Stage-specific expression of Caenorhabditis elegans ribonuclease H1 enzymes with different substrate specificities and bivalent cation requirements Hiromi Kochiwa 1,2 , Mitsuhiro Itaya 1,3 , Masaru Tomita 1,4 and Akio Kanai 1 1 Institute for Advanced Biosciences, Keio University, Tsuruoka, Japan 2 Graduate School of Media and Governance, Keio University, Fujisawa, Japan 3 Mitsubishi Kagaku Institute of Life Sciences, Machida, Japan 4 Department of Environmental Information, Keio University, Fujisawa, Japan An enzyme that specifically degrades the RNA strand of RNA–DNA hybrids was first characterized from extracts of calf thymus tissue [1] and was named ribo- nuclease H (RNase H; EC 3.1.26.4) [2]. This protein has also been identified in viruses [3], bacteria [4,5], and archaea [6,7], indicating the essential nature of its roles in cellular metabolism. Prokaryotic RNase H enzymes are divided by sequence similarity into three groups: HI, HII, and HIII. On the other hand, there are two types of eukaryotic RNase H: RNase H1 is homologous to prokaryotic RNase HI, and RNase H2 is homologous to prokaryotic RNase HII. There is a distinct difference between the two types of eukaryotic RNase H in terms of the bivalent metal ion cofactor. Eukaryotic RNase H1 requires Mg 2+ ions as a cofac- tor, cannot use Mn 2+ ions as a cofactor [8], and is inhibited by the addition of Mn 2+ ions [9,10]. In con- trast, Escherichia coli RNase HI requires both Mg 2+ and Mn 2+ ions for its activity [11] and contains one Mg 2+ -binding site or two Mn 2+ -binding sites [12,13]. Eukaryotic RNase H2 is activated by either Mn 2+ or Mg 2+ ions, but E. coli RNase HII is activated only in the presence of Mn 2+ ions [14]. Phylogenetic analysis suggests that Mn 2+ -dependent RNase H is universally present from bacteria to eukaryotes [15]. The primary structures of prokaryotic RNase HI and eukaryotic RNase H1 also differ from each other. Most eukaryotic RNase H1 enzymes consist of a non- RNase H domain at the N-terminus and an RNase H domain at the C-terminus, in contrast with prokaryotic RNase HI, which contains only the RNase H domain. This eukaryotic non-RNase H domain was first Keywords alternative splicing; C. elegans; development; metal ion; RNase H1 Correspondence A. Kanai, Institute for Advanced Biosciences, Keio University, Tsuruoka 997- 0017, Japan Fax: +81 235 29 0525 Tel: +81 235 29 0524 E-mail: akio@sfc.keio.ac.jp (Received 14 October 2005, revised 26 November 2005, accepted 1 December 2005) doi:10.1111/j.1742-4658.2005.05082.x Ribonuclease H1 (RNase H1) is a widespread enzyme found in a range of organisms from viruses to humans. It is capable of degrading the RNA moiety of DNA–RNA hybrids and requires a bivalent ion for activity. In contrast with most eukaryotes, which have one gene encoding RNase H1, the activity of which depends on Mg 2+ ions, Caenorhabditis elegans has four RNase H1-related genes, and one of them has an isoform produced by alter- native splicing. However, little is known about the enzymatic features of the proteins encoded by these genes. To determine the differences between these enzymes, we compared the expression patterns of each RNase H1-related gene throughout the development of the nematode and the RNase H activit- ies of their recombinant proteins. We found gene-specific expression patterns and different enzymatic features. In particular, besides the enzyme that displays the highest activity in the presence of Mg 2+ ions, C. elegans has another enzyme that shows preference for Mn 2+ ion as a cofactor. We char- acterized this Mn 2+ -dependent RNase H1 for the first time in eukaryotes. These results suggest that there are at least two types of RNase H1 in C. ele- gans depending on the developmental stage of the organism. Abbreviations dsRHbd, double-stranded RNA and RNA–DNA hybrid-binding domain; Ec-RNHI, Escherichia coli ribonuclease HI; Pf-RNHII, Pyrococcus furiosus ribonuclease HII; PTC, premature termination codon, RNase H, ribonuclease H. 420 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS identified in the N-terminal portion of Crithidia fas- ciculata RNase H1 in relation to the conserved domain in caulimovirus ORF VI involved in translational regu- lation [16]. The non-RNase H domain binds not only to dsRNA but also to RNA–DNA hybrids in vitro [17] and has been defined as a dsRNA and RNA–DNA hybrid-binding domain (dsRHbd) [18]. The function of the dsRHbd in eukaryotic RNase H1 has also been discussed. The dsRNA binding and RNase H activity of Saccharomyces cerevisiae RNase H1 depends on the concentration of Mg 2+ ions and the existence of dsRNA, and the activity of mutant enzymes lacking dsRHbd is not as dependent on these conditions [17]. In contrast, investigation of human RNase H1 by site- directed mutagenesis has suggested that the dsRHbd is required not for RNase H activity but to place RN- ase H in the appropriate position on the RNA primer during DNA replication [19]. However, because the dsRHbd in mouse RNase H1 contributes to RNase H processivity through formation of a dimer complex [18], there is controversy about whether the dsRHbd is in fact necessary for RNase H activity. A single RNase H1-related gene has been identified in the genomes of most eukaryotes studied, and RNase H1 enzymes of Drosophila and mice are essential for embry- ogenesis [20,21]. Unlike in other eukaryotes, four RNase H1-encoding genes have been found in the Caenorhabditis elegans genome, and cDNA sequencing analysis has revealed that one of them has an alternative splicing variant, resulting in