Portal de Periódicos da CAPES

Novel structure of an N-terminal domain that is crucial for the dimeric assembly and DNA-binding of an archaeal DNA polymerase D large subunit from Pyrococcus horikoshii

2010; Wiley; Volume: 585; Issue: 3 Linguagem: Inglês

10.1016/j.febslet.2010.12.040

1873-3468

Ikuo Matsui, Yuji Urushibata, Yulong Shen, Eriko Matsui, Hideshi Yokoyama,

RNA and protein synthesis mechanisms

DP2 binds to DP2 by circular dichroism (View interaction) D-family DNA polymerases (PolD), which were originally discovered in Euryarchaeota [1], have also been identified in Korarchaeota, which may have diverged early from the major archaeal phyla Crenarchaeota and Euryarchaeota [2]. This suggests that PolD are responsible for the replication of the ancestral genome of archaea [3]. From recent analysis of the evolution of DNA replication apparatus, it is likely that the last common ancestor of archaea had two DNA polymerases from the B-family and one from the D-family [3]. PolD from Pyrococcus horikoshii (PhoPolD) was proposed to be a heterotetramer (molecular weight: 420 kDa) consisting of two small subunits (DP1s) (PH0123, NCBI Accession No. NP_ 142131; 622 amino acids), and two large subunits (DP2s) (PH0121, NCBI Accession No. NP_ 142130; 1434 amino acids) [4]. DP2 is the catalytic subunit of DNA polymerase [4], while DP1 is the catalytic subunit of Mre11-like 3′–5′ exonuclease, which shows low but significant homology to the non-catalytic second subunit found in eukaryotic B-family DNA polymerases (Pols α, δ, and ε) [5]. Interestingly, it was reported that PolD demonstrated strong DNA polymerase and 3′–5′ exonuclease activities were acquired when the two subunits were mixed or co-expressed, even though each individual subunit demonstrated weak activity [5, 6]. The domain containing the 300 N-terminal residues of DP2 [abbreviated as DP2(1–300); similar descriptions for other fragments will be used in the present manuscript] was reported to be essential for the folding of PolD and is probably the oligomerization domain [7]. Since the molecular mechanisms of the protein folding and biochemical function of the DP2(1–300) domain are unknown, we investigated the crystal structure of the domain and its key roles in the dimeric assembly and the self-cyclization of the DP2 subunit. To prepare a selenomethionine (SeMet) derivative of DP2(1–300), Escherichia coli BL21 (DE3) Codon-Plus RIL-X (Stratagene) was transformed with the previously reported co-expression plasmid pET15b/SL(1–300) [7]. The strain was grown in LeMaster's medium [8] without methionine and supplemented with 50 mg/l L-SeMet (Sigma) and 0.8% (w/v) lactose. Crystallization drops were prepared by mixing equal volumes of protein solution (6 mg/ml) and reservoir solutions containing 22.5% PEG10 000 and 0.1 M Hepes–NaOH (pH 7.5), and 40 mM guanidine–HCl was added to the droplets. Then, crystals were grown at 20 °C according to the hanging-drop vapor-diffusion method. A crystal was cryoprotected with 15% (v/v) glycerol containing the reservoir solution, and flash-frozen at 100 K. The X-ray diffraction data of the SeMet DP2(1–300) were collected at beamline BL6A of the Photon Factory in KEK (Tsukuba, Japan) with a Quantum-4R CCD detector, and were processed and scaled with HKL2000 [9]. The structure was determined using the multiwavelength anomalous dispersion (MAD) method. Four of eight possible selenium sites were found by the direct method with SHELXS [10] and refined with SHARP [11] along with the CCP4 suite [12]. The remaining four SeMet residues were in the disordered region. The initial model was built by automated chain tracing with ARP/wARP [13]. Crystallographic refinement was carried out against remote data (Table 1 ). Iterative cycles of model building and refinement were performed with TURBO-FRODO and CNS [14]. The last stage of refinement was performed with REFMAC5 in the CCP4 suite [12]. Data collection and refinement statistics are summarized in Table 1. Fig. 2A–D was produced with PyMOL (http://pymol.sourceforge.net/). Fig. 2E was produced with CueMol (http://cuemol.sourceforge.jp/en/). The atomic coordinates and structure factors have been deposited in the RCSB Protein Data Bank (PDB) with the Accession Code 3O59. The interaction of the DP2(1–300) domain with 3′-recess DNA, single-strand DNA (ssDNA), and