Eur. J. Biochem. 210,855-865 (1992) 0FEBS 1992

Single-stranded DNA binding protein from calf thymus Purification, properties, and stimulation of the homologous DNA-polymerase-a- primase complex Alexey ATRAZHEV, Suisheng ZHANG and Frank GROSSE German Primate Center, Department of Virology and Immunology, Gottingen, Federal Republic of Germany (Received August 19/0ctober 2,1992)

-

EJB 92 1194

A binding protein for single-stranded DNA (ssDNA) was purified from calf thymus to near homogeneity by chromatography on DEAE-cellulose, blue-Sepharose, ssDNA - cellulose and FPLC Mono Q. The most purified fraction consisted of four polypeptides with molecular masses of 70, 55, 30, and 11 kDa. The polypeptide with the molecular mass of 55 kDa is most likely a degraded form of the largest polypeptide. The complex migrated as a whole on both glycerol gradient ultracentrifugation (s = 5.1 S) and gel filtration (Stokes’ radius z 5.1 nm). Combining these data indicates a native molecular mass of about 110 kDa, which is in accord with a 1: 1:1 stoichiometry for the 70 + 55/30/ 11-kDa complex. The ssDNA binding protein (SSB) covered approximately 20 - 25 nucleotides on M13mp8 ssDNA, as revealed from both band shift experiments and DNase I digestion studies. The homologous DNA-polymerase-a -primase complex was stimulated by the ssDNA binding protein 1.2-fold on poly(dA) . (dT)14 and 10-13-fold on singly primed M13mp8 DNA. Stimulation was mainly due to facilitated DNA synthesis through stable secondary structures, as demonstrated by the vanishing of many, but not all, pausing sites. Processivity of polymerase- primase was not affected on poly(dA) . (dT)14; with poly(dT) . (rA)lo an approximately twofold increase in product lengths was observed when SSB was present. The increase was attributed to a facilitated rebinding of polymerase GI to an already finished DNA fragment rather than to an enhancement of the intrinsic processivity of the polymerase. Similarly, products 300 - 600 nucleotides long were formed on singly primed M13 DNA in the presence of SSB, in contrast to 20- 120 nucleotides when SSB was absent. DNA-primase-initiated DNA replication on M13 DNA was inhibited by SSB in a concentrationdependent manner. However, with less sites available to begin with RNA priming, more homogeneous products were formed.

In the course of DNA replication, the double-stranded parental DNA has to be opened in order to provide a singlestranded template for the replicative DNA polymerases [l]. Zn vivo these single-stranded DNA (ssDNA) intermediates will be covered by single-strand-specific DNA binding proteins (SSB), mainly in order to prevent rehybridization to the energetically favored duplex structure and to prevent attacks from single-strand-specific nucleases. SSB proteins have been isolated from many different organisms, such as bacteria, bacteriophages, fungi, and viruses (for recent reviews see [2, 31). Correspondence to F. Grosse, Heinrich-Pette-Institut fur Experimentelle Virologie, Martinistrasse 52, W-2000 Hamburg 20, FRG Fax: +49 40 464709. Abbreviations. ssDNA, single-stranded DNA; ctSSB, ssDNAbinding protein from calf thymus; SV40, simian virus 40; RF-A, replication factor A = RP-A, replication protein A = HSSB protein, SSB from human cells; EcoSSB, ssDNA-binding protein from Escherichia coli; RFII, replicative form 11; RFIII, replicative form 111. Enzymes. DNA polymerase (EC 2.7.7.7); ribonuclease H (EC 3.1.26.4), DNA topoisomerase I1 (EC 5.99.1.3); DNA helicase (EC not defined), catalase (EC 1.11.1.6); lactate dehydrogenase (EC 1.1.1.27); polynucleotide kinase (EC 2.7.1.78).

The establishment of an in vitro replication system for simian virus 40 (SV40) DNA has provided an assay for the identification and isolation of SSB from human cells which is analogous to the SSB proteins from Escherichia coli and bacteriophage T4 [4 - 61. Eukaryotic SSB protein, also called RF-A, RP-A or HSSB, is purified from human cells as a tightly associated complex of three polypeptide subunits with molecular masses of about 70, 35, and 13 kDa. The DNA binding subunit has been localized at the 70-kDa polypeptide [7]. Subsequently, an SSB protein with similar physical properties has been isolated from yeast cells [8, 91. Recently, it has been demonstrated that cDNAs encoding the DNA binding proteins from both human and yeast SSB proteins share a high degree of similarity [9, 101. Furthermore, genetic studies indicated that the gene for the 70-kDa subunit is essential for viability in yeast and that cells with a disrupted gene are arrested at the G l / S boundary of the cell cycle [9]. The smaller sized subunits of eukaryotic SSB have no known function. A number of SSB proteins from heterologous sources have been tested for their ability to replace human SSB in SV40 replication, origin unwinding, and stimulation of DNA polymerases. All SSB proteins tested could substitute for human SSB in the T-antigen-dependent unwinding of SV40 origin-

856 containing plasmids, including E . coli SSB, bacteriophage T4 gene 32 protein, adenovirus DNA binding protein and herpes simplex virus ICP8 [ l l , 121 as well as yeast RF-A [S]. Furthermore, prokaryotic and viral SSB proteins can efficiently stimulate DNA polymerase 6 in the presence of proliferating cell nuclear antigen and the primer recognition protein RF-C. In contrast, these heterologous SSB proteins could not replace human SSB in stimulating the human DNA-polymerase-a primase complex [ll]. This indicates a highly species-specific interaction between the a-polymerase and SSB. In our recent work we have purified and characterized many replicative proteins from calf thymus glands, including the DNA-polymerase-a -primase complex [13, 141, the DNA polymerases 6 and F [I 51, a polymerase-a-stimulating ribonuclease H [16], DNA topoisomerases [17], and two DNA helicases [I 81. Because of the species-specific interaction between SSB and polymerase a and probably other components of the replicative apparatus as well, we attempted to purify SSB from bovine tissue. Here, we present a scheme for the large-scale purification of SSB from calf thymus (ctSSB), a protein highly similar to the previously described SSB proteins from human and yeast cells [4, 81. The rather large amounts obtained from this source should considerably facilitate further biochemical and physicochemical characterizations. We further report on the interactions of ctSSB with the homologous DNA-polymerase-a - primase complex.

