Mechanistic Studies on E. cob DNA Topoisomerase I: Divalent Ion Effects Paul L. Domanico and Yuk Ching Tse-Dinh Department of Enzymology, Glaxo Research Laboratories, Research Triangle Park, North Carolina.-YCT-D. Central Research and Development Department, E. I. du Pont de Nemours & Co., Wilmington, Delaware
PLD.
ABSTRACT E. coli DNA topoisomerase I catalyzes the hydrolysis of short, single stranded oligodeoxynucleotides. It also forms a covalent protein-DNA complex with negatively supercoiled DNA in the absence of Mg 2+ but requires Mg ‘+ for the relaxation of negatively supercoiled DNA. In this paper we investigate the effects of various divalent metals on catalysis. For the relaxation reaction, maximum enzyme activity plateaus after 2.5 mM Mg’+. However, the rate of cleavage of short oligodeoxynucleotide increased linearly between 0 and 15 mM Mg’+. In the oligodeoxynucleotide cleavage reaction, Ca’+, Mn2+, Co*+, and Zn*+ inhibit enzymatic activity. When these metals are coincubated with Mg2+ at equimolar concentrations, the normal effect of Mg ‘+ is not detectable. Of these metals, only Ca2+ can be substituted for Mg2 + as a metal cofactor in the relaxation reaction. And when Mg2+ is coincubated with Mn’+, Co’+, or Zn2+ at equimolar concentrations, the normal effect of Mg2+ on relaxation is not complex to assume a conformation detectable. We propose that Mg 2+ allows the protein-DNA necessary for strand passage and enhance the rate of enzyme turnover.
INTRODUCTION E. cofi Topoisomerase
I relaxes negatively supercoiled DNA and increases the linking number of the DNA product in units of one [l]. The enzyme is believed to catalyze this phenomena by breaking one strand of the DNA duplex in a single stranded region, passing the other, intact strand through the break, and finally rejoining the broken strand [2, 31. A model involving two binding domains has been proposed for the interaction of bacterial topoisomerase I with DNA [4-61. One domain, which forms a covalent bond to DNA, contains a nucleophilic tyrosine residue that attacks a phosphate group on the DNA substrate displacing the 3’-OH moiety of the upstream ribosyl group during the strand breakage step [7]. The other
Address reprint requests and correspondence to: Paul L. Domanico, Department of Enzymology, Research Laboratories,
Glaxo
5 Moore Drive, Research Triangle Park, NC 27709.
Journal of Inorganic Biochemistry, 42, W-96 (1991) 0 1991 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, NY, NY 10010
81 0162-0134/91/$3.50
88
P. L. Domanico and Y.-C. Tse-Dinh
domain binds noncovalently to DNA. The type of interaction between the DNA and the noncovalent binding domain is not clear. This domain interacts with the broken DNA strand and perhaps the unbroken strand and may position the 3’-OH group during the religation step. Furthermore, an energy source such as ATP or NAD is not required since the free energy within the DNA phosphodiester bond is conserved as a result of the transesterification reaction. The efficiency with which E. coli Topoisomerase I relaxes supercoiled DNA is highly dependent on the degree of 3NA supercoiling and the energy inherently stored within the supercoiled molecule. The degree of single stranded character in the duplex DNA and the free energy stored in the supercoiled molecule are proportional to the amount of negative supercoiling [8- 111. This could effect both the binding of E. coli Topoisomerase I to the DNA because the enzyme prefers single stranded DNA and the number of covalent intermediates that proceed through the strand passage step if this step was sensitive to the state of supercoiling. These arguments help explain the observation that the efficiency of relaxing negatively supercoiled DNA is higher with a more negatively supercoiled substrate (21. With a small, single stranded oligodeoxynucleotide as a substrate, the strand upstream of the point of transesterification is released after cleavage [6, 121. The small size of the strand 5’ to the cleavage site may be cause for its disassociation from the enzyme. Hydrolysis of the phosphodiester bond between the active site tyrosine residue and the single stranded DNA results in enzyme ‘turnover [6]. Hydrolysis does not occur with double stranded DNA as the substrate. This suggests that the uncleaved DNA strand assists in the transesterification reaction by anchoring that region of the cleaved DNA strand not covalently bound to the enzyme. Not all steps in the relaxation of supercoiled DNA by E. cali Topoisomerase I require metal as a cofactor 13, 121. Mg 2’ is required for relaxation of supercoiled DNA and other reactions involving strand passage. Whereas, the formation of the covalent complex does not require any exogenous Mg2 ‘. This study focuses on the role of metal in the catalysis of relaxation of supercoiled DNA, and the cleavage of short, single stranded oligodeoxynucleotides by E. coii Topoisomerase I- where cleavage of oligodeoxynucleotide is intended to mimic covalent complex formation during transesterification of supercoiled DNA. The relaxation reaction was studied with the negatively supercoiled substrate, PM2 DNA. The cleavage reaction was studied with a short oligodeoxynucleotide substrate, the thymidine octamer where cleavage by E. colt’ Topoisomerase 1 occurs at the phosphodiester linkage between the 4th and 5th nucleotides. MATERIALS Deionized formamide was from Fluka. Agarose was from BIORAD. Electrophoresis grade TEMED, 19: 1 acrylamide: N,N’-methylene bisacrylamide and urea were from BRL. Oligodeoxynucleotide pdT, was from Pharmacia. (Y-~‘P) ATP (10 mCi/ml) and NENSORB TM 20 columns were from NEN. T4 polynucleotide kinase was from New England BioLabs. Scinti Verse ITM was from Fisher, and negatively supercoiled PM2 DNA was from Boehringer Mannheim. All other organic reagents, buffers, and inorganic salts were commercially available reagent-grade chemicals. In all experiments, distilled deionized water was used. 10X Kinase bu&r: SO mM DTT, 100 mM MgCl,, and 700 mM Tris * HCl at pH 7.5. 10 X Relaxation bu@er: 1 mg/ml gelatin, 100 mM .@-mercaptoethanol, and
90
P. L. Domanico
and Y.-C.
