Investigating the Structure of Human RNase H1 by Site-directed Mutagenesis*

In this study we examine for the first time the roles of the various domains of human RNase H1 by site-directed mutagenesis. The carboxyl terminus of human RNase H1 is highly conserved with Escherichia coli RNase H1 and contains the amino acid residues of the putative catalytic site and basic substrate-binding domain of the E. coli RNase enzyme. The amino terminus of human RNase H1 contains a structure consistent with a double-strand RNA (dsRNA) binding motif that is separated from the conserved E. coli RNase H1 region by a 62-amino acid sequence. These studies showed that although the conserved amino acid residues of the putative catalytic site and basic substrate-binding domain are required for RNase H activity, deletion of either the catalytic site or the basic substrate-binding domain did not ablate binding to the heteroduplex substrate. Deletion of the region between the dsRNA-binding domain and the conserved E. coli RNase H1 domain resulted in a significant loss in the RNase H activity. Furthermore, the binding affinity of this deletion mutant for the heteroduplex substrate was ∼2-fold tighter than the wild-type enzyme suggesting that this central 62-amino acid region does not contribute to the binding affinity of the enzyme for the substrate. The dsRNA-binding domain was not required for RNase H activity, as the dsRNA-deletion mutants exhibited catalytic rates ∼2-fold faster than the rate observed for wild-type enzyme. Comparison of the dissociation constant of human RNase H1 and the dsRNA-deletion mutant for the heteroduplex substrate indicates that the deletion of this region resulted in a 5-fold loss in binding affinity. Finally, comparison of the cleavage patterns exhibited by the mutant proteins with the cleavage pattern for the wild-type enzyme indicates that the dsRNA-binding domain is responsible for the observed strong positional preference for cleavage exhibited by human RNase H1.

RNase H hydrolyzes RNA in RNA-DNA hybrids (1). RNase H activity appears to be ubiquitous in eukaryotes and bacteria (2)(3)(4)(5)(6)(7). Although RNase Hs constitute a family of proteins of varying molecular weight, the nucleolytic activity and substrate requirements appear to be similar for the various isotypes. For example, all RNase Hs studied to date function as endonucleases exhibiting limited sequence specificity and requiring divalent cations (e.g. Mg 2ϩ and Mn 2ϩ ) to produce cleavage products with 5Ј-phosphate and 3Ј-hydroxyl termini (8).
Two classes of RNase H enzymes have been identified in mammalian cells (5,9,10). These enzymes were shown to differ with respect to cofactor requirements and were shown to be inhibited by sulfhydryl reagents (10,11). Although the biological roles of the mammalian enzymes are not fully understood, it has been suggested that mammalian RNase H1 may be involved in replication and that the RNase H2 enzyme may be involved in transcription (12,13).
Recently, both human RNase H genes have been cloned and expressed (11,14,15). RNase H1 is a 286-amino acid protein with a calculated mass of 32 kDa (11). The enzyme is encoded by a single gene that is at least 10 kilobase pairs in length and is expressed ubiquitously in human cells and tissues. The amino acid sequence of human RNase H1 displays strong homology with RNase H1 from yeast, chicken, Escherichia coli, and mouse (11). The human RNase H2 enzyme is a 299-amino acid protein with a calculated mass of 33.4 kDa and has also been shown to be ubiquitously expressed in human cells and tissues (14). 1 Human RNase H2 shares strong amino acid sequence homology with RNase H2 from Caenorhabditis elegans, yeast, and E. coli (14).
The properties of the cloned and expressed human RNase H1 have been characterized recently (16). The activity of RNase H1 is Mg 2ϩ -dependent and inhibited by both Mn 2ϩ and the sulfhydryl-blocking agent N-ethylmaleimide. Human RNase H1 was also inhibited by increasing ionic strength with optimal activity for both KCl and NaCl observed at 10 -20 mM. The enzyme exhibited a bell-shaped response to divalent cations and pH, with the optimum conditions for catalysis observed to be 1 mM Mg 2ϩ and pH 7-8. The protein was shown to be reversibly denatured under the influence of temperature and destabilizing agents such as urea. Renaturation of human RNase H1 was observed to be highly cooperative and did not require divalent cations. Furthermore, RNase H1 displayed no tendency to form intermolecular disulfides or to form homomultimers. Human RNase H1 was shown to bind selectively to "A-form" duplexes with 10 -20-fold greater affinity than that observed for E. coli RNase H1 (16,17). Finally, human RNase H1 displays a strong positional preference for cleavage, i.e. the enzyme cleaves between 8 and 12 nucleotides from the 5Ј-RNA-3Ј-DNA terminus of the duplex.