20 cpm/pmol [ u - ~ ~ P I ~ T10 T Ppg, activated calf thymus DNA, 0.8% ampholytes (LKB, pH 3.5-10) and 0.1-2 units enzyme. One unit of polymerase catalyzes the incorporation of 1 nmol dTTP in 60 min at 37°C. DNA primase activity was assayed in a 50-pl reaction mixture containing 20 mM Tris/ acetate pH 7.3, 10 mM magnesium acetate, 1 mM dithiothreitol, 1 mM ATP, 0.1 mM GTP, 0.05 mM dATP, and 0.05 mM dGTP, 30 cpm/pmol [ E - ~ ~ P I ~ A0.1 T PmM , (nucleotide concentration) poly(dC,dT), 0.1 mg/ml bovine serum albumin, 1 unit E. coli DNA polymerase I (large fragment), and 0.2 - 2 units DNA primase. One unit of primase leads to the incorporation of 1 nmol dAMP/h at 37°C. Buffers The buffers used were as follows. Buffer A = 50 mM Tris/ HCl pH 7.4,25 mM NaCl, 5 mM MgCl,, 7 mM 2-mercaptoethanol, 0.5 mM phenylmethylsulfonyl fluoride, 0.5% Trasylo1 (Bayer, Leverkusen, F R G ; the protease inhibitors were added just before the homogenization of the tissue), 250 mM sucrose; buffer B = 20 mM Tris/HCl pH 7.4, 50 mM NaCl; buffer C = 20 mM Tris/HCl pH 7.4, 500 mM NaCl; buffer D = 20 mM Tris/HCl pH 7.4, 2 M NaCl, 50% (by vol.) ethylene glycol; buffer E = 20 mM Tris/HCl pH 7.4, 1.5 M NaC1, 50% ethylene glycol. Purification of calf thymus SSB protein

MATERIALS AND METHODS Column materials and chemicals DEAE-cellulose (DE52) was from Whatman. BlueSepharose was prepared by coupling Cibacron blue 3GA (Ciba-Geigy, Basel) to Sepharose 4B-CL (Pharmacia), essentially as described [19]. ssDNA-cellulose was prepared by epoxide cross-linking of denatured calf thymus DNA to cellulose powder as described elsewhere [20]. All chemicals were of at least analytical grade purity.

Preparation of cytoplasmic extracts

Thymus glands were taken from 3 - 5-months-old calves. After removal of fat and connective tissue, the glands were frozen in liquid nitrogen and stored until use at - 80 “C. The frozen tissue (about 300 g) was partially thawed in 1000 ml buffer A for 1 h and then blended for 2 min at the ‘low’ setting of a Waring blendor. The resulting cellular homogenate was passed through four layers of Miracloth and then centrifuged for 15 rnin at 15000 x g. The supernatant was taken as cytoplasmic extract (fraction I).

Nucleotides and DNA Ammonium sulfate precipitation Unlabeled NTPs and dNTPs were obtained from Boehringer (Mannheim, FRG). [ E - ~ ~ P I ~ Nand T P [sY - ~ ~ P I A T P Fraction 1 was further fractionated by slowly adding were from Amersham. Polynucleotides and oligonucleotides pulverized ammonium sulfate to 30% (saturation) [4]. The pH were obtained from Pharmacia. M13mp8 single-stranded of the solution was kept at pH 7.8 by occasional additions of DNA was prepared as described elsewhere [21]. For the prep- 2 M Tris base. After 30 rnin with continuous stirring, the aration of singly-primed ssDNA, 25 pmol M13mp8 was an- solution was centrifuged for 10 min at 3 5000 x g and the nealed to 25 pmol 5’-d(CCCAGTCACGACGTT)-3’ (Phar- supernatant discarded. The precipitate (3.1 g) was dissolved macia) in the presence of 10 mM Tris/acetate pH 7.5, 5 mM in 200 ml buffer B (fraction 11). MgC12, 50mM NaCl for 30min at 70°C and then slowly cooled down to room temperature. Unhybridized primer and DEAE-cellulose chromatography unreacted [ Y - ~ ~ P I Afrom T P the 5’-labeling reaction were removed by chromatography through a 2-ml BioGel A-5m Fraction I1 was diluted with doubly distilled water to the column. conductivity of buffer B and then loaded onto a 200-ml column (2.5 x 40 cm) of DEAE-cellulose (DE52) equilibrated with buffer B. The column was washed with 1000 ml buffer B Enzymes and step-eluted with buffer B containing 0.2 M NaCl (fraction DNA-polymerase-a -primase was purified to apparent 111). homogeneity essentially as described [14]. The final preparation had a specific activity of 19 800 units/mg. Polymerase Blue-Sepharose chromatography a catalytic subunit was prepared to a specific activity of 48000 U/mg as described elsewhere [22]. Its activity was asFraction I11 was adjusted to 0.5 M NaCl by the addition sayed in a 50-p1 reaction mixture containing 20 mM Tris/ of solid NaCl and then loaded onto a blue-Sepharose column acetate pH 7.3, 75 mM potassium acetate, 5 mM magnesium (1.5 x 30 cm, 50 ml), equilibrated in buffer C. The column was acetate, 1 mM dithiothreitol, 100 pM of each of the dNTPs, washed with 200 ml equilibration buffer and 100 ml buffer C

857 containing a final concentration of 0.8 M NaC1. SSB protein was step-eluted with 100 ml buffer D (fraction IV).

Table 1. Purification of ctSSB. The starting material was 300 g calf thymus tissue.

Purification steps

Protein

30% ammonium sulfate DEAE cellulose Blue Sepharose ssDNA cellulose FPLC Mono Q

mg 3100 280 130 4.1 0.7

ssDNA-cellulose chromatography Fraction IV was diluted with 3 vol. doubly distilled water and loaded onto a ssDNA-cellulose column (0.5 x 13 cm, 10 ml) equilibrated with buffer C. The column was extensively washed with loading buffer, then with a buffer containing 0.75 M NaC1, and finally eluted with buffer E. Protein-containing fractions were combined and dialyzed against buffer B (fraction V). Mono Q FPLC Fraction V was loaded onto a Mono Q column equilibrated with buffer B and washed with 20 ml equilibration buffer. ctSSB was eluted with a linear gradient (15 ml) of 0.05 - 0.4 M NaCl in buffer B. Protein eluted at about 0.28 M NaCl with a purity of > 95% (Fig. 2). These fractions were combined and dialyzed against 20 mM Tris/HCl pH 7.4, 0.1 M NaCI, 50% ethylene glycol (fraction VI) and stored at - 20 “C until use. Tricine/SDS gel electrophoresis Determination of the subunit structure was performed by using a Tricine-buffered SDS gel electrophoresis as described [23]. This system has the advantage of giving higher resolution of the small polypeptides and more reliable estimates of their molecular masses. Sucrose gradient ultracentrifugation Ultracentrifugation of ctSSB was performed on a 5 - 30% isokinetic sucrose gradient in 20 mM Tris/HCI pH 7.8, 0.5 M KCl, 1 mM dithiothreitol, and 1 mM EDTA. Sedimentation was Derformed in a Beckman SW-40 rotor for 66 h at 35006 rpm and 4°C. Catalase (11 S), lactate dehydrogenase (7.3 S), bovine albumin (4.8 S), and ovalbumin (3.5 S) were loaded onto parallel tubes to serve as sedimentation markers. After the run, each gradient was collected from the bottom of each tube and directly analysed by its absorbance at 254 nm. The positions of the protein bands were confirmed by SDS gel electrophoretic analysis of the individual fractions. Determination of the covering length of ctSSB protein on M13 (+) strand DNA by gel retardation and DNase I digestion experiments M13 ssDNA was hybridized to 5’-32P-labeledd(CCCAGTCACGACGTT) (1000 cpm/pmol) as described above. Then 5 p1 of reaction mixture containing 0.2 pg DNA and varying amounts of ctSSB protein in 20 mM Tris/HCl pH 7.6, 1 mM dithiothreitol, and 5% glycerol were incubated for 5 min at room temperature. After addition of xylene cyanol to a final concentration of 0.02%, reaction mixtures were loaded onto a vertical 1.5% agarose gel (150 x 80 x 1 mm). Electrophoresis was carried out at 2 V/cm for 6 h at room temperature. After electrophoresis, the gel was exposed overnight at 4°C to a Kodak XAR5 film with an intensifying screen. 0.15 pg M13 DNA was preincubated with 2.45 pg ctSSB in 5 pl 20 mM Tris/HCl pH 7.6, 1 mM dithiothreitol, 5% glycerol, 10 mM MgC12 for 5 min at 20°C. Digestion was performed by incubation with 6 U DNase I for 5 min at 37°C.

