Skip to main content
NCBI home page
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2000 May 15;28(10):2034–2042. doi: 10.1093/nar/28.10.2034

Differential functional behavior of viral φ29, Nf and GA-1 SSB proteins

Irene Gascón 1, José M Lázaro 1, Margarita Salas 1,a
PMCID: PMC105360  PMID: 10773070

Abstract

DNA replication of φ29 and related phages takes place via a strand displacement mechanism, a process that generates large amounts of single-stranded DNA (ssDNA). Consequently, phage-encoded ssDNA-binding proteins (SSBs) are essential proteins during phage φ29-like DNA replication. In the present work we analyze the helix-destabilizing activity of the SSBs of φ29 and the related phages Nf and GA-1, their ability to eliminate non-productive binding of φ29 DNA polymerase to ssDNA and their stimulatory effect on replication by φ29 DNA polymerase in primed M13 ssDNA replication, a situation that resembles type II replicative intermediates that occur during φ29-like DNA replication. Significant differences have been appreciated in the functional behavior of the three SSBs. First, the GA-1 SSB is able to display helix-destabilizing activity and to stimulate dNTP incorporation by φ29 DNA polymerase in the M13 DNA replication assay, even at SSB concentrations at which the φ29 and Nf SSBs do not show any effect. On the other hand, the φ29 SSB is the only one of the three SSBs able to increase the replication rate of φ29 DNA polymerase in primed M13 ssDNA replication. From the fact that the φ29 SSB, but not the Nf SSB, stimulates the replication rate of Nf DNA polymerase we conclude that the different behaviors of the SSBs on stimulation of the replication rate of φ29 and Nf DNA polymerases is most likely due to formation of different nucleoprotein complexes of the SSBs with the ssDNA rather than to a specific interaction between the SSB and the corresponding DNA polymerase. A model that correlates the thermodynamic parameters that define SSB–ssDNA nucleoprotein complex formation with the functional stimulatory effect of the SSB on φ29-like DNA replication has been proposed.

INTRODUCTION

Non-specific DNA–protein interactions are involved in many fundamental processes in vivo (1). Single-stranded DNA-binding proteins (SSBs) destabilize double-stranded DNA (dsDNA) and bind without sequence specificity, but selectively and cooperatively, to single-stranded DNA (ssDNA) conferring a regular structure to it, which is recognized and exploited by a variety of enzymes involved in DNA replication, repair and recombination (2). Several cases have been reported of SSBs that stimulate their cognate DNA polymerase (2–12). This modulation may be explained by the ability of the SSB to bind to ssDNA, melting out double-stranded regions by contiguous cooperative binding, but also by direct protein–protein interactions with proteins of the same replication system (3,9,13–17).

The bacteriophage φ29 genome consists of a linear dsDNA, 19 285 bp long, with a terminal protein (TP) covalently linked to each 5′-end (18,19). Replication of the viral genome starts at either DNA end non-simultaneously by a protein priming mechanism (18,19) and, after sliding back (20) and a transition step (21), proceeds by strand displacement towards the other end. Two kinds of replicative intermediates (RI) have been observed during φ29 DNA replication by electron microscopy both in vivo (22) and in vitro (23). As φ29 DNA polymerase elongates the initiation product from both DNA ends, the non-template strand is displaced by the DNA polymerase as ssDNA, giving rise to type I RI. When two growing chains, running from opposite ends, collide and separate, two type II RI are generated, where strand displacement is no longer required (18,19).

φ29 protein p5 is the SSB protein active during φ29 DNA replication (24,25). It protects ssDNA against nuclease degradation and greatly stimulates dNTP incorporation during φ29 DNA replication in vitro at incubation times where replication in its absence levels off (24). The φ29 SSB does not affect either initiation (24) or the transition stage between TP-primed initiation and normal DNA elongation (I.Gascón, unpublished results) during φ29 DNA replication but it has been described as having helix-destabilizing activity. It has been proposed that cooperative binding of the φ29 SSB to the displaced ssDNA could reduce the energy demand required for the unwinding process coupled to DNA polymerization and this could account for a stimulatory effect on type I RI (26).

