Influence of aqueous dimethyl sulfoxide on pyridoxine protonation and tautomerization

Journal of Molecular Liquids 221 (2016) 457–462

Contents lists available at ScienceDirect

Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq

Influence of aqueous dimethyl sulfoxide on pyridoxine protonation and tautomerization G.A. Gamov a,⁎, M.N. Zavalishin a, A.Yu. Khokhlova a, V.A. Sharnin a,b a b

Ivanovo State University of Chemistry and Technology, Research Institute of Thermodynamic and Kinetics of Chemical Processes, Russian Federation G.A. Krestov Institute of Solution Chemistry, Russian Academy of Sciences, Russian Federation

a r t i c l e

i n f o

Article history: Received 18 December 2015 Accepted 8 June 2016 Available online 09 June 2016 Keywords: Pyridoxine Protonation constant Tautomeric constant Gibbs energy change of transfer Binary solvent Transmembrane transport

a b s t r a c t The constants of pyridoxine (vitamin B6) protolytic and tautomeric equilibria were determined by means of potentiometry and spectrophotometry respectively in wide range of binary solvent (aqueous dimethyl sulfoxide) compositions. The free Gibbs energies of pyridoxine molecular and zwitterionic species transfer from water to aqueous DMSO were obtained using method of partition between immiscible phases. These results were employed for discussing the transmembrane transport capability of B6 vitamin. The influence of mixed solvent composition on pyridoxine protonation was analyzed taking into account the reagents solvation data. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Pyridoxine (PN) together with pyridoxal (PL) and pyridoxamine (PM) is known as B6 vitamin. Their main roles in the human tissues are to be the precursors for synthesis of pyridoxal 5′-phosphate (PLP), an important coenzyme in a variety of enzymes (transaminases, decarboxylases etc.) [1]. PLP-dependent enzymes make ~4% of all classified activities [2], influence the amino acids, lipids and carbohydrates metabolism as well as hormones, neurotransmitters and heme biosynthesis [3,4]. It is undoubtedly that protonation of pyridoxine heteronitrogen could change significantly its biological activity. For example, the influence of pyridine nitrogen protonation of more complex compound, PLP, on PLP or its derivatives reactions (e.g. transamination) is noticeable [5– 8]. However, the effect of pyridoxine heteronitrogen protonation on PN activity in biochemical reactions is still unknown. In aqueous or aqueous-organic solution, PN exists in two forms, molecular and zwitterionic [9–11] (Scheme 1, Kz): The influence of tautomeric equilibrium shift on PN reactivity had also never been investigated. The varying of the solvent composition leading to the change of the reagents solvation states is the key way to control the chemical reaction passing. Changing the ratio of mixed solvent components, one could increase the yield of desired reaction product (e.g. possessing required biological activity). ⁎ Corresponding author. E-mail address: [email protected] (G.A. Gamov).

http://dx.doi.org/10.1016/j.molliq.2016.06.031 0167-7322/© 2016 Elsevier B.V. All rights reserved.

The main aim of present paper is to study the influence of aqueous dimethyl sulfoxide solvent on protolytic and tautomeric equilibria of pyridoxine. In order to achieve the goal we studied the effect of binary solvent composition on protonation constant and tautomeric equilibrium of pyridoxine, and solvation of PN in molecular and zwitterionic form. The study of PN protonation and tautomerization in mixtures of water with other important organic solvent, ethanol, had been carried out recently [12].

2. Experimental 2.1. Chemicals Pyridoxine hydrochloride (PN·HCl) of Fisher BioReagents production (New Jersey, USA) was used without purification. The PN content in the reactive was determined by potentiometric titration by NaOH of 0.01000 mol dm−3 to be 99.8% (weight). NaOH of Reakhim production (Slavgorodnee, Russia) was used without purification. The concentrations of NaOH and Na2CO3 in its solutions were determined via consecutive titration by HClO4 of 0.1000 mol dm−3 with phenolphthalein and methyl orange. NaClO4 of Ural's Factory of Chemicals production (Verkhnyaya Pyshma, Russia) was twice recrystallized and dried at 120 °C until its weight became constant. The purity was controlled by IR-spectroscopy. Hexane of Reakhim production (Slavgorodnee, Russia) was used without purification. Its purity was controlled refractometrically.

