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

J. Biol. Chem., Vol. 277, Issue 16, 14288-14293, April 19, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/16/14288    most recent M109311200v1 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 Mücke, M. Articles by Reuter, M. Search for Related Content PubMed PubMed Citation Articles by Mücke, M. Articles by Reuter, M. Social Bookmarking                      What's this?

Asymmetric Photocross-linking Pattern of Restriction Endonuclease EcoRII to the DNA Recognition Sequence^*

Merlind Mücke, Vera Pingoud, Gerlinde Grelle^§, Regine Kraft^§, Detlev H. Krüger, and Monika Reuter^¶

From the Institut für Virologie, Medizinische Fakultät der Humboldt-Universität zu Berlin (Charité), D-10098 Berlin, the  Institut für Biochemie, Justus-Liebig-Universität, Heinrich-Buff-Ring 58, D-35392 Giessen, and the ^§ Max-Delbrück-Centrum für Molekulare Medizin, Robert-Rössle-Strasse 10, D-13122 Berlin, Germany

Received for publication, September 26, 2001, and in revised form, January 29, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The EcoRII homodimer engages two of its recognition sequences (5'-CCWGG) simultaneously and is therefore a type IIE restriction endonuclease. To identify the amino acids of EcoRII that interact specifically with the recognition sequence, we photocross-linked EcoRII with oligonucleotide substrates that contained only one recognition sequence for EcoRII. In this recognition sequence, we substituted either 5-iododeoxycytidine for each C or 5-iododeoxyuridine for A, G, or T. These iodo-pyrimidine bases were excited using a UV laser to result in covalent cross-linking products. The yield of EcoRII photocross-linked to the 5'-C of the 5'-CCAGG strand of the recognition sequence was 45%. However, we could not photocross-link EcoRII to the 5'-C of the 5'-CCTGG strand. Thus, the contact of EcoRII to the bases of the recognition sequence appears to be asymmetric, unlike that expected for most type II restriction endonucleases. Tryptic digestion of free and of cross-linked EcoRII, followed by high performance liquid chromatography (HPLC) separation of the individual peptides and Edman degradation, identified amino acids 25-49 of EcoRII as the cross-linking peptide. Mutational analysis of the electron-rich amino acids His³⁶ and Tyr⁴¹ of this peptide indicates that Tyr⁴¹ is the amino acid involved in the cross-link and that it therefore contributes to specific DNA recognition by EcoRII.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Restriction endonucleases are excellent model systems for the study of specific protein-DNA interactions because of their extraordinary specificity in DNA recognition. Type II restriction endonucleases form a homodimer to recognize their palindromic, double-stranded DNA recognition sequence. Both subunits display a characteristic 2-fold rotational symmetry when they contact their recognition sequence. This has been shown by crystal structure analyzes of several type II restriction endonucleases in complex with their DNA substrates (1, 2). The homodimeric EcoRII is a type IIE restriction endonuclease that differs from type II restriction endonucleases in that it depends on cooperative binding of two of its recognition sequences (5'-CCWGG) for DNA cleavage activity (3-7). Type IIE restriction endonucleases, like recombinases, DNA repair enzymes, and transcriptional regulators, interact simultaneously with two recognition sequences (8). Although EcoRII shows sequence homology to the Int family of recombinases (9, 10), it is not exactly clear if EcoRII and integrases also recognize DNA similarly. Another type IIE restriction endonuclease, NaeI, has been shown to have topoisomerase activity when Leu⁴³ is replaced by Lys (11). The crystal structures of NaeI and of the NaeI·DNA complex revealed a NaeI dimer that had two domains per monomer, the endo and the topo domain (12, 13). The endo domain forms a DNA binding motif that is structurally related to other type II restriction endonucleases and harbors the catalytic center. The topo domain contains a catabolite activator protein (CAP) motif for DNA binding (12). Both topo and both endo domains bind one double-stranded DNA recognition sequence (12, 13). Thus, two different structural motifs of NaeI recognize the same DNA recognition sequence and are assigned to an activator and a substrate binding site. Because an activator and a substrate binding site have also been discussed for EcoRII, it is possible that EcoRII possesses a structure similar to that of NaeI (3, 8, 14-17).

