Effect of 2 * -5 * Phosphodiesters on DNA Transesterification by Vaccinia Topoisomerase*

Vaccinia topoisomerase forms a covalent DNA-(3 * -phosphotyrosyl)-enzyme intermediate at a pentapyrimidine target site 5 * -CCCTTp 2 in duplex DNA. By introducing single 2 * -5 * phosphodiesters in lieu of a standard 3 * -5 * phosphodiester linkage, we illuminate the contributions of phosphodiester connectivity to DNA transesterification. We find that the DNA cleavage reaction was slowed by more than six orders of magnitude when a 2 * -5 * linkage was present at the scissile phosphodiester (CCCTT 2 * p 2 5 * A). Thus, vaccinia topoisomerase is unable to form a DNA-(2 * -phosphotyrosyl)-enzyme intermediate. We hypothesize that the altered geometry of the 2 * -5 * phosphodiester limits the ability of the tyrosine nucleophile to attain a requisite, presumably apical orientation with respect to the 5 * -OH leaving group. A 2 * -5 * phosphodiester located to the 3 * side of the cleavage site (CCCTTp 2 N 2 * p 5 * N) reduced the rate of transesterification by a factor of 500. In contrast, 2 * -5 * phosphodiesters at four other sites in the scissile strand (TpCGCCCTpT 2 ATpTpC) and five positions in the nonscissile strand (3 * -GGGpApApTpApA)

Vaccinia topoisomerase forms a covalent DNA-(3phosphotyrosyl)-enzyme intermediate at a pentapyrimidine target site 5-CCCTTp2 in duplex DNA. By introducing single 2-5 phosphodiesters in lieu of a standard 3-5 phosphodiester linkage, we illuminate the contributions of phosphodiester connectivity to DNA transesterification. We find that the DNA cleavage reaction was slowed by more than six orders of magnitude when a 2-5 linkage was present at the scissile phosphodiester (CCCTT 2 p2 5 A). Thus, vaccinia topoisomerase is unable to form a DNA-(2phosphotyrosyl)-enzyme intermediate. We hypothesize that the altered geometry of the 2-5 phosphodiester limits the ability of the tyrosine nucleophile to attain a requisite, presumably apical orientation with respect to the 5-OH leaving group. A 2-5 phosphodiester located to the 3 side of the cleavage site (CCCTTp2N 2 p 5 N) reduced the rate of transesterification by a factor of 500. In contrast, 2-5 phosphodiesters at four other sites in the scissile strand (TpCGCCCTpT2ATpTpC) and five positions in the nonscissile strand (3-GGGpApApTpApA) had no effect on transesterification rate. The DNAs containing 2-5 phosphodiesters were protected from digestion by exonuclease III. We found that exonuclease III was consistently arrested at positions 1 and 2 nucleotides prior to the encounter of its active site with the modified 2-5 phosphodiester and that the 2-5 linkage itself was poorly hydrolyzed by exonuclease III.
Type IB topoisomerases modulate the topological state of DNA by cleaving and rejoining one strand of the DNA duplex. Cleavage is a transesterification reaction in which the scissile Np2N phosphodiester is attacked by a tyrosine of the enzyme, resulting in the formation of a DNA-(3Ј-phosphotyrosyl)enzyme intermediate and the expulsion of a DNA strand having a 5Ј-OH terminus. In the religation step, the DNA 5Ј-OH group attacks the covalent intermediate resulting in expulsion of the active site tyrosine and restoration of the DNA phosphodiester backbone. Vaccinia topoisomerase is a prototype of the type IB topoisomerase family (1). The poxvirus enzyme is distinguished from the nuclear topoisomerase I by its compact size (314 amino acids) and its site-specificity in DNA transesterification. Vaccinia topoisomerase binds and cleaves duplex DNA at a pentapyrimidine target sequence 5Ј-(T/C)CCTT2.
The Tp2 nucleotide (defined as the ϩ1 nucleotide) is linked to Tyr-274 of the enzyme.