458

G.A. Gamov et al. / Journal of Molecular Liquids 221 (2016) 457–462

Scheme 1. Protolytic and tautomeric equilibria of pyridoxine.

Dimethyl sulfoxide of Reakhim production (Slavgorodnee, Russia) was used without purification. The residual water content was determined by K. Fischer titration to be 0.19% (weight). The absence of other impurities was proved by NMR- and IR-spectroscopy. All the binary solvents were prepared via mixing the deaerated bidistilled water (κ = 3.6 μSm cm−1, pH = 6.6) and DMSO (taking into consideration the residual water) weighted in necessary ratio with inaccuracy no N0.01 g. 2.2. Methods 2.2.1. Potentiometry We chose potentiometry for determining the protonation constants of PN. Like in previous study [12], the indicator glass electrode and silver chloride reference electrode were employed: Ag; AgCljH2 O‐S; LiCl  kH2 O‐S; NaClO4 ; PN  HCl; NaOHkglassj0:1 M HCljAgCl; Ag; where S is for organic component of binary solvent (DMSO). The potential difference between electrodes was measured with error ± 0.1 mV. The operability of potentiometrical setup was tested using HClO4 solutions with known concentrations (from 0.0001 to 0.1000 mol dm−3) in water and aqueous DMSO [13,14]. The Nernst slope value was 59.3 ± 0.5 mV. In order to eliminate diffusion potential at the boundary between reference electrode and studied solution silver chloride electrode was filled with saturated solution of KCl in water or saturated solution of LiCl in aqueous DMSO. The following equilibrium processes were considered during the experiment: þ

PN þ H ⇔PNH

þ

ð1Þ

PN–Hþ ⇔PN− þ Hþ

ð2Þ −

SolvH þ SolvH⇔Solv þ HSolvH

þ

ð3Þ

where process (1) is PN protonation, process (2) is pyridoxine acid dissociation, process (3) is autoprotolysis of mixed solvent. The protonation constants of pyridoxine were determined via titration of aqueous or aqueous-organic solution of PN·HCl of 0.009918– 0.01019 mol dm−3 and NaClO4. Aqueous or aqueous-organic solution of NaOH of 0.04958–0.1357 mol dm−3 and NaClO4 was used as a titrant. The ionic strength was 0.25 (NaClO4). The temperature within the cell was maintained at 25.0 ± 0.1 °C by external control. In addition to the

potentiometric titration, the direct pH measurements of the PN·HCl solutions set (concentration range from 0.01043 to 0.02435 mol dm−3) was employed for determining the pyridoxine protonation constants at some dimethyl sulfoxide content. Both methods gave results differing within experimental error. Potentiometric titration and direct pH measurements data were processed using PHMETR software [15], which approximates the experimental dependence (ΔE(mV) − Vtitr(ml)) by the calculated curve. During the titration the pH value of solution changed in range of 3.0– 4.5 units. Such acidity of medium was not enough for process (2) to pass with significant yield. Its exclusion from calculation scheme did not change the results. The determined macroscopic (without separation on molecular and zwitterionic species yet) constants of pyridoxine protonation are given (Table 1). The reliability of obtained data is justified by several factors: the precise equipment was employed, the accurate methods of equilibrium constants calculations were used, the statistical reproducibility of results was achieved, the different concentration conditions of the experiment gave the close values of protonation constants, and two different methods, potentiometric titration and direct pH measurements, led to the similar results. The pyridoxine macroscopic protonation constant determined in present report in water is in a satisfactory agreement with literature data (log K = 4.84 [10], log K = 4.94 [16] and log K = 5.00 [17]).

