Effect of Mutations at the Monomer-Monomer Interface of cAMP Receptor Protein on Specific DNA Binding*

To determine the thermodynamic role of binding of an operon to cAMP receptor protein (CRP) in the activation of transcription, isothermal titration calorimetry measurements were performed on the binding of three 40-base pair DNA sequences to the cyclic nucleoside complexes of CRP and its mutants at 296 K. The three 40-base pair sequences consisted of a consensus DNA (conDNA) duplex derived from the CRP-binding site sequences of the operons activated by CRP and two DNA sequences based on the CRP-binding site sequences of the lac operon (lacDNA) and of the gal operon (galDNA). The mutants of CRP consisted of a T127L mutant, a S128A mutant, and a mutant containing both mutations (CRP*) which not only alter the transcriptional activity of the CRP complexes but also are involved in the monomer-monomer interfacial interactions of the CRP dimer. The binding reactions of the DNA duplexes to the fully cNMP-ligated CRP-mutant complexes were endothermic with binding constants as high as 6.6 ± 1.1 × 106 m −1 (conDNA·CRP(cAMP)2). ConDNA binding to the unligated T127L and CRP* mutants was observed as well as conDNA and lacDNA binding to CRP with cAMP bound to only one monomer. The reduction of the binding constants with increase in KCl concentration indicated the formation of two ion pairs for the cAMP-ligated CRP and S128A complexes and four ion pairs for the cAMP-ligated T127L and CRP* complexes. Reduction of the DNA binding constants upon substitution of D2O for H2O in the buffer, the large heat capacity changes, and the enthalpy-entropy compensation exhibited by the binding reactions indicate the importance of dehydration in the binding reaction. Small angle neutron scattering measurements on the lacDNA·CRP(cAMP)2 complex in D2O/H2O mixtures show that the DNA is bent around the cAMP-ligated protein in solution.

The transcription of over 20 different operons encoded for enzymes involved in carbohydrate metabolism in Esherichia coli is activated by the binding of 3Ј,5Ј-cyclic adenosine monophosphate (cAMP) receptor protein (CRP) 1 to the operon. CRP is a 45,000 g mol Ϫ1 dimer, consisting of a 3Ј,5Ј-cyclic adenosine monophosphate (cAMP)-binding site in the amino-terminal domain and an ␣-helical DNA-binding region in the carboxylterminal domain of each monomer. Upon binding of cAMP to CRP, CRP undergoes a change to a conformation which binds specifically to the operon and results in bending of the operon around the RNA polymerase. The nature of this conformational change is not clear but may be induced by a stronger monomermonomer interaction in CRP since this interaction is enhanced in the presence of cAMP (1). X-ray crystallographic studies of the CRP(cAMP) 2 complex show that the sequence at the monomer-monomer interface is similar to that of the leucine zipper motif and that a serine (Ser 128 ) from the other monomer crosses over the interface and forms a hydrogen bond with the cAMP (2). More recent x-ray crystallographic studies of the CRP-(cAMP) 2 complexed with DNA (3,4) show that the DNA is symmetrically bound across the ␣-helices of the two monomers of CRP. Mutations at the monomer-monomer interface of CRP alter the activity of CRP in the cell as is evident by a doubly mutated (T127L/S128A) CRP mutant (CRP*) which activates transcription in the absence of cAMP, a CRP (T127L) mutant which also activates transcription in the presence of an analog of cAMP, 3Ј,5Ј-cyclic guanosine monophosphate (cGMP), and a CRP (S128A) mutant which does not activate transcription in vivo (5).
