DNA Polymerase X From African Swine Fever Virus: Quantitative Analysis of the Enzyme–ssDNA Interactions and the Functional Structure of the Complex

doi:10.1016/j.jmb.2005.10.061

J. Mol. Biol. (2006) 356, 121–141

DNA Polymerase X From African Swine Fever Virus: Quantitative Analysis of the Enzyme–ssDNA Interactions and the Functional Structure of the Complex Maria J. Jezewska, Agnieszka Marcinowicz, Aaron L. Lucius and Wlodzimierz Bujalowski* Department of Human Biological Chemistry and Genetics, Department of Obstetrics and Gynecology the Sealy Center for Structural Biology, Sealy Center for Cancer Cell Biology, The University of Texas Medical Branch at Galveston, 301 University Boulevard, Galveston, TX 77555-1053, USA

Interactions of polymerase X from African swine fever virus with singlestranded DNA (ssDNA) have been studied, using quantitative fluorescence titration and analytical ultracentrifugation techniques. Experiments were performed with a fluorescent etheno-derivative of ssDNA oligomers. Studies of unmodified ssDNA oligomers were carried out using the competition titration method. The total site-size of the pol X–ssDNA complex is 16(G1) nucleotide residues. The large total ssDNA-binding site has a complex heterogeneous structure. It contains the proper ssDNAbinding site that encompasses only 7(G1) residues. As the length of the ssDNA increases, the enzyme engages an additional binding area in interactions with the DNA, at a distance of w7–8 nucleotides from the proper site, which is located asymmetrically within the polymerase molecule. As a result, the net ion release accompanying the interactions with the DNA, increases from w1 for the proper DNA-binding site to w6 for the total DNA-binding site. Unlike in the case of the mammalian polymerase b that belongs to the same polymerase X family, the DNAbinding areas within the total DNA-binding site of pol X are not autonomous. Consequently, the enzyme does not form different binding modes with different numbers of occluded nucleotide residues, although the interacting areas are structurally separated. The statistical thermodynamic model that accounts for the engagement of the proper and the total DNA-binding site in interactions with the DNA provides an excellent description of the binding process. Pol X binds the ssDNA without detectable cooperativity and with very modest base specificity. q 2005 Elsevier Ltd. All rights reserved.

*Corresponding author

Keywords: polymerases; DNA replication; protein–DNA interactions; motor proteins; fluorescence titrations

Introduction The African swine fever virus (ASFV) is the etiological agent responsible for a highly lethal disease of domestic pigs.1–7 The DNA genome of the virus encodes two DNA polymerases, the virus Abbreviations used: ASFV, African swine fever virus; ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; 3A, etheno-adenosine; MCT method, macromolecular competition titration method; BER, base excision repair; pol, polymerase. E-mail address of the corresponding author: [email protected]

replicative polymerase, belonging to the eukaryotic family B enzymes and another polymerase (pol), a member of the pol X family referred to as ASFV pol X.1,6,7 The pol X family comprises several polymerases with different and specialized functions in the cell.6–9 The most intensively studied enzyme of this family is the mammalian pol b, an enzyme of w39 kDa, which plays a very specialized function in cell DNA repair machinery.9–21 pol b shows a typical polymerase fold, a thumb, a palm, and fingers, due to its resemblance to the human hand.10 pol b contains an additional, N-terminal 8 kDa domain, not present in the DNA replicative polymerases. The domain is the primary DNA-

0022-2836/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

122 binding site of the enzyme with a significant degree of functional autonomy and spatial separation from the rest of the enzyme molecule.15–21 On the other hand, NMR studies of ASFV pol X revealed a structure very different from pol b and other DNA replicative or repair polymerases.22,23 The enzyme is built only of the palm domain, which includes the first 105 amino acid residues from the N terminus of the protein and the C-terminal domain, built of the remaining 69 amino acid residues, as depicted in Figure 1.22 With a molecular mass of w20,000, ASFV pol X is the smallest, currently known, DNA polymerase. The palm domain contains a triad of invariant aspartate residues involved in divalent cation binding and the catalysis of the DNA synthesis by the DNA polymerases.22,23 Moreover, the palm domain has a subdomain fold, unique for ASFV pol X, that comprises a positively charged helix, aC, positioned at the surface of the domain. Similarly, the C-terminal domain contains a highly positively charged helix, aE, which does not have its counterpart in pol b.10,22,23 Due to these structural features and qualitative NMR data, both the palm and the C-terminal domain were implicated in DNA binding, although the structure of the co-complex has not been determined.22,23

Figure 1. Three-dimensional structure of ASFV pol X obtained by NMR studies.22,23 The palm domain, the first 105 amino acid residues from the N terminus of the protein, and the C-terminal domain, the remaining 69 amino acid residues, are marked in green and turquoise, respectively. Also, the lysine residues, 59, 60, and 63 of the palm domain, and the lysine residues 131, 132, and 133 contained in the aC helix, the aE helix of the palm and the C-terminal domain, respectively, are selected and marked in red and blue, respectively.

ASFV pol X–ssDNA Interactions

Besides the primary structure homology with the members of the polymerase X family, ASFV pol X shows striking functional similarities to the b-type polymerases that include template-directed DNA synthesis, distributive DNA synthesis on templateprimer DNA substrates, increased processivity on the gapped DNAs with the ssDNA gap shorter than four nucleotide residues, and particularly efficient filling of the single-nucleotide gaps.6,7,24 Such a spectrum of functional activities strongly suggests that ASFV pol X is involved in the repair processes of the viral DNA. This notion is further supported by the fact that the ASFV genome codes for another strictly replicative DNA polymerase, belonging to the B family of polymerases (see above), and several enzymes that perform most of the base excision repair (BER) pathway.4–7 In other words, pol X is a part of the ASFV DNA repair apparatus geared to repair the viral DNA, which has been damaged by the host reaction to the viral infection.6,7,22,23 Interactions of a polymerase with the DNA play a vital role in the functioning of the enzyme.25–27 Moreover, they are increasingly recognized as one of the major elements, that determine the degree of fidelity of the DNA synthesis, as the polymerase complex with the DNA constitutes the binding and recognition site for the dNTP. Furthermore, in the case of a DNA repair polymerase, elucidation of the enzyme interactions with the nucleic acid is of paramount importance for understanding the recognition mechanism of the damaged DNA by the enzyme. Thus, the mammalian pol b initiates the binding to the single-stranded (ss) or double-stranded (ds) DNA through its 8 kDa domain.19,21,28–31 The autonomous nature of the DNA-binding subsite on the 8 kDa domain allows pol b to bind small areas of the DNA substrate without engaging the catalytic 31 kDa domain of the enzyme into interactions with the nucleic acid.15–21,28–31 The subsequent association of the 31 kDa domain with the available DNA, follows this initial binding process.28–31 However, ASFV pol X does not possess the 8 kDa domain or an equivalent structural element, which would suggest a mechanism of binding similar to pol b (Figure 1). How a polymerase with such a simplified structure as ASFV pol X can still bind DNA with high affinity and recognize the damaged DNA structure is unknown. In spite of its paramount importance for understanding the DNA recognition process by the DNA repair polymerase, which is engaged in the viral defense mechanism against the host reaction to the infection, the direct and quantitative analyses of pol X interactions with the DNA has not yet been addressed. Such fundamental parameters of these interactions, like the site-size of the complex, i.e. number of nucleotide residues or base-pairs occluded by the enzyme in the complex, the functional structure of the DNA-binding site, intrinsic affinities, and cooperativity of the binding process are still unknown. Here, we describe quantitative analyses of interactions of ASFV pol X with the ssDNA.

ASFV pol X–ssDNA Interactions

We establish that the total site-size of the pol X–ssDNA complex is 16(G1) nucleotide residues. However, pol X can efficiently form a complex with ssDNA oligomer with only seven nucleotide residues, indicating the presence of the proper ssDNA-binding site, which encompasses only 7(G1) residues and is located asymmetrically within the polymerase molecule. Within the total DNAbinding site, the enzyme engages an additional binding region in interactions with the nucleic acid, spatially separated from the proper DNA-binding site. These two separated binding areas within the total DNA-binding site are not autonomous, i.e. the enzyme is not capable of engaging only one of them in interactions with longer DNA. Moreover, pol X binds the ssDNA without detectable cooperativity and with a modest base specificity.

Results Determination of the total site-size of the ASFV polymerase X–ssDNA complex: experiments with etheno-derivatives of the ssDNA oligomers Association of ASFV pol X with the ssDNA is not accompanied by an adequate change in the fluorescence of the enzyme to perform quantitative analyses of the enzyme–ssDNA interactions (data not shown). On the other hand, we have found that binding of pol X to fluorescent etheno-derivatives of the ssDNA is accompanied by a large increase of the fluorescence intensity of the nucleic acid (50–120%), providing the required signal to perform highresolution measurements of the enzyme–ssDNA complex formation. Similar large increase of the fluorescence emission of the etheno-derivative of the DNA substrates has been observed before in the case of the rat and human polymerase b, as well as Escherichia coli DnaB and PriA helicases, and explored by us to quantitatively examine the stoichiometry, energetics, and kinetics of these enzyme interactions with the DNA substrates.15–21,28–37 To address the fundamental problem of the total site-size of the pol X–ssDNA complex, i.e. the sitesize that corresponds to the maximum number of nucleotide residues occluded by the protein and obtain high resolution of the functional structure of the total DNA-binding site of the enzyme, we performed a series of quantitative fluorescence titrations, using etheno-derivatives of the ssDNA oligomers with different numbers of nucleotide residues. This is the same strategy that we have previously applied in our analyses of the interactions of the PriA and RepA helicases with the ssDNA interactions.35,36,38 Moreover, binding of pol X to the etheno-derivative of poly(dA), poly(d3A), induces precipitation of the complex at elevated concentrations of the protein and the DNA, hindering any direct quantitative analyses of the occluded total site-size of the enzyme with a polymer nucleic acid (data not shown).35,36

123 Fluorescence titrations of the 16-mer, d3A(p3A)15, with pol X at two different nucleic acid concentrations, in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, are shown in Figure 2(a). Binding of pol X to the 16-mer induces a significant increase of the nucleic acid fluorescence with the maximum relative fluorescence increase, DFmaxZ0.49G0.03. At higher nucleic acid concentrations, a given relative fluorescence increase, DFobs, is reached at a higher concentration of the polymerase, as a higher concentration of the enzyme is necessary to obtain the same total degree of binding, SQi, on the nucleic acid.15–21,32–40 The selected ssDNA oligomer concentrations provide separation of the fluorescence titration curves up to the relative fluorescence increase of w0.33. To examine the binding process independently of any assumption about the relationship between the observed signal and SQi, the fluorescence titration curves in Figure 2(a), have been analyzed, using the approach outlined in Materials and Methods.32–40 Figure 2(b) shows the dependence of DFobs on d3A(p3A)15 as a function of the total average degree of binding, SQi, of pol X. The plot is, within experimental accuracy, linear. At the selected 16-mer concentrations, the value of SQi could be determined up to w0.6. Extrapolation to the maximum relative fluorescence increase indicates that one molecule of pol X binds to the 16-mer. Analogous fluorescence titrations of the 20-mer, d3A(p3A)19, with pol X at two different nucleic acid concentrations, in the same solution conditions, are shown in Figure 2(c). Figure 2(d) shows the dependence of the observed relative fluorescence increase of the 20-mer as a function of SQi, of pol X. Extrapolation of the plot to the maximum relative fluorescence increase, DFmax, indicates that, as observed for the 16-mer, only one molecule of pol X binds to the 20-mer. However, the binding process and the maximum stoichiometry of the pol X–ssDNA is different in the case of the 24-mer, d3A(p3A)23, although this oligomer is only four nucleotide residues longer than the 20-mer. Fluorescence titrations of d3A(p3A)23 with pol X at two different nucleic acid concentrations, are shown in Figure 3(a). Because of the propensity of the pol X complexes with longer ssDNA oligomers to precipitate (see above), only the points where we did not observe precipitation are included in the titration curve in Figure 3(a), obtained at higher nucleic acid concentrations. The maximum relative fluorescence increase is larger than observed for the 16-mer and 20-mer with DFmaxZ1G0.05. Inspection of the titration curve, particularly at lower nucleic acid concentrations, where the total protein concentration is closer to its free concentration, clearly shows the presence of two binding phases.32–36,38,41 The presence of two binding phases is also reflected in the dependence of DFobs as a function of SQi of the pol X on the 24-mer, shown in Figure 3(b). The total average degree of binding, SQi, could be determined up to the value of w1.4. Extrapolation