a total of five transcripts [22]. Of these, one gene encodes an RNase H1 protein that contains both dsRHbd and RNase H domains, and its alternatively spliced transcript can be translated into an RNase H1 protein that lacks a dsRHbd at the N-ter- minus. We analyzed the expression patterns of the five transcripts, including the pair of alternative splicing variants, throughout the development of C. elegans. Furthermore, we successfully prepared and purified some recombinant C. elegans RNase H1 enzymes in sol- uble form without using denaturants such as 6 m urea or guanidine hydrochloride. This enabled us to compare the enzymatic features of each RNase H by using an in vitro reconstitution system that recapitulated the processing of Okazaki-primer RNA. Results and Discussion Primary structures of multiple RNase H1 enzymes in C. elegans Four RNase H1-related genes have been identified in C. elegans, and one of them has been found to have an alternative splicing isoform [22]. We conducted cDNA cloning of the transcripts that corresponded to each gene and confirmed their primary structures inde- pendently. In accordance with their nomenclature [22], we represented these four genes as rnh-1.0a, rnh-1.1, rnh-1.2, and rnh-1.3 and the alternative splicing iso- form of rnh-1.0a as rnh-1.0b. In contrast with rnh-1.0a, which encodes a 251-amino-acid protein (Ce-RNH1a), rnh-1.0b contains a premature termination codon (PTC) in the alternatively spliced exon. Although cis- acting nonsense-mediated mRNA decay elements in C. elegans have yet to be characterized [23], in several experiments the alternatively spliced transcript that introduces a PTC is degraded by an mRNA surveil- lance system in C. elegans [24,25]. On the other hand, despite the fact that the mRNA for mouse glutamic acid decarboxylase has a PTC inserted by alternative splicing, the N-terminal truncated protein has been shown to be produced from the downstream ORF in vivo and has been shown to exhibit enzyme activity [26], suggesting that the mRNA escapes degradation by nonsense-mediated mRNA decay. Therefore, if rnh- 1.0b also evades the mRNA surveillance system, this mRNA may produce a 41-amino-acid protein encoded by the upstream ORF and a 198-amino-acid protein (Ce-RNH1b) encoded by the downstream ORF. We defined the C. elegans RNase H1 enzymes enco- ded by rnh-1.0a, rnh-1.0b, rnh-1.1, rnh-1.2, and rnh-1.3 as Ce-RNH1a, Ce-RNH1b, Ce-RNH1A, Ce-RNH1B, and Ce-RNH1C, respectively. The domain structures of these proteins are summarized in Fig. 1. Most eukaryotic RNase H1 enzymes contain one or two dsRHbds at the N-terminus and an RNase H domain at the C-terminus. However, only Ce -RNH1a fits the typical structure, and the other RNase H1 enzymes (Ce- RNH1b, Ce-RNH1A, Ce-RNH1B, and Ce-RNH1C) each contained only one RNase H domain. Ce-RNH1α (rnh-1.0α) Ce-RNH1β (rnh-1.0β) Ce-RNH1B (rnh-1.2) Ce-RNH1A (rnh-1.1) Ce-RNH1C rnh-1.3 RNH RNH RNH RNH RNH dsRHbd 1 1 1 1 1 139 251 198 487 192 Fig. 1. Schematic diagrams of RNase H1-related gene family in C. elegans. Each domain is indicated as a shaded box. dsRHbd, dsRNA or RNA–DNA hybrid-binding domain; RNH, RNase H domain. Each gene name is in parentheses below the protein name. Numbers above boxes indicate positions of amino acids. H. Kochiwa et al. Enzymatic characterization of C. elegans RNases H1 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS 421 RNase H gene expression during C. elegans development Although eukaryotic RNase H1 enzymes are thought to be concerned with several regulatory steps, including DNA replication, DNA repair, and RNA transcription, C. elegans, unlike other model organisms, may use the appropriate RNase H1 for a specific situation. To evalu- ate this hypothesis, we first conducted RT-PCR analysis during C. elegans development [larval stages 1–2 (L1– L2), 2–3 (L2–L3), 3–4 (L3–L4), and 4 to adult (L4– adult)] to compare the expression patterns of each RNase H1. RNase H1-related gene expression can be described as one of three patterns (Fig. 2): (a) expressed constantly throughout development (rnh-1.0a and rnh- 1.0b); (b) expressed from the L3 to adult stages (rnh-1.1 and rnh-1.3); (c) preferentially expressed in the egg and adult stages (rnh-1.2). These results suggest that multiple RNase H1-encoding genes may be regulated differently throughout C. elegans development. In particular, because both Ce-RNH1a and Ce- RNH1A have RNase H activity (see below in detail), it should be noted that the expression pattern of rnh-1.0a differed considerably from that of rnh-1.1. This result raises the possibility that these two proteins have dis- tinctly different functions because of their production at different developmental stages. These gene-specific expression patterns also suggest that rnh-1.2 and rnh-1.3 are not pseudogenes, despite the fact that Ce-RNH1B and Ce-RNH1C were not detected to have RNase H activity [22], even though it has been reported that 20% of all annotated C. elegans genes may be pseudogenes [27]. The question of why rnh-1.0a and rnh-1.0b had the same expression pattern required further investigation. The third exon of rnh-1.0a and rnh-1.0b is alternatively spliced (Fig. 3A). In C. elegans, the highly conserved consensus sequence UUUUCAG ⁄ R at the 3¢ splice sites is recognized by subunits of U2AF in the process of intron removal [28]. We checked the sequences around the alternatively spliced sites of rnh-1.0a and rnh-1.0b and found that the sequences were similar to each other (AUUUAG ⁄ G and UUUUAG ⁄ A) (Fig. 3B). Hence, rnh-1.0a and rnh-1.0b have the same expression