Position-specific Suppression and Enhancement of HIV-1 Integrase Reactions by Minor Groove Benzo[a]pyrene Diol Epoxide Deoxyguanine Adducts

The viral protein HIV-1 integrase is required for insertion of the viral genome into human chromosomes and for viral replication. Integration proceeds in two consecutive integrase-mediated reactions: 3′-processing and strand transfer. To investigate the DNA minor groove interactions of integrase relative to known sites of integrase action, we synthesized oligodeoxynucleotides containing single covalent adducts of known absolute configuration derived from trans-opening of benzo-[a]pyrene 7,8-diol 9,10-epoxide by the exocyclic 2-amino group of deoxyguanosine at specific positions in a duplex sequence corresponding to the terminus of the viral U5 DNA. Because the orientations of the hydrocarbon in the minor groove are known from NMR solution structures of duplex oligonucleotides containing these deoxyguanosine adducts, a detailed analysis of the relationship between the position of minor groove ligands and integrase interactions is possible. Adducts placed in the DNA minor groove two or three nucleotides from the 3′-processing site inhibited both 3′-processing and strand transfer. Inosine substitution showed that the guanine 2-amino group is required for efficient 3′-processing at one of these positions and for efficient strand transfer at the other. Mapping of the integration sites on both strands of the DNA substrates indicated that the adducts both inhibit strand transfer specifically at the minor groove bound sites and enhance integration at sites up to six nucleotides away from the adducts. These experiments demonstrate the importance of positionspecific minor groove contacts for both the integrase-mediated 3′-processing and strand transfer reactions.

grase catalyzes insertion of cDNA copies of the viral genome into human chromosomes. During this reaction, integrase multimers are associated with both ends of the viral cDNA and several host and viral proteins that together form preintegration complexes.
The biochemical reactions catalyzed by integrase can be studied in vitro with recombinant integrase and synthetic oligonucleotides. In vivo, integrase first catalyzes a cleavage reaction releasing a dinucleotide (5Ј-p-GT-3Ј) from the 3Ј-end of each viral long terminal repeat (LTR) (3Ј-processing, 3Ј-P; Figs. 1A, 2A, and 2B). The exposed 3Ј-hydroxyl at the cleavage site is then used by integrase for nucleophilic attack upon a target DNA, resulting in insertion of the viral genome into a target chromosome (strand transfer, ST) ( Fig. 2A). Cellular factors perform gap repair and ligation between the viral and host genome junctions. In vitro, a single DNA duplex consisting of the last 21 nucleotides of the HIV-1 U5 LTR (Fig. 1A) can be used to study both 3Ј-P and ST ( Fig. 2A). The cleavage reaction (3Ј-P) is observed as accumulation of a 19-mer DNA band on a denaturing sequencing gel. Another identical duplex serves as the target DNA, and the ST reaction products (STPs) are observed as bands migrating higher than the substrate 21-mer in a denaturing sequencing gel (see Fig. 2C).
Elucidation of the molecular interactions between integrase and its DNA substrates (viral and chromosomal DNA) has proven challenging, and these interactions presently remain unknown. Models describing the arrangement of a simplified complex including integrase plus the viral and host DNA substrates have been proposed (1-4) based on biochemical and structural experiments. Biochemical studies have revealed specific contact points between several DNA bases and integrase amino acids. Integrase has an absolute requirement for the 5Ј-CA-3Ј dinucleotide sequence immediately 5Ј to the 3Ј-P site in the strand to be cleaved at the end of the viral LTR sequence (Fig. 1A, (5)(6)(7)(8)). The efficiency of 3Ј-P is dramatically decreased by changes to the G immediately 5Ј to the conserved CA dinucleotide (9,10). The deoxyadenosine within the conserved CA has been shown to form a cross-link to Lys-159 of integrase (11). This Lys-159 residue also contacts the phosphate 5Ј to the conserved deoxyadenosine in a two-domain structure of integrase (3). Mutagenesis showed that Tyr-143 and probably Gln-148 interact with the 5Ј-overhang (lower strand; Fig. 2A) resulting from upper strand 3Ј-P (11,12). Disulfide cross-linking revealed proximity of the integrase amino acid 246 when mutated from glutamic acid to cysteine and the seventh base from the 5Ј-end of the lower strand of the U5 LTR (1). Although these four contacts have been useful for general alignment of * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the viral DNA substrate relative to integrase, additional information is required for any higher resolution modeling or facilitation of structure-based drug design.
