Bioconjugate Chem. 1998, 9, 703−707
703
Conformational Transition of Restriction Endonuclease MvaI-Substrate Complex under the Influence of Mg2+ Probed by DNA-Protein Cross-Linking Studies G. Y. Sheflyan,† E. A. Kubareva, E. S. Gromova, and Z. A. Shabarova* A.N. Belozersky Institute of Physical and Chemical Biology, Department of Chemistry, M.V. Lomonosov Moscow State University, 119899 Moscow, Russia. Received February 10, 1998; Revised Manuscript Received May 22, 1998
The method of protein affinity modification by DNA analogues was used to study the characteristic features of restriction endonuclease MvaI interaction with DNA. Oligonucleotide duplexes containing a monosubstituted pyrophosphate internucleotide bond were used for cross-linking to the enzyme. The conditions of the reaction of MvaI endonuclease with these reagents were investigated. On the basis of data obtained, the model of successive inclusion of two Mg2+ ions into MvaI endonucleasesubstrate complex was proposed and confirmed by the kinetic scheme of the process.
INTRODUCTION
Restriction endonuclease MvaI recognizes the DNA sequence
The aim of this paper is to extend the application of this cross-linking technique toward the probing of conformational changes of enzyme in the enzyme-substrate complex on the example of restriction endonuclease MvaI. MATERIALS AND METHODS
and cleaves it, as shown by the arrows, Mg2+ being the cofactor of this reaction. The interaction of this enzyme with DNA was probed via the use of modified DNA substrates (1). We use the method of DNA-protein crosslinking to reveal the characteristic features of substrate scission by R.MvaI.1 In our laboratory, the novel method of affinity modification of proteins by modified DNA was developed (2). DNA analogues contain a monosubstituted pyrophosphate internucleotide bond in the definite position of sugar-phosphate backbone. This method was successfully applied for the cross-linking of restriction-modification enzymes EcoRI and RsrI (3) and restriction endonucleases EcoRII (4, 5), MvaI (5), and SsoII (6). Crosslinking follows the scheme:
* To whom correspondence should be addressed. Phone (095) 939-5411. Fax (095) 939-3181. E-mail zoeshab@ nuclacid.genebee.msu.su. † Present address: 1028 Pharmacy, College of Pharmacy, University of Michigan, Ann Arbor, MI 48109-1065. 1 Abbreviations: R.MvaI, restriction endonuclease MvaI; NuH, nucleophilic agent; MeIm, N-methylimidazole; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; Tris, tris(hydroxymethyl)aminomethane.
R.MvaI was purchased from NPO “Fermentas” (Lithuania). Its preparation had the specific activity of 250 units/µL and concentration of 46 µg/mL. Oligodeoxyribonucleotides were synthesized by T. S. Oretskaya, E. M. Volkov, and E. A. Romanova as described in ref 7. DNA duplexes with a monosubstituted pyrophosphate internucleotide bond for cross-linking to R.MvaI were synthesized as described in ref 2. DNA duplexes were 32Plabeled by T4 polynucleotide kinase. 32P label was at the disubstituted phosphate group of the modified internucleotide linkage within the DNA duplexes I and II modified strands (see formulas) and at the 5′-end of the nonmodified duplex IV unless otherwise stated. Crosslinking of R.MvaI (138 ng, concentration per monomer 2.5 × 10-7 M) to 32P-labeled DNAs (100 000 cpm, concentration per duplex, CD, 1.8 × 10-7 M) in 20 µL of buffer X (10 mM Tris-HCl, pH 8.5, 150 mM NaCl, 1 mM dithiotreitol, and 0.1 mg/mL albumin) or buffer Y (10 mM MeIm-HCl, pH 8.5, 150 mM NaCl, 1 mM dithiotreitol, and 0.1 mg/mL albumin) was performed at 37 °C as described in ref 5. Concentrations of Mg2+ ions have been varied. Incubation time was 18 h. Cross-linking reaction was monitored by SDS-PAGE (5, 6). The enzymatic hydrolysis was performed under the same conditions except that radioactivity of DNA per reaction was 20 000 cpm, CD - 3.5 × 10-7 M, incubation time varied from 1 to 18 h, and the reaction was followed by 20% PAGE containing 7 M urea. RESULTS AND DISCUSSION
We used the following modified DNA duplexes as R.MvaI substrates or their analogues:
10.1021/bc9800163 CCC: $15.00 © 1998 American Chemical Society Published on Web 10/15/1998
704 Bioconjugate Chem., Vol. 9, No. 6, 1998
Sheflyan et al.
