Interaction of Ribonuclease H from Drosophila melanogaster Embryos with DNA Polymerase-Primase *

An RNase H was purified 2,500-fold to near homogeneity from early embryos of Drosophila melanogaster. The purified enzyme has an approximate molecular weight of 180,000 and appears to consist of two 49,000and two 39,000-dalton polypeptides. The enzyme specifically hydrolyzes RNA*DNA hybrids and releases oligoribonucleotides ranging in size from 2-9 residues. The RNase H can also remove RNA primers that are synthesized and subsequently elongated by the Drosophila polymerase-primase. Preincubation of the RNase H from D. melanogaster embryos with the homologous DNA polymerase-primase results in an increased rate of DNA synthesis. The DNA chains synthesized under these conditions are shorter than those synthesized in the absence of the RNase H, and the rate of primer synthesis is increased significantly. These findings suggest that the RNase H forms a complex with the polymerase-primase, increasing its recycling capacity and thereby increasing the frequency of chain initiation.

tion of the Drosophila polymerase-primase showed one of them to be highly enriched in RNase H activity. We describe here the purification of the RNase H to near homogeneity, and consider its structure and catalytic properties. We also describe a novel feature of the Drosophila RNase H: its ability to stimulate DNA synthesis catalyzed by the Drosophila polymerase-primase.

Materials
Nucleotides and Homopolymers-Unlabeled deoxy-and ribonucleoside triphosphates and poly-and oligonucleotides were purchased from P-L Biochemicals. I3H]dTTP was purchased from New England Nuclear. Nucleic Acids-Poly(dT)1500 or poly(U)l~oo was annealed to [3H] poly (A) for 25 min a t 65 "C at a 2:l molar ratio of thymine or uracil to adenine in a buffer consisting of 10 mM Tris.HC1 (pH 8.0), 0.3 M NaCI, and 0.03 M sodium citrate. 3H-labeled double-stranded DNA was prepared by incubating activated calf thymus DNA (9) with Escherichia coli DNA polymerase I in the presence of [3H]dTTP. 3Hlabeled ssDNA' refers to a solution of 3H-labeled double-stranded DNA made 0.1 N in NaOH, then neutralized. M13mp8 and M13Goril ssDNAs were gifts from D. Soltis of this department.
Enzymes-The Drosophila DNA polymerase-primase was prepared as described previously (2). E. coli DNA polymerase 111 holoenzyme, RNase H, and single-stranded DNA binding protein were the gifts of M. O'Donnell, L. Bertsch, and D. Soltis of this department, respectively. E. coli DNA ligase and DNA polymerase I were purchased from United States Biochemical Corp. and P-L Biochemicals, respectively. Bovine serum albumin was purchased from Pentex. Thryoglobulin, bovine y-globulin, chicken ovalbumin, and bovine myoglobin were obtained from Bio-Rad. Rabbit muscle myosin, E. coli p-galactosidase, rabbit muscle phosphorylase B, and bovine carbonic anhydrase were purchased from Sigma. Buffers-All potassium phosphate buffers were at pH 7.6 and contained 1 mM 2-mercaptoethanol, 0.2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 mM sodium bisulfite, 2 pg/ml leupeptin, and 10% glycerol. The ionic strength of buffers was checked with a Radiometer conductivity meter.
Assay for Primer Synthesis-Primer synthesis was measured in a coupled assay with E. coli DNA polymerase 111 holoenzyme.' Reaction conditions were the same as those described above except that incubations were at 22 "C. At various times, aliquots (20 p l ) were withdrawn and added to a solution (5 pl) containing 1.3 pg of E. coli single-stranded DNA binding protein and 30 units of DNA polymerase 1 1 1 holoenzyme. After 15 min at 30 "C, [3ZP]dTTP incorporated into acid-insoluble material was measured. Under these conditions, any ssDNA primed by the Drosophila polymerase-primase should be converted to a full-length circle by the DNA polymerase I11 holoenzyme. Protein Determination-Protein was determined by the method of Bradford (11) with bovine serum albumin as standard.