Ethylation interference has been used to locate phosphates that are critical for integrase catalysis (13). This method highlighted specific DNA backbone contacts required for ST near the insertion site. On the donor strand, the phosphate 5Ј to the nucleophilic deoxyadenosine 3Ј-hydroxyl group, as well as the two phosphates on the complementary strand that are closest to the cleavage site, were important for ST. Within the target DNA, the two phosphates proximal to the ST site were important. Together, these observations suggest many stabilizing contacts between integrase, the tip of the cleaved LTR, and the site of ST.
We have used covalent adducts (Fig. 1B) derived from transopening of enantiomeric benzo[a]pyrene 7,8-diol 9,10-epoxides (BaP DE, isomer in which the epoxide oxygen and the benzylic 7-hydroxyl group are trans) by the exocyclic amino group of deoxyguanosine (dG) as probes of the interactions between DNA-processing enzymes, such as human and vaccinia topoisomerases (14 -16), with the minor groove of DNA. These adducts in DNA exhibit well-defined structural motifs as shown by two-dimensional NMR studies in solution (17)(18)(19), with the hydrocarbon portion lying in the minor groove and oriented toward the 5Ј-or 3Ј-terminus of the adducted strand for the trans-(S) and trans-(R) adducts, respectively, where (S) and (R) refer to the absolute configuration at the point of attachment of the 2-amino group to the hydrocarbon (schematically illustrated in Fig. 1B). The trans-opened BaP DE-dG adducts do not significantly disturb Watson-Crick base pairing or B-form DNA helix conformation (17)(18)(19). Thus such adducts, when inserted site specifically into oligonucleotides, provide valuable tools to map interactions between enzymes and their DNA substrates. In the present study, we have applied this approach to examine the effect of minor groove modifications at defined positions relative to the known sites of integrase action. Oligodeoxynucleotides containing single trans-(R) or trans-(S) BaP DE dG adducts at specific positions in the 21-bp oligonucleotide corresponding to the end of the viral U5 LTR were used (Fig. 1A). For two of these positions close to the 3Ј-P site, the importance of minor groove contacts was also tested by replacing the normal deoxyguanosine with deoxyinosine. This modification does not disturb the B-form DNA helix conformation but does remove the H-bonding potential of the 2-amino group of guanosine.
Our study demonstrates that minor groove scanning with BaP DE dG adducts and deoxyinosine substitution can reveal specific molecular interactions between integrase and the DNA minor groove. The effects of minor groove interference on 3Ј-P and ST are separable based on positioning of the minor groove modification. Adducts placed near the 3Ј-P cleavage site selectively inhibit both 3Ј-P and ST, whereas adducts placed distal to the 3Ј-P site selectively inhibit ST and thus uncouple the two integrase reactions.

MATERIALS AND METHODS
Oligonucleotide Synthesis-Unmodified and inosine-substituted oligonucleotides were commercially synthesized by IDT (Coralville, IA). Oligonucleotides containing the trans-opened BaP DE dG adducts were synthesized as described previously (20), using a manual coupling step for the adducted phosphoramidites. Oligonucleotides U17(R,S) and L4(R,S) (for notation, see Fig. 1A) were prepared from the mixed 10R/ 10S diastereomers of the BaP DE dG phosphoramidite (20) and were separated by reverse-phase high-performance liquid chromatography. Their absolute configurations were assigned from the long-wavelength CD spectra of the separated oligonucleotides, which exhibit positive bands in the 330 -360 nm region for the 10R adduct and negative bands for the 10S adduct (21). All other adducted oligonucleotides were prepared from diastereomerically pure 10R or 10S phosphoramidites de-rived from the 10R and 10S diastereomers of the BaP DE dG adducts that had been separated as their O 6 -allyl di-tert-butyldimethylsilyl triacetates (22). Characterization of the diastereomerically pure phosphoramidites is given in the Supplemental Information. All adducted oligonucleotides were purified by high-performance liquid chromatography (see Supplemental Information for conditions and retention times).
