Protein expression and co-purification
The wild-type DP1, DP2, and PolD from T. kodakarensis were prepared as described earlier . DP1 and His-tagged DP2CTD (1000-1324) were produced in combination with Escherichia coli BL21-CodonPlus (DE3)-RIL cells (Agilent Technologies, Santa Clara, CA). E. coli cells were cultured at 37 °C in Luria-Bertani medium, containing 50 μg/ml ampicillin, 34 μg/ml chloramphenicol, and 50 μg/ml kanamycin, to an OD600 = 0.3. Gene expression was induced by adding IPTG to a final concentration of 1 mM and cells were cultured at 25 °C for 16 h. The cells were collected, resuspended in buffer A, containing 20 mM HEPES, pH 7.5, 1 M NaCl, and 10 mM imidazole, and were disrupted by sonication, followed by heat treatment at 80 °C for 20 min. The soluble fraction was applied to the affinity column filled with TALON resin (Takara Bio, Shiga, Japan), and the column was washed with buffer A. The bound fraction was eluted with 20 mM HEPES, pH 7.5, 1 M NaCl, and 150 mM imidazole. Further purification procedure was described earlier with some modifications . PolDΔPIP (DP2Δ1314-1324), PolDΔKR (DP1 K326A/R328A, and DP2 K432A/R434A/R781A), and PolDΔPIPΔKR were prepared by the recombinant E. coli, containing the corresponding genes, in which each mutation was introduced by PCR-mediated site-directed mutagenesis. Each designed mutation was confirmed by nucleotide sequencing. The primers used for each mutagenesis are as follows: TK1903_R781A_F; 5′-GCCCGCCGACCTGTTAGCGCAGGCCATGGACAACC-3′, TK1903_ R781A_R; 5′-GGTTGTCCATGGCCTGCGCTAACAGGTCGGCGGGC-3′, TK1903_K432R434A_F; 5′-GACTATGAAACGGCTCTAGCGGTCGCGAACGAGGTTGACGAGATC-3′, TK1903_K432R434A_R; 5′-GATCTCGTCAACCTCGTTCGCGACCGCTAGAGCCGTTTCATAGTC-3′, TK1902_K326R328A_F; 5′-GAAGTTCGAGGTTCCCGCGGTCGCGAACGCCCAGAGCAAGG-3′, TK1902_K326R328A_R; 5′-CCTTGCTCTGGGCGTTCGCGACCGCGGGAACCTCGAACTTC-3′.
For the purifications of these PolDs, TOYOPEAL Phenyl 650S (TOSOH, Tokyo, Japan), HiTrap Heparin HP (GE Healthcare, Little Chalfont, UK), and BioPro IEX SmartSep Q10 (YMC) were used for the column chromatography. T. kodakarensis has two PCNAs, PCNA1 and PCNA2, and PCNA1 is the essential clamp molecule in the cells . Therefore, PCNA1 is called TkoPCNA in this study. The purification of TkoPCNA was done as described previously .
Preparation of PolD-PCNA-DNA complex
We reconstituted the PolD-PCNA-DNA complex by mixing the purified TkoPolD and TkoPCNA proteins, with two kinds of primed DNA, which were prepared by annealing two deoxyoligonucleotides shown in Additional file: Figure S1 (synthesized by Hokkaido System Science, Hokkaido, Japan and Sigma-Aldrich Japan, Tokyo, Japan) in 20 mM Bis–Tris, pH 7.0, and 50 mM NaCl, and incubated at 37 °C for 15 min. To prevent degradation of the primed DNAs by the 3′–5′ exonuclease of PolD, four successive phosphorothioate modifications were introduced into the 3′-terminus of both strands. The purified TkoPolD, TkoPCNA, and primed DNA (30/45) (the sequences are shown in Additional file: Figure S1) were mixed and incubated in a solution containing 50 mM Tris–HCl (pH 8.0), 300 mM NaCl, 1 mM Dithiothreitol, 0.1 mM EDTA (without MgCl2) at 20 °C for 15 min. The reconstituted PolD-PCNA-DNA (30/45) complex was loaded onto a Superdex 200 5/150 (GE Healthcare) gel filtration column equilibrated with a solution of the same composition, containing NaCl at 50 mM instead of 300 mM. For the preparation of the PolD-PCNA-DNA (25/35) complex, the mixture was loaded onto the same column, but the equilibrium and elution buffer contained 150 mM NaCl, instead of 50 mM. The protein compositions of both complexes, eluted at the peak fractions, were analyzed by SDS-PAGE.
