Thermodynamics of dT--dT base pair mismatching in linear DNA duplexes and three-arm DNA junctions.

We have used a combination of magnetic-suspension densimetry and calorimetry to derive complete thermodynamic profiles, including volume changes, for the formation of linear DNA duplexes and three-arm branched DNA junctions, from their component strands, with and without dT-dT mismatches. The formation of each type of complex at 20 degrees C is accompanied by a favorable free energy, with a favorable enthalpy term partially compensated by an unfavorable entropy. Formation is associated also with net uptake of water molecules. Using the formation of the fully-paired linear duplex or three-arm junction as reference states, we can establish a thermodynamic cycle in which the contribution of the single-strand species cancels. From this cycle, we determine that substitution of dA for dT has a differential free energy of deltadeltaG degrees of +2.4 kcal mol(-1) for mismatched duplex and +2.0 kcal mol(-1) (on the average) for the mismatched junction. These unfavorable differential free energies result from an unfavorable enthalpy, partially compensated by a favorable entropy, and a negative deltadeltaV. The free energies in the two cases have signs opposed to those of deltadeltaV, a situation that implicates hydration changes in creating the mismatch. When the deltadeltaV terms are normalized by the total number of base pairs involved, the immobilization of structural water molecules (and/or substitution of electrostricted for hydrophobic water molecules) is about 7 times greater for junctions than duplexes. This is consistent with more extensive hydrophobic hydration of branched DNA structures than of duplexes.

Differential hydration of dA.dT base pairs in parallel-stranded DNA relative to antiparallel DNA.

Parallel-stranded DNA is a novel double-stranded helical form of DNA. Its secondary structure is established by reverse Watson-Crick base pairing between the bases of the complementary strands forming a double helix with equivalent grooves. We have used a combination of magnetic suspension densitometry and isothermal titration calorimetry to obtain complete thermodynamic profiles for the formation of two DNA 25 mer duplexes. The duplexes contain exclusively dA.dT base pairs in either parallel (ps-d1.D2) or antiparallel (aps-D1.D3) orientation. At 15 degrees C, the formation of each duplex is accompanied by favorable free-energy terms resulting from the partial compensation of favorable exothermic enthalpies and unfavorable entropies and by an uptake of both counterions and water molecules. By taking into account the contribution of single-strand base-stacking interactions and using the formation of the aps-D1.D3 duplex as a reference state to establish a thermodynamic cycle in which the similar single strands cancel out, we obtained a delta delta G zero term of +10 kcal mol-1 duplex formed that results from a partial differential enthalpy-entropy compensation of +32 kcal mol-1 and a delta delta V of 257 mL mol-1. The positive sign of this enthalpy-entropy compensation together with the marginal differential counterion uptake of 0.2 mol of Na+/mol of duplex is characteristic of processes driven by differential hydration and strongly suggests that the parallel duplex is much less hydrated than its antiparallel counterpart by approximately 4 mol of water/mol of base pair.

Volume changes correlate with entropies and enthalpies in the formation of nucleic acid homoduplexes: differential hydration of A and B conformations.

We have used a combination of densimetric, calorimetric, and uv absorption techniques to obtain a complete thermodynamic characterization for the formation of nucleic acid homoduplexes of known sequence and conformation. The volume change delta V accompanying the formation of four duplexes was interpreted to reflect changes in hydration based on the electrostriction phenomenon. In 10 mM sodium phosphate buffer at pH 7, the magnitude of the measured delta V's ranged from -2.0 to +7.2 ml/mol base pair and followed the order of poly(rA).poly(dT) approximately poly(dA).poly(dT) < poly(rA).poly(dU) approximately poly(rA).poly(rU). Inclusion of 100 mM NaCl in the same buffer gave the range of -17.4 to -2.3 mL/mol base pair and the following order: poly(dA).poly(dT) < poly(rA).poly(dT) < poly(rA).poly(rU) approximately poly(rA).polyr(dU). Standard thermodynamic profiles of forming these duplexes from their corresponding complementary single strands indicated similar free energies that resulted from the compensation of favorable enthalpies with unfavorable entropies along with a similar counterion uptake at both ionic strengths. The differences in these compensating effects of entropy and enthalpy correlated very well with the volume change measurements in a manner suggesting that the homoduplexes in the B conformation are more hydrated than are those in the A conformation. Moreover, the increased thermal stability of these homoduplexes resulted from an increase in the salt concentration corresponding to larger hydration levels as reflected by the delta V results.