Since it has been shown from isothermal titration calorimetry (ITC) measurements that changes in the in vivo transcriptional activity arising from these interfacial mutations are not dependent on the binding affinity of the cyclic nucleotide monophosphate (cNMP) to the CRP-mutant (6), the transcriptional activity of cNMP-ligated CRP-mutants may depend on the influence of these interfacial mutations on the binding of the operon to CRP. It is, thus, important to determine the thermodynamics of DNA binding to CRP and its T127L, S128A, and CRP* mutants. Heretofore, indirect measurement techniques based on filter binding assays and shifts in gel electrophoresis bands have been employed to determine the operon binding affinities for CRP with often contradictory results. Equilibrium constants for binding of lac operon fragments to cAMP-ligated CRP range from 2.3 Ϯ 0.1 ϫ 10 9 M Ϫ1 for a 216-base pair fragment (7) to 8.4 ϫ 10 10 M -1 for a 203-base pair fragment (8) at ambient temperatures. Since it has been shown that the CRP-binding site on the lac operon requires a minimum of 28 base pairs for full binding affinity (9), the binding of shorter lac operon fragments consisting of 40 -41 base pairs was determined and, again, the binding constants ranged from 6 ϫ 10 6 M Ϫ1 in gel electrophoresis measurements at 310 K (9) to 3.2 Ϯ 0.1 ϫ 10 8 M Ϫ1 from filter binding assays (10). In addition, the DNA binding constants determined from these indirect techniques exhibit a large decrease with increase in salt concentration (11) in contrary to in vivo transcriptional assays which show high yields even at high salt concentrations (12).
To investigate the effect of monomeric interfacial mutations on specific DNA binding, ITC was employed to determine the binding constant (K b ) and the changes in enthalpy (⌬H b 0 ) and entropy (⌬S b 0 ) for the binding of DNA to cAMP, cGMP, and 3Ј,5Ј-cyclic inosine monophosphate (cIMP)-ligated CRP, T127L, S128A, and CRP*, DNA ϩ CRP(cNMP) 2

5Ј-GCAACGCAATAAGTGTGACATGGAATAAATTTTAGGCACCC-3Ј 3Ј-GTTGCGTTATTCACACTGTACCTTATTTAAAATCCGAGGGG-5Ј
GalDNA conDNA is a consensus sequence based on the sequences of the 20 operons while lacDNA is the CRP-binding site of the lac operon. The conDNA and lacDNA sequences are from Ebright et al. (10) and are identical from the 1 to 5Ј terminus and from 23 to the 3Ј terminus, using the numbering system from Ebright et al. (10). For comparison, galDNA was constructed so that bases from 1 to 21 are the same as the CRP-binding site of the gal operon (29). Although the DNA binding was determined at cyclic nucleoside concentrations where both sites of the CRPmutants were occupied, the binding of conDNA and lacDNA to CRP was also investigated at subsaturate concentrations of cAMP where significant concentrations of singly occupied CRP, CRP(cAMP), as calculated from previous measurements (13), were present to determine if the DNA binds to CRP(cAMP). In addition, the binding reactions of lacDNA and conDNA with CRP(cAMP) 2 were investigated as a function of salt concentration and with D 2 O substituted for water in the KPB buffer since dehydration has been shown to be involved in the binding reaction (7). A preliminary analysis of the bending of the lacDNA bound to the CRP(cAMP) 2 complex in water and D 2 O was performed by small angle neutron scattering (SANS) since the neutron scattering from the DNA can be characterized separately from the protein in H 2 O/D 2 O mixtures of the buffer. The results of this study will contribute to a better understanding of the thermodynamics of DNA binding to proteins since bending of the DNA, the ␣ helical structure of the DNA-binding site, and the presence of a leucine zipper between the protein subunits are characteristic properties of other DNA-protein complexes.