The stereochemical outcome of the net cleavage-religation reaction of vaccinia topoisomerase is a retention of configuration at the scissile phosphodiester. This suggests that the component cleavage and religation reactions entail in-line S N 2-type displacements in which the attacking nucleophile is apical to the leaving group and each transesterification results in an inversion of configuration (2). Four conserved amino acid side chains of vaccinia topoisomerase (Arg-130, Lys-167, Arg-223, and His-265) are required for catalysis (3)(4)(5). Mutational and structural data suggest that the two arginines and the histidine interact directly with the scissile phosphodiester and enhance catalysis by stabilizing the developing negative charge on a pentacoordinate phosphorane transition state (2)(3)(4)(5)(6)(7). Lys-167 serves as a general acid catalyst during the cleavage reaction, donating a proton to expel the 5Ј-OH leaving group (8).
Type IB topoisomerases engage the DNA target site circumferentially, forming a C-shaped clamp around the duplex (6,7,9). The DNA moieties that contribute to target site recognition and enable catalysis by vaccinia topoisomerase have been examined using synthetic substrates containing a single CCCTT site. Modification interference, modification protection, base and sugar analog substitution, and UV cross-linking experiments indicate that vaccinia topoisomerase makes contact with specific bases and phosphates of DNA in the vicinity of the CCCTT element (9 -12). For example, dimethyl sulfate protection and interference experiments revealed interactions with the three guanine bases of the pentamer motif complementary strand (3Ј-GGGAA) (10), and permanganate oxidation interference highlighted a functional interaction with the ϩ2T base of the scissile strand (CCCTT) (12).
Functionally relevant phosphates were initially identified by studying the effects of phosphate ethylation on topoisomerase binding (9). Ethylation of the ϩ1, ϩ2, ϩ3, and ϩ4 phosphates on the scissile strand (positions CpCpTpTp2 within the pentamer motif) and the ϩ3, ϩ4, and ϩ5 phosphates on the nonscissile strand (3Ј-GpGpGpA) interfered with topoisomerase-DNA complex formation. The relevant topoisomerase-phosphate contacts are arrayed across the minor groove of the DNA helix (9). Phosphate ethylation is a relatively crude modification interference method, insofar as it simultaneously eliminates the negative charge on the phosphate and introduces a bulky aliphatic group. Subsequent studies of the catalytic contributions of individual phosphates have entailed less drastic modifications, for example replacing the standard 3Ј-5Ј phosphodiester by a 3Ј-OH/5Ј-PO 4 nick (13). This modification interrupts the DNA backbone and introduces additional negative charge (net charge of about Ϫ1.5 at pH 7.0 at the nick versus Ϫ1.0 for the phosphodiester) but adds only minimal bulk (one extra oxygen). A 3Ј-OH/5Ј-PO 4 nick in lieu of the scissile phosphodiester abolished transesterification by vaccinia topoisomerase, but did not affect the noncovalent binding of topoisomerase to the nicked DNA (13). This result indicated that vaccinia topoisomerase is an obligate nucleotidyl-3Јphosphotransferase that cannot transesterify to a 5Ј phosphomonoester. Placement of a 3Ј-OH/5Ј-PO 4 nick at position ϩ2 (CCCTpTp2) slowed the rate of transesterification by a factor of 500. A "missing phosphate" analysis of the DNA target site entailed replacement of phosphodiesters with a 3Ј-OH/5Ј-OH nick, a maneuver that eliminates one negative charge along with potential hydrogen bonding interactions between the topoisomerase and the nonbridging phosphate oxygens. This interference method revealed a contribution of the Ϫ1 phosphate of the scissile strand (CCCTTp2NpN), which enhances the rate of transesterification by a factor of 40 (13).
Here we present a more subtle modification interference analysis in which we investigate the positional effects of introducing single 2Ј-5Ј phosphodiesters in lieu of a standard 3Ј-5Ј phosphodiester linkage (14,15). This approach imposes no alteration in net charge or size of the phosphate moiety and is better suited than previous modification methods to dissect the role of phosphate orientation in site recognition and transesterification chemistry. The instructive findings are that vaccinia topoisomerase is insensitive to single 2Ј-5Ј phosphodiester modifications, with two exceptions: the rate of single-turnover transesterification is suppressed by six orders of magnitude by a 2Ј-5Ј substitution for the scissile phosphodiester (CCCTT 2Ј p2 5Ј N) and by a factor of 500 by 2Ј-5Ј modification of the Ϫ1 phosphate (CCCTT2N 2Ј p 5Ј N). Vaccinia topoisomerase displays a more stringent requirement for natural phosphodiester geometry at the cleavage site than mammalian topoisomerase I.