double-strand DNA (dsDNA) was quantitatively analyzed on a BIAcore X apparatus (Biacore) at 25 °C. To generate biotinylated 3′-recess DNA, 500 pmol of 5′ biotinylated A strand (5′-bio-GAGCTAGATGTCGGACTCTGCCTCAAGACGGTAGTCAACGTGCACTCGAGGTCA-3′, 54mer) was boiled with 500 pmol of ssDNA (5′-TGACCTCGAGTGCACGTTGACTACCGT-3′, 27mer) in 20 μl of binding buffer (50 mM Tris–HCl buffer (pH 8.0) containing 10 mM NaCl, 5 mM EDTA, and 0.005% Tween 20) for 5 min and cooled down to room temperature for 30 min. To generate 5′ biotinylated dsDNA, 500 pmol of the biotinylated A strand was annealed with 500 pmol of the complementary ssDNA (54mer) in 20 μl of the binding buffer. The biotinylated A strand was also used as a ssDNA ligand. Each template DNA was immobilized as a ligand on a streptavidin–dextran layer on the surface of the Sensor Chip SA (Biacore) at 660–970 resonance units with the binding buffer. The flow cell was routinely equilibrated with running buffer (50 mM Tris–HCl buffer (pH 8.0) containing 50 mM NaCl, 1% glycerol, and 0.005% Tween 20). The concentrations of DP2(1–300) as analytes varied from 0.25 to 80 μM. The obtained sensorgrams were analyzed by evaluation software (Biacore) on the basis of a simple 1:1 binding model. DP2 contains the 166 residue proteinous intron (mini-intein) between Asn954 and Cys1121, as shown in Fig. 1 A. After genetically removing the mini-intein, the resultant 206-aa polypeptide of DP2(792–1163) was fused with an N-terminal histidine-tag (20 residues) and overexpressed in E. coli cells. The fused catalytic domain DP2(792–1163) was expressed as insoluble inclusion bodies and purified by Ni-affinity chromatography in 6 M urea, as described previously [15]. The catalytic domain refolding trial was started in 3 M urea by mixing the catalytic domain with an equimolar amount of the N-terminal domain of DP2 (i.e., the protein concentration of each domain was 44 μM in 3 M urea), followed by successive stepwise dialysis with 2 M urea and 50 mM Tris–HCl buffer (pH 8.0) containing 100 mM NaCl for 5 h and then with the same buffer containing 100 mM NaCl without urea overnight. The dialyzate was centrifuged at 15 000×g for 15 min to remove the precipitate and concentrated with a Centricon YM-10 (Amicon). The resultant soluble complexes were purified with a Superdex 200 (10/300 GL) gel filtration column (GE Healthcare) equilibrated with 50 mM Tris–HCl buffer (pH 8.0) containing 100 mM NaCl using the FPLC system. The intensity of the stained protein bands on SDS–PAGE gel was measured with a ChemiDoc XRS scanning imager (Bio-Rad) equipped with the Quantity One ver. 4.4. software (Bio-Rad). The molecular ratio of DP2(1–100) to DP2(792–1163) in the complex was determined by calculations involving the scanning intensity and the molecular mass of each band. The crystal structure of DP2(1–300) was determined at 2.2 A resolution according to the MAD method using a selenomethionine (SeMet) derivative (Table 1). The refined model contains the residues (48–291) and 107 water molecules; i.e., the N-terminal and C-terminal ends of DP2(1–300) are absent. Hereafter, the crystal structure is designated as DP2(48–291). Continuous 2F o − F c electron densities are shown in Supplementary Fig. 1. The stereochemistry of the model, as evaluated with PROCHECK [16], indicates that 89.6% of residues are located in the most favorable region and 10.0% are located in the additionally allowed region. DP2(48–291) mainly constitutes an α-helical structure containing nine α helices and three β strands (the amino acid sequence and secondary-structural elements are depicted in Fig. 1B). DP2(48–291) has an ellipsoidal shape, and its dimensions are approximately 45 × 30 × 30 A3 (Fig. 2 A and B). Three β strands (β1, β2, and β3) form a twisted β-sheet at the center of one face, and the β sheets are surrounded by α4, α5, and α8 helices. The β3 strand and α8 are connected by a 22-residue-long kinked loop, which is located on the surface and is in the vicinity of the β-sheet and α6 helices. The relatively long α8 helix is kinked at Ala-260. The coordinates of DP2(48–291) were submitted to the web server SSM (Secondary-Structure Matching program) [17] to search for other proteins with a similar folding pattern in the PDB. Due to its low structural similarity with other proteins, except for archaeal DP2 (highest