2.2.2. Immiscible phases partition method We chose the method of partition between immiscible phases to determine the Gibbs energy change of pyridoxine transfer from water to aqueous DMSO. Experiment was carried out analogically with prescribed one [12]. Two equal volumes (19.97 ml) of hexane and pyridoxine solution in water or aqueous DMSO were placed into hermetically closed flask. The concentration of pyridoxine solution prepared via accurate PN·HCl solution neutralization by NaOH was in range from 1.5319 · 10− 4 to 1.5365 · 10−4 mol dm−3. The flask was put on magnetic stirrer placed into air thermostat. Temperature was maintained at 25.0 ± 0.1 °C. The contents of the flask were stirred during 8 h, and then they were left to stay without stirring for 10 h. The bottom layer (aqueous of aqueous-DMSO solution) was sampled to determine the equilibrium concentration of pyridoxine using calibration graph method of spectrophotometry. Spectral measurements are described in details in Section 2.2.3. The equilibrium pyridoxine concentration in hexane phase was calculated by subtraction of the equilibrium pyridoxine concentration in bottom layer from the total concentration of PN.

Table 1 Pyridoxine protonation constants in aqueous dimethyl sulfoxide at atmospheric pressurea and T = 298.2 ± 0.1 K. XDMSO

l g Kbb

0 0.05 0.10 0.20 0.30 0.50 0.70 0.90 0.99

4.969 ± 0.016 [12] 4.602 ± 0.015 4.152 ± 0.016 3.910 ± 0.012 3.656 ± 0.016 3.461 ± 0.020 3.305 ± 0.014 4.094 ± 0.021 4.328 ± 0.022

a All experiments were carried out under atmospheric pressure which changed in the range of (0.097 to 0.104) MPa. b There are given the expanded uncertainties for every value with the confidence level of 0.95 and sample size 3–6 experiments.

G.A. Gamov et al. / Journal of Molecular Liquids 221 (2016) 457–462

459

The partition coefficients of pyridoxine and Gibbs energy change of PN transfer from water to aqueous dimethyl sulfoxide were calculated employing following equations [12]: Kp1 ¼

Kp2 ¼

½PNHex

ð4Þ

½PNH2 O ½PNHex

ð5Þ

½PNH2 O−EtOH

Δtr GðPNÞH2 O−EtOH ¼ −RT ln

½PNH2 O−EtOH ½PN

H2 O

¼ −RT ln

Kp1 Kp2

ð6Þ

where [PN]Hex, [PN]H2O, [PN]H2O −EtOH are equilibrium concentrations of distributed compound in hexane, water and binary solvent respectively, Kp1 and Kp2 are for partition coefficients of PN in hexane-water and hexane-mixed solvent systems respectively. The results of experiment determining the macroscopic Gibbs energy changes of pyridoxine (without separation on molecular and zwitterionic species yet) from water to aqueous DMSO are given (Table 2). When DMSO content is N0.5 m.f., binary solvent starts to dissolve noticeably into hexane as it was found by refractometry. Partition coefficients of pyridoxine found in present paper are close to the partition coefficients of such compounds as benzoic and salicylic acids between water and hexadecane [18]. The uncertainty of Gibbs energy change of pyridoxine transfer is caused mainly by error of PN solution preparation. The obtained thermodynamics can be suggested as standard values since they are determined at T = 298.2 ± 0.1 K, P = 0.101 ± 0.003 MPa and small ionic strength (I(NaCl) = C0(PN)). 2.2.3. Spectrophotometry The electron spectra of pyridoxine in range of 250–400 nm and absorbance interval of 0–2.5 were registered using UV-1800 spectrophotometer (Shimadzu, USA) with resolution of 1.0 nm. The inaccuracy of wavelength determination was ±0.5 nm. The error of absorbance determination did not exceed 0.004 units. Quartz cells with 10 mm absorbing layer width were used. PN concentrations changed from 5.1064 · 10−5 to 2.029 · 10−4 mol dm−3. The increasing of DMSO content in binary solvent led to the significant changes of pyridoxine spectrum (Fig. 1). The main reason of such spectrum variation is the shift of tautomeric equilibrium towards molecular form absorbing maximally at 284 nm. The line of 326 nm maxima refers to pyridoxine zwitterion. Our results