Polyacrylamide Gel Electrophoresis-Polyacrylamide gel electrophoresis in the presence of NaDodSO, was performed according to Laemmli (12).
Analysis of RNase H Digestion Products-Reaction mixtures were prepared as described above except that the specific activity of the [3H]poly(A) was 200 cpm/pmol. Following incubation at 37 "C, aliquots were applied to Whatman No. 3MM paper. Descending paper chromatography was carried out in 1 M ammonium acetate, 95% ethanol (70:30) at room temperature for approximately 16 h. Under these conditions, oligoribonucleotides containing 10 or more residues remained at the origin. Oligoribonucleotides of known chain length were added to each sample and visualized after chromatography by inspection under UV light. The spots corresponding to each oligoribonucleotide were cut out and counted in a toluene-based scintillant.
Purification of D. melanogaster RNase H All operations were performed at 0-4 "C. A summary of the purification is given in Table I. Fractions I and I1 were prepared from freshly harvested embryos as described by Kaguni et al. (2).
Ammonium Sulfate Precipitation-Solid (N&)S04 (0.31 g/ml) was added slowly with stirring to 2,700 ml of Fraction 11. The suspension was stirred for an additional 45 min and the precipitate collected by centrifugation at 32,000 X g for 45 min. (NHJ'SO, (0.159 g/ml) was added to the supernatant fluid and the resulting precipitate collected as before. The pellets were frozen in liquid nitrogen and stored at -80 "C before resuspension in 20 mM potassium phosphate buffer (Fraction 111).
DNA Cellulose Chromatography and Ammonium Sulfate Precipitation-Fraction I11 (140 ml) was dialyzed 16 h against 4 liters of 20 mM potassium phosphate (2 changes) until an ionic equivalent of 20 mM potassium phosphate containing 60 mM NaCl was reached. The S. Cotterill and I. R. Lehman, unpublished observation. dialyzed fraction was centrifuged at 24,000 X g for 20 min, and the supernatant fluid was loaded onto a column (6.6 cm' X 6 cm) of single-stranded DNA cellulose equilibrated with 20 mM potassium phosphate containing 60 mM NaCl, at the rate of 40 ml/h. The column was washed (129 ml/h) with 160 ml of buffer containing 120 mM NaCl, followed by 80 ml of buffer containing 300 mM NaCl. Active fractions, eluted with the latter buffer, were pooled (Fraction IV) and solid (NH4)'S04 was added (0.472 g/ml). The precipitate was collected as before and stored at -80 "C before resuspension in T buffer (50 mM Tris.HC1 (pH 8.5), 10% glycerol, 0.2 mM EDTA, and 1 mM 2-mercaptoethanol) (Fraction IVb). DEAE-Sephudex Chromatography and Ammonium Sulfate Precipitation-Fraction IVb (3 ml) was dialyzed against 1 liter of T buffer in collodion bags (AIr cutoff, 25,000) until an ionic equivalent of T buffer containing 50 ml of NaCl was reached. The dialyzed fraction was loaded onto a column (1.3 cm2 X 3.4 cm) of DEAE-Sephadex equilibrated with T buffer containing 50 mM NaCl at the rate of 4 ml/h. The column was washed with 9 ml of buffer and a 45-ml linear gradient from 50 to 400 mM NaCl was applied at the rate of 10 ml/h. RNase H was eluted at 170 mM NaCl and active fractions (11.8 ml) were pooled (Fraction V). Ammonium sulfate precipitation was performed as before and the precipitate was stored at -80 "C before resuspension in 0.16 ml of 50 mM potassium phosphate, 200 mM (N&)'SO& 1 mM 2-mercaptoethanol, 1 mM EDTA, and 10% glycerol (Fraction Vb).