Differential hydration of homopurine sequences relative to alternating purine/pyrimidine sequences.

The minor groove ligand distamycin A has been used to probe the relative hydration of the minor groove of eight synthetic polynucleotides of known sequence and composition. A combination of densimetric, calorimetric, and temperature-dependent spectroscopic techniques have been used to obtain complete thermodynamic profiles (delta Gzero, delta Hzero, delta Szero, and delta Vzero) for the association of distamycin A to all polymer duplexes. In 10 mM phosphate buffer, pH 7, binding of the drug to each of the polymeric duplexes resulted in characteristic negative changes in both the volume and enthalpy. Although the binding constants were found to be identical for pairs of isomer polynucleotides having identical compositions but different sequences, the values of delta Hzero, delta Szero, and delta Vzero of each such pair were remarkably different. The entropy changes were found to roughly parallel the volume changes; no such trend was seen between delta Hzero and delta Vzero. The data support the hypothesis that the volume changes observed for these systems reflect the coulombic-hydration contribution to the entropy. The heteropolymer duplexes generated much larger exothermic contributions, less favorable entropies and larger volume contractions than did the corresponding homopolymer duplexes of identical composition, and strongly suggest that polynucleotides with homopurine sequences are more hydrated than polynucleotides with alternating purine/pyrimidine sequences. In addition, it was found that duplexes containing guanine sharply reduced the affinity for the drug, also lowering the exothermicity but raising the entropy. This may be explained by the presence of an amino group in the minor groove that prevents hydrogen bonding. Substitution of the guanine with inosine reversed this trend in the thermodynamic properties. Furthermore, substitution of poly(dA) for poly(rA) in a duplex produced a similar reduction in the affinity, while raising the exothermic contribution and greatly reducing the favorable entropy effect in agreement with an apparent increase in the hydration state.

Differential hydration of dA.dT base pairing and dA and dT bulges in deoxyoligonucleotides.

The role of water in the formation of stable duplexes of nucleic acids is being studied by determining the concurrent volume change, heats, and counterion uptake that accompany the duplexation process. The variability of the volume contraction that we have observed in the formation of a variety of homoduplexes suggests that sequence and conformation acutely affect the degree of hydration. We have used a combination of densimetric and calorimetric techniques to measure the change in volume and enthalpy resulting from the mixing of two complementary strands to form (a) fully paired duplexes with 10 or 11 base pairs and (b) bulged decameric duplexes with an extra dA or dT unmatched residue. We also monitored absorbance vs temperature profiles as a function of strand and salt concentration for all four duplexes. Relative to the decamer duplex, insertion of an extra dA.dT base pair to form an undecamer duplex results in a favorable enthalpy of -5.6 kcal/mol that is nearly compensated by an unfavorable entropy term of -5.1 kcal/mol. This enthalpy difference correlates with a differential uptake of water molecules, corresponding to an additional hydration of 16 mol of water molecules/mol of base pair. Relative to the fully paired duplexes, both bulged duplexes are 12-16 degrees C less stable and exhibit marginally larger counterion uptake on forming the duplex. The enthalpy change is slightly lower for the T-bulge duplex and less still for the A-bulge duplex. The volume change results indicate that an unmatched residue increases the amount of coulombic and/or structural hydration. The combined results strongly suggest that the destabilizing forces in bulged duplexes are partially compensated by an increase in hydration levels.

Probing the hydration of the minor groove of A.T synthetic DNA polymers by volume and heat changes.

