Biochemical and Structural Study of RuvC and YqgF from Deinococcus radiodurans

Abstract

Deinococcus radiodurans possesses robust DNA damage response and repair abilities, and this is mainly due to its efficient homologous recombination repair system, which incorporates an uncharacterized Holliday junction (HJ) resolution process. D. radiodurans encodes two putative HJ resolvase (HJR) homologs: RuvC (DrRuvC) and YqgF (DrYqgF). Here, both DrRuvC and DrYqgF were identified as essential proteins for the survival of D. radiodurans. The crystal structures and the biochemical properties of DrRuvC and DrYqgF were also studied. DrRuvC crystallized as a homodimer, while DrYqgF crystallized as a monomer. DrRuvC could preferentially cleave HJ at the consensus 5'-(G/C)TC↓(G/C)-3' sequence and could prefer using Mn²⁺ for catalysis in vitro, which would be different from the preferences of the other previously characterized RuvCs. On the other hand, DrYqgF was identified as a Mn²⁺-dependent RNA 5'-3' exo/endonuclease with a sequence preference for poly(A) and without any HJR activity. IMPORTANCE Deinococcus radiodurans is one of the most radioresistant bacteria in the world due to its robust DNA damage response and repair abilities, which are contributed by its efficient homologous recombination repair system. However, the late steps of homologous recombination, especially the Holliday junction (HJ) resolution process, have not yet been well-studied in D. radiodurans. We characterized the structural and biochemical features of the two putative HJ resolvases, DrRuvC and DrYqgF, in D. radiodurans. It was identified that DrRuvC and DrYqgF exhibit HJ resolvase (HJR) activity and RNA exo/endonuclease activity, respectively. Furthermore, both DrRuvC and DrYqgF digest substrates in a sequence-specific manner with a preferred sequence that is different from those of the other characterized RuvCs or YqgFs. Our findings provide new insights into the HJ resolution process and reveal a novel RNase involved in RNA metabolism in D. radiodurans.

Keywords: Deinococcus; Holliday junction (HJ); RNase; RuvC; homologous recombination (HR); protein structure.

FIG 1

The sequence alignments of RuvC and YqgF. (A) The sequence alignments of RuvC from different organisms. (B) The sequence alignments of YqgF from different organisms. Names of species are dr, Deinococcus radiodurans; tt, Thermus thermophilus HB8; ec, Escherichia coli; pa, Pseudomonas aeruginosa; ab, Acinetobacter baylyi; mt; Mycobacterium tuberculosis; hp, Helicobacter pylori; bs, Bacillus subtilis. Secondary structural elements are depicted according to PDB files (DrRuvC, this study; TtRuvC, PDB code: 4ep4; EcRuvC, PDB code: 1hjr; PaRuvC, PDB code: 6lw3; DrYqgF, this study; EcYqgF, PDB code: 1nu0; MtYqgF, PDB code: 7ess; BsYqgF, PDB code: 1vhx) and are displayed at the top of the sequences. Similar residues are boxed in blue. Conserved key residues are written with white bold characters and are highlighted with a red background. Residues for metal binding and substrate binding are labeled at the bottom of the sequences with green and blue triangles, respectively. Key residues for protein dimerization are framed with a blue dashed box.