Many of the properties observed for Human RNase H1 are consistent with the E. coli RNase H1 isotype (e.g. the cofactor requirements, substrate specificity, and binding specificity) (16,17). In fact, the carboxyl-terminal portion of human RNase H1 is highly conserved with the amino acid sequence of the E. coli enzyme (Fig. 1, Region III). The glutamic acid and two aspartic acid residues of the catalytic site as well as the histidine and aspartic acid residues of the proposed second divalent cation-binding site of the E. coli enzyme are conserved in hu-man RNase H1 (18 -21). In addition, the lysine residues within the highly basic ␣-helical substrate-binding region of E. coli RNase H1 are also conserved in the human enzyme.
Despite these similarities, the structures of the two enzymes differ in several important ways. For example, the amino acid sequence of the human enzyme is ϳ2-fold longer than the E. coli enzyme. The additional amino acid sequence of the human enzyme extends from the amino terminus of the conserved E. coli RNase H1 region and contains a 73-amino acid region homologous with a double-strand RNA (dsRNA) 2 -binding motif (Fig. 1, region I). The conserved E. coli RNase H1 region at the carboxyl terminus is separated from the dsRNA-binding domain of the human enzyme by a 62-amino acid region (Fig. 1, region II). Although the role of both regions I and II remain unclear, the dsRNA-binding domain of human RNase H1 may account for the observed positional preference for cleavage displayed by the enzyme as well as the enhanced binding affinity of the enzyme for various polynucleotides (16). Finally, the human enzyme is a significantly more basic protein than E. coli RNase H1 with a net positive charge of ϩ15 for human RNase H1 compared with ϩ2 for the E. coli enzyme.
In this study we have explored the roles of the conserved amino acids of the catalytic site and the basic substrate-binding domain (region III), the roles of the dsRNA-binding domain (region I), and the 62-amino acid center region of human RNase H1 (region II) (Fig. 1). We have performed site-directed mutagenesis on the three conserved amino acids of the proposed catalytic site of human RNase H1 ((D145N), (E186Q), and (D210N)). In addition, the net positive charge of the basic substrate-binding domain was progressively reduced through alanine substitution of two (RNase H1(K226A,K227A)) and four (RNase H1(K226A,K227A,K231A,K236A)) of the lysines within this region. Deletion mutants were also prepared in which either the dsRNA-binding domain of region I (RNase H1(⌬I)) or the central region II (RNase H1(⌬II)) was deleted. Finally, a mutant protein representing the conserved E. coli RNase H1 region was prepared by deleting both regions I and II (RNase H1(⌬I-II)).

MATERIALS AND METHODS
Construction of Mutant Proteins-The mutagenesis of human RNase H1 was performed using a PCR-based technique derived from Landt et al. (22). Briefly, two separate PCRs were performed using a set of site-directed mutagenic primers and two vector-specific primers (11). For the RNase H1(D145N) mutant the 5Ј-oligodeoxynucleotide used for PCR was TACACTAATGGCTGCTGCTCCAGTAAT and the 3Ј-oligodeoxynucleotide was GCAGCCATTAGTGTAGACGACGACGACGAA. The PCR primers for RNase H1(E186Q) were 5Ј-AGAGCGCAAATTCATG-CAGCCTGCAAA and 3Ј-ATGAATTTGCGCTCTTTGGTTTGTCTG. The primers for RNase H1(D210N) were 5Ј-TATACAAACAGTATGTT-TACGATAAAT and 3Ј-CATACTGTTTGTATACAGAACCAGTTT. The primers for RNase H1(K226A,K227A) were 5Ј-GGTTGGGCAGCAAAT-GGGTGGAAGACAAGT and 3Ј-CCCATTTGCTGCCCAACCTTGAACC-CAGTT. The primers for RNase H1(K226A,K227A,K231A,K236A) were 5Ј-GCAGCAAATGGGTGGGCGACAAGTGCAGGCGCAGAGGTGATC-AACAAAG and 3Ј-TGCCCCTGCACTTGTCGCCCACCCATTTGCTGC-CCAACCTTGAACCCAG. The PCR primers for RNase H1(⌬I) were 5Ј-ATCTTAGGATCCTCTGCAAGCCCGGAAGTTTCA and 3Ј-ATCTTAC-TCGAGTCAGTCTTCCGATTGTTTAGCTCC. The