Effect of deuterium oxide on the thermodynamic quantities associated with phase transitions of phosphatidylcholine bilayer membranes

The bilayer phase transitions of three kinds of phospholipids, dipalmitoylphosphatidylcholine (DPPC), distearoylphosphatidylcholine (DSPC) and dihexadecylphosphatidylcholine (DHPC), in deuterium oxide (D₂O) and hydrogen oxide (H₂O) were observed by differential scanning calorimetry (DSC) under ambient pressure and light-transmittance measurements under high pressure. The DSC measurements showed that the substitution of H₂O by D₂O affected the pretransition temperatures and the main-transition enthalpies of all PC bilayers. The temperature-pressure phase diagrams for these PC bilayer membranes in both solvents were constructed by use of the data of light-transmittance measurements. Regarding the main transition of all PC bilayer membranes, there was no appreciable difference between the transition temperatures in D₂O and H₂O under high pressure. On the other hand, the phase transitions among the gel phases including the pretransition were significantly affected by the solvent substitution. The thermodynamic quantities of phase transitions for the PC bilayer membranes were evaluated and the differences in thermodynamic properties by the water substitution were considered from the difference of interfacial-free energy per molecule in the bilayer in both solvents. It was proved that the substitution of H₂O by D₂O causes shrinkage of the molecular area of phospholipid at bilayer interface due to the difference in bond strength between deuterium and hydrogen bonds and produces the great influence on the bilayer phase with the smaller area. Further, the induction of bilayer interdigitation in D₂O turned out to need higher pressures than in H₂O.     

Monomeric versus dimeric structures in ternary complexes of manganese(II) with derivatives of benzoic acid and nitrogenous bases: structural details and spectral properties

Adducts formed by [Mn(2,6-dmb)₂(H₂O)₃]_n · nH₂O, 2,6-dmb=2,6-dimethoxybenzoate(1-), Mn(2,4-dhb)₂ · 8H₂O, Mn(2,5-dhb)₂ · 4H₂O or Mn(2,6-dhb)₂ · 8H₂O, dhb=dihydroxybenzoate(1-), and 2,2^′-bipyridine (bpy), 4,4^′-dimethyl-2,2^′-bipyridine (Me₂bpy) or 4,7-dimethyl-1,10-phenanthroline (Me₂phen) were isolated in the solid state and characterised by IR, EPR and thermogravimetry. Two of them, [Mn(2,6-dhb)₂(bpy)₂] (1) and [Mn₂(2,6-dmb)₄(Me₂Phen)₂(H₂O)₂] · 2EtOH (2), were studied by single crystal X-ray diffraction. The adduct 1 is mononuclear and consists of hexa-co-ordinate manganese(II) ions bound to two bipyridine and two 2,6-dihydroxybenzoate ligands in a cis-octahedral arrangement. The complex 2 exhibits a dinuclear structure in which two manganese(II) ions share two carboxylate groups adopting a rather uncommon single-atom bridging mode. The results allow us to conclude that weak, e.g., hydrogen bonding and stacking interactions govern the type of structure, monomeric or dimeric. The spectral features of the complexes are discussed. In particular, the solid-state EPR features of the complexes are interpreted in terms of D, E and H_max, the high-field resonance. For the monomeric species, the higher is the D value, the higher is H_max.     

Solvent isotope effects on microtubule polymerization and depolymerization

The initial velocity of polymerization of purified beef brain tubulin has been determined at various values of pH or pD in water and in H₂O-D₂O mixtures. D₂O was shown to inhibit both polymerization at 37 °C and depolymerization measured at 5 °C and 37 °C. The microtubules formed in D₂O were indistinguishable from those formed in H₂O, by electron microscope examination. In 93% D₂O the pL²versus rate of polymerization curve was displaced about one unit towards higher pL values. In certain regions of the pL versus rate curve, a stimulation in the rate of polymerization by D₂O is observed. The extent of polymerization at the optimum pL value was not affected by D₂O.     

Desiccation tolerant lichens facilitate in vivo H/D isotope effect measurements in oxygenic photosynthesis

We have used the desiccation-tolerant lichen Flavoparmelia caperata, containing the green algal photobiont Trebouxia gelatinosa, to examine H/D isotope effects in Photosystem II in vivo. Artifact-free H/D isotope effects on both PSII primary charge separation and water oxidation yields were determined as a function of flash rate from chlorophyll-a variable fluorescence yields. Intact lichens could be reversibly dehydrated/re-hydrated with H₂O/D₂O repeatedly without loss of O₂ evolution, unlike all isolated PSII preparations. Above a threshold flash rate, PSII charge separation decreases sharply in both D₂O and H₂O, reflecting loss of excitation migration and capture by PSII. Changes in H/D coordinates further slow charge separation in D₂O (?23% at 120?Hz), attributed to reoxidation of the primary acceptor Q_A^?. At intermediate flash rates (5–50?Hz) D₂O decreases water oxidation efficiency (O₂ evolution) by ?2–5%. No significant isotopic difference is observed at slow flash rates (<5?Hz) where charge recombination dominates. Slower D₂O diffusion, changes in hydrogen bonding networks, and shifts in the pK_a's of ionizable residues may all contribute to these systematic variations of H/D isotope effects. Lichens' reversible desiccation tolerance allows highly reproducible H/D exchange kinetics in PSII reactions to be studied in vivo for the first time.     

On the permeability of water molecules across vesicular lipid bilayers

Summary The permeation of water molccules across single-component lecithin or lecithin-cholesterol bilayers is studied by a new technique. The new technique makes use of the different fluorescence quantum yields of appropriate molecules in D₂O and H₂O. Water-soluble indole derivatives which by experimental manipulation reside almost entirely within the aqueous (H₂O) intravesicular compartment thus can monitor D₂O molecules permeating the bilayer by virtue of an increased quantum yield of the fluorescence. In a stopped-flow instrument, a vesicle solution containing the fluorescent chromophore in the intravesicular space is rapidly mixed with the deuterated solvent. The approach to the steady state, where the intra- and extravesicular D₂O and H₂O concentrations are equal, proceeds in a single-exponential manner. Consequently, the exchange relaxation time for the D₂O molecules passing the bilayer can be deduced from the time-dependent increase of the fluorescence intensity. The method and results on lecithin and lecithin-cholesterol bilayer vesicles are discussed. The exchange relaxation times of temperature-dependent studies are interpreted within the framework of the solubility-diffusion theory. Below the crystalline to liquid-crystalline phase transition temperature and for cholesterol-free vesicles, the rate-limiting step for the D₂O permeation is attributed to the intracore diffusion. Above the phase transition and for cholesterol-containing vesicles, the intracore diffusion seems not to be rate-limiting. Deviations from the linearity below the phase transition in the Arrhenius-type presentation of the data are related to changes of the partition coefficient of water between the solvent and the lipid phase at the premelting temperature.     

Self-association of oxyhaemoglobin. A nuclear magnetic relaxation study in H2O/D2O solutions

The proton and deuterium longitudinal relaxation rates were Studied at room temperature up to the highest protein concentrations in oxyhaemoglobin solutions of different H₂O/D₂O composition. The deuterium relaxation rates followed the experimentally well known single linear dependence on protein concentration, the slopes being little influenced by solvent (D₂O/H₂O) composition. The proton ralaxation rates show two different liner dependences on haemoglobin concentration. The entire concentration range is described by two straight lines with the threshold concentration about 11 mM (in haem), The ratio of the slopes is 1.6 (high-to-low Hb-conc.). Only in the higher concentration range two T₁'s were observed if the solvent contained more than half of D₂O. The slow relaxation phase of protons has T₁'s similar to those measured in solutions with less than half of D₂O. The relaxation of the other phase was ten times faster. The ratio of the proton populations in these two phases was equal to 2 (slow-to-fast) and independent of protein concentration. The fast relaxing protons are attributed to water molecules encaged within two or more haemoglobin molecules which associate for times long enough on the PMR time-scale.     

Vibrational spectra of the diammineplatinum(II) disodium 5′-uridine monophosphate blue complex

The blue complexes produced by reaction of cis-diamminediaquoplatinum(II) nitrate, [cis-Pt(NH₃)₂(H₂O)₂](NO₃)₂, with disodium 5′-uridine monophosphate, 5′-UMP(Na₂), in H₂O and D₂O have been investigated by FT-IR spectroscopy. On the basis of the spectral changes observed in the CO stretching region during the reactions, chelation of the amidate N(3)··O(2) moiety to Pt(II) appears to be more likely than N(4)··O(4) chelation. The antisymmetric PO stretching mode of the PO_3²⁻ group of 5′-UMP splits into a triplet on complex formation indicating that PO_3²⁻ plays an important role in the structure of the platinum blue complexes. In addition, the sugar moiety of 5′-UMP apparently adopts a predominantly C(3′)-endo conformation in the solid blue complex. Finally, Raman microprobe spectroscopy of the solid provides some evidence for PtN(3) bond formation.     

Thermal and Solvent-Isotope Effects on the Flagellar Rotary Motor near Zero Load

In Escherichia coli, the behavior of the flagellar rotary motor near zero load can be studied by scattering light from nanogold spheres attached to proximal hooks of cells lacking flagellar filaments. We used this method to monitor changes in speed when cells were subjected to changes in temperature or shifted from a medium made with H₂O to one made with D₂O. In H₂O, the speed increased with temperature in a near-exponential manner, with an activation enthalpy of 52 ± 4 kJ/mol (12.0 ± 1.0 kcal/mol). In D₂O, the speed increased in a similar manner, with an activation enthalpy of 50 ± 4 kJ/mol. The speed in H₂O was higher than that in D₂O by a factor of 1.53 ± 0.14. We performed comparison studies of variations in temperature and solvent isotope, using motors operating at high loads. The variations were small, consistent with previous observations. The implications of these results for proton translocation are discussed.     

The lateral diffusion coefficient of lecithin molecules in single bilayer vesicles studied by 14N NMR

The ¹⁴N nuclear relaxation times T₁ and T₂ in egg yolk phosphatidylcholine have been observed in single bilayer vesicles dispersed in the media of different viscosities, ¹H₂O and ²H₂O. The lateral diffusion coefficient of lipid molecule D has been calculated according to the method reported earlier: D = 2.2 × 10^?8cm²s^?1 in ¹H₂O and 2.1 × 10^?8cm²s^?1 in ²H₂O at 20°C. They are in excellent agreement. This result gives a strong basis of usefulness of ¹⁴N NMR method in the evaluation of D without introducing any system perturbation.     

Purification, properties, and kinetic observations on the isoenzymes of NADP isocitrate dehydrogenase of maize.

A successful method for the purification of NADP isocitrate dehydrogenase from a plant source, Zea mays, is reported. Two mitochondrial isoenzymes were found and purified to homogeneity by a course of acetone fractionation, bulk exchange on DEAE-cellulose, cellulose hydroxylapatite column chromatography, and continuous elution electrophoresis. The mitochondrial isoenzymes are very similar with respect to kinetic properties, response to solvent perturbation, and temperature dependence of the pH/V relationship of isocitrate dehydrogenation. The Michaelis constant for isocitrate is identical for both isoenzymes. The enzymes have a molecular weight of 81,000 as estimated by permeation chromatography and an isoelectric point of 5.5 as extrapolated from gel-electrophoretic mobilities. Detectable differences are confined to differences in electrophoretic mobilities and heat denaturation. In D₂O the rate of the overall reaction from isocitrate to α-ketoglutarate and CO₂ was about 3.6 times slower than the same reaction in H₂O. Both the forward and reverse reactions, in which isocitrate is dehydrogenated or generated from oxalosuccinate, were observed to decrease by this amount in D₂O. The decarboxylation of oxalosuccinate was found to decrease by only about 25% in D₂O relative to the velocity of the reaction in H₂O. Thus the slow step in the overall reaction must be the initial dehydrogenation step rather than the decarboxylation of oxalosuccinate. The pK of the overall reaction did not change in D₂O as compared to H₂O.     

Local Control Model of Excitation–Contraction Coupling in Skeletal Muscle

The voltage-activated H⁺ selective conductance of rat alveolar epithelial cells was studied using whole-cell and excised-patch voltage-clamp techniques. The effects of substituting deuterium oxide, D₂O, for water, H₂O, on both the conductance and the pH dependence of gating were explored. D⁺ was able to permeate proton channels, but with a conductance only about 50% that of H⁺. The conductance in D₂O was reduced more than could be accounted for by bulk solvent isotope effects (i.e., the lower mobility of D⁺ than H⁺), suggesting that D⁺ interacts specifically with the channel during permeation. Evidently the H⁺ or D⁺ current is not diffusion limited, and the H⁺ channel does not behave like a water-filled pore. This result indirectly strengthens the hypothesis that H⁺ (or D⁺) and not OH⁻ is the ionic species carrying current. The voltage dependence of H⁺ channel gating characteristically is sensitive to pH_o and pH_i and was regulated by pD_o and pD_i in an analogous manner, shifting 40 mV/U change in the pD gradient. The time constant of H⁺ current activation was about three times slower (τ_act was larger) in D₂O than in H₂O. The size of the isotope effect is consistent with deuterium isotope effects for proton abstraction reactions, suggesting that H⁺ channel activation requires deprotonation of the channel. In contrast, deactivation (τ_tail) was slowed only by a factor ≤1.5 in D₂O. The results are interpreted within the context of a model for the regulation of H⁺ channel gating by mutually exclusive protonation at internal and external sites (Cherny, V.V., V.S. Markin, and T.E. DeCoursey. 1995. J. Gen. Physiol. 105:861–896). Most of the kinetic effects of D₂O can be explained if the pK _a of the external regulatory site is ∼0.5 pH U higher in D₂O.     

Determination of the proton environment of high stability Menasemiquinone intermediate in Escherichia coli nitrate reductase A by pulsed EPR

Escherichia coli nitrate reductase A (NarGHI) is a membrane-bound enzyme that couples quinol oxidation at a periplasmically oriented Q-site (Q_D) to proton release into the periplasm during anaerobic respiration. To elucidate the molecular mechanism underlying such a coupling, endogenous menasemiquinone-8 intermediates stabilized at the Q_D site (MSQ_D) of NarGHI have been studied by high-resolution pulsed EPR methods in combination with ¹H₂O/²H₂O exchange experiments. One of the two non-exchangeable proton hyperfine couplings resolved in hyperfine sublevel correlation (HYSCORE) spectra of the radical displays characteristics typical from quinone methyl protons. However, its unusually small isotropic value reflects a singularly low spin density on the quinone carbon α carrying the methyl group, which is ascribed to a strong asymmetry of the MSQ_D binding mode and consistent with single-sided hydrogen bonding to the quinone oxygen O1. Furthermore, a single exchangeable proton hyperfine coupling is resolved, both by comparing the HYSCORE spectra of the radical in ¹H₂O and ²H₂O samples and by selective detection of the exchanged deuterons using Q-band ²H Mims electron nuclear double resonance (ENDOR) spectroscopy. Spectral analysis reveals its peculiar characteristics, i.e. a large anisotropic hyperfine coupling together with an almost zero isotropic contribution. It is assigned to a proton involved in a short ∼1.6 Å in-plane hydrogen bond between the quinone O1 oxygen and the Nδ of the His-66 residue, an axial ligand of the distal heme b_D. Structural and mechanistic implications of these results for the electron-coupled proton translocation mechanism at the Q_D site are discussed, in light of the unusually high thermodynamic stability of MSQ_D.     

D2O-induced ion channel activation in Characeae at low ionic strength

Effects of D₂O were studied on internodal cells of the freshwater alga Nitellopsis obtusa under plasmalemma perfusion (tonoplast-free cells) with voltage clamp, and on Ca²⁺ channels isolated from the alga and reconstituted in bilayer lipid membranes (BLM). External application of artificial pond water (APW) with D₂O as the solvent to the perfused plasmalemma preparation led to an abrupt drop of membrane resistance (R _m = 0.12 ±0.03 kΩ · cm²), thus preventing further voltage clamping. APW with 25% D₂O caused a two-step reduction of R _m : first, down to 2.0 ± 0.8 kΩ · cm², and then further to 200 Ω · cm², in 2 min. It was shown that in the first stage, Ca²⁺ channels are activated, and then, Ca²⁺ ions entering through them activate the Cl^? channels. The Ca²⁺ channels are activated irreversibly. If 100 mm CsCl was substituted for 200 mm sucrose (introduced for isoosmoticity), no effect of D₂O on R _m was observed. Intracellular H₂O/D₂O substitution also did not change R _m . In experiments on single Ca²⁺ channels in BLM H₂O/ D₂O substitution in a solution containing 100 mm KCl (trans side) produced no effect on channel activity, while in 10 mm KCl, at negative voltage, the open channel probability sharply increased. This effect was irreversible. The single channel conductance was not altered after the H₂O/D₂O substitution. The discussion of the possible mechanism of D₂O action on Ca²⁺ and Cl^? channels was based on an osmotic-like stress effect and the phenomenon of higher D-bond energy compared to the H-bond.     

Thermal perturbation difference spectra and D2O vs H2O isothermal difference spectra of protein-model aromatic chromophores.

The thermal perturbation difference spectra of phenolic and indolic chromophores in water resemble the isothermal D₂O and H₂O spectra of these chromophores. For phenols approximately equal Δ? values are obtained in both types of spectra, but for their methyl ethers Δ? values of D₂O vs H₂O spectra are about half of those of the thermal perturbation spectra. Phenols and their methyl ethers were studied in deuterated ethylene glycol and glycerol vs the corresponding protiated solvent, and in nonprotic solvents containing 0.25–4% D₂O or H₂O. For phenols in D₂O vs H₂O, about one-third to one-half of the difference spectrum is attributed to solvent structure difference, and the remainder to the effects of replacing OH by OD and to differences in accepting hydrogen bonds from D₂O and H₂O. The refractive index difference between D₂O and H₂O was shown to be a minor contribution by means of experiments in which D₂O was at 5 dgC and H₂O at 47 dgC, conditions of equal refractive index (Na_D). D₂O vs H₂O and glycerol-d vs glycerol-h difference spectra of ribonuclease are about twice as large as expected from the known number of exposed tyrosyl side chains. Possible sources of error in D₂O vs H₂O spectra of proteins are discussed.     

Kinetic Studies on the Initial Crystallization Process of Lysozyme in the Presence of D2O and H2O

In the initial stages of the crystallization of egg-white lysozyme, monomeric lysozyme aggregates rapidly and forms a nucleus in the presence of high salt concentrations. The formation process of the aggregates was examined to make clear the difference between the situations in heavy water and in water at the same sodium ion concentration. The aggregation in both cases was observed at unsaturated and/or saturated lysozyme concentrations. The turbidity at 350 nm of lysozyme increased remarkably within 60 min under each experimental condition and showed no appreciable changes over 60 min. The increase of turbidity in H₂O was much slower than in D₂O at the same salt concentration (3%). Lysozyme showed a critical concentration for nucleus formation whose value in H₂O was lower than in D₂O at 3% salt concentration. There are two different aggregation models, depending on the concentration of lysozyme. However, similar results were not obtained at 3% sodium ions in H₂O. The initial aggregation rate was also dependent on the concentrations of both lysozyme and NaCI. Therefore, the effect of lysozyme concentration on the aggregation process in H₂O may be smaller than in D₂O.     

Resonance coherent anti-Stokes Raman scattering (CARS) spectra of flavin adenine dinucleotide,riboflavin binding protein and glucose oxidase

Coherent anti-Stokes Raman scattering spectra, in resonance with the isoalloxazine visible electronic transition, have been obtained down to 300 cm^?1 for flavin adenine dinucleotide, riboflavin binding protein and glucose oxidase, in H₂O and D₂O. Several isoalloxazine vibrational modes can be identified by analogy with those of uracil. Of particular interest is a band at ～1255 cm^?1 in H₂O, which is replaced by another at ～1295 cm^?1, in D₂O. The H₂O band appears to be a sensitive monitor of H-bonding of the N3 isoalloxazine proton to a protein acceptor group. It shifts down by 10 cm^?1 in riboflavin binding protein, and disappears altogether in glucose oxidase. Other band shifts, of 3–5 cm^?1, are similar for the two flavoproteins, and may reflect environmental changes between aqueous solution and the protein binding pockets.     

Kinetic Studies on the Initial Crystallization Process of Lysozyme in the Presence of D2O and H2O

In the initial stages of the crystallization of egg-white lysozyme, monomeric lysozyme aggregates rapidly and forms a nucleus in the presence of high salt concentrations. The formation process of the aggregates was examined to make clear the difference between the situations in heavy water and in water at the same sodium ion concentration. The aggregation in both cases was observed at unsaturated and/or saturated lysozyme concentrations. The turbidity at 350 nm of lysozyme increased remarkably within 60 min under each experimental condition and showed no appreciable changes over 60 min. The increase of turbidity in H₂O was much slower than in D₂O at the same salt concentration (3%). Lysozyme showed a critical concentration for nucleus formation whose value in H₂O was lower than in D₂O at 3% salt concentration. There are two different aggregation models, depending on the concentration of lysozyme. However, similar results were not obtained at 3% sodium ions in H₂O. The initial aggregation rate was also dependent on the concentrations of both lysozyme and NaCI. Therefore, the effect of lysozyme concentration on the aggregation process in H₂O may be smaller than in D₂O.     

Studies on the reactivity of malondialdehyde. I. Effects of acid concentration and solvent composition on deuterium exchange reactions

The rates of deuterium exchange reactions of malondialdehyde (MDA) and deuterated malondialdehyde (MDAd) have been studied as a function of acidity and the content of dimethyl sulfoxide (DMSO) in binary mixtures with D₂O . MDA incorporates deuterium from D₂O solutions in a first-order reaction with a rate constant (k_obs) that depends on the acid concentration. From this dependence, a catalytic constant, k_cat, can be derived (k_cat^MDA = 2.25 × 10⁵M^?s^?1). Similar kinetic behavior was found for MDAd in H₂O solutions, and in this case, k_cat^MDA = 1.56 × 10⁵M^?1s^?1. Results from reactions of MDA and MDAd in identical H₂OD₂O mixtures show that primary and secondary isotope effects are small (k_H/k_D = 1.13) and that solvent isotope effects cause most of the differences found between reactions in D₂O and H₂O. Reactions in binary DMSOd₆D₂O mixtures show a six-fold rate increase as the proportion of DMSOd₆ increases from 50% to 90%. These results also illustrate the relatively high reactivity of MDA at pH values well above its pK_a and the importance of medium composition on its reaction rate.     

Investigation of γE-crystallin target protein binding to bovine lens alpha-crystallin by small-angle neutron scattering

α-Crystallin, one of the main constituent proteins in the crystalline lens, is an important molecular chaperone both within and outside the lens. Presently, the structural relationship between α-crystallin and its target proteins during chaperone action is poorly understood. It has been hypothesised that target proteins bind within a central cavity. Small-angle neutron-scattering (SANS) experiments in conjunction with isotopic substitution were undertaken to investigate the interaction of a target lens protein (γE-crystallin) with α-crystallin (α_H) and to measure the radius of gyration (Rg) of the proteins and their binary complexes in solution under thermal stress. The size of the α_H in D₂O incubated at 65 °C increased from 69 ± 3 to 81 ± 5 Å over 40 min, in good agreement with previously published small-angle X-ray scattering (SAXS) and SANS measurements. Deuterated γE-crystallin in H₂O buffer (γE_D/H₂O) and hydrogenous γE-crystallin in D₂O buffer (γE_H/D₂O) free in solution were of insufficient size and/or too dilute to provide any measurable scattering over the angular range used, which was selected primarily to investigate γE:α_H complexes. The evolution of the aggregation size/shape as an indicator of α_H chaperone action was monitored by recording the neutron scattering in different H:D solvent contrasts under thermally stressed conditions (65 °C) for binary mixtures of α_H, γE_H, and γE_D. It was found that Rg(α_H:γE_D/D₂O) > Rg(α_H:γE_H/D₂O) > Rg(α_H/D₂O) and that Rg(α_H:γE_H/D₂O) ≈ Rg(α_H/D₂O). The relative sizes observed for the complexes weighted by the respective scattering powers of the various components imply that γE-crystallin binds in a central cavity of the α-crystallin oligomer, during chaperone action.     

Partitioning of taxifolin-iron ions complexes in octanol-water system

The composition of taxifolin-iron ions complexes in an octanol-water biphasic system was studied using the method of absorption spectrophotometry. It was found that at pH 5.0 in an aqueous biphasic system the complex of [Tf · Fe₂(OH) _k (H₂O)_{8 ? k} ] is present, but at pH 7.0 and 9.0 the complexes of [Tf₂ · Fe(OH) _k (H₂O)_{2 ? k} ] and [Tf · Fe(OH) _k (H₂O)_{4 ? k} ] are predominantly observed. The formation of a stable [Tf₃ · Fe] complex occurred in octanol phase. The charged iron ion of this complex is surrounded by taxifolin molecules, which shield the iron ion from lipophilic solvent. During transition from water to octanol phase the changes of the composition of complexes are accompanied by reciprocal changes in portion of taxifolin and iron ions in these phases. It was shown that the portion of taxifolin in aqueous solution in the presence of iron ions is increased at high pH values, and the portion of iron ions is minimal at pH 7.0. In addition, the parameters of solubility limits of taxifoliniron ions complexes in an aqueous solution were determined. The data obtained gain a better understanding of the role of complexation of polyphenol with metal of variable valency in passive transport of flavonoids and metal ions across lipid membranes.