A Novel Single-Strand Specific 3′–5′ Exonuclease Found in the Hyperthermophilic Archaeon, Pyrococcus furiosus

Nucleases play important roles in all DNA transactions, including replication, repair, and recombination. Many different nucleases from bacterial and eukaryotic organisms have been identified and functionally characterized. However, our knowledge about the nucleases from Archaea, the third domain of life, is still limited. We searched for 3′–5′ exonuclease activity in the hyperthermophilic archaeon, Pyrococcus furiosus, and identified a protein with the target activity. The purified protein, encoded by PF2046, is composed of 229 amino acids with a molecular weight of 25,596, and displayed single-strand specific 3′–5′ exonuclease activity. The protein, designated as PfuExo I, forms a stable trimeric complex in solution and excises the DNA at every two nucleotides from the 3′ to 5′ direction. The amino acid sequence of this protein is conserved only in Thermococci, one of the hyperthermophilic classes in the Euryarchaeota subdomain in Archaea. The newly discovered exonuclease lacks similarity to any other proteins with known function, including hitherto reported 3′–5′ exonucleases. This novel nuclease may be involved in a DNA repair pathway conserved in the living organisms as a specific member for some hyperthermophilic archaea.


Introduction
The stability of DNA largely depends on the accuracy of its repair systems, which remove DNA damage induced by exogenous and endogenous agents or introduced by DNA metabolism, such as replication [1,2]. If unrepaired or misrepaired, DNA damage can cause cell death. To mend the many different types of DNA damage, cells utilize DNA repair systems, including homologous recombination repair (HR), mismatch repair (MMR), and excision repair (NER, BER) [3,4].
Archaea, the third domain of life, has DNA replication factors that resemble those in Eukarya with respect to the amino acid sequences, suggesting that the mechanisms of DNA replication would be similar in eukaryotic and archaeal cells [5][6][7]. The DNA repair proteins are also conserved between Archaea and Eukarya, and structural studies of the archaeal homologs have yielded numerous important insights into the structures and functions of eukaryotic DNA repair proteins [8][9][10][11][12]. In addition, the DNA repair systems in the hyperthermophilic archaea are especially attractive to investigate, for understanding the phenomena allowing life to exist at extremely high temperatures. It has been suggested that thermophiles must have extremely efficient and specialized DNA repair systems to withstand not only the high temperatures but also other damaging factors, such as ionizing and ultraviolet radiation and chemical agents. The spontaneous mutation rate is accelerated at elevated temperatures, due to the formation of lesions in the DNA bases [13][14][15].
Pyrococcus furiosus is a hyperthermophilic archaeon that optimally grows at over 100uC under anaerobic conditions, and its DNA transactions have been partially characterized [9,16,17]. The exposure of P. furiosus cells to ionizing radiation reportedly resulted in chromosomal fragmentation, but a following incubation of the cells at 95uC resulted in chromosome reassembly [14]. These findings suggested that P. furiosus must have a highly efficient DNA repair system for DNA strand breaks, but the details about the components and mechanism of the system still remain unknown.
Deoxyribonucleases (DNases) operate as central components of most DNA repair systems, by executing or initiating DNA damage removal to promote cell survival and ensure genetic integrity [18,19]. DNases can be classified into structure-, damage-, or sequence-specific families, with respect to their substrate preferences. They can also be divided into exonucleases, which hydrolyze nucleic acids from either the 59 or 39 end, and endonucleases, which hydrolyze internal phosphodiester bonds without the requirement of a free DNA end. Many exonucleases have been identified in Eukarya and Bacteria [18], and their functions have been analyzed both in vitro and in vivo. Singlestranded DNA specific exonucleases are particularly abundant in E. coli, where they play important roles in several repair processes within the cell, such as MMR [4], HR [20], and tandem repeat stabilization [21]. Although similar functional exonucleases are expected to be involved in the repair systems in P. furiosus, no single-stranded DNA specific 39-59 exonuclease has been identified yet, except for the DNA polymerase-associated exonuclease activity [22,23]. To understand the DNA repair systems in more detail in the hyperthermophilic archaea, the unknown factors, including DNases, must be identified.
In this study, we identified a novel nuclease in P. furiosus cells. This enzyme degrades single-stranded DNA, but not doublestranded DNA, from the 39 to 59 direction. No conserved sequence motifs for the 39-59 exonucleases were detected in the deduced amino acid sequence of this protein. Furthermore, genes encoding the homologous sequence of this enzyme are present only in the Thermococci in Archaea annotated as hypothetical protein in the public databases, and therefore, it is a novel nuclease.