pattern because these alternatively spliced sites may be chosen evenly when the splicing reaction occurs, suggesting that a specific alternative splicing factor or an exonic enhancer may not regulate this alternative splicing for each transcript. At this point it is not certain whether rnh-1.0b is the result of regula- ted alternative splicing or of aberrant splicing, because the mRNA contains a PTC in the alternatively spliced exon. Cleavage specificity of C. elegans RNase H1 enzymes on RNA–DNA ⁄ DNA hybrid In light of the results of the gene expression analysis, we assumed that the multiple RNase H1 enzymes in C. elegans had distinct enzymatic characteristics. Although Arudchandran et al . [22] detected RNase H activity by a renaturation gel assay, we wanted to make more detailed comparisons by using purified soluble enzymes. Therefore, we overexpressed the recombinant RNase H1 enzymes and purified them to rnh-1.0α rnh-1.0β rnh-1.3 rnh-1.1 rnh-1.2 eft-3 Egg L1-L2 L2-L3 L3-L4 L4-Adult RNA(-) Fig. 2. Expression patterns of RNase H1-related genes during the developmental stages of C. elegans. RT-PCR was carried out using gene-specific primers (see Experimental procedures). Negative con- trol without addition of cDNA template is indicated as RNA (–). eft-3 encoding a translation elongation factor 1-alpha homolog was the positive control. The same results were obtained in at least two inde- pendent experiments for each gene. rnh-1.0α rnh-1.0β α β A UGAAAAUUUAGGAGUCAAACGUUGUUUUAGAUUCAAGAAA αβ B Fig. 3. Alternative splicing of rnh-1.0a and rnh-1.0b. (A) Schematic presentation of alternative splicing. Exons are represented as boxes, and the characters a and b above the third exons indicate alternatively spliced sites of rnh-1.0a and rnh-1.0b. (B) Premature mRNA sequences around alternatively spliced sites. The underlined sequences labeled with a or b correspond to the sequences around the 3¢ exon–intron junction of rnh-1.0a or rnh-1.0a. AGs enclosed in squares represent splice acceptor sites. Arrows indicate 3¢ exon– intron junctions. Enzymatic characterization of C. elegans RNases H1 H. Kochiwa et al. 422 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS near-homogeneity. The molecular masses of the puri- fied recombinant proteins Ce-RNH1a, Ce-RNH1b, Ce-RNH1A, Ce-RNH1B, and Ce-RNH1C were esti- mated to be 33, 28, 62, 25, and 17 kDa, respectively, by SDS ⁄ PAGE (Fig. 4). The enzymatic activity of each recombinant C. ele- gans RNase H1 was analyzed by using two different 30-mer RNA–DNA ⁄ DNA hybrids as substrates so that the cleavage site of the RNA strand could be determined. RNase HI from the bacterium E. coli (Ec-RNHI) and RNase HII from the archaeon Pyro- coccus furiosus (Pf-RNHII) also were used to compare cleavage specificity with those of C. elegans RNase H1 enzymes. When the RNase H assay was performed using 6-carboxyfluorescein (FAM) labeling at the 5¢ end of the RNA–DNA strand, the degradation patterns of Ce-RNH1a and Ce-RNH1b were completely the same, but the other proteins exhibited different patterns (Fig. 5A). On the other hand, when the RNase H assay was performed using fluorescein isothiocyanate labeling at the 3¢ end of the RNA–DNA strand, in contrast with Ec-RNHI, which cleaved the 5¢ phosphodiester bond of the third ribonucleotide from the RNA–DNA junction, C. elegans RNase H1 enzymes and Pf-RNHII cleaved the 5¢ phosphodiester bond of the last ribonucleotide at the RNA–DNA junction (Fig. 5B). The cleavage pat- terns of Ce -RNH1a and Ce-RNH1b were closer to that of Pf-RNHII than to that of Ec-RNHI, whereas 250 150 100 75 50 37 25 15 kDa 234 516 20 Fig. 4. Purified recombinant RNase H1 enzymes of C. elegans. Samples were analyzed by SDS ⁄ PAGE on a 10–20% gel and stained with Coomassie Brilliant Blue. Lane 1, molecular mass markers; lane 2, Ce-RNH1a; lane 3, Ce-RNH1b; lane 4, Ce-RNH1A; lane 5, Ce-RNH1B; lane 6, Ce-RNH1C. Dots indicate positions of each recombinant protein. Ce-RNH1α Ce-RNH1β Ce-RNH1A Ec-RNHIPf-RNHII 30 mer A G C U G G G A U A G C G U A 3' 5' 1 2 3 4 5 6 7 8 9 10111213141516 A 5'-GCGAAUUUAGGGCGAgagcaaacttctcta-3' 5'-GCGAAUUUAGGGCGAgagcaaacttctcta-3' 5'-GCGAAUUUAGGGCGAgagcaaacttctcta-3' 5'-GCGAAUUUAGGGCGAgagcaaacttctcta-3' Ce-RNH1α, Ce-RNH1β Ce-RNH1A Ec-RNHI Pf-RNHII C B 3' 5' Ce-RNH1α Ce-RNH1β Ce-RNH1A Ec-RNHIPf-RNHII 30 mer A G C G 12345678910111213141516 Fig. 5. Cleavage specificity of RNase H enzymes on (A) 5¢ FAM- labeled or (B) 3¢ fluorescein isothiocyanate-labeled RNA–DNA ⁄ DNA hybrid. RNase H digestion products were analyzed using denaturing polyacrylamide gel (see Experimental procedures). Lane 1, no enzyme control; lanes 2–4, 0.05, 0.25, 1 n M Ce-RNH1a; lanes 5–7, 20, 100, 400 n M Ce-RNH1b; lanes 8–10, 10, 50, 200 nM Ce-RNH1A; lanes 11–13, 0.02, 0.1, 0.4 n M Pf-RNHII; lanes 14–16, 0.02, 0.1, 0.4 U Ec-RNHI. Metal ion concentrations are: lanes 1–7 and 14–16, 1m M MgCl 2 ; and lanes 8–13, 1 mM MnCl 2 . Positions of nucleo- tides are indicated on the right. (C) Summary of cleavage sites. Ribonucleotides are represented as uppercase letters and deoxy- ribonucleotides as lower case letters. Arrows indicate cleavage sites. H. Kochiwa et al. Enzymatic characterization of C. elegans RNases H1 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS 423 Ce-RNH1A exhibited a unique digestion pattern (Fig. 5C). Above all, the finding that Ce-RNH1a and Ce-RNH1b showed the same cleavage patterns is in contrast with the fact that the dsRHbd deletion mutant of human RNase H1 exhibits a degradation pattern different from that of the wild-type [19], because Ce-RNH1b also