The herpes simplex virus type-1 single-strand DNA-binding protein, ICP8, increases the processivity of the UL9 protein DNA helicase.

Herpes simplex virus type-1 UL9 protein is a sequence-specific DNA-binding protein that recognizes elements in the viral origins of DNA replication and possesses DNA helicase activity. It forms an essential complex with its cognate single-strand DNA-binding protein, ICP8. The DNA helicase activity of the UL9 protein is greatly stimulated as a consequence of this interaction. A complex of these two proteins is thought to be responsible for unwinding the viral origins of DNA replication. The aim of this study was to identify the mechanism by which ICP8 stimulates the translocation of the UL9 protein along DNA. The data show that the association of the UL9 protein with DNA substrate is slow and that its dissociation from the DNA substrate is fast, suggesting that it is nonprocessive. ICP8 caused maximal stimulation of DNA unwinding activity at equimolar UL9 protein concentrations, indicating that the active species is a complex that contains UL9 protein and ICP8 in 1:1 ratio. ICP8 prevented dissociation of UL9 protein from the DNA substrate, suggesting that it increases its processivity. ICP8 specifically stimulated the DNA-dependent ATPase activity of the UL9 protein with DNA cofactors that allow translocation of UL9 protein and those with secondary structure. These data suggest that UL9 protein and ICP8 form a specific complex that translocates along DNA. Within this complex, ICP8 tethers the UL9 protein to the DNA substrate, thereby preventing its dissociation, and participates directly in the assimilation and stabilization of the unwound DNA strand, thus facilitating translocation of the complex through regions of duplex DNA.

The origin-binding protein (UL9 protein) is a 94-kDa protein that recognizes specific elements in the HSV-1 origins of DNA replication (4,5). The UL9 protein also possesses intrinsic DNA helicase activity, and associated nucleoside triphosphatase (ATPase) activity, that is presumably required for unwinding the origins of DNA replication (6 -9). The HSV-1 SSB, henceforth referred to as ICP8 (infected cell polypeptide 8), is a 128-kDa protein, capable of binding single-stranded DNA (ssDNA) cooperatively and with high affinity (10). ICP8 forms a specific complex with the UL9 protein by interacting with its extreme C terminus (11)(12)(13). In addition, ICP8 has been shown to interact with the HSV-1 DNA polymerase and helicaseprimase (14 -18). The ability of ICP8 to participate in multiple protein-protein interactions suggests that it fulfills several roles during viral DNA replication.
The interaction between ICP8 and UL9 protein greatly stimulates the rate and extent of DNA unwinding catalyzed by the UL9 protein, enabling it to unwind long stretches of DNA (9,19). It has been shown that disruption of the ICP8-UL9 protein complex by deletion of the 27 C-terminal amino acid of the UL9 protein greatly reduces origin-specific DNA replication (12). Presumably, the ICP8-UL9 protein complex promotes efficient unwinding of the HSV-1 origins of DNA replication (20,21).
This study examines the mechanism by which ICP8 stimulates the DNA helicase and ATPase activities of the UL9 protein. The results show that ICP8 increases the processivity of the UL9 protein, facilitating its translocation along DNA and through regions of secondary structure.
Proteins-T4 polynucleotide kinase was obtained from Boehringer Mannheim and Promega. Bovine serum albumin (DNase-free) was purchased from Pharmacia Biotech Inc. Rabbit muscle L-lactic dehydrogenase and pyruvate kinase, as solutions in 50% glycerol, were obtained from Sigma. Escherichia coli SSB (E-SSB) was purchased from U. S. Biochemical Corp. E-SSB concentrations are expressed in moles of tetrameric protein.
ICP8 and UL9 protein were purified from Spodoptera frugiperda Sf21 cells infected with Autographa californica nuclear polyhedrosis virus recombinant for the HSV-1 UL29 and UL9 genes, respectively (11). The concentrations of ICP8 and UL9 protein were determined by using extinction coefficients of 82,720 and 89,220 M Ϫ1 cm Ϫ1 at 280 nm, respectively. ICP8 and UL9 protein concentrations are expressed in moles of monomeric protein.
ATPase assays were performed essentially as described (17). The hydrolysis of ATP to ADP and P i was coupled to the oxidation of NADH and was measured by a decrease in the absorbance at 340 nm as a function of time. Unless otherwise stated, reactions were performed at 37°C and contained 20 mM EPPS-NaOH, pH 8 software. Absorbance readings were taken at 1-s intervals. Rates of ATP hydrolysis were calculated by converting the absorbance change at 340 nm to moles of ATP hydrolyzed using an extinction coefficient of 6,220 M Ϫ1 cm Ϫ1 for NADH at 340 nm. Rate measurements in the absence or presence of ICP8 or E-SSB were taken during the same time interval for each reaction.

