Báo cáo khoa học: Purification and characterization of organellar DNA polymerases in the red alga Cyanidioschyzon merolae

Purification And Characterization Of Organellar DNA Polymerases In The Red Alga Cyanidioschyzon Merolae

Tác giả: Takashi Moriyama, Kimihiro Terasawa, Makoto Fujiwara, and Naoki Sato

Lĩnh vực: Khoa học Đời sống, Trường Cao học Nghệ thuật và Khoa học, Đại học Tokyo, Nhật Bản

Nội dung tài liệu: Nghiên cứu này tập trung vào việc phân lập và đặc tính hóa các DNA polymerase có nguồn gốc từ bào quan (POPs) trong tảo đỏ Cyanidioschyzon merolae. Các nhà nghiên cứu đã xác định hai gen mã hóa các protein có họ hàng xa với Escherichia coli DNA polymerase I (PolA và PolB). Phân tích phát sinh loài cho thấy PolB là một gen tương đồng với POPs, ám chỉ một nguồn gốc cổ xưa trước sự phát triển của quá trình quang hợp ở sinh vật nhân chuẩn. PolA, mặt khác, có họ hàng gần với E. coli DNA polymerase I. Nghiên cứu đã thực hiện phân lập PolB từ tế bào C. merolae và protein tái tổ hợp PolA. Các kết quả cho thấy sự khác biệt về độ nhạy với các chất ức chế tổng hợp DNA giữa PolA, PolB và E. coli DNA polymerase I. Phân tích miễn dịch và nghiên cứu định vị protein huỳnh quang xanh lá cây đã chứng minh PolA được định vị trong lục lạp, trong khi PolB có mặt ở cả lục lạp và ty thể. Biểu hiện của PolB được điều hòa bởi chu kỳ tế bào. Các phát hiện này gợi ý rằng PolB đóng vai trò trong quá trình sao chép của lục lạp và ty thể.

Mục lục chi tiết:

• Purification and characterization of organellar DNA polymerases in the red alga Cyanidioschyzon merolae
• Keywords
• Correspondence
• (Received 28 January 2008, revised 30 March 2008, accepted 1 April 2008)
• doi:10.1111/j.1742-4658.2008.06426.x
• Plants and algae have two types of organelles that originate from two distinct endosymbionts, namely, mitochondria and plastids [1]. Each organelle has its own genome, which is replicated by the replication machinery that resides in the respective organelle. The enzymes involved in replication of organellar genomes have been supposed to be encoded by the nuclear genome and transported to the organelles after their synthesis [2]. In animals and yeasts, DNA polymerase γ is involved in mitochondrial replication [3,4]. In photosynthetic organisms, purification of DNA polymerases has so far been attempted in chloroplasts [5-11] and mitochondria [6,12-15]. However, in the light of recent genomic data (see next paragraph), all of them were partially purified preparations of partially degraded enzymes. Findings based on such preparations suggested some similarity of the purified enzymes to the γ-type DNA polymerases with respect to their optimal conditions, inhibitor sensitivities, and template preferences. Despite such postulates, no gene encoding a homolog of DNA polymerase γ has been found in the completely sequenced genomes of plants and algae, including Arabidopsis thaliana [16]. It should also be noted that the presence of polymerase γ is not universal, and this is being verified by the accumulating data on completely sequenced genomes: Polymerase γ has been found in only opisthokonts (fungi and animals, or organisms with a single flagellum) but not in
• Abbreviations
• ACTD, activated calf thymus DNA; dd, dideoxy; GFP, green fluorescent protein; PAA, phosphonoacetate; Poll, DNA polymerase I; POP, plant organellar DNA polymerase; SiR, sulfite reductase; Trx, thioredoxin.
• FEBS Journal 275 (2008) 2899-2918© 2008 The Authors Journal compilation 2008 FEBS
• 2899
• Plant organellar DNA polymerase
• T. Moriyama et al.
• bikonts (organisms with two flagella: see [17] for the evolution of these major divisions of organisms).
• Sakai et al. [18-20] isolated nucleoid-enriched frac-tions from chloroplasts and mitochondria of tobacco leaves. They detected DNA synthetic activity in the nucleoid fraction, and showed that the apparent molecular mass of the polypeptide bearing this activ-ity was similar in the nucleoid fractions of mitochon-dria and chloroplasts. They suggested that a previously unknown DNA polymerase(s) is involved in the DNA synthesis in higher plant organelles.
• DNA polymerases with distant homology to Escheri-chia coli DNA polymerase I (Poll) have been isolated in rice, A. thaliana, and tobacco [21-24]. Localization to both plastids and mitochondria, enzymatic activity or expression analysis of these enzymes were reported in these articles. The authors of these papers consider that the enzyme is the replicase of the organellar gen-ome in plants, but we still need evidence that these enzymes are really acting as the replication enzymes in plant organelles. In addition, higher plants have two enzymes with high homology to each other, so the functions of individual enzymes within the cell, such as replication, repair, or both, remained unclear. These polymerases have been called ‘Poll-like’ or given similar names, but the enzymes are distinct from bacterial Poll [25] and a new name is needed. We called these DNA polymerases plant organellar DNA polymerases (POPs). It has not yet been possible to purify POPs from plant tissues or organelles without degradation. There is no report on the enzymatic characteristics of the native POPs.
• We attempted to purify a POP from Cyanidioschy-zon merolae, a unicellular red alga originally isolated from Italian volcanic hot springs [26]. The C. merolae cell has a single plastid and a single mitochondrion, and is easily manipulated in laboratory experiments. Recently, the 100% complete genome sequence of this organism has been determined, and two potential genes encoding DNA polymerases with distant homology to E. coli Poll have been found in its nuclear genome [25,26]. We named them PolA and PolB. PolB is a homolog of POPs, but PolA is more similar to Poll.
• In the present study, phylogenetic analysis was first performed to establish their evolutionary origin. We also examined the organellar localization of PolA and PolB by immunoblotting and green fluorescent protein (GFP) expression analysis. We examined the enzymatic activity of PolB purified from C. merolae and recombi-nant PolA and PolB proteins in vitro, and compared their properties with those of E. coli DNA Poll (Klenow fragment) as well as of nucleoids of plastids and mitochondria of C. merolae cells. Finally, expres-sion of PolA and PolB in synchronous culture was examined.
• Results
• DNA polymerase genes in C. merolae
• The C. merolae genome project (http://merolae. biol.s.u-tokyo.ac.jp/ [26,27]) revealed 4775 ORFs on the 20 chromosomes, and various genes for DNA syn-thesis-related proteins were predicted. These included genes for subunits of nuclear DNA polymerases, a (CMF127C, CMI176C), δ (CMB052C, CMN199C, CMT350C), ε (CMH082C, CMQ098C, CMR360C), ζ (CMR103C), 1 (CMA062C), λ (CMQ426C), and REV1 (CMS282C). However, no gene for DNA poly-merase γ, the mitochondrial replication enzyme in yeasts and mammals, was detected. The absence of putative genes for polymerase γ homolog in the gen-ome is also true for a green alga, Chlamydo-monas reinhardtii [28] (Joint Genome Initiative http:// www.jgi.doe.gov/), a diatom Thalassiosira pseudonana [29], and the higher plants A. thaliana [16] and Oryza sativa [30]. Two distant homologs of bacterial Poll genes were found in C. merolae (Fig. 1A,B) [25]. We named the gene products PolA (on chromosome 20; CMT462C) and PolB (on chromosome 15; CMO270C). Both genes have no introns and are