primers for RNase H1(⌬II) were 5Ј-TTTGTCAGGAAAATGGGAGACTTCGTCGTC and 3Ј-GAAGTCTCCCATTTTCCTGACAAAGGCCCA. The PCR primers for RNase H1(⌬I-II) were 5Ј-ATCTTAGGATCCATGGGAGACTTCGTCG-TCGTCTA and the same 3Ј-primer as the RNase H1(⌬I) mutant. Approximately 1 g of human RNase H1 cDNA was used as the template for the first round of amplification of both the amino-and carboxyl-terminal portions of the cDNA corresponding to the mutant site. The fragments were purified by agarose gel extraction (Qiagen, Germany). PCR was performed in two rounds consisting of, respectively, 15 and 25 amplification cycles (94°C, 30 s; 55°C, 30 s; 72°C, 180 s). The purified fragments were used as the template for the second round of PCR using the two vector-specific primers. The final PCR product was purified and cloned into the expression vector pET17b (Novagen) as described previously (11). The incorporation of the desired mutations was confirmed by DNA sequencing.
Protein Expression and Purification-The plasmid was transfected into E. coli BL21(DE3) (Novagen). The bacteria were grown in M9ZB medium (23) at 32°C and harvested at A 600 of 0.8. The cells were induced with 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside at 32°C for 2 h. The cells were lysed in 8 M urea solution, and the recombinant protein was partially purified with nickel-nitrilotriacetic acid-agarose (Qiagen, Germany).
Synthesis of Oligonucleotides-The oligoribonucleotides were synthesized on a PE-ABI 380B synthesizer using 5Ј-O-silyl-2Ј-O-bis(2-acetoxyethoxy)methylribonucleoside phosphoramidites and procedures described elsewhere (25). The oligoribonucleotides were purified by reverse-phase HPLC. The oligodeoxyribonucleotides were synthesized on a PE-ABI 380B automated DNA synthesizer and standard phosphoramidite chemistry. The oligodeoxyribonucleotides were purified by precipitation 2 times out of 0.5 M NaCl with 2.5 volumes of ethyl alcohol. The concentration of the oligonucleotides was determined by UV adsorption at A 260 and previously published extinction coefficients (26).
Preparation of the Heteroduplex-The heteroduplex substrate was prepared in 100 l containing unlabeled oligoribonucleotide ranging from 1 to 1000 nM, 10 5 cpm of 32 P-labeled oligoribonucleotide, 2-fold excess complementary oligodeoxyribonucleotide and hybridization buffer (20 mM Tris, pH 7.5, 20 mM KCl). Reactions were heated at 90°C for 5 min, cooled to 37°C, and 60 units of Prime RNase Inhibitor (5 Prime 3 3 Prime, Inc., Boulder, CO) and MgCl 2 at a final concentration of 1 mM were added. Hybridization reactions were incubated 2-16 h at 37°C, and ␤-mercaptoethanol was added at a final concentration of 20 mM.
Multiple Turnover Kinetics-The heteroduplex substrate was digested with 1.5 ng of human RNase H1 at 37°C. A 10-l aliquot of the cleavage reaction was removed at time points ranging from 2 to 120 min and quenched by adding 5 l of stop solution (8 M urea and 120 mM EDTA). The aliquots were heated at 90°C for 2 min and resolved in a 12% denaturing polyacrylamide gel, and the substrate and product bands were quantitated on a Molecular Dynamics PhosphorImager. The concentration of the converted product was plotted as a function of time. The initial cleavage rate was obtained from the slope (mol of RNA cleaved/min) of the best fit line for the linear portion of the plot, which consists, in general, Ͻ10% of the total reaction and data from at least five time points. The initial cleavage rates were plotted as a function of the substrate concentration (n Ն 4), and the data were fit to the Michaelis-Menten equation using the program Ultrafit (Biosoft, NJ). The K m value corresponds to the heteroduplex substrate concentration at half-maximum rate, and the k cat ϭ V max /(total RNase H1), where V max corresponds to the horizontal asymptote of the hyperbolic curve.