DNase I was destroyed by a 10-min incubation at 96”C, the resulting fragments were labeled with [y-32P]ATPby using the phosphate exchange reaction of polynucleotide kinase [24]. Electrophoresis and autoradiography was as described above. Processivity measurements Processivities of DNA polymerase were measured on either poly(dA) . oligo(dT),, and poly(dT) . oligo(rA)lo as described [25] or on single-primed M13 ssDNA. In the latter case 20 pl of a solution containing 20 mM Tris/acetate pH 7.3, 5 mM magnesium acetate, 75 mM potassium acetate, 1 mM dithiothreitol, 0.1 mg/ml bovine albumin, 0.3 pg singleprimed M13mp8 ssDNA (6.3 nM 3’-OH primer ends) and 0.5 U (4 nM) polymerase - primase or 0.5 U (4 nM) u-subunit of polymerase u. The reaction was started by adding a mixture of 10 pM [E-~’P]~TTP (4000 cpm/pmol), 50 pM each of dATP, dGTP, and dCTP. After 20 s, 1 min, and 3 min, 6-p1 samples were taken and the reaction was stopped by the addition of EDTA to a final concentration of 10 mM. ctSSB and polymerase u were digested by incubation with 4 pg proteinase K (13.7 mg/ml; Sigma) for 2 h a t 37°C. Thereafter, the lengths of the newly formed products were analyzed by denaturing polyacrylamide gel electrophoresis, as described [25]. Nitrocellulose filter binding measurements Filter binding experiments were performed as described elsewhere [20]. M13 ssDNA was hybridized to 5’-32P-labeled d(CCCAGTCACGACGTT) (1000 cpm/pmol) as described above. Then 5 p1 of reaction mixture containing 0.1 pg DNA and varying amounts of ctSSB protein in 20 mM Tris/HCl pH 7.6,50 mM NaC1,l mM EDTA, 1 mM dithiothreitol, and 5% glycerol were incubated for 5 min at room temperature. Thereafter, the reaction mixture was passed through alkalinetreated nitrocellulose membranes (BA85, 7.5 mm diameter, Schleicher & Schull, Dassel, FRG) by applying a water stream vacuum. The filters were washed with 2 ml reaction buffer, dried under red light, and the retained radioactivity was determined by liquid scintillation counting.

RESULTS Purificationof calf thymus single-stranded DNA binding protein (ctSSB) Analogous to previously published purification schemes for yeast and human SSB proteins [4-6, 81 and based on the visual detection of the characteristic trimeric/tetrameric protomer bands as derived from SDS gels, we have developed a fast and efficient purification scheme for ctSSB. The purification is summarised in Table 1 and an outline of the individ-

858

A 280

DEAE Cellulose

5 0.2 M

1M

NaCl

NaCl

A,,,

Blue Sepharose

4

NaCl 50% EG?

+ $

- 1

200

3lutedVol.

400

I

I

600 ml I

... :.::: ... _.:. ... ... ..... .:. ... :,.... .... ... .... .... ....

200

Eluted VolumelYO

I

300 ml I

/-

0.4 M NaCl

A280

FPLC Mono Q

ssDNA Cellulose

2

0.75M NaCl

Eluted Volume

5?

1.5 M NaCl 50%EG

-1

100 ml Eluted Volume

10

20 ml I

Fig. 1. Calf thymus SSB protein at various stages of purification. The hatched areas under the absorption peaks indicate ctSSB-containing fractions, as detected by their polypeptide pattern on Tricine/SDS/polyacrylamide gels.

ual purification steps is given in Fig. 1. In order to scale up the purification of a homogeneous ctSSB protein, it was necessary to include an ammonium sulfate precipitation step and a DEAE-cellulose step. However, because the first three chromatographic steps can be performed in a batch-wise procedure, milligram amounts of homogeneous ctSSB are obtainable within 3 days. Structure of ctSSB protein SDS gel electrophoresis of the FPLC Mono-Q-purified ctSSB protein revealed four protomeric bands with molecular masses of 70, 55, 30, and 11 kDa with an estimated purity of more than 95% (Fig. 2). Since there was an inverse relationship between the occurrence of the 70-kDa and 55-kDa proteins, the latter form is most likely a proteolytic product of the former, as has been shown earlier with human SSB protein RP-A [5]. All our attempts to avoid this degradation have failed so far. In preparative ultracentrifugal runs, ctSSB protein migrated at 5.1 S as a homogeneous, distinct band. The relative intensities of the subunits remained constant over the absorbance peak, indicating a strong physical interaction between the three subunits (Fig. 3). Since, after ultracentrifugation we did not observe a separation of the 55-kDa and 70-kDa forms (Fig. 3, insert), we conclude that both are able to form complexes with the two smaller subunits of ctSSB. The Stokes’ radius for the native enzyme was estimated from runs on a calibrated Sephacryl S200 column as 5.1 nm (not shown). Together with the sedimentation data, this suggests a native molecular mass of around 110 kDa, in-

Fig. 2. SDS gel analysis of purified ctSSB. Lane 1, ctSSB protein after chromatography on an FPLC Mono Q column; lane 2, ctSSB after chromatography on a ssDNA-cellulose column; lane 3, molecular mass standards. To achieve a high resolution of the small-molecularmass band(s), gel electrophoresis was carried out on a 16% Tricine/ SDS/polyacrylamide gel, as described in Materials and Methods.

dicating a 1 : 1:1 stoichiometry for the three protomeric subunits. Very similar data have been found for the homologous enzymes from yeast [S] and man [5, 6, 261.