ASFV pol X–ssDNA Interactions

Relative Fluorescence Increase

Relative Fluorescence Increase

124

0.6 (c)

Relative Fluorescence Increase

Relative Fluorescence Increase

Figure 2. (a) Fluorescence titrations (l ex Z325 nm, l em Z 410 nm) of the ssDNA 16-mer, d3A(p3A)15, with pol X in buffer 0.4 C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, 0.3 at two different concentrations of 0.2 the nucleic acid: (&) 1.0!10K6 M; (,) 8.0!10K6 M. The continuous lines are non-linear least-squares 0 0 fits of the fluorescence titration –7 –6 –5 –4 –7 –6 –5 –4 curves, using the single bindingLog [Pol X] Total Log [Pol X] Total site isotherm, as defined by equation (1), using a single set of parameters: KZ4.1!105 MK1 and 0.6 (b) (d) DFmaxZ0.49 (details in the text). 0.6 (b) The dependence of the relative fluorescence quenching, DF, upon 0.4 the average degree of binding, 0.4 SQi, of pol X on the 16-mer (&). The total average degree of 0.2 0.2 binding, SQi, has been quantitatively determined using the method described in Materials 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 and Methods. The continuous, Degree of Binding SQ i Degree of Binding SQ i straight line follows the points and does not have a theoretical basis. The broken line is an extrapolation of the binding density to the maximum value of the relative fluorescence change, DFmaxZ0.49, which provides SQiZ0.9G0.1, corresponding to the stoichiometry of one pol X per 16-mer. (c) Analogous fluorescence titrations of the 20-mer, d3A(p3A)19, with pol X at two different concentrations of the nucleic acid: (&) 1.0!10K6 M; (,) 8.0!10K6 M. The continuous lines are non-linear least-squares fits of the fluorescence titration curves, using the single binding-site isotherm, as defined by equation (1), using a single set of parameters: KZ6.8!105 MK1 and DFmaxZ0.58 (details in the text). (d) The dependence of the relative fluorescence quenching, DF, upon the average degree of binding, SQi, of pol X on the 20-mer (&). The continuous, straight line follows the points and does not have a theoretical basis. The broken line is an extrapolation of the binding density to the maximum value of the relative fluorescence change, DFmaxZ0.58, which provides SQiZ1.0G0.1, corresponding to the stoichiometry of one pol X per 20-mer. (a)

of the plot in Figure 3(b) to DFmaxZ1G0.05 provides SQiZ2.0G0.3. In other words, the presence of four additional nucleotide residues in the 24-mer, as compared to the 20-mer, enables the binding of two pol X molecules to the ssDNA 24-mer. However, further increase of the length of the ssDNA oligomer by six additional nucleotide residues does not lead to an increase of the maximum stoichiometry of the pol X–ssDNA oligomer complex. Fluorescence titrations of the 30-mer, d3A(p3A)29, with the polymerase at two different nucleic acid concentrations are shown in Figure 3(c). The two binding phases are still visible in the titration curve obtained at the lower nucleic acid concentration, although they are less pronounced than observed in the case of the 24-mer, due to the lower value of the maximum relative fluorescence increase, DFmaxZ0.83G0.03 (Figure 3(a) and (c)). The dependence of the relative fluorescence increase of the 30-mer, upon SQi of pol X on the 30-mer, is shown in Figure 3(d). Separation of the titration curves allowed us to determine the average degree of binding, SQi, up to w1.4, corresponding to the relative fluorescence quenching DFobsz0.65. Short extrapolation of the plot

in Figure 3(d) to DFmax Z0.83G0.03 provides SQiZ2.0G0.3. Analogous quantitative analysis of the stoichiometry of pol X–ssDNA complexes has been performed for the entire series of ssDNA oligomers (Table 1). The dependence of the maximum stoichiometry of pol X for the examined ssDNA oligomers, obtained from quantitative fluorescence titrations, as a function of the length of the ssDNA oligomer is shown in Figure 4. The data indicate that for the ssDNA oligomers ranging from 7 to 20 residues, only a single pol X molecule binds to the nucleic acid. A sudden jump in the maximum stoichiometry of the complex, from one to two, occurs between the 20-mer and 24-mer. The maximum stoichiometry does not increase even for the oligomer containing 37 residues, which is 13 residues longer than the 24-mer (Figure 4). These data provide the first indication that a total site-size of the pol X–ssDNA complex, i.e. the number of the nucleotide residues corresponding to the total sitesize of the enzyme in the complex with the ssDNA, is less than 19 residues, but it must contain at least 12 nucleotide residues per bound protein molecule (see below).

125

Relative Fluorescence Increase

Relative Fluorescence Increase

ASFV pol X–ssDNA Interactions

0.9

(c)

Relative Fluorescence Increase

Relative Fluorescence Increase

Figure 3. (a) Fluorescence titrations (lex Z325 nm, l em Z 410 nm) of the ssDNA 24-mer, d3A(p3A)23, with pol X in buffer 0.6 C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, 0.5 at two different concentrations of 0.3 the nucleic acid: (&) 1.0!10K6 M; (,) 8.0!10K6 M. The continuous lines are non-linear least-squares 0 0 fits of the fluorescence titration –7 –6 –5 –4 –8 –7 –6 –5 –4 Log [Pol X] Total Log [Pol X] Total curves, using the statistical thermodynamic model, as defined by equations (7)–(9), using a single set of parameters: 0.9 (d) 1 (b) Kq Z1.5!10 6 MK1, KpZ3.8! 105 MK1, DF1Z0.60 and DF2Z 0.40 (details in the text). (b) The 0.6 dependence of the relative fluorescence quenching, DFobs, upon 0.5 the average degree of binding, 0.3 SQi, of pol X on the 24-mer (&). The total average degree of binding, SQi, has been quantitatively 0 0 0 1 2 0 1 2 determined using the method Degree of Binding SQ i Degree of Binding SQ i described in Materials and Methods. The continuous straight line is the theoretical dependence of DFobs as a function of SQi, using equations (7)–(9). The line (— - —) marks the slope of the initial part of the plot. The line (- - -) is an extrapolation of the plot to the maximum value of the relative fluorescence change, DFmaxZ1.0G0.05, which provides SQiZ2.1G0.2, corresponding to the maximum stoichiometry of two pol X molecules per 24-mer. (c) Analogous fluorescence titrations of the 30-mer, d3A(p3A)29, with pol X at two different concentrations of the nucleic acid: (&) 1.0!10K6 M; (,) 8.0!10K6 M. The continuous lines are non-linear least-squares fits of the fluorescence titration curves, using the statistical thermodynamic model, as defined by equations (7)–(9), using a single set of parameters: KqZ1.7!106 MK1, KpZ2.1!105 MK1, DF1Z0.53 and DF2Z0.29 (details in the text). (d) The dependence of the relative fluorescence quenching, DFobs, upon the average degree of binding, SQi, of pol X on the 30-mer (&). The total average degree of binding, SQi, has been quantitatively determined using the method described in Materials and Methods. The continuous straight line is the theoretical dependence of DFobs as a function of SQi, using equations (7)–(9). The line (— - —) marks the slope of the initial part of the plot. The line (- - -) is an extrapolation of the plot to the maximum value of the relative fluorescence change, DFmaxZ0.83G0.05, which provides SQiZ2.0G0.2, corresponding to the maximum stoichiometry of two pol X per 30-mer. 1 (a)

The dependence of the stoichiometry of the pol X–ssDNA complex, upon the length of the ssDNA examined using the sedimentation velocity technique Further analysis of the stoichiometry of the pol X–ssDNA complexes as a function of the length of the nucleic acid has been carried out using an independent sedimentation velocity technique.16,34,37,38 The sedimentation coefficient of pol X is s20,wZ 2.5(G0.1) S, which is significantly larger than the sedimentation coefficient of any examined ssDNA oligomer (w1.4–1.7 S) (data not shown). The sedimentation experiments were performed with the ssDNA thymine oligomers containing a fluorescein residue at the 5 0 end of the nucleic acid (Materials and Methods).37,38 Because the fluorescein residue is introduced through phosphoramidate chemistry, with the marker replacing the base at the 5 0 end of the ssDNA, an oligomer, e.g. 5 0 Fl-dT(pT)12 contains an extra phosphate group and is referred to as a 14-mer. The same convention is applied to all fluorescein-ssDNA oligomers studied in this work. The presence of fluorescein

residue allows us to exclusively examine the sedimentation of the ssDNA by monitoring the absorption band of the marker, without any interference of the protein absorbance. The complex formation between the ssDNA oligomer and pol X will be manifested by a large increase of the apparent s20,w of the ssDNA. The observed increase of s20,w will be larger if more than one pol X molecule associates with the nucleic acid.37,38 Examples of sedimentation velocity profiles (40,000 rpm) of the ssDNA 14-mer, 5 0 Fl-dT(pT)12, in the presence of pol X and recorded at the fluorescein absorption band (495 nm) are shown in Figure 5(a). The protein concentration is 5!10K5 M and corresponds to the saturating concentration observed for the corresponding ssDNA oligomers (Table 1). The plots in Figure 5(a) show a moving boundary characterized by s20,wZ2.65(G0.11) S, which is close to the value of s20,w obtained for pol X alone (data not shown). Thus, the results indicate that a single pol X molecule associates with the ssDNA 14-mer. However, significantly faster movement of the sedimentation profile is observed for the 24-mer, 5 0 Fl-dT(pT)22, in the presence of

126

Maximum Stoichiometry [Pol X] / [ssDNA Oligomer]

1 6.8(G0.7)!105 – 9.3(G2.1)!104 0.58G0.03 1 5.9(G0.6)!105 – 9.3(G2.1)!104 0.60G0.03 1 4.8(G0.5)!105 – 9.3(G2.1)!104 0.61G0.03

1

40

Figure 4. The dependence of the total average degree of binding, SQi, of the pol X–ssDNA complex upon the length of the ssDNA oligomer in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2. The continuous lines follow the experimental points and do not have a theoretical basis.

Errors are standard deviations determined using 3–4 independent titration experiments. a Obtained using the KN(N) function in equation (17) (see the text for details).

1 3.2(G0.4)!105 5.4(G0.8)!104 – 1.05G0.07 1 2.3(G0.3)!105 5.4(G0.8)!104 – 0.63G0.03 1 1.1(G0.1)!105 5.4(G0.8)!104 – 0.43G0.01 1 5.5(G0.6)!104 5.4(G0.8)!104 – 0.35G0.03 n KN (MK1) Kpa (MK1) Kqa (MK1) DFmax

2

10 20 30 ssDNA Oligomer Length (Nucleotides)

1 4.0(G0.4)!105 – – 0.62G0.03

1 4.1(G0.4)!105 – – 0.49G0.03

d3A(p3A)19 20-mer d3A(p3A)18 19-mer d3A(p3A)17 18-mer d3A(p3A)15 16-mer d3A(p3A)13 14-mer d3A(p3A)11 12-mer d3A(p3A)9 10-mer d3A(p3A)7 8-mer d3A(p3A)6 7-mer

Table 1. Thermodynamic and spectroscopic parameters for the binding of pol X to ssDNA oligomers, that can accept only a single pol X molecule, in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2

ASFV pol X–ssDNA Interactions

the saturating concentration of pol X, as shown in Figure 5(b). The sedimentation boundary of this system is characterized by s20,wZ3.76(G0.15) S. Analogous studies have been performed for the entire series of the ssDNA oligomers. The dependence of the apparent sedimentation coefficient s20,w of the ssDNA oligomer, obtained in the presence of saturating concentrations of pol X, upon the length of the nucleic acid, is shown in Figure 5(c). There is only a slight linear increase of the s20,w value from w2.6 S, for the pol X–ssDNA 7-mer complex, to w2.7 S, observed for the complex with the 16-mer, indicating that a single molecule of pol X is associated with the ssDNA oligomers in this range of the length of the nucleic acid.38 The very modest increase of the sedimentation coefficient results from the increasing size of the ssDNA in the complex. However, for the 24-mer, the value of s20,w increases suddenly to w3.8 S, indicating an abrupt change in the stoichiometry of the formed complex, i.e. the binding of two pol X molecules to the nucleic acid (Figure 4). Further increase of the length of the ssDNA oligomer from 24 to 37 residues results in only a small progressive increase of the value of the sedimentation coefficient, indicating that the maximum stoichiometry of two pol X molecules bound to the examined ssDNA is preserved. It is evident that the plot in Figure 5(c) reflects the same behavior as the analogous stoichiometric plot in Figure 4. Structure of the ssDNA-binding site of pol X: minimum number of nucleotide residues engaged in interactions with the polymerase There are two pieces of experimental data that already indicate a complex structure of the DNAbinding site of pol X. First, a single pol X molecule