Competition experiments were performed as described for the determination of multiple turnover kinetics with the exception that 20 nM oligodeoxyribonucleotide, 10 nM oligoribonucleotide, and 60 ng of the mutant RNase H1 protein was used. Reactions were digested with 6 ng of wild-type human RNase H. The reactions were quenched, analyzed, and quantitated as described for multiple turnover kinetics.
Determination of Dissociation Constants (K d )-Binding affinities were determined by inhibition analysis (17). Here the cleavage rate is determined for the heteroduplex substrate at a variety of concentrations in both the presence and absence of a competing noncleavable substrate analog. The heteroduplex substrate was prepared as described above except in a final volume of 50 l and with equimolar oligoribonucleotide and oligodeoxyribonucleotide ranging in concentration from 10 to 500 nM. The competing noncleavable substrate analog was prepared in 50 l of hybridization buffer containing equimolar oligodeoxyribonucleotide and complementary 2Ј-fluoro-modified oligonucleotide. The concentration of the noncleavable substrate analog was in excess of the heteroduplex substrate and ranged from 1 to 5 M.
Reactions were heated at 90°C for 5 min and cooled to 37°C. Prime RNase inhibitor and MgCl 2 were added to the reactions as described above. Reactions were incubated at 37°C for 2-16 h, and the noncleavable substrate analog was added to the heteroduplex substrate. ␤-Mercaptoethanol was added at a final concentration of 20 nM, and the combined reaction digested with 1.5 ng of human RNase H1. The reactions were quenched, analyzed, and quantitated as described for multiple turnover kinetics.

RESULTS
The mutant proteins of human RNase H1 prepared for this study are described in Fig. 1. Analysis of human RNase H1 and the mutant proteins by SDS-polyacrylamide gel electrophoresis is shown in Fig. 2. As expected, mutant proteins containing amino acid substitutions (e.g. D145N, E186Q, D210N, K226A,K227A, and K226A,K227A,K231A,K236A) exhibited molecular weights similar to the 32-kDa wild-type enzyme (Fig.  2, lanes 1-6). The RNase H1(⌬I) mutant in which the dsRNAbinding domain was deleted resulted in a 213-amino acid protein with an approximate molecular mass of 23 kDa (lane 7). The deletion of the 62-amino acid center portion of human RNase H1 (RNase H1(⌬II)) resulted in a 224-amino acid protein with an approximate molecular mass of 25 kDa (lane 8). Finally, the deletion of both the dsRNA-binding domain and the central region of the enzyme (RNase H1(⌬I-II)) resulted in a 151-amino acid protein containing the conserved E. coli RNase H1 sequence and with an approximate molecular mass of 17 kDa (lane 9).
The enzymatic activities of the human RNase H1 enzyme and the mutant proteins were determined using a 17nucleotide-long oligoribonucleotide/oligodeoxyribonucleotide heteroduplex ( Table I). Substitution of any one the three amino acids comprising the proposed catalytic site of human RNase H1, (e.g. D145N, E186Q, and D210N) ablated the cleavage activity of the enzyme. In addition, alanine substitution of two (RNase H1(K226A,K227A)) or four (RNase H1(K226A,K227A,K231A,K236A)) lysine residues within the basic substrate-binding domain also ablated cleavage activity.
The kinetic constants for the deletion mutants are shown in Table I resulted in a comparable increase in both k cat and K m values when compared with the wild-type enzyme. Conversely, deletion of region II of human RNase H1 resulted in a reduction in both k cat and K m . In this case, the K m and k cat observed for the wild-type enzyme was ϳ2-3-fold greater than the kinetic constants observed for the RNase H1(⌬II) mutant. Finally, the k cat for the mutant protein in which both regions I and II were deleted (RNase H1(⌬I-II)) was ϳ2-fold faster than the k cat observed for the wild-type enzyme, whereas the K m for both enzymes was comparable.
The positions of the cleavage sites for the wild-type and mutants of human RNase H1 in the heteroduplex substrate are shown in Fig. 3. As observed previously, human RNase H1 exhibited a strong positional preference, i.e. 8 -12 nucleotides from the 5Ј-RNA/3Ј-DNA terminus of the duplex (Fig. 3A). A similar cleavage pattern was observed for the RNase H1(⌬II) deletion mutant. The RNase H1(⌬I) and H1(⌬I-II) deletion mutants exhibited broader cleavage patterns on the heteroduplex substrate, with cleavage sites ranging from 7 to 13 nucleotides from the 5Ј terminus of the RNA (Fig. 3B).