Figure 1. The dependence of relative cross-linking yields of R.MvaI to duplex I on the CD ratio of duplexes I and III. CD of duplex I is 1.8 × 10-7 M. Cross-linking yield in the absence of duplex III was referred to as 100%. Table 1. Cleavage of Duplexes I, II, and IV by R.MvaI at 37 °C in Buffer X, Containing 10 mM MgCl2, for 1 h duplex I
In duplex I, the modified group was located in the middle of R.MvaI recognition site and coincided with the scissile bond. In duplex II, we probed protein contacts with sugar-phosphate backbone on the border between the flanking sequence and the recognition site (5). Nonmodified duplexes III and IV were used in control experiments. The monosubstituted pyrophosphate internucleotide bond in DNA duplexes does not interfere with their binding to the enzyme (5). Previously, we demonstrated (5) that the cross-linking reaction of R.MvaI to duplex I required Mg2+ ions in the concentration of 37.5 µM. At higher as well as lower concentrations of Mg2+, modification of R.MvaI proceeded with low yield. The increase of nucleophilicity of reaction buffer replacing Tris-HCl (buffer X) by MeIm (buffer Y) did not result in the increase of a cross-linking product unlike the data previously observed for R.EcoRII (4). The cross-linking extent of duplex II to R.MvaI did not exceed 2% in the absence of Mg2+ ions nor in their presence till the Mg2+ concentration 150 µM (5). The specificity of covalent attachment of duplex I to R.MvaI was established through the standard method of competitive inhibition of the affinity modification reaction by the increasing amounts of duplex III (Figure 1). This reaction proved to be specific. Therefore, it can reflect the structure of DNA-protein interface and its changes during the course of enzymatic hydrolysis. To explain the results of R.MvaI affinity modification by duplexes I and II we checked (i) the ability of R.MvaI to hydrolyze duplex I and II in comparison with duplex IV; (ii) R.MvaI activity dependence on the Mg2+ concentration on the example of duplex IV; and (iii) R.MvaI activity dependence on the buffer nucleophilicity. The results of hydrolysis of duplexes I, II and IV by R.MvaI are adduced in Table 1. Hydrolysis of duplex I in buffer containing 10 mM MgCl2 results in the number of products corresponding to the weak enzymatic hydrolysis of intact strand Ib as well as the monosubstituted pyrophosphate bond degradation. Since the active group coincides with the R.MvaI scissile bond, it is difficult to explain the degradation of the modified strand Ia by either enzymatic or chemical cleavage. However, in ref 1, it was shown that R.MvaI cleaves only the intact strand of the DNA duplex containing a pyrophosphate group instead of the scissile phosphodiester bond. This allows us to suggest that the enzymatic hydrolysis of a
II IV a
strand
enzymatic cleavage, %
a b a b a+b
-a 8 9; 51b 84 88
Modified bond coincides with the scissile bond. b For 18 h.
Figure 2. The dependence of duplex IV hydrolysis by R.MvaI on Mg2+ concentration. CD of duplex IV is 3.5 × 10-7 M. The Mg2+ concentration is adduced in the logarithmic scale.