φ29-related phage contain linear dsDNA and TP and genome replication proceeds as in φ29 (19). The φ29-like genus of the family Podoviridae has been classified into three evolutionary branches, both by early (27) and recent studies based on comparisons of nucleotide and amino acid sequences of selected DNA regions and proteins (28). Thus, one branch consists of phages BS32, φ15, φ29 and PZA; the second branch is composed of phage B103, M2 and Nf; the third branch has phage GA-1 as its sole member (28).

In the present work, the SSBs of phage φ29, Nf and GA-1, which constitute a representative example of each of the three branches, have been purified and functionally characterized. The helix-destabilizing activity and the ability of the three SSBs to eliminate non-productive binding of φ29 DNA polymerase to ssDNA have been analyzed. Furthermore, the stimulatory effect of the three SSBs on φ29 DNA polymerase in primed M13 ssDNA replication, a situation that resembles type II RI that occur during viral DNA replication, has also been studied. Significant differences have been appreciated in their functional behavior. We propose the existence of a correlation between the thermodynamic parameters that define each SSB–ssDNA nucleoprotein complex and the differential stimulatory effect of the SSBs in the two types of RI generated during viral DNA replication of this family of phages.

MATERIALS AND METHODS

Nucleotides and DNA

Oligonucleotides for sequencing were obtained from Boehringer Ingelheim. [α-32P]dATP (3000 Ci/mmol) and [γ-32P]ATP (3000 Ci/mmol) were obtained from Amersham International plc. Unlabeled nucleotides and M13mp18 ssDNA were from Pharmacia Biotech. For the helix-destabilizing assay, M13mp18 ssDNA was hybridized to a [γ-32P]ATP 5′-labeled universal primer (17mer oligonucleotide) at a ratio 1.5:1, in the presence of 0.2 M NaCl, as described (26). For the M13 DNA replication assay, M13mp18 ssDNA was hybridized to the unlabeled universal primer at a ratio of 10:1, in the presence of 0.2 M NaCl, as described (29). Oligonucleotide sp1 (5′-GATCACAGTGAGTAC) was 5′-labeled with [γ-32P]ATP and T4 polynucleotide kinase and further purified by electrophoresis in 20% polyacrylamide gels in the presence of 8 M urea. The universal primer and sp1 were supplied by Isogen.

Proteins

Sequenase v.2.0 T7 DNA polymerase was from Amersham Life Science. T4 polynucleotide kinase was from Boehringer Mannheim. φ29 TP and DNA polymerase were overproduced in Escherichia coli and purified as described (30,31). The φ29 SSB was purified by differential ammonium sulfate fractionation of φ29-infected cellular extracts followed by phosphocellulose and DEAE–cellulose chromatography, essentially as described (26). The Nf and GA-1 SSBs were purified as described below.

Determination of the DNA sequences of the Nf and GA-1 ssb genes

Sequencing of gene 2 (González-Huici, unpublished results), gene 3 (32), gene 4 (33) and gene 6 (34) of bacteriophage Nf and of genes 2 and 3 (35), gene 4 (36) and gene 6 (34) of bacteriophage GA-1 suggest that Nf and GA-1 early genes lie in the same order as in φ29. Therefore, the nucleotide sequence of gene 5 of phages Nf and GA-1, which codes for their respective SSBs, was obtained by genomic sequencing and the amino acid sequence of the corresponding proteins was deduced. DNA sequencing of the corresponding genes was done using genomic DNA as template, oligonucleotides corresponding to the flanking regions of genes 4 and 6 as primers (R.Freire and J.A.Horcajadas, personal communication) and the dideoxynucleotide chain termination method (37) with Sequenase v.2.0 T7 DNA polymerase.

Sequence accession numbers

The sequences reported have been submitted to the EMBL Nucleotide Sequence Database under accession nos AJ244025 (Nf SSB) and AJ244026 (GA-1 SSB).