Z score = 1.4), the folding of DP2(48–291) was considered to be unique. According to molecular surface representations, the front and bottom of DP2(48–291) are rich in non-charged residues (Fig. 2C). On the other hand, the back of DP2(48–291) is rich in charged residues (Fig. 2D); therefore, the front and bottom are likely to interact with other domains. According to secondary-structure prediction using the amino acid sequence and the PSIPRED program [18], it was considered that the region between residues 8 and 33 is likely to form an α-helix. In the remaining space of the crystal packing shown in Fig. 2E, we manually built a model of residues 1–47 of DP2(1–300) (unpublished data). Helical wheel analysis of DP2(8–33) was carried out with the DNASIS-Mac ver. 2.0 software (http://catalog.takara-bio.co.jp/product/). Half of the wheel is hydrophobic and covered with 11 hydrophobic residues (Met10, Tyr13, Phe14, Met16, Leu17, Ile21, Ala24, Tyr25, Ile27, Ala28, Ala31), and the other side is hydrophilic. These hydrophobic characteristics were well conserved in the N-terminal regions of archaeal DP2 as shown in Fig. 2F. As we reported previously the interaction between the catalytic domain, DP2(792–1163) and the N-terminal 100-amino acid region of DP2, it was expected that the denatured DP2(792–1163) might be refolded when it was mixed with DP2(1–100). Accordingly, refolding of the hydrophobic domain DP2(792–1163) was investigated by mixing it with an equimolar amount of the N-terminal domain of DP2 in 3 M urea followed by successive stepwise dialysis to remove urea. Sampling was carried out before and after dialysis, and refolding efficiency was examined by SDS–PAGE after the precipitate had been removed. Fig. 4 A shows the SDS–PAGE pattern of the supernatant and precipitate separated by centrifugation. For the refolding of DP2(792–1163) with DP2(1–50), approximately half of DP2(792–1163) remained in the supernatant due to complex formation with DP2(1–50), judging from band intensity as shown in Fig. 4A, lane 2; although the complex of DP2(792–1163) with DP2(1–50) formed a precipitate to some extent (in Fig. 4A, lane 3). As a control, the refolding of DP2(792–1163) in the absence of DP1(1–50) was done by the same procedure, however, DP2(792–1163) was recovered completely as a precipitate (data not shown). For the refolding of DP2(792–1163) with DP2(50–300), DP2(792–1163) disappeared completely from the supernatant without urea (Fig. 4A, lane 5) and was recovered as a precipitate (Fig. 4A, lane 6). These results suggest that DP2(1–50) is essential for complex formation and the refolding of DP2(792–1163) in soluble form, although it should be addressed that the folding observed is not in physical condition. Since the tight interaction between DP2(1–100) and the hydrophobic domain DP2(792–1163) was described previously [15], the putative α-helix (8–33) might associate with the hydrophobic domain DP2(792–1163) through its hydrophobic characteristics as shown in Fig. 2F. Furthermore, the complexes DP2(1–300)DP2(792–1163) and DP2(1–100)DP2(792–1163) were recovered as a soluble form without any precipitate and purified by gel filtration. In the case of DP2(1–100)DP2(792–1163), the elution profile was investigated in detail. The molecular weight and the structural uniformity of the purified complex, DP2(1–100)DP2(792–1163), were analyzed by gel filtration with a Superdex 200 column. The protein was eluted as a sharp peak, suggesting a uniform molecular mass, as shown in Fig. 4B, whose molecular mass was estimated to be 72 kDa according to the elution profiles of marker proteins. The N-terminal 100 residues, DP2(1–100), are sufficient to refold the catalytic DP2(792–1163) domain into a soluble complex. The peak fraction was analyzed by SDS–PAGE and Western blotting with anti-histidine tag antibody as shown in lanes 1 and 2 of Fig. 4C, respectively. The molar ratio of DP2(1–100) to DP2(792–1163) in the complex was estimated to be 1 to 1 according to their band intensities and molecular weights, as shown in lane 1 of Fig. 4C. A faint 50 kDa band, which is shown in lane 1 of Fig. 4C, was confirmed to be dimeric DP2(792–1163) because of the Western signal for his-tagged DP2(792–1163) (lane 2 of Fig. 4C), although the signal intensity of the dimer band was markedly higher than that of the monomer. Denaturation