Mechanism of lesion verification by the human XPD helicase in nucleotide excision repair

Abstract

In nucleotide excision repair (NER), the xeroderma pigmentosum D helicase (XPD) scans DNA searching for bulky lesions, stalls when encountering such damage to verify its presence, and allows repair to proceed. Structural studies have shown XPD bound to its single-stranded DNA substrate, but molecular and dynamic characterization of how XPD translocates on undamaged DNA and how it stalls to verify lesions remains poorly understood. Here, we have performed extensive all-atom MD simulations of human XPD bound to undamaged and damaged ssDNA, containing a mutagenic pyrimidine (6-4) pyrimidone UV photoproduct (6-4PP), near the XPD pore entrance. We characterize how XPD responds to the presence of the DNA lesion, delineating the atomistic-scale mechanism that it utilizes to discriminate between damaged and undamaged nucleotides. We identify key amino acid residues, including FeS residues R112, R196, H135, K128, Arch residues E377 and R380, and ATPase lobe 1 residues 215-221, that are involved in damage verification and show how movements of Arch and ATPase lobe 1 domains relative to the FeS domain modulate these interactions. These structural and dynamic molecular depictions of XPD helicase activity with unmodified DNA and its inhibition by the lesion elucidate how the lesion is verified by inducing XPD stalling.

Graphical Abstract

MD simulations of XPD–ssDNA complexes show that XPD opens its pore to allow undamaged DNA to translocate and narrows its pore to stall the lesion.

Figure 1.

Initial structures of XPD complex with the lesion-containing and lesion-free ssDNA and the interactions of the XPD–ssDNA near the entry pore. (A) Initial structure of XPD complex with the lesion (red stick)-containing ssDNA (grey). This initial structure is the first snapshot in the production run of the MD simulation. The XPD protein includes four domains: the Arch, FeS, ATPase lobe 1 and lobe 2 domains, which are color coded and labelled according to constituent subunits. XPD has a central tunnel that accommodates ssDNA with its 3′- and 5′-end extended toward the entry and exit pore, respectively. Inset box showing a view into the DNA entry pore, formed by the FeS, Arch, and ATPase lobe 1 domains. The iron-sulfur cluster (FeS cluster) adjacent to the DNA entry pore is indicated as orange and yellow spheres, respectively. The lesion is the pyrimidine (6−4) pyrimidone photoproduct (6−4PP). The chemical and crystal structures of the 6−4PP lesion, which is taken from a Rad4-Rad23 structure bound to a 24-bp DNA containing a 6−4PP lesion (PDB ID 6CFI (64)) are shown here; the structure shows that the two thymine rings are linked by a single C6−C4 covalent bond and are roughly perpendicular. (B) Zoom-in view of the DNA entry pore depicts the initial structures of the XPD complex with lesion-containing and lesion-free ssDNA. Note that initial structures for each model are the first snapshot in the production run in the MD simulations. Full details concerning the preparation of the initial models for the MD simulations of XPD–ssDNA complex are given in Materials and Methods and Supplementary Figures S1 and S2. (C) Schematic representation of interactions of XPD with the nucleotides at positions 0 to 2 near the entry pore for each initial structure as described in (B). The 5T and 3T of the lesion indicate the pyrimidine and pyrimidone of 6−4PP, respectively. These XPD-DNA interactions mainly involve R112, Y192, R196 and Y211 that anchor the DNA backbone moieties.

Figure 2.

Superposition of the most representative structure (red) from the equilibrated ensemble with the initial structure (grey) and RMSDs of the DNA near the XPD pore for each XPD complex, revealing a one-nucleotide 3′→5′ backbone translocation and base reorientation in the unmodified dT1, while the backbone translocation is absent in the lesion-containing XPDs. (A) In XPD-unmod, during the equilibrated state, the dT1 exhibits a large displacement from its initial position within XPD (with heavy atom RMSD of 7.5 ± 0.26 Å). This displacement is dominated by its backbone 3′→5′ movement, as viewed along the entry pore (with P atom RMSD of 6.6 ± 0.54 Å), as well as the base reorientation. A view into the entry pore shows that the base of dT1 is reoriented from the ATPase lobe 1 domain to the interface between the Arch and the FeS domains. (B) In XPD-lesionOUT, during the equilibrated state, the 6−4PP lesion reveals a large deviation from its initial conformation (with heavy atom RMSD of 6.5 ± 0.20 Å). This deviation is mostly due to the reorientation of the modified bases that transit from state 1 (0−1 μs) to state 2 (1−2 μs), and to the equilibrated state (2−3 .5 μs) (Supplementary MovieS5); while one-nucleotide backbone translocation is not achieved as revealed in the P atom RMSD of 2.2 ± 0.15 Å. (C) In XPD-lesionIN, the lesion retains a stable conformation after ∼0.6 μs with heavy-atom RMSDs of 3.0 ± 0.12 Å, indicating that both backbone and base orientations are similar to the initial conformations. The backbone is nearly immobilized with P atom RMSD of 1.7 ± 0.17 Å. For each XPD–ssDNA complex, we computed the time-dependent heavy atom and backbone P atom RMSDs of the 6−4PP lesion in the lesion-containing XPDs as well as the corresponding RMSDs of the dT1 in the lesion-free XPD.

Figure 3.