859 kDa

A 254

- 70 - 55

I

I

6

7

I

I

I I l l

8 9 1 0 1 1 1213

3.5 s Sucrose Gradient 30% 4 5% I I I I I I I I I I I I I I I I I I I I I I I 1 3 5 7 9 11 13 15 17 19 21 Fraction Number

Fig. 3. Sucrose gradient ultracentrifugationof purifiedctSSB. Ultracentrifugation of ctSSB was performed on a 5 - 30% isokinetic sucrose gradient in a Beckman SW-40 rotor for 66 h at 35000 rpm and 4°C. The pQSltiOnS of sedimentation rate markers lactate dehydrogenase (7.3 S) and ovalbumin (3.5 S) are indicated by arrows. The insert shows an SDSjPAGE analysis of the fractions taken from the absorbance peak. The trail of material absorbing at 254 nm which follows the peak is due to non-proteineous contaminants of the loading buffer.

Table 2. Molecular properties of calf thymus SSB.

Property

Value

Molecular mass (a) SDSjPAGE (b) gel filtration Sedimentation coefficient (sz0,,J Stokes' radius Amount (molecules/cell)" Covering length per trimer

70 (55), 30, 11 kDa 110 kDa 5.1 s 5.1 nm 30000 22 - 26 nucleotides

a Based on the assumption of 4.6 x lo9 cellsjg thymus [40] and a 30% purification yield for ctSSB.

Determination of the covering length of ctSSB protein on MI3 (+) strand DNA The covering lengths of SSB protein on natural ssDNA were determined by titrating M13 DNA with various amount of ctSSB and subsequent analysis of the binding products by agarose gel electrophoresis [20,27]. When increasing amounts of SSB were added to M13 DNA, slower migrating product bands were formed until ctSSB was in an 12.5 - 15-fold mass excess over DNA (Fig. 4). Under the assumption that 1 mg/ ml ctSSR has an absorption at 280 nm of 1 cm-' and that the native protein has a molecular mass of 111 kDa, a covering length of 22 - 26 nucleotides/trimer can be calculated. DNase I digestions of M13 DNA coated with a 16-fold mass excess of ctSSB revealed oligomer products in the range of 10-30 nucleotides with a center of density at about 20 nucleotides, whereas naked DNA yielded oligomers of 6 - 14 nucleotides in length with most of the DNA degraded to a

Fig. 4. Gel retardation assay for demonstrating interactions between ctSSB and M13 ssDNA. A mixture of 0.2 pg primer-labeled M13mp8 ssDNA (see Materials and Methods) was incubated with ctSSB in 20 mM TrisjHCl pH 7.0,1 mM EDTA for 10 min at 20°C to give the indicated ctSSBjssDNA mass ratios and then applied onto a 1.5% agarose gel. Electrophoresis was carried out at 2 Vjcm in 50 mM Tris/ phosphate pH 8.3, 1 mM EDTA. After electrophoresis, the gel was exposed to a Kodak XAR5 film.

form that was not accessible to the subsequent polynucleotide kinase reaction (Fig. 5).

Stimulation of the homologous DNA-polymerase-a -primase complex by calf thymus SSB Human SSB has been shown to stimulate its homologous polymerase -primase on poly(dA) . oligo(dT) about 10-fold [ l l , 26,28,29]. Under the same reaction conditions [ll], ctSSB stimulated calf thymus polymerase a only marginally (Fig. 6 A). On the other hand, there was an about 10- 13-fold stimulation when singly-primed M13 DNA served as templateprimer (Fig. 6A). An electrophoretic analysis of the replication products on M13 DNA revealed that most of the pausing sites of polymerase a were removed (Fig. 6B). From this it is concluded that a facilitated DNA synthesis through stable secondary structures of the template DNA is mainly responsible for the stimulatory effect of ctSSB on the homologous DNA-polymerase-a -primase complex.

Processivity of DNA-polymerase-a - primase in the presence of ctSSB The processivity of the bovine polymerase - primase was measured on both poly(dA) . (dT)14 and poly(dT) . (rA)lo and compared to products synthesized during very short times of incubation on single-primed M13 ssDNA. With poly(dA) . oligo(dT) there was no apparent increase of the enzyme's processivity in the presence of up to a 20-fold mass excess of SSB over the synthetic template-primer (data not shown). This might be attributed to the weak binding of this particular synthetic template-primer to ctSSB (see below). Therefore, we analyzed the processivity of polymerase a on poly(dT) . (rA)lo.

860 A

, 2

4

6

8

1

0

ctSSB/ DNA

Fig. 5. DNase I digestion of free and ctSSB-coated M13 ssDNA. Lane 1, the 5’-labeled 15-nucleotide universal primer of MI3 DNA; lane 2, alkaline-digested poly(rA) as length marker; lane 3, M I 3 DNA that had been incubated with a 16.7-fold mass excess of ctSSB and then subjected to DNase I digestion (see Materials and Methods); lane 4, free MI3 DNA that had been digested with DNase I under identical conditions as the SSB-coated material shown in lane 3. (20 mer indicates a 20-residue oligonucleotide.)

In the absence of SSB, products centered at about 20 and 30 nucleotides were formed (Fig. 7, lanes 1- 4) as has been shown previously [22, 251. At a 17-fold mass excess of SSB over DNA, the 20-residue product became less apparent and 30and 40-residue products showed up (Fig. 7, lanes 6 - 8). Furthermore, after a 3-min incubation, a considerable amount of the DNA template was converted to fully dsDNA of about 300 nucleotides in length (Fig. 7, tops of lanes 6 and 14). Similar effects were observed when the free DNA-synthesizing subunit (a-subunit) was used instead of the four-subunit polymerase-primase complex (Fig. 7, lanes 9- 16). With the free a-subunit, even longer products were synthesized in the presence of SSB (Fig. 7, lanes 14-16). Since there were discrete steps of increasing product lengths, these results suggest that the intrinsic processivity of polymerase a might not be affected by SSB; rather the reinitiation of DNA synthesis at an already finished product may be favored over initiation at a hitherto unelongated primer. On single-primed M13 ssDNA, DNA polymerase a generated products of less than 120 nucleotides in length with three strong pausing sites (Fig. 8, lanes 1 - 3). A very similar pattern