It has been shown by membrane-bound peptide libraries that one EcoRII monomer contains at least two regions that are involved in DNA recognition (18), DNA binding regions I (amino acids 88-102) and II (amino acids 256-273). Although restriction endonucleases usually show little sequence homology to each other (1, 2), DNA binding region II is homologous to a sequence of the type II restriction endonuclease SsoII (18) that recognizes the DNA sequence 5'-CCNGG (19). Moreover, amino acids have been identified that are conserved in DNA binding region II of EcoRII and in several other restriction endonucleases that recognize C:G and G:C base pairs in their recognition sequences (18). In addition to DNA binding regions I and II, recent, more comprehensive sequence alignments of restriction endonucleases that recognize terminal CC/GG base pairs within their recognition sequence suggest a third region of EcoRII (starting from amino acid 298) that could be involved in DNA recognition and catalysis.¹ Although specific regions involved in DNA recognition have been identified, it is not known which particular amino acids of EcoRII contact specific bases of the DNA recognition sequence. Furthermore, it is not known if contacts are identical for both strands of the recognition sequence. To answer these questions using a He/Cd laser we photocross-linked EcoRII to oligonucleotide duplexes that contained a single recognition sequence for EcoRII. In each of these oligonucleotide duplexes, either the base analogues 5-iododeoxyuridine (X)² or 5-iododeoxycytidine (Y) replaced one base of the recognition sequence. This photocross-linking method allows the identification of contacts between bases and amino acids in DNA-protein as well as in RNA-protein complexes (20-28).

Here we report the cross-linking of EcoRII to the 5'-C of the 5'-CCAGG strand that was replaced by 5-iododeoxycytidine. Because we did not find a cross-link to the 5'-C position of the opposite 5'-CCTGG strand, the cross-linking pattern of EcoRII to the DNA recognition sequence appears asymmetric. Moreover, RP-HPLC and Edman degradation of tryptic peptides of cross-linked and free EcoRII revealed amino acids 25-49 of EcoRII to be the cross-linked peptide. Mutational analysis of position Tyr⁴¹ of this peptide in the EcoRII amino acid sequence led us to conclude that this amino acid contributes to specific DNA recognition of EcoRII.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Wild-type and Mutant EcoRII Preparations-- Site-directed mutagenesis was performed as described previously (18). The mutant enzymes H36A, H36G, and Y41A as well as the wild-type enzyme contained a His₆-tag at the N terminus. The enzymes were expressed in Escherichia coli JM109 (pDK1rm⁺) and purified to >95% homogeneity as described previously (18).

Oligonucleotides-- Oligonucleotides of 20-bp length were purchased from Interactiva Biotechnologie GmbH. The sequence of the unmodified oligonucleotide was 5'-GTTAGAGCCA GGTTGGCAGC-3'. Modifications consisted either in substitutions of 5-iododeoxycytidine for each C, or 5-iododeoxyuridine for each A, G, or T of the recognition sequence. When indicated, oligonucleotides were radioactively labeled using T4 polynucleotide kinase (New England Biolabs) and [-³²P]ATP (Hartmann Analytic).

To test the influence of the base pairs flanking the recognition sequence (bold letters) on the DNA recognition and on the cross-linking pattern, we used the following unmodified oligonucleotide 5'-GTT AGA GCC AGG CTC GCA GC-3'. In this oligonucleotide duplex, we replaced either the 5'-C of the A- or of the T-strand of the recognition sequence (underlined) by Y. Additionally, we introduced a central T/T mismatch in the modified oligonucleotide duplex that contained Y for the 5'-C of the A-strand to search for the reasons of asymmetric cross-linking.

To analyze the influence of the central A/T pair on the asymmetry of the cross-linking pattern, we performed further cross-linking reactions with a self-complementary oligonucleotide 5'-CTCCCA/TGGGAG duplex that contained X for T at the central position or X for both, A and T.