EXPERIMENTAL PROCEDURES
DNA Substrates-Modified oligonucleotides containing a single 2Ј-5Ј phosphodiester bond were synthesized as described (14 -17). Unmodified oligonucleotides were purchased from BIOSOURCE International.
Vaccinia Topoisomerase-Recombinant vaccinia topoisomerase was produced in Escherichia coli(BL21) by infection with bacteriophage CE6 and then purified from the soluble bacterial lysate by phosphocellulose chromatography as described previously (18).
Free DNA migrated near the dye front. Covalent complex formation was revealed by transfer of radiolabeled DNA to the topoisomerase polypeptide. The extent of covalent adduct formation was quantified by scanning the dried gel using a Fujix BAS2500 Phosphorimager. A plot of the percentage of input DNA cleaved versus time established the end point values for cleavage. The data were then normalized to the end point value (defined as 100%), and the cleavage rate constants (k obs ) were calculated by fitting the normalized data to the equation 100 Ϫ %cleavage (norm) ϭ 100e Ϫkt . The values of k obs at the saturating level of input topoisomerase (either 75 or 150 ng) and the actual end point cleavage values are listed in Tables I and II.
Cleavage Site Specificity-Reaction mixtures (20 l) containing 50 mM Tris-HCl (pH 7.5), 0.3 pmol of standard or 2Ј-5Ј modified 18-mer/ 30-mer DNA, and 75 ng of vaccinia topoisomerase were incubated for 15 min at 37°C. A mixture containing DNA modified at Ϫ1 phosphodiester on the 18-mer was incubated for 240 min at 37°C. The reactions were quenched with 1% SDS. Half of the sample was digested for 2 h at 37°C with 10 g of proteinase K and the other half was not digested. The mixtures were adjusted to 47% formamide, heat-denatured, and electrophoresed through a 20% denaturing polyacrylamide gel containing 7 M urea in TBE (90 mM Tris-borate, 2.5 mM EDTA). The reaction products were visualized by autoradiographic exposure of the gel.
Exonuclease III Digestion-5Ј 32 P-labeled standard and 2Ј-5Ј modified oligonucleotides were hybridized to complementary 60-mer DNAs at a 1:4 molar ratio of labeled strand to 60-mer. Exonuclease III reaction mixtures (90 l) containing 66 mM Tris-HCl (pH 8.0), 0.66 mM MgCl 2 , 3 pmol of 18-mer/60-mer or 30-mer/60-mer DNA, and 2.5 units of E. coli exonuclease III (New England Biolabs) were incubated at 22°C. Aliquots (10 l) were withdrawn at the times specified and quenched by adding EDTA to 30 mM final concentration. The samples were adjusted to 47% formamide, heat-denatured, and electrophoresed through a 20% denaturing polyacrylamide gel containing 7 M urea in TBE. The reaction products were visualized by autoradiographic exposure of the gel.

Effects of Single 2Ј-5Ј Phosphodiester Modifications within
the Scissile CCCTT Strand-A series of 18-mer scissile strands containing a single 2Ј-5Ј phosphodiester at positions ϩ8, ϩ2, ϩ1, Ϫ1, Ϫ2, and Ϫ3 were 5Ј 32 P-labeled and annealed to an unlabeled 30-mer strand to form a "suicide" substrate for vaccinia topoisomerase (Fig. 1). Cleavage results in covalent attachment of a 5Ј 32 P-labeled 12-mer (5Ј-pCGTGTCGCCCTTp) to the enzyme via Tyr-274. The unlabeled 6-mer leaving strand 5Ј HO ATTCCC dissociates spontaneously from the protein-DNA complex. Loss of the leaving strand drives the reaction toward the covalent state.
The reaction of topoisomerase with the unmodified control substrate attained an end point of 90% covalent adduct formation, and the reaction was completed within 20 s (Fig. 1). The extent of transesterification after 5 s was 80% of the end point value. From this datum, we calculated a single-turnover cleavage rate constant of 0.3 s Ϫ1 (Table I). Introduction of a 2Ј-5Ј phosphodiester at positions ϩ8, ϩ2, Ϫ2, or Ϫ3 had no significant effect on the reaction end point or the cleavage rate con-  (Table I). In contrast, a 2Ј-5Ј phosphodiester at position Ϫ1 reduced the rate of transesterification by a factor of 500 (k obs ϭ 6.2 ϫ 10 Ϫ4 s Ϫ1 ) (Fig. 1, Table I). The k obs for the Ϫ1 modified substrate did not increase when the concentration of topoisomerase in the reaction mixture was increased 2-fold (not shown). This indicated that the slowed cleavage rate was not caused by a defect in the initial binding of topoisomerase to the substrate.