Fig. 1. Variation of pyridoxine (C = 2.029 · 10−4 mol dm−3) electron spectrum at DMSO concentration growth in binary solvent.

are in accordance with spectral data on aqueous solution of PN (see e.g. [10,19]). It is possible to determine the tautomeric equilibrium constant of pyridoxine basing its electron absorption spectra in aqueous and aqueous-organic solutions. The method is described in details in [9,10]. The proposed way suggests finding the areas of peaks referring to molecular and zwitterionic species of pyridoxine. Their ratio divided by a constant found via linearization of dependence of PN± peak area on PN0 peak area in the entire studied range of solvent compositions gives Kz [9,10]. The deconvolution of pyridoxine spectra was carried out employing MagicPlot software [20]. An example of spectrum deconvolution is given (Fig. 2): The calculation of microscopic equilibrium constants is evident from Hess's law knowing macroscopic ones and tautomeric equilibrium constants (see Scheme 1). The calculation of the Gibbs energy change of transfer of pyridoxine molecular and zwitterionic forms also makes no challenge knowing total ΔtrGPN and Kz. Data on Kz, Kb1 and Kb2 are given (Table 3).

Table 2 Initial concentrations of pyridoxine (C0(PN)), equilibrium concentrations of pyridoxine in hexane, aqueous and binary solvent phase ([PN]Hex, [PN]H2O and [PN]H2O–DMSO respectively), partition coefficients of pyridoxine between water or aqueous DMSO and hexane (Kp1 and Kp2 respectively), and Gibbs energy changes of pyridoxine transfer from water to aqueous DMSO at atmospheric pressurea and T = 298.2 ± 0.1 K. XDMSO 0

4

−3b

C (PN)·10 , mol dm [PN]Hex·104, mol dm−3 [PN]H2O·104, mol dm−3 [PN]H2O-DMSO·104, mol dm−3 Kp1c Kp2c ΔtrGo(PN) ± 0.20, kJ mol−1d

0

0.05

0.1

0.2

0.3

0.5

1.5365 0.0581 1.4784 – 0.0393 – 0

1.5355 0.0818 – 1.4537 – 0.0563 0.89

1.5360 0.1847 – 1.3513 – 0.1367 3.09

1.5360 1.5360 1.5360 0.0582 0.0578 0.0652 – – – 1.4778 1.4782 1.4708 – – – 0.0392 0.0391 0.0443 0 −0.01 0.30

a All experiments were carried out under atmospheric pressure which changed in the range of (0.097 to 0.104) MPa. b Standard uncertainty is 10−6 mol dm−3. c Standard uncertainty is 0.0001. d There are given the expanded uncertainties with the confidence level of 0.95 and sample size of 3 experiments.

Fig. 2. Experimental electron absorption spectrum of aqueous-dimethyl sulfoxide solution (XDMSO = 0.1) of pyridoxine of 2.029 · 10−4 mol dm−3 (solid line) and its approximation by Gauss lines (dashed line).

460

G.A. Gamov et al. / Journal of Molecular Liquids 221 (2016) 457–462

Table 3 Tautomeric equilibrium constants (Kz), microscopic protonation constants of pyridoxine in molecular and zwitterionic forms (lg Kb1 and lg Kb2 respectively) in aqueous DMSO, and Gibbs energy change of transfer of zwitterionic and molecular form of pyridoxine (ΔtrG(PN±) and ΔtrG(PN0) respectively) from water to DMSO at atmospheric pressurea and T = 298.2 ± 0.1 K. XDMSO

0

0.05

0.1

0.2

0.3

0.5

Kz ± 0.10b lg Kb1 ± 0.03c lg Kb2 ± 0.03c ΔtrG(PN±) ± 0.3 kJ mol−1d ΔtrG(PN0) ± 0.3 kJ mol−1d

2.35 5.19 4.82 0 0

3.51 5.03 4.48 1.0 0.8

2.51 4.44 4.04 3.2 3.3

2.28 4.16 3.80 0.1 0.1

0.28 3.54 4.09 1.4 −0.8

0.03 3.42 4.94 4.3 −0.3

a All experiments were carried out under atmospheric pressure which changed in the range of (0.097 to 0.104) MPa. b There are given the expanded uncertainties for every value with the confidence level of 0.95 and sample size of 3 experiments. c There are given uncertainties including errors of macroscopic Kb and Kz determination. d There are given uncertainties including errors of macroscopic ΔtrG(PN) and Kz determination.

