Entrapment of hOGG1 complex with intrahelical DNA

Numerous structures of hOGG1 complexed with DNA have been reported, but all represent states with the target nucleotide in the course of being extruded from the DNA helix^{16,23,31,32,33}. To entrap the complex of hOGG1 with intrahelical G or oxoG, we identified several potential sites for DXL in the protein–DNA interface based on the native lesion-recognition complex (LRC) of hOGG1¹⁶ and prepared the corresponding mutant proteins, containing a single cysteine at the relevant positions. Given that, a panel of oligonucleotides shown in Fig. 1c containing a thiol-tether attached to the major grove, the minor grove, or the DNA backbone were screened for DXL. In this study, we developed 5 and 8 atom long new tethers, –CH₂CH₂OCH₂CH₂–, –CH₂CH₂OCH₂CH₂ OCH₂CH₂–, hereafter referred as X₅ and X₈ (see Supplementary Fig. 1 for synthesis) respectively, that do not suffer from the insolubility problems that hampered past attempts to incorporate long hydrocarbon linkers. Tethers were systematically varied in length and composition from C₂ (–CH₂CH₂–) to C₃ (–CH₂CH₂CH₂–), C₄ (–CH₂CH₂CH₂CH₂–), X₅ (–CH₂CH₂OCH₂CH₂–), and X₈ (–CH₂CH₂OCH₂CH₂OCH₂CH₂–). On the basis of this screen, we have selected the Y207C variant, because it crosslinks efficiently but is substantially distant from the active site of the enzyme to participate in the catalysis and does not interact with DNA in native LRC, and crosslinked it through the X₈ linker to the minor groove N2-exocyclic amine of 3′ guanine of the target nucleobase (Fig. 1b–d). The long X₈ linker (~17 Å long) has the flexibility necessary for the capture of the hOGG1 complex with intrahelical DNA. The complex produced a crystal structure at 2.35 Å (Supplementary Table 1) resolution with an intrahelical target G•C base-pair, hereafter referred to as the interrogation complex (IC).

To entrap the structure of hOGG1 encountering the intrahelical oxoG•C lesion, a point mutation, C253W, was introduced in the active site to sterically block the entrance of oxoG to the active site³⁴, since a similar strategy had proved successful in related studies with MutY. With this and Y207C variant tethered to X8 linker, the sequence-matched structure of hOGG1 encountering an intrahelical oxoG•C lesion, referred to as the EC, was obtained and refined to 2.38 Å resolution (Supplementary Table 1). In both IC and EC structures, the target G•C and oxoG•C base-pairs are unambiguously intrahelical (Fig. 2b, c and Supplementary Fig. 2a, b), with a root-mean-square deviation (RMSD) of 0.244 Å between the two structures, in which the RMSD superposition was performed for all atoms. These structures reveal the nature of early events in DNA inspection of hOGG1.

In parallel, we solved the structure of the control complex (2.37 Å resolution, Supplementary Table 1) to show that DXL between Y207C and the X8 linker does not interfere with oxoG extrusion and recognition by hOGG1. The structure of this complex, herein xLRC (with catalytically inactive K249Q), reveals that the active site as well as the overall structure is identical within an RMSD of 0.201 Å for Cα atoms (a total of 2058 atoms) to those of the native LRC¹⁶. In the xLRC structure the oxoG is extrahelical and deeply inserted into the enzyme’s lesion recognition pocket as observed in the native LRC complex¹⁶ (Fig. 2a and Supplementary Fig. 3a), confirming that the introduced mutations and crosslinking through the X8 linker do not interfere with the extrusion of oxoG into the active site of hOGG1.

Unique Interactions of hOGG1 with DNA

Several features of the IC and EC structures support the conclusion that the two structures represent the state of the enzyme at its initial encounter with the DNA. They are characterized below in terms of three key elements, which are different from the native LRC: (1) unique conformation of the DNA backbone, (2) rearrangement of the active site, and (3) different interaction with the C opposite oxoG (hereafter referred to as the estranged C).

First, the DNA conformation of the IC and EC structures differs substantially from that of LRC. The least-squares superposition of the IC and EC structures with LRC, using only the protein component in the superposition, clearly shows a well-defined anchor point on the 3′ side of the target strand (Fig. 3). The backbone of the target strand is held in place by the main-chain hydrogen bonds with G245, Q/K249 and V250 of the signature helix-hairpin-helix motif, which are found in all hOGG1–DNA structures solved to date^{16,23,31,32,33}. On the other hand, the 5′ side of the target strand and the non-target strand of the IC and EC structures are noticeably different from those of LRC (Fig. 3).

