Purification and Functional Characterization of Adenovirus tsll1A DNA-binding Protein FLUORESCENCE STUDIES OF PROTEIN-NUCLEIC ACID BINDING*

The adenovirus single-stranded DNA (ssDNA)-bind-ing protein (DBP) is necessary for the elongation step in viral DNA replication. In an attempt to characterize the putative ssDNA-binding domain of the DBP, we purified and characterized the Ad2tslllA DBP, which contains a glycine-to-valine substitution at amino acid 280. This mutation is adjacent to that in the previously studied Ad2+NDlts23. Ad2+NDlts23 exhibits a tem-perature-sensitive defect in DNA replication, and its DBP has previously been shown to bind ssDNA with reduced affinity. at addition the Reactions 0.5 20% The acid-insoluble incorporation of radioactive label was measured Cerenkov counting. FluorescenceSpectroscopy of DBP-Nucleic Acid Znteraction-Titra-tions were monitored by observing the increase of the fluorescence of poly(l,N’)-ethenoadenylic acid (poly(reA)) as the result of binding by DBP. These experiments were performed by adding aliquots of protein to a 0.69 pM solution of nucleic acid in a starting sample volume of 2 ml. In some experiments, the “salt back-titration” pro- cedure was performed, in which case saturation of a fixed amount of poly(rtA) by the addition of DBP under tight binding conditions was followed by adding a concentrated NaCl solution to dissociate the DBP-nucleic acid complex. Dissociation of the complex was moni- tored by measuring the resulting decrease in poly(reA) fluorescence. Tris-HCI (10 mM), pH 8.0, Na2EDTA (0.1 mM) was used as the buffer in all experiments, with the salt concentration as specified for indi- vidual experiments. Poly(rfA)

The adenovirus single-stranded DNA (ssDNA)-binding protein (DBP) is necessary for the elongation step in viral DNA replication.
In an attempt to characterize the putative ssDNA-binding domain of the DBP, we purified and characterized the Ad2tslllA DBP, which contains a glycine-to-valine substitution at amino acid 280. This mutation is adjacent to that in the previously studied Ad2+NDlts23. Ad2+NDlts23 exhibits a temperature-sensitive defect in DNA replication, and its DBP has previously been shown to bind ssDNA with reduced affinity. Ad2tslllA DBP, like Ad2+NDlts23, does not support adenovirus DNA replication in vitro at elevated temperatures. However, the Ad2tslllA DBP binds ssDNA more tightly than does Ad2+NDlts23 and is not temperature sensitive in this function.
To determine the nucleic acid-binding properties of DBP, we applied spectrofluorometric techniques, which had not been used previously to study adenovirus DBP. Using the homopolynucleotide poly(l,iV')ethenoadenylic acid (poly(rcA)), we have determined that the binding site size is approximately 16 nucleotides. In 20 mM NaCl, the Ad2wt, AdatslllA, and Ad2+ND 1 ts23 DBP proteins all bound stoichiometrically to poly(rcA) with overall apparent affinities above 10' M-'. Based on titrations carried out at higher salt concentrations, however, the stability of these complexes did appear to increase in the order Ad2+NDlts23 < Ad2tslllA < Ad2wt. By these techniques, we have confirmed also that the DBP of another temperature-sensitive mutant, H5ts107, like the Ad2tslllA DBP, retains its ability to bind ssDNA even at a restrictive temperature utilizing the salt concentration compatible with adenovirus DNA replication in vitro. The H5ts107 DBP, which contains an amino acid substitution at position 413, is defective for in vitro replication at nonpermissive temperature but is not temperature sensitive for binding to ssDNA. ). The costs of publication of this article were defrayed in part bv the payment of page charges. This article must therefore be-hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The adenovirus (Ad)' single-stranded (ss) DNA-binding protein (DBP) is a multi-functional protein that is required for 1) viral DNA replication (van der Vliet et al., 1975;Horwitz, 1978;Friefeld et al., 1983), 2) regulation of viral gene expression at both transcriptional (Carter and Blanton, 1978;Nevins and Winkler, 1980) and post-transcriptional levels (Babich and Nevins, 1981), 3) determination of host range for viral infectivity (Klessig and Grodzicker, 1979), and 4) viral assembly (Nicolas et al., 1983). DBP was isolated initially from adenovirus-infected cells based upon its ability to bind to ssDNA (van der Vliet and Levine, 1973). Studies of the roles of DBP in viral DNA synthesis have been facilitated by the development of in vitro DNA replication systems (Horwitz, 1978;Challberg and Kelly, 1979;Friefeld et al., 1983), which have allowed a more detailed analysis of the individual protein requirements at each of the different stages of the replicative reaction. In vitro analysis of temperature-sensitive DBP mutants has provided genetic and biochemical evidence that DBP is essential for elongation of nascent adenovirus DNA but not for initiation of replication (Kaplan et al., 1979;Friefeld et al., 1983).