Purification of the Nf and GA-1 SSBs

The Nf and GA-1 SSBs were purified from Bacillus subtilis strains 110NA (38) and Bacillus sp. G1R (39) infected with phages Nf and GA-1, respectively. Cells were infected at a multiplicity of 5 and harvested 45 min after infection. They were disrupted by grinding with alumina (2:1 w/w) and resuspended in a buffer containing 50 mM Tris–HCl, pH 7.5, 1 mM EDTA, 7 mM β-mercaptoethanol, 5% glycerol (buffer A) and 0.8 M NaCl. DNA was removed by addition of polyethyleneimine to 0.3% after adjusting the absorbance at λ 260 nm to 120 U/ml and centrifugation for 15 min at 11 000 g. The supernatant was made 0.2 M NaCl with buffer A and the SSBs were recovered in the supernatant after centrifugation as above. The samples were precipitated with 70% ammonium sulfate and the Nf and GA-1 SSBs were washed with 45 and 50% ammonium sulfate, respectively. The samples were dissolved in buffer A and applied to the following columns: in the case of Nf SSB, DEAE–cellulose (elution in buffer A, 40 mM NaCl) and hydroxyapatite (elution in 250 mM Na3PO4, pH 7.5); in the case of GA-1 SSB, phosphocellulose (elution in buffer A, 60 mM NaCl), heparin–agarose (elution in buffer A, 0.1 M NaCl) and hydroxyapatite (elution in 250 mM Na3PO4, pH 7.5). Protein samples were precipitated with ammonium sulfate up to 60%, dissolved in buffer A containing 50% glycerol and dialyzed against the same buffer.

In all purification steps, proteins were followed by SDS–PAGE (10–20% acrylamide gradients). At the final purification step, the Nf and GA-1 SSBs were at least 99 and 95% homogeneous, respectively.

N-terminal sequencing of both proteins was performed in an Applied Biosystems 473A pulsed-liquid protein sequencer for 10 cycles. The N-terminal sequences of the purified proteins proved to be identical to those obtained from the nucleotide sequences (see below).

Protein concentration was determined both by the Lowry method (40) and by comparison with a known amount of BSA in polyacrylamide gel electrophoresis.

Helix-destabilizing assay

The incubation mixture contained, in 12.5 µl, 125 ng of primed M13mp18 ssDNA, 50 mM Tris–HCl, pH 7.5, 4% glycerol, 0.1 mg/ml BSA and the indicated amounts of φ29, Nf or GA-1 SSBs or the corresponding buffer. After 40 min at 37°C, reactions were stopped with 1.25 µl of 0.25% (w/v) bromophenol blue, 0.25% (w/v) xylene cyanol, 30% glycerol and 0.5% SDS. Samples were subjected to electrophoresis at 4°C in an 8% polyacrylamide gel containing 0.1% SDS. The gel was dried and autoradiographed.

ssDNA degradation assay

The preincubation mixture contained, in 12.5 µl, 50 mM Tris–HCl, pH 7.5, 10 mM MgCl2, 1 mM dithiothreitol, 4% glycerol, 0.1 mg/ml BSA, 17.5 ng of φ29 DNA polymerase and 0.25 µg of M13mp18 ssDNA. The indicated amounts of the φ29, Nf or GA-1 SSB, or the corresponding buffer, were added to the preincubation mixture and then 0.075 ng of the 5′-labeled sp1 oligonucleotide (15mer) was used as ssDNA substrate for the 3′→5′ exonucleolytic activity of φ29 DNA polymerase. Samples were incubated at 30°C for 75 s and quenched by adding 3 µl of sequencing gel loading buffer. Reactions were analyzed by electrophoresis in 20% polyacrylamide gels in the presence of 8 M urea and autoradiography.

Primed M13 DNA replication

The incubation mixture contained, in 25 µl, 250 ng of primed M13mp18 ssDNA, 50 mM Tris–HCl, pH 7.5, 1 mM dithiothreitol, 4% glycerol, 0.1 mg/ml BSA, 10 µM each dCTP, dGTP, dTTP and [α-32P]dATP (2 µCi), 12 mM MgCl2, the indicated amount of φ29 or Nf DNA polymerase and the φ29, Nf or GA-1 SSB, or the corresponding buffer. After incubation for the indicated times at 30°C, the reaction was stopped by adding EDTA up to 10 mM and SDS up to 0.1% and the samples were filtered through Sephadex G-50 spin columns in the presence of 0.1% SDS. Stimulation of dNTP incorporation was calculated by counting the Cèrenkov radiation corresponding to the excluded volume. To determine the DNA elongation rate, the DNA replication products of the samples taken at different times were denatured by treatment with 0.7 M NaOH and analyzed by alkaline agarose gel electrophoresis in the presence of size markers as described (41). After electrophoresis, the position of unit length M13mp18 DNA was detected by ethidium bromide staining and the gels were dried and autoradiographed. The size of the newly synthesized DNA was estimated after densitometric scanning of the autoradiograms.