Placement of a 2Ј-5Ј phosphodiester directly at the cleavage site virtually abrogated the transesterification reaction. A mere 3% of the input DNA was cleaved after a 6-day reaction of topoisomerase with the ϩ1 modified substrate (not shown). Although an end point was clearly not attained, we calculated an initial rate of cleavage of the ϩ1 modified strand of 4.2 ϫ 10 Ϫ6 % s Ϫ1 . This value reflects a rate decrement of at least 10 Ϫ6.5 compared with the initial rate of cleavage of the unmodified control DNA. Thus we can estimate that cleavage rate constant for the ϩ1 modified substrate was Ͻ1 ϫ 10 Ϫ7 s Ϫ1 .
To address whether the 2Ј-5Ј phosphodiester substitutions altered the site of cleavage within the 18-mer scissile strand, the reaction products were digested with proteinase K in the presence of SDS to remove the covalently linked topoisomerase. The radiolabeled DNA reaction products were then analyzed by denaturing polyacrylamide gel electrophoresis (Fig. 2). Reaction of topoisomerase with the unmodified control substrate resulted in the appearance of a cluster of radiolabeled species migrating faster than the input 32 P-labeled 18-mer strand but slower than a free 32 P-labeled 12-mer 5Ј-pCGTGTCGCCCTTp. (The free 12-mer was generated by peroxidolysis of the covalent topoisomerase-DNA intermediate, Ref. 19.) The cluster consists of the 12-mer 5Ј-pCGTGTCGCCCTTp linked to one or more amino acids of the topoisomerase. Detection of the covalent oligonucleotide-peptide complex was completely dependent on prior digestion of the sample with proteinase K (Fig. 2). This is because the labeled DNA does not migrate into the polyacrylamide gel when it is bound covalently to the topoisomerase polypeptide. The instructive finding was that the same cluster was produced by proteinase K digestion of the covalent adduct formed by topoisomerase on the scissile strands containing 2Ј-5Ј phosphodiesters at positions Ϫ3, Ϫ2, Ϫ1, and ϩ2 (Fig. 2). Thus, the site of covalent adduct formation was unchanged by the modifications. Any shift in the cleavage site, and hence the size of the covalently bound oligonucleotide, would have been readily detected by an altered mobility of the labeled oligonucleotide-peptide adducts. Additional experiments verified that the ϩ8 modification also had no effect on the site of cleavage (not shown).
Effects of Single 2Ј-5Ј Phosphodiester Modifications within the Nonscissile Strand-A series of 30-mer nonscissile strands containing a single 2Ј-5Ј phosphodiester at positions ϩ3, ϩ2, ϩ1, Ϫ1, and Ϫ2 were annealed to the unmodified 5Ј 32 P-labeled 18-mer scissile strand (Fig. 1). None of these 2Ј-5Ј phosphodiester modifications had a significant effect on the reaction end point or the cleavage rate constant (Table I). The modifications of the nonscissile strand did not alter the site of cleavage within the 18-mer scissile strand (not shown).

Effects of Combining Scissile Strand 2Ј-5Ј Modifications with Nonscissile Strand 2Ј-5Ј
Modifications-A series of doubly modified substrates containing one 2Ј-5Ј phosphodiester in the scissile strand and one 2Ј-5Ј phosphodiester in the nonscissile strand was prepared by annealing the modified 32 P-labeled 18-mer strands to each of the modified unlabeled 30-mer strands. The substrate combinations and the results of the analysis of transesterification kinetics are shown in Table II.