Although DBP migrates on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) with a mobility that corresponds to a 72,000-dalton protein, it is a 529-amino acid protein which has a molecular mass of 59,049 daltons (Kruijer et al., 1981(Kruijer et al., , 1982. In an attempt to localize the functions of DBP relative to its structure, proteolytic fragments of DBP have been purified and characterized (Klein et al., 1979;Schechter et al., 1980;Linne and Philipson, 1980;Ariga et al., 1980;Friefeld et al., 1983;Krevolin and Horwitz, 1987). Limited proteolysis with a variety of enzymes has indicated that the DBP is composed of two relatively discrete structural  -sensitive  DBPs, infected  cells were incubated  for 5 days  at 32 "C prior to harvesting;  for Ad2 wild-type  DBP, infected  cells  were incubated for 40-48 h at 37 "C. Ad2 wild-type and H5ts107 DBP were purified by a previously described method (Schechter et al., 1980), which was modified as detailed below. Because it was observed that a significant quantity of wild-type DBP eluted from the ssDNAcellulose when the column was washed with buffer containing 0.5 M NaCl, a 10.column volume (150 ml) 0.2-0.5 M NaCl gradient was used instead.
Subsequently, the bulk of the DBP was eluted with 2.0 M NaCl and 0.01 M Tris-HCl, pH 8.0, in 3-ml fractions. Fractions containing the DBP were identified by SDS-PAGE (Laemmli, 1970); nrotein bands were visualized bv staining with Coomassie Blue R-250. I ".
terminal and carboxyl-terminal domains consist of approximately 173-and 356-amino acid residues, respectively (Tsernoglou et al., 1985). Chymotryptic digestion produces nested 44-and 34-kDa fragments, representing the carboxyl-terminal domain; each fragment is capable of binding DNA and is active in the in uitro complementation of adenovirus DNA replication (Ariga et al., 1980). The isolation of conditional mutants and second site revertants mapping within the carboxyl-terminal region has been important in establishing the relationships between the structure and function of DBP. All mutations in DBP which affect viral DNA replication have been localized to the carboxyl region. The localization of these mutations to discrete clusters within this domain has suggested sites which may be integrally involved in the DNA replication process. H5ts125 and H5ts107 are independently isolated replication-defective mutants which have identical serine-for-proline substitutions at amino acid residue 413 of the DBP (Kruijer et al., 1983). The H5ts125 DBP has been reported to exhibit decreased binding to ssDNA-cellulose, but this altered phenotype was elicited at a relatively high salt concentration (0.25 M) (van der Vliet et al., 1975). Analysis of H5ts107 has suggested that the capacity to bind DNA is retained at nonpermissive temperatures at the salt concentration (20 mM) used during the in vitro synthesis reaction (Krevolin and Horwitz, 1987). However, the complementing activity of the H5ts107 and H5ts125 DBPs for in vitro replication is unequivocally defective at elevated temperatures.
This result suggested that the replication defect in H5tsI07 is not the result of impaired DNA binding by the mutant DBP.