• Fig. 1. Structure of the two Poll-like proteins, PolA and PolB, from C. merolae. Predicted amino acid sequences of PolA (A) and PolB (B), and a schematic comparison of the structures of DNA polymerases (C). In (A) and (B), the 5’→3′ exonuclease domain (pink background), the 3’→5′ exonuclease domain (blue background) and the polymerase domain (orange background) are indicated on the sequences. The 3’→5′ exonuclease domain of PolA is tentatively assigned here on the basis of very low similarity. The protein regions used for protein-tar-geting experiments (see Fig. 4) are indicated by underlines. In (C), the colored boxes indicate domains estimated from the Pfam database: pink, 5’→3′ exonuclease domain; blue, 3’→5′ exonuclease domain; orange, DNA polymerase domain; purple, primase domain; green, heli-case domain. Yellow boxes indicate characteristic motifs in the 3’→5′ exonuclease domain or the DNA polymerase domain. Crossed boxes indicate conserved sequences in bikonts and amoebozoa. Dotted boxes and boxes with stripes indicate conserved sequences in apicom-plexa and opisthokonts, respectively. In Pol y of Homo sapiens, the 3’→5′ exonuclease domain is not indicated, because the domain was not found by Pfam, although 3’→5′ exonuclease activity was reported for Pol y [3,4].
• 2900
• FEBS Journal 275 (2008) 2899-2918© 2008 The Authors Journal compilation 2008 FEBS
• Plant organellar DNA polymerase
• T. Moriyama et al.
• A PolA
• B PolB
• C 5′-3′ exo 3′-5′ exo DNA polymerase
• Escherichia coli_Pol I
• Anabaena sp. PCC 7120_Poll
• C. merolae_PolA
• C. merolae_PolB
• A. thaliana_At1g50840
• A. thaliana_At3g20540
• Oryza sativa
• Ostreococcus tauri
• Dictyostelium discoideum
• Tetrahymena thermophila
• Paramecium tetraurelia
• Plasmodium falciparum
• Primase
• Theileria annulata
• Homo sapiens_Poly
• Bacteria
• Bikonts and amoebozoa
• Opisthokonts Apicomplexa
• FEBS Journal 275 (2008) 2899-2918© 2008 The Authors Journal compilation 2008 FEBS
• 2901
• Plant organellar DNA polymerase
• T. Moriyama et al.
• transcribed, and 5′-RACE analysis identified three mRNA-capping sites for PolA and one for PolB (supplementary Fig. S1). On the basis of this map-ping, the PolA and PolB genes are predicted to encode polypeptides of 1451 and 905 amino acids, corresponding to molecular masses of 159 kDa and 102 kDa, respectively.
• PolA looks similar to bacterial Poll, because it has a complete set of domains normally present in Poll, namely, a 5’→3′ exonuclease domain, a 3’→5′ exonu-clease domain, and a DNA polymerase domain (Fig. 1A,C). However, the central domain has only very weak similarity to the authentic 3’→5′ exonucle-ase domain (E-value was 0.77) in Pfam analysis. We will describe the lack of 3’→5′ exonuclease activity later in Figure 9. PolB has a 3’→5′ exonuclease domain and a DNA polymerase domain (Fig. 1B,C). PolB is a homolog of POPs in A. thaliana (AtPolI-like A and AtPolI-likeB (At1g50840 and At3g20540) [23], rice [22,31], and tobacco (NtPolI-likel and NtPol-I-like2 [24]. POPs are also conserved in the Alveolata (Tetrahymena thermophila and Paramecium tetraurelia), and also in the Amoebozoa (Dictyostelium discoideum). In A. thaliana, the 5’→3′ exonuclease domain exists as distinct polypeptides, At1g34380 and At3g52050 [25], although it is not known whether these putative exonucleases function within the organelles with the POPs. The POP of D. discoideum only has a 5’→3′ exonuclease domain (Fig. 1C). Many of these putative polymerases bear a putative targeting sequence [21] for import into mitochondria or plastids. According to such comparisons, as well as other avail-able data, the POPs are the most probable candidates for the replication enzyme in plastids and mitochon-dria in plants and algae, in which no polymerase γ is found. However, the final identification is hampered by the fact that no native DNA polymerase has been purified from plants and algae. C. merolae enzymes could be better suited for this purpose, because the cells grow rapidly and the enzymes might be more stable because the temperature of growth is higher.
• Phylogenetic analysis of organellar DNA polymerases in plants and algae
• The putative organellar DNA polymerases identified as POP homologs exhibited some homology to bacterial Polls, which are members of family A DNA polyme-rases, as are DNA polymerases γ, θ, and v [32]. Struc-tural comparison indicates that the identified organellar DNA polymerases (PolA and PolB, as well as POPs) are distinct from DNA polymerases γ, θ, and v (Figs 1C and 2), but their phylogenetic relationship to bacterial enzymes needs to be carefully analyzed, because they might originate from bacterial endos-ymbionts, either cyanobacteria or a-proteobacteria. Multifunctional enzymes with a similar polymerase region were also reported in malaria parasites, and the possible relationship with secondary endosymbiosis is also an important question. Figure 2 shows a maxi-mum likelihood tree. An essentially similar minimal evolution tree was reported with a smaller number of sequences [25]. It is clear that the POPs, including C. merolae PolB, rice OsPOLP, and A. thaliana At1g50840 and At3g20540, as well as homologs of var-ious photosynthetic eukaryotes, belong to a well-defined clade, separated from any bacterial enzymes. Interestingly, putative polymerases of Tetrahymena, Phytophthora, Paramecium and Dictyostelium are also present in this clade. This indicates that the origin of organellar DNA polymerase of this type might pre-date the diversification of photosynthetic eukaryotes. The tree in Fig. 2 also shows that the plant and algal POPs did not originate from DNA polymerases of cyanobacteria (blue) or a-proteobacteria (red). In contrast, PolA of C. merolae is associated with Rhodothermus Poll. Another plant enzyme, A. thaliana At4g32700 (recently named TEBICHI [33]), as well as some eukaryotic enzymes, are included in a separate clade (DNA polymerases 0 and v). The two enzymes of malaria parasites (Pf1 and Pf2) are present in a sis-ter clade to POPs. These results suggest that PolB of C. merolae is an ortholog of POPs, and that the repli-cation enzymes of organelles were not provided by the
• Fig. 2. Phylogenetic tree of POPs and Poll. This is a maximum likelihood tree calculated by TREEFINDER version 3.6. Most sequences were obtained from cluster 473 of the dataset ALL95_4 in the Gclust database (http://gclust.c.u-tokyo.ac.jp/ [50]). Sources of additional sequences: Thermus aquaticus, J04639; Streptococcus pneumoniae, AE007320; Rickettsia felis, AJ238763; Rhodothermus marinus, AF121780; Aq. aeo-licus, AE000765; OsPoIP (formerly OsPoll-like), AB047689; tobacco, AB174898 (NtPoll-like1) and AB174899 (NtPoll-like2); Ostreococcus tau-ri, CAL55440; Caenorhabditis elegans W03A3.2 (theta), U50184; T7 phage, V01146; P. falciparum, AAN36724 (Pf1) and CAG25066 (Pf2); human, NP0002684 (gamma) and 075417 (theta). The two numbers (left/right) on each branch represent edge support values obtained by maximum likelihood analysis (left) and bootstrap values obtained by neighbor-joining analysis (right), respectively, and both values are shown in percentiles. An essentially similar tree was also obtained by the neighbor