RESULTS

Amino acid sequence comparison of the φ29, Nf and GA-1 SSBs

Amino acid sequences of the SSBs of phages Nf and GA-1 were determined by viral genomic sequencing as described in Materials and Methods and compared to that of the φ29 SSB (42). Figure 1 shows the amino acid sequence alignment of the three SSBs. The φ29 and Nf SSBs share 61.4% identity and 71.9% similarity; the φ29 and GA-1 SSBs 21% identity and 44.7% similarity; the Nf and GA-1 SSBs 18.4% identity and 46.5% similarity. The calculations were made omitting the N-termini of the proteins, where the GA-1 SSB contains an insert. These values are in good agreement with previous comparative studies of φ29-like phages (27,28) which showed that while phages belonging to groups A and B (φ29 and Nf, respectively) share many similarities, phage GA-1 (the sole member of group C) diverges from all other φ29-like phages.

Figure 1.

Figure 1

Open in a new tab

Sequence alignment of the φ29, Nf and GA-1 SSBs. Identical residues in the three SSBs are white on a black background. Similar residues of the φ29 and/or Nf SSBs with respect to the GA-1 SSB, the least related of them, are black on a grey background. Similarities between the φ29 and Nf SSBs are not highlighted. Numbers indicate the amino acid position.

Helix-destabilizing activity of the φ29, Nf and GA-1 SSBs

The φ29, Nf and GA-1 SSBs were purified as indicated in Materials and Methods and their helix-destabilizing activities were examined on a substrate consisting of a full-length M13 ssDNA molecule to which a 5′-labeled 17mer oligonucleotide had been hybridized, under conditions in which virtually all the labeled oligonucleotide remained annealed to the M13 ssDNA. This substrate was incubated with increasing amounts of either SSB or the corresponding buffer, as indicated in Materials and Methods, and the reaction products analyzed by native polyacrylamide gel electrophoresis. In agreement with previous results, the φ29 SSB was able to displace the labeled oligonucleotide from M13 ssDNA (26; see also Fig. 2). The Nf and GA-1 SSBs were also able to efficiently displace the labeled oligonucleotide from M13 ssDNA in a protein concentration-dependent fashion, as shown in Figure 2. However, a smaller amount of the GA-1 SSB than of the φ29 and Nf SSBs was required to obtain a similar unwinding extent (10 versus 40 µM). Under the test conditions, incubation in the absence of SSB did not result in release of the labeled oligonucleotide, indicating that the hybrid substrate was stable throughout the course of the experiment.

Figure 2.

Figure 2

Open in a new tab

Helix-destabilizing activity of the φ29, Nf and GA-1 SSBs. An M13 ssDNA molecule to which a 5′-labeled 17mer oligonucleotide was hybridized was incubated with increasing amounts of either the φ29, Nf or GA-1 SSB or the corresponding buffer. After 40 min at 37°C, reactions were stopped and fractionated on a polyacrylamide gel, as indicated in Materials and Methods. Positions of the hybrid substrate and the displaced oligonucleotide are indicated. C is the control of the heat-denatured substrate.

The φ29, Nf and GA-1 SSBs eliminate non-productive binding of φ29 DNA polymerase to ssDNA to different extents