Experiments were performed to determine whether the inactive mutants of human RNase H1 competitively inhibit the cleavage activity of the wild-type enzyme. These experiments were performed under single turnover kinetics with the enzyme concentration in excess of the substrate concentration and with the concentration of the mutant protein in excess of the wild-type enzyme concentration. To ensure that inhibition of human RNase H1 activity by the mutant proteins was competitive and not due to nonspecific protein-protein interactions, competition experiments were also performed under multiple turnover kinetics with the substrate concentration in excess of the wild-type and mutant RNase H1 concentrations. Under these conditions, no reduction in human RNase H1 activity was observed in the presence of 10-fold excess mutant RNase H1 protein (data not shown). In contrast, all three of the mutant proteins tested under single turnover kinetics were observed to inhibit competitively the cleavage activity of human RNase H1 (Fig. 4). For example, the initial cleavage rate of human RNase H1 alone was determined to be 6-fold faster than the initial cleavage rate for human RNase H1 in the presence of the RNase H1(D145N) mutant. The initial cleavage rate of human RNase H1 in the presence of the region II deletion mutant (RNase H1(⌬II)) was ϳ50% slower than the rate observed for human RNase H1 alone. Finally, the initial cleavage rate for human RNase H1 in the presence of the RNase H1(K226A,K227A,K231A,K236A) mutant was ϳ60% slower than the rate observed for human RNase H1 alone.
The binding affinities of human RNase H1 and the RNase H1(⌬I-II) mutant were determined indirectly using a competition assay as described previously (17). Briefly, the cleavage rate of the oligodeoxyribonucleotide/oligoribonucleotide heteroduplex was determined at a variety of substrate concentrations in both the presence and absence of competing noncleavable oligodeoxyribonucleotide/2Ј-fluoro-modified oligonucleotide heteroduplex. Lineweaver-Burk and Augustinsson analysis of the data were used to determine the inhibitory constant (K i ) for the competing noncleavable heteroduplex. The K i is equivalent to the dissociation constant (K d ) of the enzyme when a noncleavable heteroduplex is used.
The dissociation constant (K d ) of human RNase H1 for the oligodeoxyribonucleotide/2Ј-fluoro-modified oligonucleotide a ND ϭ cleavage rates below the detection limit of the assay (e.g. Ͻ1% of the heteroduplex substrate cleaved over 60 min).
b The Michaelis-Menten kinetics for E. coli RNase H1 was determined as described under "Materials and Methods" with the exception that 500 pg of enzyme was used.  (Table II). The RNase H1(⌬I), H1(⌬II), and H1(⌬I-II) mutants exhibited dissociation constants (K d ) for the noncleavable heteroduplex of, respectively, 410, 34, and 126 nM.

Structure of Human RNase H1-
The human RNase H1 protein can be divided into three regions (Fig. 1). Region I, located at the amino terminus of the enzyme, contains a structure consistent with a dsRNA-binding motif. Region II consists of a 62-amino acid region between the dsRNA-binding domain and the conserved E. coli RNase H1 region. Finally, region III is situated at the carboxyl terminus of human RNase H1 and contains an amino acid sequence that is highly conserved with the amino acid sequence of E. coli RNase H1. Included within region III are the conserved amino acid residues that form the putative catalytic site, the second divalent cation-binding site, and the basic substrate-binding domain of the E. coli enzyme.
Catalytic Triad-The three amino acids (Asp-10, Glu-48, and Asp-70) that make up the catalytic site of E. coli RNase H1 were identified by site-directed mutagenesis (20). These amino acid residues have also been shown to be involved with the coordination of the requisite divalent cation cofactor (28). Comparison of the amino acid sequence of E. coli RNase H1 with the amino acid sequences of the RNase H domain of retroviruses and RNase H1 from yeast, chicken, human, and mouse indicates that these three amino acid residues are conserved among all type 1 sequences (11).