Electrophoretic Mobility Shift Assays with EcoRII and Modified Oligonucleotides-- ³²P-Labeled, 20-bp oligonucleotide duplexes (0.2 µM) were incubated with wild-type EcoRII (0.4 µM) in 1× universal buffer (Stratagene) and 5% glycerol in a total reaction volume of 20 µl at 0 °C for 15 min. The reaction mixtures were then loaded on a native 5% polyacrylamide gel. Gels were run in 1× TBE (90 mM Tris-HCl, pH 8.0, 2 mM EDTA, 90 mM H₃BO₃) at I = 15 mA and 8 °C. After electrophoresis, gels were dried, and radioactive bands were analyzed using a phosphorimager.

Cleavage of Modified Oligonucleotides-- ³²P-Labeled 20-bp oligonucleotide duplexes (0.25 µM) were incubated with EcoRII (0.5 µM) in 1× universal buffer in a total reaction volume of 20 µl at 37 °C for 1 h. Cleavage products were separated on a denaturing 20% polyacrylamide gel containing 7 M urea and run in 1× TBE at I = 40 mA. Gels were dried, and radioactive bands were analyzed by a phosphorimager.

Photocross-linking Reaction with EcoRII-- To test which modified bases form a cross-link to the oligonucleotide duplexes, 3.75 µM EcoRII and 7.5 µM of a mono-substituted ³²P-labeled oligonucleotide duplex were mixed in 1× universal buffer in a total reaction volume of 50 µl at 0 °C. The reaction mixture was then irradiated with a 40-milliwatt He/Cd laser at 325 nm (Laser 2000) at 0 °C for 45 min, and for kinetic experiments for 0-60 min, respectively (25, 27). Two-µl samples were analyzed by 15% SDS-polyacrylamide gel electrophoresis. Gels were silver-stained (29), and radioactive bands were detected using a phosphorimager.

Identification of the Photocross-linking Peptide of EcoRII-- Bands of free EcoRII and of cross-linked EcoRII corresponding to ~15 pmol of protein were cut from a 12% SDS-polyacrylamide gel. Gel slices were minced to cubes of around 1 mm³ and then washed three times with 50/50 (v/v) 0.5 M Tris-HCl, pH 8.5/acetonitrile at 30 °C for 20 min in a volume corresponding to that of the gel slices. Gel slices were dried and then re-hydrated in 0.1 M (NH₄)₂CO₃, 10 mM dithiothreitol, 1 mM EDTA at 60 °C for 45 min. The supernatant was discarded, and the proteins were alkylated "in-gel" by 100 mM iodoacetamide in 0.1 M (NH₄)₂CO₃ at room temperature for 30 min. After alkylation, gel slices were washed and dried as described above. For trypsin digestion, gel slices were re-hydrated in 0.05 M Tris-HCl, pH 8.5, 10% acetonitrile, 1 mM CaCl₂ and incubated with trypsin at a w/w ratio of 1:1 to 1:10 (trypsin:EcoRII) at 37 °C for 16 h. To extract the peptides from the gel, gel slices were incubated in 2 volumes of 2% trifluoroacetic acid at 60 °C for 30 min. The peptide mixture obtained in the supernatant was separated by RP-HPLC on a µRPC C2/C18 SC2.1/10 column using the SMART system (Amersham Biosciences). Peptides were eluted with a gradient of 0.5%/min buffer B (0.1% trifluoroacetic acid, CH₃CN) in buffer A (0.1% trifluoroacetic acid, H₂O) at a flow rate of 100 µl/min. The peptide of EcoRII that corresponded to the peptide that disappeared in the RP-HPLC chromatogram of cross-linked EcoRII was sequenced on a Procise Protein^TM sequencer (Applied Biosystems Inc.).

Electrophoretic Mobility Shift Assays with Mutant Enzymes-- The ³²P-labeled 191-bp DNA fragment was obtained by PCR from the recombinant plasmid-DNA pMM71/1 (30) using the primers 1233 and 1211 (New England Biolabs). This PCR product (0.012 pmol), which contained a single recognition sequence for EcoRII, was incubated with increasing amounts of the mutant enzymes H36A, H36G, and Y41A (0-2.4 nmol). Incubation was performed in 1× universal buffer with 5% glycerol in a total reaction volume of 20 µl at 4 °C for 20 min. Samples were separated on a 5% polyacrylamide gel containing 0.1% SDS. The gel was run in 1× TBE, 0.1% SDS at I = 20 mA and at 8 °C. The apparent K_D value of mutant enzymes was estimated as described previously (18).

