The full-length structure of Thermus scotoductus OLD defines the ATP hydrolysis properties and catalytic mechanism of Class 1 OLD family nucleases

Abstract

OLD family nucleases contain an N-terminal ATPase domain and a C-terminal Toprim domain. Homologs segregate into two classes based on primary sequence length and the presence/absence of a unique UvrD/PcrA/Rep-like helicase gene immediately downstream in the genome. Although we previously defined the catalytic machinery controlling Class 2 nuclease cleavage, degenerate conservation of the C-termini between classes precludes pinpointing the analogous residues in Class 1 enzymes by sequence alignment alone. Our Class 2 structures also provide no information on ATPase domain architecture and ATP hydrolysis. Here we present the full-length structure of the Class 1 OLD nuclease from Thermus scotoductus (Ts) at 2.20 Å resolution, which reveals a dimerization domain inserted into an N-terminal ABC ATPase fold and a C-terminal Toprim domain. Structural homology with genome maintenance proteins identifies conserved residues responsible for Ts OLD ATPase activity. Ts OLD lacks the C-terminal helical domain present in Class 2 OLD homologs yet preserves the spatial organization of the nuclease active site, arguing that OLD proteins use a conserved catalytic mechanism for DNA cleavage. We also demonstrate that mutants perturbing ATP hydrolysis or DNA cleavage in vitro impair P2 OLD-mediated killing of recBC-Escherichia coli hosts, indicating that both the ATPase and nuclease activities are required for OLD function in vivo.

Figure 1.

Purification and biochemical characterization of Ts OLD. (A) SDS-PAGE gel showing purified full-length Ts OLD (Ts^FL). (B) SEC-MALS analysis of Ts^FL. UV trace (black) and calculated molecular weight based on light scattering (blue) are shown. (C) Metal-dependent nuclease activity of Ts^FL on linear λ DNA with quantification of DNA digestion. (D) Metal-dependent nuclease activity of Ts OLD C-terminal region (Ts^CTR) on linear λ DNA. (E) Metal-dependent nuclease activity of Ts^FL on supercoiled pUC19 DNA. N’, ‘L’ and ‘S’ denote the positions of ‘nicked’, ‘linearized’, and ‘supercoiled’ DNA respectively. DNA topology is also illustrated by the small cartoons to the left of the gel. (F) Metal-dependent nuclease activity of Ts OLD C-terminal region (Ts^CTR) on supercoiled pUC19 DNA. Graphs represent the average of three independent trials with error bars representing the standard error of the mean.

Figure 2.

Architectural organization of full-length Ts OLD. (A) Structure of Ts OLD dimer. Coloring is as follows: ATPase domain orange; dimerization domain, light pink; Toprim domain, teal. Dimer is oriented orthogonal to the two-fold axis with one subunit rendered transparent for clarity. Cartoon of domain the architecture with numbered boundaries is included on the right. Secondary structure elements are numbered sequentially within each domain with the superscripts ‘A’, ‘D’, and ‘T’ denoting the ATPase, dimerization, and Toprim domains respectively (See Supplementary Figure S5A). (B) Top down view of dimerization domain helices. Helices from Subunit 1 and Subunit 2 are colored light pink and warm pink respectively. (C) Dimerization domain interactions. Residues contributing to the dimerization interface are shown as sticks and labeled. (D) Interactions stabilizing the ATPase-Toprim interface. Contributing side chains are shown as sticks and labeled. Dashed black line denotes salt bridge between R353 and E413.

Figure 3.

Catalytic determinants of Ts OLD ATP hydrolysis. (A) Michaelis–Menten kinetics of Ts^FL. (B) Structural superposition of the ATPase active site loops from Ts OLD (light orange), Methanococcus jannashii Rad50 (light blue; PDB: 5DNY), and Bacillus subtilis SMC (light green; PDB: 5H66). P loop, Walker B, Q loop, D loop and H loop are labeled with the identity and position of the catalytic residues for each protein species. (C) Initial ATP hydrolysis rates of wildtype (WT) Ts OLD compared to mutants with single alanine substitutions for each of the residues involved in ATP binding and hydrolysis. Data points are the average of three independent trials with error bars representing the standard error of the mean.

Figure 4.

Ts OLD ATPase domains adopt a non-productive conformation. (A) Structural superposition of the ATPase domain dimers from Ts OLD (light orange) and Methanococcus jannashii (Mj) Rad50 (light blue; PDB: 5DNY). Bottom (left) and side (left) views are depicted. Aligned domains are labeled as ‘Subunit 1’ with the second domain in each structure labeled as ‘Subunit 2’ to highlight alternative conformations. Mj Rad50 adopts the canonical, head-to-head dimer arrangement necessary for productive hydrolysis in ABC ATPases (See Supplementary Figures S10 and S11). Catalytic motifs in each structure are colored as follows: P loop, red; Q loop, blue; Walker B, green; D loop, yellow; ABC signature sequence, orange; H loop, purple. Arrows mark the Cα positions of the trans-acting catalytic D loop residues (H283 in Ts OLD; D293 in Mj Rad50) in Subunit 2 of each structure. (B) Stereo view of the aligned ATPase domains (Subunit 1) from dimer superposition in A. The P loop, Q loop, D loop, and H loop are labeled. Position of the α2^D helix in Ts OLD is marked for reference. (C) Stereo view of non-aligned ATPase domains (Subunit 2) from dimer superposition in A. Subunits are shown in the same orientation as in B. Relative positions of the P loop, D loop, and H loop are labeled in each structure. Position of the α2^D helix in Ts OLD is marked for reference. Dashed black arrow indicates rotation of the second Ts ATPase domain relative second subunit Mj Rad50, which adopts a hydrolysis-competent conformation. Dashed black line indicates displacement (measured from Cα to Cα) of the trans-acting H283 in the Ts D-loop relative to its counterpart (D293) in Mj Rad50.

Figure 5.

Class 1 and Class 2 OLD proteins share a conserved mechanism for nuclease cleavage. (A) Structural superposition of OLD CTR structures from Ts (teal) and Burkholderia pseudomallei (Bp, PDB: 6NK8; yellow and purple). (B) Zoomed view of active sites identifying catalytic machinery required for nuclease function. Coloring as in A. Bound magnesium ions in Bp^CTR are shown as green spheres with ‘A’ and ‘B’ denoting the positions of ‘metal A’ and ‘metal B’ respectively. (C) Conservation of active site residues among Class 1 OLD homologs. Coloring generated using the ConSurf server (15) and the alignment in Supplementary Figure S6. (D) Cleavage activities of Ts OLD active site mutants on linear λ DNA. Mutant abbreviations are as follows: 3A, E377A/D431A/D433A; 3B, D381A/S478A/E480A; 2A/2B, D431A/D433A/S478A/E480A. (E) Nicking and cleavage activities of Ts OLD mutants on supercoiled pUC19 DNA. All cleavage assays were performed in the presence of magnesium and calcium as described in the Materials and Methods. Graphs represent the average of three independent trials with error bars representing the standard error of the mean.

Figure 6.

ATPase and nuclease activities are required for P2 OLD function in vivo. (A) Representative images of spot assay for E. coli carrying a temperature-sensitive recBC^ts allele transformed with arabinose-inducible P2 OLD wild type as well as P2 OLD carrying mutations in the ATPase and nuclease domains. Strains were grown at 37°C overnight, which is not permissive for the RecBC^ts function. P2 OLD was induced with 0.1% arabinose or repressed with 0.1% glucose. (B) Quantification of colony forming units under P2 OLD induction compared to P2 repression. Graphs represent the average of three independent trials with error bars representing the standard error of the mean.