lacks a dsRHbd at the N-terminus. We also checked the activity of a dsRHbd deletion mutant of Ce-RNH1a: the mutant proteins showed exactly the same cleavage pattern as Ce-RNH1a and Ce-RNH1b (data not shown). This result also supports the idea that the dsRHbd of Ce-RNH1a does not affect the specificity of the cleavage site. To compare the enzymatic activities of Ce-RNH1a and Ce-RNH1b, we determined the kinetic parameters of both enzymes and calculated the relative K m and k cat values as in a previous study [29]. The results are summarized in Table 1. The K m value of Ce-RNH1a was comparable to that of Ce-RNH1b, whereas the k cat value of the former enzyme was  91 times higher than that of the latter. Consequently, we can presume that the N-terminus portion of Ce-RNH1a helps to enhance the hydrolysis rate but affects neither the clea- vage site nor the binding affinity for the substrate. The activities of Ce-RNH1B and Ce-RNH1C were also examined by the same RNase H assay, but no activity was detected (data not shown), in agreement with the results of a previous report describing the inactivity of Ce-RNH1B and Ce-RNH1C [22]. The fact that yeast two-hybrid analysis revealed that Ce-RNH1C formed a complex with several other pro- teins [30] suggests that other factors may be necessary for the activation of Ce-RNH1C. The inactivity of Ce-RNH1B may be due to protein misfolding, because we used urea as a denaturant to prepare and purify the recombinant protein. Metal ion preferences of C. elegans RNase H1 enzymes We also compared the RNase H1 enzymes in C. ele- gans in terms of their preferences for bivalent ions. For this purpose, we performed RNase H assays in the presence of Mg 2+ or Mn 2+ ions at concentrations ran- ging from 0.01 to 20 mm. Increased RNase H activity of Ce-RNH1a was associated with an increase in the concentration of Mg 2+ ions but not of Mn 2+ ions (Fig. 6A). On the other hand, although Ce-RNH1b has an optimum concentration of Mg 2+ ions similar to that of the mutant human RNase H1 with deleted dsRHbd at the N-terminus [18], this enzyme was also Table 1. Kinetic parameters of Ce-RNH1a and Ce-RNH1b. The kin- etic parameters were determined from two independent experi- ments. Relative K m and k cat values were calculated by dividing the values for Ce-RNH1b by those for Ce-RNH1a. Enzyme Relative K m value Relative k cat value Ce-RNH1a 1.0 1.0 Ce-RNH1b 0.86 0.011 MgCl2 MnCl2 30 mer 0.01 0.1 0.5 1 5 10(-) 0.01 0.1 0.5 1 5 10 (mM)2020 C Ce-RNH1A 30 mer B MgCl2 MnCl2 0.01 0.1 0.5 1 5 10(-) 0.01 0.1 0.5 1 5 10 (mM)2020 Ce-RNH1β 30 mer A MgCl2 MnCl2 0.01 0.1 0.5 1 5 10(-) 0.01 0.1 0.5 1 5 10 (mM)2020 Ce-RNH1α Fig. 6. Metal ion preferences of C. elegans RNase H1 enzymes. Reactions were carried out in the presence of 0.01–20 m M MgCl 2 or MnCl 2 using 5¢ FAM-labeled RNA–DNA ⁄ DNA hybrid. No enzyme control is indicated as (–). Concentrations of each recombinant pro- tein were (A) 0.5 n M Ce-RNH1a,(B)100nM Ce-RNH1b, and (C) 150 n M Ce-RNH1A. Enzymatic characterization of C. elegans RNases H1 H. Kochiwa et al. 424 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS activated in the presence of Mn 2+ ions (Fig. 6B). To investigate the dependence of enzymatic activity on metal ions more precisely, we also determined the spe- cific activities of Ce-RNH1a and Ce-RNH1b in the presence of Mg 2+ or Mn 2+ ions, as shown in Table 2. Although the specific activity of Ce-RNH1a in the presence of Mg 2+ was higher than in the presence of Mn 2+ , Ce-RNH1a could use Mn 2+ as a cofactor for cleavage of Okazaki fragment substrates. This result is inconsistent with those of previous reports in the calf thymus [8] and humans [10]; however, Bacillus halodu- rans RNase HI, which, like eukaryotic RNase H1, contains a dsRHbd at the N-terminus, is also activated in the presence of Mn 2+ (Wei Yang, personal commu- nication). Therefore, this feature of Ce-RNH1a is thought to be more similar to that of bacterial RNase HI than to that of mammals. On the other hand, the specific activity of Ce-RNH1b is only about one-hundredth that of Ce-RNH1a in the presence of Mg 2+ and about one-fortieth in the presence of Mn 2+ . The difference in specific activities between Ce-RNH1a and Ce-RNH1b suggests that the existence of the dsRHbd may be related to RNase H activity and supports the idea that eukaryotic RNase H1 enzymes act processively by interactions through the dsRHbd, leading to dimerization of the protein [18]. However, Ce-RNH1b function in vivo is still unclear, because this protein has low RNase H activity. It is likely that the rnh-1.0b encoding Ce-RNH1b may not produce a functional protein; instead, aberrant splicing or alternative splicing may contribute to the regulation of gene expression in combination with the nonsense- mediated mRNA decay system [31]. In contrast with the activities of Ce-RNH1a and Ce-RNH1b, that of Ce-RNH1A was enhanced in the presence of Mn 2+ ions rather than Mg 2+ ions (Fig. 6C). We also found that the pattern of digestion by Ce-RNH1A differed with the metal ion. A previous report defined eukaryotic RNase H1 as an enzyme that requires Mg 2+ ions for activity but cannot use Mn 2+ ions as cofactors [8]. However, our study revealed that this description does not apply to Ce-RNH1A, which can be classified as a eukaryotic RNase H1 on the basis of amino-acid sequence similarity. Phylogram analysis shows that Ce-RNH1a is orthologous to human RNase H1 and that Ce-RNH1A is out-grouped with S. cerevisiae and S. pombe RNase H1 enzymes [22], but that S. cerevisiae RNase H1 prefers Mg 2+ ions as a