Secondly, in the LRC, F319 and C253 interact with both π-faces of the extruded oxoG, which sandwich the base in the active site of hOGG1. The extruded oxoG is further stabilized by hydrogen bonding with the sidechain carbonyl of Q315 and also with the backbone carbonyl of G42 (Supplementary Fig. 3a). The interaction with G42 is specific to oxoG and contributes to the discrimination of oxoG in the active site^16,23. These interactions are not possible in the IC and EC structures as there is no base extrusion. In addition, the αO-helix, which includes F319 and Q315 in the IC and EC structures, retracts away from that of LRC (Supplementary Fig. 3b, c), leaving the active site open for lesion binding. This active site rearrangement was observed previously in the exo-site structure of hOGG1²³ (Supplementary Fig. 3d), suggesting that the rearrangement of the αO-helix relative to LRC takes place after or in concert with the insertion of the base into the active site^16,23.

Finally, hOGG1 in the IC and EC structures interacts differently with the estranged C (Fig. 4). In LRC¹⁶, Y203 wedges into the DNA helical stack on the 5′-side of the estranged C stabilizing a helical bend. N149 enters the space left vacant by the oxoG extrusion and hydrogen bonds with the Watson-Crick face of the estranged C. In addition, the estranged C is stabilized by R154 and R204 that form bidentate hydrogen bonds together with N149. On the target strand, the 3′ and 5′ phosphoryl groups of oxoG are anchored to hOGG1 by main chain hydrogen bonds with N150 of the conserved NNN motif (Fig. 4a). This mode of interaction was observed in all the hOGG1–DNA complex structures published to date, including the exo-site structure²³.

In the IC and EC structures, due to the presence of oxoG/G in the helical stack, the residues mentioned above engage in a different mode of interactions (Fig. 4b, c). For example, R154 and R204 contact the non-target strand backbone, instead of interacting with the C opposite to oxoG (i.e., the estranged C in LRC). In addition, Y203 does not invade the helical stack and remains at the periphery of the minor-groove face of DNA. Similarly, N149 rests on the minor-groove face of the target G•C (in IC)/oxoG•C base-pair (in EC) and hydrogen bonds with both bases (Fig. 4b, c and Supplementary Fig. 4a–c). In these hydrogen bonds, the interaction with N2 of G/oxoG is specific to G and oxoG. Adenine (A) cannot interact with N149 in this orientation (Supplementary Fig. 5). For C and T, although their C2=O carbonyl can form a hydrogen bond with N149, its sidechain is too short to reach C and T on the target strand. OxoG can also adopt a syn conformation and mis-pair with adenine (Supplementary Fig. 5d). Although N149 could in principle interact with the C8=O in the syn conformation, similar to C and T, the side chain of N149 is too short to reach the C8=O group of oxoG (syn), so it is unlikely that N149 will make a productive engagement with the oxoG (syn):A pair. Biochemical data also support that hOGG1 is specific to oxoG(anti):C and does not catalyze oxoG (syn):A pair effeciently³⁵.

Intrahelical lesion recognition by hOGG1

Between IC and EC, while the target strand follows a similar backbone trajectory, the non-target strand of IC has translocated a half-nucleotide step toward its 3′ end relative to EC (Fig. 3b). Despite the conformational difference of DNA between the IC and EC structures, their DNA backbone structures around the target nucleotide are similar to each other (Supplementary Fig. 6a). The only notable difference is in the longer distance between C8 and C5′ of oxoG compared to the corresponding distance of IC (5.4 Å in EC vs 4.4 Å in IC).

The X-ray structures also reveal that, in EC, a water molecule bridges the oxoG to its 5′-backbone phosphate (Supplementary Fig. 6b) so as to attenuate the repulsion between them. Previously, it had been shown that the repulsion between C8=O of oxoG and its backbone phosphate plays a key role in the oxoG-specific intrahelical recognition by MutM²². The repulsion forces the oxoG ribose to adopt an alternative sugar pucker and/or a rotation of phosphodiester groups around oxoG. In the EC structure of hOGG1, the longer distance and the bridging water molecule between C8=0 of oxoG and the C5′ and 5′-backbone phosphate, respectively, could help bend the DNA at the target site, bringing these repulsive functional groups (i.e., oxoG and backbone phosphate) close to each other, thereby initiating base extrusion. Through this repulsion the enzyme discriminates oxoG and G, even at the initial encounter of the (intrahelical) lesion, prior to base extrusion and the conformational change to LRC.

MD simulations reveal an important consequence of the structural differences at the target site between EC and IC. As presented in Supplementary Fig. 7, in the presence of oxoG, hOGG1 establishes a stable contact with the target strand, while in the case of G it fluctuates back-and-forth. Consistent with this, the target strand with oxoG shows smaller root-mean-square fluctuation (RMSF) of atomic position than that with G (Supplementary Fig. 8). This suggests that the enzyme can quickly translocate to the next base pair in IC, while in the case of oxoG, it is locked at the target site.