127

ASFV pol X–ssDNA Interactions

0.3 Absorbance (495 nm)

(a)

0.2

0.1

0

6.6

6.7

6.8

6.9

7

6.9

7

Radial Distance (cm) 0.15 Absorbance (495 nm)

(b)

0.1

0.05

0

6.6

6.7

6.8

Radial Distance (cm) (c)

s 20,w

4

3

2 0

10

20

30

40

ssDNA Oligomer Length (Nucleotides)

Figure 5. (a) Examples of absorption profiles, recorded at 495 nm, of analytical sedimentation velocity runs of the ssDNA 14-mer, 5 0 -Fl-dT(pT)12, in the presence of pol X in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2. (b) Examples of absorption profiles, recorded at 495 nm, of analytical sedimentation velocity runs of the 24-mer, 5 0 -Fl-dT(pT)22, in the presence of the pol X in the same buffer as in (a). In both panels, the concentration of pol X is 5!10K5 M and the concentration of the nucleic acid is (a) 2.5!10K6 M and (b) 1.2!10K6 M (oligomer), respectively; 14 min time interval; 40,000 rpm. (c) The dependence of the average sedimentation coefficient, s20,w, of the pol X-ssDNA complex upon the length of the ssDNA oligomer in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2. The continuous lines follow the experimental points and do not have a theoretical basis.

binds to the ssDNA 7-mer and to the 20-mer, indicating that the total site-size of the enzyme– ssDNA complex contains a binding region significantly smaller than the total ssDNA-binding site, that can still form efficient interacting contacts with the nucleic acid.35,36,38 Both oligomers can bind only a single pol X molecule, yet the 20-mer is almost three times longer than the 7-mer (see Discussion). Moreover, the fact that the enzyme binds only one 7-mer molecule indicates that there is only one such binding region (Figure 4).32–38 Second, the maximum fluorescence increase of the ssDNA oligomers at saturation with the polymerase, DFmax, shows a very peculiar dependence upon the length of the oligomers, reflecting a changing structure of the pol X–ssDNA complex as the length of the oligomer increases32–38 (Figures 2 and 3) (see below). Fluorescence titrations of the selected oligomers, 6, 7, 8, 10, and 12-mer, d3A(p3A)5, d3A(p3A)6, d3A(p3A)7, d3A(p3A)9, and d3A(p3A)11, with pol X in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, are shown in Figure 6(a). The affinity and the value of the relative fluorescence increases observed for the 6-mer are too low, if any, to quantitatively determine the binding of the enzyme to this oligomer in the examined solution conditions. However, as the length of the oligomer increases, the value of DFmax increases strongly for this set of the ssDNA oligomers, with the largest value of DFmax observed for the 12-mer. Such a large change in the etheno-derivative fluorescence in the complex indicates significant and progressive changes in the structure of the DNA as the length of the nucleic acid increases32–38 (see Discussion). Moreover, the titration curves shift toward the lower [pol X] for the longer nucleic acids, indicating an increased macroscopic affinity of the polymerase for the longer ssDNAs. Analogous fluorescence titrations of the selected ssDNA oligomers 14, 16, and 20-mer, d3A(p3A)13, d3A(p3A)15, and d3A(p3A)19, with pol X, are shown in Figure 6(b). For comparison, the titration of the 12-mer is also included in Figure 6(b). It is evident that further increase in the length of the ssDNA oligomer beyond 12 nucleotide residues does not lead to any further increase of the value of DFmax. Contrary, the maximum relative fluorescence increase dramatically drops for the 14-mer, although this oligomer is only two nucleotide residues longer than the 12-mer. The value of DFmax is also significantly lower for the 16-mer and 20-mer as compared to the 12-mer and does not show any clear correlation with the length of the nucleic acid, as observed for the shorter ssDNAs (Figure 6(a)). Such behavior of homo-oligomers of etheno-adenosine indicates that the structure of the ssDNA oligomers, longer than w12 nucleotide residues, is different from the structure of the shorter oligomers in the complex with pol X (see Discussion). Recall, a single pol X molecule binds to 7, 8, 10, 12, 14, 16, and 20-mer (Figure 4). Therefore, the titration

ASFV pol X–ssDNA Interactions

1

0.5

0

–7

–6 –5 Log [Pol X] Total

–4

8

(b)

(a)

1 6 KN x 10 – 5

Relative Fluorescence Increase

curves to equation (1). Values of KN and DFmax for the examined oligomers are included in Table 1. The dependence of the macroscopic binding constant, KN, for pol X binding to the ssDNA oligomers, containing from seven to 20 residues, upon the length of the ssDNA oligomer, is shown in Figure 7(a). The striking feature of the plot is its unusual structure, built of two linear regions separated by an intermediate plateau. For the oligomers from 7 to 12-mer, the values of KN increase linearly with the length of the ssDNA oligomers. The simplest explanation of such

(a)

0.5

4

2

0

0 –7

–6 –5 Log [Pol X] Total

–4

Figure 6. (a) Fluorescence titrations of 6, 7, 8, 10, and 12mer, d3A(p3A)6, d3A(p3A)7, d3A(p3A)9 and d3A(p3A)11 (lexZ325 nm, lemZ410 nm) with the pol X, in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2. Concentrations of all ssDNA oligomers are 1! 10K6 M; (:) d3A(p3A)5, (B) d3A(p3A)6, (C) d3A(p3A)7, (,) d3A(p3A)9, (&) d3A(p3A)11. The continuous lines are non-linear least-squares fits using the single-site binding isotherm (equation (1)) with binding parameters KN and DFmax included in Table 1. (b) Fluorescence titrations of 14, 16, and 20-mer, d3A(p3A)13, d3A(p3A)15, and d3A(p3A)19, (lexZ325 nm, lemZ410 nm) with the pol X, in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2. For comparison, the fluorescence titration of the 12-mer from (a) is also included (&). Concentrations of all ssDNA oligomers are 1!10K6 M; (,) d3A(p3A)13, (B) d3A(p3A)15, (C) d3A(p3A)19. The continuous lines are non-linear least-squares fits using the single-site binding isotherm (equation (1)) with binding parameters KN and DFmax included in Table 1.

curves in Figure 6(a) and (b) have been analyzed using the single site-binding isotherm defined as:
DF Z DFmax

K N PF 1 C KN PF

 (1)

where KN is the macroscopic binding constant characterizing the affinity for a given ssDNA oligomer containing N nucleotide residues. The continuous lines in Figure 6(a) and (b) are nonlinear least squares fits of the experimental titration

5

10 15 20 ssDNA Oligomer Length (Nucleotides)

5

10 15 20 ssDNA Oligomer Length (Nucleotides)

8 (b) 6 KN x 10 –5

Relative Fluorescence Increase

128

4

2

0

Figure 7. (a) The dependence of the macroscopic binding constant KN, upon the length of the ssDNA for oligomers containing seven to 20 residues, i.e. the oligomers that can accommodate only one pol X molecule. The continuous line for the part of the plot corresponding to oligomers from seven to 12 nucleotide residues is a linear least-squares fit to equation (3). Extrapolation of the line to KNZ0 intercepts the DNA length axis at NpZ6.0G0.5. The continuous line for the part of the plot corresponding to oligomers from 18 to 20 nucleotide residues is a linear least-squares fit to equation (5). Extrapolation of the line intercepts the DNA length axis at NqZ13.2G1.0. (b) The same plot of the dependence of the macroscopic binding constant KN, upon the length of the ssDNA for oligomers containing seven to 20 residues as in (a), but with non-linear least-squares fit (continuous line) of the experimental curve, using the analytical expression of KN as a function of N as defined by equation (16) with the KpZ5.4(G0.7)!104 MK1, KqZ 9.3(G2.1)!104 M K1, Kp0 Z 2:4ðG0:7Þ !104 MK1 , a pZ7, ainterZ13, and aqZ17 (details in the text).

129

ASFV pol X–ssDNA Interactions

empirical linear behavior of KN as a function of the length of the ssDNA oligomers is that there is a small, discrete binding region within the total DNA-binding site of pol X that experiences the presence of several potential binding sites on the ssDNA oligomers.42–44 We refer to this binding region as the proper DNA-binding site.35,36,38 The proper DNA-binding site engages in interactions a number, p, of the nucleotide residues, referred to as the site-size of the proper DNA-binding site, which must be smaller than the length of the examined ssDNA oligomers.35,36,38 Therefore, the values of KN contain a statistical factor that can analytically be defined in terms of the site-size, p, and the intrinsic binding constant, Kp, of the proper DNA-binding site, as:38,45 KN Z ðNKp C 1ÞKp

(2)

Expanding equation (2) provides a dependence of the KN upon the length of the oligomer, N, as: KN Z NKp Kðp C 1ÞKp

(3)

Extrapolation of the plot in Figure 7(a) to the zero value of the macroscopic equilibrium constant, KN, intercepts the abscissa at the DNA length, Np, corresponding to the length of the ssDNA oligomer too short to be able to form a complex with the polymerase.38,45 Introducing KNZ0 into equation (2) provides the value of Np as: Np Z pK1

oligomers indicates that different areas of the total DNA-binding site of the polymerase, beyond the proper DNA-binding site, become involved in the interactions with the longer nucleic acid38,45 (see Discussion). The second linear region of the plot in Figure 7(a) can be described by expressions analogous to equations (3) and (4), as: KN Z NKq Kðq C 1ÞKq

(5)

Nq Z qK1

(6)

and:

However, the parameters in equations (5) and (6) now have different physical meanings. The quantity, q, is the minimum length of the ssDNA oligomer that can engage the total DNA-binding site of the polymerase and Kq is the intrinsic binding constant for the total DNA-binding site. Extrapolation of this region to the KNZ0 provides NqZ13.3G1.0 and qZ14.3G1.0. Thus, the data indicate that in order to engage the total DNA-binding site of pol X, the ssDNA must have, at least, w14–15 nucleotide residues. Notice, this is a minimum estimate, not the actual site-size of the total DNA-binding site, which can be even larger (see below). The slope of the linear region, for the ssDNA oligomers with the length between 18 and 20 residues, provides Kq Z ð9:3G2:1Þ !104 MK1 . This value is higher than the corresponding value of the intrinsic binding constant characterizing the interactions of the short ssDNA oligomers with the proper DNA-binding site of the enzyme (see Discussion).

(4) Model of the pol X–ssDNA complex

The plot in Figure 7(a) gives NpZ6G0.5. Thus, an oligomer that has only six residues will not form efficient binding contacts with the enzyme, in excellent agreement with direct fluorescence titrations (Figure 6(a)). Therefore, the proper DNA-binding site of pol X requires pZ7(G1) nucleotide residues of the ssDNA to engage in energetically efficient interactions with the nucleic acid. The intrinsic binding constant, Kp, can be determined from the slope of the first linear region in Figure 7(a), for the ssDNA oligomers with the length between seven and 12 residues (equation (3)), with the average value of 5.4(G0.8)!104 MK1 (Table 1) (see Discussion). Structure of the ssDNA-binding site of pol X: binding of the ssDNA oligomers comparable to the site-size of the total DNA-binding site For the ssDNA oligomers with the length exceeding 12 nucleotide residues, the plot in Figure 7(a) becomes non-linear and passes through an intermediate plateau into the second linear region for the oligomers with the length higher than w16–17 residues. Such transition between two linear regions of the plot of the macroscopic binding constant as a function of the length of the ssDNA

As discussed above, the transition of the maximum stoichiometry of the pol X–ssDNA oligomer complex, from a single pol X molecule bound per a ssDNA oligomer to two molecules bound per oligomer, occurs between 20 and 24-mers (Figure 4). On the other hand, the 37-mer can still accommodate only two pol X molecules (Table 2). However, binding studies with short oligomers indicate that the polymerase can bind the ssDNA and engage only 7(G1) nucleotides in interactions with its proper DNA-binding site (see above). These data and the analysis in the previous section indicate that the total site-size of the pol X–ssDNA complex is 16(G1) residues per bound protein (see Discussion). To address the structure of the total DNA-binding site of pol X, we first consider the following limiting model of the pol X–ssDNA complex depicted schematically in Figure 8(a). The proper ssDNAbinding site, which engages seven nucleotides, is located in the central part of the protein molecule. However, if this model applies, it would require that the protein molecule possesses two parts, each part protruding over w8 nucleotides on both sides of the proper DNA-binding site giving the total DNA-binding site size of 23 nucleotides. This is because the minimum number of nucleotides