Fig. 6. Stimulation of DNA polymerase a on singly-primed M13 singlestranded DNA. (A) The elongation of a 15-residue oligonucleotide hybridized to M13mp8 ssDNA was measured in the presence of the indicated mass ratios of ctSSB/DNA. Polymerase c( activity was assayed in 50 p1 20 mM Tris/acetate pH 7.3, 75 mM potassium acetate, 5 mM magnesium acetate, 1 mM dithiothreitol, 100 pM of each of the dNTPs, 20 cpm/pmol [ce3’P]dTTP, 0.1 pg M13mp8 ssDNA, primed with 5’-d(CCCAGTCACGACGTT)-3’, and 0.5 U polymerase c(. After a 5-min incubation at 3 7 T , samples were withdrawn and the incorporated radioactivity was determined by scintillation counting of acid-insoluble material. DNA synthesis on poly(dA) . oligo(dT) was measured exactly as described by Kenny et al. [l 11. (B) Reaction mixtures (10 pl) were the same as described above but, instead of [a3ZP]dTTP,[5’-3zP]d(CCCAGTCACGACGTT)-3’ was used. After a 5-min incubation at 37”C, samples were withdrawn and the reaction was stopped by the addition of EDTA to a final concentration of’ 10 mM. Proteins were digested with 4 pg proteinase K (2 h, 37”C), then formamide was added to a final concentration of 50% (by vol.). The samples were loaded onto a 10% polyacrylamide gel containing 7 M urea. After electrophoresis for 3 h at 35 V/cm, the gel was exposed overnight at -70°C by using an intensifying screen. The numbers on the right indicate fragment lengths in nucleotides, as derived from HueIII-digested @XI74 DNA as size marker.

was observed for DNA synthesis catalyzed by the free a-subunit of polymerase a (Fig. 8, lanes 7 - 9). In the presence of saturating amounts of ctSSB, the strong pausing sites at about 35, 50, and 65 nucleotides downstream of the primer were more or less completely removed, whereas the weak pausing site at about 7 5 nucleotides and the strong one at about 125 nucleotides were much less affected (Fig. 8, lanes 4-6 and 10- 12). Generally, much longer products were produced with SSB present, independent of whether DNA synthesis was catalyzed by the four-subunit polymerase - primase complex

861

Fig. 7. Processivity of DNA polymerase a on poly(dT) * (rA),o in the presence and absence of ctSSB. Lanes 1-4, oligonucleotide products formed by the four-subunit polymerase-a - primase complex on poly(dT) . (rA)lo within 1, 3, 10, and 30min, respectively; lanes 5 - 8, oligonucleotides formed by the four-subunit polymerase-a primase complex on ctSSB-coated poly(dT) . (rA)lo within the same time intervals; lanes 9- 12, DNA fragments generated by the x-subunit of polymerase a on poly(dT) . (rA)lo within 1,3,10, and 30 min; lanes 13- 16, oligonucleotide products synthesized by the a-subunit of polymerase a on ctSSB-covered poly(dT) . (rA)lo within the indicated incubation times; lane 17, alkaline-digested poly(rA) as length standard; lane 18, the 15-nucleotide universal primer of MI3 as size marker. (35 mer indicates a 35-residue oligonucleotide.)

or the free cc-subunit (Fig. 8, lanes 4-6 and 10- 12). Most strikingly, during an incubation period of 20 s and in the presence of saturating amounts of ctSSB, products of 300 600 nucleotides in length were formed (Fig. 8, lanes 4 and 10). From this an elongation rate for polymerase CI of at least 1020 s-’ can be derived, which comes close to the migration rate of the eukaryotic replication fork [30]. Since, under the conditions chosen, only 1% of the primers have been elongated (data not shown, but see the strong 15-nucleotide band representing unelongated primer in Fig. 8), the observed fragments of 300 - 600 nucleotides have most likely been formed en route, with an unaltered intrinsic processivity of polymerase CI of 10-20 and a largely facilitated rebinding to an already synthesized DNA fragment.

Fig.8. DNA fragments synthesized by DNA polymerase a on singly primed M13 ssDNA in the absence and presence of ctSSB. The reaction mixture (20 PI) contained 20 mM Tris/acetate pH 7.3,75 mM potassium acetate, 5 mM magnesium acetate, 1 mM dithiothreitol, 0.1 mg/ ml bovine serum albumin, 100 pM dNTP, 0.3 pg [5’-32P]d(CCCAGTCACGACGTT)-primed M13mp8 ssDNA, in the absence of ctSSB (lanes 1-3, 7--9) or with 5 pg ctSSB (lanes 4-6, 10-12). 0.5 U polymerase-primase (lanes 1-6) or 0.5 U free a-subunit (lanes 712) were used to catalyze DNA synthesis. Samples were withdrawn after 20 s (lanes 1,4, 7, lo), 1 min (lanes 2, 5 , 8, 1l), and 3 min (lanes 3, 6, -9, 12) incubation at 37°C. HueIII-digested @X174 DNA was taken as nucleotide length standard (numbers on right).

ctSSB inhibits the primase activity of the homologous DNA-polymerase-a - primase complex

On single-stranded DNA templates, the eukaryotic DNApolymerase-a - primase complex can perform DNA synthesis de n o w by using its intrinsic primase activity for the formation of RNA primers and elongating these primers by the DNA polymerase activity. Since in the living cell probably most, if not all, of the ssDNA is covered with SSB, it was of interest to study primer formation in the presence of various amounts of SSB. Surprisingly, primase-primed DNA synthesis was completely inhibited when the ssDNA template was saturated with SSB (data not shown). From this, it is concluded that primer formation rather than DNA synthesis is inhibited by ctSSB. This was shown directly by measuring the incorporation of [ E - ~ ~ P I C T into P acid-precipitable material in the presence of various amounts of ctSSB protein. As expected, rNTP incorporation was no longer detectable when the ssDNA had been saturated with ctSSB. At a fourfold mass excess of SSB over DNA (about 33% saturation), the rate of primer formation was inhibited by 50% ; at 66% saturation RNA priming

862

A

0

2 4 ctSSB/M13 DNA

6

I

Fig. 9. Influence of ctSSB on the self-primed replication of M13 ssDNA by the DNA-polymerase-a- DNA-primase complex. (A) Primer synthesis by polymerase-primase was measured at various mass ratios of ctSSB/M13 DNA. Reaction mixtures contained 20 mM Tris/acetate pH 7.3, 75 mM potassium acetate, 5 mM magnesium acetate, 1 mM dithiothreitol, 100 pM of each of the dNTPs, 500 pM each of ATP, GTP and UTP, 50 pM [x-~*P]CTP(100 cpm/pmol), 0.5 pg M13mp8 ssDNA and 5 U DNA-polymerase-a- DNA-primase complex. Incubation time was 30 min at 37 “C. Thereafter, samples were withdrawn and acid-insoluble product formation was measured by scintillation counting. (B) Product analysis of primase-primed MI3 replication on a 1% agarose gel under non-denaturing conditions (50 mM Tris/phosphate pH 8.3, 1 mM EDTA). The gel was stained with ethidium bromide. (C) Autoradiograph of the newly synthesized replication products after strand separation by alkaline agarose gel electrophoresis (30 mM NaOH, 50 mM NaC1). Reaction mixtures were the same as described above but, RFII indicates the open circle double-stranded form of M13 DNA in B and C, [x-~’P]CTPwas replaced by 20 cpm/pmol [U-~’P]~TTP. (replicative form II), RFIII indicates the linear duplex form of M I 3 DNA (replicative form 111). The latter form arises because our preparations of M13 ssDNA contain up to 20% of nicked, i.e. linear, ssDNA.