We found that the modifications of the Ϫ3 and Ϫ2 positions of the scissile strand, which by themselves had no deleterious effect on transesterification, also had no significant effect on the rate of cleavage when combined with any of the five singly substituted complementary strands. (Our operational definition of a significant effect is one that elicits at least a 4-fold change in k obs .) In contrast, the ϩ2 modification of the scissile strand, which also had no deleterious effect per se, was inhibitory when combined with 2Ј-5Ј phosphodiester modifications on the complementary strand that were also benign by themselves (Table II). A hierarchy of synergistic effects was evident, whereby "cross-strand" combination with a 2Ј-5Ј phosphodiester at position ϩ2 on the nonscissile strand elicited a 100-fold decrement in cleavage rate (k obs ϭ 0.002 s Ϫ1 ), combination with modifications at Ϫ1 (k obs ϭ 0.02 s Ϫ1 ) or ϩ3 (k obs ϭ 0.013 s Ϫ1 ) resulted in an order of magnitude rate decrement, and combination with a 2Ј-5Ј linkage at ϩ1 slowed cleavage by a factor of 4 (k obs ϭ 0.051 s Ϫ1 ).
The already severe negative effect of the 2Ј-5Ј phosphodiester at position Ϫ1 of the scissile strand (k obs ϭ 6.2 ϫ 10 Ϫ4 s Ϫ1 ) was exacerbated by a factor of 8 in combination with 2Ј-5Ј phosphodiesters at positions Ϫ1 or ϩ2 on the complementary strand (k obs ϭ 7.5 ϫ 10 Ϫ5 s Ϫ1 ) (Table II). Other modifications of the nonscissile strand did not have a significant impact on cleavage of the Ϫ1 modified 18-mer.
Effects of 2Ј-5Ј Phosphodiesters on Exonuclease III-E. coli exonuclease III catalyzes unidirectional digestion of duplex DNA from the 3Ј-end to liberate 5Ј dNMP products. The phosphodiesterase activity of exonuclease III is impeded by chemical modifications of the phosphate backbone (20,21), including the 2Ј-5Ј phosphodiester modification studied here (14,15). To confirm the incorporation of the 2Ј-5Ј phosphodiesters during chemical synthesis of the topoisomerase substrate strands and further explore the effects of this modification on the phosphodiesterase activity of exonuclease III, we annealed the 5Ј 32 Plabeled 30-mers containing the complement of the topoisomerase cleavage site to a complementary 60-mer strand and incubated the duplexes with exonuclease III. The 5Ј-labeled digestion products were resolved by denaturing gel electrophoresis (Fig. 3). All of the unmodified 30-mer was converted after 15-30 min to 5Ј-labeled species 8 -10 nucleotides in length. A ladder of partially digested strands was evident at 5 min. Introduction of a 2Ј-5Ј phosphodiester at position Ϫ2 resulted in a kinetic roadblock to exonuclease III digestion located 1 nucleotide 3Ј of the site of the modification (ϪA 2Ј p 5Ј ApT). The paused species persisted at 30 min and was only slowly converted to a species arrested at the site of modification (ϪA 2Ј p 5Ј A). There was almost no digestion of the 32 Plabeled DNA past this point. Changing the position of the 2Ј-5Ј Vaccinia DNA Topoisomerase phosphodiester modification elicited a corresponding shift in the site of the impediment to exonuclease III digestion (Fig. 3).
In the case of the ϩ1 phosphate modification, the progress of exonuclease III was arrested after 5 min at points 1 and 2 nucleotides 3Ј of the modified phosphate. After 15 and 30 min, the major products were arrested at the site of modification and 1 nucleotide upstream and only a minority of the DNA strands were degraded past the modified phosphodiester. The upstream block to digestion was apparent, but less pronounced, on the Ϫ1 phosphate-modified strand, so that most of the DNA was digested up to the site of the 2Ј-5Ј phosphodiester and no further (Fig. 3). The 5Ј 32 P-labeled 18-mers containing the CCCTT topoisomerase cleavage site were also annealed to a complementary 60-mer strand and digested with exonuclease III (Fig.  4). Here again, the 2Ј-5Ј modifications arrest exonuclease III at sites 1 and 2 nucleotides 3Ј of the modified phosphodiester and, as the reaction proceeds, at the modified phosphodiester itself. These results indicate that exonuclease III can sense the conformation of the DNA backbone in advance of the position of its active site.