130

ASFV pol X–ssDNA Interactions

Table 2. Thermodynamic and spectroscopic parameters for the binding of pol X to ssDNA oligomers, that can accept two pol X molecules in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2

n Kq (MK1) Kp (MK1) nq np u DF1 DF2

d3A(p3A)23 24-mer

d3A(p3A)25 26-mer

d3A(p3A)29 30-mer

d3A(p3A)34 35-mer

d3A(p3A)36 37-mer

2 1.5(G0.5)!106 3.8(G1.1)!105 16 7 1.0G0.02 0.60G0.03 0.40G0.03

2 1.5(G0.5)!106 2.6(G0.7)!105 16 7 1.0G0.2 0.80G0.03 0.45G0.03

2 1.7(G0.6)!106 2.1(G0.5)!105 16 7 1.0G0.2 0.53G0.03 0.29G0.03

2 1.3(G0.2)!106 1.8(G0.3)!105 16 7 1.0G0.2 0.39G0.03 0.40G0.03

2 1.4(G0.4)!106 2.1(G0.5)!105 16 7 1.0G0.2 0.55G0.03 0.55G0.03

Errors are standard deviations determined using 3–4 independent titration experiments.

engaged in interactions with the total DNA-binding site must be, at least, w14–15, independent of where the proper DNA-binding site is located on the nucleic acid (Figure 7(a)) (see above). Besides the fact that such total site-size would be much larger than estimated from the fluorescence titration and sedimentation data, the first bound pol X molecule would block w15 nucleotides and the second bound molecule would also require an additional w15 residues, giving the total requirement of 30 residues. As a result, the ssDNA 24-mer would be able to accommodate only one pol X molecule. This is not experimentally observed (Figures 4 and 5). Next, we consider a limiting model where the proper DNA-binding site of pol X engages seven nucleotide residues and is asymmetrically located on one side of the polymerase molecule as depicted in Figure 8(b). Part of the enzyme protrudes over the nucleic acid and engages an additional area in interactions at a distance from the proper binding site corresponding to w8–9 nucleotide residues of the ssDNA. The total site-size of the complex is now 16 residues. With such a functional structure of the total DNA-binding site, the ssDNA oligomers with 20 or less residues would bind only one pol X molecule, while oligomers from 24 to 37 residues would be able to accommodate two pol X molecules (Figure 4). This is exactly what is experimentally observed. One pol X molecule binds to the 20-mer or shorter oligomers and only two enzyme molecules bind to oligomers with the length between 24 and 37 residues (Figure 4). Therefore, the model of the pol X–ssDNA complex, depicted in Figure 8(b), adequately describes all experimentally determined stoichiometries of the pol X complexes with the series of ssDNA oligomers examined here. In other words, these data and analyses of the two alternative limiting models indicate that the actual total site-size of the pol X–ssDNA complex is 16(G 1) nucleotide residues and includes seven residues encompassed by the proper DNA-binding site of the enzyme, as well as nine residues occluded by the protein only on one side of the proper DNAbinding site and engaged in interactions with the nucleic acid (Figure 8(b)). Notice, if the total DNAbinding site were only 15 nucleotides then the 37mer would be able to accommodate three pol X molecules, which is not experimentally observed

(Table 2). This is because two bound pol X molecules would block only w30 residues, leaving seven residues to accommodate an additional polymerase molecule bound through its proper DNA-binding site.

(a)

(b)

ssDNA 24 -mer

Figure 8. (a) Schematic model for the binding of pol X to the ssDNA 24-mer, based on the site-size of the proper DNA-binding site of the complex pZ7 nucleotide residues and the minimum site-size of the total DNAbinding site of the polymerase, qZ16. The proper DNAbinding site, which encompasses seven nucleotide residues, is located in the center of the enzyme molecule. In such a model the protein matrix would have to protrude on both sides of the proper DNA-binding site over w7–8 residues in order to provide the minimum total DNA-binding site-size of w15 residues, independent of the location of the proper DNA binding site on the nucleic acid lattice. However, this model would allow the binding of only one molecule of pol X to the ssDNA 24-mer, which is not experimentally observed (Figures 4 and 5(c)). (b) Schematic model for the binding of the pol X to the ssDNA 24-mer, based on the site-size of the proper DNA-binding site of the complex pZ7 nucleotide residues and the site-size of the total DNA-binding site of the polymerase, qZ16. The polymerase binds the ssDNA in a single orientation with respect to the polarity of the sugar-phosphate backbone of the ssDNA. The proper DNA-binding site, which encompasses seven nucleotide residues, is located on one side of the enzyme molecule, with the rest of the protein matrix protruding over the extra nine nucleotide residues. When bound at the ends of the nucleic acid, or in its center, the enzyme can occlude from seven to 16 nucleotides. This model would allow the binding of two molecules of pol X to the ssDNA 24-mer and is in excellent agreement with the stoichiometries of all studied ssDNA oligomers in this work (details in the text).

131

ASFV pol X–ssDNA Interactions

Statistical thermodynamic model for the pol X binding to the ssDNA oligomers that can accommodate two pol X molecules Estimation of the site-sizes of the proper and total DNA-binding site of the pol X–ssDNA complex allows us to address the binding of the enzyme to the ssDNA oligomers that can accommodate two polymerase molecules. However, the complex structure of the total DNA-binding site of the pol X–ssDNA complex makes such analysis significantly more difficult than for the oligomers where only a single molecule of the enzyme associates with the DNA (Figure 2(a) and (c)) or, in general, for the cases where a protein forms only a single type of complex with the nucleic acid.35,36,38 The partition function, ZN, for the pol X–ssDNA oligomer systems must account for different site-sizes of the formed complexes, overlap of the binding sites, and possible cooperative interactions between the bound pol X molecules.42–44 We consider the simplest statistical thermodynamic model that accounts for the experimental observations, using a combinatorial approach. The single bound pol X molecule can form two types of complexes occluding qZ16 or pZ7 nucleotide residues, i.e. can form a complex engaging the total or only the proper DNA-binding site (see above). These two complexes are characterized by different intrinsic binding constants, Kq and Kp, respectively. Engagement of the total DNA-binding site in interactions with the DNA is favorable, because of its higher intrinsic affinity, i.e. KqOKp. In the case of the examined 24, 26, and 30-mer, the second pol X molecule can only engage its proper DNA-binding site, i.e. it occludes only seven nucleotide residues. Taking into account the presence of complexes with different site-sizes and different intrinsic affinities, the overlap of the binding sites, and possible cooperative interactions, the partition function for the considered pol X–ssDNA systems, ZN1, is:

ZN1 Z 1 C ½ðNKq C 1ÞKq C ðqKpÞKp PF C ½ðNKqKp C 1ÞðNKqKp C 2Þ K2ðNKqKp C 1ÞKq Kp P2F C 2ðNKqKp C 1Þ !Kq Kp uP2F

ð7Þ

where u is the parameter that characterizes cooperative interactions. This form of the partition function applies to all pol X–ssDNA complexes with the ssDNA oligomers ranging from 23 to 32 nucleotide residues, i.e. where one of the bound pol X molecules can engage only its proper DNAbinding site in interactions with the nucleic acid. The total average degree of binding, SQi, is then obtained by the standard statistical thermodynamic

expression, SQiZvln ZN1/vln PF, as:46 f½ðNKq C 1ÞKq C ðqKpÞKp PF C2½ðNKqKp C 1ÞðNKqKp C 2Þ K2ðNKqKp C 1ÞKq Kp P2F C4ðNKqKp C 1ÞKq Kp uP2F g SQi Z ZN1

(8)

The observed relative fluorescence increase of the ssDNA, DFobs, is then:
DFobs Z DF1

ðNKq C 1ÞKq C ðqKpÞKp PF ZN1



C ðDF1 C DF2 Þ 3 ½ðNKqKp C 1ÞðNKqKp C 2Þ 2 7 6 K2ðNKqKp C 1ÞKq Kp PF 7 6 6 C2ðNKqKp C 1ÞK K uP2 7 F 7 q p 6 !6 7 7 6 ZN1 7 6 5 4 2

(9)

where DF1 and DF2 are the relative fluorescence increases accompanying the binding of the first and the second pol X to the nucleic acid. The inclusion of only one spectroscopic parameter, DF1, to characterize the binding of a single pol X molecule is dictated by the fact that the dependence of DFobs upon SQi is, within experimental accuracy, linear, indicating very similar molar fluorescence increase for the pol X bound through the total or the proper DNA-binding site (Figure 3(b) and (d)). The situation is different in the case of the 35 and 37-mers, where both pol X molecules bound to the nucleic acid can engage the total or the proper DNA-binding site in interactions with the nucleic acid. The partition function for these oligomers, ZN2, that takes into account the presence of complexes with different site-sizes (qZ16 and pZ7) for both bound enzyme molecules is then: ZN2 Z 1 C ½ðNKq C 1ÞKq C ðqKpÞKp PF C f½ðNK2q C 1ÞðNK2q C 2ÞK2ðNK2q C 1Þ C 2ðNK2q C 1ÞugðKq PF Þ2 (" C

ðqKpÞðNKqKpÞ C 1K

iZqK Xp

#

)

i C ðqKpÞu

iZ1

!ðKq Kp P2F Þ

ð10Þ

In general, this form of the partition function applies to all pol X–ssDNA systems with the ssDNA length ranging from 33 to 47 nucleotide residues. The average degree of binding, SQi, is defined by SQ i ZvlnZ N2/vlnPF,46 while the

132

ASFV pol X–ssDNA Interactions

observed relative fluorescence increase is:
 ðNKqC1ÞKq CðqKpÞKp PF DFZDF1 ZN2

the obtained uz1 indicates the lack of significant cooperative interactions between the bound pol X molecules (see Discussion).

CðDF1 CDF2 Þ 3 2 f½ðNK2qC1ÞðNK2qC2ÞK2ðNK2qC1Þ 7 6 C2ðNK2qC1ÞugðKq PF Þ2 7 6 8" #9 7 6 iZqKp 7 6 > P > > 7 6 > = < ðqKpÞðNKqKpÞC1K i 7 6 iZ1 6C ðKq Kp P2F Þ7 7 6 > > > 7 6 > CðqKpÞu ; 7 6 : 7 !6 7 6 ZN2 7 6 7 6 7 6 7 6 7 6 7 6 7 6 7 6 5 4 (11) Notice that because DFmaxZDF1CDF2, and both DF1 and DFmax are experimentally determined there are only three independent parameters, Kq, Kp, and u. The continuous lines in Figure 3(a) and (c) are non-linear least-squares fits of the experimental isotherms using equations (7)–(9) for the 24, 26, and 30-mer. Analogously, equations (10) and (11) were used for the analyses of binding of pol X to the 35 and 37-mer, respectively (data not shown). The spectroscopic and binding parameters for the examined ssDNA oligomers, which can accommodate two pol X molecules, are included in Table 2. The values of Kq and Kp are similar for all examined ssDNA oligomers with the averages of w1.5!106 MK1 and w2.5!105 MK1, respectively (Table 2). This is an expected result as both binding constants characterize the intrinsic binding processes. However, the data indicate that the intrinsic affinity of the complex where the polymerase associates with the DNA using its total DNA binding site, with the site-size qZ16, is only by a factor w6 larger than the intrinsic affinity of the complexes where only the proper DNA-binding site, with the site-size, pZ7, is engaged in interactions with the nucleic acid (see Discussion). In the case of the 24, 26, and 30-mers, the relative increase of the nucleic acid fluorescence is larger for the first bound pol X molecule, predominantly associated with the nucleic acid using the total DNA-binding site, than for the second pol X bound through its proper DNA-binding site. The difference between the values of DF1 and DF2 is less pronounced for the 35 and 37-mers where both pol X molecules can engage the total binding site (Table 2). Interestingly, the cooperative interaction parameter uZ1G0.2 for all examined ssDNA oligomers. Although the values of Kp and u are strongly correlated in the considered statistical thermodynamic model (equations (10) and (11)),

Salt effect on pol X–ssDNA interactions To obtain further insight into the pol X interactions with the ssDNA within the proper and the total DNA-binding site, we examined the salt effect on the binding of the 10, 16, and 20-mer, d3A(p3A)9, d3A(p3A)15, and d3A(p3A)19 to pol X. The experiments have been performed in the absence of magnesium to avoid the competition between the NaC and Mg2C cations for the nucleic acid.47,48 Fluorescence titrations of d3A(p3A)9 with pol X in buffer C (pH 7.0, 10 8C), containing 0.1 mM EDTA and different NaCl concentrations, are shown in Figure 9(a). Analogous fluorescence titrations of the 16-mer, and 20-mer with the polymerase are shown in Figure 9(b) and (c). At higher concentrations of NaCl, the titration curves shift toward higher total protein concentrations, indicating a decreasing macroscopic affinity of the polymerase–nucleic acid complex for all examined oligomers. Also, the maximum relative fluorescence change accompanying the binding decreases with increasing salt concentration. The dependence of the logarithm of the macroscopic binding constants, K10, K16, and K20 upon the logarithm of [NaCl] (log–log plot)47,48 is shown in Figure 9(d). The plots are clearly linear in the examined salt concentration ranges. However, there are dramatic differences between the values of the slopes of the plots obtained for different oligomers, which correlate with the length of the nucleic acid. In the case of the 10-mer, the slope of the plot vlogK 16/vlog[NaCl], is K1.3G0.3 indicating that association of the ssDNA oligomer that can engage only the proper DNA-binding site is accompanied by the net release of w1 ion. The slope of the log–log plot obtained for the 16-mer is v log K16 =v log½NaClZK3:2G0:3, indicating that the binding of the ssDNA oligomer that has a minimum length to engage the total DNA-binding site is accompanied by the net release of w3 ions, a number significantly higher than observed for the proper DNA-binding site. On the other hand, the slope, v log K20 =v log½NaClZK5:9G0:3. Thus, binding of the 20-mer, which is larger than the estimated total site-size of the pol X–ssDNA complex, but still can bind only one pol X molecule, is accompanied by the net release of w6 ions. The observed dramatic differences between the net ion releases accompanying the binding of different ssDNA oligomers, which span the proper or the entire total DNA-binding site, clearly indicate that an additional area of the total DNA-binding site, beyond the proper DNA-binding site, is being engaged in interactions with the longer ssDNA oligomers (see Discussion).