Oligonucleotides were further purified on denaturing 20% polyacrylamide gels. Single-stranded oligonucleotides were 5Ј-labeled by T4 polynucleotide kinase (Invitrogen, Rockville, MD) with [␥-32 P]ATP (Amersham Biosciences, Piscataway, NJ) or 3Ј-labeled by terminal transferase (NEB, Beverly, MA) with cordycepin [␣-32 P]dATP (Perkin Elmer Life Sciences) according to the manufacturers' instructions. Unincorporated nucleotide was removed by mini Quickspin oligo column (Roche, Indianapolis, IN). The duplex DNA was annealed by addition of an equal concentration of the complementary strand, heating to 95°C, and slow cooling to room temperature.
Integrase Reactions-Recombinant wild-type HIV-1 integrase was purified from Escherichia coli as described (23) with the addition of 10% glycerol to all buffers. Integrase was incubated with DNA substrates for 1 h at 37°C. The reaction conditions were 600 nM integrase, 20 nM duplex DNA, 7.5 mM MnCl 2 , 5 mM NaCl, 14 mM 2-mercaptoethanol, and 20 mM 4-morpholinepropanesulfonic acid, pH 7.2. This high concentration of integrase was used to maintain the multimer complex required for enzymatic activity. Results similar to those reported here with manganese were also observed with magnesium in the reaction mixture. Reactions were quenched by the addition of an equal volume of gel loading dye (formamide containing 1% SDS, 0.25% bromphenol blue, and xylene cyanol). Products were separated on 20% polyacrylamide denaturing sequencing gels. Dried gels were visualized using a Molecular Dynamics 445 SI phosphorimager (Sunnyvale, CA). Densitometric analysis was performed using ImageQuant software from Molecular Dynamics.
The Schiff base cross-linking experiments were performed as described (24). Briefly, oligonucleotides containing a uracil at position U19 (annealed to unmodified or adducted lower strands) or L7 (annealed to unmodified or adducted upper strands) were 5Ј-32 P-labeled as described above. After annealing, uracil DNA glycolylase was added to create an abasic site at the uracil position. The abasic site facilitates ribose opening and formation of a Schiff base cross-link between an aldehyde group on the ribose and a nearby integrase lysine. The crosslinks were stabilized by addition of 100 mM sodium borohydride. The cross-linked integrase-DNA products were separated from the substrate DNA by SDS-PAGE using 16% tricine gels (Invitrogen, Carlsbad, CA).

RESULTS
We placed a single trans-opened BaP DE dG adduct at each of five positions within the terminal 21 bases of the HIV-1 U5 LTR outlined in Fig. 1A. The locations chosen for the BaP DE dG adducts are near the site of 3Ј-P and the prominent ST sites (see mapping of ST sites in Fig. 5). By convention, modified (dG adducted) oligonucleotides were named (U) and (L) when the adduct was attached on the upper or lower strand, respectively. The position of the adduct was counted from the 5Ј-end of the strand bearing the adduct. For instance, U17(S) refers to the modified duplex oligonucleotide with a BaP DE trans-(S) dG adduct at the 17th position relative to the 5Ј-end of the upper strand (see Fig. 1).
Binding of Integrase to Modified Oligonucleotides Containing BaP DE dG Adducts-Before investigating integrase catalytic reactions, we examined the binding of integrase to the adducted oligonucleotides using the Schiff base cross-link assay between integrase and its DNA substrate (24). None of the BaP DE dG adducts affected integrase binding to the modified DNAs (Fig. 1S, Supplemental Information).