Results
Identification of a Nuclease Activity in the Heat-stable Protein Library from P. furiosus We searched the databases for an ORF bearing a sequence homologous to those of the known 39-59 exonucleases in the P. furiosus genome, to detect homologs of known exonucleases. However, no homologous sequence was found. Therefore, we tried to identify a novel enzyme containing a 39-59 exonuclease by screening for the activity from the heat-stable protein libraries derived from P. furiosus, as described in the Materials and Methods section [24]. Among 500 independent heat extracts of E. coli clones transformed by the cosmid-based gene library of P. furiosus, we isolated a clone producing a protein that degrades the synthetic oligonucleotide. The cosmid DNA containing a P. furiosus genomic DNA of 35 kbp was recovered from the E. coli clone exhibiting the activity. The inserted genomic DNA fragment was subcloned into the pUC118 plasmid vector after PstI digestion. Then, as a second screening, we searched for the nuclease activity in the heatresistant cell extract from each transformant. One clone with the target activity was found (lane 8 in Fig. 1A). The plasmid was prepared from clone #8, and the inserted fragment (about 5 kbp) was sequenced. As shown in Fig. 1B, 7 open reading frames (ORFs), among which 5 were full-length and 2 were truncated, were identified. Therefore, we cloned these 5 full-length ORFs individually into the pET21a expression vector, to determine which ORF is responsible for the degradation of ssDNA. As shown in Fig. 1C, the heat extract of the E. coli transformant containing PF2046 clearly expressed the nuclease activity.

Purification of the Recombinant Protein Encoded by PF2046
The recombinant protein encoded by PF2046 was successfully overproduced, by cultivating the E. coli cells bearing the expression plasmid, pPF2046, as described above, with IPTG induction. The protein was purified to near homogeneity ( Fig. 2A) by the three sequential chromatography steps described in the Materials and Methods section. From a one liter culture (3.4 g cells), 1.5 mg of homogeneous protein was obtained. The relative molecular mass of the purified protein determined from its migration position in the SDS-PAGE gel corresponded to 25,596, calculated from the deduced amino acid sequence. We designated the protein as PfuExo I.

The Oligomeric State of PfuExo I
To investigate the oligomeric state of PfuExo I in solution, analytical gel filtration chromatography was performed (Fig. 2B). The molecular weight of PfuExo I was estimated from the elution profile to be approximately 73,400 (Fig. 2C). As the theoretical molecular mass of PfuExo I is 25,596, as described above, the gel filtration results suggested that PfuExo I exists in a homotrimeric form in solution.
Detection of PfuExo I in the P. furiosus Cells Using the highly purified PfuExo I protein, a polyclonal antibody was prepared. A western blotting analysis revealed that a protein band corresponding to the size of the recombinant PfuExo I (25,596), which specifically reacted with the antibody, was detected in the cell extract prepared from an exponential phase P. furiosus cell culture (Fig. 3A). This band was not detected in the cell extract prepared from the stationary phase culture of the cells, probably because of physiological proteolysis. The smaller bands, in addition to the main band, detected in the exponential phase may indicate that the function of PfuExo I is regulated with relatively short turnover in the P. furiosus cells. The appearance of an immunoreactive band at around 80,000 (Fig. 3A) is consistent with the results from the SDS-PAGE of the recombinant PfuExo I without heat treatment (Fig. 3B). The protein at 80,000 is also likely to be PfuExo I that was not completely denatured. The trimeric structure of this protein should be extremely stable, even in the presence of SDS. Properties of the Exonuclease Activity of PfuExo I The nuclease assay was performed using highly purified PfuExo I and a 31 nucleotide (nt)-long 59-32 P-labeled deoxyoligonucleotide. As shown in Fig. 4A, the deoxyoligonucleotide substrate became shorter as the reaction time progressed. This result clearly supports the proposal that the purified PF2046 protein actually has 39-exonuclease activity. Similar to other nucleases, the activity of PfuExo I was dependent upon a divalent cation, and Mg 2+ was the most preferable (data not shown). Ladder-like bands were detected on the gel, indicating the progress of the cleavage reaction. It is interesting that the ladder bands corresponded to exactly every two nucleotides judged from the positions of the size marker bands, 7, 10, and 14 mers with the sequence same as that of substrate DNA. Since the DNA substrate with the (AC)-repeated sequence was used for the nuclease assay, some nucleotide preference may exist for the cleavage by PfuExo I. Therefore, homooligomers were used as the substrates for the PfuExo I reaction. As shown in Fig. 4B, the homo dT 30 nt-long was the most preferable substrate for PfuExo I. The dC 30 nt-long DNA was the second best substrate, and the dA 30 nt-long was the least preferred. The dG 30 nt-long oligonucleotide could not be tested, because it could not be appropriately synthesized. These results support the idea that the ladder bands in Fig. 4A result from cleavage at the position of dC. However, the ladder bands in Fig. 4B also correspond to every two nucleotides. One more notable property of PfuExo I is that the degradation of the DNA strand does not proceed to the 59-end, but stops at the 4 th -6 th nucleotide from the 59-end. This probably occurs because PfuExo I needs to grasp a 4-6 nt-long DNA strand to express its nucleolytic activity.
To obtain further evidence that PfuExo I degrades the singlestranded DNA in the 39 to 59 direction, the nuclease assays were performed using 39 and 59-overhang DNAs. As shown in Fig. 5, the reaction products were detected only from the 39-overhang DNA substrate, and the cleavage reaction stopped at the 30-mer. This is the position of the 3 rd nucleotide from the end of the double-stranded region. This result supports the proposal that PfuExo I is an exonuclease with 39 to 59 polarity, and its substrate DNA must be at least a few nucleotides long for PfuExo I to bind.
It is interesting to analyze the degradation mode of the PfuExo I from the enzymological aspect. We did several experiments investigating how processive this enzyme is for its cleavage of the DNA strand. Constant PfuExo I and increasing amounts of substrate DNA were reacted. The cleavage pattern showed that the amounts of DNA strands with shorter sizes decreased with the