Recently, studies describing mutations located toward the NHP-end of the carboxyl domain have been reported. Purification and biochemical characterization of the Ad2+NDlts23 DBP (substitution of phenylalanine for leucine at amino acid 282 (Kruijer et al., 1982)) suggested a linkage in this mutant between its defective DNA replication activity and its reduced affinity to bind single-stranded DNA even at permissive temperature (Prelich and Stillman, 1986). The defective DNA binding of the Ad2+NDlts23 DBP has resulted in speculation that the mutation is localized to the binding site. Another DBP mutant, AdBtslllA, maps only 2 amino acid residues away from that in Ad2+NDlts23 and contains a valine substitution for glycine at position 280, as demonstrated by nucleotide sequencing (Prelich and Stillman, 1986 were kept at 32-33 "C (permissive temperature) for 5-6 days (Horwitz, 1978). Ad2, H5tsld7, and AdPtslllA virions were nurified as ureviouslv described (Horwitz. 1978;Maize1 et al., 1968;Green and Pina, 1964 and 10% alvcerol) containina either 0.5 M NaCl or no added NaCl.' Concentrated DBP was stored at -70 "C, which prevented most of the rapid degradation found at 4 "C, even in 2 M NaCl.
Protein concentrations were determined by amino acid analysis on a Beckman 6300 analyzer. In Vitro Assay for Adenovirus-specific DNA Replication-The in vitro "end fragment" assay for origin-specific adenovirus DNA replication has been described previously (Horwitz and Ariga, 1981 temperature for 30 min with-the reaction components, excluding the template-primer and adenovirus pol. The poly(dT) (0.33 pg) and oligo(dA) (0.33 pg) were incubated at 65 "C for 10 min, then added to the above reaction mixture, and allowed to incubate at the reaction temperature for 15 min. The 60-min reaction was initiated by addition of the adenovirus pol. Reactions were terminated by the addition of 0.2 ml of chilled sodium pyrophosphate (0.2 M), 150 pg each of bovine -y-globulin and yeast tRNA, and 0.5 ml of 20% trichloroacetic acid. The acid-insoluble incorporation of radioactive label was measured by Cerenkov or liquid scintillation counting.
FluorescenceSpectroscopy of DBP-Nucleic Acid Znteraction-Titrations were monitored by observing the increase of the fluorescence of poly(l,N')-ethenoadenylic acid (poly(reA)) as the result of binding by DBP. These experiments were performed by adding aliquots of protein to a 0.69 pM solution of nucleic acid in a starting sample volume of 2 ml. In some experiments, the "salt back-titration" procedure was performed, in which case saturation of a fixed amount of poly(rtA) by the addition of DBP under tight binding conditions was followed by adding a concentrated NaCl solution to dissociate the DBP-nucleic acid complex. Dissociation of the complex was monitored by measuring the resulting decrease in poly(reA) fluorescence. Tris-HCI (10 mM), pH 8.0, Na2EDTA (0.1 mM) was used as the buffer in all experiments, with the salt concentration as specified for individual experiments. Poly(rfA) was purchased from Pharmacia LKB Biotechnology Inc. Fluorescence of poly(reA) was monitored using the SLM 8000 and SLM 8000C (SLM Urbana, IL) spectrofluorometers with an excitation wavelength of 309 nm and an emission wavelength of 406 nm. The SLM 8000C was controlled and data acquisition processed by a user-defined language executed on an IBM PC XT. In each fluorescence titration, the protein concentration was kept below 20 rg/ml in order to prevent aggregation (van der Vliet et al., 1978). For the calculation of binding site size, the experimental data were analyzed by a nonlinear least-squares technique (Johnson and Frasier. 1985) implemented on a MicroVax II assuming an e25, of 3.7 x lo3 M'l/moi of phosphate for the poly(rcA) (Karpel et al., i987;Ledneva et al., 1978).