FIG 2

Structure analysis of DrRuvC and DrYqgF. (A) The overall structure of the DrRuvC dimer (this study). The structure elements are numbered and labeled. (B) The zoomed-in view of α2 of DrRuvC (the dimerization area). The key residues for dimerization are labeled and shown as sticks. (C) The electrostatic potential of the DrRuvC surface. The potential was determined using the Adaptive Poisson-Boltzmann Solver (APBS) and is shown as a solvent excluded surface (range = ±5) by PyMOL. The catalytic centers are marked with yellow stars. The HJ substrate, which was extracted from the TtRuvC-HJ complex (PDB code: 6s16), was docked into the DrRuvC apo structure by PyMOL, shown as a cartoon, colored cyan. (D) The zoomed-in view of the aligned catalytic center of DrRuvC (orange; this study), TtRuvC (red; PDB code: 4ep4), EcRuvC (blue; PDB code: 1nmn), and PaRuvC (brown; PDB code: 6lw3). The key putative catalytic residues are labeled and shown as sticks. (E) The overall structure of the MtYqgF dimer (PDB code: 7ess). Each monomer is colored differently. The structure elements are numbered and labeled. The putative catalytic residues and the key residues for dimerization are labeled and shown as sticks. (F) The overall structure of the BsYqgF dimer (PDB code: 1vhx). Each monomer is colored differently. The structure elements are numbered and labeled. The putative catalytic residues and key residues for dimerization are labeled and shown as sticks. (G) The overall structure of EcYqgF (PDB code: 1ovq). The structure elements are numbered and labeled. The putative catalytic residues are labeled and shown as sticks. (H) The overall structure of DrYqgF (this study). The structure elements are numbered and labeled. The putative catalytic residues are labeled and shown as sticks. (I) The electrostatic potential of the DrYqgF surface. The potential was determined using the APBS and is shown as a solvent excluded surface (range = ±5) by PyMOL. The catalytic center is marked with a yellow star. (J) The zoomed-in view of the aligned catalytic center of DrYqgF (orange; this study), EcYqgF (blue; PDB code: 1ovq), MtYqgF (violet; PDB code: 7ess), and BsYqgF (red; PDB code: 1vhx). The key putative catalytic residues are labeled and shown as sticks.

FIG 3

Dimerization analysis of DrRuvC and DrYqgF. (A) The gel filtration analysis of DrRuvC and DrYqgF. A set of protein standards of known molecular mass, such as aprotinin (6.5 kDa), RNase A (13.7 kDa), carbonic anhydrase (29 kDa), ovalbumin (43 kDa), and conalbumin (75 kDa), were used to calibrate the Superdex 75 10/300GL. DrRuvC eluted around 10 mL, reflecting a dimer form. DrYqgF eluted around 14 mL, reflecting a monomer form. (B) The protein cross-linking assays of DrRuvC and DrYqgF. The upper gel showed the cross-linking result of DrRuvC, and the lower gel showed the cross-linking result of DrYqgF. Lanes 1 and 8: DrRuvC or DrYqgF protein without the cross-link reagent treatment. Lanes 2 to 7: proteins were treated with different concentrations of glutaraldehyde (0.01, 0.02, 0.04, 0.08, 0.16, and 0.32%). Lanes 9 to 14, proteins were treated with different concentrations of BS3 (0.125, 0.25, 0. 5, 1, 2, and 4 mM).

FIG 4

Holliday junction resolvase assays of DrRuvC and DrYqgF. (A) Analysis of the HJR activities of DrRuvC and DrYqgF over short HJ (HJ31) with a short mobile junction. Together with three other unlabeled strands (J31-2, 3 and 4), a 5′ 6-FAM labeled strand (J31-1) which contains the putative RuvC cleavage site (5′-ATTC-3′) at the mobile junction area was annealed into the HJ substrate HJ31. The underlined bases in the HJs correspond to the homologous core. 200 nM DNA was mixed with 1 μM protein, 10 mM metal (Mg²⁺, Mn²⁺, Ca²⁺, or Zn²⁺), and products were resolved by 10% native TBE-PAGE (middle) and 12% TBE-urea denaturing gels (lower) in the same time. (B) Analysis of the HJR activities on long HJ substrates (HJ98) with a long mobile junction. 5′ 6-FAM labeled HJ substrates with a 66 nt homologous core were synthesized (labeled at strand J98-1 or J98-2) to monitor the different cutting patterns by DrRuvC. 200 nM DNA was mixed with 1 μM protein and 10 mM metal (Mg²⁺ or Mn²⁺) and then incubated at 37°C for 30 min. Products were resolved by 15% TBE-urea denaturing gel. (C) Analysis of the preferred cleavage sequences of DrRuvC. Holliday junctions are different only in the homologous core sequence indicated as N¹N²N³N⁴. Reactions for each substrate were performed at the same conditions. 2.5 μM DNA was mixed with 5 μM DrRuvC and 10 mM Mn²⁺ and then incubated at 37°C for 30 min. Products were resolved by 10% native TBE-PAGE and stained by Stains-all. (D) and (E) Mapping the cleavage sites of DrRuvC. 5′ 6-FAM labeled HJ31 (D) and unlabeled HJ24-TTCG (E) substrates were used for a cleavage sites analysis, and reactions were performed at the same conditions as in (A) and (C), respectively. The gels were imaged by fluorescence mode or after staining with Stains-all. The cleavage sites of DrRuvC were determined by comparing the position of the product bands with the marker bands and the product bands created by EcRuvC.