Cleavage Activities of Mutant Enzymes-- Cleavage activity was analyzed by incubating 200 ng of BamHI-linearized pBR322 dcm DNA (6.7 pmol sites) with 3.35 pmol of the EcoRII wild-type and the mutant enzyme H36G or with increasing amounts of the mutant enzymes H36A and Y41A due to low cleavage activity. Reactions were performed in 1× universal buffer in a total volume of 20 µl at 37 °C for 60 min. Cleavage was monitored by 0.8% agarose gels stained with ethidium bromide.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

EcoRII Binding and Cleavage of Oligonucleotide Duplexes That Contain a Modified Recognition Sequence-- To identify base-specific contacts of EcoRII to the DNA recognition sequence by photocross-linking, we designed ten modified oligonucleotide duplexes that were 20 bp in length and mono-substituted at each base of the EcoRII recognition sequence. We substituted either Y for C or X for T, A, or G. Consequently, the substitutions for A and G generated a mismatch. Before photocross-linking, we tested how these substitutions influenced DNA binding and cleavage by EcoRII. Electrophoretic mobility shift assay of the ³²P-labeled oligonucleotide duplexes showed that the substitution of iodo-pyrimidines for bases of the recognition sequence interfered with binding of EcoRII at certain positions (Fig. 1). EcoRII formed complexes with oligonucleotide duplexes, where Y was substituted for the 5'-C of both strands of the recognition sequence and where X was substituted for the central T. Apparently EcoRII also bound to the oligonucleotide duplex where X was substituted for the central A (not shown), because EcoRII cleaved this oligonucleotide duplex (Fig. 2). Moreover, EcoRII formed a weak complex with the oligonucleotide duplex, where X was substituted for the 3'-G of the 5'-CCAGG strand. Cleavage of the modified oligonucleotide duplexes by EcoRII supported the results of the shift assays (Fig. 2). Except for the oligonucleotide duplexes, where X was substituted for the central A or T, EcoRII did not cleave any modified oligonucleotide duplex efficiently. EcoRII cleaved oligonucleotide duplexes, where Y was substituted for the 5'-C in both strands with very low efficiency. The same low efficiency was observed when X was substituted for the internal G in the CCAGG strand (5'-CCAXG). These data illustrate the extremely high specificity of DNA recognition by the restriction endonuclease EcoRII. Inhibition of DNA binding of EcoRII due to the substitution of Y for the inner C of the recognition sequence, which is the methylation position of the EcoRII methyltransferase, can be explained by the similar size and polarity of the iodine atom to the methyl group. Mismatches that were caused by substituting X for G reduced or even prevented DNA recognition by EcoRII. In contrast, EcoRII was not sensitive to mismatches (X:T) or modifications (A:X) at the central A:T pair. We conclude from these data that modified oligonucleotide duplexes that cannot be bound by EcoRII are unlikely to be cross-linked to the enzyme efficiently.

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Electrophoretic mobility shift assay of the modified oligonucleotide duplexes and EcoRII. X = 5-iododeoxyuridine, Y = 5-iododeoxycytidine. The double bands of the EcoRII-DNA complexes are due to the heterogeneity of the enzyme preparation.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   Cleavage assay with modified oligonucleotide duplexes and EcoRII. X = 5-iododeoxyuridine, Y = 5-iododeoxycytidine.

Photocross-linking of EcoRII to the DNA Recognition Sequence-- The base analogues X and Y have often been used successfully for studying protein-DNA binding (20-28). Both X and Y are zero-length cross-linkers that can be activated by UV light at  = 325 nm. Zero-length cross-linkers minimize DNA cross-links to amino acids of the protein that are within a distance of 3-5 Å to the DNA. In addition, irradiation at  = 325 nm, which excites electrons of the carbon-iodine bond in X and Y, does not excite other nucleic acid and protein chromophores. X and Y cross-link preferentially to electron-rich amino acid residues such as Phe, Tyr, Trp, His, and Met (31, 32). Aside from the presence of a reactive amino acid residue, the orientation of the reacting group influences the yield of cross-link reactions.