EXPERIMENTAL PROCEDURES
Materials-The DNA sequences were synthesized by Oligos, Inc. and purified by high performance liquid chromatography. The DNA was analyzed by capillary electrophoresis and an analytical ion exchange column and found to be at a purity level Ͼ90%. Gel electrophoresis of the single strands revealed essentially one intense band at 12,000 g mol Ϫ1 , the molecular mass of the strand. A procedure similar to that described by Ebright et al. (10) was followed to prepare the duplexes from the single strand DNA sequences. Briefly, milligram quantities of the lyophilized powdered DNA strands were dissolved in 4 ml of 10 mM Tris-HCl buffer containing 1 mM MgCl 2 and 0.5 M NaCl at pH 7.5. Equal amounts of the DNA strand and its complementary strand were mixed together, heated up slowly to 368 K, held at 368 K for about 10 min, and then cooled very slowly at a rate of 10 K h Ϫ1 to ambient temperature. Gel electrophoresis measurements on aliquots of the DNA solution showed the presence of only one band at the molecular mass of the duplex. The DNA duplex solutions were then concentrated up to 0.3-0.6 mM by centrifugation, dialyzed in the KPB buffer used for dialysis of the protein, and stored in a freezer at 253 K. Prior to the ITC measurements, the DNA duplex concentrations were determined from optical density measurements at 260 nm using the extinction coefficient of 1.32 ϫ 10 4 M Ϫ1 cm -1 per base pair (8). The production from E. coli of CRP and mutants and their purification have been described previously (5) and their activities were checked by an in vitro transcription assay as described by Zhang et al. (14). The concentration of the CRP-mutants was determined from UV measurements at 280 nm using an extinction coefficient of 3.5 ϫ 10 4 M Ϫ1 cm -1 (15). The potassium phosphate salts, KCl, Tris, MgCl 2 , and sodium salts of cAMP, cGMP, and cIMP were reagent grade from Sigma. The dithiothreitol was Ultra-pure brand from Life Technologies, Inc. and the NaEDTA was from Serva Co. The D 2 O was from Cambridge Isotope Laboratories.
ITC Measurements-All calorimetric titrations were preformed according to the methods of Wiseman et al. (16) and Schwarz et al. (17) using a Microcal Omega titration calorimeter. The Omega titration calorimeter consists of a matched pair of sample and reference vessels (1.374 ml) containing the protein solution in the phosphate buffer and the buffer solution, respectively. Four to 10-l aliquots of the DNA solution at concentrations ϫ 10 -20 the protein concentration of 0.01-0.05 mM in the sample vessel were added 3-4 min apart. A separate titration of the DNA solution into the buffer was performed to determine any DNA heat of dilution which was then subtracted from the heats obtained during the titration of the DNA solution into the protein solution. A nonlinear, least squares minimization performed by Microcal Origin scientific plotting software was used to fit the incremental heat of the ith titration (⌬Q(i)) of the total heat, Q t , to the total DNA titrant concentration, X t , according to the following equations (16,18), where n is the stoichiometry of the binding reaction, C t is the total protein concentration in the sample vessel, and V is the vessel volume. Binding entropies, ⌬S b 0 , were calculated using the following equation of thermodynamics, and r ϭ 8.31451 J mol Ϫ1 K Ϫ1 . Two methods were employed to determine the dependence of DNA binding on the subsaturate concentration of cAMP in the solution. In one method, aliquots of cAMP in KPB buffer were titrated into CRP solutions containing the DNA duplex and the amount of heat exchanged was monitored as a function of the relative CRP(cAMP) and CRP-(cAMP) 2 concentrations. Separate ITC runs with just CRP in the protein solution showed very little contribution of the heat of binding of cAMP to CRP under these titration conditions and any observed heat exchange would be from the binding of DNA to CRP(cAMP) 2 and perhaps to CRP(cAMP). In the second method, the amount of cAMP added to the CRP solution prior to the DNA titration was well below saturation so that the CRP contained about equal concentrations of CRP-(cAMP) and CRP(cAMP) 2 . In both methods, the concentrations of CRP-(cAMP) and CRP(cAMP) 2 were determined by solving the following equations, where K 1 and K 2 are the macroscopic cAMP to CRP binding constants from Gorshkova et al. (13), (Eq. 6) So that and [cAMP] t is the total cAMP concentration in the sample. The concentration of CRP(cAMP) 2 thus determined was then employed with the total DNA concentration in the sample to determine if the total observed amount of heat exchanged could be attributed solely to the CRP(cAMP) 2 complex. In this latter calculation, the binding constant and binding enthalpy for DNA binding to CRP(cAMP) 2 obtained at saturating concentrations of cAMP were employed. In the alternative procedure where DNA is titrated into solutions of CRP at subsaturate cAMP concentrations, the DNA binding constant and binding enthalpy were determined employing Equations 2 and 3 and compared with the DNA binding quantities determined at fully saturated CRP(cAMP) 2 concentrations. The standard uncertainties in the values of K b and ⌬H b 0 were determined by combining in a quadrature the standard deviations of several determinations of each value with estimates of their systematic standard uncertainties. The systematic standard uncertainties are 0.03 K b and 0.03 ⌬H b 0 resulting from uncertainty in the solution concentrations, 0.01 K b and 0.01⌬H b 0 resulting from uncertainty in the ITC calibration, and 0.005 K b and 0.005 ⌬H b 0 resulting from uncertainty in the titrant and cell volumes.