Kye-Dinh
tration was 2 mM. The ionic strength of the cleavage reaction wa! maintained at SO mM with NaCl. The reaction solution minus the enzyme, and the stock enzyme solution were pre-incubated for five minutes at 37°C. The cleavage reaction was initiated by the addition of enzyme (10-80 ,ug/ml) and maintained at 37”C, and then quenched after 30 min by the addition of 5 ~1 SIX stop solutir~n Analysis
of Supercoiled
PM2 DNA Relaxation
Twenty ~1 of each sample was loaded into a 0.7% agarose gel and electrophoresed at 70 V, constant voltage. for I& 16 h. The gels were then lightly stained with ethidium bromide (2.5 pg/ml) for 5 min. The covalcntly closed PM2 DNA circles were then nicked by exposure to long wavelength UV irradiation t;>r JS set [ 15, 161. The gels were then saturated with ethidium bromide by staining for 3 h, and the background was reduced by destaining in H,O for 3 h. The gel> were photographed under short wavelength W illumination and the negatives were scanned with a LKB Ultroscan XL Laser Densitometer and analyzed by the GelScan XL Laser E&n\:.. tometer Program.
RESULTS Oligodeoxynucleotide
Cleavage
Analysis
The oligodeoxynucleotide (32P)pdT, is hydrolysis by E. coli Topoisomerase f to (32P)pdT4 and unlabeled pdTa. The initial rates of hydrolysis were monitored. The fraction of product formed at each time point was determined by analyzing the distribution of radioactivity between substrate and product after separation on a sequencing gel. The amount of product formed is the product of the initial substrate concentration and the fraction of product formed. The cleavage rate, were determined by least-squares analysis of product vs time plots and are reported as the average of triplicates. Figure i shows a positive, linear relationship between the concentration of Mg’ ‘-_ from 0 to 15 mM, and the rate of pdT, cleavage. Linear regression yields a slope of 5.33 x 1O-3 ~Mjmin,/mM Mgzc. and a y-intercept of !).OS5 yM lmin with a correlation of 0.96. Table I presents the effects of various metals at 2 mM on the rate of oligodeoxynucleotide cleavage and the effects of co.-incubating equimolar (2 mM) concentration of Mg2 + with each of the other metals. Mg’-+ accelerated the rate of cleavage and was the only metal to show increased activity relative to the rate of cleavage in the absence of all divalent metals. Mn’ ’ dnd Ca’ ’ showed slight inhibition but both Co’ ’ and Zs-? showed nearly complete inhibition of cleavage activity. In the co-incubation studies, the presence of Mg” i had no effect on the degree of inhibition of the cleavage rate caused by the other metals except in the case of [ME’++ Mn*+ 1. Relaxation
of Supercoiled
DNA Analysis
E. co/i Topoisomerase I partially relaxes negatively supercoiled DNA. The effects of various metals on the relaxation activity was determined by agarose gel electrophoresis where the migration rate of the DNA is related to its degree of supercoil. Photographs of the ethidium bromide stained agarose gels were analyzed by laser densitometry
DIVALENT
ION EFFECTS
ON
E. coli TOPOISOMERASE
91
0.14-
0 O.lOI % 0.06.2 e *: a 0.06-
12
3
4
5
6
7
W(II)l,
6
9
10
11.
12
13
14
15
mt4
cleavage. FIGURE 1. The effect of Mg2+ concentration on the rate of oligodeoxynucleotide The graph is a plot of PM (S-“P)dT, formed per minute as a function of Mg*+ concentration from 0 to 15 mM. The reaction was run at 37°C with 20 PM (5’-32P)pdTe and 0.5 PM enzyme at a constant ionic strength of 50 mM. See text for further details.