Conformational Requirements at the Scissile Phosphodiester Differ between Poxvirus and Cellular Type IB Topoisomerases-
The rate of DNA transesterification by vaccinia topoisomerase was reduced by more than six orders of magnitude by a 2Ј-5Ј linkage at the scissile phosphodiester (CCCTT 2Ј p2 5Ј N). Thus, vaccinia topoisomerase is essentially unable to form a DNA-(2Ј-phosphotyrosyl)-enzyme intermediate. The 2Ј-5Ј modification is relatively subtle because it does not alter the charge on the scissile phosphate or introduce bulk. We hypothesize that the altered geometry of the 2Ј-5Ј phosphodiester limits the ability of the nucleophile Tyr-274 to attain its requisite apical orientation with respect to the 5Ј-OH leaving group. Additional perturbations of contacts of the phosphate oxygens with the catalytic Arg-130, Arg-223, His-265, and Lys-167 side chains in the ground state or the transition state may also contribute to the inability of the poxvirus enzyme to cleave the 2Ј-5Ј phosphodiester.
Arslan et al. (14,15) showed that mammalian topoisomerase I is quite capable of cleaving a 2Ј-5Ј phosphodiester to form a DNA-(2Ј-phosphotyrosyl)-enzyme intermediate, although the extent of the reaction at the modified phosphodiester was reduced because the enzyme was diverted to an alternative (unmodified) cleavage site 2 nucleotides upstream. Apparently, the active site of the mammalian type IB topoisomerase is better able to accommodate the altered geometry of a 2Ј-5Ј phosphodiester. Note that the mammalian enzyme is also much less fastidious than the vaccinia topoisomerase with respect to the nucleotide sequence at the cleavage site. The cellular and poxvirus topoisomerases have a common fold and the same constellation of catalytic side chains (6, 7), but the structural nuances that account for the greater stringency of the poxvirus topoisomerase active site are entirely uncharted.
Interference Effects at the Phosphodiesters Immediately 3Ј and 5Ј of the Cleavage Site-A 2Ј-5Ј modification of the Ϫ1 phosphate (CCCTTp2N 2Ј p 5Ј N) reduced the rate of transesterification by vaccinia topoisomerase by a factor of 500. This effect was an order of magnitude more deleterious than simply removing the Ϫ1 phosphate and replacing it by a 3Ј-OH/5Ј-OH nick (13). We infer that the 2Ј-5Ј modification interference is not caused solely by a perturbation of functional contacts between the topoisomerase and the Ϫ1 phosphate. Rather, we invoke an additional effect of the unconventional 2Ј-5Ј linkage on the conformation of either the adjacent Ϫ2 nucleoside (CCCTTp2A 2Ј p 5Ј T) or the Ϫ1 nucleoside (CCCTTp2A 2Ј p 5Ј T) or both. Note that removal of the Ϫ1 phosphate plus the Ϫ2T nucleoside reduces the cleavage rate by three orders of magnitude (13), which is similar to the effect elicited by the 2Ј-5Ј modifi-  cation of the Ϫ1 phosphate. It is also conceivable that a 2Ј-5Ј modification of the Ϫ1 phosphate affects the conformation of the adjacent (unmodified) scissile phosphodiester in such a way as to make it unfavorable for transesterification chemistry. Mammalian topoisomerase I also displays profound interference of cleavage by a 2Ј-5Ј phosphodiester immediately 3Ј of the cleavage site (14,15). A 2Ј-5Ј modification of the ϩ2 phosphate (CCCT 2Ј p 5 Tp2N) had no effect on vaccinia topoisomerase. This result was remarkable, given that the introduction of a 3Ј-OH/5Ј-PO 4 nick at the ϩ2 position slowed the rate of cleavage by a factor of 500 (13). Thus, a modification that preserves the electrostatics on the ϩ2 phosphate and the continuity of the backbone is benign compared to one that breaks the backbone and imposes an extra negative charge. Covalent intermediate formation was also suppressed by introduction of a single 2Ј-O-methyl moiety on the ϩ2 sugar (10). Apparently, the combination of a 2Ј-OMe with the standard 3Ј-5Ј phosphodiester was more detrimental than substitution of the 2Ј carbon (as a phosphate ester) in the context of a 3Ј deoxy sugar. In the crystal structure of vaccinia topoisomerase, the side chains of Arg-84, Ser-268, and Ser-270 coordinate a sulfate FIG. 3. Exonuclease III digestion of 2-5 phosphodiester-modified nonscissile strands. 5Ј 32 P-labeled control and 2Ј-5Ј phosphodiester-containing 30mer oligonucleotides were annealed to a complementary 60-mer to form the tailed duplex molecule shown below the autoradiogram. The DNAs were digested with exonuclease III for 0, 5, 15, or 30 min, and the products were analyzed by polyacrylamide gel electrophoresis. The nucleotide sequences of the 30-mers are displayed next to the cleavage ladders with each letter specifying the 3Ј nucleotide of the indicated radiolabeled species. The sequence of the ladder was verified by coelectrophoresis of the digests with defined oligonucleotide markers (not shown). Filled diamonds (ࡗ) are inserted into the sequences at the sites of the 2Ј-5Ј phosphodiester modifications.