Examination of the Endonuclease Activity of PfuExo I
To determine whether PfuExo I also has endonuclease activity, endonuclease assays were performed using the circular forms of single-and double-stranded DNAs. As shown in Fig. 6, no cleavage was detected with either ssDNA or dsDNA, even after a 30-min incubation. This result indicated that PfuExo I requires the terminus to express the nuclease activity.

Comparison of the Binding Affinities of PfuExo I to Various DNA Substrates
To further characterize PfuExo I, gel mobility shift assays were performed using different types of DNA in the absence of Mg 2+ , to prevent substrate loss due to degradation, as shown in Fig. 7. The shifted bands should correspond to the DNA-PfuExo I complexes. Multiple shifted bands were observed with the ssDNA substrate (Fig. 7A). As slower migrating complexes were observed with increasing protein concentrations, PfuExo I may bind ssDNA nonspecifically at higher concentrations. Slight shifts were observed when dsDNA was used (Fig. 7B). Shifted bands were distinctly observed from both the 39-and 59-overhang DNAs, but only one band was observed (Fig. 7C, and 7D), suggesting that the preferential binding of PfuExo I occurred at the single-stranded region, which is not long enough for the second binding. These observations are consistent with the results of the nuclease assays, showing the single-stranded specific cleavage. Various lengths of ssDNAs were used for the gel shift assays and distinct shift band was observed from the ssDNA of 7 nt-long (data not shown), supporting that 7 nt-long is sufficient for PfuExo I to stably grasp DNA.

PfuExo I Homologs in Archaea
The deduced amino acid sequence of PfuExo I is not similar to any proteins from eukaryotic and bacterial organisms in the public databases. From a comprehensive search of the public databases, ORFs with highly similar sequences were detected only from the organisms that belong to the Thermococcaceae (6 Pyrococcus and 10 Thermococcus species) in the Archaeal domain. The multiple alignment of the ORF sequences from P. furiosus, P. horikoshii, P. abyssi, and T. kodakarensis is shown in Fig. 8. The sequence identities among these proteins are 70% on average. The newly discovered nuclease likely functions in DNA repair in Thermococcales lineage. It would be interesting if these organisms have a specific DNA repair system including this protein for the DNA damages occurring due to the high temperatures of their habitat. activity of PfuExo I were performed, using dAC31 ssDNA labeled with 32 P at the 59-end. PfuExo I (10 nM, as a trimer) was incubated with 5 nM DNA at 65uC. For each time point, aliquots were removed from the reaction and quenched. These samples were subjected to PAGE on an 18% gel containing 8M urea, and the degradation products were visualized by autoradiography using TyphoonTrio (GE Healthcare). The 7-nt, 10-nt, and 14-nt oligonucleotides with the same sequence as dAC31, and [c-32 P]ATP were loaded alongside to provide markers. (B) dA30, dC30, and dT30 were used as substrates with 5 nM of PfuExo I. The samples were subjected to PAGE on a 12% gel containing 8M urea, and the degradation products were visualized by autoradiography. The 10-nt, 15-nt, and 20-nt poly dAs were loaded alongside to provide markers. doi:10.1371/journal.pone.0058497.g004

