristic of Type II REases is their exquisite sequence specificity (36). R.KpnI, in contrast, exhibits high degree of promiscuous DNA cleavage in the presence of Mg2+ and Mn2+. The kinetic parameters of R.KpnI with canonical and non-canonical sequences suggest that the enzyme is not able to discriminate non-canonical sequences effectively from the canonical sequence. In Mg2+-catalyzed reactions, R.KpnI exhibits comparable kcat value (4 min−1) for canonical substrate as that of other REases, such as RsrI, EcoRV, EcoRI and BamHI (4.2, 8.4, 2.4 and 20.8 min−1 respectively) (19,22,31,37). However, in contrast to those enzymes, it exhibits very low degree of discrimination at non-canonical sites (Table 2). Our earlier analysis showed that R.KpnI belongs to the HNH endonuclease superfamily (15) and follows the mechanism similar to that of I-PpoI and I-CmoeI in that all of them utilize an invariant histidine residue as a general base during catalysis (38,39). I-PpoI is capable of cleaving sequences with single, double or multiple substitutions within the homing site (40,41). Kinetic analysis with I-CmoeI and I-PpoI showed that these enzymes exhibit kcat values of 0.26 and 0.046 min−1 respectively. These values are 20–100 times lower than that of R.KpnI (4 min−1). The specificity constant value of R.KpnI (3.1 × 106 M−1 s−1) is 14 and 24 times lower than that of I-CmeoI (4.3 × 107 M−1 s−1) and I-PpoI (7.5 × 107 M−1 s−1) respectively. The higher values with the latter enzymes could be due to their binding to the long, extended DNA substrate with high affinity (39,42). Thus, as in the case of PD … D/EXK family enzymes, the members of HNH superfamily exhibit variations in kinetic properties. The present studies reveal that the lower degree of discrimination between canonical and non-canonical sequence for DNA binding by R.KpnI largely accounts for its higher promiscuity in Mg2+-catalyzed reactions.

From the vast body of literature and this article, it is apparent that the nucleases exhibit a wide range in discrimination of the non-canonical sequences, manifested at the level of DNA binding and/or catalysis. Non-specific nucleases would, by and large, be non-discriminatory at the level of binding and catalysis. In contrast, highly specific REases and enzymes with moderate sequence specificity may discriminate at the level of binding or during catalysis or at both the steps. Kinetically well-characterized REases such as EcoRV and PvuII discriminate at the level of DNA cleavage and not at the level of binding (23,26). In a recent structural and kinetic study, binding of EcoRV to GAATTC (EcoRI recognition site) was about two orders of magnitude lower compared to GATATC (EcoRV recognition site). The enzyme, however, was shown to discriminate GAATTC site for cleavage. The cleavage at this non-canonical sequence was reduced five orders of magnitude (43). EcoRI, on the other hand, discriminates non-canonical sequences at the level of both binding and catalysis (20). In contrast, I-PpoI, a member of HNH superfamily, fails to show appreciable discrimination while binding to homing sites and its variant sequences (40,41). Studies described here reveal that R.KpnI is a poor discriminator while binding to the non-canonical sequences. However, during catalysis, it exhibits remarkably varied property. Although bound to non-canonical sequences, DNA cleavage is not observed in the presence of Ca2+. Further, the extent of sequence discrimination in Mg2+-catalyzed reactions is very low compared to any other site-specific, high-fidelity REases. Whether the low degree of discrimination and high promiscuity of R.KpnI is an intrinsic design of the organism is not yet known, and is an important question that needs to be addressed.

In conclusion, we have carried out analysis of DNA binding and cleavage by R.KpnI, a ββα-Me finger nuclease containing HNH active site. We have demonstrated that the enzyme has comparable specificity constant with Mg2+, Mn2+ and Ca2+, a unique property not seen with other REases. In addition, our study has established an unusually low discriminatory property of R.KpnI in Mg2+-catalyzed reactions, atypical to other REases. Further, the Ca2+-mediated discrimination towards non-canonical sequences is exerted only at the step of cleavage. The presence of HNH catalytic motif and the associated unusual properties of R.KpnI suggest its closer link to the enzymes which recognize degenerate recognition sequences and a distinct evolutionary origin from that of typical Type II REases.

ACKNOWLEDGEMENTS
We thank S.Chandrashekaran for pETRK and pACMK clones, S.Arivudai Nambi for technical assistance, and Molecular Biophysics Unit, Indian Institute of Science, for the fluorescence spectrometry facility. M.S is a recipient of Senior Research Fellowship of Council of Scientific and Industrial Research, Government of India. Funding to pay the Open Access publication charge was provided by institutional support.

Conflict of interest statement. None declared.