The relaxation activity as a function of Mg2+ concentration on agarose is shown in Figure 2. Figure 3 is a representative laser densitometric scan of the relaxation activity at 0.25 and 2.5 mM Mg 2+ . In the figure, all the DNA topoisomers can be distinguished. The percent relaxation was determined by dividing the distance between the supercoiled band, S. C.; and the weighted center of the partially relaxed band, P. R. ; by the distance between the supercoiled band, S. C. ; and the fully relaxed band, F. R. or (S. C.-P. R.)/S. C.-F. R.). Under the conditions of the relaxation assay, the rate of relaxation is constant; therefore, the percent relaxation, as described above, is proportional to a rate constant for relaxation. TABLE 1. The Effects of Various Divalent Metals on the Rate of Oligodeoxynucleotide Cleavage by E. coli Topoisomerase I
Metal
Rate’
Relative Rate*
Mg2+ No Metal MI?+ Ca2+ co2+ Zn2+ Mn’++ Mg2+ Ca’++ Mg*+ Co*++ Mg2+ Zn2++ Mg2+
1.46 1.07 0.610 0.362 0.053 0.005 0.840 0.335 0.045 0.004
1.370 1.000 0.573 0.339 0.038 0.005 0.789 0.314 0.042 0.004
‘nmol of (Y3’P)pdT4 formed per minute per mg of enzyme. ‘The ratio of the rate in the presence of that entries metal conditions to the rate in the absence of all metals.
92
P. L. Domanico
and Y.-C. Tse-Dinh
FIGURE 2. The dependency of relaxation activity on Mg2’ concentration: agarose gel analysis. The gel shows the relaxation activity as a function of Mg”. concentration from 0.25 to 10 mM. The reaction was run at 37°C with 40 PM PM2 supercoiled DNA (in base pairs) and 6 nM enzyme at a constant ionic strength of 50 mM. S. C. represents the fully supercoiled DNA species. P. R. represents the partially relaxed DNA species. F. R. represents the fully relaxed and nicked DNA species. W. represent the location of the well.
Figure 4 is a plot of the percent relaxation as a function of the concentration of Mg”+ at 4 and 6 nM enzyme concentrations. The figure indicates a positive relationship between the concentration of Mg’+, from 0 to 10 mM, and the percent relaxation as well as the sensitivity of the percent relaxation to the amount of enzyme. The figure also demonstrates near saturation of the metal requirement in the relaxation reaction is achieved by Mg2+ at approximately 2 .S mM. FIGURE 3. Representative laser densitometry scan of agarose gel in
P.R
FR
relaxation activity assay. A laser densitometric scan was taken of lanes I (0.25 mM Mg”, curve A) and lane 6 (2.5 mM Mg"' , curve B) from Figure 2. The gel was scanned down from the well and the densitometry scan is read from right to left. S. C. represents the fully supercoiied DNA species. P. R. represents the partially relaxed DNA species. F. R. :epresents the fully relaxed and nicked DNA species. The optical density was determined at 633 nm
94
P. L. Domanico and Y.-C. Tse-Dinh
DISCUSSION By studying the effects of various metals on both the relaxation of negatively supercoiled DNA and the cleavage of the oligodeoxynucleotide substrate pdT,, which is single stranded, the importance of the overall DNA structure and the various DNA regions that the enzyme recognizes may be determined. Fourier Transform infra-red spectroscopy has been used to show that Mg** and Co’+, and high NaCl concentrations cause transitions such as a B to Z conformational change in DNA [17]. These metals induce changes in the ribose sugar ring pucker, the rotomer configuration around the C4’-C5’ bond, as well as more subtle conformational changes involving interaction between the metal and nitrogen atoms in the bases of DNA, In addition, the structure of DNA can be characterized by two circular dichroism spectral signatures-namely the (+) or ( - ) PHI-structures. The alkali earth metals which only bind to the phosphates in the DNA backbone, and univalent metals at high concentrations have little impact upon the pre-existing PHl-structure except to enhance that signature [I 81. On the other hand, the transition metals which show affinity for the phosphate backbone and the purine bases. and to a lesser extent, the pyrimidine bases can have a profound impact upon the PHI-structure where different transition metals promoting different PHI-structures [19, Xl]. These deformations are believed to arise primarily from disruption of hydrogen bonds between bases and/or disruption of base stacking upon metal complexation with the bases 121). The extent of long range deformation is dependent on the metal to DNA concentration ratio since the amount of metal coordinated to the bases of DNA increases as the molar ratio of the metal to the DNA monomer increases 