Mutant proteins of human RNase H1 were prepared in which each of the three conserved catalytic residues Asp-145, Glu-186, and Asp-210 was substituted with, respectively, Asn, Gln, and Asn. The complete ablation of cleavage activity observed for the RNase H1(D145N), (E186Q), and (D210N) mutants indicates that all three of the conserved residues in human RNase H1 are required for catalytic activity (Table I). The RNase H1(D145N) mutant competitively inhibited the activity of human RNase H1 suggesting that the loss in cleavage activity observed for this mutant protein was not due to a loss in the binding affinity for the heteroduplex substrate (Fig. 4). Taken together these data suggest that, consistent with the E. coli enzyme, the three conserved residues likely form the catalytic site of the enzyme and are not involved in the substrate-binding interaction.
Basic Substrate-binding Domain-The amino acid sequence of the basic substrate-binding region of E. coli RNase H1 is highly conserved in the human enzyme (11). The basic substrate-binding domain of E. coli RNase H1 has been extensively characterized and has been shown to comprise the ␣-helix III and following loop region of the enzyme (18, 20, 21). The crystal structure of E. coli RNase H1 indicates that this region forms a relatively independent sub-domain with the loop region composed of a cluster of basic amino acid residues. These basic amino acid residues are believed to bind electrostatically to the phosphate backbone of the heteroduplex substrate.
Mutant proteins of human RNase H1 were prepared to determine whether these conserved basic amino acids served a similar function in the human enzyme. The RNase H1 mutants in which two (RNase H1(K226A,K227A)) and all four lysine residues (RNase H1 (K226A,K227A,K231A,K236A)) were substituted with alanine residues resulted in the complete loss of RNase H activity (Table I). Furthermore, the RNase H1(K226A,K227A,K231A,K236A) mutant was shown to inhibit competitively the cleavage activity of the wild-type human enzyme (Fig. 4), suggesting that the observed loss of RNase H activity for the mutant protein was not due to a loss in the overall binding affinity of the mutant protein for the substrate. However, it is possible that these lysine residues are important for the proper positioning of the catalytic domain.
The properties observed for the basic amino acid residues of human RNase H1 differ from those observed for the E. coli enzyme. First, unlike the human enzyme the alanine substitution of any two of the basic residues within the substratebinding domain of E. coli RNase H1 did not affect the cleavage activity of the mutant E. coli proteins, i.e. the cleavage rates for the alanine-substituted mutants were comparable to the rates observed for the wild-type E. coli enzyme (18). Second, alanine substitution of four basic amino acids within the substratebinding domain of E. coli RNase H1 resulted in a 5-10-fold reduction but not the ablation of the RNase H activity of the E. coli mutants. In addition, a 30 -60-fold increase in the K m was observed for these mutant proteins suggesting that a reduction in the binding affinity for the substrate was responsible for the observed reduction in cleavage activity. In the case of human RNase H1 mutant, the complete loss in cleavage activity for the RNase H1(K226A,K227A,K231A,K236A) mutant did not coincide with an appreciable loss in binding affinity for the substrate. Whether differences in assay conditions or methods used to prepare the proteins can account for the observed disparity between the human and E. coli enzymes is not clear. The fact that the alanine substitution of the conserved lysine residues resulted in the ablation of the RNase H activity indicates that these residues are essential for the catalytic processes of the human enzyme. Finally, our data also suggest that other regions within human RNase H1 are contributing to the binding affinity of the enzyme. One likely region that may be involved in substrate binding is the dsRNA-binding domain of the human enzyme.
Region I-Human RNase H1 has been observed to contain the canonical ␣-␤-␤-␤-␣ structure consistent with the dsRNAbinding motif (11). The position of the double-strand RNAbinding domain at the amino terminus of the enzyme is consistent with the structure observed for RNase H1 from Saccharomyces cerevisiae (29). The human RNase H1 dsRNAbinding domain differs from the yeast enzyme in that the human sequence appeared to correspond to a more complete dsRNA-binding motif. The properties of the dsRNA-binding domain of human RNase H1 were also observed to differ from the yeast enzyme in that the dsRNA-binding domain of the human enzyme was not modulated by divalent cation concentration. For example, RNase H1 from S. cerevisiae was observed to bind to dsRNA at Mg 2ϩ concentrations below those required to activate the enzyme and was inhibited from binding to dsRNA at Mg 2ϩ concentrations that activated the enzyme. In other words, binding to dsRNA and the RNase H activity of the S. cerevisiae enzyme were observed to be mutually exclusive. a The net charge of the RNase H1 proteins was calculated by subtracting the number of acidic amino acids from the basic residues.
b The K d value for E. coli RNase was determined as described under "Materials and Methods" with the exception that 500 pg of enzyme was used.