133

Relative Fluorescence Increase

Relative Fluorescence Increase

ASFV pol X–ssDNA Interactions

Figure 9. (a) Fluorescence titrations of the ssDNA 10-mer, d3A(p3A)9, (lexZ325 nm, lemZ 1 410 nm) with pol X in buffer C (pH 7.0, 10 8C), containing differ1 ent NaCl concentrations: (&) 0.5 50 mM, (,) 63.3 mM, (C) 75.7 mM, (B) 101.3 mM, (:) 126 mM. (b) Fluorescence titrations of the ssDNA 16-mer, 0 0 d3A(p3A)9, (lexZ325 nm, lemZ –8 –7 –6 –5 –4 –8 –7 –6 –5 –4 Log [Pol X] Total Log [Pol X] Total 410 nm) with pol X in buffer C (pH 7.0, 10 8C), containing different NaCl concentrations: (&) 50 mM, (,) 75.7 mM, (C) 2 (b) 8 (d) 101.3 mM, (B) 126 mM, (:) 154.5 mM, (6) 202 mM. (c) Ana1.5 logous fluorescence titrations of 7 the ssDNA 20-mer, d3A(p3A)19, 1 with pol X in buffer C (pH 7.0, 6 10 8C), containing different NaCl 0.5 concentrations: (&) 50 mM, (,) 5 76.5 mM, (C) 101.3 mM, (B) 126 mM, (:) 154.5 mM. The 0 –1.4 – 1.2 –1 – 0.8 –7 –6 –5 –4 concentrations of the ssDNA Log [Pol X] Total Log [NaCl] oligomers in all panels are 1! 10K6 M (oligomer). The continuous lines in (a), (b), and (c) are non-linear least-squares fits of the titration curves, using the single binding site isotherm as defined by equation (1). (d) The dependence of the logarithm of the macroscopic binding constants, K10, K16, and K20 or the binding of the pol X to d3A(p3A)9, d3A(p3A)15, and d3A(p3A)19 upon the logarithm of NaCl concentrations (log–log plots).47,48 The continuous lines are the linear least-squares fits of the plots, which provide the slopes, vlogK9/vlog[NaCl]Z1.3G0.3 vlogK16/vlog[NaCl]Z3.2G0.5, and vlogK19/vlog[NaCl]Z5.9G 0.8 (details in the text). 2

(c)

Log K N

Relative Fluoresecence Increase

(a)

Base specificity of pol X: binding of the enzyme to unmodified ssDNA oligomers using macromolecular competition titrations method (MCT) Determination of the pol X affinity for the unmodified ssDNA oligomers has been performed using the macromolecular competition titration (MCT) method (Materials and Methods).41 In these studies the fluorescent 16-mer, d3A(p3A)15, is used as a reference lattice and competition studies are performed with the unmodified ssDNA 20-mers, which can only accommodate a single pol X molecule (see above). Thus, the entire binding system is composed of two short ssDNA lattices, reference and unmodified oligomer, competing for pol X.41 The partition function, ZC, of the entire binding system is then defined as: ZC Z 1 C K16 PF C K20 PF

(12)

where K16 is the macroscopic binding constant for pol X binding to the 16-mer, d3A(p3A)15 (Table 1). The concentration of the bound protein, Pb, is then:

 K16 PF K20 PF Pb Z (13) N C N 1 C K16 PF TR 1 C K20 PF TS Pb Z ðSQi ÞR NTR C ðSQi ÞS NTR

(14)

and: PF Z PT – Pb

(15)

where PT is the total concentration of pol X, NTR and NTS are the total concentrations of the fluorescent reference 16-mer and the unmodified ssDNA 20-mer, respectively. Fluorescence titrations of d3A(p3A)15, with the pol X, in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, in the presence of two different concentrations of the 20-mer, dA(pA)19, are shown in Figure 10(a). For comparison, the titration curve in the absence of the competing unmodified 16-mer is also included. Analogous fluorescence titrations of d3A(p3A)15, with pol X in the presence of two different concentrations of the 20-mer, dT(pT)19, are shown in Figure 10(b). A large shift of the binding isotherms, in the presence of dA(pA)19 or dT(pT)19 indicates an efficient competition between d3A(p3A) 15 and the 20-mers for pol X. The continuous lines in Figure 10(a) and (b) are non-linear least-squares fits of the experimental titration curves, with a single fitting parameter, K20, using equations (12)–(15) and K16Z4.1! 105 MK1, obtained independently for the reference 16-mer. The obtained binding constants are K20 Z for 3:0ðG0:9Þ !107 MK1 and 8:0ðG1:8Þ !106 MK1

Relative Fluorescence Increase

134

ASFV pol X–ssDNA Interactions

0.4

Total ssDNA-binding site of the ASFV pol X has heterogeneous structure 0.2

0 –8

Relative Fluorescence Increase

Discussion

(a)

–7

–6 –5 Log [Pol X] Total

–4

–7

–6 –5 Log [Pol X] Total

–4

(b) 0.4

0.2

0 –8

Figure 10. (a) Fluorescence titrations of the ssDNA 16-mer, d3A(p3A)15, (lexZ325 nm, lemZ410 nm) with the pol X in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, in the absence (&) and the presence of the unmodified ssDNA 20-mer, dA(pA)19. The concentration of d3A(p3A)15 is 1!10K6 M. The concentration of unmodified ssDNA oligomer, dA(pA)19 are: (,) 1!10K6 M (C) 5!10K6 M (oligomer). The continuous lines are non-linear least-squares fits of the experimental fluorescence titration curves, using the MCT method (Materials and Methods) (details in the text). (b) Fluorescence titrations of the ssDNA 16-mer, d3A(p3A)15, (lexZ325 nm, lemZ410 nm) with the pol X in buffer C (pH 7.0, 10 8C), containing 100 mM NaCl and 1 mM MgCl2, in the absence (&) and the presence of the unmodified ssDNA 20-mer, dT(pT)19. The concentration of d3A(p3A) 15 is 1!10 K6 M. The concentrations of unmodified ssDNA oligomer, dT(pT)19 are: (,) 1! 10K6 M (C) 5!10K6 M (oligomer). The continuous lines are non-linear least-squares fits of the experimental fluorescence titration curves, using the MCT method (Materials and Methods) (details in the text).

dA(pA)19 and dT(pT)19, respectively. Analogous studies with dC(pC)19 provided K20 Z 3:0ðG0:6Þ ! 107 MK1 (data not shown). Thus, pol X does not have a preference for the type of the base. However, all unmodified oligomers have their macroscopic binding constant larger by a factor of w30 than the binding constant obtained for the ethenoderivative of the dA(pA)19, d3A(p3A)19 (Table 2) (see Discussion).

Elucidation of the stoichiometries and interactions of the protein–nucleic acid complex is of paramount importance for any quantitative analysis of the energetics, kinetics, and structure of the protein–nucleic acid complexes.35,36,38,41–48 Studies described here provide, for the first time, an insight into the complex energetics of the ASFV pol X interactions with the ssDNA. The major element of the experimental strategy applied in this work to examine the total DNA-binding site of pol X and its functional structure is the use of a large series of ssDNA oligomers, differing in number of nucleotide residues.35,36,38 It allows us to determine the total site-size of the pol X–ssDNA complex, q, heterogeneity of the total DNA-binding site, i.e. the presence of the proper DNA-binding site with the site-size, p, the location of the proper DNA-binding site within the polymerase molecule, the presence of an additional interaction area within the total DNA-binding site, and its spatial separation from the proper DNA-binding site. The crucial information about the functional structure of the total DNA-binding site of pol X comes from the analysis of the macroscopic affinities of the enzyme as a function of the length of different ssDNA oligomers that can accommodate only one pol X molecule (Figure 7(a)). The value of p corresponds to the length of the shortest ssDNA oligomer that still forms a complex with the enzyme, with the intrinsic affinity similar to the intrinsic affinity of longer oligomers. In the case of pol X, the shortest ssDNA oligomer that efficiently binds to the polymerase has 7(G1) nucleotides, a value significantly shorter than the total site-size of the pol X–ssDNA complex, qZ16G1 (see below). Notice, the observation that macroscopic binding constants for the set of the ssDNA oligomers, much shorter than the total site-size, contain a statistical factor (Figure 7(a)) indicates an essential aspect of the proper ssDNA-binding site of pol X. The statistical effect would appear only if there was not interference from the remaining areas of the total DNA-binding site. Thus, the proper DNAbinding site must be structurally separated from the rest of the total binding site. Moreover, this spatial separation must be larger than merely the site-size of the proper DNA-binding site. For instance, the ssDNA 12-mer experiences the same form of the statistical factor in its macroscopic binding constant, KN, as the 7-mer (Figure 7(a)). As a result, the ssDNA oligomers with seven to w12 residues have the same intrinsic binding constant, Kp, and, the plot of KN as a function of N is strictly linear (Figure 7(a)). However, the presence of the extra two or three nucleotide residues over 12 eliminates the strictly linear dependence of KN upon N for the examined