The tautomeric equilibrium constant of pyridoxine in water found in the present report is in a satisfactory agreement with literature data (Kz = 2.3 [16] and Kz ~ 2.6 [9]). 3. Discussion 3.1. Pyridoxine protonation in aqueous dimethyl sulfoxide The macroscopic protonation constant of pyridoxine continuously decreases until XDMSO = 0.7 and shows some increasing at higher content of organic solvent (see Table 1). Even small amount of DMSO added to the system significantly reduces basicity of pyridoxine. The pyridine and its other derivatives studied before behave similarly. Pyridine, pyridine-3-carboxamide (nicotinamide) and isonicotinylhydrazine (isoniazid) show the minimum of protonation constant at 0.5, 0.3 and 0.7 m.f. of DMSO respectively [21–23 respectively]. In order to find the reasons of observed basicity change of pyridoxine in aqueous dimethyl sulfoxide one should analyze the protonation reaction thermodynamics employing solvation and thermodynamic approach. It suggests reviewing the solvation thermodynamics of all participants of equilibrium process, namely, hydrogen ion, base, and conjugated acid. The Gibbs energy change of proton transfer from water to aqueous DMSO is taken from [24]. ΔtrG0(PN) is determined in present paper. The thermodynamics of reaction transfer are calculated using the following well-known equations: ΔGrH2 Oð−SÞ ¼ −RT ln Kb

ð7Þ

Δtr Gr ¼ ΔGrH2 O−S −ΔGH2O r

ð8Þ

2O(−S) where ΔGH is Gibbs energy change of the reaction in water or bir nary solvent, ln Kb is for equilibrium constant in corresponding solvent (Table 2). The results are given (Fig. 3a). The changes of microscopic constants Kb1 and Kb2 should be analyzed separately (Fig. 3b, c). The analysis of Fig. 3a–c data shows molecular, zwitterionic and protonated species of pyridoxine to have positive maximum of Gibbs energy change of transfer at XDMSO = 0.1. This maximum is probably caused by strengthening the three-dimensional spatial net of water hydrogen bonds occurring at first additions of organic liquid to the solvent [25,26]. The difference between ΔtrG(PN0) and ΔtrG(PN±) is negligible, and ΔtrG(PNH+) is lower by ~1.5 kJ mol−1. It is a consequence of solvation of proton bonded with pyridoxine more efficient than that of free base. When DMSO concentration rises up to 0.2 m.f., both pyridoxine species are solvated as efficiently as in water, and their ΔtrG ~ 0. The

Fig. 3. Gibbs energy changes of pyridoxine protonation reaction and reagents [24] transfer from water to aqueous dimethyl sulfoxide. (a) – macroscopic thermodynamics are given; (b) – molecular species protonation is given; (c) – zwitterionic species protonation is given.