Effect of Minor Groove BaP DE Adducts Close to the 3Ј-P Site-Two of the sites chosen to incorporate BaP DE dG adducts are located near the integrase 3Ј-P site: L4 and U17 (see Fig. 1A). The effects of trans-(R) and trans-(S) BaP DE adducts at the L4 position were evaluated on both the 3Ј-P and the ST reactions (Fig. 2). Integrase reactions are shown schematically in A, B, E, and F. The trans-(R) adduct (L4(R)) partially inhib-ited the formation of the 3Ј-P product of the normal length (19-mer) (Fig. 2C, compare lanes 3 and 2). Additionally, the L4(R) adduct caused an enhanced additional cleavage because of abnormal 3Ј-P (gray arrowhead in B), leading to accumulation of 20-mer DNA. The trans-(S) adduct (L4(S)) blocked 3Ј-P to a greater degree than the trans-(R) adduct (Fig. 2C, compare lanes 4 and 3). Enhanced accumulation of the 20-mer DNA was not observed for the L4(S) adduct. 3Ј-P products were quantified to compare the cleavage efficiency between the control and adducted DNAs (Fig. 2D). For the L4(R) adduct, 3Ј-P at the normal site (generating the 19-mer product) was ϳ35% of 3Ј-P for the unmodified oligonucleotide. The abnormal 3Ј-P (generating the 20-mer product) was greater than the normal 3Ј-P for the L4(R) adduct. However, if the 19-and 20-mer products are summed, total normal and abnormal 3Ј-P was similar for the L4(R) adduct and for control DNA (35% versus 38%, respectively).
Because inhibition of 3Ј-P precludes the formation of STP, a precleaved substrate was used to examine the ST reaction independently of 3Ј-P (see scheme in The effects of the U17(S) adduct on the ST reactions were examined directly using a precleaved substrate (see Fig. 2E) containing the trans-(S) BaP DE dG adduct at the U17 position to isolate the ST reaction. Using this adducted DNA (Fig. 3C), ST was not observed (Fig. 3D, lane 4).
Together, the data obtained with the L4 and U17 BaP DE adducted DNAs demonstrate that the presence of a BaP DE dG minor groove adduct two or three nucleotides 5Ј from the 3Ј-P site and spanning the minor groove adjacent to the 3Ј-P site interferes with both 3Ј-P and ST.
Role of the Guanine 2-Amino Group at the U17 and L4 Positions-Because the U17 and L4 BaP DE dG adducts were linked to the guanine 2-amino position in the minor groove, oligonucleotides containing inosine substitutions at the L4 (Fig. 4, A and C) and U17 (Fig. 4E) positions were examined to determine the role of the guanine 2-amino group on 3Ј-P and ST. The L4 inosine caused a marked decrease in 3Ј-P (81%) and ST (92%) (Fig. 4B, lane 3). ST was directly tested using a precleaved upper strand (Fig. 4C). The L4 inosine did not block ST (Fig. 4D, lane 3), indicating that it selectively inhibited 3Ј-P. In the presence of the U17 inosine (Fig. 4E), 3Ј-P was similar to the control DNA, and ST was only partially reduced (by ϳ24%, Fig. 4F, lane 4). These results demonstrate the critical importance of the minor groove guanine 2-amino group at position L4 for 3Ј-P and its dispensability at an adjacent position in the upper strand (U17).
Mapping of the Integrase ST Sites in the U5 LTR Sequence-To examine the effects of BaP DE dG adducts at positions L8, L10, and U6 (see Fig. 1A) on ST reactions, we first had to map the ST sites within the U5 LTR. With the unmodified U5 LTR oligonucleotide labeled at the 5Ј-end of the upper strand, clusters of ST sites are consistently observed (see Figs. 2 and 3). However, the STP bands cannot be assigned unambiguously because they arise from both upper and lower strand STP (see Fig. 2A). To map the integration sites and therefore the effects of BaP DE adducts, we used 3Ј-end-labeling with cordycepin either on the upper (Fig. 5A) or the lower (Fig. 5B) strand. For consistency in referring to the 3Ј-and 5Ј-end-labeled products, we did not include the cordycepin in numbering the residues in the 3Ј-end-labeled oligonucleotides. Placing the radioactive label at the 3Ј-end of one of the strands of the DNA allows the mapping of strand-specific integration products. For example, only the STP on the left side of the diagram in Fig. 5A will be observed when 3Ј-labeled upper strand DNA is used (Fig. 5C). These STPs result from ST into the upper strand. The STP on the right side of Fig. 5A, resulting from ST into the lower strand, do not contain a radioactive label and therefore cannot be observed. Likewise, only the STPs illustrated on the right of Fig. 5B are detectable in the gel shown in Fig. 5D.