was diminished to about 10% of the value measured on naked DNA (Fig. 9 A). An analysis of the replication products on neutral agarose gels revealed the production of completely replicated M13 DNA, i.e. RFII and RFIII, when ctSSB was present up to a fourfold mass excess (Fig. 9 B). On alkaline agarose gels, the newly formed replication products displayed lengths of 2000 - 4000 nucleotides (Fig. 9C). Since the 7228nucleotide M 13mp8 genome had been completely replicated, and the newly formed fragments were 2000 - 4000 nucleotides long, 2-4 primers/genome must have been formed under conditions where 0 - 33% of the DNA was covered with SSB. At 66% saturation with SSB, primase-primed DNA synthesis was still detectable but already strongly inhibited. Under this condition, about 20% of the starting material remained unreplicated and most of the other template molecules were only partially replicated (Fig. 9B). Product analysis on alkaline agarose gels revealed a rather sharp band of about 3800 nucleotides in length (Fig. 9C). Since under these conditions fully replicated RFIIiRFIII was not detected (Fig. 9B), only one or two priming events have taken place on about 80% of the ssDNA template molecules, whereas the remaining 20%, which most likely represent completely covered templates (which in turn points to a weak positive cooperative binding of SSB), were not accessible for primase-primed DNA synthesis.

A

B

M13 (+)DNA

B

0

I

-11’

/

20

0

2

poiy(dA).oligo(dT)

4

6 8 10 CtSSBissDNA

12

14

16

18

20

0 20 40 60 80 100 poly~dA).oligo(dT)lMl3(+) DNA

Binding of ctSSB to M13 ssDNA and poly(dA) (dT)14

Fig. 10. Nitrocellulose filter binding assay of ctSSB binding to M13 ssDNA and poly(dA) * (dT)14.(A) 0.1 pg [5’-32P]d(CCCAGTCACGACGTT)-3’-primed M13mp8 ssDNA or [5’-32P](dT)14-primed poly(dA) was incubated with the indicated amounts of ctSSB in 5 pl 20 mM Tris/HCl pH 7.6, 3 5 mM NaCl, 1 mM EDTA, 2 mM 2mercaptoethanol, 5% glycerol (buffer F) for 10 min at 20°C and then transferred onto a nitrocellulose filter as described in Materials and Methods. After washing with 2 ml buffer F, the filters were dried and the retained radioactivity was measured by scintillation counting. (B) Concurrent binding of ctSSB to [5’-32P]primer-labeledM13mp8 ssDNA and non-radioactive poly(dA) . oligo(dT). To a solution of 0.1 pg primed M13 DNA in 5 p1 buffer F the indicated excess of poly(dA) . oligo(dT) and 0.5 pg ctSSB were added. After a 5-min incubation at 20”C, the retained radioactivity was determined as described above.

As has been shown above, the DNA synthesizing activity of polymerase a was largely stimulated when SSB-covered singly primed M13 DNA served as template-primer. However, on poly(dA) . oligo(dT) there was only a marginal stimulatory effect. This is in some contrast to observations with SSB protein and polymerase -primase from human cells [ 11, 261. In order to understand this discrepancy, we have compared the binding of ctSSB to both M13 ssDNA and poly(dA) . oligo(dT) by using the nitrocellulose filter binding technique

[20]. In the absence of ctSSB, both DNA substrates passed through the filters, whereas in the presence of a fivefold mass excess of SSB over M13 ssDNA more than 95% of M13 DNA was retained (Fig. 10A). By contrast, even at a 20-fold excess of ctSSB over poly(dA) . oligo(dT), only 5% of the synthetic polymer was withheld by the filter. Thus, ctSSB binds poly(dA) . oligo(dT) much more weakly than M13 ssDNA.

-

863 The difference in the binding affinity between M13 DNA and poly(dA) . oligo(dT) was quantified by competing out the retention of labeled M13 DNA by the addition of various amounts of unlabeled poly(dA) . oligo(dT). As shown in Fig. 10B, a 100-fold mass excess of poly(dA) . oligo(dT) was required to inhibit the retention of M13 ssDNA by 50%. From this it is concluded that ctSSB binds poly(dA). oligo(dT) 100-fold less efficiently than M13 ssDNA. Furthermore, under the reaction conditions used for measuring the stimulatory effect on both the activity and the processivity, the covering of poly(dA) . oligo(dT) was less than 5%. This might explain the low stimulatory effects on this particular synthetic template-primer.

DISCUSSION Eukaryotic single-stranded DNA binding proteins have been isolated so far from both human and yeast cells [5, 6, 81. Calf thymus is an excellent source for the large-scale purification of various replicative proteins from mammals. Therefore, we have devised a purification procedure for calf thymus SSB, mainly to provide sufficient material for subsequent biophysico-chemical studies and for analyzing interactions with the homologous DNA-polymerase-a -primase complex. By using a procedure analogous to that described for the purification of yeast RF-A [8], about 2 mg homogeneous SSB was obtained from 1 kg thymus. SDSjPAGE of the most purified fractions revealed four polypeptides with molecular masses of 70, 55,30, and 11 kDa. From the sedimentation coefficient of 5.1 S and the Stokes’ radius of 5.1 nm, a native molecular mass of 111 kDa was calculated [31]. Hence, the bovine SSB consists of three polypeptides of 70 or 55 kDa, 30 kDa and 11 kDa in a 1: 1:l stoichiometry. Similarly, the human SSB protein RP-A, has been isolated as a complex consisting of 70 + 53-kDa, 32-kDa 14-kDa proteins [5]. In this case, partial proteolysis has been made responsible for the production of the 53-kDa form from the 70-kDa subunit [7].We observed an increase of the relative amounts of the 55-kDa form and a decrease of the 70-kDa form in lengthy purifications. Therefore, we suspect that the former is a proteolyzed form of the latter. By using a gel retardation assay for M13 ssDNA in the presence of various amounts of ctSSB, a saturation of M13 DNA was achieved at a 13-15-fold mass excess of SSB trimers over DNA. From this, a covering length of about 22 25 nucleotides/trimer can be calculated. The covering length was confirmed by DNase I digestion experiments of M13 DNA in the presence of a 16-fold mass excess of ctSSB, which gave products with an average length of around 20 nucleotides. After this work had been completed, a covering length of about 30 nucleotides was determined for SSB proteins from human and yeast cells by band shift experiments with (dT)sO and (dT)70, where (dT)sO bound one molecule SSB/molecule, whereas (dT)70 bound two molecules SSB/molecule [32]. From the same study, noncooperative or only weak cooperative binding of SSB to DNA was concluded [32]. In our band shift studies with M13 DNA, we observed the occurrence of a discrete and fully covered DNA band when only half-saturating amounts of SSB were added to ssDNA. Furthermore, RNA primer formation (which is completely inhibited on fully covered ssDNA) was already abolished in 20% of the template molecules at 66% saturation of DNA with ctSSB. Taken together, these results indicate a weak, but still detectable, positive cooperative mode of binding.