of DP2(792–1163) by boiling with SDS-buffer might be insufficient, resulting the presence of the dimer band. A weaker signal against the monomer band than the dimer one in lane 2 might be due to the weaker affinity of the antibody to the monomer protein. These findings suggest that the native 72 kDa molecule has a dimer structure, [DP2(1–100)DP2(792–1163)]2. The circular dichroism (CD) spectra of the purified dimeric form, [DP2(1–100)DP2(792–1163)]2, were measured at 25 °C and 85 °C as shown in Supplementary Fig. 2A. Both CD spectra showed a double-minimal shape that is characteristic of helical polypeptides [19]. The difference in ellipticity at 220 nm between 25 °C and 85 °C was not marked as shown in Supplementary Fig. 2B, suggesting that the dimeric structure is stable at 85 °C. Using the fluorometric method, the thermal transition of the dimer form was also investigated between 20 °C and 85 °C, as shown in Supplementary Fig. 2C. As one tryptophan residue and twelve tyrosine residues were present in the monomeric form, the fluorometric analysis focused on the abundant aromatic residue Tyr. The temperature-dependence of the emission at 300 nm of Tyr residues was measured with excitation at 250 nm. The fluorescence intensity of Tyr residues at 300 nm decreased linearly from 0.44 to 0.16 as the temperature increased. This result is likely to reflect the temperature dependence of the emission of Tyr residues at 300 nm, as reported previously [20], without dissociation or conformation changes, indicating that the dimer [DP2(1–100)DP2(792–1163)]2 is stable at 85 °C with no dissociation to monomers or drastic conformational changes. An asymmetric dimeric structure is believed to be reasonable for the coupling of both leading and lagging strand DNA synthesis at the DNA replication fork [21, 22]. The dimeric structure, [DP2(1–100)DP2(792–1163)]2, seems to be an important structure in the catalytic center of heterotetrameric PhoPolD, due to its thermostability and molecular uniformity. As a way of investigating the evolutional relationships between PolD and eukaryotic B-family DNA polymerases [3, 7, 23], solving the molecular structure of the dimer [DP2(1–100)DP2(792–1163)]2 will help us to understand the structure of archaeal-eukaryotic type DNA replicase and its DNA replication mechanism. We thank H. Morii of AIST for performing the fluorometric analysis and are also grateful to K. Hiramoto for their technical support. This work was supported in part by a Grant-in-Aid for Science Research (18608005 to I.M.) from the . Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2010.12.040. Fig. S1. Stereo representation of the 2F o − F c electron densities of DP2(1–300). Electron densities are contoured at 1.5σ. Each residue around Pro218, Trp275, and Trp277 is shown as a stick model (C atom, yellow; N atom, blue; O atom, red). Water molecules are shown as red spheres. The figure was produced with CueMol (http://cuemol.sourceforge.jp/en/). Fig. S2. The dimer structure, [DP2(1–100)DP2(792–1163)]2 is heat stable. (A) CD spectra of the dimeric complex [DP2(1–100)DP2(792–1163)]2 were measured at 25 °C (black line) and 85 °C (gray line). The secondary structure of the complex was measured using an AVIV model 62DS circular dichroism (CD) spectrometer (Shimazu, Kyoto, Japan) and a quartz cuvette with a path length of 1.0 mm. The protein concentration was 4.67 μM in 100 mM Tris–HCl buffer (pH 8.0) containing 250 mM NaCl. Spectra were obtained between 200 nm and 250 nm with a bandwidth of 1.0 nm. (B) Temperature dependency of the ellipticity of the dimeric [DP2(1–100)DP2(792–1163)]2 at 220 nm between 25 °C and 85 °C. (C) Temperature dependency of the emission of Tyr residues at 300 nm for the dimer, [DP2(1–100)DP2(792–1163)]2. The conformation changes that occurred in the dimer complex under increasing temperature were analyzed by the fluorometric method using an FP-750 analytical fluorometer (JASCO Co., Japan) equipped with an ETC-272T temperature controller (JASCO) and a quartz cuvette with a path length of 10 × 10 mm. The protein concentration in the dimer form was 0.82 μM in 50 mM Tris–HCl buffer (pH 8.0) containing 100 mM NaCl. The temperature-dependence of the emission at 300 nm of Tyr residues was measured with excitation at 250 nm. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.