Discussion
Archaea are unique organisms that are evolutionally different from bacterial and eukaryotic organisms. To understand their DNA repair systems, several archaeal proteins homologous to the eukaryotic and bacterial proteins involved in the DNA repair processes have been identified and biochemically analyzed to date. The hyperthermophilic archaea should have especially efficient DNA repair systems, due to their habitation in extreme environments. However, the DNA repair system in Archaea is still not well understood, and many more efforts are required to determine the similarities and differences between the archaeal DNA repair system and those of Bacteria and Eukarya. For example, the thermophilic archaea lack the NER damagerecognition proteins (eukaryotic XPA and XPC or bacterial UvrA and UvrB) and the MMR-specific proteins (MutS and MutL) [25]. Every factor that seems to be involved in DNA repair systems must be identified to elucidate the molecular mechanisms of damaged DNA repair in Archaea, and the exonucleases definitely play important roles in the processes. It would be especially interesting to determine whether any unique repair ability is present in the third domain of life.
In E. coli cells, several proteins function as single-stranded DNA specific exonucleases. RecJ, Exo I, Exo VII, and Exo X are involved in MMR and HR. In Archaea, a 59-39 exonuclease, NurA, is conserved only in the thermophilic archaea, and the nurA gene is organized in an operon structure with rad50 and mre11, which are involved in double-stranded break repair [26]. This nuclease probably plays an important role in generating 39 singlestranded DNA during archaeal HR, together with Mre11 and Rad50. HerA, a bipolar DNA helicase, is also present in the operon, and is involved in this DNA processing system [27]. In addition, several genes with sequences similar to that of the bacterial RecJ nuclease are present in the archaeal genomes [28]. A recent report showed that one of the RecJ homologs in T. kodakarensis stably interacts with the GINS complex, an essential factor for both the initiation and elongation processes in DNA replication, and its 59-39 exonuclease activity is stimulated by the interaction with GINS [29]. The authors designated this protein as GAN (GINS-associated nuclease), and proposed that GAN is involved in lagging strand processing. It is still not known if the bacterial RecJ-like proteins are involved in some repair system in the archaeal cells.
This is the first report to describe a single-stranded specific 39-59 exonuclease in Archaea. The amino acid sequence of the identified protein lacks obvious similarity to the known 39-59 exonucleases, which have some conserved motifs [18], and therefore, it is a new nuclease family member. At this point, it is not easy to predict the exact function of this nuclease, since it has no homolog in either Bacteria or Eukarya. The genes encoding sequences homologous to this enzyme are found only in the Thermococcales, although more than 140 archaeal genomes have been completely sequenced. It is most likely that the DNA repair systems are conserved in the living organisms, but the diverse members are involved in these processes in various organisms. However, because of the specific habitation, the organisms in Thermococcales may have a unique pathway for nucleic acid metabolism. The DNA of hyperthermophilic archaea is known to be extremely resistant to breakage in vivo by radiolysis and thermolysis. DiRuggiero et al. reported that the amount of mRNA for PF2046, corresponding to PfuExo I, increased after ionizing irradiation [30]. The fact that the chromosomal fragmentation occurring upon the exposure of P. furiosus cells to ionizing radiation was quickly ameliorated by an incubation of the cells at 95uC [14] suggests that P. furiosus must have a highly efficient DNA repair system for DNA strand breaks. PfuExo I may be one of the crucial enzymes in this pathway. Ionizing radiation, radiomimetic drugs, and to some extent, all free radical-based genotoxins induce DNA double-strand breaks by oxidative fragmentation of DNA sugars. Most of the breaks bear terminal 39-phosphate or 39-phosphoglycolate moieties [31][32][33]. Although we examined the end-processing activity of PfuExo I using synthetic oligonucleotides with a phosphate at the 39-end, the enzyme could not excise ssDNA (data not shown). Therefore, another unknown factor, such as a phosphatase, may be required  to remove the 39 phosphate before PfuExo I functions, if this nuclease participates in end-processing.
To prove that PfuExo I is actually involved in some DNA repair system in P. furiosus, genetic studies using a mutant strain lacking or deficient in the encoding gene are necessary. A few genetic investigations have been performed on the DNA repair genes in Archaea. RadA functions in the repair of DNA damaged by UV and methylmethane sulfonate [34], and Mre11 is important for the repair of DNA double-strand breaks, whereas Rad50 is dispens-able [35] in Haloarchaea. In addition, the uvrA, uvrB and uvrC genes are required for the repair of ultraviolet light-induced DNA photoproducts in Haloarchaea [36]. In hyperthermophilic Archaea, gene disruption methods and genetic tools were first developed for the crenarchaea of the genus Sulfolobus [37]. These genetic tools were used to analyze the functions of the genes involved in the DNA replication and repair processes in S. islandicus [38]. For anaerobic hyperthermophilic euryarchaea, genetic tools for gene targeting have been established in T.  kodakarensis [39,40]. We obtained knock-out mutants of the genes encoding the proteins Xpb, Xpd, Hef, Hjc, and Hjm, and investigated the sensitivity of all of these mutants to several different types of DNA damaging agents, including UV, c-rays, mitomycin C, and methylmethane sulfonate [41]. In the same study, we also proposed that the radA, mre11, rad50, herA, and nurA genes are essential, because deletion strains for these genes could not be obtained. The HR event may be especially important for repairing the DNA damage caused by the high temperature required for T. kodakarensis viability. Using this technique, we could make a deletion mutant of the homolog of PfuExo I in T. kodakarensis (TK1646) and characterize it. This experiment is now underway. P. furiosus cells with high transformation efficiency have been described recently [42,43], and therefore, the creation of PF2046-disrupted cells is also possible.