Unmodified nucleotides near the pore are translocated in the 3′→5′ direction passing through the entry pore. (A) Stage 1 (during 0−0.3 μs), the nucleotide dT at position 1 (dT1) is placed outside the pore with its base pointing toward the ATPase lobe 1 domain. R196 (FeS) anchors the backbones of dT1 and dT2. Y192 (FeS) interacts with backbone of dA0. (B) Stage 2 (during 0.3−0.5 μs), the side-chain of R380 (Arch) is inserted between dT1 and dA0. The positively-charged side-chain of R380 is positioned on the 5′-side of dT1 and forms cation-pi interactions with the base of dT1 (Supplementary Figure S9); thus, it aids in pulling the base of dT1 toward the Arch domain and away from the ATPase lobe 1 domain. Interactions of R196 with dT1 and Y192 with dA0 remain unchanged. (C) Stage 3 (during 0.5−1 μs), R380 (Arch) and K128 (FeS), both positioned on the 5′-side of dT1, together pull the phosphate group of dT1 to outcompete its backbone interactions with R196; as a result, the backbone of dT1 is translocated toward its 5′ direction and is pulled toward the Arch domain and lifted away from the FeS and the ATPase lobe 1 domains. A bend is formed between dA0 and dT1 because the phosphate group of dA0 is still anchored stably by Y192 after the backbone of dT1 is translocated in the 3′−5′ direction. Now, R196 interacts with dT2 and the side-chain of R380 on the 5′-side of dT1 retains its cation-pi interactions with the base of dT1. (D) Stage 4 (during 1−3.0 μs), the base of dT1 is flipped into the pore with its base sandwiched between H135 and R380; in this way, the base of dT1 avoids a steric clash with the side-chain of R380. The bend between dT1 and dA0 is retained. Contacts between R196 and dT2, Y192 and dA0, and K128 and the dT1 backbone remain unchanged. (E) Stage 5 (during 3.0−3.5 μs), R380 has moved to the 3′-side of dT1 and interacts with the backbone of dT2 and dT3. The base of dT1 is still well-stacked with H135. The DNA bend between dT1 and dA0 is retained (also see Supplementary Figure S11). Interactions of R196 with dT2, Y192 with dA0, and K128 with the dT1 backbone remain unchanged. Note that the best representative frames from each stage are displayed, revealing the pathway of the nucleotides’ translocation in the 3′→5′ direction along the entry pore (also see Supplementary MovieS1). Left panel shows the XPD–ssDNA interactions near the entry pore; the hydrogen bonds are denoted by the yellow dashed lines. The middle panel is a view along the entry pore (surface rendering), showing the positioning of the nucleotides along the pore during the MD simulation. Right panel shows a view into the entry pore; 3′-end d(TTG) nucleotides are not shown for clarity. The structures are color-coded as in Figure 1. H135 and R380 are labeled and displayed as sphere rendering, highlighting their relative positions with respect to the undamaged dT1.

Figure 4.

Interactions of the 6−4PP lesion with surrounding XPD residues in XPD-lesionOUT. (A) The hydrogen bonding interactions of XPD residues with the backbone and bases of the lesion are shown in the left and middle panels, respectively. Right panel, schematic representation of hydrogen bonds (dashed lines) of XPD residues with the 6−4PP lesion; the hydrogen bond numbers of each residue with nucleotides are also displayed. Total hydrogen bond numbers of each XPD residue−nucleotide pair are the summation of the values of all the hydrogen bond donor-acceptor pairs within each XPD residue−nucleotide pair, as listed in Supplementary Table S2. For details see hydrogen bonding analysis in Supplementary Structural Analyses. The backbone of the 6−4PP lesion is anchored stably by the FeS residues R196, Y192 and R112, with hydrogen bond number (total HBs) ∼5.42; the bases of the lesion interact with the ATPase lobe 1 residues P215, D219, and L220, with total HBs ∼1.95 and the Arch residue E377 (Arch) with total HBs ∼0.30. (B) The 5T of the lesion is clamped narrowly between the Arch (E377and R380) and the ATPase lobe 1 (residues 215–221) domains. (C) The N3 and O2 atoms of the 5T base (rendered in spheres) are captured by a small pocket formed by the ATPase lobe 1 helix. Furthermore, the loop with residues P215 and K216 of the ATPase lobe 1 domain is inserted between dA0 and the 5T base of the lesion, aiding in blocking the 5T base of the lesion from entering the pore (also see Supplementary MovieS2).

Figure 5.

Interactions of the 6−4PP lesion with surrounding XPD residues in XPD-lesionIN. (A) The hydrogen bonding interactions of the XPD residues with the backbone and bases of the lesion are shown in the left and middle panels, respectively. Right panel, schematic representation of hydrogen bond interactions of XPD with the nucleotides near the pore. The backbone of the 6−4PP lesion is anchored by the FeS residues R196, Y192 and R112, with total HBs ∼5.02. The bases of the lesion contact K216 (ATPase lobe 1) and N402 (Arch) with total HBs ∼1.37. The O2 atom of the 3T base hydrogen bonds stably with K216 (ATPase lobe 1). (B) Other non-H-bond interactions of XPD with the 3T base of the lesion contribute to the stabilization of the 6−4PP lesion at the pore. Left panel reveals that K216 (ATPase lobe 1) which hydrogen bonds with the O2 atom of 3T forms a network of interactions with E377 and R380; furthermore, the methyl group of the 3T base of the lesion interacts with the aromatic residue H135 via methyl−π interactions. (C) The 3T base (rendered in spheres) of the lesion is confined between all three XPD domains: FeS, Arch and ATPase lobe 1; E377 and R380 (Arch), K216 (ATPase lobe 1), and H135 (FeS) are in close contact with the 3T base of the lesion (also see Supplementary MovieS3).