Electron microscopy and single-particle image analysis
For the negatively stained specimen, a 3-μl aliquot of sample solution was applied to a glow discharged, continuous thin-carbon film supported by a copper grid, left for 1 min, and then stained with three drops of on-ice-cooled 2% uranyl acetate. Stained samples are examined by a Tecnai T20 electron microscope (FEI, Hillsboro, OR) operated at 200 kV accelerating voltage. Images were recorded using an Eagle 2k CCD camera (FEI). Holey carbon (Quantifoil R1.2/1.3 Au 200) grids were used for frozen-hydrated specimens. Grids were glow discharged for 1 min by an HDT-400 hydrophilic treatment device (JEOL, Tokyo, Japan) before usage. Two-three μL of sample solutions were applied on holey grids for rapid freezing. Rapid freezing was performed using EM-GP (Leica Wetzlar, Germany) or Vitrobot (FEI) freezing robots.
Cryo-EM data collection
All the frozen-hydrated samples were first examined by a Polara electron microscope (FEI) operated at 200 kV accelerating voltage, to optimize the sample preparation conditions. Images were taken using either an UltraScan 2 k CCD camera with a Gatan GIF Tridiem energy filter (GATAN, Pleasaston, CA) or an UltraScan 4 k CCD camera (GATAN). Electron microscopic image datasets of PolD-PCNA-DNA (30/45) complex for structure analyses were collected on a Titan Krios electron microscope (Thermo Fisher Scientific, Waltham, MA), equipped with a spherical aberration (Cs) corrector (CEOS GmbH, Heidelberg, Germany), under 300 kV accelerating voltage. An EPU software (Thermo Fisher Scientific) was used for fully automated data collection. Images were recorded by Falcon II direct electron detector (Thermo Fisher Scientific), in the dose-fractionation mode (2 s exposure/32 fractions), at a nominal magnification of × 59,000, corresponding to 1.10 Å/pixel on the specimen. The nominal defocus was in the range of − 1.0 μm ~ − 1.4 μm. A total electron dose of 36 electrons/Å2 was used for each image recording. A Volta phase plate (VPP) was used to enhance the contrast of particle images. Electron microscopic images of the PolD-PCNA-DNA (25/35) complex for structure analysis were collected on a Talos Arctica electron microscope (Thermo Fisher Scientific) using a Falcon 3EC direct electron detector (Thermo Fisher Scientific). Images were recorded in the electron counting mode, at a nominal magnification of × 92,000, corresponding to 1.12 Å/pixel on the specimen, with 47 s total exposure time. The nominal defocus range of the data was − 1.0 μm ~ − 3.0 μm. Intermediate frames (total 1821 movies) were recorded every 0.64 s, giving an accumulated dose of 40 electrons/Å2 and a total of 74 fractions per image (i.e., 0.54 electrons/Å2 dose per fraction).
Cryo-EM image processing
Movie frames of PolD-PCNA-DNA (25/35) and PolD-PCNA-DNA (30/45) were aligned to correct the dose-induced and dose-weighted motions of the specimens using MotionCor2 . The contrast transfer function was determined for each image using the CTFFIND4 program . About one thousand particle images were picked manually from the images using the Relion  manual picking tools. This initial small dataset was subjected to a reference-free 2D classification in Relion. Selected “good” class average images were used as references to automatically pick particles, using Gautomatch (http://www.mrc-lmb.cam.ac.uk/kzhang/Gautomatch/). The output coordinates of the picked particles were used for the particle extraction program of Relion .
A total of 240,256 particles were extracted from the PolD-PCNA-DNA (30/45) complex images, and after 2D classification procedure, particles classified as “bad” were removed from the data set. In total, 171,739 particles were subjected to the 3D classification procedure in Relion (Class3D), assuming eight classes. Particles classified as the best (class 8, 30,889 particles) and second-best (class 1, 19,356 particles) 3D classes were further used for refining (Refine3D in RELION) each of the two class of 3D maps. The refined map of class 1 (corresponding to the Form B), was sharpened by applying a negative B-factor (− 150) and corrected for the MTF of the Falcon II detector.
The extracted 309,113 images of the PolD-PCNA-DNA (25/35) complex were subjected to 2 turns of reference-free 2D classification (Class 2D, Relion) and particle selection procedure to remove bad images. In total, 230,965 images of the PolD-PCNA-DNA (25/35) complex were subjected to 3D classification procedure in Relion assuming eight classes of 3D structures. The initial volume for this procedure was obtained by the “3D initial model” procedure in Relion. Among the eight classes, only one 3D class (Class 2) corresponding to the Form A, exhibited interpretable high-resolution structure. Also, 41,073 particle images, assigned to this class, were used for “3D auto refinement” in Relion. After refinement, “Movie refinement” and “Particle polishing” were applied to improve resolution. The final map was obtained after 2nd “3D auto refinement” and “Post Process” which includes the 3D masking, correction for the MTF of the Falcon 3EC detector, and B-factor sharpening (B = − 500). The resolution of the final maps was estimated by Gold standard FSC using FSC = 0.143 criteria.