RESULTS
Effects of Preincubation and ICP8 on the DNA Helicase Activity of the UL9 Protein-Consistent with previously published results (9), the time course of UL9 protein DNA unwinding shows an appreciable lag in the reaction (Fig. 1A). This lag period was eliminated upon preincubation of UL9 protein with DNA substrate (Fig. 1B). These results suggest that UL9 protein needs to assemble on the DNA substrate and that this association is slow and possibly rate-limiting.
Addition of ICP8 to an ongoing DNA helicase reaction led to rapid stimulation of the rate of DNA unwinding (Fig. 1A). Maximal DNA unwinding was observed 10 min after the addition of ICP8. Similarly, rapid stimulation of DNA unwinding was observed when UL9 protein and ICP8 were preincubated with DNA substrate (Fig. 1B). However, when UL9 protein was added to ICP8-coated DNA substrate, there was a slower rate of reaction (Fig. 1B). Preincubation of UL9 protein and ICP8 in the absence of DNA led to an intermediate level of DNA unwinding (Fig. 1B), suggesting that their association with the DNA substrate was rate-limiting or that the ICP8-UL9 protein complex is unstable. Collectively, these data indicate that ICP8 exerts its maximal effect once UL9 protein has assembled on the DNA substrate. Fig. 2 shows that ICP8 stimulated the DNA helicase activity of the UL9 protein on both the M13:100-mer and 38:53-mer DNA substrates. Moreover, it shows that optimal stimulation of DNA unwinding with the M13:100-mer DNA substrate occurred at an equimolar ratio of UL9 protein to ICP8 ( Fig. 2A). The concentration of ICP8 (150 nM) required for maximal stimulation is ϳ3-fold lower than would be required to coat the DNA substrate (assuming a site size of ϳ15 nucleotides of winding activity (Fig. 3A). Furthermore, preincubation of the UL9 protein with DNA substrate to allow formation of a UL9 protein-DNA complex followed by addition of competitor DNA did not prevent competition (Fig. 3B). The ability of competitor DNA to reduce DNA unwinding activity, regardless of the prior formation of a UL9 protein-DNA complex, suggests that the UL9 protein DNA helicase is nonprocessive and that dissociation of the UL9 protein from the DNA substrate is fast.
To examine the effect of ICP8 on the processivity of the UL9 protein DNA helicase, competition experiments were performed in the presence of ICP8. Fig. 4A shows that addition of coating concentrations of ICP8 (1.0 M) immediately after com-petitor DNA almost eliminated the challenge with M13 mp18 ssDNA since near maximum stimulation of DNA unwinding was observed. In contrast, equimolar concentrations of ICP8 (175 nM) added immediately after competitor DNA did not elicit maximal stimulation of DNA unwinding activity. Presumably, equimolar concentrations of ICP8 stimulated and stabilized UL9 protein that had not dissociated from the DNA substrate, whereas coating concentrations of ICP8 largely prevented dissociation of UL9 protein and stabilized its interaction with the DNA substrate.
Addition of equimolar (175 nM) or coating (1.0 M) concentrations of ICP8 immediately prior to competitor DNA elimi- nated the challenge with M13 mp18 ssDNA since maximum rates of DNA unwinding were observed (Fig. 4B). Since the 30-s interval between addition of ICP8 and competitor DNA is insufficient to allow the reaction to go to completion (see Fig. 1, and Refs. 9 and 12), this result suggests that ICP8 prevents dissociation of the UL9 protein rather than representing rapid and maximum stimulation of DNA unwinding activity. Furthermore, in experiments with the 38:53-mer DNA substrate, subsaturating concentrations of ICP8 (1.14 M, 0.4 molar eq of ICP8-binding sites) stimulated DNA unwinding and prevented dissociation of UL9 protein from the DNA substrate when added immediately prior to M13 mp18 ssDNA competitor (3fold molar excess of ICP8-binding sites over 38:53-mer DNA substrate) (Fig. 4C). There was no effect of ICP8 on DNA unwinding with either DNA substrate in the absence of UL9 protein. These data suggest that ICP8 stabilizes the interaction of the UL9 protein with the DNA substrate and increases its processivity.