SANS Measurements-The lacDNA⅐CRP(cAMP) 2 product from an ITC measurement was concentrated up to a concentration of 0.085 mM and 0.5-ml aliquots of the solution were dialyzed in 100, 70, 30, and 0% (v/v) D 2 O KBS buffer. The neutron scattering intensity from each of the four solutions was measured as a function of the scattering angle on the NG3 30-m SANS instrument at the NIST Center for Neutron Research. At high water concentrations the major component of the scattering intensity is from the nucleic bases of the DNA and at high D 2 O concentrations the major component is from the cAMP-ligated CRP complex. Thus, by solving a set of simultaneous equations defined by the measured neutron intensities as a function of water concentration, the neutron intensities of the CRP and DNA can be separately identified (19). Details of the subsequent analysis of the data have been described previously by Krueger et al. (20). The x-ray crystallographic structure of the DNA complexed to the cAMP-ligated CRP determined by Schultz et al. (3) was used as the scattering model for the lacDNA⅐CRP(cAMP) 2 .
Solvent-accessible Surface Areas-The change in the solvent-accessible surface on formation of the DNA⅐CRP(cAMP) 2 complex (⌬A s ) was calculated with "NACCESS," a computer program from Hubbard (21), which conveniently divides the change into polar (⌬Asp) and non-polar area changes(⌬Asn). The ⌬A s was calculated as, Here A s (CRP in complex) refers to the solvent-accessible surface area of the CRP in the DNA⅐CRP(cAMP) 2 complex and A s (CRP(cAMP) 2 ) refers to the solvent-accessible area of the unligated CRP. Asp(CRP) and Asn(CRP) were calculated from the atomic coordinates from the Protein Data Bank file 3GAP; while Asp(complex) and Asn(complex) were calculated by using both PDB file 1CGP and 1RUN. Average solvent-accessible surface area changes were determined to be Ϫ626 (angstrom) 2 ϭ ⌬Asn(complex-CRP) and Ϫ2008 (angstrom) 2 ϭ ⌬Asp(complex-CRP) yielding a total solvent accessible surface area change of Ϫ2633(angstrom) 2.