φ29-like DNA replication takes place with the generation of high amounts of ssDNA (18,19). SSB binding to ssDNA could avoid or even eliminate non-productive binding of the viral DNA polymerase to ssDNA, directing it to the replication origins and thus increasing effective re-usage of the DNA template. To test this possibility, we analyzed the ability of the φ29, Nf and GA-1 SSBs to eliminate non-productive binding of φ29 DNA polymerase to ssDNA. Thus, we incubated φ29 DNA polymerase with a 5′-labeled oligonucleotide that can be degraded by the 3′→5′ exonucleolytic activity of φ29 DNA polymerase. When the enzyme was preincubated with M13 ssDNA, all DNA polymerases were trapped and no exonucleolytic degradation of the oligonucleotide could be detected (Fig. 3). Increasing amounts of each SSB or the corresponding buffer were added to the preincubation mixture containing φ29 DNA polymerase bound to M13 ssDNA and then the oligonucleotide was added, as indicated in Materials and Methods. As can be observed in Figure 3, addition of increasing amounts of φ29 SSB to the preincubation mixture enabled φ29 DNA polymerase to degrade the oligonucleotide up to the initial levels. This degradation was not due to any nuclease activity present in the φ29 SSB preparation. When increasing amounts of the Nf and GA-1 SSBs were added, there was also an increase in oligonucleotide degradation, although degradation levels were lower than those obtained in the absence of M13 ssDNA (Fig. 3). Addition of higher amounts of the Nf or GA-1 SSBs to the preincubation mixture exceeding those required to saturate all the ssDNA present resulted in the same oligonucleotide degradation level (not shown). Addition of a protein that does not bind ssDNA, like BSA, instead of the SSBs, did not produce an increase in degradation of the oligonucleotide (data not shown). Therefore, it can be concluded that the oligonucleotide degradation increase upon SSB addition is due to its ability to bind ssDNA, releasing the titration of φ29 DNA polymerase by the ssDNA. In any case, some specificity between the DNA polymerase and its own SSB cannot be ruled out at present, as indicated by the different oligonucleotide degradation levels by φ29 DNA polymerase upon addition of the φ29 or Nf and GA-1 SSBs.

Figure 3.

Figure 3

Open in a new tab

Ability of the φ29, Nf and GA-1 SSBs to prevent non-productive binding of φ29 DNA polymerase to ssDNA. The assays were carried out as described in Materials and Methods, using a 5′-labeled 15mer oligonucleotide as substrate for the 3′→5′ exonucleolytic activity of φ29 DNA polymerase. When the enzyme was preincubated with M13, all DNA polymerases were trapped and no degradation of the oligonucleotide was detected. The indicated amounts of the φ29, Nf or GA-1 SSB were added to the preincubation mixture of φ29 DNA polymerase and M13 ssDNA. After incubation for 75 s at 30°C, degradation of the labeled DNA was analyzed by electrophoresis in 20% polyacrylamide gels in the presence of 8 M urea and autoradiography. Control of each SSB alone, at the highest amount employed in the assay, is also shown in each last lane. Positions of the 15mer oligonucleotide and of the degradation products are indicated.

Effects of the φ29, Nf and GA-1 SSBs on primed M13 ssDNA replication by φ29 DNA polymerase

To determine the effect of the three SSBs on φ29 DNA polymerase in a type II RI replication-like situation in which strand displacement is not required, we carried out a replication assay with primed M13 ssDNA as template. After incubation with increasing amounts of each SSB, φ29 DNA polymerase was added to the preincubation mixture and reactions proceeded as indicated in Materials and Methods. In all three cases, the stimulation of dNTP incorporation by φ29 DNA polymerase as well as the increase in its replication rate were measured. Stimulation of dNTP incorporation was determined by counting the Cèrenkov radiation of the newly synthesized DNA, while the replication rate was calculated from the length of the DNA synthesized at different reaction times, estimated after densitometric scanning of alkaline agarose gels in the presence of size markers. The intensity of the radioactive bands was not relevant in the replication rate calculations.

The functional behavior displayed by the three SSBs in this replication assay was very different, as can be observed in the corresponding agarose gels (see Fig. 4A, B and C for the φ29, Nf and GA-1 SSBs, respectively). Graphs that represent the length of the synthesized DNA versus time at different SSB concentrations, together with tables of the corresponding replication rates are also shown (see Fig. 5A, B and C for the φ29, Nf and GA-1 SSBs, respectively). On the one hand, it has to be highlighted that the GA-1 SSB was able to stimulate dNTP incorporation by φ29 DNA polymerase even at the lowest SSB concentration assayed (3.25 µM), conditions under which the other two SSBs were unable to produce any stimulation of dNTP incorporation by the DNA polymerase (Fig 4). On the other hand, as can be observed in Figure 4A, the φ29 SSB was able to increase the replication rate of φ29 DNA polymerase, even at concentrations of SSB that produced hardly any stimulation of dNTP incorporation (3.5 and 7 µM; Fig. 5A). Addition of a concentration of SSB enough to cover all the ssDNA (28 µM) produced a 20-fold increase in dNTP incorporation and a 2-fold increase in the replication rate, calculated under conditions in which no strand displacement occurs. The increase in dNTP incorporation that is not due to stimulation of the replication rate is probably due to prevention of non-productive binding of φ29 DNA polymerase to ssDNA (see the ssDNA degradation experiment above). Neither the Nf nor GA-1 SSB was able to stimulate the replication rate of φ29 DNA polymerase, even at the highest concentrations assayed (Figs 4B and 5B for the Nf SSB, and 4C and 5C for the GA-1 SSB), enough to produce a 10-fold increase in dNTP incorporation due to the increase in the number of φ29 DNA polymerase molecules.