cofactor, as do other eukaryotic RNase H enzymes [32]. To our knowledge, Ce-RNH1A is the only eukaryotic RNase H1 that prefers Mn 2+ ions for activity. These results suggest that at least two types of RNase H1, Ce-RNH1a and Ce-RNH1A, occur in C. elegans. Comparative analysis of RNase H1-encoding genes in C. elegans and C. briggsae C. elegans has multiple RNase H1-related genes, unlike other eukaryotes, and Ce-RNH1A seems to be the only exception found so far among eukaryotic RN- ase H1 enzymes from the perspective of ionic prefer- ence, as described in the previous section. Are these features limited to C. elegans? To clarify this, we con- ducted a comparative analysis of the RNase H1-enco- ding genes in C. elegans and C. briggsae, which diverged from a common ancestor 100 million years ago, because the complete C. briggsae genome was published recently [33] and its protein database was useful for this analysis. Comparative analysis showed that Ce-RNH1a, Ce-RNH1A, and Ce-RNH1C had independent orthologous proteins in C. briggsae (Table 3), leading us to conclude that these RNase H1-related genes were generated before the two species diverged. On the other hand, the fact that Ce-RNH1B is more similar to Ce-RNH1C than to any other eukaryotic RNase H1 enzymes (data not shown) suggests that these two genes may have been generated as a result of gene duplication within the C. elegans genome. In summary, C. elegans RNase H1 enzymes Table 2. Specific activities of Ce-RNH1a and Ce-RNH1b in the pres- ence of Mg 2+ or Mn 2+ ions. One unit of enzymatic activity was defined as the amount of enzyme hydrolyzing 1 lmol substrate per minute, and the specific activity was defined as the enzymatic activity per mg protein. The activity derived from Ce-RNH1a in the presence of MgCl 2 was set as 100. The specific activity values rep- resent means from two separate experiments. Enzyme Metal ion Specific activity (UÆmg )1 ) Relative activity (%) Ce-RNH1a MgCl 2 7.0 100 MnCl 2 2.1 30 Ce-RNH1b MgCl 2 0.070 1.2 MnCl 2 0.057 0.8 Table 3. Protein conservation among three species. Numerical val- ues indicate sequence identities (%) between proteins. Ortholo- gous proteins are underlined. C. elegans enzymes are indicated along the top, and C. briggsae enzymes down the side. Ce-RNH1a Ce-RNH1A Ce-RNH1B Ce-RNH1C CBP16719 82.9 28.4 30.2 38.8 CBP03109 28.4 76.1 29.0 28.0 CBP19944 39.2 26.2 32.1 56.1 CBP07225 37.1 30.8 22.9 32.6 Human RNase H1 45.8 33.1 26.4 34.9 H. Kochiwa et al. Enzymatic characterization of C. elegans RNases H1 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS 425 can be classified into three groups: (a) Ce-RNH1a may have functions common to other eukaryotic RNase H1 enzymes; (b) Ce-RNH1A and Ce-RNH1C may provide a lineage-specific function for C. elegans and C. brigg- sae; (c) Ce-RNH1B may be specific to C. elegans. Comparative analysis has shown the possibility that the N-terminal sequences of C. elegans RNase H1 enzymes serve as localization signals. Alteration of the N-terminal portion contributes to the subcellular local- ization of RNase H1 in the mouse [21] and Cr. fascicu- lata [34], and protein diversity in these cases may be caused by translation from different start codons. In C. elegans, phylogenetic profiling of eukaryotic proteins has also determined that there are 660 nucleus-encoded mitochondrial genes, and C. elegans RNase H1 enzymes are also predicted to be mitoch- ondrial [35]. In particular, we found that rnh-1.1 of C. elegans has two potential start codons at the 5¢ ends of the ORF, and the 17-amino-acid sequence (MIR- WFRNFGALFKKPRG) from the first methionine was conserved in the gene orthologous to rnh-1.1 in C. briggsae, with high similarity (88%). The amino- acid sequence of C. briggsae is MIRWFRNL- GTLFKKPRG (amino acid residues identical with those of Ce-RNH1A are underlined). Because the mit- ochondrial localization signal of mouse RNase H1 is 27 amino-acid residues from the first methionine and is conserved in several vertebrates [21], the N-terminal portion of Ce-RNH1A may also serve as some sort of localization signal. From this result, we propose that the multiple RNase H1 enzymes in C. elegans are regu- lated not only at a transcriptional level but also at a post-transcriptional level. Taking into consideration the conservation of RNase H1 enzymes between C. elegans and C. briggsae and the enzymatic features described in the previous sections, we suggest that C. elegans may have obtained various types of RNase H1 in a phased manner and that the roles of C. elegans RNase H1 enzymes may have diverged in accordance with their evolution. Experimental procedures Strain maintenance The N2 nematode strain and E. coli strain OP50 used in this work were provided by the Caenorhabditis Genetics Center, which is funded by the NIH National Center for Research Resources (NCRR). To synchronize worm devel- opmental staging, eggs were collected from adult worms and incubated on nematode growth medium agar plates overnight at 25 °C. L1 larval stage worms were collected from plates with M9 buffer and plated on to nematode growth medium agar plates inoculated with E. coli strain OP50. L1 worms were incubated at 25 °C, and worms at each developmental stage were harvested after the appropri- ate incubation times [36]. RT-PCR analysis Total RNAs from eggs and from larval stages 1–2 (L1–L2), 2–3 (L2–L3), 3–4 (L3–L4), and 4 to adult (L4 –adult) were prepared with TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). To detect the expression of each gene during C. ele- gans development, RT-PCR