1211. Finally, the effect a particular metal has on DNA conformation is highly dependent on whether the DNA is double or single stranded 1221. Since E. cob Topoisomerase I binds to both single and double stranded regions of DNA, the effect of different metals on the relaxation of supercoiled DNA will be due t.o the combined effect these metals have on double and single stranded DNA. The effects of MgZ + concentration on the relaxation of supercoiled DNA and the cleavage of the oligodeoxynucleotide by E. coii Topoisomerase I are quite dissimilar. The increase in the rate of the cleavage reaction is linearly proportional to the Mg’+ concentration over the 0 to 15 mM range. However, the relaxation activity increases as the Mg’ ’ concentration is raised but then plateaus off--with the relaxation activity nearly insensitive to the Mg*’ concentration in the 2.5 and 10 mM range. This is understandable since the reactants in these two reactions are not the same. Figure 5 depicts a model emphasizing one of the differences between these two reactions. In the relaxation reaction, the 3’-OH group of the DNA backbone which is 5’ of the point of transesterification is the final nucleophile in the religation step of the reaction scheme. But, in the cleavage reaction, that part of the oligodeoxynucleotide which is 5’ of the point of transesterification disassociates from the enzyme [6] and water is the final nucleophiie in the hydrolysis step of the reaction scheme. This may be the fundamental difference between the relaxation and the cleavage reactions with respect to their Mg ’ ’ requirements; that is, the concentration of Mg2 + required to saturate the relaxation reaction, where a ribose hydroxyl group is the nucleophile, is much lower than the concentration of Mg2 ’ required to saturate the cleavage reaction, where water is the nucleophile. There are general similarities in the abilities of metals other than Mg*+ when
96
P. L. Dornanico and Y.-C. Tse-Dinh
that the metal acts as an allosteric activator which initiates isomerization of the complex to an active state. We are currently investigating which steps in the hydrolysis of oligodeoxynucleotides are Mg *+ sensitive and the extent to which the conformation of the enzyme itself and protein-DNA complex is sensitive to the Mg’+ present. The authors are grateful IO Rita Beran-Steed for supplying pure enzyme, and to Caroline Braun and Helen Tabb for technical assistance. We also thank Drs. Stephen Brenner and James B. Matthew for critical reading of the manuscript.
REFERENCES Science 206, 1081-1083 (1979); J. C. Wang, Ann. Rev. Biochem. 54, 665-697 (1985). 3 3. C. Wang, J. Mol. Biol. 55, 523-533 (1971). i. 3. M. Gellert, Annu. Rev. Biochem. 50, 879-910 (1981). 4. P. 0. Brown and N. R. Cozzarelli, Proc. Natl. Acad. Sci. USA 78, 843-847 (1981). 5. K. Kirkegaard, G. Pleufelder. and J. C. Wang. ColOr Spring Harbor Symp. on Quant. 1. P. 0. Brown and N. R. Cozzarelli,
6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Biol. Y.-C. Y.-C. R. E. D. E.
49, 411-419 (1984). Tse-Dinh, J. Biol. Chem. 261, 10931-10935 (1986). Tse, K. Kirkegaard, and J. C. Wang, J. Biol. Chem. 255. 5560-5565 (1980). Depew and J. C. Wang. Proc. Natl. Acad. Sci. USA 72, 4275-4279 (1975). Pulleyblank. M. Shure, D. Tang, J. Vinograd, H.-P. Vosberg, Proc. Natl. Acad. Sci. USA 72, 4280-4284 (1975). W. Bauer, J. Vinograd, J. Mol. Biol. 47, 419-435 (1970). T.-S. Hsieh and J. C. Wang, Biochemistry 14. 527-535 (1975). Y.-C. Tse-Dinh, B. G. H. McCarron, R. Arentzen, and V. Chowdhry, Nucleic Acids Res. 11, 8691-8701 (1983). Y.-C. Tse-Dinh and R. K. Beran-Steed, J. Biof. Chem. 263, 15857-1.5859 (1988). T. Maniatas, S. G. Kce, E. Lacy. J. Lauer, C. O’Connell, I>. Quon, I). K. Sim. and A. Efstratiadis, Cell 8, 1630 (1978). I. S. Deniss and A. R. Morgan, Nucleic Acids Res. 3, 315-323 (1976). E. Goldstein and K. Drlica, Proc. Natl. Acad. Sci. USA 81, 4046-4050 (1984). T. Theophanides and H. A. Tajmir-Riahi, .I. Biomol. Struci. Dyer. 2,995- 1004 (1985). W. F. Dove and N. Davidson, J. Mol. Biol. 5, 467-478 (1962). R. M. Izatt, J. J. Christensen, and J, H. Rytting, Chem. Kev. 71, 439-481 (1971). G. L. Eichhorn, in inorganic Biochemistry, G. L. Eichhorn. Ed., Elsevier, New York, 1973, Vol. 2, pp. i121O-1243. H. Shirai, Y. Itoh, A. Kurose, K. Hanabusa, K. Abe. and N. Hojo, Polym. .I. (Tokyo) 16, 207-215 (1984). J. E. Morgan, J. W. Blankenship, and H. R. Matthews, Arch. Biochem. Biophys. 246,
225-232 (1986). 23. J. A. McClarin, C. A. Rederick, B.-C. Wang, P. Greene, H. W. Boyer, J. Grable. and J. M. Rosenberg, Science 234, 1526-1541 (1986).