Effect of ICP8 on the ATPase Activity of the UL9 Protein- Fig. 5 shows that ICP8 had no effect on the DNA-independent ATPase activity of the UL9 protein, suggesting that ICP8 does not directly affect the active site of the UL9 protein.
Consistent with previously published results (8), ICP8 specifically stimulated the DNA-dependent ATPase activity of the UL9 protein. Fig. 6 shows that subsaturating concentrations of ICP8 (500 nM, 75% coating concentration) stimulated the rate of ATP hydrolysis 5.4-fold, whereas equivalent concentrations of E-SSB (215 nM) (assuming a site size of 35 nucleotides of ssDNA/tetramer of protein; Ref. 23) resulted in inhibition of activity. Consistent with the observation made for the DNA helicase activity, excess concentrations of ICP8 (1.0 M, 150% coating concentration) reduced the stimulatory effect. Likewise, E-SSB at 150% coating concentration (430 nM) further inhibited the rate of reaction. These results indicate that SSBsaturated DNA is less accessible as a cofactor to the UL9 protein. In the case of ICP8, specific protein-protein interactions presumably allow continued access of the UL9 protein to the DNA even at saturating ICP8 concentrations.
Addition of 500 nM ICP8 to a reaction containing 215 nM E-SSB led to partial reversal of the inhibitory effect of E-SSB, providing further evidence for the ability of ICP8 to specifically stimulate the DNA-dependent ATPase activity of the UL9 protein. In the reciprocal experiment, addition of 215 nM E-SSB to a reaction containing 500 nM ICP8 led to stimulation of activity but only to the level observed with 1.0 M ICP8 (150% coating concentration), confirming that the stimulatory effect is dependent on the extent to which the DNA cofactor is coated with SSB.

Stimulation of the DNA-dependent ATPase Activity of the UL9 Protein by ICP8 Is Proportional to DNA Length and Is
Affected by the Secondary Structure Content of the DNA Cofactor-To examine the effect of ICP8 on the processivity of the UL9 protein, ATPase assays were performed with (dT) polymers of varying length (Fig. 7A). These DNA cofactors lack secondary structure and therefore allow unobstructed translocation of the UL9 protein along the DNA. Measurements on the effect of DNA length on the DNA-dependent ATPase activity of the UL9 protein showed that the minimum DNA length required to elicit activity is 14 nucleotides with activity increasing up to ϳ60 nucleotides (8). The data in Fig. 7A show that (dT) 15 was sufficient to elicit the DNA-dependent ATPase activity of the UL9 protein. However, there was a significantly higher rate of ATP hydrolysis with (dT) 20 , (dT) 60 , and poly(dT) as cofactors, presumably due to the ability of the UL9 protein to bind and translocate along the longer DNA polymers while only capable of binding to (dT) 15 . Addition of ICP8 to a reaction with (dT) 15 did not stimulate the rate of ATP hydrolysis. Similarly, ICP8 did not stimulate the rate of ATP hydrolysis with a 17-mer oligodeoxyribonucleotide (data not shown). In contrast, there was detectable stimulation of activity by ICP8 with (dT) 20 and extensive stimulation with (dT) 60 and poly(dT) (also see Fig. 8). The ability of ICP8 to stimulate ATP hydrolysis with DNA cofactors that allow translocation of UL9 protein suggests that ICP8 confers a higher degree of processivity upon the UL9 protein, presumably by preventing its dissociation from the DNA substrate. Fig. 7B shows the effect of ICP8 on the ATPase activity of the UL9 protein with DNA cofactors that contain secondary structure. Consistent with previously published results (8) bonucleotide) that allow translocation of the UL9 protein.
Addition of ICP8 to a reaction with double-stranded activated calf thymus DNA did not stimulate the rate of ATP hydrolysis. In contrast, the rates of ATP hydrolysis with heat-denatured activated calf thymus DNA, M13 mp18 ssDNA, and 60-mer hairpin oligodeoxyribonucleotide were greatly stimulated by ICP8. These results suggest that ICP8 not only stimulates ATP hydrolysis by increasing the processivity of the UL9 protein but also by eliminating secondary structure in the DNA, thereby allowing the UL9 protein to efficiently translocate along the DNA substrate and through regions of duplex DNA.