DNA Substrates
The following oligonucleotides were synthesized by Sigma-Aldrich Japan K. K.

Cloning and Identification of the PfuExo I Gene
The cosmid-based genomic library, in which each of clone contains the P. furiosus DNA fragment with 35-40 kbp, was prepared as described previously [24]. Then, heat-stable protein extracts were obtained from 500 independent clones by the heat treatment at 80uC for 10 min. The heat-stable extracts were used for the screening of DNase activity. Cosmid DNA was prepared from the clone exhibiting the heat-stable DNase activity, and was partially digested by PstI. The DNA fragments were inserted into pUC118, and the resultant plasmids were introduced into E. coli JM109. The heat-stable extracts were prepared from each clone, and were used for the DNase assay. The DNA fragment from the positive clone was analyzed by DNA sequencing, which revealed the presence of 5 full-length ORFs in the 5-kbp fragment. These 5 ORFs were amplified by PCR directly from the genomic DNA of P. furiosus. Each amplified gene was cloned into the pGEM-T easy vector (Promega). The cloned genes were digested by either NdeI-BamHI or NcoI-BamHI, and were inserted into the corresponding sites of pET21a or pET21d. The constructed plasmids were designated as pPF2045, pPF2046, pPF2047, pPF2048, and pPF2049, respectively. E. coli cells were transformed with the plasmids, and heat-stable extracts were prepared from each clone and used for the DNase assay.
Production and Purification of PfuExo I E. coli BL21-codon plus (DE3)-RIL cells (Novagen) harboring pPF2046 were grown at 37uC, in 1 liter of LB medium containing 50 mg/ml ampicillin and 34 mg/ml chloramphenicol. When the E. coli culture reached an A 600 of 0.52, expression of the gene encoding PfuExo I was induced by adding IPTG to 0.1 mM. After cultivation at 18uC for 24 hours, the cells were harvested by centrifugation, and the cell lysate was prepared by sonication in 40 ml of buffer A (50 mM Tris-HCl, pH 8.0, 10% (v/v) glycerol, 0.5 mM DTT, and 0.1 mM EDTA) containing 0.5 M NaCl and 1 mM PMSF. After centrifugation for 20 min at 23,7006g, the supernatant was incubated at 80uC for 30 min to remove most of the E. coli proteins, and then centrifuged again. To the heated supernatant, polyethyleneimine was added to 0.15% (w/v). The solution was stirred on ice for 30 min, and centrifuged for 10 min at 23,7006g. Ammonium sulfate was added to the supernatant to 80% saturation, and the solution was stirred on ice for 1 hour. The precipitate from the ammonium sulfate treatment was suspended in 8 ml of buffer A containing 1 M ammonium sulfate, and applied to a hydrophobic column (HiTrap phenyl, GE Healthcare UK Ltd), which was developed with a linear gradient of 1.0-0 M ammonium sulfate. The PfuExo I-containing fractions were applied to an anion-exchange column (HiTrapQ, GE Healthcare UK Ltd), after dialysis against buffer A containing 0.1 M NaCl. The column was developed with a 0.1-1.0 M NaCl gradient. The active fraction eluted at 0.15 M NaCl was diluted to 0.015 M NaCl, and applied to an affinity column (HiTrap heparin, GE Healthcare UK Ltd), which was developed with a linear gradient of 0.1-0.8 M NaCl. The eluted fractions were concentrated and stored at 4uC. The protein concentrations were determined by measuring the absorbance at 280 nm, with an extinction coefficient of 35,410 M 21 cm 21 , which was obtained by the method described earlier [44].