modified strand of duplex I does not occur. It was also demonstrated in ref 8 that R.MvaI is capable of cutting certain single-stranded oligodeoxyribonucleotides, for example, strand IVa, having the same sequence as Ia, but not strand IVb (Ib). This eliminates the possibility that the weak cleavage of strand Ib in duplex I is due to the formation of an incomplete duplex by strand Ib alone. When monosubstituted pyrophosphate internucleotide bond is located on the border between the recognition site and the flanking nucleotide sequence (duplex II), the hydrolysis of modified strand IIa at the Mg2+ concentration equal to 10 mM occurs in a time-dependent manner. The intact strand IIb is efficiently cleaved by the enzyme (Table 1). At low Mg2+ concentrations (till 7.5 µM), R.MvaI practically does not cleave duplex IV (Figure 2). At Mg2+ concentration equal to 37.5 µM, the hydrolysis extent of duplex IV is equal to 20%. The following addition of Mg2+ ions causes the fast growth of enzymatic activity. At the Mg2+ concentration 75 µM, duplex IV is completely digested by R.MvaI. The efficiency of hydrolysis does not depend on incubation time (1 and 18 h, data not shown). We demonstrated that Tris-HCl replacement for MeImHCl in the reaction buffer did not decrease R.MvaI enzymatic activity in cleaving duplex IV (data not
R.MvaI cross-linking to DNA
shown). The weak enzymatic hydrolysis of only strand Ib in duplex I made it impossible to investigate the influence of Mg2+ on the enzymatic cleavage of this duplex by R.MvaI. All the reactions with duplex IV, having the identical sequence as duplex I, were performed under the same conditions as cross-linking experiments, with CD of duplex IV being in the same range as the CD of the cross-linking reagent (350 nM for duplex IV and 180 nM for duplex I). The necessity of the presence of Mg2+ in low concentration for the R.MvaI covalent attachment to duplex I and the hydrolysis studies is believed to show the following features of DNA recognition by R.MvaI: (1) In the absence of Mg2+ ions, there are no nucleophilic amino acid residues in the proximity to the scissile bond modified in duplex I. (2) The introduction of Mg2+ in an enzyme-substrate complex results in a conformational change of the enzyme. As a consequence of this transition, a nucleophilic amino acid residue is moved toward the scissile bond. This is demonstrated by the increase of the cross-linking extent of the enzyme to duplex I in these conditions. (3) The Mg2+ concentration equal to 37.5 µM is not sufficient for the effective substrate hydrolysis (Figure 2), though Mg2+ ion is already incorporated into the enzyme-substrate complex as revealed by cross-linking studies. Hence, more than one Mg2+ ion is necessary for the R.MvaI catalytic act. (4) With the further increase of Mg2+ concentration, the catalytically active complex is formed and the substrate is cleaved. In the case where duplex I is the substrate, the covalent attachment is inhibited due to enzymatic hydrolysis. This explains the reduction in the cross-linking extent of duplex I to R.MvaI at Mg2+ concentrations above 150 µM. (5) The ability of duplex II to modify R.MvaI demonstrates the contact of a nucleophilic group with the phosphate group adjacent to 5′-terminus of recognition sequence. The absence of the dependence of the crosslinking yield of duplex II to R.MvaI on Mg2+ concentration 150 µM confirms that the enzyme conformational change has a local character and occurs only in the surroundings of the scissile bond. Thus, R.MvaI requires more than one Mg2+ ion for catalysis. The same feature was supposed in ref 9 for EcoRV restriction endonuclease and confirmed in ref 10. One can conclude that these enzymes share the same mode of action. From the data obtained, as well as taking into account (10), we proposed the following scheme of the enzymatic reaction together with the parallel crosslinking process:
Bioconjugate Chem., Vol. 9, No. 6, 1998 705
dissociation into the enzyme and products. If only complex B with one Mg2+ ion can be transformed into the cross-linked product, the cross-linking yield is proportional to the complex B concentration in the reaction mixture. Complex C can also retain the conformation favorable for cross-linking, but the degradation of this complex through substrate cleavage proceeds faster than the affinity modification of the enzyme. Using the assumption of steady-state concentrations of intermediate compounds as well as material balance for enzyme concentration, we get the system of equations:
d[A] ) k1[E][S] - k2[A] - k3[A][Mg2+] + k4[B] ) 0 dt d[B] ) k3[A][Mg2+] - k5[B][Mg2+] - k4[B] ) 0 dt d[C] ) k5[B][Mg2+] - k6[C] ) 0 dt [E] ) [E0] - [A] - [B] - [C] To calculate the concentration of B we use the following assumptions: (1) the consumption of the complex B in covalent attachment is too small to influence its concentration and (2) analysis of the Mg2+ binding was performed in terms of the concentration of free Mg2+ ions. The concentration of free Mg2+ ions was obtained by the subtraction of the phosphate group concentration from the initial concentration of Mg2+. We used an assumption that one phosphate group binds one cofactor ion. The concentration of the phosphate groups in the crosslinking reactions was equal to 4.5 µM. Since the concentrations of E and S are fixed, the concentration of B is adduced as a function of Mg2+ concentration. The final equation is
[B] )
l[Mg2+] m[Mg2+]2 + n[Mg2+] + p
where
l ) k1k3k6[S][E0] m ) k1k3k5[S] + k3k5k6 n ) k1k5k6[S] + k1k3k6[S] + k2k5k6 p ) k1k4k6 + k2k4k6 Since we suppose that γ (the extent of affinity modification) is proportional to a concentration of B with the coefficient q determining the probability of formation of covalent conjugate, we get
where E is the enzyme, S is the general substrate or duplex I in the case of covalent attachment, and [E-S]CL is the product of attachment of R.MvaI to duplex I. On the basis of this scheme, we calculated the theoretical dependence of R.MvaI cross-linking to duplex I on Mg2+ concentration and compared it with the experimental data. To simplify the calculation, we assumed that the formation of the complex with two Mg2+ ions is irreversible since it undergoes quick substrate hydrolysis and
l[Mg2+] γ)q m[Mg2+]2 + n[Mg2+] + p We picked the coefficients via computer simulations. The function obtained is in a good agreement with the experimental data (Figure 3). The requirement of two Mg2+ ions for successful cleavage of substrate demonstrates the similarity of mechanisms of R.MvaI and EcoRV restriction enzyme, the later being extensively studied by different methods including kinetics, X-ray structure analysis, and stop flow technique. As proposed in ref 11 on the basis of the
706 Bioconjugate Chem., Vol. 9, No. 6, 1998
Figure 3. Theoretical dependence of covalent attachment of R.MvaI to duplex I on the Mg2+ concentration (solid line) in comparison with the experimental data ([) (5).
kinetic order of enzymatic hydrolysis on Mg2+ concentration, R.EcoRV involves only one ion in catalysis. But as mentioned in ref 12, the kinetic equation for R.EcoRV would have the second order on Mg2+ concentration only in the case when both ions are attached with high cooperativity in one action. By kinetic means, it is possible to determine only the order of the slowest step of the process. One can conclude the possibility of successive binding of Mg2+ ions to sites with different affinity to cofactor. In ref 12, on the basis of the previously obtained X-ray structure of the R.EcoRV complex with substrate, two Mg2+-binding sites formed by carboxyl-groups of Asp90, Asp74, and Glu45 residues were proposed. Both Asp residues originate from consensus sequence common to a number of restriction endonucleases Pro-Asp...Asp-XLys, where X is the hydrophobic amino acid residue. Glu residue is in close contact with scissile phosphodiester bond only in the case of R.EcoRV compared to R.EcoRI, for which one Mg2+ ion was proved to be sufficient for DNA cleavage. Modeling the crystal structure of R.EcoRV with substrate and one cofactor ion also allowed a confirmation of the binding site and the catalytic significance of the second Mg2+ ion (13). Though the participation of two Mg2+ ions does not raise any doubts now, the location of the second cofactor binding site is still under discussion. EcoRV triple mutant Glu45Ala/Asp74Ala/Asp90Ala was obtained in ref 14, which retained the ability to discriminate cognate DNA only in the presence of Mg2+ being catalytically inactive. By cleavage experiments with phosphorothioate bond containing substrates as well as site-mutagenesis experiments, it was revealed that the second Mg2+binding site is probably formed by Tyr 129 residue and a phosphate group between dG and dA residues of the EcoRV recognition site (14). But Asp residues from the consensus sequence are still critical for cofactor and substrate binding. In the R.MvaI amino acid sequence, the motif 49 50 117 118 119 Pro -Asp...Asp-Phe-Lys exists, which can play an analogous role in Mg2+ binding. Complex B can be a covalent one, being an intermediate compound during enzymatic hydrolysis. To test this possibility we used duplex IV containing the 32P-label within the scissile bond. The mixtures of R.MvaI and duplex IV were incubated under various Mg2+ concentrations during 2.5 or 18 h and analyzed by SDS-PAGE.