High Performance Liquid Chromatography-Fraction Vb (0.16 ml) was centrifuged for 5 min in a Brinkman microfuge at full speed, and approximately 0.14 ml of the supernatant was applied to a Bio-Si1 TSK-250 column (7.5 mM X 300 mM) equilibrated with. 50 mM potassium phosphate, 200 mM (NH4)2S04, 1 mM 2-mercaptoethanol, 1 mM EDTA, 10% glycerol. The column was run at a flow rate of 0.5 ml/min and 0.25-ml fractions were collected. The peak fractions (1.0 ml) were pooled (Fraction VI) and stored at 4 "C.

Physical Properties of D. melanogaster RNase H
NaDodS04-polyacrylamide gel electrophoresis of Fraction VI yielded two major polypeptides, with M, values of 49,000 and 39,000 ( Fig. 1). Densitometric scanning of the gel stained with Coomassie Blue showed the relative abundance of the M, 49,000 and 39,000 polypeptides to be 1.0/1.5 ( Fig. 1, right). The two polypeptides together accounted for approximately 75% of the protein applied to the gel.
When the peak fractions obtained by the high performance liquid chromatography were subjected to NaDodS04-polyacrylamide gel electrophoresis and stained with Coomassie Blue, the M, 49,000 and 39,000 polypeptides corresponded closely to the RNase H activity (data not shown).
The molecular weight of the native RNase H was determined by Bio-Si1 TSK-250 gel filtration in the presence of 0.2 M (NH,),SO,. RNase H activity eluted as a major peak with a molecular weight of 186,000 (Fig. 2). Based on its behavior during NaDodS04-polyacrylamide gel electrophoresis and Bio-Si1 TSK-250 gel filtration, the Drosophila RNase H appears to be a heterotetramer composed of two 49,000dalton and two 39,000-dalton subunits.

Characterization of RNase H Activity
Reaction Requirements-The pH optimum in 50 mM Tris. HCl was 7.5; the reaction rates at pH 7.2 and 8.5 were 38 and 45%, respectively, of the rate at pH 7.5. A divaIent cation was absolutely required; the optima for MgCl, and MnC1, were 10 mM and 0.4 mM, respectively; the former stimulating 3-fold more efficiently than the latter. RNase H activity was optimal at 10 mM (NH4),S04; however, varying the concentration of (NH4),S04 from 0-40 mM resulted in less than a 30% change in the rate of reaction.
Substrate Specificity-The substrate specificity of the Drosophila RNase H is summarized in Table 11. The enzyme was specific for an RNA. DNA hybrid. There was a linear depend-  FIG. 2. Co-elution of RNase H and stimulatory activity. The purified RNase H (36 pg, Fraction VI) was applied to a column of Bio-Si1 TSK-250 (7.5 X 300 mM) that was equilibrated with 50 mM potassium phosphate (pH 7.6), 200 mM (NH&S04, 1 mM 2-mercaptoethanol, 1 mM EDTA, and 10% g!ycerol. The flow rate was 0.5 ml/ min, and 0.5-min fractions were collected and assayed for both RNase H and stimulatory activities. Marker proteins run under identical conditions and indicacec! by their molecular weights (X were thyroglobulin, bsvine y-globulin, ovalbumin, and myoglobin. ence on enzyme .concentration with [3H]poly (A) .poly(dT) as substrate, and essentially no (less than 0.1%) activity with either [3H]poly (A) or [3H]poly(A) -poly(U). Similarly, the enzyme was inactive with either doubleor single-stranded DNA.
The presumed substrate for the Drosophila RNase H in vivo 1s an RNA primer that had initiated Okazaki fragment synthesis. Circular ssDNA, partial.ly replicated by the Grosophila polymerase-primase in the presence of [~u~~P I A T P , was treated with increasing levels of RNase H, and the products were analyzed by neutral agarose gel electrophoresis (Fig. 3A).