Various DNA size markers (M) matching several integration products were included so that the sizes of the STPs could be determined (Fig. 5, C and D). The upper 24, lower 24, and lower 25 markers match the sequence of the integration products and therefore have the same electrophoretic migration as the 24 and 25 STPs. The upper 30, upper 34, lower 30, and lower 32 markers differ in sequence from the predicted STPs and migrate slightly faster (by approximately half a base) than the corresponding STPs.
A map of preferred integration sites on unmodified DNA was created by analyzing the strand-specific pattern of integration bands (Fig. 5E). The intensity of a band on a sequencing gel corresponds to the frequency of integration events at a particular site. Densitometric analysis was used to create a graph illustrating band migration for each STP (Fig. 5E). The height of each peak was transferred into a bar-graph format and aligned with the substrate DNA sequence (Fig. 5E)  ST is an approximation. On the lower strand, two STP clusters were observed peaking around 24-and 32-mers, respectively. The 32-mer STP peak was adjacent to the L8 and L10 positions, where BaP DE dG adducts were located in the experiments to be described below. Strand-specific mapping was used for analyzing the effects of each adduct.

Effect of the L8, L10, and U6 BaP DE dG Adducts on ST-
The L8 and L10 positions are near the prominent lower strand STP peak 32 (identified in Fig. 5). The data in Fig. 6, B and C, show the results of labeling of the 5Ј-end of the upper strand of the L8(R) and L8(S) adducted duplexes and provide a composite of the STP in the upper and lower strands. In contrast, 3Ј-end labeling of the upper strand (shown in Fig. 6, D and E) allows detection of upper strand STP only and may be compared with Fig. 5, C and E (U3Ј). None of the upper strand STPs from the L(8) adducted oligonucleotides contain adducts, and thus they migrate at the same positions as the unmodified controls.
In the unadducted control (Fig. 6B, lane 2), the triplet of STPs with the most prominent peak at 32 (indicated by the double-headed arrow) is attributable to integration into the lower strand. Decreased intensity was observed for this 32 STP cluster with the L8(R) adduct, and to a lesser extent with the L8(S) adduct, when compared with the control. Suppression of STPs at this site is presumably attributable to proximity of the adducts. Changes to the ST pattern resulting from integration into the intact upper strand induced by the BaP DE adducts at the L8 position are shown in Fig. 6, D and E. The L8(S) adduct slightly enhanced the upper strand 33, whereas the L8(R) adduct suppressed the 24 STPs and resulted in two new STPs at 23 and 25. The strong 23 STPs (Fig. 6D, lane 3) are not apparent in the composite products seen on 5Ј-end-labeling of the upper strand (Fig. 6B, lane 3) and may result from an integration that occurs only when the upper strand is extended by one base by the cordycepin label. The L8(S) adduct gave an anomalous, unidentified upper-strand integration product that migrated between the bands corresponding to 25 and 24 STPs.
We next investigated the effects of the BaP DE adducts at the L10 position (Fig. 7A), two nucleotides away from the L8 adducts described above. In the composite of upper and lower strand STPs (Fig. 7B, upper strand 5Ј-end-labeled), the 30 STPs are enhanced by the L10(S) adduct and completely suppressed by the L10(R) adduct. The 30 STP results from integration into the lower strand on the 3Ј-side of the adducted G and hence migrates normally because it does not contain an adduct. Integration into the lower strand on the 5Ј-side of the L10 adducts would result in STPs with retarded migration (e.g. 32 † ) because of the presence of the adduct (Fig. 7, B and C). The L10(R), and to a lesser extent the L10(S), adduct inhibited formation of these 32 † STP. The pattern of STPs derived from integration into the upper strand near its 3Ј-end is quite similar for the L10 (Fig. 7, D and E) and L8 (Fig. 6, D and E) adducts. Similar to the L8(R) adduct (compare Fig. 6D), the L10(R) adduct enhanced the 23 and 25 STP, but in contrast to the L8(R) adduct, the L10(R) adduct did not significantly inhibit formation of the 24 STP. Similarly, the L10(S) adduct (Fig. 7D, lane 4), similar to the L8(S) adduct (Fig. 6D, lane 4), produced an anomalous band migrating slightly slower than the 24-mer marker.