was performed with ReverTra Dash (Toyobo, Osaka, Japan) using gene-specific primers (Table S1) to amplify the entire coding sequences of rnh-1.1 (GenBank accession no. ZK1290.6), rnh-1.2 (GenBank accession no. ZK938.7), and rnh-1.3 (GenBank accession no. C04F12.9), the partial coding sequences of rnh-1.0a (GenBank accession no. F59A6.9), the alternative spliced transcript (defined as rnh-1.0b)ofrnh-1.0a, and eft-3 (Gen- Bank accession no. F31E3.5). PCRs were performed under the following conditions: 3 min at 94 °C; 10 s at 98 °C, 2 s at 60 ° C, and 20 s (rnh-1.0a, rnh-1.0b, rnh-1.2, rnh-1.3, and eft-3) or 1 min (rnh-1.1)at74°C for 30 cycles; and 5 min at 74 °C. These conditions were set up to make a linear dose–response relationship between each RNA and its PCR product. PCR products were separated by 1.2% agarose gel electrophoresis and stained with ethidium bromide. cDNA cloning To obtain cDNA clones of rnh-1.1, rnh-1.2, and rnh-1.3, the amplified PCR products described in the previous sec- tion were used. For cloning of rnh-1.0a and rnh-1.0b alter- natively spliced from the same transcripts, the coding sequences were amplified by RT-PCR using H0AB-S (5¢-CCAGTTACTCAAGATTTTGAACGC-3¢) as a for- ward primer and H0AB-A (5¢-CGTTTAATGAACAT TTGGGCTCC-3¢) as a reverse primer. PCR products were purified with GFX PCR DNA and a Gel Band Purification Kit (Amersham Biosciences, Piscataway, NJ, USA) and cloned into pPCR-Script Amp SK(+) vectors (Stratagene, La Jolla, CA, USA). Plasmid DNAs were transformed into E. coli strain DH5a competent cells (Toyobo) and purified with a QIAprep Spin Miniprep Kit (Qiagen GmbH, Hilden, Germany). The nucleotide sequences of each insert DNA were determined and confirmed to be identical with those in the database. Expression and purification of recombinant proteins The ORFs of each gene were PCR-amplified from plasmids containing each cDNA by using ReverTra Dash (Toyobo) and gene-specific primers containing NdeI and XhoIor Enzymatic characterization of C. elegans RNases H1 H. Kochiwa et al. 426 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS NotI sites (Table S2). The amplified PCR fragments were treated with appropriate restriction enzymes and subcloned into pET-23b expression vector (Novagen, Darmstadt, Ger- many) and sequenced to confirm their correct nucleotide sequences. The resulting plasmids were transformed into E. coli strain BL21(DE3)pLysS (Novagen). The transform- ants were incubated at 37 °C for  6 h in Luria–Bertani medium containing 100 lgÆmL )1 ampicillin and 40 lgÆmL )1 chloramphenicol and then subjected to induction at 20 °C overnight with 0.4 mm isopropyl b-d-thiogalactopyrano- side. The recombinant proteins were extracted by sonica- tion in buffer A [20 mm sodium phosphate (pH 7.4), 10 mm imidazole, and 500 mm NaCl] and centrifuged at 12 000 g for 10 min. Except for Ce -RNH1B, the superna- tant of each recombinant protein was loaded on to a nickel–Sepharose column (Amersham Biosciences) and elut- ed with buffer B [20 mm sodium phosphate (pH 7.4), 500 mm imidazole, and 500 mm NaCl] by AKTA FPLC (Amersham Biosciences). The partly purified recombinants were loaded on to a HiTrap Desalting Column (Amersham Biosciences) and desalted with buffer C, containing 50 mm Tris ⁄ HCl (pH 7.5), 0.02 mm EDTA (pH 8.0), 0.05% 2- mercaptoethanol, 0.02% Tween 20, and 10% glycerol. For extraction of Ce-RNH1B, after sonication and centrifuga- tion, the pellet was washed with buffer D [0.5% Triton X- 100, 1 mm EDTA (pH 8.0)] several times and resolved in buffer A containing 6 m urea, loaded on to a nickel–Seph- arose column, and eluted with buffer B containing 6 m urea. The eluted sample was dialyzed against buffer C con- taining 6 m urea, loaded on to a HiTrap Desalting Col- umn, and desalted with buffer C. Purified recombinant proteins were analyzed by SDS ⁄ PAGE on a 10–20% gradi- ent gel and stained with Quick-CBB (Wako, Osaka, Japan). In vitro assay for RNase H activity RNase H activity was assayed by analyzing the stability of an RNA–DNA hybrid in the presence of enzymatic sam- ples. A 5¢ FAM-labeled or 3¢ fluorescein isothiocyanate- labeled 30-nucleotide-long RNA–DNA oligonucleotide (5¢-GCGAAUUUAGGGCGAgagcaaacttctcta-3¢) and its cDNA oligonucleotide (5¢-tagagaagtttgctctcgccctaaattcgc-3¢) (ribonucleotides denoted by uppercase letters and deoxy- ribonucleotides by lowercase letters) were chemically syn- thesized by Hokkaido System Science (Hokkaido, Japan) and annealed for use as a substrate. RNase H reactions were performed in 20 lL reaction buffer containing 20 mm Tris ⁄ HCl (pH 8.0), 1 mm dithiothreitol, 50 mm KCl, 0.01– 20 mm MgCl 2 or MnCl 2 ,1mgÆmL )1 BSA, 100 nm sub- strate, and 0.05–400 nm purified recombinant C. elegans RNase H1 or purified recombinant Pf-RNHII (provided by A. Sato) [37], or Ec-RNHI (purchased from Toyobo, Osaka, Japan). Samples were incubated for 15 min at room temperature (C. elegans RNase H1 enzymes) or at 37 °C (Ec-RNHI) or 50 °C(Pf-RNHII). We then added an equal volume of stop solution [8 m urea ⁄ 1 m Tris ⁄ HCl, pH 7.2, and a small amount of blue dextran (Sigma Chemical, St Louis, MO, USA)] to stop the reactions. The reaction mix- tures were heated at 70 °C for 3 min, loaded on to a 20% polyacrylamide gel containing 8 m urea, and run for 20 min at 2000 V and 50 min at 2200 V. The reaction products were visualized with a Molecular Imager FX Pro (Bio-Rad Laboratories, Hercules, CA, USA). Kinetic analysis To determine the kinetic parameters, the enzymatic activity was observed in the presence of 1 mm MgCl 2 using the RNA–DNA ⁄ DNA hybrid