Figure 4.

Figure 4

Open in a new tab

Dose-dependent effect of the φ29 (A), Nf (B) and GA-1 (C) SSBs on φ29 DNA polymerase in primed M13 ssDNA replication. The assay was carried out as described in Materials and Methods using 25 ng of φ29 DNA polymerase and the indicated amounts of SSB or the corresponding buffer. After incubation at 30°C for the indicated times, dNTP incorporation was calculated by counting the Cèrenkov radiation of the synthesized DNA and the replication rate was measured by analysis of the replication products by alkaline 0.7% agarose gel electrophoresis and autoradiography in the presence of size markers. Two different expositions are shown for the highest protein concentration used in the case of φ29 and Nf SSBs. The size of the newly synthesized DNA was estimated after densitometric scanning of the autoradiograms. Positions of the molecular size markers are indicated.

Figure 5.

Figure 5

Open in a new tab

Graphs of the length of the synthesized DNA versus time and tables with the respective replication rates corresponding to the M13 DNA replication assays shown in Figure 4. (A) Stimulatory effect of increasing amounts of the φ29 SSB on φ29 DNA polymerase. Values corresponding to 0 (open circles), 3.5 (closed triangles), 7 (open triangles), 14 (open squares) and 28 µM (closed circles) φ29 SSB are shown. (B and C) Replication rate of φ29 DNA polymerase in the presence of increasing amounts of the Nf and GA-1 SSBs, respectively. Values corresponding to 0 (open circles) and the respective highest protein concentrations tested (closed circles) (25.2 and 13 µM) are shown, as intermediate concentrations did not display different behavior.

The specific stimulation of the replication rate of φ29 DNA polymerase by its own SSB could be due either to a specific stimulatory interaction between the two proteins or else to differences in the nucleoprotein complex formed by each SSB with ssDNA.

The φ29 SSB but not Nf SSB stimulates the replication rate of Nf DNA polymerase

As already shown, the φ29 SSB is the only SSB that stimulates the replication rate of φ29 DNA polymerase in primed M13 ssDNA replication. To determine if there is specific stimulation of the replication rate of the DNA polymerase by its cognate SSB, we studied the effect of the φ29 and Nf SSBs on the replication rate of Nf DNA polymerase. Figure 6 shows the effect of the φ29 and Nf SSBs on the replication rate of φ29 and Nf DNA polymerases in primed M13 ssDNA replication. Graphs of the length of the synthesized DNA versus time and tables with the corresponding replication rates are also displayed. In the absence of the SSB, the replication rate of both DNA polymerases was 1400 nt/min. The replication rates of φ29 DNA polymerase in the presence of the Nf and φ29 SSBs were 1500 and 2700 nt/min, respectively (Fig. 6A). The replication rates of Nf DNA polymerase in the presence of the Nf and φ29 SSBs were 1500 and 2400 nt/min, respectively (Fig. 6B). These results indicate that there is not specific stimulation of the replication rate of the viral DNA polymerase by its own SSB. In contrast, the differential stimulatory effect of each SSB on the replication rate of the DNA polymerases could be modulated by the thermodynamic parameters that define the nucleoprotein complex formed by each SSB with the ssDNA (43), as discussed below.

Figure 6.

Figure 6

Open in a new tab

Effect of the φ29 and Nf SSBs on the replication rate of φ29 and Nf DNA polymerases. The assay was carried out as described in Materials and Methods. Primed M13 ssDNA was incubated with 25 µM of the indicated SSB or the corresponding buffer. After addition of 25 ng of φ29 (A) or Nf (B) DNA polymerases, samples were incubated at 30°C for the indicated times and subjected to alkaline 0.7% agarose gel electrophoresis. Positions of the molecular size markers are indicated. Graphs of the length of the synthesized DNA versus time and the tables with the respective replication rates corresponding to both (A) and (B) are also shown. Values corresponding to the DNA synthesized in the absence (open circles) and in the presence of the φ29 (closed circles) and Nf (triangles) SSBs are displayed.