Binding parameters were estimated by visually comparing the experimental binding isotherms with theoretical curves that were generated by the use of equation 15 in McGhee and von Hippel, 1974. Typically, a constant value of the cooperativity parameter, w, was assumed and the intrinsic affinity &, was allowed to vary until a best fit was obtained. With this constant value of Ki.,, the cooperativity parameter was then varied until this parameter was also optimized. The overall apparent affinity, Z&e, was then calculated as the product of the Ki,, and w.
Regardless of which parameter was being varied, only data points in the range from 0 to 50% maximal fluorescence enhancement were considered in determining the "best fit" as recommended by Kowalczykowski et al. (1981) and Newport et al. (1981).

Purification of Ad2tslllA
DBP-When the Ad2'NDlts23 DBP was purified without modifying the procedure described by Schecter et al. (1980), it eluted from the ssDNA-cellulose column in the 0.5 M NaCl wash and was contaminated by other proteins (Prelich and Stillman, 1986). Because of the possibility that the AdBtslllA DBP might also elute with the 0.5 M wash and not be sufficiently purified, the DBP was loaded onto the ssDNA cellulose column in 0.1 M NaCl and resolved by elution with a 20-column volume, 0.1-1.0 M NaCl linear gradient. As shown in Fig. 1, the elution of Ad2tslllA DBP peaked at approximately 0.6 M NaCl and was well separated from contaminants which eluted at lower salt concentrations. By densitometry of SDS-PAGE, DBP fractions eluting at greater than 0.50 M NaCl were judged to be greater than 99% pure with respect to protein.
Using conditions similar to those described above for the purification of AdBtslllA DBP, we observed that Ad2 wildtype and H5ts107 DBPs eluted from ssDNA cellulose at NaCl concentrations similar to AdZtslllA. However, when we purified Ad2+NDlts23 DBP by the same protocol, peak elution occurred at 0.2 M NaCl, and the protein had almost totally The DBP-containing pool eluted from the phosphocellulose column was loaded onto the ssDNA-cellulose column in a buffer containing 0.1 M NaCl. The proteins were resolved by elution with a 0.1-1.0 M NaCl gradient. Fractions were analyzed by 12% SDS-PAGE, the gels were stained by Coomassie Blue R-250 and each gel lane scanned densitometrically. The amount of DBP in a fraction is expressed in terms of the area of each densitometric peak, and this value is plotted against the concentration of NaCl of the fraction as determined by measurement of conductivity. The assay for specific synthesis of adenovirus DNA on the terminal fragments of Ad35 DNA-pro (B and G) was performed as described under "Materials and Methods." As described, DBP was preincubated at the specified reaction temperature for 60 min, prior to initiation of the reaction by the addition of the SmaI-cleaved adenovirus DNA-pro. Letters on the left side of the panel indicate the positions of the restriction fragments of Ad35 on the gel. Lanes l-4 present reactions incubated at 30 "C, whereas lanes 5-8 present reactions incubated at 38.5 "C. In lanes I and 5, DBP was omitted from the reactions. Lanes 2 and 6 contained Ad2 wild-type DBP; lanes 3 andf 7, AdZtslllA DBP; lanes 4 and 8, H5ts107 DBP. eluted by 0.5 M NaCl (data not shown), confirming the results of Prelich and Stillman (1986). These results suggested that Ad2tslllA DBP binds with a higher affinity to singlestranded DNA than does Ad2+NDlts23 DBP, even though the mutations in these two DBPs are located only 2 amino acid residues apart. Actiuity of Ad2tslllA DBP in Adenouirus-specific DNA Replication in Vitro-The end fragment assay which measures adenovirus DNA synthesis proceeding from specific initiation sequences at both ends of the molecule was used to study the activity of Ad2tslllA DBP. The results are presented in Fig.  2. As expected, the reaction was dependent on exogenous DBP (lanes 1 and 5). Purified AdPtslllA DBP was active in the elongation assay at 30 "C (lane 3) but was inactive at 38.5 "C (lane 7). In contrast, Ad2 wild-type DBP was active at both temperatures (lanes 2 and 6). Purified H5ts107 DBP, like Ad2tslllA DBP, was inactive at the nonpermissive tem-perature (lane 8). The