In the absence of RNase H (1st lane), there was a distribution of partially replicated ssDNA circles. Upon addition of increasing levels of RNase H, the number of labeled molecules decreased (2nd through 4th lanes). In contrast, RNase H treatment of ssDNA that was partially replicated by the polymerase-primase in the presence of [32P]dTTP, had no effect on the size distribution of the newly synthesized DNA chains (Fig. 3B) (14). The replicated DNA was treated with increasing levels of RNase H and analyzed by neutral agarose gel electrophoresis. The gel was stained with ethidium bromide (0.5 pg/ml) to visualize DNA markers, dried under vacuum, and exposed to Kodak XAR-5 x-ray film. B, DNA synthesis was carried out as described above except that [cx-~'P]~TTP was the labeled nucleotide. The replicated DNA was treated with increasing levels of RNase H and analyzed by alkaline agarose gel electrophoresis. Molecular weight markers, whose sizes are indicated in nucleotide residues, were produced by HaeIII digestion of 4x174 supercoiled duplex DNA. They were run in an adjacent lane and visualized by ethidium bromide staining.  (15). Hydrolysis of the last ribonucleotide does, however, appear to be rather inefficient (approximately 10% the rate at which the bulk of the ribonucleotide residues are cleaved).
Analysis of Products-The products in an exhaustive digest of [3H]poly (A) .poly(dT) were analyzed by descending paper chromatography. As summarized in Table 111, a distribution of oligonucleotides ranging from 2-9 residues was generated. Although 85% of the [3H]poly(A) was degraded to acid-soluble products by the Drosophila RNase H, like similar enzymes (16-23), less than 0.5% was in the form of [3H]AMP. subsequent incubation at 30 "C (Fig. 4). Stoichiometric amounts of RNase H were required for the stimulation. Thus, an increase in the molar ratio of RNase H to polymeraseprimase from 0.8 to 8.0 produced an approximate &fold stimulation in DNA synthesis. The stimulation occurred only with unprimed ssDNA and was not observed with either singly primed 4x174 ssDNA or with activated duplex DNA (data not shown). The stimulation was also specific for the Drosophila RNase H. Under the same conditions, RNase H from E. coli at molar ratios ranging from 1 to 10 had no effect (data not shown).

Identity of RNase H and Stimulatory Activity
To determine whether the RNase H and stimulatory activities resided in the same protein, the purified RNase H (Frac- tion VI) was passed through a Bio-Si1 TSK-250 gel filtration column. As shown in Fig. 2, RNase H and stimulatory activity co-eluted perfectly as a 186,000-dalton protein. This finding, together with the fact that the RNase H has been purified some 2,500-fold, to near homogeneity, makes it unlikely that the two activities reside in different proteins.

Requirements for RNase H Stimulation of DNA Synthesis by Polymerase-Primase
Preincubation-Less than 2-fold stimulation in DNA synthesis was observed without preincubation of the RNase H and polymerase-primase (Fig. 5). As the time of preincubation was increased the level of stimulation increased correspondingly, so that at 30 min there was an approximate 8-fold stimulation. The requirement for preincubation suggests that there is a time-dependent formation of a complex between the RNase H and polymerase-primase. In fact, a fraction of the Drosophila RNase H remains associated with polymeraseprimase during purification of the RNase H. When the DEAE-Sephadex fraction (Fraction V) of RNase H was applied to a high performance liquid chromatography gel filtration column, approximately 10-15% of the RNase H activity recovered eluted as a protein of greater than 380,000 daltons, in association with the polymerase-primase.3 Ionic Strength-Increasing the concentration of (NH4),S04 from 1 to 41 mM had only a slight (inhibitory) effect on DNA synthesis catalyzed by the polymerase-primase. In contrast, there was a significant effect of ionic strength during the preincubation. At low (NH4)2S04 concentrations (1-11 mM) there was an approximate %fold stimulation in DNA synthesis; maximum stimulation occurred at 21 mM (approximately 4-fold). However, at 41 mM (NH4)2S04 the stimulation fell to When the polymerase-primase and RNase H were preincubated in the absence of (NH4)2S04, stimulation of DNA synthesis (upon addition of 21 mM (NH4)&04) was also only 2-fold. R. DiFrancesco and I. R. Lehman, unpublished observation.
2-fold. The interaction between the RNase H and polymeraseprimase thus appears to be influenced by ionic strength.