DISCUSSION

Since the first SSB was isolated (44), there has been great interest in elucidating the mechanism of action of these proteins. SSBs constitute essential components of the DNA replication machinery of φ29-related phages, as viral DNA replication takes place with the generation of large amounts of ssDNA (18,19). The genome of φ29-related phages consists of a linear dsDNA with a TP covalently linked to each 5′-end and replicates via a protein-primed initiation step catalyzed by the viral DNA polymerase, followed by elongation of the initiation product. As the DNA polymerase elongates the initiation product from both DNA ends, the non-template strand is displaced by the DNA polymerase as ssDNA, giving rise to type I RI (Fig. 7). When two growing chains, running from opposite ends, collide and separate, two type II RI are generated. The viral SSB forms a nucleoprotein complex with the displaced DNA strand in the type I RI, which has to dissociate ahead of the DNA polymerase in the type II RI (Fig. 7). The thermodynamic parameters that define this nucleoprotein complex are: K, the intrinsic binding constant, which represents the affinity of a single protein monomer for an isolated binding site on the nucleic acid lattice; ω, the cooperativity parameter, which specifies the relative affinity of an incoming ligand for a contiguous as opposed to an isolated binding site; Keff, the effective binding constant, which is the product of K and ω; n, the nucleotide binding site size, which is the number of nucleotides covered by a protein monomer (45,46).

Figure 7.

Figure 7

Open in a new tab

Association–dissociation equilibrium of SSB binding to DNA in type I and II RIs generated during φ29-like DNA replication. Primer TP is indicated in black and parental TP is shaded. DBP p6, dsDNA-binding protein; SSB p5, ssDNA-binding protein. Only the left DNA end has been drawn except for type I and type II molecules, where the two DNA ends are shown.

The φ29 SSB does not display any stimulatory effect in the early replication steps, but greatly stimulates φ29 DNA replication at incubation times where replication in its absence levels off (24). To further elucidate the SSB mechanism of action in the φ29-like replication system, a comparative functional characterization of the SSBs of φ29 and of the related phages Nf and GA-1 has been carried out. Functional characterization of the three SSBs has focused on their stimulatory effect in primed M13 ssDNA replication, a situation that resembles type II RI that occur during φ29-like DNA replication, as well as on their helix-destabilizing activity and ability to eliminate non-productive binding of DNA polymerase to ssDNA. In the present study, we provide evidence that binding of the SSB to ssDNA not only prevents non-productive binding of the viral DNA polymerase to ssDNA, as demonstrated by the increase in dNTP incorporation in primed M13 ssDNA replication assays, but can also release DNA polymerase molecules that are already titrated by the ssDNA, as shown in the oligonucleotide degradation assay. This effect would be of particular importance in φ29-like DNA replication systems, where large amounts of ssDNA are generated (18) and SSB binding to ssDNA could favor efficient re-usage of templates.

Additionally, significant differences in the functional behavior of the three SSBs have been appreciated. First, the GA-1 SSB is able to display helix-destabilizing activity as well as to stimulate dNTP incorporation by φ29 DNA polymerase in the M13 DNA replication assay, even at SSB concentrations at which the φ29 and Nf SSBs do not show these effects (3.25 µM). This correlates with the different behaviors of the three SSBs in gel retardation assays and fluorescence measurements (43): a smaller concentration of the GA-1 SSB than of the other two SSBs is required to saturate a similar amount of ssDNA, due to the higher global affinity (Keff) for ssDNA and larger binding site size (n) of the former.