temperature-sensitive phenotypes of the H&107 and Ad2tslllA DBPs were manifest only if the DBP was preincubated for 30-60 min prior to initiation of the reaction and were not reversed when incubation at nonpermissive temperature was followed by equilibration and incubation of the reactions at permissive temperature. Intragenie complementation between Ad2tslllA DBP (amino acid 282) and H5ts107 DBP (amino acid residue 413) was attempted at nonpermissive temperature, but the defect in adenovirus DNA replication was not corrected. In contrast, at permissive temperature, identical mixtures of Ad2tslllA and H5ts107 DBPs were able to replicate adenovirus DNA effectively (data not shown). Neither AdBtslllA DBP nor H5ts107 DBP was inhibitory to Ad2wt DBP function at elevated temperature. Stimulation of DNA Synthesis on Poly(dT). Oligo(dA) by Temperature-sensitive DBPs Is Temperature Sensitive-A recently developed assay has proven useful in studying the isolated role of DBP in the elongation of DNA by the adenovirus polymerase (Field et al., 1984). This assay, which utilizes poly(dT) . oligo(dA) as a primed template, requires only the adenovirus polymerase and the DBP as protein components; the reaction does not require the pTP or any of the nuclear factors from uninfected cells. Using this template, the DBP greatly stimulates DNA synthesis activity of its cognate adenovirui polymerase, suggesting that DNA synthesis on poly(dT) .oligo(dA) requires a specific, functional interaction between adenovirus DBP and adenovirus pol.
We have utilized this assay to further characterize the activities of the AdPtslllA and H5ts107 DBPs. Representative results are shown in Table I. Although each of the temperature-sensitive DBPs was able to stimulate the activity of the adenovirus polymerase at 30 "C, neither AdPtslllA or H5ts107 DBP was active at 38.5 "C. Similarly, the Ad2+NDlts23 DBP, which was active in this assay at 30 "C, was inactive at 38.5 "C (data not shown).
Partial Proteolysis of Ad2tslllA DBP-Partial proteolysis of the wild-type DBP by chymotrypsin generates relatively stable, nested 44-and 34-kDa polypeptides corresponding to the carboxyl-terminus of the protein. The primary aminoterminal peptide (27 kDa) is degraded under these conditions and does not appear as a band on the gel. Other proteases generate similar sized fragments, suggesting a bi-domain tertiary structure for wild-type DBP. Partial chymotryptic proteolysis of AdBtslllA DBP was performed at permissive and nonpermissive temperatures to determine whether the altered activity of the mutant protein might be the result of conformational changes affecting the global two-domain structure of the DBP. As shown in Fig. 3, similar peptide fragments are produced at both temperatures from the wild-type (lanes 3 and 5) and Ad2tslllA (lanes 4 and 6) DBPs indicating that major conformational changes resulting in the accessibility of other potential chymotryptic sites for Ad2tslllA DBP do not occur at nonpermissive temperatures. However, by increasing the protease/DBP ratio (data not shown) or by extending the reaction time, we were able to examine the intrinsic stability of the 34-kDa carboxyl-terminal domains of the various DBPs. When the proteolytic reaction was extended to 60 min at 30 "C, the 34-kDa proteolytic fragments from Ad2tslllA DBP (Fig. 3, lane 8) and from wild-type DBP (Fig. 3, lane 7) were approximately equally stable. However, at 37 "C, the major chymotryptic fragment (34 kDa) of AdBtslllA DBP is almost completely degraded (lane IO), while that of the wildtype DBP (lane 9) remains at a level similar to that observed at the lower temperature (lane 7). Similarly, the 34-kDa proteolytic fragment of the H5ts107 DBP was observed to be more susceptible to proteolysis than the corresponding wildtype fragment at nonpermissive temperature (data not shown). The increased susceptibilities of the 34-kDa proteolytic fragments of the Ad2tslllA and H5ts107 DBPs to further proteolysis at 37 "C indicate that at nonpermissive temperature the carboxyl-terminal domains of these temperature-sensitive DBPs are either destabilized or exist in different, more protease-sensitive configurations than the corresponding region of the wild-type DBP.