The U6 position (Fig. 8) is in the center of a strong cluster of upper-and lower-strand STPs (see Fig. 5E). Analysis of STPs with this adducted substrate 5Ј-end labeled on the upper strand is complicated by the possible formation of STPs containing one or two BaP DE adducts (from lower-and upperstrand STPs, respectively). Consistently, integration products for both the U6(R) and U6(S) adducts (Fig. 8B, lanes 5 and 6) in the 32 cluster were retarded on electrophoresis (shifted up) relative to the unadducted STPs (Fig. 8B, lane 4). These bands may correspond to the same positions of integration, producing products with altered mobility attributable to the hydrocarbon. For both adducts the 31 † STPs appeared to be enhanced at the expense of the 32 † STPs. In addition, the U6 adducts decreased the amount of STPs corresponding to the cluster around the 24 † STPs (arrows in Fig. 8, B and C). We could not map precisely the integration sites in the 24 † cluster observed for the U6 adducted DNAs. Nevertheless, these data are consistent with inhibition of integration into the lower strand at sites 24 and 25 opposite to the position of adduct attachment (Fig. 8A).

DISCUSSION
Chiral BaP DE adducts were incorporated into five different positions of individual oligonucleotides derived from the HIV-1 U5 LTR to explore interactions between HIV-1 integrase and the minor groove of substrate DNA. Because noncovalent binding of DNA to HIV-1 integrase was unaffected by the presence of the adducts, they could be used to study minor groove contacts between integrase and DNA. We observed position-dependent effects of the adducts on both 3Ј-P and ST. Specifically, 1) 3Ј-P and ST were blocked by adducts near the 3Ј-P site; 2) ST was blocked at the position of adduct minor groove occupancy; and 3) some ST sites were enhanced by adducts positioned remotely from those sites.
Importance of the L4 Minor Groove Contacts for 3Ј-P-Stereo-specific inhibition of integrase 3Ј-P on the upper strand was observed for the lower strand L4(R) and L4(S) adducts that lie in the minor groove adjacent to the conserved 5Ј-CA-3Ј nucleotides, which are required for integrase activity (5)(6)(7)(8). Furthermore, we found that the L4 guanine 2-amino group is also critical for efficient 3Ј-P because inosine substitution blocked integrase 3Ј-P.
Orientation of the L4 adducts proximal to the 3Ј-P site (*) is illustrated in Fig. 9. These space-filling illustrations were derived from two-dimensional NMR structures of adducted duplex DNAs with a different sequence from the present integrase substrate (17)(18)(19). They serve only to show the location of the scissile phosphate in relation to the BaP DE adducts for discussion purposes. Inhibitory effects of the L4 adducts on 3Ј-P were in agreement with the directionality of adduct extension in the minor groove relative to the cleavage site. Specifically, 3Ј-P was selectively blocked by the L4(S) adduct, which directly overlaps the minor groove at the 3Ј-P site (Figs. 1A and  9). This is in contrast to the L4(R) adduct, which extends away from the 3Ј-P site and allows some 3Ј-P. This defines the importance of the minor groove region that is covered by the L4(S) adduct for 3Ј-P. Stereospecific inhibition by trans-(R) and trans-(S) BaP DE adducts was also observed for DNA cleavage by vaccinia and human topoisomerases (15,16). Both of these type IB topoisomerases perform site-specific cleavage and religation of DNA via a DNA-(3Ј-phosphotyrosyl)-enzyme intermediate. For both topoisomerases, the non-interfering diastereomer extended away from the scissile phosphate, whereas the inhibitory diastereomer extended toward the cleavage site. These results imply the importance of minor groove or guanine 2-amino interactions between the enzymes and DNA and show that BaP DE adducts are useful tools for examining enzymatic mechanisms.