RESULTS
Thermodynamics of DNA Binding-Results of a typical ITC measurement consisting of titrating 5.0-l aliquots of 0.250 mM conDNA into a solution of 0.020 mM CRP(cAMP) 2 complex at 295.6 K along with the binding isotherm are presented in Fig.  1. Prior to this titration, a 10 mM cAMP solution was first titrated into the 0.020 mM CRP solution to completely (99%) ligate the CRP with cAMP. The stoichiometry of the binding reaction as determined by the binding results was always close to one (0.90 Ϯ 0.05). Similar results were obtained with the other cNMP-ligated CRP-mutant complexes and are summarized in Table I for all 12 complexes including binding of conDNA to unligated L127 and unligated CRP*. Binding of the conDNA duplex with unligated CRP and S128A was not observed. The thermodynamic quantities in Table I are average values of at least two ITC scans with different concentrations of ligand and protein and the average stoichiometry was again close to one. As shown in Table I, all the conDNA binding reactions are endothermic with binding enthalpies ranging from 84.1 Ϯ 0.3 kJ mol Ϫ1 for binding to CRP(cAMP) 2 to 166 Ϯ 12 kJ mol Ϫ1 for binding to L127(cGMP) 2 and binding constants ranging from 8.1 Ϯ 0.9 ϫ 10 4 M Ϫ1 for binding to S128A(cGMP) 2 to 6.6 Ϯ 1.1 ϫ 10 6 M Ϫ1 for binding to CRP(cAMP) 2 at room temperature. The enthalpy change for binding of conDNA to CRP(cAMP) 2 is close to the literature value of 80 Ϯ 17 kJ mol Ϫ1 obtained from a van't Hoff analysis of the binding constants determined from filter binding assays in a 10 mM MOPS buffer at pH 7.3 containing 0.2 mM NaCl, 0.1 mM dithiothreitol, 50 g ml -1 serum albumin, and 0.2 mM cAMP, while the corresponding binding constant is much higher, 3.9 Ϯ 0.3 ϫ 10 10 M Ϫ1 (10). The reason for this large discrepancy is unknown except to emphasize that ITC is a direct measurement of the binding event. The heat capacity change accompanying the conDNA to cAMP-CRP binding reaction, determined from the dependence of the binding enthalpy on the temperature (Fig. 2), is Ϫ4.0 Ϯ 0.4 kJ mol Ϫ1 K Ϫ1 , close to an estimate of Ϫ2.85 kJ mol Ϫ1 K Ϫ1 (22) from filter binding assays by Ebright et al. (10).
For comparison, the thermodynamics of binding of the lacDNA and galDNA duplexes to the fully cAMP-ligated CRP, T127L, S128A, and CRP* complexes and to T127L(cGMP) 2 and CRP*(cIMP) 2 were determined from ITC measurements and are presented in Table II Table II. However, the binding constant of a 41base pair lacDNA fragment of 6 ϫ 10 6 M Ϫ1 determined at 310 K from gel electrophoresis measurements (9) is also less by a factor of 50 than the value of 3.2 Ϯ 0.1 ϫ 10 8 M Ϫ1 determined by Ebright et al. (10). For galDNA, the binding constants range from 1.7 Ϯ 0.4 ϫ 10 4 M Ϫ1 for binding to CRP*(cIMP) 2 to 1.11 Ϯ 0.21 ϫ 10 6 M Ϫ1 for binding to CRP(cAMP) 2 , while the binding enthalpies ranged from 85 Ϯ 3 kJ mol Ϫ1 for binding to CRP-(cAMP) 2 to 174 Ϯ 4 kJ mol Ϫ1 for binding to T127L(cAMP) 2 . For lacDNA binding to CRP(cAMP) 2 , the heat capacity change of Ϫ5.7 Ϯ 1.6 kJ mol Ϫ1 K Ϫ1 is the same as the heat capacity change observed for binding of conDNA to CRP(cAMP) 2 (Ϫ4.0 Ϯ 0.4 kJ mol Ϫ1 K Ϫ1 ). All the DNA binding reactions exhibited enthalpy-entropy compensation as shown in Fig. 5 where the slopes of the T⌬S b 0 versus ⌬H b 0 plots are 0.95 Ϯ 0.5 for conDNA, 0.95 Ϯ 0.06 for lacDNA, and 0.98 Ϯ 0.06 for galDNA binding to CRP-mutant(cNMP) 2 . A comparison of all the DNA binding constants is shown in Fig. 6 where it is observed that the naturally occurring sequences exhibit a greater range of values, i.e. specificity than the synthetic conDNA.