135

ASFV pol X–ssDNA Interactions

set of the ssDNA oligomers, indicating that the 14mer already experiences some interference of the remaining part of the total DNA-binding site (Figure 7(a)). This is accompanied by a strong decrease of the observed maximum fluorescence increase, DFmax, accompanying the complex formation with the ssDNA oligomers longer than 12 nucleotide residues, indicating also that the longer oligomers assume a different structure in the complex with the polymerase (Table 1) (see below). The plot in Figure 7(a) shows an intermediate region for oligomers with 14 to 16 residues. The region has a slope of w0, indicating that the intrinsic binding constant for the oligomers in this intermediate region does not depend upon the length of the bound nucleic acid. The simplest explanation for such independence of the intrinsic binding constant in the intermediate region of the plot in Figure 7(a) is that, although the oligomers in this region interact with the proper DNA-binding site and experience an additional interacting area of the total DNA-binding site, particularly the 16-mer, they are allowed to assume energetically more favorable conformation than the conformation induced upon the exclusive binding to the proper binding site, as indicated by the abrupt and large decrease of their observed maximum fluorescence increase, DF max upon saturation with the enzyme (Table 1). In other words, the favorable unique conformation of the DNA oligomers in the intermediate region would affect the intrinsic affinity by eliminating the unfavorable free energy required to assume a different DNA structure imposed on the nucleic acid by the structure of the proper and the total DNA-binding site (see below). On the other hand, for the ssDNA oligomers with 16 or more residues, the dependence of KN upon N becomes once again apparently linear (Figure 7(a)). The intrinsic binding constant is higher than observed for the shorter ssDNAs because, now, the oligomers are long enough to more efficiently engage an additional interaction area of the total DNA-binding site, which overcomes the unfavorable energy required for the structural changes of the nucleic acid imposed by the total binding site (see above). The maximum fluorescence increase for the oligomers containing from w16 to 20 residues, DFmaxz0.5–0.6, is similar for all examined ssDNAs, indicating similar structure of the DNA in the complex (Table 1). Notice, the proper DNA-binding site encompasses w 7 nucleotide residues. Yet, the 12-mer cannot engage the additional area of interactions and does not assume a conformational state of the longer ssDNA oligomers, as indicated by its much larger DFmax than observed for the oligomers containing 14 or more residues (Table 1). Only when the ssDNA contains w16–17 nucleotides, the plot in Figure 7(a) becomes linear once again and the binding is characterized by a different intrinsic affinity. These data indicate that the additional interaction area is spatially separated from the proper DNA-binding site by a distance

corresponding to the length of w7–8 nucleotide residues of the ssDNA (see below). Further information about the heterogeneous structure of the total-DNA-binding site comes from the salt effect on the enzyme binding to the ssDNA oligomers of different lengths that accommodate only one pol X molecule. The net number of ions released upon the complex formation with the 10, 16, and 20-mer is w1, w3, and w6 (Figure 9(d)). Although the degree of cation binding to the oligomers increases with the length of the nucleic acid, such an increase cannot account for the experimentally observed large difference in the net number of released ions among the studied oligomers.47,48 Rather, the data indicate that while the length of the ssDNA increases, the nucleic acid engages in interactions an additional area of the total DNA-binding site, leading to the increased net number of the released ions (see below). pol X binds the ssDNA with several different intrinsic affinities reflecting a heterogeneous structure of the total DNA-binding site The dependence of the macroscopic binding constant, KN, upon N in Figure 7(a), for the ssDNA oligomers that can only accommodate one pol X molecule, shows three different regions, indicating that there are at least three different intrinsic interactions of the ssDNA with the total DNA-binding site of pol X. The changes in the interactions occur abruptly within one or two nucleotide residues of the nucleic acid between different regions. Such behavior suggests that the observed macroscopic binding constant can be analytically described by a unit step function, U(NKax), that defines all three different regions of the functional dependence of KN upon N, characterized by different intrinsic affinities, as: KN Z ½NKp KðpK1ÞKp ½UðNKap ÞK½UðNKainter Þ C Kinter ½UðNKainter ÞKUðNKaq Þ C ½NKq KðqK1ÞKq C ðqKpÞKp0 UðNKaq Þ (16) where U(NKax) is a unit step function defined as 0 for 0%N!ax and 1 for NRax.49 The parameter ax defines the range of the ssDNA oligomer length with a given intrinsic binding constant that characterizes the interactions with pol X. For instance, the function, [U(NKap)KU(NKainter)], describes the range of lengths of the ssDNA oligomers that can only engage the proper DNA-binding site of pol X with intrinsic binding constant, Kp. The plot in Figure 7(a) indicates that apZ7 and ainterZ13 residues. The function, [U(NKainter)KU(NKaq)], describes the range of lengths of the ssDNA oligomers in the intermediate region, characterized by the intrinsic binding constant, Kinter, which does not indicate the presence of any statistical factor in this region (Figure 7(a)). Thus, the binding constant, Kinter, is equivalent to the macroscopic binding

136 constant, KN, for the ssDNA oligomers in the intermediate region in the plot in Figure 7(a) and (b). Finally, the function, U(NKaq), describes the range of lengths of the ssDNA oligomers that can efficiently engage the total DNA-binding site with the intrinsic binding constant Kq. The inclusion of the term (qKp)Kp0 is dictated by the fact that the longer oligomers, although they can efficiently engage the total binding site, can also form qKp number of complexes, using only the proper DNAbinding site, although the conformation of the bound nucleic acid is different (see below). In general, the intrinsic binding constant for such complexes does not have to be equal to Kp and is designated as Kp0 . Because, Kp, Kq, and Kinter can be determined from the plot in Figure 7(c), there are only two unknown parameters in equation (16), Kp0 , and aq. For clarity and easy comparison, the data from Figure 7(a) have been plotted in Figure 7(b) together with the non-linear least-squares fit of the experimental curve, using equation (16), which provides Kp0 Z ð2:4G0:7Þ !104 MK1 , Kq Z 9:3ðG2:1Þ !104 MK1 , and aqZ17. Notice, the value of Kinter is the highest among all intrinsic affinities (Table 1). As discussed above, in the region where only the proper DNAbinding site engages in interactions with DNA, the binding site imposes the most rigid and constrained conformation on the nucleic acid, as reflected by the large and progressive change in the fluorescence increase of the etheno-oligomers (see below). In the intermediate region, the conformation of the nucleic acid is less constrained, as indicated by a large and abrupt decrease of DFmax (Table 1). The lack of the detectable statistical effect suggests that there is only one dominant state and/or orientation of the oligomers with w14 to 16 residues where the DNA can assume the most favorable conformation in the total DNA-binding site of pol X. The intrinsic binding constant, Kq Z ð9:3G2:1Þ! 104 MK1 , is lower by a factor of w4–5 than Kinter, characterizing the intermediate region, strongly suggesting that interactions with the total DNAbinding site impose energetically unfavorable constrains on the nucleic acid as compared to the intermediate region. As mentioned above, in such complexes, the nucleic acid interacts with the proper DNA-binding site and with an additional area within the total DNA-binding site. However, the nucleic acid is in a conformation imposed by the structure of the total DNA-binding site, leading to a lower value of the intrinsic binding constant. On the other hand, the presence of Kp0 indicates that oligomers, which can efficiently engage the entire total DNA-binding site of pol X, exist in equilibrium with the complexes with the intrinsic binding constant lower than Kp and, particularly, Kq. This additional equilibrium, embedded in equation (16), also explains why the extrapolation of the second linear region of the plot of KN as a function of N provides only a minimum value of the total sitesize. At KNZ0, the extrapolated value of Nq is not equal to qK1, as in the simpler equation (6), which

ASFV pol X–ssDNA Interactions

does not consider the presence of the additional equilibrium process, but to a lower value, Nq0 . Using equation (16), this lower value is defined as: Nq0 Z ðqK1ÞKðqKpÞ

Kp0 Kq

(17)

With the obtained values of qZ16, pZ7, Kq Z 9:3ðG2:1Þ !104 MK1 and Kp0 Z 2:4ðG0:7Þ !104 MK1 , equation (17) provides Nq0 z12:7 as compared to the experimental value of w13.2 (Figure 7(a)). The value of aqZ17 is slightly higher than the determined total site-size, qZ16G1, although it is in the error range of the determination of this parameter (G1). The 16-mer is on the border of the two intrinsic affinity regions in Figure 7(a). It is possible, particularly, in the context of the equilibrium between two states in pol X complexes with longer ssDNA oligomers (see above), that although the total site-size of the pol X–ssDNA complex is 16(G1) residues, the 16-mer alone is too short to assume a complete conformational state imposed by the total DNA-binding site. In other words, such behavior indicates that the enzyme affects the conformation of the nucleic acid outside of the occluded fragment of 16(G1) residues. The salt effect on the 16-mer binding to the polymerase indicates that the nucleic acid does engage an additional area of the total DNA-binding site, as the net number of the ions strongly released increases from w1 to w3, as compared to the 10-mer that binds exclusively to the proper DNA-binding site (Figure 9(d)). Nevertheless, the net number of released ions is significantly lower than the net release of w6 ions observed for the 20-mer, corroborating the conclusion that the ssDNA 16-mer alone does not efficiently form all interaction contacts with the total DNA-binding site of the polymerase. The structure of the ssDNA is different in the proper and total DNA-binding site of pol X A striking feature of the pol X binding to the etheno-derivative of the ssDNA oligomers, which can accommodate only one pol X molecule, is the dependence of the maximum change of the nucleic acid fluorescence, DFmax, upon the length of the oligomers (Figure 6(a) and (b)). For the oligomers ranging from seven to 12 nucleotide residues, the value of DFmax increases progressively with the length of the oligomer, while such correlation does not exist for the longer oligomers with 14 to 20 nucleotide residues in length. At the excitation wavelength applied (lexZ325 nm), the observed increase predominantly results from an increase of the quantum yield of the nucleic acid in the complex with pol X. A peculiar and very useful aspect of the fluorescence of 3A is that it is not very sensitive to the polarity of the environment.50,51 However, as compared to the free 3AMP, the emission of etheno-oligomers is quenched (eightto tenfold), primarily via the dynamic process of an intramolecular collision, i.e. it is principally affected

ASFV pol X–ssDNA Interactions

by the structure of the nucleic acid, not by the polarity of the environment.32–37,50,51 Thus, the usefulness of the etheno derivative oligomers in protein–nucleic acid studies results from the fact that their fluorescence increase is generated by the conformational change of the nucleic acid and reflects the increased separation and restricted mobility of the nucleic acid bases.32–37 The obtained data indicate that the conformation of the ssDNA oligomers that can only engage the proper DNA-binding site of pol X is different from the conformation of the oligomers with intermediate lengths, or those encompassing the total DNAbinding site of the enzyme. Progressive increase of DFmax indicates that as the length of the oligomer increases, up to w12 nucleotide residues, the separation and immobilization of the bases strongly increases. The energetic cost of such conformational changes would partially compensate favorable interactions with the binding site (see above). However, once the length of the oligomers reaches w14–16 residues, the structure of the bound DNA becomes less constrained, allowing the more energetically favorable conformation of the nucleic acid and more efficient dynamic quenching between the bases to occur.32–37,50,51 This conclusion is supported by the fact that the intrinsic binding constant characterizing the proper DNA-binding site, Kp, is significantly lower than Kinter. On the other hand, the analogous parameter for the total DNA-binding site, Kq, is only by a factor of w4 lower than Kinter (Tables 1 and 2), as the interactions with the additional interacting area in the total DNA-binding site are compensated by the less stringent structural requirement of the total site. Nevertheless, the nucleic acid that fully encompasses the total DNA-binding site does not form a single binding complex with pol X. Although the complex is dominated by Kq, it contains equilibrium with the complex characterized by Kp0 (see above). Pol X binds the ssDNA without detectable cooperative interactions The obtained data indicate that binding of pol X is characterized by uZ1.0G0.2, i.e. the binding does not involve any detectable cooperative interaction between the bound pol X molecules. This behavior is different from the rat or human pol b, where interactions with the ssDNA are characterized by, albeit weak, nevertheless, positive cooperative interactions characterized by uz2–16.15–18 On the other hand, strong positive cooperative interactions occur in the pol b binding to the dsDNA, where u reaches the value of w150, at high salt concentration.52 The difference between the strength of the cooperative interactions on two different conformations of the DNA indicates that pol b can bind the ss and dsDNA in different orientations enabled by the structure of the DNA-binding site located on the 8 kDa domain. We have previously indicated that the cooperative

137 interactions could play a role in finding the ssDNA gap of the damaged DNA by pol b, by allowing the enzyme to examine longer stretches of the dsDNA without engaging the catalytic 31 kDa domain in interactions with the nucleic acid and compensate for the decreasing intrinsic affinity of the polymerase at higher salt concentration. As such, they may also be a part of the mechanism to increase the processivity of the synthesis of the DNA on the gapped DNA.52 Cooperative interactions are not available for the much simpler ASFV pol X. The lack of cooperative interactions in the case of pol X strongly suggests that the viral repair polymerase should be less processive as indicated by biochemical studies.6,7,22–24 Also, pol X should be much less efficient in finding the ssDNA gap of the damaged DNA. Our laboratory is currently addressing these issues. Correlation between the functional structure of the total DNA-binding site of ASFV pol X and the structure of the enzyme The obtained functional structure of the total DNA-binding site of pol X and the observed complex binding process correlate very well with the determined NMR structure of the polymerase (Figure 1).22,23 Two major interaction areas that include positively charged helix aC or aE, already implicated in the DNA binding, are located on two spatially separated domains of the enzyme. Each ˚ long and can accommodate the area is w35 A ssDNA up to 10–12 nucleotide residues, but not 16 residues. Thus, each domain can bind the ssDNA oligomer containing seven to 12 residues without the interference from the other domain. The spatial separation of the aC and aE helices is such that the engagement of both areas in interactions with the nucleic acid requires a DNA fragment containing at least 16 nucleotide residues, clearly indicating that the total DNA-binding site must include both binding areas. In other words, the palm and the C-terminal domain contain the DNA-binding subsites of the total DNA-binding site of pol X. Each proposed binding area is located asymmetrically at one or the other end of the polymerase molecule, as deduced from the thermodynamic studies in this work (Figure 8(b)). The observed large difference in the salt effects on the 10-mer, versus 16-mer, and 20mer binding to the polymerase also corroborates with the presence of two binding areas that provide electrostatic contacts with the nucleic acid. However, at present, it is not known from the structural and thermodynamic data alone, which of these two areas is the proper DNA-binding site. It is tempting to assign the catalytic palm domain with its aC helix as the location of the proper DNA-binding site. However, the lesson from the studies of the analogous polymerase b indicates that this is the 8 kDa domain, but not the catalytic 31 kDa domain, that is the primary DNA-binding site of the enzyme.15–19,28–31