Received September 14, 1990; accepted October I?. 1990
PLD.
ABSTRACT E. coli DNA topoisomerase I catalyzes the hydrolysis of short, single stranded oligodeoxynucleotides. It also forms a covalent protein-DNA complex with negatively supercoiled DNA in the absence of Mg 2+ but requires Mg ‘+ for the relaxation of negatively supercoiled DNA. In this paper we investigate the effects of various divalent metals on catalysis. For the relaxation reaction, maximum enzyme activity plateaus after 2.5 mM Mg’+. However, the rate of cleavage of short oligodeoxynucleotide increased linearly between 0 and 15 mM Mg’+. In the oligodeoxynucleotide cleavage reaction, Ca’+, Mn2+, Co*+, and Zn*+ inhibit enzymatic activity. When these metals are coincubated with Mg2+ at equimolar concentrations, the normal effect of Mg ‘+ is not detectable. Of these metals, only Ca2+ can be substituted for Mg2 + as a metal cofactor in the relaxation reaction. And when Mg2+ is coincubated with Mn’+, Co’+, or Zn2+ at equimolar concentrations, the normal effect of Mg2+ on relaxation is not complex to assume a conformation detectable. We propose that Mg 2+ allows the protein-DNA necessary for strand passage and enhance the rate of enzyme turnover.
INTRODUCTION E. cofi Topoisomerase
I relaxes negatively supercoiled DNA and increases the linking number of the DNA product in units of one [l]. The enzyme is believed to catalyze this phenomena by breaking one strand of the DNA duplex in a single stranded region, passing the other, intact strand through the break, and finally rejoining the broken strand [2, 31. A model involving two binding domains has been proposed for the interaction of bacterial topoisomerase I with DNA [4-61. One domain, which forms a covalent bond to DNA, contains a nucleophilic tyrosine residue that attacks a phosphate group on the DNA substrate displacing the 3’-OH moiety of the upstream ribosyl group during the strand breakage step [7]. The other
Address reprint requests and correspondence to: Paul L. Domanico, Department of Enzymology, Research Laboratories,
Glaxo
5 Moore Drive, Research Triangle Park, NC 27709.
Journal of Inorganic Biochemistry, 42, W-96 (1991) 0 1991 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, NY, NY 10010
81 0162-0134/91/$3.50
88
P. L. Domanico and Y.-C. Tse-Dinh
domain binds noncovalently to DNA. The type of interaction between the DNA and the noncovalent binding domain is not clear. This domain interacts with the broken DNA strand and perhaps the unbroken strand and may position the 3’-OH group during the religation step. Furthermore, an energy source such as ATP or NAD is not required since the free energy within the DNA phosphodiester bond is conserved as a result of the transesterification reaction. The efficiency with which E. coli Topoisomerase I relaxes supercoiled DNA is highly dependent on the degree of 3NA supercoiling and the energy inherently stored within the supercoiled molecule. The degree of single stranded character in the duplex DNA and the free energy stored in the supercoiled molecule are proportional to the amount of negative supercoiling [8- 111. This could effect both the binding of E. coli Topoisomerase I to the DNA because the enzyme prefers single stranded DNA and the number of covalent intermediates that proceed through the strand passage step if this step was sensitive to the state of supercoiling. These arguments help explain the observation that the efficiency of relaxing negatively supercoiled DNA is higher with a more negatively supercoiled substrate (21. With a small, single stranded oligodeoxynucleotide as a substrate, the strand upstream of the point of transesterification is released after cleavage [6, 121. The small size of the strand 5’ to the cleavage site may be cause for its disassociation from the enzyme. Hydrolysis of the phosphodiester bond between the active site tyrosine residue and the single stranded DNA results in enzyme ‘turnover [6]. Hydrolysis does not occur with double stranded DNA as the substrate. This suggests that the uncleaved DNA strand assists in the transesterification reaction by anchoring that region of the cleaved DNA strand not covalently bound to the enzyme. Not all steps in the relaxation of supercoiled DNA by E. cali Topoisomerase I require metal as a cofactor 13, 121. Mg 2’ is required for relaxation of supercoiled DNA and other reactions involving strand passage. Whereas, the formation of the covalent complex does not require any exogenous Mg2 ‘. This study focuses on the role of metal in the catalysis of relaxation of supercoiled DNA, and the cleavage of short, single stranded oligodeoxynucleotides by E. coii Topoisomerase I- where cleavage of oligodeoxynucleotide is intended to mimic covalent complex formation during transesterification of supercoiled DNA. The relaxation reaction was studied with the negatively supercoiled substrate, PM2 DNA. The cleavage reaction was studied with a short oligodeoxynucleotide substrate, the thymidine octamer where cleavage by E. colt’ Topoisomerase 1 occurs at the phosphodiester linkage between the 4th and 5th nucleotides. MATERIALS Deionized formamide was from Fluka. Agarose was from BIORAD. Electrophoresis grade TEMED, 19: 1 acrylamide: N,N’-methylene bisacrylamide and urea were from BRL. Oligodeoxynucleotide pdT, was from Pharmacia. (Y-~‘P) ATP (10 mCi/ml) and NENSORB TM 20 columns were from NEN. T4 polynucleotide kinase was from New England BioLabs. Scinti Verse ITM was from Fisher, and negatively supercoiled PM2 DNA was from Boehringer Mannheim. All other organic reagents, buffers, and inorganic salts were commercially available reagent-grade chemicals. In all experiments, distilled deionized water was used. 10X Kinase bu&r: SO mM DTT, 100 mM MgCl,, and 700 mM Tris * HCl at pH 7.5. 10 X Relaxation bu@er: 1 mg/ml gelatin, 100 mM .@-mercaptoethanol, and
90
P. L. Domanico
and Y.-C.