The ability of ICP8 to stimulate the ATPase activity of the UL9 protein in a manner that is proportional to the length of the DNA cofactor is shown in Fig. 8. The data show that increasing concentrations of ICP8 had no effect on the rate of ATP hydrolysis with (dT) 15 . In contrast, there was a 4-fold FIG. 7. Effect of ICP8 on the ATPase activity of the UL9 protein in the presence of various DNA cofactors. Reactions were performed as described under "Experimental Procedures" with 10 M (nucleotide) of the indicated DNA cofactors. The arrows indicate the points at which UL9 protein and ICP8 were added. A, effect of ICP8 with (dT) polymers. Thickest trace, (dT) 15 ; thinnest trace, (dT) 20 ; second thickest trace, (dT) 60 ; second thinnest trace, poly(dT). B, effect of ICP8 with DNA cofactors that contain secondary structure. Thickest trace, activated calf thymus DNA; thinnest trace, heat-denatured activated calf thymus DNA; second thickest trace, M13 mp18 ssDNA; second thinnest trace, 60-mer hairpin oligodeoxyribonucleotide. stimulation in the rate of reaction with (dT) 20 , and a 5-fold stimulation with (dT) 60 . Similarly, increasing concentrations of ICP8 led to stimulation of ATP hydrolysis with poly(dT) (data not shown). Consistent with the data shown in Fig. 7B, there was even greater stimulation (6.5-fold) in the rate of ATP hydrolysis with the 60-mer hairpin oligodeoxyribonucleotide. In this case, 100 nM ICP8 (15% coating concentration and equimolar to UL9 protein concentration) was sufficient to stimulate ATP hydrolysis almost 5-fold, presumably due to the formation of a specific ICP8-UL9 protein complex that efficiently translocated along the DNA and disrupted the hairpin (duplex) region of the DNA substrate.

DISCUSSION
The aim of this study was to identify the mechanism by which the HSV-1 SSB, ICP8, stimulates the DNA helicase and DNA-dependent ATPase activities of the HSV-1 origin-binding protein (UL9 protein).
The results show that subsaturating concentrations of ICP8 were sufficient to stimulate the DNA helicase activity of the UL9 protein, and that maximal stimulation occurred at an ICP8 to UL9 protein ratio of 1:1. This observation is consistent with the existence of a specific ICP8-UL9 protein complex that translocates along the DNA and is active during DNA unwinding. In addition, the stoichiometry of 1:1 inferred from the functional interaction between ICP8 and UL9 protein is identical to that observed for the physical complex (13).
ICP8 also stimulated the DNA-dependent ATPase activity of the UL9 protein in a species-specific manner. A heterologous SSB, E-SSB, led to inhibition of activity, presumably by preventing access of the UL9 protein to the DNA substrate. Furthermore, ICP8 partially reversed the inhibitory effect of E-SSB. These observations are indicative of a specific tertiary complex that consists of UL9 protein, ICP8, and ssDNA.
Interestingly, the stimulatory effect of ICP8 on the DNA helicase and DNA-dependent ATPase activities of the UL9 protein was dependent on the ICP8/nucleotide ratio. Maximal stimulation was observed at subsaturating concentrations of ICP8, whereas coating concentrations produced less of an effect. Assuming that the level of stimulation correlates with the physical association of ICP8 and UL9 protein, these results suggest that the ICP8-UL9 protein complex is less stable at high ICP8/nucleotide ratios. This finding may explain why Gustafsson et al. (13) failed to detect a physical complex of UL9 protein, ICP8, and ssDNA at high ICP8/nucleotide ratios. It is possible that, when ICP8 is in excess and all ssDNA sites are occupied, ICP8 undergoes a conformational change and loses its affinity for UL9 protein.
The existence of a lag in the initial phase of the DNA unwinding reaction catalyzed by the UL9 protein suggests that the rate-limiting step is its association with the DNA substrate. This conclusion is substantiated by the observation that the lag was eliminated by preincubating the UL9 protein and DNA substrate. Moreover, rapid and maximal stimulation of DNA unwinding by ICP8 was only observed when UL9 protein was preincubated with DNA substrate or with ongoing DNA helicase reactions. Consequently, ICP8 exerts its maximal effect on a preassembled UL9 protein-DNA complex.
The observation that ICP8 had no effect on the DNA-independent ATPase activity of the UL9 protein suggests that it does not directly affect the catalytic site of the UL9 protein.
Therefore, experiments were designed to examine how ICP8 influences the interaction of the UL9 protein with its DNA substrate during DNA unwinding and DNA-dependent ATP hydrolysis. Both activities entail translocation of the UL9 protein along ssDNA and through regions of duplex DNA. Consequently, these experiments addressed how ICP8 affects the translocation of UL9 protein along DNA.