138 The DNA-binding subsites of ASFV pol X are not autonomous: comparison with mammalian pol b Recall, pol b binds the ssDNA in two binding modes, which differ in the number of occluded nucleotide residues, the (pol b)16 and (pol b)5 binding modes.15–18,28–31 In the (pol b)16 binding mode, both the 8 kDa and the 31 kDa domains are involved in interactions with the ssDNA, i.e. the total DNA-binding site of the enzyme is engaged in the complex. In the (pol b)5 binding mode, only the 8 kDa domain is engaged in interactions with the ssDNA. Similarly, pol b engages only its 8 kDa domain in interactions with the dsDNA.52 The very conspicuous difference between the binding of pol X and pol b to the ssDNA is the lack of the formation of different binding modes by pol X with longer ssDNAs, in which the enzyme occludes different numbers of nucleotide residues.15–18,28–31 The incapability of pol X to engage only one of its DNA-binding subsites in interactions with the DNA, indicates a very strict mutual orientation of the two binding subsites of the enzyme resulting in the lack of autonomy of the binding subsites. Once the nucleic acid is bound to one of the subsites, it is strictly oriented toward the other binding subsite that must protrude over the adjacent nucleotide residues. In other words, none of the pol X DNAbinding subsites can bind the DNA in different orientations that would enable the enzyme to change its orientation with respect to the nucleic acid lattice and diminish the number of the occluded nucleotides. The very strict mutual orientation of the binding subsites in pol X most probably results from the much less flexible connection between the two domains as compared to pol b. The 8 kDa domain of pol b is connected with the rest of the molecule by a flexible 14 amino acid residue linker, enabling the domain to assume multiple orientations with respect to the catalytic domain. The connection between the palm and C-terminal domain in pol X is much more restrictive, strongly suggesting much less independent movement of both domains with respect to each other, i.e. much less autonomy in the interactions with the nucleic acid (Figure 1). Additional functional implication Another fundamental difference between pol b and pol X is that when pol X engages the total DNAbinding site, the polymerase–DNA complex exists in equilibrium between two states characterized by intrinsic affinities Kq and Kp0 (see above). Although the state characterized by Kq dominates the equilibrium, nevertheless, the release of the DNA in transition from one complex to another will contribute to its less rigid structure in the binding site, as indicated by low fluorescence increase characterizing the complexes with the total binding site (see above). Moreover, one of these states, characterized by Kp0 , should have strongly diminished, if any, catalytic activity as the DNA is

ASFV pol X–ssDNA Interactions

partially released from the interactions with the total DNA-binding site of the enzyme (see above). Such equilibrium was not observed in the case of pol b.15–18,28–31 Less rigid structure and/or partial release of the DNA from interactions with the total binding site in the complex with pol X is most probably one of the factors behind the observed low processivity of the ASFV pol X6,7,24 (see above). Interactions between the polymerase and the DNA have been recognized as a key element of the fidelity of the enzymes.25–27 The most paramount aspect of these interactions, neglected for a long time, is that the complex with the nucleic acid defines to a large extent the structure of the binding site of the incoming nucleotide. The presence of the less rigid structure of the enzyme–DNA complex and particularly, the unique equilibrium between different states of the complex, may also contribute to the observed exceptionally low fidelity of pol X.6,7,24 Finally, the fact that the total DNA-binding site must engage in interactions with the DNA indicates that the mechanism of the nucleic acid recognition by pol X must be different as compared to the mechanism of the DNA recognition by the well-studied pol b.28–31 Such a mechanism cannot rely on fast search of the gap using an analog of the 8 kDa domain because such an analog is missing in the pol X structure.

Materials and Methods Reagents and buffers All chemicals were reagent grade. All solutions were made with distilled and deionized O18 MU (Milli-Q Plus) water. Buffer C is 10 mM sodium cacodylate adjusted to pH 7.0 with HCl, 1 mM MgCl2, 10% (v/v) glycerol. The temperature, concentration of NaCl, and MgCl2 in the buffer are indicated in the text. ASFV Pol X The plasmid harboring the gene of the ASFV pol X was a generous gift from Dr Maria L. Salas (Universidad Autonoma, Madrid, Spain). The gene of the enzyme has been placed in plasmid pET30a under control of the T7 polymerase system. Isolation and purification of the protein was performed, with slight modifications, as described.6,7 The protein was O98% pure as judged by SDS-PAGE with Coomasie brilliant blue staining. The concentration of the protein was determined spectrophotometrically using the extinction coefficient 3280Z 1.656!105 cmK1 MK1, obtained with the approach based on Edelhoch’s method.53,54 Nucleic acids All unmodified and modified ssDNA oligomers, dA(pA) 5, dA(pA)6 , dA(pA)7, dA(pA)9, dA(pA)11, dA(pA)13, dA(pA)15, dA(pA)17, dA(pA)18, dA(pA)19, 5 0 Fl-dT(pT)5, 5 0 Fl-dT(pT)6, 5 0 Fl-dT(pT)7, 5 0 Fl-dT(pT)8, 5 0 Fl-dT(pT)9, 5 0 Fl-dT(pT)14, 5 0 Fl-dT(pT)19, 5 0 Fl-dT(pT)24, 5 0 Fl-dT(pT)29, 5 0 Fl-dT(pT)39, 5 0 Fl-dT(pT)44, 5 0 Fl-dT(pT)49, 5 0 Fl-dT(pT)54, and unmodified ssDNA oligomers were

139

ASFV pol X–ssDNA Interactions

purchased from Midland Certified Reagents (Midland, Texas). The labeled ssDNA oligomer contains a fluorescent label, fluorescein, attached at the 5 0 end through phosphoramidate chemistry. Thus, the fluorescein residue is placed as an analog of an additional base in each modified DNA oligomer. The etheno-derivatives of homo-adenosine oligomers were obtained by modification with chloroacetaldehyde as described by us.35,36 Concentrations of all ssDNA oligomers have been spectrophotometrically determined.16,20,30,31,38,55,56

protein concentration, PF, must be the same. The values of SQi and PF are then related to the total protein concentrations, PT1 and PT2, and the total nucleic acid concentrations, NT1 and NT2, at the same value of DFobs, by: SQi Z

PF Z PTX KðSQi ÞNTX where xZ1 or 2.

Fluorescence measurements Steady-state fluorescence titrations were performed using the SLM-AMINCO 8100C. In order to avoid possible artifacts due to the fluorescence anisotropy of the sample, polarizers were placed in excitation and emission channels and set at 908 and 558 (magic angle), respectively15–21,28–36,38,57. The binding was followed by monitoring the fluorescence of the nucleic acid (lexZ 325 nm, lemZ410 nm). Computer fits were performed using Mathematica (Wolfram, IL) and KaleidaGraph (Synergy Software, PA). The nucleic acid relative fluorescence increase, DFobs, upon the pol X binding is defined as DFobs Z ðFi KFo Þ=Fo , where Fi is the fluorescence of the nucleic acid at a given titration point i, and Fo is the initial value of the fluorescence of the sample.15–21,28–36 Sedimentation velocity and equilibrium measurements

PT2 KPT1 NT2 KNT1

(19) (20)

15–18,32–39,40,60

Quantitative determination of binding isotherms for pol X association with unmodified ssDNAs using the MCT method Determination of the interaction parameters for the pol X–unmodified nucleic acid complexes has been performed using the MCT method.41 In this method, the fluorescent reference nucleic acid (e.g. etheno-derivative of a ssDNA oligomer) at total concentration, NTR , is titrated with the protein in the presence of a competing non-fluorescent nucleic acid of the total concentration, NTS . The total concentration of the protein, PT, at which the same value of the relative fluorescence increase, DFobs, of the reference nucleic acid is observed in the absence of the unmodified ssDNA, PTR, and in the presence of the unmodified ssDNA, PTS, are defined as:35,36,38,41 PTR Z ðSQi ÞR NTR C PF

(21)

PTS Z ðSQi ÞR NTR C ðSQi ÞS NTS C PF

(22)

Sedimentation velocity experiments were performed with an Optima XL-A analytical ultracentrifuge (Beckman Inc., Palo Alto, CA), using double-sector charcoal-filled 12 mm centerpieces. Sedimentation velocity scans were collected at the absorption band of the fluorescein residue (495 nm) for the fluorescein-modified ssDNA oligomers. Time derivative analyses of the sedimentation scans were performed with the software supplied by the manufacturer using averages of five to eight scans for each concentration.58–60 The reported values of sedimentation coefficients were corrected to s20,w for solvent viscosity and temperature to standard conditions.61–63

where (SQi)R, (SQi)S, and PF are the degree of binding of pol X on the reference nucleic acid, the degree of binding of the enzyme on the non-fluorescent, competing nucleic acid, and the free protein concentration, respectively. Solving the set of equations (21) and (22), for (SQi)S and PF provides the thermodynamically rigorous degree of binding of pol X on an unmodified ssDNA and the free protein concentration defined as:35,36,38,41

Quantitative determination of binding isotherms and stoichiometries of the pol X–ssDNA complexes

and:

In this work, we followed the binding of pol X to the ssDNAs by monitoring the fluorescence increase, DFobs, of the etheno-derivative of ssDNA oligomers upon the complex formation. Quantitative estimates of the total average degree of binding, SQi, (number of pol X molecules bound per oligomer) and the free protein concentration, PF, has been described in detail by us.15–18,32–39,40,60 Briefly, the experimentally observed DFobs has a contribution from each of the different possible i complexes of pol X with the ssDNA. Thus, the observed relative fluorescence increase is functionally related to SQi by:

Acknowledgements

DFobs Z SQi DFimax

(18)

where DFimax is the molecular parameter characterizing the maximum fluorescence increase of the nucleic acid with pol X bound in complex i. The same value of DFobs, obtained at two different total nucleic acid concentrations, NT1 and NT2, indicates the same physical state of the nucleic acid, i.e. the degree of binding, SQi, and the free

ðSQi ÞS Z

PTS KPTR NTS

PF Z PTS KðSQi ÞS NTS KðSQi ÞR NTR

(23)

(24)

We thank Mrs Betty Sordahl for reading the manuscript. This work was supported by NIH grant GM-58565 (to W.B.).

References 1. Yanez, R. J., Rodriguez, J. M., Nogal, M. L., Yuste, L., Enriquez, C., Rodriguez, J. F. & Vinuela, E. (1995). Analysis of the complete nucleotide sequence of African Swine Fever Virus. Virology, 208, 249–278.