Sheflyan et al.
No radioactive bands with mobility less than that of initial DNA were detected. So, we did not manage to reveal the covalent intermediate between R.MvaI and duplex IV in the course of enzymatic hydrolysis. The catalysis of phosphodiester bond hydrolysis possibly occurs without the formation of enzyme-DNA linkage, which is in accordance with the data of ref 15. As computer simulations showed, the appropriate coordination of cofactor ions could cause the catalysis of internucleotide bond cleavage (16). The possibility of conformational changes for restriction endonucleases in the complexes with substrates is now discussed for a number of enzymes. By X-ray analysis, the existence of structural disturbances of substrate is revealed on the example of R.EcoRI and R.EcoRV complexes with DNA (17). The further conformational changes of these complexes are suggested under the influence of Mg2+ ions. It is particularly essential for R.EcoRV, since only in the presence of Mg2+, the enzyme can distinguish its recognition site in DNAs (18). Analogous data were recently obtained for R.TaqI and R.PaeR7 (19). Retardation assays of R.EcoRV-substrate complexes demonstrated that Mg2+ ions introduced into these complexes form contacts with both protein and DNA. Changes in DNA-protein interface lead to close contact of active amino acid residue and the scissile bond (20). Mg2+ ions induce conformational changes not only in the enzyme but in DNA as well (21). So, by means of covalent attachment of R.MvaI to activated substrate analogues, we established the conformational perturbation of the enzyme under the influence of a cofactor. At the first stage of the catalytic act the complex is formed with one Mg2+ ion, which is a structural precursor of the catalytic active complex. In this complex, the initial conformational changes occur, induced by the cofactor, and are completed by the subsequent addition of the second Mg2+ ion to form the full catalytic unit. ACKNOWLEDGMENT
This work was supported by the Russian Foundation of Fundamental Investigation (Grant 9404-12649a). The authors are thankful to Molchanova Ya.V. for the help with mathematical simulations. LITERATURE CITED (1) Gromova, E. S., Kubareva, E. A., Vinogradova, M. N., Oretskaya, T. S., and Shabarova, Z. A (1991) Peculiarities of recognition of CCA/TGG sequences in DNA by restriction endonucleases MvaI and EcoRII. J. Mol. Recognit. 4, 133141. (2) Kuznetsova, S. A., Ivanovskaya, M. G., and Shabarova, Z. A. (1990) Chemical reactions within the double-stranded DNA. IX. Directed introduction of substituted pyrophosphate bonds in the DNA structure. Bioorgan. Khim. (Russ.) 16, 219-225. (3) Purmal, A. A., Shabarova, Z. A., and Gumport, R. I. (1992) A new affinity reagent for the specific, covalent attachment of DNA to active-site nucleophiles: application to the EcoRI and RsrI restriction and modification enzymes. Nucleic Acids Res. 20, 3713-3719. (4) Shabarova, Z. A., Sheflyan, G. Ya., Kuznetsova, S. A., Kubareva, E. A., Sysoev, O. N., Ivanovskaya, M. G., and Gromova, E. S. (1994) Affinity modification of restriction endonuclease EcoRII by DNA duplex containing monosubstituted pyrophosphate internucleotide bond. Bioorgan. Khim. (Russ.) 20, 413-419. (5) Sheflyan, G. Ya., Kubareva, E. A., Volkov, E. M., Oretskaya, T. S., Gromova, E. S., and Shabarova, Z. A. (1995) Chemical cross-linking of MvaI and EcoRII enzymes to DNA duplexes