The L4 guanine 2-amino group appears to be an important contact for efficient 3Ј-P, because decreased cleavage was observed with both the adducted and inosine-substituted DNA substrates. We observed similar binding of HIV-1 integrase to the control DNA and both of the L4 adducted DNAs (Fig. 1S), implying that the L4 free amino group is not required for binding by HIV-1 integrase. In contrast, modified bases showed that the absence of the L4 guanine 2-amino group resulted in a loss of DNA binding and 3Ј-P by human T-cell leukemia virus type II integrase (25). It was hypothesized that human T-cell leukemia virus type II integrase may initiate binding of the viral DNA end through a minor groove interaction but may use required major groove contacts for integrase catalysis (25). Unlike human T-cell leukemia virus type II 2, HIV-1 integrase may exhibit different binding requirements because DNA binding was unaffected by any of the minor groove BaP DE adducts. However, 3Ј-P was almost completely inhibited by the L4(S) adduct (and both U17(R) and (S) adducts, see below) in the minor groove, as well as the L4 inosine substitution.
Role of Minor Groove Contacts Adjacent to U17 for 3Ј-P-The U17 guanine, 2 bp removed from the 3Ј-P site, is critical for activity by HIV-1 integrase. An abasic site at this position prevents 3Ј-P (10). In contrast, we found that the U17 guanine 2-amino group is not required for 3Ј-P because substitution for this base by inosine (lacking the 2-amino group) did not affect 3Ј-P. Moreover, inosine substitution at U17 decreased ST by only 24%, implying that the 2-amino group is not required for the ST reaction. Therefore, the inhibition of 3Ј-P and ST by U17(R) and U17(S) adducts may be ascribed to the presence of the minor groove adducts rather than to a direct involvement of the U17 guanine 2-amino group.
The relative orientations of the adduct and the phosphodiester bond that is cleaved in 3Ј-P are illustrated in Fig. 9 for R and S adducts at U17 as well as at L4. Interestingly, both U17(R) and U17(S) adducts inhibit 3Ј-P, although U17(S) does not directly overlap the scissile phosphodiester bond (*). Thus, contacts distal to this cleaved bond must be critical for 3Ј-P. The only adduct of these four that permits significant 3Ј-P is L4(R), which occupies approximately the same linear space in the minor groove as the inhibitory U17(R) (Fig. 9). Therefore, L4(R) and U17(R) are not equivalent for integrase interaction. Integrase might be more inhibited when the adduct is covalently linked to the scissile strand (U17 position) than the non-scissile strand (L4 position), as was suggested recently for DNA topoisomerase II (26). Also, because they are on opposite strands, the U17(R) and L4(R) adducts have opposite orientations relative to the site of 3Ј-P. Specific steric or hydrogen bonding interactions mediated by the bulkier and more exposed trihydroxy tetrahydrobenzo-ring (relative to the pyrene moiety) may play a role in the differential effects of these two adducts. For example, if minor groove interference is inhibitory in a region one or two base pairs 5Ј to the 3Ј-P cleavage site (to the lower left in Fig. 9), then this bulkier portion of the adduct may present less of a block to the enzyme when closer to the 3Ј-P site as in L4(R) as compared with L17(R). Interestingly, an opposite effect of BaP DE dG adduct orientation was observed for DNA cleavage by vaccinia topoisomerase, where greater inhibition of cleavage was observed when the bulky trihydroxy tetrahydrobenzo-ring was oriented toward the cleavage site (16).
Global Versus Site-specific Inhibition of ST by the BaP DE Minor Groove Adducts-Unlike 3Ј-P, ST was globally inhibited by BaP DE adducts positioned adjacent to the DNA 3Ј-hydroxyl group, which acts as the nucleophile for the ST reactions. Binding of the substrate for the cleavage reaction was not perturbed by the presence of the BaP DE-dG adducts, as demonstrated by Schiff base cross-linking experiments (see Fig. 1S, Supplemental Material). However, inhibition of ST suggests that these bulky adducts prevent the proper binding of this target DNA to integrase and/or the catalysis of the ST reaction.