Effect of RNase H on DNA Products Synthesized by the Polymerase-Primase
To investigate the site (chain initiation or elongation) at which the RNase H exerts its stimulatory effect, the products of DNA synthesis were examined by alkaline agarose gel electrophoresis (Fig. 6). At 15 min in the absence of RNase H, the newly synthesized DNA chains were 250-900 nucleotides in length (Fig. 6, lane 1). As the reaction proceeded, the size of the products increased, and at 60 min they ranged from 500 to 1600 nucleotides (Fig. 6, lanes 2 4 ) . In the presence of RNase H, DNA chains in the size range 100-500 nucleotides were observed at 15 min (Fig. 6, lune 5). This size distribution is approximately 2-fold smaller than that seen with the polymerase-primase alone (compare lanes 1 and 5). After 60 min, the products increased to 400-1000 nucleotides (Fig. 6, lanes  6-8), but were still smaller than those synthesized by the polymerase-primase alone. Thus, under conditions where the Alkaline agarose gel electrophoresis of products of DNA synthesis. DNA synthesis was carried out in the absence or presence of RNase H using the conditions described in the legend to Fig. 4. Samples were removed at the indicated times and were quenched by the addition of EDTA and NaDodSOl to a final concentration of 20 mM and 1%, respectively. Aliquots, containing approximately equal amounts of 32P, were electrophoresed in a 1% alkaline agarose gel. Lanes 1-4, products synthesized by the polymeraseprimase after 15,30,45, and 60 min of incubation; lanes 5-8, products synthesized by the polymerase-primase in the presence of RNase H after 15, 30, 45, and 60 min. Molecular weight markers, whose sizes are indicated in nucleotide residues, were produced by HpaII digestion of 6x174 supercoiled duplex DNA. They were run in an adjacent lane and visualized by ethidium bromide staining.
RNase H stimulates the polymerase-primase, the DNA chains are significantly shorter than those synthesized by the polymerase-primase alone. This finding suggests that the RNase H stimulates DNA synthesis by increasing the number rather than the length of DNA chains synthesized (i.e. by increasing the number of primers).

Stimulation of Primer Synthesis by RNase H
Since degradation of RNA primers occurs in the coupled primase-polymerase reaction carried out in the presence of RNase H (data not shown), we were unable to measure primer synthesis directly. However, the coupled assay for primer synthesis, which scores the number of primed M13mp8 ssDNA circles synthesized by the polymerase-primase, does permit accurate measurement of the rate of primer synthesis. As shown in Fig. 7, a burst of primer synthesis occurred in the first 10 min of the reaction performed with polymeraseprimase alone. However, during the course of the next 50 min the number of primers increased by only 50%. When primer synthesis was carried out in the presence of RNase H, there was also an initial burst of primer synthesis. In this case, however, synthesis continued a t a linear rate for approximately 30 min. Thus, the RNase H appears to increase the recycling capability of the polymerase-primase, thereby allowing it to increase the number of primers synthesized.
To test this hypothesis, a DNA challenge experiment using the coupled assay was performed. Primer synthesis was initiated with M13mp8 ssDNA as template; at 10 min, M13Goril ssDNA was added and the reaction allowed to proceed for an additional 60 min. As shown in Fig. 8u, RNase H stimulated primer synthesis approximately %fold. To determine which of the ssDNA templates was utilized by the polymeraseprimase, aliquots were removed from the two reactions (+ RNase H) and analyzed by agarose gel electrophoresis (Fig.   8b). With polymerase-primase alone, the M13mp8 ssDNA was replicated preferentially; however, a low level of the challenging M13Goril ssDNA also underwent replication (Fig. 8b, 1st through 6th lanes)  autoradiograph revealed that 10 min after addition of the challenge DNA, 10% of the replicated DNA was M13G&il, which increased to approximately 25% after 60 min. In priming reactions carried out in the presence of RNase H, approximately 20% of the replicated DNA was M13Goril at 10 min after the addition of the challenge ssDNA, which increased to 50% after 60 min (Fig. 8b, 7th through 12th lunes). Thus, the RNase H increases the recycling capacity of the polymeraseprimase, thereby increasing the number of primers synthesized.