We screened the modified as well as the unmodified, radioactively labeled oligonucleotide duplexes that EcoRII bound to for cross-linking after irradiation using a He/Cd laser at  = 325 nm. Using a temperature of 0 °C during irradiation results in all cross-links being formed with substrate DNA, because EcoRII does not cleave DNA at 0 °C. We found that only the oligonucleotide duplex that contained Y at the 5'-C position of the 5'-CCAGG strand cross-linked to EcoRII with a yield of around 45% of the total EcoRII protein. Irradiation times of 45-60 min resulted in the most photocross-linked product (Fig. 3). Longer irradiation times led to a decrease of photocross-linked and free EcoRII because of photolysis. Substitution of the central A or T by X resulted in cross-link yields of around 5%, which we observed in SDS-polyacrylamide gel electrophoresis and as radioactive bands after phosphorimaging (data not shown). The other modified oligonucleotide substrates did not yield a cross-link (data not shown) due to impaired DNA binding or due to the absence of reactive amino acid residues in close proximity and/or suitable orientation to the photoactive base analogue. As expected, the unmodified oligonucleotide duplex also did not yield a cross-link due to the absence of a photoactive iodo-substitution (data not shown). Therefore, EcoRII cross-linked specifically to the 5'-C position of the 5'-CCAGG strand. The specific photocross-links of EcoRII to the bases of its recognition sequence are summarized in Fig. 4.

View larger version (39K): [in this window] [in a new window]   Fig. 3.   Top panel, time course of the photocross-linking (CL) reaction of EcoRII to the modified oligonucleotide duplex that contained Y at the 5'-C position of the 5'-CCAGG-strand of the recognition sequence. Aliquots of the photocross-linking reaction were analyzed at the time points indicated at the top. The double bands are due to the heterogeneity of the EcoRII enzyme preparation. Marker proteins are shown on the left. Bottom panel, photocross-linking yield depending on irradiation time. The data are based on phosphorimaging.

View larger version (9K): [in this window] [in a new window]   Fig. 4.   Photocross-linking positions for EcoRII to its recognition sequence. The bold arrow symbolizes a cross-linking yield of around 50%. Thin arrows symbolize cross-linking yields of less than 5%. Horizontal line, EcoRII recognition sequence; vertical lines, cleavage position for EcoRII.

The cross-linking pattern is asymmetric, as we did not find a cross-link to the 5'-C position of the 5'-CCTGG strand. To analyze the reason for this asymmetry, we searched for the influence of base pairs flanking the recognition sequence or of the central A/T pair. We did not find a cross-link for the 5'-C in the T-strand even if three identical bases flank the recognition sequence in the A- as well as in the T-strand (data not shown). In addition, we found that EcoRII formed a weak cross-link of around 5% to the 5'-C of the recognition sequence on a DNA substrate with a central T/T mismatch. This cross-link only appeared in the absence of magnesium ions. The cross-link almost completely disappeared in the presence of magnesium ions due to the increased DNA binding specificity of EcoRII in the presence of magnesium. Based on these observations, we conclude that the asymmetric cross-linking pattern was not caused by base pairs flanking the recognition sequence.

Aside from the modified and unmodified oligonucleotide duplexes, we also analyzed a self-complementary oligonucleotide duplex that contained X at the central A/T position. For this duplex, the base pairs flanking the recognition sequence were identical due to the self-complementation. When we replaced T by X in this oligonucleotide duplex, we found a cross-link of around 5%. This result agrees with the results obtained for the oligonucleotide duplexes that did not contain identical flanking base pairs. Additionally, we analyzed the self-complementary oligonucleotide duplex when X replaced both A and T in the recognition sequence. If EcoRII recognizes both DNA strands of its recognition sequence symmetrically, the cross-linking yield should reduplicate, because EcoRII could form cross-links to both DNA strands then. However, we did not observe this for EcoRII (data not shown). Hence, we assume that the asymmetry of the recognition sequence at the central A/T pair is responsible for the asymmetric cross-linking pattern of EcoRII. Thus, EcoRII seems to differ from type II restriction endonucleases with respect to DNA recognition. In general, type II restriction endonucleases recognize their DNA target sequences symmetrically and therefore should show symmetrical cross-linking patterns.