ctSSB stimulated the activity of the purified homologous DNA-polymerase-a -primase complex on long stretches of natural single-stranded DNA 10 - 13-fold. This is comparable to the stimulatory effect of human SSB exerted on the human polymerase-a -primase complex [28]. An analysis of the products formed on single-primed M13 ssDNA revealed that most of the pausing sites disappeared in the presence of ctSSB. If one assumes that most of these sites are halting points of the polymerase -primase complex in front of stable hairpin structures [33], the vanishing of these sites can be interpreted as melting out of short segments of double-stranded DNA by ctSSB. The removal of pausing sites is certainly an important contributor to the 10- 13-fold stimulatory effect of ctSSB on the DNA synthesizing activity of polymerase a. However, other effects, such as a more processive synthesis of polymerase M on ctSSB-covered DNA, might also contribute to the stimulatory effect. Therefore, we measured the processivity of poly(dA) . oligo(dT)14 and poly(dT) . (rA)lo. With poly(dA) . oligo(dT)14, no effect on the processivity of polymerase a was observed, which might be attributed to the poor binding of ctSSB to this homopolymer. A comparable low binding affinity for this template has been also observed with EcoSSB, which binds poly(dA) 100-fold less efficiently than bacteriophage fd DNA [34,35]. On the other hand, there was a significant increase in product length formation on poly(dT) . (rA)lo or single-primed M13 ssDNA. Interestingly, the products formed on poly(dT) were multiples of the inherent processivity of polymerase a, which strongly suggests that ctSSB may stimulate reinitiation events at the same already synthesized DNA fragment after one processive cycle has been finished. This may be accomplished by reducing the binding affinity of polymerase - primase to SSB-coated ssDNA. This in turn would diminish the ‘challenger’ action of the large excess of ssDNA otherwise present. A similar effect seems also to be responsible for the longer products observed on singly primed M13 DNA when ctSSB is present. In this case, products of more than 600 nucleotides in length were observed, whereas in the absence of SSB product lengths did not exceed 120 nucleotides. Very similar findings have been recently reported by Tsurimoto and Stillman: in the presence of human SSB, the homologous polymerase a generated 300 400-nucleotide replication products, while in the absence of SSB less than 20 nucleotides were synthesized from the initiating primer molecule [28,29]. Tsurimoto and Stillman attributed this to an increased processivity of polymerase M on SSB-coated templates but, for the reasons discussed above, the longer products may reflect DNA synthesis en route, i.e. polymerase M synthesizes DNA with an intrinsic processivity of 10 - 20 nucleotides, but rebinding to an already finished DNA fragment is largely increased over binding to a primer molecule not yet elongated. In any case, it is interesting to note that the products formed by polymerase a in the presence of human or bovine SSB exceed the lengths of eukaryotic Okazaki fragments on the lagging strand of the replication fork. Furthermore, the in vitro rate of DNA synthesis by polymerase M on SSB-coated ssDNA was 10 -20 nucleotides/ s, which is close to the rate of fork propagation in the living cell [30]. Thus, polymerase a alone is able to synthesize DNA on SSB-covered natural templates in vitro with a rate and product length that strongly resemble Okazaki fragment formation in vivo. Beside its DNA synthesizing activity, eukaryotic DNA polymerase SI contains an inherent DNA primase activity, which is localized on the two smallest subunits of this enzyme

864 [36]. Somewhat unexpectedly, SSB-covered ssDNA is not an efficient substrate for the primase. When M13 ssDNA was saturated with ctSSB, RNA-primer formation, and thus primase-initiated DNA replication, were no longcr detectable. When about 66% of the single-stranded M13 DNA was covered, RNA primer formation was diminished by 85%. Under this condition, obviously only one primer has been formed and subsequently elongated to a 4000-nucleotide fragment. Very similar results were obtained earlier by studying primase-initiated DNA replication on EcoSSB-covered M13 DNA [37]. Hence, a reduced binding of polymerase -primase to SSB-coated ssDNA may be responsible for the strong inhibition of the primase activity of the polymerase - primase complex. Recently, Collins and Kelly have shown that the human SSB inhibits the priming reaction of the human polymerase -primase complex. Moreover, this effect could be partially reversed by the addition of SV40 large T-antigen [38]. T-antigen is an SV40 origin binding protein that unwinds DNA by its inherent DNA helicase activity and delivers polymerase-cc - primase to the lagging strand of the replication fork [39]. In analogy, we suppose that there are cellular protein factors that stimulate RNA primer formation on SSB-covered ssDNA. The system consisting of ssDNA, purified ctSSB, and purified polymerase - primase should be very helpful for the identification of the hypothesized cellular priming factors.

12.

13. 14. 15.

16. 17. 18. 19. 20.

This work was supported by the Deutsche Forxhungsgemeinschuft. A. A. gratefully acknowledges the receipt of fellowships from EMBO and the Max-Planck Society. 21.

REFERENCES 1. Kornberg, A. & Baker, T. (1991) DNA replication, 2nd edn, W. H. Freeman, San Francisco CA. 2. Meyer, R. R. & Laine, P. S. (1990) The single-stranded DNAbinding protein of Escherichia coli, Microhiol. Rev. 54, 342 380. 3. Chase, J. W. & Williams, K. R. (1986) Single-stranded DNA binding proteins required for DNA replication, Annu. Rev. Biochem. 55, 103- 136. 4. Wobbe, C. R., Weissbach, L., Borowiec, J. A., Dean, F. B., Murakami, Y . , Bullock, P. & Hurwitz, J. (1987) Replication of simian virus 40 origin-containing DNA in vitro with purified proteins, Proc. Natl Acud. Sci. USA 84, 1834- 1838. 5. Wold, M. S. & Kelly, T. (1988) Purification and characterization of replication protein A, a cellular protein rcquired for in vitro replication of simian virus 40 DNA, Proc. Nut1 Acad. Sci. USA 85,2523 -2527. 6. Fairman, M. P. & Stillman, B. (1988) Cellular factors required for multiple stages of SV40 DNA replication in vitro, EMBO J. 7, 1211-1218. 7. Wold, M. S., Weinberg, D. H., Virshup, D. M., Li, J. J. & Kelly, T. J. (1989) Identification of cellular proteins required for simian virus 40 DNA replication, J. Biol. Chem. 264,2801 -2809. 8. Brill, S. J. & Stillman, B. (1989) Yeast replication factor-A functions in the unwinding of the SV40 origin of DNA replication, Nature 342, 92 - 9 5. 9. Heyer, W.-D., Rao, M. R. S., Erdile, L. F., Kelly, T. J. & Kolodner, R. D. (1990) An essential Saccharomyces cerevisiae single-stranded DNA binding protein is homologous to the large subunit of human RP-A, EMBO J . 9,2321 -2329. 10. Erdile, L. F., Heyer, W.-D., Kolodner, R. & Kelly, T. J. (1991) Characterization of a cDNA encoding the 70-kDa singlestranded DNA-binding subunit of human replication protein A and the role of the protein in DNA replication, J . Biol. Chem. 266, 12090-12098. 11. Kenny, M. K., Lee, S.-H. & Hurwitz, J. (1989) Multiple functions of human single-stranded DNA binding protein in simian virus 40 DNA replication : Single-strand stabilization and stimu-

22. 23.