Thermodynamics of DNA Binding to CRP as a Function of cAMP Concentration-The results of titrating 0.435 mM cAMP in KPB into a solution containing 0.023 mM CRP and 0.023 mM conDNA in KPB are presented in the form of a binding isotherm in Fig. 7 along with the calculated concentrations of CRP(cAMP) and CRP(cAMP) 2 , the calculated heats absorbed by the CRP(cAMP) 2 alone, and the calculated heats absorbed by both CRP(cAMP) and CRP(cAMP) 2 using the binding constant and binding enthalpy values from Table I. Similar results were obtained with titrating the cAMP solution into a solution containing lacDNA instead of the conDNA. As shown in Fig. 7, at least 90% of the total amount of heat absorbed with each addition of the cAMP solution can only be attributed to DNA binding to both CRP(cAMP) 2 (Table II). An ITC measurement under similar concentration conditions with conDNA yielded a binding constant of 6.3 Ϯ 3.7 ϫ 10 6 M Ϫ1 and a binding enthalpy of 88 Ϯ 2 kJ mol Ϫ1 close to the results in Table I. The ITC results, thus, show that the DNA duplex can also bind to CRP(cAMP) with the same binding affinity and enthalpy as to the CRP(cAMP) 2 species. Equal binding affinities of specific DNA binding to both CRP-(cAMP) and CRP(cAMP) 2 were also observed earlier by Takahashi et al. (23) in equilibrium dialysis measurements on specific DNA binding to CRP as a function of cAMP concentration.
Solvent Effects on the Thermodynamics of DNA Binding-Since DNA binding to CRP involves electrostatic complementarity (24), ITC measurements were performed on the binding of conDNA to CRP-mutants as a function of KCl concentration from 0.15 to 0.60 M and are summarized in Table III  only a partial compensating increase in the binding entropy. The decrease in the binding constant could result from competitive anion binding to the DNA-binding site on CRP or electrostatic screening of the ion pair interactions between the DNA bases and the CRP amino acid residues. For B-DNA binding to proteins based solely on ion pair interactions, it has been shown (25) that, where K b (1.0 M KCl) is the binding constant at 1.0 M KCl and m is the number of CRP-mutant/DNA ion pairs. The results in Table III (20). The radius of gyration for the lacDNA duplex in the complex was 32 Ϯ 1 angstroms which is close to the value of 29.7 angstroms calculated for the DNA component in the x-ray crystal structure of the DNA⅐CRP(cAMP) 2 complex (3). The distance distribution function, P(r), (20) for the lacDNA component of the lacDNA⅐CRP(cAMP) 2 complex in the 0.085 mM solution is shown in Fig. 9 along with the distance distribution function of the DNA component of the x-ray crystal structure. The shape of the two curves are similar in that they both contain a maximum in distance distribution function at a radius around 20 angstroms, distance distribution function decreases slowly to a shoulder around 60 angstroms, and then distance distribution function decreases rapidly to 0 -95 angstroms. This last rapid decrease occurs when "rigid" rods are bent into a "U" shape, thereby eliminating the larger distances in the distribution function. However, the lacDNA in the lacDNA⅐CRP(cAMP) 2 complex appears to exhibit a broader distribution around the maximum than the DNA in the DNA⅐CRP(cAMP) 2 crystal structure. This is probably due to a globular component in the distance distribution function from aggregates of the complex in the 70 and 100% (v/v) D 2 O-buffered solutions. It can be concluded that the SANS data show that the DNA in the lacDNA⅐CRP(cAMP) 2 complex is bent, independent of the concentration of D 2 O in the sample.

DISCUSSION
As observed for other protein-DNA associations (26,27), the specific binding of DNA to cNMP-ligated CRP-mutant complexes is endothermic, thus, entropically driven, and involves dehydration. The binding enthalpy can be attributed to endothermic bending of the DNA (⌬H bn 0 ) and to the formation of ionic pair interactions, van der Waals interactions, and hydrogen bonding interactions between the amino acid residues at the CRP-binding site and the DNA bases (direct interactions) and the ribose-phosphate backbone (indirect interactions) all of which would be exothermic (⌬H bx 0 ), Contributions to ⌬H bn 0 would principally arise from the energy absorbed in the bending of the DNA duplex about 90 o around the CRP and would expected to be the same for the different DNA sequences bending around the CRP(cAMP) 2 complex. Differences in the direct and indirect interactions based on the x-ray crystallographic structure of the DNA⅐CRP(cAMP) 2 complex by Parkinson et al. (4) are itemized in Table V. Since the number of A, T, G, and C bases is the same for conDNA, lacDNA, and galDNA, it can be assumed that the interaction between the isolated DNA duplexes and water is approximately the same for all three DNA sequences and, thus, differences in the intermolecular binding interactions can be conceptualized in terms of differences between the CRP-binding site sequences of the DNA duplexes.