TkoPolD-PCNA-DNA model construction
The model coordinates of the TkoPolD-PCNA-DNA complex were constructed by assembling several crystal structures and applying the homology modeling method. The crystal structures of P. abyssi DP1 (PDB: 5IHE) and DP2 (PDB: 5IJL)  were first fit into the density map of Form A as described in the text. Then, the crystal structure of the TkoPCNA trimer (PDB: 3LX1) and a model of primed DNA (30/45) in standard B-form parameters were fit into the density map .
Next, the parts of the model, which were not presented in the above-mentioned close homologs, were constructed. A density predicted to be a helix bundle was observed at DP1-DP2 interface, which was presumed to be similar to that of human Polα B-subunit–Polα catalytic subunit (PDB: 4Y97) or human Polε subunit 2–catalytic subunit A (PDB: 5VBN) interfaces [33, 34]. Thus, Phe1322-Phe1440 of Polα catalytic subunit (chain B of 4Y97) was introduced into the model by superposing DNA polymerase B α-subunit (chain A of 4Y97) on the DP1 subunit. A density predicted to be a short helix was observed at the PIP-binding site of PCNA. TkoDP2 had PIP-box peptide (1316-ISLDEFFGS-1324) at the C-terminus, which was similar to that of human RNase H2 subunit B (295-KSIDTFF-301; chain B of PDB: 3P87) or DNA annealing helicase and endonuclease ZRANB3 (518-KQHDIRSFFV-527; chain B of PDB: 5MLO), in which the consecutive Phe residues provided most of the interactions with the PCNA [47, 48]. Thus, the PIP-box peptide (chain B of 3P87) was introduced into the model by superposing PCNAs (chain A of 3P87 upon chain E of the T. kodakarensis model).
Then, the sequences of the modeled parts were converted into that of T. kodakarensis subunits via homology modeling using MODELLER . The model coordinates were further refined by using PHENIX suite , repeatedly applying real-space refinement by using phenix.real_space_refine , manual model refinement by using COOT , and geometry optimization by using phenix.geometry_minimization. The residues absent from the original model and mainly forming helices were added to the model where they were consecutive to the modeled parts if significant densities were observed. They were consistent with secondary structure predictions, namely, Asn319–Val327 of DP1, and Ile331–Pro346, Val454–Asn457, Arg1007–Val1018, Gly1080–His1103, and Thr1282–Arg1294 of DP2. The model finally showed a map correlation coefficient of 0.76 for whole unit cell (0.70 for masked region), and the MolProbity score of 1.87, with 0% Ramachandran outlier, 0.15% rotamer outlier, and 16.03 clash score . The unmodeled segments in the final model were the N-terminal domain (Met1–Ala285) of DP1, and Lys289–Asn314, Arg365–Thr371, Asp383–Lys399, Ala663–Met676, Arg1048–Leu1079, Glu1199–Leu1281, and Gly1295–Lys1314 of DP2. The model of Form B was constructed based on the Form A model by applying real-space refinement, manual refinement, and geometry optimization as for the Form A model. The final model showed a map correlation coefficient of 0.87 for whole unit cell (0.78 for masked region), and the MolProbity score of 1.63, with 0% Ramachandran outlier, 0.35% rotamer outlier, and 11.73 clash score (Additional file: Table S1). The coordinates of Form A and Form B models have been deposited to the Protein Data Bank with the accession codes 6KNB and 6KNC, respectively.
Yeast two-hybrid assay
A yeast two-hybrid (Y2H) detection system (Matchmaker™ Gold Yeast Two-hybrid System, Matchmaker GAL4 Two-Hybrid System 3, Takara Bio) was used to screen for DP1-DP2 interacting region. The plasmid pGBKT7, encoding the GAL4 DNA binding region, and the plasmid pGADT7, encoding the activation domains, were, respectively, used to prepare plasmids containing the gene encoding DP1 and various DP2 fragment. Co-transformations of the yeast Y2H Gold cells with pGBKT7-DP1 and the pGADT7-DP2 fragments were performed according to the manufacturer’s protocol (Clontech Matchmaker manual). Cell suspensions (3 μl of 2 × 106 cell/ml) of each strain were spotted onto synthetic defined (SD) plates without Leu and Trp for non-selection plate and Leu, Trp, and His, or Leu, Trip, His, and Ade for two different selection strengths plates. The agar plates were incubated at 30 °C for 4 days, and growing cells indicated the interactions of the two proteins produced from the two plasmids used for the co-transformation.