24.

25. 26.

27. 28. 29.

30. 31.

32.

lation of DNA polymerases M and 6, Proc. Nut1 Acud. Sci.USA 86,9757-9761. Virshup, D. M. & Kelly, T. J. (1989) Purification of replication protein C, a ccllular protein involved in the initial stages of simian virus 40 DNA replication in vitro, Proc. Nut1 Acad. Sci. USA 86, 3584-3588. Grosse, F. & Krauss, G. (1981) Purification of a 9 S DNA polymerase M species from calf thymus, Biochemistry 20,5470 5475. Nasheuer, H.-P. & Grosse, F. (1987) Immunoaffinity-purified DNA polymerase M displays novel properties, Biochemistry 26, 8458 - 8466. Nickel, W., Austermann, S., Bialek, G. & Grosse, F. (1992) Interactions of azidothymidine triphosphate with the cellular DNA polymerases M , 6, and E and with DNA primase, J . Biol. Chem. 267, 848 - 854. Hagemeier, A. & Grosse, F. (1 989) A distinct form of ribonuclease H from calf thymus stimulates its homologous DNA-polymerase-a -primase complex, Eur. J . Biochem. 185, 621 -628. Schomburg, U. & Grosse, F. (1986) Purification and characterization of DNA topoisomerase I1 from calf thymus with polypeptides of 175 and 150 kDa, Eur. J . Biochem. 160,451 -457. Zhang, S. & Grosse, F. (1991) Purification and characterization of two DNA helicases from calf thymus nuclei, J . Biol. Chem. 266,20483-20490. Bohme, H. J., Kopperschlager, G., Schulz, J. & Hofmann, E. (1972) J . Chromatogr. 69,209 -214. Grosse, F., Nasheuer, H. P., Scholtissek, S. & Schomburg, U. (1986) Lactate dehydrogenase and glyceraldehyde-phosphate dehydrogenase are single-stranded DNA-binding proteins that affect the DNA-polymerase-cc- primase complex from calf thymus, Eur. J . Biochem. 160,459-467. Grosse, F. & Krauss, G. (1984) Replication of M13mp7 singlestranded DNA in vitro by the 9-S DNA polymerase a from calf thymus, Eur. J . Biochem. 141, 109- 114. Bialek, G. & Grosse, F. (1992) The DNA synthesizing subunit of polymerase -primase from calf thymus, J . Bid. Chem. 267, 2915-2919. Schlgger, H. & von Jagow, G. (1987) Tricine-sodium dodecylsulfate-polyacrylamidegel electrophoresis for the separation of proteins in the range of l to 100 kDa, Anal. Biochem. 166, 368 - 379. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular cloning: a laboratory manual, Cold Spring Harbor Laboratories, Cold Spring Harbor NY. Hohn, K. T. & Grosse, F. (1987) Processivity of the DNA polymerase a primase complex from calf thymus, Biochemistry 26, 2870-2878. Kenny, M. K., Schlegel, U., Furneuax, H. & Hurwitz, J. (1990) The role of human single-stranded DNA binding protein and its individual subunits in simian virus 40 DNA replication, J . Bid. Chem. 265,7693 - 7700. Lohman, T. M., Overman, L. B. & Datta, S. (1986) Salt-dependent changes in the DNA binding co-operativity of Escherichiu coli single-strand binding protein, J . Mol. Biol. 187, 603 -616. Tsurimoto, T. & Stillman, B. (1989) Multiple replication factors augment DNA synthesis by the two eukaryotic DNA polymerases, a and 6, EMBO J . 8, 3883 - 3889. Tsurimoto, T. & Stillman, B. (1991) Replication factors required for SV40 DNA replication in vitro. 11. Switching of DNA polymerase a and 6 during initiation of leading and lagging strand, J . Biol. Chem. 266, 1961- 1968. Yurov, Y. B. (1980) Rate of DNA replication fork movement within a single mammalian cell, J . Mol. Biol. 136, 339 - 342. Siegel, L. M. & Monty, K. J. (1966) Determination of molecular weight and frictional ratios of proteins in impure systems by use of gel filtration and density gradient centrifugation: Application to crude preparations of sulfite and hydroxylamine reductase, Biochim. Biophys. Acta 112, 346-362. Kim, C., Snyder, R. 0. & Wold, M. S. (1992) Binding properties ofreplication protein A from human and yeast cells, Mol. Cell. Bid. 12, 3050-3059.

33. Reckmann, B., Grosse, F., Urbanke, C., Frank, R., Blocker, H. & Krauss, G. (1985) Analysis of secondary structures in MI 3mp8 (+) single-stranded DNA by the pausing of DNA polymerase a, Eur. J . Biochem. 152,633-643. 34. Krauss, G., Sindermann, H., Schomburg, U. & Maass, G. (1981) Escherichiu coli single-stranded deoxyribonucleic acid binding protein: stability, specificity, and kinetics of complexes with oligonucleotides and deoxyribonucleic acids, Biochemistry 20, 5346 - 5352. 35. Greipel, J., Urbanke, C. & Maass, G. (1989) The ssDNA binding protein of Escherichiu coli: physiological properties and biological functions, in Topics in nucleic acid reseurch (Neidle, T. S., ed.) pp. 61 -86, Macmillan, London. 36. Roth, Y.-F. (1987) Eukaryotic primase, Eur. J . Biochem. 165, 473 - 481.

37. Grosse, F. & Krauss, G. (1985) The primase activity of DNA polymerase CI from calf thymus, J . Biol. Chem. 260, 18811888. 38. Collins, K. L. & Kelly, T. J. (1991) Effects of T antigen and replication protein A on the initiation of DNA synthesis by DNA polymerase a-primasc, Mol. Cell. Biol. 11,2108-2115. 39. Dornreiter, I., Erdile, L. F., Gilbert, I. U., von Winkler, D., Kelly, T. J. & Fanning, E. (1992) Interaction of DNA polymeraser-primase with cellular replication protein A and SV40 Tantigen, EMBO J . I I , 769 - 776. 40. Bollum, F. J. (1975) Mammalian DNA polymerases, Proxr. Nucleic Acids Res. Mol. Biol. 15, 109 - 144.