Attempts have been made to correlate the binding affinities (Ϫ⌬G b 0 ) of the catabolite operons with the 22-base pair sequence of the CRP-binding site in the operon (28,29). In all the catabolite operons, the 4 TGTGA sequence is conserved and there is a center of pseudo 2-fold symmetry between bases 11 and 12 (28). ConDNA with the highest binding affinity exhibits perfect 2-fold symmetry and this 2-fold symmetry is reduced in the more naturally occurring lacDNA and galDNA duplexes which also exhibit weaker binding affinities. By determining the frequency a specific nucleotide appeared in each of the 1-22 bases of the CRP-binding site in 26 operons, Berg and von Hippel (29) derived a quantitative model predicting the binding affinity of an operon from the difference between its CRPbinding site sequence and that of conDNA. The greater the deviation of each half of the 22-base sequence from the perfect 2-fold sequence of conDNA, i.e. the lower the symmetry, the weaker the CRP binding affinity. However, their model predicted a weaker binding affinity for galDNA than for lacDNA, contrary to the ITC results. In addition to a correlation between sequence and binding affinity, changes in the relative number of direct and indirect interactions between the sequences, as determined from the known crystal structures of the DNA⅐ CRP(cAMP) 2 complexes (3,4), would be also expected to contribute to changes in the binding affinities. As shown in Table  V, the number of base changes between conDNA and lacDNA or galDNA are seven and more of the base changes in galDNA are involved in direct interactions with CRP, while more of the base changes in lacDNA are involved in indirect interactions. The lower binding affinity for lacDNA could be accounted for by the change in more of the indirect interactions between the phosphate backbone. This implies that the indirect interactions contribute more than the direct interactions to the binding affinity.
The variation of the binding affinities of a particular DNA sequence to the different cNMP⅐CRP mutant complexes results   (3,4). Since cAMP enhances the monomer-monomer interaction in CRP (1), the closed form must be favored by a stronger interaction between the monomers in CRP. The amino acid sequence along the monomer-monomer helical interface exhibits an almost perfect leucine zipper motif consisting of a heptad repeat of leucine residues along one face of the helix with the exception of a Met 120 and Thr 127 (6). In the T127L mutant, Thr 127 is replaced by Leu, thereby creating a more perfect leucine zipper which would lead to stronger association between the monomeric units as is observed in other leucine zipper proteins (30). Hence, the stronger association in the T127L and CRP* mutants would alter the topography of the DNA-binding site more toward the closed form. The tendency toward the closed form in CRP containing the T127L CRP mutation is further substantiated by SANS measurements (20) which show, unlike CRP, little structural difference between the unligated and cAMP-ligated forms of CRP*. According to a structural model for DNA binding proposed by Garges and Adhya (31), the closed form is more favorably disposed toward DNA binding and, thus, conDNA binds to the unligated T127L and CRP*. The difference in ⌬G b 0 of Ϫ30 kJ mol -1 for conDNA binding to T127L and CRP* is about 80% of the average ⌬G b 0 of Ϫ37.8 kJ mol -1 for conDNA bonding to the CRP(cAMP) 2 and L127(cAMP) 2 complexes. This "tightness" of the monomermonomer interaction in T127L and CRP* can account for contributing at least 80% to the DNA binding affinity. Interestingly, the S128A mutation in S128A and CRP eliminates the hydrogen bond to cAMP formed in the crossover of the Ser 128 residue from the other monomer, thereby reducing the monomer-monomer interaction. This reduction results in a decrease of the DNA binding affinity in S128A and CRP* by about 4 kJ mol -1 . Finally, DNA can also bind to the singly cAMP occupied CRP and, thus only one cAMP is apparently enough to strengthen the monomer-monomer association in CRP, perhaps through formation of the Ser 128 -cAMP hydrogen bond.