Kye-Dinh
tration was 2 mM. The ionic strength of the cleavage reaction wa! maintained at SO mM with NaCl. The reaction solution minus the enzyme, and the stock enzyme solution were pre-incubated for five minutes at 37°C. The cleavage reaction was initiated by the addition of enzyme (10-80 ,ug/ml) and maintained at 37”C, and then quenched after 30 min by the addition of 5 ~1 SIX stop solutir~n Analysis
of Supercoiled
PM2 DNA Relaxation
Twenty ~1 of each sample was loaded into a 0.7% agarose gel and electrophoresed at 70 V, constant voltage. for I& 16 h. The gels were then lightly stained with ethidium bromide (2.5 pg/ml) for 5 min. The covalcntly closed PM2 DNA circles were then nicked by exposure to long wavelength UV irradiation t;>r JS set [ 15, 161. The gels were then saturated with ethidium bromide by staining for 3 h, and the background was reduced by destaining in H,O for 3 h. The gel> were photographed under short wavelength W illumination and the negatives were scanned with a LKB Ultroscan XL Laser Densitometer and analyzed by the GelScan XL Laser E&n\:.. tometer Program.
RESULTS Oligodeoxynucleotide
Cleavage
Analysis
The oligodeoxynucleotide (32P)pdT, is hydrolysis by E. coli Topoisomerase f to (32P)pdT4 and unlabeled pdTa. The initial rates of hydrolysis were monitored. The fraction of product formed at each time point was determined by analyzing the distribution of radioactivity between substrate and product after separation on a sequencing gel. The amount of product formed is the product of the initial substrate concentration and the fraction of product formed. The cleavage rate, were determined by least-squares analysis of product vs time plots and are reported as the average of triplicates. Figure i shows a positive, linear relationship between the concentration of Mg’ ‘-_ from 0 to 15 mM, and the rate of pdT, cleavage. Linear regression yields a slope of 5.33 x 1O-3 ~Mjmin,/mM Mgzc. and a y-intercept of !).OS5 yM lmin with a correlation of 0.96. Table I presents the effects of various metals at 2 mM on the rate of oligodeoxynucleotide cleavage and the effects of co.-incubating equimolar (2 mM) concentration of Mg2 + with each of the other metals. Mg’-+ accelerated the rate of cleavage and was the only metal to show increased activity relative to the rate of cleavage in the absence of all divalent metals. Mn’ ’ dnd Ca’ ’ showed slight inhibition but both Co’ ’ and Zs-? showed nearly complete inhibition of cleavage activity. In the co-incubation studies, the presence of Mg” i had no effect on the degree of inhibition of the cleavage rate caused by the other metals except in the case of [ME’++ Mn*+ 1. Relaxation
of Supercoiled
DNA Analysis
E. co/i Topoisomerase I partially relaxes negatively supercoiled DNA. The effects of various metals on the relaxation activity was determined by agarose gel electrophoresis where the migration rate of the DNA is related to its degree of supercoil. Photographs of the ethidium bromide stained agarose gels were analyzed by laser densitometry
DIVALENT
ION EFFECTS
ON
E. coli TOPOISOMERASE
91
0.14-
0 O.lOI % 0.06.2 e *: a 0.06-
12
3
4
5
6
7
W(II)l,
6
9
10
11.
12
13
14
15
mt4
cleavage. FIGURE 1. The effect of Mg2+ concentration on the rate of oligodeoxynucleotide The graph is a plot of PM (S-“P)dT, formed per minute as a function of Mg*+ concentration from 0 to 15 mM. The reaction was run at 37°C with 20 PM (5’-32P)pdTe and 0.5 PM enzyme at a constant ionic strength of 50 mM. See text for further details.