Lesion discrimination and base extrusion mechanism

In the IC and EC structures, hOGG1 does not form any direct interaction with the discriminatory major groove face of oxoG and G (Fig. 1a). This raises the question: by what mechanism does hOGG1 discriminate G and oxoG in the early phase of extrusion? To answer this question, we simulated the extrusion of G and oxoG out of the DNA helix into an extrahelical state and determined the associated free energy change by use of the string method³⁶. The free energy profiles determined for the oxoG and G extrusions in this event are presented in Fig. 5a, in which the progression of the base extrusion is described by a normalized reaction coordinate α. The simulations show that the extrusion proceeds in three steps: (1) the target base is extruded out of a helical stack through a major groove of DNA, (2) the extruded base binds transiently at the exo-site, and (3) it then enters the active site of the enzyme with concomitant closure of the active site (see Supplementary Movie 1).

The MD simulation shows that G has a higher barrier than oxoG (13.3 kcal/mol for G versus 8.1 kcal/mol for oxoG; Fig. 5a). Moreover, the free energy of G continues to increase to 21.2 kcal/mol near the exo-site (Supplementary Fig. 9). Two protein residues, H270 and K249, stabilize the extruded oxoG through C8=O but not G (Fig. 5b); both residues are indispensable for the oxoG cleavage³⁷. In particular, H270 forms a hydrogen bond with C8=O of the extruded oxoG between α = 0.5 and 0.9, through its backbone amine (Fig. 5b); Supplementary Fig. 9 shows the change of the distance between H270 and oxoG C8=O group along the entire base extrusion process. As the oxoG extrusion continues, the C8=O group begins to interact with the K249 sidechain. These interactions lead to a relatively flat free energy profile between α = 0.5 and 1.0 (Fig. 5a). Since these interactions are not possible for G, its free energy remains high, thus G return quickly to the intrahelical position. This difference suggests that hOGG1 kinetically discriminates the DNA lesion during its extrusion as shown in Fig. 6.

Despite the significant difference of the free energy barrier between oxoG and G, the free energy of the two systems increases very similarly at the beginning of the base extrusion between 0 < α < 0.4 (Fig. 5a). This is consistent with the fact that hOGG1 does not interact directly with the N7 and C8 of oxoG and G in the IC and EC structures, thus not discriminating between them in the early phase of base extrusion. Poor discrimination in the early phase of base extrusion was also proposed from the stopped-flow fluorescence measurements^38,39. Nevertheless, compared with similar events in water^40,41, the entire process is accelerated by DNA bending and the extent of protein–DNA contacts on the minor groove face of DNA. For example, hOGG1 uses a non-specific breakage of the target base-pair, assisted by N149. In addition, K249 contacts the 3′-side phosphate of oxoG/G at the beginning of the process (Figs. 5b, 6), thereby establishing a pivot for base extrusion. H270 is the first residue that specifically interacts with extruded oxoG, followed by K249 with C8=O of oxoG. This suggests that H270 and K249 function as a “cherry-picking” residue in hOGG1, with a similar role of R112 in MutM via significantly different mechanisms²². By contrast, in the case of G, its extrusion cannot be stabilized by the two residues and competes with the translocation of the enzyme along the DNA strand.

Figure 5 also suggests that hOGG1 extrudes the oxoG through the major groove, in accordance with the previously determined hOGG1/DNA complex structure³² with a barrier of 8.1 kcal/mol. The free energy profile for the minor-groove oxoG extrusion is also presented in Fig. 5a, and the free energy profiles along the entire base extrusion process are shown in Supplementary Fig. 9. The barrier for the minor groove extrusion is 17.9 kcal/mol. This result can be compared with the different results reported for MutM between the major^30,42 and minor groove base extrusions^22,29.

In summary, we present X-ray crystallographic structures of human DNA glycosylase hOGG1 interrogating DNA lesions in their intrahelical position, achieved by covalent trapping of an ordinarily transient state in DNA recognition. They reveal how hOGG1 discriminates oxoG from G while both are embedded in the DNA duplex. Specifically, the enzyme utilizes unique protein/DNA contacts to induce DNA bending at the target site. This bending brings the repulsive functional group of oxoG to the immediate vicinity of the DNA backbone, resulting in an oxoG specific distortion of the DNA backbone in its intrahelical orientation. In silico molecular dynamics simulations and free energy calculations corroborate the structural results and help to elucidate the role of the human enzyme in discriminating oxoG from G prior to a complete extrusion from the DNA stack. The results presented here broaden our understanding of one of the earliest events that occur as this extraordinary enzyme patrols genome in its surveillance of DNA damage.