The binding entropy can be attributed to entropy changes arising from hydrophobic interactions (dehydration) (⌬S bh 0 ), and the conformational change in the DNA (⌬S bc 0 ), as well as smaller conformational changes in the DNA-binding site. Following the treatment of Spolar and Record (22), at a temperature (Ts) where the binding entropy is zero then,  the entropy loss due to the loss of rotational and translational degrees of freedom upon binding is the same for protein-ligand interactions, ⌬S brt 0 ϭ Ϫ209 J mol -1 K Ϫ1 , and the entropy change due to the number of ion pairs formed (z ϭ 2) and the salt concentration [KCl] ϭ 0.15 mol liter -1 , ⌬S be 0 , is Ϫ0.88 zRln[KCl] ϭ 27 J mol -1 K Ϫ1 . The change in the solvent accessible surface area of the nonpolar residues is Ϫ626 (angstroms) and, thus, ⌬S bb 0 ͑T͒ ϭ 1.34͑ Ϫ 626͒ln͕T/386͖ ϭ Ϫ 852 ln͕T/386͖J mol Ϫ1 K Ϫ1 (Eq. 16) so that at a Ts ϭ 328 K determined from ⌬C p ϭ Ϫ4.0 Ϯ 0.4 kJ mol Ϫ1 K Ϫ1 and ⌬S b 0 ϭ 415 J mol -1 K Ϫ1 for conDNA binding to CRP(cAMP) 2 , the entropy of bending of the DNA is estimated to contribute only 139 J mol -1 K Ϫ1 to the binding entropy of 415 J mol -1 K Ϫ1 . Thus, the main contribution to the binding entropy must be from the desolvation process involving the removal of water between the DNA and the polar residues at the CRP interface upon binding which had not been taken into account in Equation 15.
Dehydration upon DNA binding to cAMP-ligated CRP has been observed by Vossen et al. (7) and is evident in DNA binding to the other cNMP-ligated CRP-mutants by the reduction of the binding constant in D 2 O due to an increase in the endothermicity of the binding reaction. The increase in binding enthalpy of 43 Ϯ 2 kJ mol -1 for conDNA and 34 Ϯ 9 kJ mol -1 for lacDNA arises from enhanced stabilization of the DNA and CRP in D 2 O relative to the DNA⅐CRP(cAMP) 2 complex in D 2 O. If Ϫ8.8 J mol Ϫ1 (angstrom) -1 accompanies the transfer of nonpolar amino acid groups from H 2 O to D 2 O (32, 33), then the enhanced stabilization of the nonpolar binding surface area (626 (angstrom) 2 on the protein alone in D 2 O contributes about 5.5 kJ mol Ϫ1 to the difference between the binding enthalpies in D 2 O and H 2 O. Apparently, as shown in the entropy calculation, the larger buried polar surface area on the protein of 2008 (angstrom) 2 must also contribute to the enhanced interaction between the protein and solvent in D 2 O. Also, if the effect of D 2 O on the solvent-reactant interaction involved solely solvent reorganization, the binding constant would be unaffected (34) because of enthalpy-entropy compensation. The release of water upon complex formation is also evident in the enthalpyentropy compensation exhibited by the binding reaction at ambient temperatures and by the temperature dependence of the DNA-binding reaction where the large negative heat capacity changes minimize changes in ⌬G b 0 as the temperature is increased from 288 K to 303 K.  9. Distance distribution function of lacDNA in lacDNA⅐CRP-(cAMP) 2 complex from SANS data (E) compared with that calculated for the DNA components in the crystal structure (line).