The relaxation activity as a function of Mg2+ concentration on agarose is shown in Figure 2. Figure 3 is a representative laser densitometric scan of the relaxation activity at 0.25 and 2.5 mM Mg 2+ . In the figure, all the DNA topoisomers can be distinguished. The percent relaxation was determined by dividing the distance between the supercoiled band, S. C.; and the weighted center of the partially relaxed band, P. R. ; by the distance between the supercoiled band, S. C. ; and the fully relaxed band, F. R. or (S. C.-P. R.)/S. C.-F. R.). Under the conditions of the relaxation assay, the rate of relaxation is constant; therefore, the percent relaxation, as described above, is proportional to a rate constant for relaxation. TABLE 1. The Effects of Various Divalent Metals on the Rate of Oligodeoxynucleotide Cleavage by E. coli Topoisomerase I
Metal
Rate’
Relative Rate*
Mg2+ No Metal MI?+ Ca2+ co2+ Zn2+ Mn’++ Mg2+ Ca’++ Mg*+ Co*++ Mg2+ Zn2++ Mg2+
1.46 1.07 0.610 0.362 0.053 0.005 0.840 0.335 0.045 0.004
1.370 1.000 0.573 0.339 0.038 0.005 0.789 0.314 0.042 0.004
‘nmol of (Y3’P)pdT4 formed per minute per mg of enzyme. ‘The ratio of the rate in the presence of that entries metal conditions to the rate in the absence of all metals.
92
P. L. Domanico
and Y.-C. Tse-Dinh
FIGURE 2. The dependency of relaxation activity on Mg2’ concentration: agarose gel analysis. The gel shows the relaxation activity as a function of Mg”. concentration from 0.25 to 10 mM. The reaction was run at 37°C with 40 PM PM2 supercoiled DNA (in base pairs) and 6 nM enzyme at a constant ionic strength of 50 mM. S. C. represents the fully supercoiled DNA species. P. R. represents the partially relaxed DNA species. F. R. represents the fully relaxed and nicked DNA species. W. represent the location of the well.
Figure 4 is a plot of the percent relaxation as a function of the concentration of Mg”+ at 4 and 6 nM enzyme concentrations. The figure indicates a positive relationship between the concentration of Mg’+, from 0 to 10 mM, and the percent relaxation as well as the sensitivity of the percent relaxation to the amount of enzyme. The figure also demonstrates near saturation of the metal requirement in the relaxation reaction is achieved by Mg2+ at approximately 2 .S mM. FIGURE 3. Representative laser densitometry scan of agarose gel in
P.R
FR
relaxation activity assay. A laser densitometric scan was taken of lanes I (0.25 mM Mg”, curve A) and lane 6 (2.5 mM Mg"' , curve B) from Figure 2. The gel was scanned down from the well and the densitometry scan is read from right to left. S. C. represents the fully supercoiied DNA species. P. R. represents the partially relaxed DNA species. F. R. :epresents the fully relaxed and nicked DNA species. The optical density was determined at 633 nm
94
P. L. Domanico and Y.-C. Tse-Dinh
DISCUSSION By studying the effects of various metals on both the relaxation of negatively supercoiled DNA and the cleavage of the oligodeoxynucleotide substrate pdT,, which is single stranded, the importance of the overall DNA structure and the various DNA regions that the enzyme recognizes may be determined. Fourier Transform infra-red spectroscopy has been used to show that Mg** and Co’+, and high NaCl concentrations cause transitions such as a B to Z conformational change in DNA [17]. These metals induce changes in the ribose sugar ring pucker, the rotomer configuration around the C4’-C5’ bond, as well as more subtle conformational changes involving interaction between the metal and nitrogen atoms in the bases of DNA, In addition, the structure of DNA can be characterized by two circular dichroism spectral signatures-namely the (+) or ( - ) PHI-structures. The alkali earth metals which only bind to the phosphates in the DNA backbone, and univalent metals at high concentrations have little impact upon the pre-existing PHl-structure except to enhance that signature [I 81. On the other hand, the transition metals which show affinity for the phosphate backbone and the purine bases. and to a lesser extent, the pyrimidine bases can have a profound impact upon the PHI-structure where different transition metals promoting different PHI-structures [19, Xl]. These deformations are believed to arise primarily from disruption of hydrogen bonds between bases and/or disruption of base stacking upon metal complexation with the bases 121). The extent of long range deformation is dependent on the metal to DNA concentration ratio since the amount of metal coordinated to the bases of DNA increases as the molar ratio of the metal to the DNA monomer increases 1211. Finally, the effect a