as substrate. The concentrations of the substrate varied from 0.1 to 1.0 lm and the amount of enzyme was controlled such that the cleavage rate of the substrate did not exceed 30% of the total, as previously described [38]. Hydrolysis of the substrate with the enzyme followed Michaelis–Menten kinetics, and the kinetic param- eters were obtained from the Lineweaver–Burk plot. k cat was calculated from k cat ¼ V max ⁄ [E]. To determine the spe- cific activity, one unit of enzymatic activity was defined as the amount of enzyme hydrolyzing 1 lmol substrate per minute, and the specific activity was defined as the enzy- matic activity per mg protein. Enzymatic activity was observed in the presence of 1 mm MgCl 2 or 5 mm MnCl 2 using the RNA–DNA ⁄ DNA hybrid as substrate, and the substrate concentration was 0.1 lm. Each value given is the mean from two separate experiments. Comparative analysis of C. elegans and C. briggsae BLAST analysis [39] was conducted against a C. briggsae protein database provided by WormBase [40], by using the amino-acid sequence of human RNase H1 (GenBank acces- sion no. NP_002927) as a query sequence. Four proteins of C. briggsae (WormBase protein IDs CBP16719, CBP03109, CBP19944, CBP07225) were detected as similar to human RNase H1. The RNase H domain of each C. elegans and C. briggsae RNase H1 was identified by using hmmpfam [41], and we extracted the amino-acid sequences corres- ponding to the RNase H domain. fasta analysis [42] was performed to analyze protein similarity by comparing the amino-acid sequences of each RNase H domain. Acknowledgements We thank Asako Sato (Keio University, Japan) for technical assistance with the RNase H assay and Dr Naoto Ohtani, Azusa Kuroki, Koji Numata (Keio University, Japan), Dr Robert J. Crouch (National Institutes of Health, Bethesda, MD, USA), Dr Yuji H. Kochiwa et al. Enzymatic characterization of C. elegans RNases H1 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS 427 Kohara, and Dr Hideaki Hiraki (National Institute of Genetics, Japan) for their helpful discussions. We also appreciate the help of Dr Wei Yang, Dr Marcin Now- otny, and Dr Sergei A. Gaidamakov (National Insti- tutes of Health, USA) for providing unpublished data and suggestions. This research was supported in part by: a Grant-in-Aid for Scientific Research on Priority Areas; a Grant-in-Aid from the 21st Century Center of Excellence (COE) Program, entitled ‘Understanding and Control of Life’s Function via Systems Biology (Keio University)’; and grants from the Japan Society for the Promotion of Science (JSPS) and Keio Univer- sity. References 1 Stein H & Hausen P (1969) Enzyme from calf thymus degrading the RNA moiety of DNA-RNA Hybrids: effect on DNA-dependent RNA polymerase. Science 166, 393–395. 2 Hausen P & Stein H (1970) Ribonuclease H. An enzyme degrading the RNA moiety of DNA-RNA hybrids. Eur J Biochem 14, 278–283. 3 Hansen J, Schulze T, Mellert W & Moelling K (1988) Identification and characterization of HIV-specific RNase H by monoclonal antibody. EMBO J 7, 239–243. 4 Itaya M (1990) Isolation and characterization of a sec- ond RNase H (RNase HII) of Escherichia coli K-12 encoded by the rnhB gene. Proc Natl Acad Sci USA 87, 8587–8591. 5 Henry CM, Ferdinand FJ & Knippers R (1973) A hydridase from Escherichia coli. Biochem Biophys Res Commun 50, 603–611. 6 Ohtani N, Yanagawa H, Tomita M & Itaya M (2004) Identification of the first archaeal Type 1 RNase H gene from Halobacterium sp. NRC-1: archaeal RNase HI can cleave an RNA–DNA junction. Biochem J 381, 795–802. 7 Ohtani N, Haruki M, Morikawa M & Kanaya S (1999) Molecular diversities of RNases H. J Biosci Bioeng 88, 12–19. 8 Vonwirth H, Frank P & Busen W (1989) Serological analysis and characterization of calf thymus ribonu- clease H IIb. Eur J Biochem 184, 321–329. 9 Wu H, Lima WF & Crooke ST (1998) Molecular clon- ing and expression of cDNA for human RNase H. Antisense Nucleic Acid Drug Dev 8, 53–61. 10 Frank P, Albert S, Cazenave C & Toulme JJ (1994) Purification and characterization of human ribonuclease HII. Nucleic Acids Res 22, 5247–5254. 11 Berkower I, Leis J & Hurwitz J (1973) Isolation and characterization of an endonuclease from Escherichia coli specific for ribonucleic acid in ribonucleic acid– deoxyribonucleic acid hybrid structures. J Biol Chem 248, 5914–5921. 12 Katayanagi K, Okumura M & Morikawa K (1993) Crystal structure of Escherichia coli RNase HI in com- plex with Mg 2+ at 2.8 A ˚ resolution: proof for a single Mg(2+)-binding site. Proteins 17, 337–346. 13 Goedken ER & Marqusee S (2001) Co-crystal of Escherichia coli RNase HI with Mn 2+ ions reveals two divalent metals bound in the active site. J Biol Chem 276, 7266–7271. 14 Ohtani N, Haruki M, Muroya A, Morikawa M & Kanaya S (2000) Characterization of ribonuclease HII from Escherichia coli overproduced in a soluble form. J Biochem (Tokyo) 127, 895–899. 15 Ohtani N, Haruki M, Morikawa M, Crouch RJ, Itaya M & Kanaya S (1999) Identification of the genes encoding Mn 2+ -dependent RNase HII and Mg 2+ -dependent RNase HIII from Bacillus subtilis: classification of RNases H into three families. Bio- chemistry 38, 605–618. 16 Mushegian AR, Edskes HK & Koonin EV (1994) Eukaryotic RNAse H shares a conserved domain with caulimovirus proteins that facilitate translation of poly- cistronic RNA. Nucleic Acids Res 22 , 4163–4166. 17 Cerritelli SM & Crouch RJ (1995) The non-RNase H domain of Saccharomyces cerevisiae RNase H1 binds double-stranded RNA: magnesium modulates the switch between double-stranded RNA binding and RNase H activity. RNA 1, 246–259. 