R.MvaI cross-linking to DNA containing monosubstituted pyrophosphate internucleotide bond. Gene 157, 187-190. (6) Sheflyan, G. Y., Kubareva, E. A., Kuznetsova, S. A., Karyagina, A. S., Nikolskaya, I. I., Gromova, E. S., and Shabarova, Z. A. (1996) Cross-linking of SsoII restriction endonuclease to cognate and noncognate DNAs. FEBS Lett. 390, 307-310. (7) Volkov, E. M., Romanova, E. A., Krug, A., Oretskaya, T. S., and Potapov, V. K. (1988) Automated synthesis of oligodeoxyribonucleotides with terminal phosphate groups. Bioorgan Khim. (Russ.) 14, 1034-1039. (8) Kubareva, E. A., Pein, C. D., Gromova, E. S., Kuznezova, S. A., Tashlitzki, V. N., Cech D., and Shabarova, Z. A. (1988) The role of modifications in oligonucleotides in sequence recognition by MvaI restriction endonuclease. Eur. J. Biochem. 175, 615-618. (9) Winkler, F. K., Banner, D. W., Oefner, C., Tsernoglou, D., Brown, R. S., Heathmen, S. P., Bryan, R. K., Martin, P. D., Petratos, K., and Wilson, K. S. (1993) The crystal structure of EcoRV endonuclease and of its complexes with cognate and noncognate DNA fragments. EMBO J. 12, 1785-1795. (10) Vipond, I. B., Baldwin, G. S., and Halford, S. E. (1995) Divalent metal ions at the active sites of the EcoRV and EcoRI restriction endonucleases. Biochemistry 34, 697-704. (11) Jeltsch, A., Alves, J., Wolfes, H., Maass, G., and Pingoud, A. (1993) Substrate assisted catalysis in the cleavage of DNA by the restriction enzymes EcoRI and EcoRV. Proc. Natl. Acad. Sci. U.S.A. 90, 8499-8503. (12) Baldwin, G. S., Vipond, I. B., and Halford, S. E. (1995) Rapid reaction analysis of the catalytic cycle of the EcoRV restriction endonuclease. Biochemistry 34, 705-714. (13) Jeltsch, A., Maschke, H., Selent, U., Wenz, C., Kohler, E., Conolly, B. A., Thorogood, H., and Pingoud, A. (1995) DNA binding specificity of the EcoRV restriction endonuclease is
Bioconjugate Chem., Vol. 9, No. 6, 1998 707 increased by Mg2+ binding to a metal ion binding site distinct from the catalytic center of the enzyme. Biochemistry 34, 6239-6246. (14) Kostrewa, D., and Winkler, F. K. (1995) Mg2+ binding to the active site of EcoRV endonuclease: a crystallographic study of complexes with substrate and product DNA at 2 Å resolution. Biochemistry 34, 683-696. (15) Pingoud, A., Alves, J., and Geiger, R (1993) Restriction enzymes. Methods in Molecular Biology, v. 16, Enzymes of Molecular Biology (Burell, M. M., Ed.) pp 107-200, Humana Press Inc., Totowa, NJ. (16) Uchimaru, T., Uebasi, M., Tanabe, K., and Taira, K. (1993) Theoretical analyses of the role of Mg2+ ions in ribozyme reactions. FASEB J. 7, 137-142. (17) Winkler, F. K. (1992) Structure and function of restriction endonucleases. Curr. Opin. Struct. Biol. 2, 93-99. (18) Taylor, J. D., Badcoe, I. G., Clarke, A. R., and Halford, S. E. (1991) EcoRV restriction endonuclease binds all DNA sequences with equal affinity. Biochemistry 30, 8743-8753. (19) Zebala, J. A., Choi, J., and Barany, F. (1992) Characterization of steady state, single-turnover and binding kinetics of the TaqI restriction endonuclease. J. Biol. Chem. 267, 80978105. (20) Thielking, V., Selent, U., Kohler, E., Landgraf, A., Wolfes, H., Alves, J., and Pingoud, A. (1992) Mg2+ confers DNA binding specificity to the R‚EcoRV restriction endonuclease. Biochemistry 31, 3728-3732. (21) Stover, T., Kohler, E., Fagin, U., Wende, W., Wolfes, H., and Pingoud, A. (1993) Determination of the DNA bend angle induced by the restriction endonuclease R.EcoRV in the presence of Mg2+. J. Biol. Chem. 268, 8645-8650.
BC9800163