FIG. 4. Exonuclease III digestion of 2-5 phosphodiester-modified scissile strands. 5Ј 32 P-labeled control and 2Ј-5Ј phosphodiester-containing 18-mer oligonucleotides were annealed to a 60mer to form the tailed duplex molecule shown below the autoradiogram. The DNAs were digested with exonuclease III for 0, 2, 5, 15, or 30 min and the products were analyzed by polyacrylamide gel electrophoresis. The nucleotide sequences of the 18-mers are displayed next to the cleavage ladders with each letter specifying the 3Ј nucleotide of the indicated radiolabeled species. Filled diamonds (ࡗ) are inserted into the sequences at the sites of the 2Ј-5Ј phosphodiester modifications.
proposed to correspond to the ϩ2 phosphate of the scissile strand (6). In the mammalian topoisomerase-DNA cocrystal (7), the ϩ2 phosphate interacts with Lys-433, which is the mammalian counterpart of vaccinia Arg-84. Mammalian topoisomerase I is also unaffected by a 2Ј-5Ј linkage immediately 5Ј of the cleavage site. Thus, the relevant contacts to the ϩ2 phosphate appear not to be disrupted by the 2Ј-5Ј phosphodiester.
Cross-strand 2Ј-5Ј Modification Interference-Single modifications at five phosphodiesters on the nonscissile strand had no significant effect on transesterification by vaccinia topoisomerase. However, some of the nonscissile strand modifications were inhibitory in concert with a 2Ј-5Ј phosphodiester at positions ϩ2 or Ϫ1 on the scissile strand. Indeed, modification of the Ϫ1 or ϩ2 phosphodiesters of the nonscissile strand elicited 10-fold synergistic effects in combination with scissile strands containing either ϩ2 or Ϫ1 phosphate modifications. The most severe synergistic effect (100-fold) was observed with 2Ј-5Ј modifications at the ϩ2 positions of both DNA strands. Insofar as the observed negative cross-strand effects involve phosphates located on opposite faces of the B-form DNA duplex, a simple interpretation of the data is that certain local distortions of the duplex in the vicinity of the scissile phosphate adversely affect the transesterification reaction.
2Ј-5Ј Phosphodiester Effects on Exonuclease III-Exonuclease III employs a one-step in-line mechanism in which an activated water attacks the scissile phosphodiester (which is coordinated to an essential divalent cation) and expels the 3Ј-O of the upstream nucleotide of the DNA strand (22). Exonuclease III is clearly impeded from hydrolyzing a 2Ј-5Ј phosphodiester. We presume that the altered geometry affects the correct positioning of the attacking nucleophile relative to the new 2Ј-O leaving group. The block to cleavage at the modified phosphodiester, which is not surprising per se, is overshadowed by two other aspects of our findings. First, we observed that that exonuclease III, though slowed by the 2Ј-5Ј linkage, is nonetheless capable of degrading a fraction of the input DNA strands beyond the modification site. This result raises the interesting question of whether exonuclease III traverses the block directly by hydrolyzing the 2Ј-5Ј phosphodiester linkage or by skipping over the modified linkage and cleaving the DNA at the next available standard phosphodiester bond. (It is also possible that a contaminating endonuclease in the commercial exonuclease III preparation is responsible for traversal of the modified phosphodiester.) Second, we observed that exonuclease III is consistently arrested at positions 1 and 2 nucleotides prior to the encounter of its active site with the modified 2Ј-5Ј phosphodiester. This result implies that exonuclease III surveys or senses the phosphate backbone in advance of the active site.