18 Gaidamakov SA, Gorshkova II, Schuck P, Steinbach PJ, Yamada H, Crouch RJ & Cerritelli SM (2005) Eukaryotic RNases H1 act processively by interactions through the duplex RNA-binding domain. Nucleic Acids Res 33, 2166–2175. 19 Wu H, Lima WF & Crooke ST (2001) Investigating the structure of human RNase H1 by site-directed mutagen- esis. J Biol Chem 276, 23547–23553. 20 Filippov V, Filippov M & Gill SS (2001) Drosophila RNase H1 is essential for development but not for pro- liferation. Mol Genet Genomics 265, 771–777. 21 Cerritelli SM, Frolova EG, Feng C, Grinberg A, Love PE & Crouch RJ (2003) Failure to produce mitochon- drial DNA results in embryonic lethality in Rnaseh1 null mice. Mol Cell 11, 807–815. 22 Arudchandran A, Cerritelli SM, Bowen NJ, Chen X, Krause MW & Crouch RJ (2002) Multiple ribonuclease H-encoding genes in the Caenorhabditis elegans genome contrasts with the two typical ribonuclease H-encoding genes in the human genome. Mol Biol Evol 19, 1910– 1919. 23 Lejeune F & Maquat LE (2005) Mechanistic links between nonsense-mediated mRNA decay and pre- mRNA splicing in mammalian cells. Curr Opin Cell Biol 17, 309–315. 24 Morrison M, Harris KS & Roth MB (1997) smg mutants affect the expression of alternatively spliced SR Enzymatic characterization of C. elegans RNases H1 H. Kochiwa et al. 428 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS protein mRNAs in Caenorhabditis elegans. Proc Natl Acad Sci USA 94, 9782–9785. 25 Mitrovich QM & Anderson P (2000) Unproductively spliced ribosomal protein mRNAs are natural targets of mRNA surveillance in C. elegans. Genes Dev 14, 2173– 2184. 26 Szabo G, Katarova Z & Greenspan R (1994) Distinct protein forms are produced from alternatively spliced bicistronic glutamic acid decarboxylase mRNAs during development. Mol Cell Biol 14, 7535–7545. 27 Mounsey A, Bauer P & Hope IA (2002) Evidence suggesting that a fifth of annotated Caenorhabditis elegans genes may be pseudogenes. Genome Res 12, 770–775. 28 Zorio DA & Blumenthal T (1999) Both subunits of U2AF recognize the 3¢ splice site in Caenorhabditis ele- gans. Nature 402, 835–858. 29 Chai Q, Qiu J, Chapados BR & Shen B (2001) Archaeo- globus fulgidus RNase HII in DNA replication: enzymo- logical functions and activity regulation via metal cofactors. Biochem Biophys Res Commun 286, 1073–1081. 30 Li S, Armstrong CM, Bertin N, Ge H, Milstein S, Boxem M, Vidalain PO, Han JD, Chesneau A, Hao T, et al. (2004) A map of the interactome network of the meta- zoan C. elegans. Science 303, 540–543. 31 Baek D & Green P (2005) Sequence conservation, rela- tive isoform frequencies, and nonsense-mediated decay in evolutionarily conserved alternative splicing. Proc Natl Acad Sci USA 102, 12813–12818. 32 Wintersberger U & Frank P (2001) Ribonucleases H of the budding yeast, Saccharomyces cerevisiae. Methods Enzymol 341, 414–429. 33 Stein LD, Bao Z, Blasiar D, Blumenthal T, Brent MR, Chen N, Chinwalla A, Clarke L, Clee C, Coghlan A, et al. (2003) The genome sequence of Caenorhabditis briggsae: a platform for comparative genomics. Plos Biol 1, 166–192. 34 Engel ML, Hines JC & Ray DS (2001) The Crithidia fasciculata RNH1 gene encodes both nuclear and mito- chondrial isoforms of RNase H. Nucleic Acids Res 29, 725–731. 35 Marcotte EM, Xenarios I, van Der Bliek AM & Eisen- berg D (2000) Localizing proteins in the cell from their phylogenetic profiles. Proc Natl Acad Sci USA 97, 12115–12120. 36 Lewis JA & Fleming JT (1995) Basic culture methods: Caenorhabditis elegans. Modern Biological Analysis of an Organism, pp. 4–9. Academic Press, San Diego, CA. 37 Sato A, Kanai A, Itaya M & Tomita M (2003) Coop- erative regulation for Okazaki fragment processing by RNase HII and FEN-1 purified from a hyperthermophi- lic archaeon, Pyrococcus furiosus. Biochem Biophys Res Commun 309, 247–252. 38 Haruki M, Tsunaka Y, Morikawa M & Kanaya S (2002) Cleavage of a DNA-RNA-DNA ⁄ DNA chimeric substrate containing a single ribonucleotide at the DNA–RNA junction with prokaryotic RNases HII. FEBS Lett 531, 204–208. 39 Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215, 403–410. 40 Chen N, Harris TW, Antoshechkin I, Bastiani C, Bieri T, Blasiar D, Bradnam K, Canaran P, Chan J, Chen CK, et al. (2005) WormBase: a comprehensive data resource for Caenorhabditis biology and genomics. Nucleic Acids Res 33, 383–389. 41 Eddy SR (1998) Profile hidden Markov models. Bio- informatics 14, 755–763. 42 Pearson WR (1990) Rapid and sensitive sequence com- parison with FASTP and FASTA. Methods Enzymol 183, 63–98. Supplementary material The following supplementary material is available online: Table S1. Oligonucleotides used for RT-PCR analysis. Table S2. Oligonucleotides used for expression clo- ning. Restriction sites are underlined. This material is available as part of the online article from http://www.blackwell-synergy.com H. Kochiwa et al. Enzymatic characterization of C. elegans RNases H1 FEBS Journal 273 (2006) 420–429 ª 2005 The Authors Journal compilation ª 2005 FEBS 429 . Stage-specific expression of Caenorhabditis elegans ribonuclease H1 enzymes with different substrate specificities and bivalent cation requirements Hiromi Kochiwa 1,2 ,. duplication within the C. elegans genome. In summary, C. elegans RNase H1 enzymes Table 2. Specific activities of Ce-RNH1a and Ce-RNH1b in the pres- ence of Mg 2+ or Mn 2+ ions. One unit of enzymatic. the N-terminus and an RNase H domain at the C-terminus. However, only Ce -RNH1a fits the typical structure, and the other RNase H1 enzymes (Ce- RNH1b, Ce-RNH1A, Ce-RNH1B, and Ce-RNH1C) each contained