Fluorescence Studies-Fluorescence spectroscopy has been used to study the binding of other ssDNA-binding proteins to nucleic acids. Such studies can be used to define the binding site size (n), expressed in terms of nucleotides occluded/DBP molecule, the apparent binding constant (Kapp) which is equal to the product of the intrinsic binding constant Ki":,,t, and the cooperativity parameter w. The binding of protein to nucleic acid can be monitored either by measuring the quenching of the intrinsic fluorescence of the protein or by measuring the enhancement of fluorescence of a fluorescent nucleic acid. In experiments in which the quenching of the inherent fluorescence of the wild-type Ad DBP was utilized as a measure of adenovirus DBP-poly(dT) binding, we found that the protein was unusually photosensitive. The extensive bleaching of control protein samples in the absence of nucleic acid upon repeated excitation at 282 nm made the titration curves erratic and difficult to reproduce reliably (data not shown). Therefore, we utilized the enhancement of fluorescence of poly(rtA) induced by DBP binding to investigate the interaction of DBP with nucleic acid. The use of poly(rtA) permitted excitation at a longer wavelength where considerably less photobleaching occurred. Our initial experiments indicated that when poly(rtA) was saturated with wild-type DBP, the fluo-  (Kuil et al., 1989). Titrations at 20 mM NaCl were performed for another important reason; the in uitro replication assays used to characterize the DBP are typically performed in the presence of 20 mM NaCl. Activity in these assays is sensitive to increased salt concentration, with virtually no DNA synthesis above 40 mM NaCl.
The titrations for each DBP at higher NaCl concentrations, at which binding was weaker, were utilized to more accurately determine the nucleic acid-binding affinities of the various DBPs (Kowalczykowski et al., 1986). The values for Kapp of the respective DBPs in 100 mM NaCl at 25 "C are compiled in Table II and indicate that increasing the [NaCl] from 20 to 100 mM results in at least a 4-40-fold decrease in Kapp.
Sets of titrations at varied NaCl concentrations were performed to determine the effects of NaCl concentration on the nucleic acid-binding of the various DBPs. The results of a characteristic series of titrations with Ad2tslllA DBP are displayed in Fig. 4A. At increasing concentrations of NaCl, the titration curves are progressively shifted to the right indicating that Ad2tslllA binds poly(rcA) less tightly at higher NaCl concentrations.
The relationship between NaCl concentration and K,, for Ad2tslllA DBP is displayed in Fig. 4B in the form of a log-log plot. The slope of this plot, as well as similar plots for wild-type and H5ts107 DBPs, is reported in Table II. The increase in salt sensitivity at 25 "C on going from the Ad2wt to either the AdZtslllA or H5ts107 DBP proteins is not significant. At each salt concentration (from-20 to 300 mM), 0.69 FM poly(reA) was titrated with AdZtslllA DBP at 25 "C. and binding parameters were obtained from curves as described under "Materials and Methods." The theoretical binding curves (solid lines aboue) were generated using a site size of 16 nucleotides and the following values of the overall apparent affinity, I&,, and the cooperativity parameter, w: 20 mM NaCI, I&, = 7.0 X 10s M-', w = 7; 50 mM NaCl, I&,, = 5.2 X lo7 M-', w = 13; 100 mM NaCl, I&, = 1.9 X 10' M-l, w = 9; 300 mM NaCl, Kapp = 2.3 X lo6 M-', w = 45. B, effect of NaCl concentration on Kapp for binding of Ad2tslllA DBP to poly(reA) derived from data presented in titration curves of A.
An alternative method for studying the dependence of binding upon salt concentration is the salt back-titration procedure in which the dissociation of the protein-poly(rtA) complexes is measured as a function of salt concentration.