Addition of excess challenger DNA to ongoing DNA unwinding reactions showed that UL9 protein was effectively competed from the DNA substrate, implying that it dissociates rapidly from the DNA substrate and that it is nonprocessive. The inhibitory effect of the competitor DNA was eliminated by ICP8. Specifically, equimolar, subsaturating concentrations of ICP8 added immediately prior to competitor DNA prevented the effect of the challenger DNA. These results suggest that ICP8 prevents the dissociation of UL9 protein from the DNA substrate and therefore increases its processivity. Further evidence for the processivity enhancing function of ICP8 is provided by previous experiments in which ICP8 enabled the UL9 protein to unwind long regions of DNA (up to ϳ3 kilobase pairs) (9,19).
(dT) 15 is sufficient to elicit the DNA-dependent ATPase activity of the UL9 protein (8). Presumably, this cofactor is sufficiently long to allow the UL9 protein to bind but is too short to allow translocation of the UL9 protein along the DNA. The extent of ATP hydrolysis seen with this cofactor should therefore be representative of UL9 protein binding only. Consistent with the assumption that ICP8 increases the processivity of the UL9 protein, there was no effect of ICP8 on the rate of ATP hydrolysis with (dT) 15 . In contrast, ATP hydrolysis with longer DNA cofactors, which allow both binding and translocation of the UL9 protein, was greatly stimulated by ICP8. Furthermore, ICP8 had an even greater stimulatory effect on ATP hydrolysis with DNA cofactors that contain secondary structure.
In conclusion, the ability of ICP8 to prevent competition with challenger DNA during DNA unwinding, and to stimulate ATP hydrolysis with DNA cofactors that allow translocation, implies that it prevents dissociation of the UL9 protein from the DNA substrate and increases its processivity. In addition, ICP8 appears to facilitate translocation of the UL9 protein through regions of duplex DNA. This is probably a manifestation of the helix-destabilizing activity of ICP8 (24). Fig. 9 shows a model of how ICP8 stimulates the translocation of the UL9 protein along ssDNA and through regions of duplex DNA. Both molecules within the UL9 protein dimer (6,7,19,20) make contact with ssDNA via their N-terminal ssDNA-binding regions (25), enabling it to translocate 3Ј to 5Ј (7,9). One molecule of ICP8 is bound to the C terminus of each UL9 protein molecule (11,12), allowing one ICP8 molecule to bind the template strand, tethering the UL9 protein to the DNA substrate, while the second molecule of ICP8 assimilates and stabilizes the unwound primer strand, thereby facilitating the translocation of the UL9 protein through regions of duplex DNA.
Previous studies have shown that a site-specific cisplatin lesion impairs the DNA helicase activity of the UL9 protein (22). Furthermore, it was shown that ICP8 relieved the inhibitory effect imposed by the lesion. Based on the findings in this study, it is likely that ICP8 enables the UL9 protein to bypass the lesion by tethering it to the DNA substrate, thereby preventing its dissociation.
The effects of cognate and noncognate SSBs have been documented for numerous DNA helicases (reviewed in Refs. 26 -28). E-SSB has been shown to affect the activities of E. coli PriA (29,30), DNA helicase IV (31), and Rep (32, 33) DNA helicases. Likewise, replication protein-A (RP-A) has been shown to affect several eukaryotic DNA helicases including: Saccharomyces cerevisiae Hel B (HCSB) (34,35), calf thymus DNA helicases A-D and F (36 -38), human DNA helicases isolated from HeLa cells (39,40), and simian virus 40 large T antigen (41)(42)(43). In HSV-1, apart from the interaction between ICP8 and the UL9 protein discussed in this report, it has also been shown that ICP8 can specifically stimulate the DNA helicase activity of the DNA helicase-primase (17,18,44). In most cases, SSBs have been shown to enhance DNA unwinding by preventing nonproductive binding of the DNA helicase to the DNA substrate. Specific protein-protein interactions between DNA helicases and cognate SSBs have also been shown to increase the length of DNA unwound. However, no previous studies have addressed the exact mechanism by which an SSB stimulates DNA unwinding.
Previous studies have indicated the importance of the ICP8-UL9 protein interaction for HSV-1 origin-specific DNA replication (12,21). The properties of the UL9 protein and ICP8 and those of the ICP8-UL9 protein complex, described in this and previous studies (7,9,10,20,21,24), illustrate how this complex is suited for its predicted role of recognizing and unwinding the HSV-1 origins of DNA replication.