140 2. Dixon, L. K., Abrams, C. C., Bowick, G., Goatley, L. C., Kay-Jackson, P. C., Chapman, D. et al. (2004). African swine fever virus proteins involved in evading host defense systems. Vet. Imm. Immunopathol. 100, 117–134. 3. Jouvnet, N. & Wileman, T. (2005). African swine fever virus infection disrupts centrosome assembly and function. J. Gen. Virol. 86, 589–594. 4. Takamatsu, H., Denyer, M. S., Oura, C., Childerstone, A., Andersen, J. K., Pullen, L. & Parkhouse, R. M. E. (1999). African Swine Fever Virus: a B cell-mitogenic virus in vitro and in vivo. J. Gen. Virol. 208, 249–278. 5. Rubio, D., Alejo, A., Rodriguez, I. & Salas, M. L. (2003). Polyprotein processing protease of African swine fever virus: purification and biochemical characterization. J. Virol. 77, 4444–4448. 6. Oliveros, M., Yanez, R. R., Salas, M. L., Slas, J., Vinuela, E. & Blanco, L. (1997). Characterization of an African fever virus 20 kDa DNA polymerase involved in DNA repair. J. Biol. Chem. 272, 30899–30910. 7. Garcia-Escudero, R., Garcia-Diaz, M., Salas, M.L., Blanco, L., & Salas, J. (2003). DNA Polymerase X of African fever virus: insertion fidelity on gapped DNA substrates and AP lyase activity support a role in base excision repair of viral DNA. J. Mol. Biol. 326, 1403–1412. 8. Knopf, C. W. (1998). Evolution of viral DNAdependent DNA polymerases. Virus Genes, 16, 47–58. 9. Hubscher, U., Nasheuer, H.-P. & Syvaoja, J. E. (2000). Eukaryotic DNA polymerases, a growing family. Trends Biochem. Sci. 25, 143–147. 10. Pelletier, H., Sawaya, M. R., Wolfle, W., Wilson, S. H. & Kraut, J. (1996). A structural basis for metal ion mutagenicity and nucleotide selectivity in human DNA polymerase b. Biochemistry, 35, 12762–12777. 11. Matsumoto, Y., Kim, K., Katz, D. S. & Feng, J. A. (1998). Catalytic center of DNA polymerase b for excision of deoxyribose phosphate groups. Biochemistry, 37, 6456–6464. 12. Masumoto, Y. & Kim, K. (1995). Excision of deoxyribose phosphate residues by DNA polymerase b during DNA repair. Science, 269, 699–702. 13. Sweasy, J. B. (2003). Fidelity mechanisms of DNA polymerase b. Prog. Nucl. Acid Res. 73, 137–168. 14. Budd, M. E. & Campbell, J. L. (1997). The roles of the eucaryotic DNA polymerases in DNA repair synthesis. Mutat. Res. 384, 157–167. 15. Rajendran, S., Jezewska, M. J. & Bujalowski, W. (1998). Human DNA polymerase b recognizes singlestranded DNA using two different binding modes. J. Biol. Chem. 273, 31021–31031. 16. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (1998). Transition between different binding modes in rat DNA polymerase b-ssDNA complexes. J. Mol. Biol. 284, 1113–1131. 17. Rajendran, S., Jezewska, M. J. & Bujalowski, W. (2001). Recognition of template primer and gapped DNA substrates by human DNA polymerase b. J. Mol. Biol. 308, 477–500. 18. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (2001). Energetics and specificity of Rat DNA polymerase b interactions with template-primer and gapped DNA substrates. J. Biol. Chem. 276, 16123–16136. 19. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (2001). Interactions of the 8-kDa Domain of Rat DNA Polymerase b with ssDNA. Biochemistry, 40, 3295–3307. 20. Jezewska, M. J., Galletto, R. & Bujalowski, W. (2003). Tertiary conformation of the template-primer and

ASFV pol X–ssDNA Interactions

21.

22.

23.

24. 25.

26.

27.

28.

29.

30.

31. 32.

33.

34.

gapped DNA substrates in complexes with rat polymerase b. Fluorescence energy transfer studies using the multiple donor-acceptor approach. Biochemistry, 42, 11864–11878. Galletto, R., Jezewska, M. J. & Bujalowski, W. (2005). Kinetic mechanism of rat polymerase b-dsDNA interactions. Fluorescence stopped-flow analysis of the cooperative ligand binding to a two-site onedimensional lattice. Biochemistry. 44, 1251–1267. Maciejewski, M., Shin, R., Pan, B., Marintchev, A., Denninger, A., Mullen, M. A. et al. (2001). Solution structure of a viral DNA repair polymerase. Nature Struct. Biol. 8, 936–941. Showwalter, A. K., Byeon, I.-J., Su, M.-I. & Tsai, M-D. (2003). Solution structure of a viral DNA polymerase X and evidence for mutagenic function. Nature Struct. Biol. 8, 942–946. Showalter, A. K. & Tsai, M.-D. (2001). A DNA polymerase with specificity for five base pairs. J. Am. Chem. Soc. 123, 1776–1777. Washington, M. T., Wolfle, W. T., Spratt, T. S., Prakash, L. & Prakash, S. (2003). Yeast DNA polymerase h makes functional contacts with the DNA minor groove only at the incoming nucleotide triphosphate. Proc. Natl Acad. Sci. USA, 100, 5113–5118. Trinacao, J., Johnson, R. E., Wolfle, W. T., Escalante, C. R., Prakash, S., Prakash, L. & Aggarwal, A. K. (2004). Dpo4 is hindered in extending a G-T mismatch by a reverse wobble. Nature Struct. Mol. Biol. 11, 457–462. Morales, J. C. & Kool, E. T. (2000). Functional hydrogen-bonding map of the minor groove binding tracks of six DNA polymerases. Biochemistry, 39, 12979–12988. Jezewska, M. J., Galletto, R. & Bujalowski, W. (2003). Rat polymerase b gapped DNA interactions: antagonistic effects of the 5 0 terminal PO4K group and magnesium on the enzyme binding to the gapped DNAs with different ssDNA gaps. Mol. Cell. Biophys. Biochem. 38, 125–160. Jezewska, M. J., Rajendran, S., Galletto, R. & Bujalowski, W. (2001). Kinetic mechanisms of rat polymerase b-ssDNA interactions: quantitative fluorescence stopped-flow analysis of the formation of the (Pol b)16 and (Pol b)5 binding mode. J. Mol. Biol. 313, 977–1002. Rajendran, S., Jezewska, M. J. & Bujalowski, W. (2001). Multiple-step kinetic mechanisms of the ssDNA recognition process by human polymerase b in its different ssDNA binding modes. Biochemistry, 40, 11794–11810. Jezewska, M. J., Galletto, R. & Bujalowski, W. (2002). Dynamics of gapped DNA recognition by human polymerase b. J. Biol. Chem. 277, 20316–20327. Bujalowski, W. & Jezewska, M. J. (1995). Interactions of Escherichia coli primary replicative helicase DnaB protein with single-stranded DNA. The nucleic acid does not wrap around the protein hexamer. Biochemistry, 34, 8513–8519. Jezewska, M. J., Kim, U.-S. & Bujalowski, W. (1996). Binding of Escherichia coli primary replicative helicase DnaB protein to single-stranded DNA. Long-range allosteric conformational changes within the protein hexamer. Biochemistry, 35, 2129–2145. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (1998). Complex of Escherichia coli primary replicative helicase DnaB protein with a replication fork. Recognition and structure. Biochemistry, 37, 3116–3136.

141

ASFV pol X–ssDNA Interactions

35. Jezewska, M., Rajendran, S. & Bujalowski, W. (2000). Escherichia coli helicase PriA protein-single stranded DNA complex. J. Biol. Chem. 275, 27865–27873. 36. Jezewska, M., Rajendran, S. & Bujalowski, W. (2000). Interactions of Escherichia coli replicative helicase PriA protein with single-stranded DNA. Biochemistry, 39, 10454–10467. 37. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (1998). Complex of Escherichia coli primary replicative helicase DnaB protein with a replication fork. Recognition and structure. Biochemistry, 37, 3116–3136. 38. Jezewska, M. J., Galletto, R. & Bujalowski, W. (2004). Interactions of the RepA helicase hexamer of plasmid RSF1010 with the ssDNA. Quantitative analysis of stoichiometries, intrinsic affinities, cooperativities, and heterogeneity of the total ssDNA-binding site. J. Mol. Biol. 343, 115–136. 39. Lohman, T. M. & Bujalowski, W. (1991). Thermodynamic methods for model-independent determination of equilibrium binding isotherms for protein–DNA interactions: spectroscopic approaches to monitor binding. Methods Enzymol. 208, 258–290. 40. Bujalowski, W. & Jezewska, M. J. (2000). Quantitative determination of equilibrium binding isotherms for multiple ligand-macromolecule interactions using spectroscopic methods. In Spectro-photometry, and spectrofluorimetry. A Practical Approach (Gore, M. G., ed.), pp. 141–165, Oxford University Press, Oxford, UK. 41. Jezewska, M. J. & Bujalowski, W. (1996). A general method of analysis of ligand binding to competing macromolecules using the spectroscopic signal originating from a reference macromolecule. Application to Escherichia coli replicative helicase DnaB proteinnucleic acid interactions. Biochemistry, 35, 2117–2128. 42. McGhee, J. D. & von Hippel, P. H. (1974). Theoretical aspects of DNA–protein interactions: cooperative and non-cooperative binding of large ligands to a onedimensional homogeneous lattice. J. Mol. Biol. 86, 469–489. 43. Epstein, I. R. (1978). Cooperative and non-cooperative binding of large ligands to a finite one-dimensional lattice. A model for ligand–oligonucleotide interactions. Biophys. Chem. 8, 327–339. 44. Bujalowski, W., Lohman, T. M. & Anderson, C. F. (1989). On the cooperative binding of large ligands to a one-dimensional homogeneous lattice: the generalized three-state lattice model. Biopolymers, 28, 1637–1643. 45. Galletto, R., Jezewska, M. J. & Bujalowski, W. (2004). Multi-step sequential mechanism of E. coli helicase PriA protein-ssDNA interactions. Kinetics and energetics of the active ssDNA-searching site of the enzyme. Biochemistry, 43, 11002–11016. 46. Hill, T.L., (1985). Cooperativity Theory in Biochemistry. Steady State and Equilibrium Systems, pp. 167– 234, Springer, New York. 47. Record, M. T., Lohman, T. M. & deHaseth, P. L. (1976). Ion effects on ligand–nucleic acid interactions. J. Mol. Biol. 107, 145–158. 48. Record, M. T., Jr, Anderson, C. F. & Lohman, T. M. (1978). Thermodynamic analysis of ion effects on

49. 50. 51.

52.

53. 54. 55.

56.

57. 58.

59.

60.

61. 62. 63.

the binding and conformational equilibria of proteins and nucleic acids: the roles of ion association or release, screening, and ion effects on water activity. Quart. Rev. Biophys. 11, 103–178. Habermann, R. (1998). Elementary Applied Partial Differential Equations, pp. 381–383, Prentice-Hall, Upper Saddle River, NJ. Baker, B. M., Vanderkooi, J. & Kallenbach, N. R. (1978). Base stacking in a fluorescent dinucleoside monophosphate: 3Ap3A. Biopolymers, 17, 1361–1372. Tolman, G. L., Barrio, J. R. & Leonard, N. J. (1974). Chloroacetaldehyde-modified dinucleoside phosphates. Dynamic fluorescence quenching and quenching due to intramolecular complexation. Biochemistry, 13, 4869–4878. Jezewska, M. J., Galletto, R. & Bujalowski, W. (2003). Rat polymerase b binds double-stranded DNA using exclusively the 8-kDa domain. Stoichiometries, intrinsic affinities, and cooperativities. Biochemistry, 42, 5955–5970. Edelhoch, H. (1967). Spectroscopic determination of tryptophan and tyrosine in proteins. Biochemistry, 6, 1948–1954. Gill, S. C. & von Hippel, P. H. (1989). Calculation of protein extinction coefficients from amino acid sequence data. Anal. Biochem. 182, 319–326. Jezewska, M. J., Rajendran, S. & Bujalowski, W. (1998). Functional and structural heterogeneity of the DNA binding of the E. coli primary replicative helicase DnaB protein. J. Biol. Chem. 273, 9058–9069. Jezewska, M. J., Rajendran, S., Bujalowska, D. & Bujalowski, W. (1998). Does ssDNA pass through the inner channel of the protein hexamer in the complex with the E. coli DnaB helicase? Fluorescence energy transfer studies. J. Biol. Chem. 273, 10515–10529. Azumi, T. & McGlynn, S. P. (1962). Polarisation of the luminescence of phenantrene. J. Chem. Phys. 37, 2413–2420. Correia, J. J., Chacko, B. M., Lam, S. S. & Lin, K. (2001). Sedimentation studies reveal a direct role of phosphorylation in Smad3:Smad4 homo- and heterodimerization. Biochemistry, 40, 1473–1482. Stafford, W., III (1992). Boundary analysis in sedimentation transport experiments: a procedure for obtaining sedimentation coefficient distributions using the time derivative of the concentration profile. Anal. Biochem. 203, 295–301. Galletto, R., Jezewska, M. J. & Bujalowski, W. (2003). Interactions of the Escherichia col DnaB helicase hexamer with the replication factor the DnaC protein. Effect of nucleotide cofactors and the ssDNA on protein-protein interactions and the topology of the complex. J. Mol. Biol. 329, 441–465. Tanford, C. (1961). Physical Chemistry of Macromolecules, pp. 364–390, Wiley, New York. Cantor, R. C. & Schimmel, P. R. (1980). Biophysical Chemistry, vol. II, W.H. Freeman, New York pp. 591–641. Bujalowski, W., Klonowska, M. M. & Jezewska, M. J. (1994). Oligomeric structure of Escherichia coli primary replicative helicase DnaB protein. J. Biol. Chem. 269, 31350–31358.

Edited by D. E. Draper (Received 9 September 2005; received in revised form 18 October 2005; accepted 22 October 2005) Available online 10 November 2005