FIG 5

Catalytic metal concentrations analysis of DrRuvC. (A) Comparison of the HJR efficiencies of wild type DrRuvC, H139D mutant, and H139A mutant. Different concentrations of protein (0.5, 1, 2, and 4 μM) were incubated with 200 nM HJ31 and 10 mM Mn²⁺, and the reaction products were resolved by 10% native TBE-PAGE. (B) Analysis of the metal preference of wild type DrRuvC and H139D mutant. Different concentrations (0.31, 0.625, 1.25, 2.5, 5, 10, and 20 mM) of Mg²⁺ or Mn²⁺ were added into the reaction system, and the products were resolved by 8% native TBE-PAGE (see Fig. S4A for one of the representative gel results). The digestion fractions were calculated by Image J from three repeats and displayed as a line chart using GraphPad Prism 9.

FIG 6

Analysis of the digestion efficiency and the binding affinity of DrRuvC. (A) The digestion efficiency and the binding affinity of DrRuvC toward different DNA structures. 200 nM substrate was mixed with 0.5 or 1 μM DrRuvC and 10 mM Mn²⁺ and then incubated at 37°C for 30 min. The products were resolved by 15% TBE-urea denaturing gels. Different DNA structures were shown below the gel. The corresponding binding K values of each structure were calculated from the EMSA results shown in Fig. S6. (B) The model of the DrRuvC-HJ complex. Based on the TtRuvC-HJ complex structure, the HJ substrate (cyan) was docked into the aligned binding motifs of DrRuvC (orange; this study), TtRuvC (red; PDB code: 6s16), EcRuvC (blue; PDB code: 1nmn), and PaRuvC (brown; PDB code: 6lw3) in PyMOL. The key putative residues for interactions are labeled and shown as sticks. A cartoon model for half of the DrRuvC-HJ complex was built in the right-hand corner. The cyan dots indicate the HJ binding sites on DrRuvC. The cleavage sites of HJ are colored pink. (C) The influence of divalent metal ions on the HJ binding affinity of DrRuvC. 100 nM HJ31x was incubated with 250 nM DrRuvC in the presence of different concentrations (0, 1.25, 2.5, and 5 mM) of EDTA, Mg²⁺, Mn²⁺, or Ca²⁺. The products were resolved by 5% TB-native gel.

FIG 7

RNase activity analysis of DrYqgF. (A) The total RNA digestion assays of DrYqgF and DrRuvC. 2 μg of total RNA extracted from D. radiodurans were incubated with 1 μM wild type DrYqgF, site-directed DrYqgF mutants, and wild type DrRuvC in the presence of different kinds of metal ions. The products were resolved by 5% TBE-urea denaturing gel. (B) The analysis of the substrate sequence specificity of DrYqgF. 200 nM 20 nt RNA substrates with different sequences were incubated with 1 μM wild type DrYqgF and 10 mM metal (Mg²⁺ or Mn²⁺) at 37°C for 30 min. (C) Comparisons of the digestion efficiencies of wild type DrYqgF and D22A mutant on poly(dA) or poly(A) substrates. Different substrates (200 nM) were incubated with gradient concentrations (0, 0.01, 0.1, 1, and 10 μM) of protein and 10 mM Mn²⁺ at 37°C for 30 min. (D) Analysis of the exonuclease digestion direction of DrYqgF. 200 nM 20 nt poly(A) substrates, labeled at either the 5′ end or the 3′ end, were incubated with gradient concentrations (0, 1, and 2 μM) of DrYqgF and 10 mM Mn²⁺ at 37°C for 30 min. (E) Analysis of the preferred sequence for the endonuclease activity of DrYqgF. 200 nM 20 nt poly(U), which contained 0, 1, 2, 3, or 4 adenine bases within its sequence, were incubated with gradient concentrations (0, 1, and 2 μM) of DrYqgF and 10 mM Mn²⁺ at 37°C for 30 min. The products of (B), (C), (D) and (E) were resolved by 15% TBE-urea denaturing gel.