As in the titrations reported above at low salt, tight-binding conditions, a fixed quantity of poly(reA) is added to a cuvette and is titrated by addition of aliquots of DBP until the increase in poly(rtA) fluorescence reaches the plateau indicating saturation. Then, aliquots of concentrated NaCl (5 M) are added stepwise, and the decreases in fluorescence due to dissociation of the protein-nucleic acid complex are monitored until poly(rcA) fluorescence reaches the minimum value, indicating that the polynucleotide exists only in its free form. The curves for such a series of salt titrations comparing wild-type, Ad2tslllA, and Ad2+NDlts23 DBPs are presented in Fig. 5. Ad2+NDlts23 DBP was observed to dissociate from singlestranded polynucleotide at a lower salt concentration than AdBtslllA DBP, a finding which is qualitatively consistent with results obtained by ssDNA-cellulose chromatography. Fluorescence spectroscopy was also utilized to characterize the effects of elevated temperature upon the abilities of Ad2tslllA, H5ts107, and wild-type DBPs to bind to singlestranded nucleic acid in 20 mM NaCl. In Fig. 6  ' These data are the average of two or three determinations. The cooperativity factor w is expressed as a range of values calculated from these titrations. The values of w presented in the remainder of the table were determined from single titrations. However, these values of w would also be expected to vary 2-h-fold. Comparison of Ad2'NDlts23, Ad2tslllA, and Ad2wt DBPs. After saturation of the poly(rcA) lattice with the specified DBP in 20 mM NaCl at 25 "C, 5 M NaCl was added in stepwise fashion and the decrease in fluorescence indicative of dissociation of the protein-nucleic acid complexes was monitored.

DISCUSSION
AdBtslll is a nitrous-acid induced mutant (Martin et al., 1978), which contains two independent mutations, one of which is a temperature-sensitive DNA replication defect which maps to the DBP (D'Halluin et al., 1984;Stillman et al., 1984). The DBP mutation was separated from the second mutation, which maps to the  protein DBP and is more similar to wild-type DBP or H5ts107 DBP in its binding parameters.
The DNA synthesis assay utilizing a poly(dT).oligo(dA) template-primer for adenovirus pol has allowed for the further characterization of the role of DBP in DNA replication. Unlike the end fragment assay, in which elongation proceeds from an obligate adenovirus-specific initiation event and for which at least five proteins are required, the synthesis of poly(dA) from the poly(dT) .oligo (dA) template-primer by the adenovirus polymerase is dependent only on the presence of DBP (Field et al., 1984). Studies using this assay suggest that the DBP functions not only by binding single-stranded DNA but also by directly and specifically interacting with and stimulating the cognate adenovirus polymerase. DBP does not stimulate the activity of HeLa DNA polymerase (Y. Moreover, the E. coli ssDNA-binding protein (SSB) alone does not stimulate the adenovirus pol in this reaction; but, saturating levels of SSB do increase the degree of stimulation produced by low levels of DBP (Lindenbaum et al., 1986 PoIy(rcA) at 0.69 fiM in 20 mM NaCl was titrated with the specified DBP at 25 or at 42 "C. Toppanel, AdPtslllA DBP. Middkp&zel, Ad2 wild-type DBP.
Bottom Danel. H5ts107 DBP. The theoretical curves (sdolid lines) 1 were generated as described under "Materials and Methods" using a site s&e of 16 nucleotides and for the 42 "C titration, apparent affinities, Kepp, and cooperativity parameters, w, given in Table II. In the case of the 25 "C titrations the theoretical curves were generated with the following Kapp and w: tsllla, K,,,, = 1.8 X lOa M-', w = 12; AdBwt, Kepp = 7.0 X lOa M-', w = 35; and ts107, K,,, = 2.0 X 10' M-l, w = 22. sorption measurements, The site sizes calculated for the AdBtslllA DBP and the H5ts107 DBP did not differ significantly from that calculated for the wild-type protein.
Binding of DBP to poly(rtA) at 20 mM NaCl was nearly stoichiometric, and as a result, only minimum estimates of the apparent binding constant Kapp could be obtained at this salt concentration.