Primer extension assay
The primer extension ability of TkoPolD was measured by counting incorporated radioactivity into DNA strands using dNTP containing [methyl-3H]-dTTP as substrates, and the activities were compared among the WT, ΔPIP, ΔKR, and ΔPIPΔKR in the presence and absence of PCNA. The reaction was performed in 25 μl containing 20 mM Tris–HCl, pH 8.0, 100 mM NaCl, 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgCl2, 0.1% Triton X-100 and 0.1 mg/mL BSA, 10 nM template primer substrate (prepared by annealing M13mp18ssDNA and a deoxyoligonucleotide, M13-63; 5′-dTGCCAAGCTTGCATGCCTGCAGGTCGACTCTAGAGGATCCCCGGGTACCGAGCTCGAATTCGT-3′), 0.2 mM dNTPs including 0.13 μM [methyl-3H]-dTTP (PerkinElmer, MA), 20 nM PCNA, and 5 nM PolD, at 72 °C for 1, 2, and 4 min. The reaction mixture was pre-incubated for 3 min, and PolD was added to initiate the reaction. After incubation, aliquots (8 μl) were fractionated and spotted onto DE81 filters (GE Healthcare). The filters were washed with 5% Na2HPO4 solution thrice and dried. Incorporated radioactivity was measured with a scintillation counter Tri-Carb 3110TR (PerkinElmer).
3′–5′ exonuclease activity assay
The exonuclease reaction was performed in a 20 μl containing 20 mM Tris–HCl, pH 8.0, 10 mM KCl, 10 mM (NH4)2SO4, 1 mM DTT, 0.1 mg/ml BSA, 2 mM MgCl2, and 10 nM Cy5-labeled DNA substrates (prepared by annealing two deoxyoligonucleotides, 5′ Cy5 pri32 and temp45, shown in Additional file: Figure S1, and 50 nM of the recombinant proteins at 65 °C for indicated times). The reaction mixture was pre-incubated for 2 min, and the proteins were added to initiate the reaction. After incubation, an aliquot (4 μl) was fractionated and an equal volume of stop solution (98% formamide and 0.01% orange G) was added followed by 8 M urea-15% PAGE. The gel images were visualized using a Typhoon Trio + imager (GE Healthcare).
Electrophoretic mobility shift assay
Electrophoretic mobility shift assay was performed in 20 μl containing 20 mM Tris–HCl, pH 8.0, 10 mM KCl, 10 mM (NH4)2SO4, 1 mM DTT, 0.1 mg/ml BSA, 2 mM MgCl2, 2, 5 nM Cy5-labeled DNA (5′ Cy5 pri32 for ssDNA, 5′ Cy5 pri32, and temp45 for primed DNA, shown in Additional file: Figure S1) substrates, and indicated concentrations of proteins at 40 °C for 10 min. The protein–DNA complexes were fractionated by 0.8% agarose gel electrophoresis in 0.1 × TAE buffer and visualized with a Typhoon Trio + imager. The binding affinities were calculated as described previously .
Surface plasmon resonance (SPR) analysis
The Biacore J system (GE Healthcare) was used to test the physical interactions of PCNA with various PolD mutants. PCNA was fixed on a CM5 Sensor Chip (GE Healthcare), according to the manufacturer’s recommendations. To measure the kinetic parameters, purified PolDs (PolDwt, PolDΔPIP, PolDΔKR, and PolDΔPIPΔKR) in a running buffer (10 mM HEPES–NaOH pH 7.4, 0.15 M NaCl, 0.1% Tween 20), with six different concentrations (50, 100, 200, 300, 500, 1000 nM), were applied for 120 s to the PCNA-immobilized chip, at a continuous flow rate of 30 μl/min at 25 °C. The bound analytes were removed by washing with regeneration buffer (10 mM HEPES–NaOH pH 7.4, 1 M NaCl, 0.1% Tween 20) at the end of each cycle. The apparent equilibrium constants (KD) of the interactions were determined from the association and dissociation curves of the sensorgrams, using the BIAevaluation (ver. 4.1) software (GE Healthcare).
Analytical gel filtration chromatography
Analytical gel filtration chromatography was performed using the SMART system (Amersham Pharmacia, Buckinghamshire, UK). PolD (wt, ΔPIP, ΔKR, ΔPIPΔKR), PCNA, and DNA (each 5.4 μM, as a heterodimer: PolD, homotrimer: PCNA in 30 μl) were mixed and incubated for 3 min at 60 °C. The protein solutions were applied to a Superose 6 PC 3.2/30 column (GE Healthcare) and were eluted with buffer containing 50 mM Tris–HCl, pH 8.0, and 0.3 M NaCl. Aliquots (5 μl) of applied solution and aliquots (8 μl) of each fraction from the eluates were subjected to 10% SDS-PAGE, followed by Coomassie Brilliant Blue staining. The standard marker proteins, including thyroglobulin (670,000), γ-globulin (158,000), ovalbumin (44,000), and myoglobin (17,000), were also subjected to gel filtration as controls.