DISCUSSION
We have purified an RNase H from early embryos of D. melunogaster approximately 2,500-fold, to near homogeneity. It has an approximate molecular weight of 180,000 and appears to consist of two 49,000-and two 39,000-dalton polypeptides. The eukaryotes that have been examined thus far (e.g. yeast, calf thymus, and rat liver) possess several species of RNase H, which fall into two size classes: a low molecular weight form with M , values ranging from 20,000 to 40,000, and a high molecular weight form with M, values of 70,000 to 90,000 (17-23). In contrast, we have observed only a single RNase H in Drosophila embryos which is considerably higher than even the high molecular forms found in other eukaryotes.
Like the analogous enzymes from yeast, calf thymus, and KB cells, the Drosophila RNase H is specific for RNA. DNA hybrids. Activity with either single-or double-stranded RNA is approximately 3 orders of magnitude lower than that with the hybrid substrate. Moreover, neither single-nor doublestranded DNA are susceptible to hydrolysis by the Drosophila enzyme. Similarly, upon digestion of an RNA.DNA hybrid the enzyme releases oligoribonucleotides ranging in size from 2-9 residues.
The physiological role of RNase H in eukaryotes has yet to be firmly established. It has been suggested that the multiple forms of RNase H have different functions in uiuo. Biisen et al. (24) have proposed that RNase H IIb from calf thymus, a low molecular weight form, is involved in "RNA metabolism", while the RNase H1, an 80,000-dalton protein, is required for DNA synthesis. In the case of yeast, an RNase H activity has been identified that is associated with RNA polymerase I, suggesting that it is involved in transcription of ribosomal RNA genes (25). A yeast RNase H activity has also been described that stimulates DNA polymerase I activity (19). We have shown that the Drosophila RNase H is capable of removing RNA primers that were synthesized and subsequently elongated by the Drosophila polymerase-primase, suggesting that it may play a role in DNA replication.
A novel feature of the purified RNase H from Drosophila is its ability to stimulate DNA synthesis by the homologous DNA polymerase-primase. The stimulation is unique in that it is specific for a coupled reaction in which both priming and subsequent chain elongation are catalyzed by the polymeraseprimase. Karwan et al. (19)  represent an intact RNase H-polymerase-primase complex that had not undergone dissociation during purification. The fact that only a relatively small fraction (10-15%) of the RNase H remains associated with polymerase-primase up to the final step in the purification suggests that the interaction between these proteins is relatively weak. Attempts to isolate a complex following incubation of the two enzymes under conditions that yield maximal stimulation, by either gel filtration or glycerol gradient sedimentation, have thus far been unsuccessful.
The interaction of RNase H and polymerase-primase may be analogous to the situation with the bacteriophage T4 DNA polymerase and its various polymerase-accessory factors (gene 44/62 and gene 45 proteins). In the presence of these proteins both the rate and processivity of the polymerase are markedly increased (31) despite their relatively weak interaction. Similarly, the C1C2 primer recognition proteins isolated from monkey cells form a complex with the homologous DNA polymerase 01, thereby enhancing the ability of the polymerase to locate primers by eliminating nonproductive binding to ssDNA (27,28). In our efforts to reconstitute a complex of enzymes from D. metamgmterembryos that can catalyze the efficient synthesis of Okazaki fragments, we have isolated a DNA polymerase-primase and an RNase H, activities that should be essential components of such a complex. The DNA polymerase-primase is essential for Okazaki fragment synthesis; the synthesis and subsequent extension of an RNA primer. The RNase H should then remove the RNA primers, permitting joining of the Okazaki fragments by DNA ligase. Although these proteins can catalyze the basic reactions required for Okazaki fragment synthesis, other replication factors (e.g. helix destabilizing proteins, processivity factors, etc.) that should increase the efficiency of this reaction must also be involved. Efforts are currently under way to identify such proteins in Drosophila embryos.