==== Refs
REFERENCES
1 Roberts RJ  Halford SE   Linn SM  Lloyd RS  Robert RJ   Type II restriction enzymes Nucleases 1993 Cold Spring Harbor, NY Cold Spring Harbor Laboratory Press 35 88 
2 Anderson JE   Restriction endonucleases and modification methylases Curr. Opin. Struct. Biol. 1993 3 24 30 
3 Bujnicki JM   Phylogeny of the restriction endonuclease-like superfamily inferred from comparison of protein structures J. Mol. Evol. 2000 50 39 44 10654258 
4 Pingoud A  Jeltsch A   Recognition and cleavage of DNA by type II restriction endonucleases Eur. J. Biochem. 1997 246 1 22 9210460 
5 Aiken CR  Fisher EW  Gumport RI   The specific binding, bending, and unwinding of DNA by RsrI endonuclease, an isoschizomer of EcoRI endonuclease J. Biol. Chem. 1991 266 19063 19069 1918025 
6 Vipond IB  Halford SE   Specific DNA recognition by EcoRV restriction endonuclease induced by calcium ions Biochemistry 1995 34 1113 1119 7827059 
7 Zebala JA  Choi J  Barany F   Characterization of steady state, single-turnover, and binding kinetics of the TaqI restriction endonuclease J. Biol. Chem. 1992 267 8097 8105 1569066 
8 Lagunavicius A  Siksnys V   Site-directed mutagenesis of putative active site residues of MunI restriction endonuclease: replacement of catalytically essential carboxylate residues triggers DNA binding specificity Biochemistry 1997 36 11086 11092 9287151 
9 Hsu M  Berg P   Altering the specificity of restriction endonuclease: effect of replacing Mg2+  with Mn2+  Biochemistry 1978 17 131 138 201281 
10 Tomassini J  Roychoudhury R  Wu R  Roberts RJ   Recognition sequence of restriction endonuclease KpnI from Klebsiella pneumoniae  Nucleic Acids Res. 1978 5 4055 4064 82935 
11 Hammond AW  Gerard GF  Chatterjee DK   Cloning the KpnI restriction-modification system in Escherichia coli  Gene 1991 97 97 102 1995432 
12 Chandrashekaran S  Saravanan M  Radha DR  Nagaraja V   Ca2+  mediated site-specific DNA cleavage and suppression of promiscuous activity of KpnI restriction endonuclease J. Biol. Chem. 2004 279 49736 49740 15375161 
13 Bujnicki JM   Pingoud A   Molecular phylogenetics of restriction endonucleases Restriction Endonucleases. 2004 Berlin Springer 63 93 
14 Pingoud A  Jeltsch A   Structure and function of type II restriction endonucleases Nucleic Acids Res. 2001 29 3705 3727 11557805 
15 Saravanan M  Bujnicki JM  Cymerman IA  Rao DN  Nagaraja V   Type II restriction endonuclease R.KpnI is a member of the HNH nuclease superfamily Nucleic Acids Res. 2004 32 6129 6135 15562004 
16 Friedhoff P  Kolmes B  Gimadutdinow O  Wende W  Krause KL  Pingoud A   Analysis of the mechanism of the Serratia  nuclease using site-directed mutagenesis Nucleic Acids Res. 1996 24 2632 2639 8758988 
17 Li C  Hor L  Chang Z  Tsai L  Yang W  Yuan H   DNA binding and cleavage by the periplasmic nuclease Vvn: a novel structure with a known active site EMBO J. 2003 22 4014 4025 12881435 
18 Mate MJ  Kleanthous C   Structure-based analysis of the metal-dependent mechanism of H-N-H endonucleases J. Biol. Chem. 2004 279 34763 34769 15190054 
19 Aiken CR  McLaughlin   Gumport RI   The highly homologous isoschizomers RsrI endonuclease and EcoRI endonuclease do not recognize their target sequence identically J. Biol. Chem. 1991 266 19070 19078 1918026 
20 Thielking V  Alves J  Fliess A  Maass G  Pingoud A   Accuracy of the EcoRI restriction endonuclease: Binding and cleavage studies with oligonucleotide substrates containing degenerate recognition sequences Biochemistry 1990 29 4682 4691 2372551 
21 Alves J  Selent U  Wolfes H   Accuracy of the EcoRV restriction endonuclease: binding and cleavage studies with oligodeoxynucleotide substrates containing degenerate recognition sequences Biochemistry 1995 34 11191 11197 7669777 
22 Sam MD  Horton NC  Nissan TA  Perona JJ   Catalytic efficiency and sequence selectivity of a restriction endonuclease modulated by a distal manganese ion binding site J. Mol. Biol. 2001 306 851 861 11243793 