The values obtained for the K,,, values of the wild-type, H5ts107, Ad2tslllA, and Ad2+NDlts23 DBP proteins were all above 10' M-' in 20 mM NaCl. In 100 mM NaCl, binding was less tight, and binding constants could be more reliably calculated. The KBpp for the wild-type DBP was 1.0 X 10' M-l, whereas those for AdPtslllA and H5ts107 DBPs were 3.0 X lo7 M-' and 2.1 X lo7 M-', respectively, in 100 mM NaCl. Thus, in terms of binding affinity at permissive temperature, the Ad2tslllA and H5ts107 DBPs differ by less than a factor of five from the wild-type in 100 mM NaCl. When AdBtslllA DBP was preincubated at elevated temperature, the shape of the binding curve was not altered, indicating that the affinity of binding was not decreased. However, the absolute increase in poly(rcA) fluorescence at saturation was only 150% as compared with about 250% for the wild-type DBP titrations (data not shown), suggesting that the temperature-sensitive DBP had imposed a slightly different, probably less extended, conformation upon the poly(rtA) at elevated temperature. A series of titration curves was generated at varying salt concentrations for each of the DBPs to compare the salt dependence of the Kapp values for poly(rtA).
The slopes of the resulting log Kapp uersus log [NaCl] curves ranged from -1.02 to -2.34. These experiments demonstrate that binding of the DBP is significantly less dependent upon salt concentration than is the binding of gp32, for which the corresponding slope is equal to -7 (Newport et al., 1981). Thus, a lo-fold increase in salt concentration would be expected to decrease the affinity of gp32 for ssDNA by seven orders of magnitude as opposed to only about 1 to 2 orders of magnitude in the cases of the DBP proteins examined.
From the results presented above, it is apparent that the temperature-sensitive replication defects of AdPtslllA and H5ts107 DBPs cannot be readily attributed to their nucleic acid-binding affinities, which do not differ significantly from those of the wild-type protein at 42 "C under conditions necessary for adenovirus replication in uitro. This result confirms and extends previous studies on H5ts107 utilizing different experimental techniques for measuring DNA-binding (Krevolin and Horwitz, 1987). The Ad2'NDlts23 DBP, with its phenylalanine for leucine substitution at amino acid 282, had been reported to be defective for DNA binding because it eluted from a ssDNA-cellulose column at lower NaCl concentration than the wild-type (Prelich and Stillman, 1986), and we have reproduced this result. The results of our spectrofluorometry experiments investigating the dissociation of the DBP . poly(reA) complexes as a function of salt concentration confirm by a second method that the binding of Ad+NDlts23 DBP to single-stranded polynucleotide is weaker than that of Ad2tslllA DBP, as well as the wild-type DBP. In summary, the focus of the work reported here has been the delineation of structure-function relationships of the DBP for single-stranded nucleic acid binding and complementation of DNA replication.
We have shown that the H5ts107 and Ad2tslllA DBP mutant proteins, which are defective in replication assays in uitro, are not temperature sensitive for binding under the same conditions. These results reinforce the postulate that the ability of DBP to bind single-stranded DNA is by itself not sufficient for DNA replication. The results of the limited proteolysis studies probably signal small but significant changes in the conformation of the COOHterminal domains of Ad2tslllA and H5ts107 DBP at nonpermissive temperature.
Therefore, it appears that the amino acid substitution in AdBtslllA, as well as that in H5ts107, affects the protein's activity by decreasing the thermal stability of the protein structure rather than by directly disrupting the nucleic acid binding. The conformational changes that are induced in the Ad2ts107 and H5tslllA proteins at restrictive temperatures are not sufficient to substantially lower the affinity of these proteins for single-stranded nucleic acids. However, they may disrupt a presumed DBP:DNA polymerase interaction or perhaps alter the topology of the bound nucleic acid. In fact, the decreased extent of enhancement of the poly(rcA) fluorescence caused by binding of the AdPtslllA DBP at nonpermissive temperature suggests that the