Identification of the Photocross-linked EcoRII Peptide-- To analyze the region of EcoRII that formed the cross-link to the 5'-C position of the 5'-CCAGG strand, we separated cross-linked EcoRII and free EcoRII by cutting respective bands from an SDS-polyacrylamide gel used to separate the photocross-linking reaction mixture. Cross-linked EcoRII and free EcoRII were alkylated by iodoacetamide and subsequently digested by trypsin in-gel. After extracting the peptide mixture from gel pieces, peptides were separated by RP-HPLC (Fig. 5). As shown in the chromatograms in Fig. 5, we analyzed the absorption (A) of the peptide fractions at 214, 260, and 280 nm to detect any differences between the A₂₆₀/A₂₈₀ ratios. We found that the chromatogram of cross-linked EcoRII differed from that of free EcoRII by the absence of a single peak. In addition, all of the peptides retained by the RP material showed no differences between their A₂₆₀/A₂₈₀ ratios, indicating that all separated peptide fractions contained no DNA. The peak missing in the cross-linked EcoRII chromatogram appears, therefore, to consist of the cross-linked peptide, which could not be isolated by RP-HPLC. This peptide did not bind to the RP material, presumably due to the hydrophilicity of the cross-linked nucleic acid as verified using a ³²P-labeled cross-linking product.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Chromatograms of cross-linked EcoRII and free EcoRII. The diagram shows the absorption units (AU, left y axis) at wavelengths 214, 260, and 280 nm versus the retention time (min). The wavelengths are indicated on the graphs. The right y axis shows the gradient B (%) (0.1% trifluoroacetic acid in acetonitrile) in A (0.1% trifluoroacetic acid in water). Arrow, only peptide peak that differs between both chromatograms.

Because the cross-linked peptide could not be isolated by RP-HPLC, we decided to perform Edman degradation of the first ten N-terminal amino acids of the EcoRII peptide that corresponded to the missing peptide in the chromatogram of cross-linked EcoRII. Sequencing the peptide would allow us to place this peptide within the amino acid sequence of EcoRII. We determined that the sequence corresponded to an N-terminal peptide of EcoRII, namely to amino acids 25-49: LSANDTGATG GHQVGLYIPS GIVEK. Mass spectrometry (matrix-assisted laser desorption ionization-mass spectrometry) verified the molecular weight of the peptide (data not shown). Based on these results, we inferred that an amino acid residue of this peptide cross-linked to the Y at the 5'-C position of the 5'-CCAGG strand.

Mutational Analysis of Electron-rich Amino Acids of the Photocross-linked EcoRII Peptide-- To find out which amino acid of the identified EcoRII peptide formed the photocross-link, we decided to mutate the electron-rich amino acids, because electron-rich amino acids are the preferred candidates for cross-linking to halogenated pyrimidines (Phe, Tyr, Trp, His, and Met), (31, 32). Thus, we replaced His³⁶ by Ala and Gly and Tyr⁴¹ by Ala by site-directed mutagenesis. All three His₆-tagged mutant proteins were purified to 95% homogeneity by affinity chromatography. Electrophoretic mobility shift assays for the mutants showed that Y41A did not form a detectable enzyme-DNA complex (Fig. 6). Mutant enzymes H36A and H36G formed an enzyme-DNA complex with the same electrophoretic mobility as the wild type. However, DNA binding affinity of the mutant enzymes H36G was 2 orders of magnitude lower than that of the wild-type enzyme (K_D = 5 nM, Ref. 18) and that of H36A was at least 5 orders of magnitude lower than that of the wild-type enzyme. Nonetheless, mutant enzymes H36A and Y41A retained a low cleavage activity. Therefore, these mutant enzymes should be able to bind to DNA substrates with low affinity.

View larger version (33K): [in this window] [in a new window]   Fig. 6.   Electrophoretic mobility shift assays of the EcoRII mutants H36A, H36G, and Y41A. Enzyme concentrations are written at the top of each lane. The lanes marked as wild type show positive controls of wild-type EcoRII enzyme at a concentration of 50 µM. For comparison, the KD for the wild-type EcoRII is 5 nM (18).