Contrary to the yeast enzyme, the binding of human RNase H1 to dsRNA was determined not to be affected by Mg 2ϩ concentrations required to activate the enzyme.
Deletion mutants of human RNase H1 were prepared in order to investigate the role of the dsRNA-binding domain, i.e. region I. These mutants included the deletion of the dsRNAbinding domain (RNase H1(⌬I)) and deletion of both the dsRNA-binding domain and region II between the dsRNAbinding domain and the conserved E. coli RNase H1 region (RNase H1(⌬I-II)). Both mutants in which the dsRNA-binding domain was deleted cleaved the heteroduplex substrate at rates faster than the rate observed for the wild-type enzyme (Table I). A mutant of RNase H1 from S. cerevisiae in which the dsRNA-binding domain was deleted also exhibited RNase H activity (29). In light of the fact that the RNase H activity and dsRNA-binding properties of the yeast enzyme are mutually exclusive, it is not surprising that the yeast enzyme would remain active with the dsRNA-binding domain deleted. On the other hand, the robust cleavage activity of the human deletion mutants is surprising, particularly when considering that the RNase H1(K226A,K227A,K231A,K236A) mutant was able to bind to the heteroduplex substrate. The cleavage activity of the RNase H1(⌬I) and (⌬I-II) mutants suggests that the enzyme does not require the dsRNA-binding domain in order to bind to the heteroduplex substrate, although the dsRNA-binding region appears to contribute to the overall binding affinity of the human RNase H1 as the K d for the wild-type enzyme was ϳ5-fold tighter than the K d of the RNase H1(⌬I) mutant (Table  II). Clearly, both the basic substrate-binding domain and the dsRNA-binding domain contribute to the overall binding affinity of human RNase H1, because the elimination of either domain did not affect the ability of the enzyme to bind to the substrate.
It is important to note that these human RNase H1 proteins are His tag fusion proteins, and it is possible that the binding properties of the proteins may be enhanced by the His tag. Numerous studies have shown that a His tag does not interfere with nucleic acid-binding properties as it is very small (few amino acids), and its pK is near neutral (30,31). Furthermore, examination of the contribution of the His tag to the overall net charge of the various human RNase H1 proteins suggest that the His tag is likely not contributing to the binding affinity of the proteins (Table II). For example, the contribution of the His tag to the net charge of the human RNase H1 proteins was greatest with the RNase H1(⌬I) mutant, yet this mutant exhibited the weakest binding affinity for the heteroduplex. In contrast, the RNase H1(⌬II) mutant exhibited the tightest binding affinity even though the contribution of the His tag to the net charge of this protein was the smallest.
The cleavage pattern for the mutants in which the dsRNAbinding region was deleted (RNase H1(⌬I) and (⌬I-II)) differed from the pattern observed for the wild-type human enzyme. In fact the cleavage pattern for the RNase H1(⌬I) and -(⌬I-II) mutants resembled the cleavage pattern of the E. coli RNase H1 enzyme that does not contain a dsRNA-binding domain. Taken together these data suggest that the dsRNA-binding domain is responsible for the observed strong positional preference for cleavage exhibited by human RNase H1 (16) and further suggest that this region contributes to the overall binding affinity of the enzyme for the substrate and the regulation of the sites of cleavage. Finally, the broad cleavage pattern observed for the RNase H1(⌬I-II) mutant further suggests that the strong positional preference for cleavage displayed by human RNase H1 is not responsible for slower catalytic rate of the human enzyme compared with E. coli RNase H1. The catalytic rate observed for E. coli RNase H1 was ϳ30-fold faster than the rate observed for the human enzyme (Table I). The strong positional preference for cleavage displayed by human RNase H1 in effect limits the number of productive binding interactions for a given substrate. Considering that RNase H1(⌬I-II) mutant displayed a similar cleavage pattern to the E. coli enzyme, the slower cleavage rate observed for human RNase H1 is likely not due to the strong positional preference for cleavage. Again, it is not clear whether differences in the purification and renaturation methods used to prepare the E. coli and human enzyme can account for the observed rate differences. The fact that the renaturation of human RNase H1 was observed to be highly cooperative and reversible suggests that the slower cleavage rate of the human enzyme compared with E. coli RNase H1 is likely not due to the misfolding of the human enzyme (16). Furthermore, numerous studies indicate that in the majority cases the native structure of proteins (e.g. amino acid code) naturally generate the folding pathway of proteins (for review see Ref. 32).