Another significant difference in the functional behavior of the three SSBs is that the φ29 SSB, but not the Nf or GA-1 SSB, increases the replication rate of φ29 DNA polymerase in primed M13 DNA replication, even at very low protein concentrations that produce hardly any stimulation of dNTP incorporation (3.5–7 µM). The stimulation of the replication rate of the DNA polymerase at protein concentrations not high enough to stimulate dNTP incorporation could be due to the formation of patches of φ29 SSB that straighten the ssDNA in front of the DNA polymerase, favoring its advance. At φ29 SSB concentrations high enough to saturate the ssDNA present in the assay (28 µM) the increase in dNTP incorporation by the DNA polymerase is due both to the increase in the replication rate as well as to the prevention of non-productive binding of DNA polymerase molecules to the ssDNA. Several SSBs have been reported to stimulate their cognate DNA polymerases during DNA replication (3,9,13–17). In systems where a direct interaction between the SSB and the DNA polymerase has not been detected, the stimulatory effect has been proposed to be due to the ability of the SSB to melt hairpin structures that block progression of the DNA polymerase. However, the Nf and GA-1 SSBs, which also bind ssDNA and melt secondary structures, as demonstrated by their helix-destabilizing ability, do not stimulate the replication rate of φ29 DNA polymerase. Furthermore, stimulation of the replication rate of Nf DNA polymerase by φ29 SSB but not by its own SSB rules out the hypothesis of a specific interaction between the DNA polymerase and its own SSB and strengthens the importance of the kind of nucleoprotein complex formed by each SSB with ssDNA. Based on our results, we raise the hypothesis that the highly dynamic dissociation of the φ29 SSB–ssDNA nucleoprotein complex ahead of the DNA polymerase could be responsible for specific stimulation of the replication rate of the DNA polymerase by the φ29 SSB in type II RI-like DNA replication (Fig. 7). In fact, the φ29 SSB displays a lower K than the Nf and GA-1 SSBs (43).

This functional comparative study of three related SSBs that participate in the same replication process has allowed us to suggest the existence of a correlation between the binding parameters that define the SSB–ssDNA nucleoprotein complex and the different steps of viral DNA replication. Thus, cooperative binding of the SSB to the displaced ssDNA in type I RI would likely be governed by the global affinity of the protein for ssDNA (Keff), while a low intrinsic binding constant (K) of the SSB–ssDNA nucleoprotein complex ahead of the DNA polymerase would probably favor its advance in type II RI (Fig. 7). Therefore, nucleoprotein complex formation with type I RI would be more easily accomplished by the GA-1 SSB due to its higher Keff and n, while SSB–ssDNA complex dissociation by the advancing DNA polymerase in type II RI would be more favorable in the case of the φ29 SSB due to its lower K. In fact, the three SSBs display very different stimulatory effects on φ29 TP–DNA amplification in vitro (43). Thus, at a protein concentration of 3 µM, the GA-1 SSB stimulates dNTP incorporation by φ29 DNA polymerase 12-fold, while the φ29 and Nf SSBs produce 1.8- and 1.1-fold stimulation of dNTP incorporation, respectively. Furthermore, stimulation of dNTP incorporation under the test conditions reached a plateau at a protein concentration of 6 µM (25-fold stimulation) in the case of the GA-1 SSB and 14 µM in the case of the φ29 SSB (15- to 20-fold stimulation), while at 15 µM Nf SSB concentration only a 7.2-fold stimulation was obtained.

Amino acid sequence comparisons of different SSBs show little detectable similarity. However, resolution of the crystal structures of very different SSBs (47–53) suggests that these proteins have striking similarities at the structural level, with their DNA-binding domain composed of an OB (oligonucleotide/oligosaccharide binding) fold (54). Our work shows that the φ29, Nf and GA-1 SSBs share a high degree of amino acid sequence similarity, but in spite of this, they have differential stimulatory effects in the same replication system. This is probably due to the different structural complexes they form with ssDNA and supports the idea that it is both the protein structure and the thermodynamic parameters that define the ssDNA–SSB nucleoprotein complex, rather than amino acid sequence similarities, that have to be taken into account to understand the function of an SSB.

Acknowledgments

ACKNOWLEDGEMENTS

We are grateful to M. Elías-Arnanz, J. Saturno and M. de Vega for their help, to L. Blanco for his orientation in the amino acid sequence alignments and to the Protein Sequencing Unit of the Centro de Biología Molecular ‘Severo Ochoa’. This investigation was aided by research grant 2RO1 GM27242-20 from the National Institutes of Health, by grant PB93-0173 from the Dirección General de Investigación Científica y Técnica, by grant ERBFMRX CT970115 from the European Union and by an institutional grant from the Fundación Ramón Areces. I.G. was holder of a pre-doctoral fellowship from the Fundación Ramón Areces.

DDBJ/EMBL/GenBank accession nos AJ244025, AJ244026

REFERENCES

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES