On the application of polyelectrolyte limiting laws to the helix–coil transition of DNA. V. Ionic effects on renaturation kinetics

The bimolecular rate constant k₂ for the association of complementary polynucleotide strands has been observed to increase strongly with increasing ionic strength—in fact, proportional to its third or fourth power. This effect is here interpreted quantitatively by means of polyelectrolyte theory starting with the Wetmur–Davidson postulate of a pre-equilibrium between separated strands and aligned segments close to one another but unbonded. The correct form, a power dependence of k₂ on ionic strength, is predicted. Comparison of the theoretical exponent with data allows the conclusion that each of the two single-stranded segments in the aligned but unbonded configuration consists of about 13–16 nucletides (not to be confused with the much smaller number of bonded base pairs in the nucleus), and that this number, denoted by Q, is possibly correlated either with a minimum length for duplex stability or with the persistence length of a single polynucleotide strand. It is suggested that experimental determination of the dependence of Q on (G+C)-content may distinguish between these possibilities. It is also suggested that addition of sufficient amounts of divalent metal ions such as Mg²⁺, Ca²⁺, or Co²⁺ may reverse the dependence of k₂ on ionic strength; under these conditions, k₂ is predicted to decrease with about the first power of ionic strength. At fixed ionic strength, k₂ should increase with increasing concentration of divalent metal ion, and, in fact, the published observation that the formation of poly(A)·2 poly(U) from poly(A)·poly(U) and poly(U) is second order in Mg²⁺ concentration is here correctly predicted from a priori molecular considerations. Finally, published association rate data for oligonucleotides are discussed in the present theoretical context.     

DNA orientation in shear flow

The linear dichroism (LD) has been measured for DNA molecules 239–164,000 base pairs long oriented in shear flow over a large range of velocity gradients (30–3,000 s ^?1) and ionic strengths (2–250 mM). At very low gradients, the degree of DNA orientation increases quadratically with the applied shear as predicted by the Zimm theory [J. Zimm, (1956) Chemical Physics, Vol. 24, p. 269]. At higher gradients, the orientation of fragments ≥ 7 kilobase pairs (kbp) increases linearly with increasing shear, whereas the orientation of fragments ≥ 15 kbp shows a more complicated dependence. In general, the orientation decreases with increasing ionic strength throughout the studied ionic strength interval, owing to a decrease in the persistence length of the DNA. The effect is most dramatic at ionic strengths below 10 mM, and is more pronounced for longer DNA fragments. For fragments ≥ 15 kbp and velocity gradients ≥ 100 s^?1, the orientation can be adequately described by the empirical relation: LD^r= –(k₁-G)/(k₂ + G), where k₁is a linear function of the square root of the ionic strength and k₂ depends on the DNA contour length. Since the DNA persistence length can be represented as a linear function of the reciprocal square root of the ionic strength [D. Porschke, (1991) Biophysical Chemistry, Vol. 40, p. 169], extrapolation of the empirical relation provides information about the stiffness of the DNA fibers. © 1993 John Wiley & Sons, Inc.     

Ionic strength effects on macroion diffusion and excess light-scattering intensities of short DNA rods

Quasielastic and static light-scattering measurements were made on DNA isolated from chicken erythrocyte mononucleosomes as a function of ionic strength between 6 × 10^?4 and 1.0M. A transition from single-exponential autocorrelation functions to markedly non-single-exponential decays was observed around 10^?2M ionic strength and was accompanied by a large decrease in the excess light-scattering intensity. Autocorrelation functions recorded below 10^?2M salt were well fit by the sum of two exponential relaxation which differed by as much as 100-fold in time constants. Apparent diffusion coefficients for the fast and slow processes plateaued around 10^?3M with numerical values approximately 10-fold and 1/10, respectively, of the translational diffusion coefficient for mononucleosome DNA at high ionic strength. This behavior is similar to that observed with poly(L -lysine), for which the slow decay has been associated with a transition to an extraordinary phase. The strong and complex salt dependence observed here illustrates potential difficulties in deriving structural information from scattering by polyions at low ionic strength.     

The reaction between cytochrome C1 and cytochrome C

The kinetics of electron transfer between the isolated enzymes of cytochrome c₁ and cytochrome c have been investigated using the stopped-flow technique. The reaction between ferrocytochrome c₁ and ferricytochrome c is fast; the second-order rate constant (k₁) is 3.0 · 10⁷ M^?1 · s^?1 at low ionic strength (I = 223 mM, 10°C). The value of this rate constant decreases to 1.8 · 10⁵ M^?1 · s^?1 upon increasing the ionic strength to 1.13 M. The ionic strength dependence of the electron transfer between cytochrome c₁ and cytochrome c implies the involvement of electrostatic interactions in the reaction between both cytochromes. In addition to a general influence of ionic strength, specific anion effects are found for phosphate, chloride and morpholinosulphonate. These anions appear to inhibit the reaction between cytochrome c₁ and cytochrome c by binding of these anions to the cytochrome c molecule. Such a phenomenon is not observed for cacodylate. At an ionic strength of 1.02 M, the second-order rate constants for the reaction between ferrocytochrome c₁ and ferricytochrome c and the reverse reaction are k₁ = 2.4 · 10⁵ M^?1 · s^?1 and k_?1 = 3.3 · 10⁵ M^?1 · s^?1, respectively (450 mM potassium phosphate, pH 7.0, 1% Tween 20, 10°C). The ‘equilibrium’ constant calculated from the rate constants (0.73) is equal to the constant determined from equilibrium studies. Moreover, it is shown that at this ionic strength, the concentrations of intermediary complexes are very low and that the value of the equilibrium constant is independent of ionic strength. These data can be fitted into the following simple reaction scheme: cytochrome c²⁺₁ + cytochrome c³⁺ai cytochrome c³⁺₁ + cytochrome c²⁺.     

Interactions des ions métalliques avec le DNA. IV. Fixation de i'ion cuivrique sur le DNA

The binding of cupric ion (Cu⁺⁺) to DNA was followed by spectrophotometry, melting profiles, and hydrodynamic techniques, in 0. 1M NaClO₄ and at pH 5. 6. A small amount of Cu⁺⁺ is bound specifically to bases (about 1 Cu⁺⁺ per 20 nucleotides), in agreement with polarographic and EPR data. A preferential stabilization of G–C pairs and only a slight increase of the flexibility of the molecule were observed. In 5 × 10^?3M NaClO₄, a higher number of nonhomogeneous binding sites is found by spectrophotometry. It is concluded that at least two types of sites are available for Cu⁺⁺. The first one, where Cu⁺⁺ is chelating N₇ of purines to phosphate, is observed only at low ionic strength and destabilizes the double helix. The second exists mainly at 0, 1M or higher ionic strength. All the sites are identical and could be attributed to two successive guanine residues in the same strand. Similar behavior was found for other divalent cations, e. g., Fe⁺⁺, Mn⁺⁺, and Co⁺⁺.     

Helix-coil transition in nucleoprotein. Effect of ionic strength on thermal denaturation of polylysine-DNA complexes

Thermal denaturation of direct-mixed and reconstituted polylysine–DNA complexes in 2.5 × 10^?4 M EDTA, pH 8.0 and various concentrations of NaCl has been studied. For both complexes, increasing ionic strength of the solution raises T_m, the melting temperature of free base pairs. The linear dependence of T_m on log Na⁺ indicates that the concept of electrostatic shielding on phosphate lattice of an infinitely long pure DNA by Na⁺ can be applied to short free DNA segments in a nucleoprotein. For a direct-mixed polylysine–DNA complex, the melting temperature of bound base pairs T_m′ remains constant at various ionic strengths. On the other hand, the T_m′ in a reconstituted polylysine–DNA complex is shifted to lower temperature at higher ionic strength. This phenomenon occurs for reconstituted complex with long polylysine of one thousand residues or short polylysine of one hundred residues. It is shown that such a decrease of T_m′ is not due to a reduction of coupling melting between free and bound regions in a complex when the ionic strength is raised. It is also not due to intermolecular or intramolecular change from a reconstituted to a direct-mixed complex. It is suggested that this phenomenon is due to structural change on polylysine-bound regions by ionic strength. It is suggested further that Na⁺ may replace water molecules and bind polylysine-bound regions in a reconstituted complex. Such a dehydration effect destabilizes these regions and lowers T_m′. This explanation is supported by circular dichroism (CD) results.     

IMPROVED PROCEDURES FOR THE CLONING AND PURIFICATION OF MICROCYSTIS CULTURES (CYANOPHYTA)1

A cloned axenic culture of Microcystis Kützing was obtained by combining two procedures: a) the disaggregation of multicellular Microcystis colonies by dilution into deionized water, and b) the selective growth of Microcystis in agar media containing Na₂S, which inhibited or killed the associated contaminants. Microcystis growth was stimulated by 0.3–1 mM Na₂SO₃, but not by 0.1–33 mM Na₂SO₄. Although Microcystis cells survived temporary exposure to high Na₂S concentrations, their growth was not stimulated by 1 × 10^?5 to 1.0 M Na₂S. Possible metabolic roles of reduced sulfur compounds are considered. Microcystis colonies disaggregated to unicells at ionic concentrations below 1 mM for univalent cations, 10–100 μM for the divalent cations, and 3–10 μM for Fe³⁺. Higher cation concentrations prompted cell aggregation. With > 100 mM Fe³⁺, the Microcystis capsule appeared rust-colored. Neither nonionic solutes nor anions detectably influenced aggregation. These observations suggest cation interactions with the Microcystis capsule and are discussed with regard to: a) possible siderochrome activity, cation chelation or luxury uptake of cations, b) the questionability of using cell aggregation as a criterion for identifying Microcystis in samples of unknown ionic strength, c) the utility of low ionic strength media in releasing contaminating bacteria from the capsule and in obtaining algal unicells for cloning, and d) a model for cation interactions with the capsule.     

Osmotic Pressure of DNA Solutions and Effective Diameter of the Double Helix

Abstract

A simple osmometer with nuclear filters (polymer films with pores of a preset diameter) were used to measure the osmotic pressure of Col El plasmid DNA solutions in the concentration range of 1–4 mg/ml DNA. Linear and open circular DNA forms proved to have the same osmotic pressure within the experimental accuracy. The results of the measurements were used for calculating the second virial coefficient A ₂ of the solution of DNA segments and the effective chain diameter d _eff in the ionic strength range of 10^?2-0.1 M, As the ionic strength is lowered from 0.1 to 10^?2 M the effective diameter of DNA increases from 80 to 220 A. The results are in rather good agreement with theory and with other experimental data.     

Effects of pH,Ionic Strength,Temperature, and Complexing Anions on the Sorption Behavior of Cobalt on Hydrous Silica

Sorption of Co(II) on SiO₂.xH₂O (silica gel) has been investigated as a function of time, amount of silica gel (0.10–1.00g), cobalt concentration (5.00 × 10^?5–1.20 × 10^?3 M), ionic strength (0.20–1.40 M NaClO₄), pH (～6.80–10.80), and temperature (273–318 K). Using the sorption kinetics data, the diffusion coefficient of Co(II) was calculated to be 6.86(±0.44) × 10^?12 m²sec^?1 under particle diffusion-controlled conditions. The sorption rate was determined as 2.61(±0.19) × 10^?3 sec^?1 at 298 K, pH 6.70(±0.05) and 0.20 M NaClO₄. The sorption data followed the Freundlich, Langmuir, and Dubinin-Radushkevich (D-R) isotherms. Cobalt sorption decreased with increased ionic strength. A gradual decrease in pH with increased ionic strength supported the sorption of Co(II) by an ion exchange mechanism. The effects of different ligands such as , F^?, and on the sorption of Co(II) were studied in the pH range 6.50 to 8.50. The sorption of cobalt on silica gel increased with increased temperature and had an endothermic enthalpy change (ΔH = 23.60(±0.57) kJ/mol).     

Characterization of group C meningococcal polysaccharide by light-scattering spectroscopy. III. Determination of molecular weight, radius of gyration, and translational diffusional coefficient.

Group-specific polysaccharides isolated by means of a cetavlon procedure are immunogenic in man and induce protective immunity against meningococcal meningitis. Minute quantities of the polymers in solution can act as vaccines. We now report the first characterization of a fractionated (C-1) group C polysaccharide in 0.4KM KCl and 0.05M sodium acetate by means of light-scattering spectroscopy. Independent measurements of refractive index increments, absolute scattered intensities, angular scattering intensities and line widths as a function of scattering angles and delay times at different concentrations using incident wavelengths of 632.8 nm from a He–Ne laser and of 488 nm from an argon–ion laser yield information on aggregation properties, molecular weight (M_r), radius of gyration 〈r⁰_g〉^1/2_z, translational diffusion coefficient 〈D〉⁰_z, and second virial coefficients A₂ and B₂ of C-1 polysaccharide. At relatively high ionic strength (0.04M KCl + 0.05M sodium acetate), we obtain for the C-1 polysaccharide in solution M_r = 5.15 × 10⁵, 〈r²_g〉^1/2_z = 345 Å, A₂ = 1.25 × 10^?4 ml/g, 〈D〉 = 1.16 × 10^?7 cm²/sec with a corresponding Stokes radius of 240 Å and B₂ = 4.4 ml/g. A₂ and B₂ are the second virial coefficients from intensity- and diffusion-coefficient measurements. The C-1 polysaccharide aggregates in solution and behaves hydrodynamically like random coils. Viscosity and sedimentation studies further confirm our conclusions that the fractioned C-1 polysaccharide aggregates in solution and EDTA can partially break up those aggregates. However, the system remains polydisperse even after adding an excess amount of EDTA. The weight-average molecular weight of the C-1 polysaccharide in solution depends upon ionic strength and exhibits a minimum at ～0.2M KCl. Finally, viscosity, light-scattering, and sedimentation results all show that the aggregated macromolecular system behaves like random-coiled polymers with no measurable shape factors.     

Quasielastic light scattering: Effect of ionic strength on the internal dynamics of DNA

Quasielastic light scattering is used to study the effect of ionic strength on the dynamic behaviour of DNA. In a first approach the spectrum of scattered light is analyzed in terms of a single relaxation process. The large difference between the observed behaviour and that expected according to a pure diffusional process reflects the contribution associated with internal modes, which increases with decreasing ionic strength. Such behaviour is better analyzed in terms of a double relaxation process by using two relaxation times, the reciprocals of which are equal to DK² and DK² + τ_i^?1 (K), respectively, where τ_i (K) is an average value describing the set of modes observed at a given K value. Relative intensity and relaxation times, which are the more accurate parameters, were used to interpret the results. The observed increase of the relative contribution of internal modes with decreasing ionic strength is actually a relative decrease of the diffusional contribution induced by a corresponding increase of the radius of gyration R_G. On the other hand, the reciprocal τ_i^?1 (K) of the relaxation time is a linear function of K² in the analyzed KR_G range and is insensitive to ionic strength between 10^?2M and 1M. These results, when discussed according to Rouse's model, lead to define for each value of τ_i^?1 (K) a corresponding mean-squared equilibrium length 〈μ〉 which is found to be a linear function of K^?2.     

Electric birefringence of native DNA in an alternating field

The relaxation of the birefringence of native DNA in solution was investigated in a pulsed sine-wave electric field. Relaxation times were calculated from the degree of damping of the birefringence signal and were studied as a function of the strength and frequency of the applied field, the molecular weight of the DNA, and the viscosity and ionic strength of the solvent. Relaxation times decrease with increasing field strength. For high-molecular weight DNA (>10⁶ daltons), the relaxation times decreased with frequency and increased less than linearly with viscosity. For low-molecular-weight DNA (<6 × 10⁵ daltons), the relaxation times were independent of frequency, increased linearly with viscosity, and varied with the 1.65 ± 0.1 power of the molecular weight. The average birefringence of high-molecular-weight DNA decreased with frequency in 0.001M Na₂ EDTA plus NaOH, pH 7.0, but is much less frequency-dependent if the EDTA concentration is reduced tenfold, while the average birefringence of sonicated DNA increases in both solvents with increasing frequency.     

An assessment of thermodynamic reaction constants for simulating aqueous environmental monomethylmercury speciation

ABSTRACT

Monomethylmercury (CH₃Hg ⁺) is both the most ecologically significant and the least well characterized species of mercury in environmental settings. Our understanding of the environmental speciation behavior of this compound is limited both as the result of lesser available laboratory data (when compared to inorganic mercury) as well as the uncertainties associated with our understanding of the properties of environmental ligands. A careful examination and synthesis of data reported in the technical literature led to the following findings: (1) a 25°C, zero ionic strength bicarbonate ion complexation constant estimate is remarkably close to an earlier reported value at 0.4 M: CH₃Hg⁺ + HCo₃^-?CH₃HgHCO₃,log₁₀K = 2.6 (±0.22, 1 SD), (2) three 25°C zero ionic strength reaction constants reported by DeRobertis et al.(1998) were confirmed to within ~±0.1 log₁₀K units: CH₃Hg ⁺+ OH^-?CH₃HgOH, log₁₀K = 9.47; 2CH₃Hg ⁺ + H₂O?(CH₃Hg)₂OH ⁺ + H⁺, log₁₀K =?2.15; CH₃Hg ⁺+ Cl^-?CH₃HgCl, log₁₀K = 5.45, (3) “best estimate” literature complexation constants corrected to zero ionic strength include: CH₃Hg ⁺ + F^-?CH₃HgF, log₁₀K = 1.75 (20°C corr. Schwart-zenbach and Schellenberg, 1965); CH₃Hg ⁺ + Br^-?CH₃HgBr, log₁₀K = 6.87 (20°C corr. Schwartzenbach and Schellenberg, 1965); CH₃Hg ⁺ +1^-?CH₃HgI, log₁₀K = 8.85 (20°C corr. Schwartzenbach and Schellenberg, 1965); and CH₃Hg ⁺+ SO₄^2-?CH₃HgSO₄^-,log₁₀K = 2.64 (25°C, DeRobertis et al., 1998), (4) literature reported values for simulating monomethylmercury complexation with the carbonate ion may be too low: CH₃Hg ⁺+ CO₃^2-?CH₃HgCO₃^-, log₁₀K = 6.1 (Rabenstein et al., 1976; Erni, 1981), and (5) ‘‘best estimate’’ constants for simulating methyl mercury complexation with reduced environmental sulfur species include: CH₃Hg ⁺ + S^2-?CH₃HgS ^-, log₁₀K = 21.1; CH³Hg ⁺+ SH ^-? CH3HgSH, log₁₀K = 14.5 (H ⁺ + SH^-?CH₂S, log₁₀K = 6.88; Dyrssen and Wedborg, 1991); CH₃Hg ⁺ + RS^-?CH₃HgSR, log₁₀K = 16.5 (H + + RS^-?RSH, log₁₀K = 9.96; Qian et al., 2002); and CH₃Hg ⁺+ CH₃HgS^{1 -}?(CH₃Hg)₂S, log₁₀K = 16.32 (Schwartzenbach and Schellenberg, 1965; Rabenstein et al., 1978; and Erni, 1981).     

Mixed Phase Solid‐State Plastic Crystal Electrolytes Based on a Phosphonium Cation for Sodium Devices

Na batteries are seen as a feasible alternative technology to lithium ion batteries due to the greater abundance of sodium and potentially similar electrochemical behavior. In this work, mixed phase electrolyte materials based on solid‐state compositions of a tri methylisobutylphosphonium (P_111i4) bis(tri fluromethanesulphonyl)amide (NTf₂) organic ionic plastic crystal (OIPC) and high concentration of NaNTf₂ that support safe, sodium metal electrochemistry are demonstrated. A Na symmetric cell can be cycled efficiently, even in the solid state (at 50 °C and 60 °C), for a 25 mol% (P_111i4NTf₂)–75 mol% NaNTf₂ composition at 0.1 mA cm^?2 for 100 cycles. Thus, these mixed phase materials can be potentially used in Na‐based devices under moderate temperature conditions. It is also investigated that the phase behavior, conductivity, and electrochemical properties of mixtures of NaNTf₂ with this OIPC. It is observed that these mixtures have complex phase behavior. For high compositions of the Na salt, the materials are solid at room temperature and retain a soft solid consistency even at 50 °C with remarkably high conductivity, approaching that of the pure ionic liquid at 50 °C, i.e., 10^?3–10^?2 S cm^?1.     

Stable Seamless Interfaces and Rapid Ionic Conductivity of Ca–CeO2/LiTFSI/PEO Composite Electrolyte for High‐Rate and High‐Voltage All‐Solid‐State Battery

Stable and seamless interfaces among solid components in all‐solid‐state batteries (ASSBs) are crucial for high ionic conductivity and high rate performance. This can be achieved by the combination of functional inorganic material and flexible polymer solid electrolyte. In this work, a flexible all‐solid‐state composite electrolyte is synthesized based on oxygen‐vacancy‐rich Ca‐doped CeO₂ (Ca–CeO₂) nanotube, lithium bis(trifluoromethanesulfonyl)imide (LiTFSI), and poly(ethylene oxide) (PEO), namely Ca–CeO₂/LiTFSI/PEO. Ca–CeO₂ nanotubes play a key role in enhancing the ionic conductivity and mechanical strength while the PEO offers flexibility and assures the stable seamless contact between the solid electrolyte and the electrodes in ASSBs. The as‐prepared electrolyte exhibits high ionic conductivity of 1.3 × 10^?4 S cm^?1 at 60 °C, a high lithium ion transference number of 0.453, and high‐voltage stability. More importantly, various electrochemical characterizations and density functional theory (DFT) calculations reveal that Ca–CeO₂ helps dissociate LiTFSI, produce free Li ions, and therefore enhance ionic conductivity. The ASSBs based on the as‐prepared Ca–CeO₂/LiTFSI/PEO composite electrolyte deliver high‐rate capability and high‐voltage stability.     

Characterization of the binding of paylean and DNA by fluorescence,UV spectroscopy and molecular docking techniques

The interaction of paylean (PL) with calf thymus DNA (ctDNA) was investigated using fluorescence spectroscopy, UV absorption, melting studies, ionic strength, viscosity experiments and molecular docking under simulated physiological conditions. Values for the binding constant K_a between PL and DNA were 5.11 × 10³, 2.74 × 10³ and 1.74 × 10³ L mol^–1 at 19, 29 and 39°C respectively. DNA quenched the intrinsic fluorescence of PL via a static quenching procedure as shown from Stern–Volmer plots. The relative viscosity and the melting temperature of DNA were basically unchanged in the presence of PL. The fluorescence intensity of PL–DNA decreased with increasing ionic strength. The value of K_a for PL with double‐stranded DNA (dsDNA) was larger than that for PL with single‐stranded DNA (ssDNA). All the results revealed that the binding mode was groove binding, and molecular docking further indicated that PL was preferentially bonded to A–T‐rich regions of DNA. The values for ΔH, ΔS and ΔG suggested that van der Waals forces or hydrogen bonding might be the main acting forces between PL and DNA. The binding distance was determined to be 3.37 nm based on the theory of Förster energy transference, which indicated that a non‐radiation energy transfer process occurred. Copyright © 2015 John Wiley & Sons, Ltd.     

Orientational interactions in low-concentration DNA solutions

The rotational relaxation tiem τ₃ of DNA molecules (M_w ? 5 × 10⁶) in solution has been determined by the transient electric birefringence method. The analysis of the birefringence decay makes it possible to study only the higher-molecular-weight fraction, the molecules being considered as rigid elongated particles in a short time scale. A marked concentration dependence of the relaxation time has been observed for DNA in low ionic strengths. Above a critical concentration c*, τ₃ increases with the DNA concentration, c. The value of c* increases with the ionic strength. For 10^?3 ionic strength (with NaCl), c* is about 10 μg/mL; then we observe the same strong concentration dependence of rotational relaxation times as recently reported for rodlike M-13 viruses [Maguire, J. F., McTague, J. P. & Rondelez, F. (1980) Phys. Rev. Lett. 45 , 1891–1894]. These results may be discussed in terms of the Doi-Edwards theory for rotational relaxation time of rigid macromolecules [Doi, M. (1975) J. Phys. 36 , 607–611; Doi, M. & Edwards, S. F. (1978) J. Chem. Soc. Faraday Trans. 74 , 918–932] and the critical concentration above which the interactions between the molecules begin to appear allows determining the corresponding molecular length. We observe a very good agreement between the DNA lengths obtained from the c* values and by using the infinite dilution value of τ₃ and Broersma's equation. Therefore, only highly diluted solutions can be used if intrinsic molecular properties based on the rotational diffusion of high-molecular-weight elongated molecules are studied.     

“Ordered” structure in solutions and gels of a globular protein as studied by small angle neutron scattering

Small-angle neutron scattering measurements were used for structural investigation of β-lactoglobulin solutions and heat-set gels in conditions of strong double layer repulsions. At pH 9 and low ionic strength, a correlation peak was observed on the scattering curves of the solutions whatever the protein concentration C used (in the range C = 0.02–0.10 g/mL). The wave vector value q_max corresponding to these maxima scaled as C^0.25. This exponent value is in agreement with those reported in the literature for other globular proteins in the same concentration range. Increasing the ionic strength decreased the peak which vanished without changing position at 0.1M NaCl. This polyelectrolyte-like behaviour suggests a local structure in the protein solution due to double layer repulsions. In the case of heat-set aggregates and gels (0.02–0.13 g/mL) formed at pH 9 and low ionic strength, a peak in the scattering curves was also observed indicating that even after gelation a correlation is still present; q_max varied as C^0.5. As in the case of the solutions, the correlation peak decreased with increasing ionic strength, and it vanished at 0.06M NaCl. The dilution of the aggregates in order to determine their intraparticle structure factor showed that the correlations were lost and that the aggregates displayed the same internal structure as the elementary subunit in the gels. At high ionic strength, fractal structures of the aggregates down to a length scale of about 40 Å were observed with d_f = 1.3–1.75 ± 0.05, increasing with protein concentration. Subsequent dilution didn't change the fractal dimension of these structures. © 1996 John Wiley & Sons, Inc.