We could detect the low DNA binding affinity of the mutants H36A and Y41A by performing shift assays without SDS. For these assays, we observed diffuse bands of complexes for both mutant enzymes in a concentration range of 0.5-1 µM of enzyme. Mutant enzyme H36G cleaved DNA with an activity comparable with that of wild-type EcoRII (data not shown). When we tried to cross-link the mutant enzymes to the oligonucleotide duplex substituted at the 5'-C position of the 5'-CCAGG strand (Fig. 7), we found that around 19% of the mutant enzyme H36A and around 33% of the mutant H36G cross-linked to the oligonucleotide duplex. In contrast, mutant enzyme Y41A did not yield a cross-link, although we tested enzyme concentrations of 1, 5, and 10 µM. Based on these findings, we conclude that the photocross-linking amino acid of EcoRII is Tyr⁴¹. Tyr⁴¹ should therefore play a role in base-specific target recognition by the enzyme.

View larger version (98K): [in this window] [in a new window]   Fig. 7.   12% SDS-polyacrylamide gel electrophoresis of the photocross-linked mutant enzymes H36A, H36G, Y41A and of the wild-type enzyme. The enzymes were incubated with the oligonucleotide duplex that contained Y at the 5'-C position of the 5'-CCAGG strand and then irradiated (CL-) or not irradiated. The double band of wild-type CL-EcoRII is due to the heterogeneity of the enzyme preparation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this photocross-linking study we searched for base-specific contacts of EcoRII to the recognition sequence 5'-CCWGG. In agreement with results that were found for other nucleic acid base substitutions (33, 34), EcoRII recognized in particular those substrates that were substituted at the 5'-C pair and at the central A/T of both strands. Some of the X and Y substitutions that we introduced in the recognition sequence for photocross-linking inhibited or even prevented DNA binding by EcoRII. Although this limited the possible cross-linking substrates to those that EcoRII bound to, we found that most probably amino acid Tyr⁴¹ of EcoRII is involved in the cross-link to the oligonucleotide duplex that contained Y at the 5'-C position of the 5'-CCAGG strand of the recognition sequence. For this particular photocross-link, we obtained a yield of around 45%.

Although EcoRII bound to the oligonucleotide duplex that contained Y at the 5'-C position of the 5'-CCTGG strand, we did not find a cross-link to this position. Thus, the cross-linking pattern of EcoRII is asymmetric. This asymmetry was not caused by the base pairs flanking the recognition sequence, but rather by the asymmetric central A/T pair within the otherwise palindromic recognition sequence. Therefore, the partial asymmetry at the central A/T pair of the EcoRII recognition sequence induces the asymmetric DNA binding by the restriction endonuclease EcoRII.

Because of the high cross-linking yield and because Y is a zero-length cross-linker, we propose that the Tyr⁴¹ residue of EcoRII is in close proximity of the DNA recognition sequence in the EcoRII-DNA complex. It is known from co-crystal structures of type II restriction endonucleases that the OH group in the side chain of Tyr can form a hydrogen bond to the phosphate backbone of DNA (Tyr⁹⁹ and Tyr¹⁴⁴ of BglII) (35). Furthermore, Tyr is able to form water-mediated contacts to the phosphate backbone of DNA (Tyr¹⁶⁵ of BamHI) (36) or water-mediated contacts to the bases of the recognition sequence (Tyr¹⁹⁰ of BglII) (35). Similarly, Tyr⁴¹ of EcoRII could interact with the DNA substrate.

Residues of the conserved catalytic centers of restriction endonucleases are typically charged amino acids, namely two acidic (Asp, Glu) and one basic (Lys) amino acid. Thus, Tyr⁴¹, which is not charged, should not be one of the active site residues (2). Rather Tyr⁴¹ is involved in DNA binding. Recent sequence homology studies of restriction endonucleases that recognize CC/GG pairs within their different recognition sequences, including EcoRII support this hypothesis.¹ Based on these homology studies, the C-terminal sequence motif ²⁹⁸PDX₂₄KX₁₁E has been predicted to be the putative active site of EcoRII.