The role of the dsRNA-binding domain is not clear. Obviously the dsRNA-binding domain of human RNase H1 is not required for RNase H activity, and consequently, this region likely serves another function. RNase H1 enzymes have been proposed to participate in DNA replication and are believed to aid in the removal of the RNA primers during the DNA replication of the lagging strand. The strong positional preference for cleavage exhibited by the human RNase H1 proteins containing the dsRNA-binding domain is consistent with the average length of the RNA primers that have been shown to range from 7 to 14 nucleotides (33). Therefore, the role of the dsRNAbinding domain of human RNase H1 may be to place the enzyme in the appropriate position on the RNA primer in order to ensure efficient removal of the primer.
Region II-Region II comprises the amino acid sequence between the dsRNA-binding domain (region I) and the conserved E. coli RNase H1 domain (region III). Deletion of this region (RNase H1(⌬II)) resulted in a significant reduction in both the k cat and K m values when compared with the wild-type enzyme (Table I). In addition, the RNase H1(⌬II) mutant also exhibited a 2-fold tighter K d value than the wild-type enzyme suggesting that the loss in RNase H activity did not appear to be due to a reduction in the binding affinity of the RNase H1(⌬II) mutant for the heteroduplex substrate (Table II). Consequently, one possibility for the reduction in k cat for the RNase H1(⌬II) mutant may be due to a reduction in turnover rates as a result of the tighter binding affinity of this mutant for the heteroduplex. Alternatively, deletion of region II places the dsRNA-binding domain immediately adjacent to the conserved E. coli RNase H1 region of the human enzyme that may result in steric hindrance of the catalytic site by the dsRNA-binding domain.
The loss of cleavage activity observed for the RNase H1(⌬II) mutant is consistent with that observed for RNase H1 of Trypanosoma brucei in which the deletion of this region also resulted in the loss of enzymatic activity. 3 Furthermore, this region in the T. brucei enzyme has been shown to contain numerous acidic residues, and site-directed mutagenesis of the acidic amino acids within this region also resulted in the ablation of the enzymatic activity. The highly acidic nature of this region is consistent in the human enzyme and is found in RNase H1 proteins of Crithidia fasciculata, Drosophila melanogaster, and S. cerevisiae. It is unclear how these acidic residues contribute to the enzymatic activity of human RNase H1, but this region appears to play a critical role in the structure of the enzyme. Clearly, understanding the role of this region with respect to the enzymatic activity of human RNase H1 warrants further investigation.
Region III-Region III, as represented by the H1(⌬I-II) mutant, contains the conserved E. coli RNase H1 domain. The k cat observed for the H1(⌬I-II) mutant was ϳ2-fold faster than the catalytic rate observed for wild-type human enzyme but approximately 1 order of magnitude slower than the catalytic rate observed for E. coli RNase H1 (Table I). Again, differences in the methods used to prepare the two proteins may account for the observed rate differences. But the robust activity of the RNase H1(⌬I-II) mutant indicates that region III is capable of folding into an active structure independent of regions I and II and further suggests that region III constitutes an autonomous sub-domain of the human enzyme. Folding of the E. coli RNase H1 enzyme has been shown to follow a two-step process involving a core folding intermediate (34). The amino acid sequence of the core folding intermediate is highly conserved in the human enzyme, which shares 41% amino acid identity with the E. coli enzyme, and suggests a similar folding pathway for human RNase H1(⌬I-II) mutant. Whether the wild-type human enzyme follows a similar folding pathway remains to be determined.
The binding affinity of the RNase H1(⌬I-II) mutant for the heteroduplex substrate was determined to be ϳ28-fold tighter than the binding affinity observed for the E. coli enzyme (Table  II). Furthermore, the binding affinity of the wild-type human enzyme was Ͻ2-fold tighter than that observed for the RNase H1(⌬I-II) mutant. Taken together, these data suggest that region III is providing a significant contribution to the increased binding affinity observed for human RNase H1. The tighter binding affinity observed for both the wild-type and human RNase H1(⌬I-II) mutant likely limits the turnover rate of the human enzyme and may account for the slower k cat of the human proteins compared with E. coli RNase H1 (Table I).