particular metal has on DNA conformation is highly dependent on whether the DNA is double or single stranded 1221. Since E. cob Topoisomerase I binds to both single and double stranded regions of DNA, the effect of different metals on the relaxation of supercoiled DNA will be due t.o the combined effect these metals have on double and single stranded DNA. The effects of MgZ + concentration on the relaxation of supercoiled DNA and the cleavage of the oligodeoxynucleotide by E. coii Topoisomerase I are quite dissimilar. The increase in the rate of the cleavage reaction is linearly proportional to the Mg’+ concentration over the 0 to 15 mM range. However, the relaxation activity increases as the Mg’ ’ concentration is raised but then plateaus off--with the relaxation activity nearly insensitive to the Mg*’ concentration in the 2.5 and 10 mM range. This is understandable since the reactants in these two reactions are not the same. Figure 5 depicts a model emphasizing one of the differences between these two reactions. In the relaxation reaction, the 3’-OH group of the DNA backbone which is 5’ of the point of transesterification is the final nucleophile in the religation step of the reaction scheme. But, in the cleavage reaction, that part of the oligodeoxynucleotide which is 5’ of the point of transesterification disassociates from the enzyme [6] and water is the final nucleophiie in the hydrolysis step of the reaction scheme. This may be the fundamental difference between the relaxation and the cleavage reactions with respect to their Mg ’ ’ requirements; that is, the concentration of Mg2 + required to saturate the relaxation reaction, where a ribose hydroxyl group is the nucleophile, is much lower than the concentration of Mg2 ’ required to saturate the cleavage reaction, where water is the nucleophile. There are general similarities in the abilities of metals other than Mg*+ when
96
P. L. Dornanico and Y.-C. Tse-Dinh
that the metal acts as an allosteric activator which initiates isomerization of the complex to an active state. We are currently investigating which steps in the hydrolysis of oligodeoxynucleotides are Mg *+ sensitive and the extent to which the conformation of the enzyme itself and protein-DNA complex is sensitive to the Mg’+ present. The authors are grateful IO Rita Beran-Steed for supplying pure enzyme, and to Caroline Braun and Helen Tabb for technical assistance. We also thank Drs. Stephen Brenner and James B. Matthew for critical reading of the manuscript.
REFERENCES Science 206, 1081-1083 (1979); J. C. Wang, Ann. Rev. Biochem. 54, 665-697 (1985). 3 3. C. Wang, J. Mol. Biol. 55, 523-533 (1971). i. 3. M. Gellert, Annu. Rev. Biochem. 50, 879-910 (1981). 4. P. 0. Brown and N. R. Cozzarelli, Proc. Natl. Acad. Sci. USA 78, 843-847 (1981). 5. K. Kirkegaard, G. Pleufelder. and J. C. Wang. ColOr Spring Harbor Symp. on Quant. 1. P. 0. Brown and N. R. Cozzarelli,
6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Biol. Y.-C. Y.-C. R. E. D. E.
49, 411-419 (1984). Tse-Dinh, J. Biol. Chem. 261, 10931-10935 (1986). Tse, K. Kirkegaard, and J. C. Wang, J. Biol. Chem. 255. 5560-5565 (1980). Depew and J. C. Wang. Proc. Natl. Acad. Sci. USA 72, 4275-4279 (1975). Pulleyblank. M. Shure, D. Tang, J. Vinograd, H.-P. Vosberg, Proc. Natl. Acad. Sci. USA 72, 4280-4284 (1975). W. Bauer, J. Vinograd, J. Mol. Biol. 47, 419-435 (1970). T.-S. Hsieh and J. C. Wang, Biochemistry 14. 527-535 (1975). Y.-C. Tse-Dinh, B. G. H. McCarron, R. Arentzen, and V. Chowdhry, Nucleic Acids Res. 11, 8691-8701 (1983). Y.-C. Tse-Dinh and R. K. Beran-Steed, J. Biof. Chem. 263, 15857-1.5859 (1988). T. Maniatas, S. G. Kce, E. Lacy. J. Lauer, C. O’Connell, I>. Quon, I). K. Sim. and A. Efstratiadis, Cell 8, 1630 (1978). I. S. Deniss and A. R. Morgan, Nucleic Acids Res. 3, 315-323 (1976). E. Goldstein and K. Drlica, Proc. Natl. Acad. Sci. USA 81, 4046-4050 (1984). T. Theophanides and H. A. Tajmir-Riahi, .I. Biomol. Struci. Dyer. 2,995- 1004 (1985). W. F. Dove and N. Davidson, J. Mol. Biol. 5, 467-478 (1962). R. M. Izatt, J. J. Christensen, and J, H. Rytting, Chem. Kev. 71, 439-481 (1971). G. L. Eichhorn, in inorganic Biochemistry, G. L. Eichhorn. Ed., Elsevier, New York, 1973, Vol. 2, pp. i121O-1243. H. Shirai, Y. Itoh, A. Kurose, K. Hanabusa, K. Abe. and N. Hojo, Polym. .I. (Tokyo) 16, 207-215 (1984). J. E. Morgan, J. W. Blankenship, and H. R. Matthews, Arch. Biochem. Biophys. 246,
225-232 (1986). 23. J. A. McClarin, C. A. Rederick, B.-C. Wang, P. Greene, H. W. Boyer, J. Grable. and J. M. Rosenberg, Science 234, 1526-1541 (1986).
Received September 14, 1990; accepted October I?. 1990