The properties of the mutant enzyme Y41A supported our conclusion that Tyr⁴¹ is involved in specific DNA binding, because substitution of Tyr⁴¹ by Ala reduced DNA binding affinity by more than five orders of magnitude (i.e. beyond the detection limit of the electrophoretic mobility shift assay). However, the mutant enzyme Y41A retained some cleavage activity. This low cleavage activity is most probably due to impaired DNA binding. If one compares the three mutant enzymes that we have studied (H36G, H36A, Y41A), it can be inferred that the more DNA binding affinity is reduced, the more DNA cleavage activity is reduced. In contrast to the mutant enzymes H36A and H36G, the mutant enzyme Y41A did not yield a cross-link. This confirms our assumption that Tyr⁴¹ is the cross-linking amino acid.

Amino acid Tyr⁴¹, as well as the amino acids 25-49 of the cross-linking peptide, are not part of DNA binding regions I or II that had been found by membrane-bound peptide libraries. Because membrane-bound peptide libraries emphasize the contribution of electrostatic binding at the expense of other more specific protein-DNA interactions, these libraries are not suitable to identify individual amino acid residues of EcoRII that contact the specific DNA recognition sequence (18). Therefore, the identification of Tyr⁴¹ as the main cross-linking amino acid has added a new element involved in DNA recognition by EcoRII. Furthermore, it confirms that the N-terminal part of EcoRII substantially contributes to DNA-binding of EcoRII.

If on assumes a two-domain structure for EcoRII, as reported for NaeI, the N terminus of EcoRII, including amino acids 25-49, might be a DNA binding domain corresponding to the topo domain of NaeI (12, 13). The N terminus of EcoRII may correspond to the C-terminal topo domain of NaeI, because all known mutations of the N-terminal DNA binding region I as well as the substitutions of His³⁶ and Tyr⁴¹ by Ala have primarily affected DNA binding and consequently DNA cleavage by EcoRII (18). On the other hand, mutations in the C-terminal DNA binding region II of EcoRII affected DNA cleavage but never severely affected DNA binding (18). Therefore, we propose that Tyr⁴¹ is part of the DNA binding site of the effector (activator) domain.

	ACKNOWLEDGEMENTS

We thank Dr. Alfred Pingoud for stimulating discussions. We thank Dr. Angelika Hofmann and Dr. Martin Raftery for critical reading of the manuscript. We gratefully acknowledge Petra Mackeldanz, Ulrike Marzahn, and Ursula Scherneck for skillful technical assistance.

	FOOTNOTES

^* This work was supported by the Deutsche Forschungsgemeinschaft (Re 879/2-3) and the Universitäre Forschungsförderung of the Humboldt University.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. Tel.: 49-30-450-52-52-01; Fax: 49-30-450-52-59-07; E-mail: monika.reuter@charite.de.

Published, JBC Papers in Press, February 6, 2002, DOI 10.1074/jbc.M109311200

¹ V. Pingoud and V. Siksnys, personal communication.

	ABBREVIATIONS

The abbreviations used are: X, 5-iododeoxyuridine; Y, 5-iododeoxycytidine; RP-HPLC, reversed phase high performance liquid chromatography.

	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:

				S. Grazulis, E. Manakova, M. Roessle, M. Bochtler, G. Tamulaitiene, R. Huber, and V. Siksnys Structure of the metal-independent restriction enzyme BfiI reveals fusion of a specific DNA-binding domain with a nonspecific nuclease PNAS, November 1, 2005; 102(44): 15797 - 15802. [Abstract] [Full Text] [PDF]

				R. Sabatini, N. Meeuwenoord, J. H. van Boom, and P. Borst Site-specific Interactions of JBP with Base and Sugar Moieties in Duplex J-DNA. EVIDENCE FOR BOTH MAJOR AND MINOR GROOVE CONTACTS J. Biol. Chem., July 26, 2002; 277(31): 28150 - 28156. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/16/14288    most recent M109311200v1 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 Mücke, M. Articles by Reuter, M. Search for Related Content PubMed PubMed Citation Articles by Mücke, M. Articles by Reuter, M. Social Bookmarking                      What's this?