23 Baldwin GS  Vipond IB  Halford SE   Rapid reaction analysis of the catalytic cycle of the EcoRV restriction endonuclease Biochemistry 1995 34 705 714 7819266 
24 Hensley P  Nardone G  Chirikjian JG  Wastney MF   The time resolved kinetics of superhelical DNA cleavage by BamHI restriction endonuclease J. Biol. Chem. 1990 265 15300 15307 2203774 
25 Sasnauskas G  Jeltsch A  Pingoud A  Siksnys V   Plasmid DNA cleavage by MunI restriction enzyme: single-turnover and steady state kinetic analysis Biochemistry 1999 38 4028 4036 10194315 
26 Nastri HG  Evans PD  Walker IH  Riggs PD   Catalytic and DNA binding properties of PvuII restriction endonuclease mutants J. Biol. Chem. 1997 272 25761 25767 9325303 
27 Chandrashekaran S  Babu P  Nagaraja V   Characterization of DNA binding activities of over-expressed KpnI restriction endonuclease and modification methylase J. Biosci. 1999 24 269 277 
28 Segal IH   Biochemical Calculations 1976 2nd edn New York John Wiley & sons Inc 234 Enzymes 
29 Lehrer SS   Solute perturbation of protein fluorescence. The quenching of the tryptophyl fluorescence of model compounds and of lysozyme by iodide ion Biochemistry 1971 17 3254 3263 5119250 
30 Cowan JA   Pingoud A   Role of metal ions in promoting DNA binding and cleavage by restriction endonucleases Restriction Endonucleases. 2004 Springer, Berlin 339 360 
31 Jeltsch A  Alves J  Oelgeschlager T  Wolfes H  Maass G  Pingoud A   Mutational analysis of the function of Gln115 in the EcoRI restriction endonuclease, a critical amino acid for recognition of the inner thymidine residue in the sequence -GAATTC- and for coupling specific DNA binding to catalysis J. Mol. Biol. 1993 229 221 234 8421302 
32 Taylor JD  Halford SE   Discrimination between DNA sequences by the EcoRV restriction endonuclease Biochemistry 1989 28 6198 6207 2675966 
33 Pingoud A   Spermidine increases the accuracy of type II restriction endonucleases: Suppression of cleavage at degenerate, non-symmetrical sites Eur. J. Biochem. 1985 147 105 109 2982605 
34 Cao W  Mayer AN  Barany F   Stringent and relaxed specificities of TaqI endonuclease: interactions with metal cofactors and DNA sequences Biochemistry 1995 34 2276 2283 7857938 
35 Alves J  Selent U  Wolfes H   Accuracy of the EcoRV restriction endonucleases: binding and cleavage studies with oligodeoxynucleotide substrates containing degenerate recognition sequences Biochemistry 1995 34 11191 11197 7669777 
36 Pingoud A  Fuxreiter M  Pingoud V  Wende W   Type II restriction endonucleases: structure and mechanism Cell. Mol. Life Sci. 2005 62 685 707 15770420 
37 Kang YK  Lee B  Nob MJ  Cho N-Y  Yoo OY   Different effects of base analogue substitutions in BamHI restriction site on recognition by BamHI endonuclease and BamHI methylase Biochem. Biophys. Res. Commun. 1995 206 997 1002 7832816 
38 Galburt EA  Chevalier B  Tang W  Jurica MS  Flick KE  Monnat RJ Jr.  Stoddard BL   A novel endonuclease mechanism directly visualized for I-PpoI Nature Struct. Biol. 1999 6 1096 1099 10581547 
39 Drouin M  Lucas P  Otis C  Lemieux C  Turmel M   Biochemical characterization of I-Cmoe I reveals that this H-N-H homing endonuclease shares functional similarities with H-N-H colicins Nucleic Acids Res. 2000 28 4566 4572 11071947 
40 Wittmayer PK  McKenzie JL  Raines RT   Degenerate DNA recognition by I-PpoI endonuclease Gene 1998 206 11 21 9461409 
41 Argast GM  Stephens KM  Emond MJ  Monnat RJ Jr.   I-PpoI and I-CreI homing site sequence degeneracy determined by random mutagenesis and sequential in vitro enrichment J. Mol. Biol. 1998 280 345 353 9665841 
42 Wittmayer PK  Raines RT   Substrate binding and turnover by the highly specific I-PpoI endonuclease Biochemistry 1996 35 1076 1083 8547243 
43 Hiller DA  Rodriguez AM  Perona JJ   Non-cognate enzyme-DNA complex: structural and kinetic analysis of EcoRV endonuclease bound to the EcoRI recognition site GAATTC J. Mol. Biol. 2005 354 121 136 16236314