ST was also inhibited at and near adducts at the L8, L10, and possibly the U6 positions. These adducts could inhibit ST in three ways: 1) by blocking appropriate positioning of the nucleophilic 3Ј-hydroxyl of the donor DNA; 2) by blocking a required protein-minor groove contact; or 3) by blocking target DNA binding. In contrast to the site-specific 3Ј-P cleavage, selection of a ST site occurs in an apparently sequence-independent manner. Analysis of the crystal structures of 60 protein-DNA complexes to determine minor groove components important for recognition of DNA by drugs and proteins (27) showed that purine N 3 , pyrimidine O 2 , deoxyguanine N 2 , and deoxyribose O 4Ј were the only minor groove groups involved in interactions between DNA and proteins, drugs, and water. Additionally, sequence-independent hydrogen-bond acceptors in the minor groove lie in identical positions for all four Watson-Crick base pairs (28). Therefore, it is likely that the BaP DE adducts may block integrase side-chain interaction with minor groove hydrogen-bonding groups required to position the acceptor DNA in a sequence-independent manner, prohibiting ST at the site of adduction. For example, the deoxyguanosine N 2 amino group forms many water-mediated contacts in protein/DNA complexes (27) that may be blocked by BaP DE dG adducts.
Enhancement of ST at Sites Flanking BaP DE Adducts-ST was enhanced at several insertion sites remote from the BaP DE dG adducts at the L8, L10, and U6 positions, notably to give the 23 and 25 STPs with L8(R) and L10(R) adducts and the 31 STPs with U6(S) and (R) adducts This enhanced ST may be a result of a shift of the total ST activity of integrase to new sites. The low sequence specificity of integrase target site selection may promote integration at less preferred ST sites when preferred sites are blocked by the BaP DE adducts or by structural perturbations in their immediate vicinity. NMR structures of BaP DE dG adducted DNA have shown that the hydrocarbon occupies the minor groove while the DNA maintains essentially a B-form helix conformation with normal base pairing (17)(18)(19). However, recent molecular dynamics calculations on these structures, based on the NMR data, have indicated that the minor groove is significantly (2-4 Å) widened at the site of attachment of these dG adducts and 1-2 bp away, particularly for the S adduct (29). A similar effect on human topoisomerase I specificity was observed with BaP DE dG adducts at or near a normal cleavage site, which resulted in suppression of DNA cleavage at the adduct site and enhancement of cleavage at distal sites (14,15).
Inhibition of HIV-1 by Targeting the LTR Minor Groove-Integrase is an important clinical target for the inhibition of HIV replication. Synthetic polyamines such as the lexitropsins selectively bind the minor groove of the conserved AAAAT stretch of the HIV-1 LTRs and inhibit integrase at nanomolar concentrations (30). The results presented here suggest that targeting the minor groove of the HIV-1 DNA near the integrase 3Ј-P site may also effectively inhibit integrase activity. Polyamines specifically binding the AϩT-rich tip of the Moloney murine leukemia virus LTR exhibit nanomolar inhibition of integrase and subnanomolar affinity for the Moloney murine leukemia virus LTR (31), supporting the viability of this approach for HIV. Both the HIV-1 U3 and U5 LTRs terminate in sequences that are site specifically bound by integrase (5Ј-GCAGT-3Ј for the U5 LTR and 5Ј-CCAGT-3Ј for the U3 LTR). Additionally, simultaneously targeting of integrase and the DNA minor groove with the same inhibitor may increase inhibitor specificity. Minor groove-binding inhibitors recognizing this sequence, or other types of inhibitors that specifically prevent interaction between integrase and the minor groove of the LTR tip, could interfere with provirus integration and therefore with viral replication. FIG. 9. Relative orientation of the hydrocarbon moiety (yellow) and the sites of 3-P ‫)ء(‬ for L4 and U17. Models of the trans-opened 10R and 10S BaP DE dG adducts, based on two-dimensional NMR studies (17,18), are aligned so that the 3Ј-end of the upper strand (the unmodified strand for L4 and the modified strand for U17) is at the top right to give the same spatial orientation of the 3Ј-P site in all cases. Brackets indicate the two phosphodiester linkages in the modified strand immediately flanking the adducted dG. Although the adducts occupy approximately the same linear space in the minor groove, their directional projection from the groove is dependent on their R versus S configuration. In all cases, the tetrahydrobenzo-ring is more exposed than the pyrene moiety, which is deeper in the groove.