Exonuclease Assay
The heat-stable protein extracts or purified PfuExo I was incubated with 5 nM of various 32 P-labeled substrate DNAs in a standard reaction mixture, containing 20 mM Tris-HCl (pH 8.8), 5 mM MgCl 2 , 50 mM NaCl, and 0.1 mg/ml BSA. Reaction time and temperature are described in each figure legend. The reactions were stopped with an equal volume of deionized formamide. These samples were subjected to polyacrylamide-8 M urea gel electrophoresis, and the degradation products were visualized by autoradiography.

Endonuclease Activity
Purified PfuExo I (100 nM) was added to the reaction mixture containing 5 ng/ml of pBR322 or M13 mp18 ssDNA (Takarabio), in the standard reaction condition for the exonuclease assay for 30 min at 65uC. The reactions were terminated by the addition of an equal volume of phenol. The reaction products were fractionated by electrophoresis on a 1% agarose gel, and the DNA was visualized by ethidium bromide staining.
Western Blot Analysis P. furiosus cells (4.5610 11 cells) were disrupted by sonication in 15 ml of the buffer containing 50 mM Tris-HCl, pH 7.0, 0.5 mM dithiothreitol, 0.1 mM EDTA, and 10% glycerol with proteinase inhibitor (Complete TM , Roch Diagnostics GmbH), and the extract was obtained by centrifugation. The P. furiosus cell extract and the purified PfuExo I protein were separated by 14% SDS-PAGE, blotted onto PVDF membranes, and reacted with anti-PfuExo I antiserum, prepared by immunizing a rabbit with the recombinant PfuExo I protein. The bands were detected by using ECL (GE Healthcare UK Ltd), according to the supplier's recommendations.
DNA Binding Activity of PfuExo I Different concentrations (0-1,000 nM) of purified recombinant PfuExo I were incubated with 2.5 nM of various 32 P-labeled substrate DNAs, in a reaction mixture containing 20 mM triethanolamine, 50 mM NaCl, 0.1 mg/ml BSA, and 0.1 mM EDTA, for 20 min at 37uC. The reaction mixtures were treated with 0.1% glutaraldehyde, to fix the protein and DNA. The protein-DNA complexes were fractionated by 4.5% native PAGE and visualized by autoradiography, using an FLA5000 bioimage analyzer (FUJIFILM).

Computer Analysis of the Amino Acid Sequences
Search for the homologous sequences in the databases with BLAST was carried out at a website (http://blast.ncbi.nlm.nih. gov/Blast.cgi). CLUSTALW (http://www.genome.jp/tools/ clustalw/) was used for the amino acid sequence comparison of the PfuExo I and other Thermococcal homologs. It should be noted that the genome sequence encoding the homolog in P. horikoshii lacked one nucleotide at 514 th from the start site of the ORF (PH0067, accession number; NP_142085.2), and this deletion caused a frame shift in the downstream of the ORF. Therefore, we used our own sequence data, which codes for the amino acid sequence with high homology in the entire ORF, as PH0067 for the comparison. The other sequences were picked up from the database as follows. PF2046; NP_5797751, PAB2293; NP_125767.1, TK1646; YP_184059.1.