The minor-groove ligand netropsin provides a sensitive probe of the hydration difference between poly(dA).poly(dT) and poly[d(AT)].poly[d(AT)]. We have measured the volume change delta V accompanying binding of netropsin to these polymers, using an improved magnetic suspension densimeter. For poly(dA).poly(dT) we find delta V = +97 mL/mol of bound netropsin at pH 7.0 and 10 mM sodium phosphate buffer. For poly[d(AT)].poly[d(AT)] we find delta V = -16 mL/mol of bound netropsin. This striking differential effect suggests that the poly(dA).poly(dT) duplex compresses more water (or is more extensively hydrated). From our enthalpy and entropy results we estimate the approximately 10 water molecules, immobilized in the minor groove of this system, are displaced by each netropsin bound. The volume increase, however, is substantially larger than can be explained by a simple melting of these immobilized water molecules in the minor groove. A decompression of at least 40 water molecules must attend the complexation to the poly(dA).poly(dT) duplex. This suggests that the conformation change attending the binding of the drug to this polymer duplex causes a further dehydration, whereas no such change in dehydration and configuration for the heteropolymer system is indicated.

Volume changes in the binding of lanthanides to peptide analogues of loop II of calmodulin.

The solution expansion accompanying coordination of lanthanide ions to synthetic peptide analogues of a metal-binding loop in calmodulin was determined by a density method. This study was designed to further test the hypothesis that the nonlinear expansions observed upon sequential addition of Ca2+ to intracellular calcium-binding proteins reflect principally upon the coordination event at specific binding sequences. Three peptides of 13 residues each were synthesized as analogues of binding loop II in mammalian calmodulin: Peptide I was the native analogue; peptide II contained an aspartyl in place of an asparaginyl residue at position 5 from the N-terminus; for peptide III, the aspartyl residue in position 3 of the native analogue was interchanged with the asparaginyl residue in position 5. Thus, the number of charged-oxygen donor atoms for coordination was the same in I and in III, but the latter peptide could permit two pairs of acidic groups to converge toward the metal ion as in some loops of these proteins. The observed expansions with different lanthanide ions to the same peptide varied appreciably, suggesting dissimilar structures [Gariépy et al. (1983) Biochemistry 22, 1765-1772]; coordination to the simpler tetracarboxylate sequestrants, on the other hand, generated an expansion profile approximately as expected from the properties of the lanthanide series. The largest expansions were generated with peptide II (having the additional acidic group) for all lanthanides tested.(ABSTRACT TRUNCATED AT 250 WORDS)

Sequential volume changes on a small sample by density.

A procedure is described for generating a sequence of volume changes by density on the same sample of a solution. The entire volume-change profile over the reactant concentration range of interest can be accomplished with 0.1-0.2 mumol of a protein in 100 microliters of sample. The calcium-binding proteins, calmodulin and skeletal troponin-C, were employed to determine the volume increases attending the sequential addition of calcium ions. Both volume-change profiles exhibited nonlinear increases over the first four equivalents of calcium ion added.

Volume changes upon addition of Ca2+ to calmodulin: Ca2+-calmodulin conformational states.

Measurement of the volume change by a rapid density method upon sequential addition of calcium ion to calmodulin showed relatively large, nonuniform increases for the first 4 moles Ca2+ per mole calmodulin. Substantially larger volume increases (approximately 15 ml/mol protein) were observed upon addition of the second and fourth moles Ca2+ relative to the first and third moles added per mole calmodulin. A total volume increase of approximately 170 ml/mol protein attended the addition of 4 moles Ca2+, as expected for multidentate carboxylate coordination to metal ion. Marginal changes in volume were observed upon further additions, the data showing a remarkably sharp transition after [Ca2+]/[calmodulin] = 4. The results are consistent with an ordered binding of Ca2+ in which pair-wise additions produce similar volume changes; the volume change behavior, however, does not indicate an absence of distinct conformational states for a Ca2+(1)-calmodulin and a Ca2+(3)-calmodulin complex as has been proposed on the basis of 1H-NMR evidences.

Simultaneous measurements of viscosity and density in solutions undergoing change.

The magnetic suspension densimeter-viscometer has been modified so that the variations of viscosity and density in solutions undergoing change are determined continuously as a function of time. The changes in values for both of these properties are also obtained when solutions of varying composition are passed through the sample cell (<0.5 ml). The new method measures the torque on the magnetically suspended buoy which is held stationary in a stable velocity gradient; this procedure also facilitates the measurement of the density with greater accuracy. Preliminary results relating to the conformation changes of ribonuclease A in the presence of guanidinium chloride and a disulfide cleaving agent are presented.

Solid-like character of virus solutions.

The solid-like behavior of turnip yellow mosaic virus solutions following the extrusion of viral RNA in alkali was observed with a torsion-fiber balance developed for the purpose. This method provided a direct measurement of the yield stresses required to break or liquefy these solutions. The yield stresses were found to increase and to be less time dependent with increasing concentrations of the virus and they were maximal at room temperatures. If the virus had been damaged, as by freeze-thaw, little or no solid-like behavior could be demonstrated. Purified viral capsids, with or without added RNA, were also inactive. The values for the yield stresses were of the same order as the value reported previously with the use of a magnetic suspension viscometer; hence, the apparent coherency appears unrelated to the magnetic fields generated by the latter instrument. These solutions behaved as typical liquids after the required stress was applied [about 0.005 to 0.17 dyne cm-2 (0.05 to 1.7 micronN cm-2)], these forces being smaller than those usually conferred by ordinary handling.

Quasi-elastic behavior of solutions of viral capsid and RNA at very low shearing stresses.

By the application of shearing stresses on the order of 10(-3) dyne cm-2 (10(-2) muN cm-2), via the magnetic viscodensimeter, extremely high relative viscosities (greater than 500) were observed when turnip yellow mosaic virus was degraded in alkali into its capsid and RNA. The solutions, however, possessed a watery consistency at this stage and exhibited a quasi-elastic character by rotor-recoil experiments. The development of this curious behavior was concentration and temperature dependent; it was not seen less than 0.5% nor at 8 degrees, and appeared sooner at 30 degrees than at 20 degrees. The time of appearance was delayed as the pH was lowered; however, the effect was still observed when the pH was as low as 9. Whereas reversibility was demonstrated when the shearing stresses exceeded the elastic resistance [0.17 dyne cm-2 (1.7 muN CM-2)], thorough mixing usually resulted in a normal behavior of the solutions thereafter. Values for the modulus of rigidity at 20 degrees for about 1% virus concentration was less than 2 X 10(-2) dyne cm-2 rad-1 (0.2 muN cm-2 rad-1), which, while extremely small, was reproducible. A porous structure, possibly involving a capsid and RNA complex, is envisioned.

Equilibrium sedimentation of turnip yellow mosaic virus.

Sedimentation equilibrium was achieved with turnip yellow mosaic virus at low speeds (600 rpm) in a magnetic ultracentrifuge. The experiments were carried out in the newly installed constant-speed rotor, equipped with automatic control and an electromagnetic drive. A particle mass of 5.55 x 10(6) daltons was calculated for the virus at vanishing concentrations, in essential agreement with the earlier results using the more tedious procedure with a freely coasting rotor. In order to interpret the observed departure from ideal behavior (nonassociating conditions), preferential interaction experiments were carried out by magnetic densimetry. These showed that a strong Donnan effect exists at pH 7, which contributed > 10(3) times more to the nonideality than did the excluded volume effect. A net negative charge of about 3.5 x 10(3) per particle was estimated at this pH.

A magnetic suspension osmometer.

The magnetic suspension balance is used for the measurement of very small osmotic pressures. The apparatus is essentially the same as that previously used for the measurement of the density and viscosity of protein solutions except that the magnetically suspended buoy is modified to make it pressure sensitive. The method is especially useful for the measurement of osmotic pressure in small samples of dilute solutions (about 10(-6) M) or of substances with molecular weights greater than 10(5). A height-sensing device has been developed which is not dependent upon the visual precision of the operator.