G.A. Gamov et al. / Journal of Molecular Liquids 221 (2016) 457–462

difference between solvation of neutral and zwitter-ionic forms of pyridoxine manifests itself at higher concentrations of DMSO. Gibbs energy change of neutral species transfer becomes slightly negative at XDMSO N 0.2 m.f. while affinity of zwitter-ion towards binary solvent decreases noticeably. Neutral species contribute into Gibbs energy change of reaction transfer at XDMSO = 0.1 m.f. only. At higher organic solvent content, the ΔtrGr is determined by solvation difference between protonated pyridoxine and H+ (Fig. 3b). The decreasing of zwitter-ionic species solvation effectiveness at XDMSO N 0.2 m.f. provides growth of the protonation constant (Fig. 3c). Thus, different species of pyridoxine behave differently in the binary solvent water-dimethyl sulfoxide. Basicity of neutral form decreases continuously at DMSO concentration increasing while protonation constant of zwitter-ionic species goes through the minimum at XDMSO = 0.2. The main reason of this dissimilarity is the difference between solvation contributions of species. In the aqueous ethanol, the solvation contributions of charged reagents were predominant at low content of alcohol (contribution of PNH+ at XEtOH = 0.1 and H+ at XEtOH = 0.2) [12]. In the range of 0.3– 0.5 m.f. of ethanol, the neutral or zwitter-ionic species of pyridoxine contributed mainly into ΔtrGr [12]. 3.2. Permeation of pyridoxine in model lipid membranes The direct measurements of organic or inorganic compounds solubility in lipid membranes are difficult. For this reason, organic solvents are often used as these membranes models. n-octanol is usually employed for estimating the biodistribution of drugs [27,28]. Hexane, non-polar solvent mimicking the surrounding of distributed compound in the center of phospholipid bilayer, is also used [29]. Hexane and noctanol using gives the same partition coefficients for some molecules at least [29]. It is possible to estimate the permeability coefficient of hexane for pyridoxine (Pm) using equations [18,29]: Pm ¼

Kp  Dm dx

ð9Þ

where Kp is for partition coefficient of compound, Dm is for diffusion coefficient of compound in a model membrane, dx is for width of bilayer (membrane). Kp are reported in this work, we take dx equal 4 nm [29]. The only difficulty is determining the diffusion coefficient of PN. In order to estimate this parameter well-known Stokes-Einstein equation could be used: D¼

kB T 6πηr

461

Table 4 Estimated hexane permeability coefficients (Pm) for pyridoxine from aqueous dimethyl sulfoxide at atmospheric pressurea and T = 298.2 ± 0.1 K. XDMSO

0

0.05

0.1

0.2

0.3

0.5

Pm, m·s−1b

0.022

0.031

0.075

0.022

0.022

0.024

a All experiments we carried out under atmospheric pressure which changed in the range of (0.097 to 0.104) MPa. b Standard uncertainty caused by partition coefficients determination error is 0.001 m s − 1 .

value of pyridoxine diffusion coefficient obtained from Stokes-Einstein equation. Regardless, whatever the true value of Dm is, Pm remains maximal at XDMSO = 0.1 m.f. This medium is the most appropriate to transport the B6 vitamin through membrane (see Scheme 2). Thus, in order to increase the pyridoxine comprehensibility it should be probably dissolved in aqueous dimethyl sulfoxide with XDMSO = 0.1 (or at least XDMSO = 0.05, where Pm = 0.031, which is still higher than in water).

4. Conclusions In the present paper, the constants of pyridoxine protolytic and tautomeric equilibria were determined in the wide range of binary solvent water-dimethyl sulfoxide compositions. The growth of DMSO content in the solvent resulted in extremal change of protonation constant (minimum at XDMSO = 0.7). The fraction of pyridoxine zwitterionic species increased first with DMSO content growth (until XDMSO = 0.1 m.f.), then decreased at XDMSO ≥ 0.2 m.f. The Gibbs energy change of pyridoxine transfer from water to aqueous DMSO was determined by immiscible phases (water/hexane or binary solvent/hexane) partition method. Employing tautomeric equilibrium data, ΔtrG0(PN) was divided by thermodynamics of molecular and zwitterionic species transfer. ΔtrG0(PN0) and ΔtrG0(PN±) showed positive maxima in the initial range of water-DMSO solvent composition (0.1 m.f.). Their further variation was quite different: a little stabilization of neutral species at XDMSO N 0.2 and growth of ΔtrG0(PN±). The thermodynamic analysis of pyridoxine protonation reaction and reagents transfer revealed neutral species contributing the Gibbs energy change of reaction transfer at XDMSO = 0.1 m.f. only. At higher organic solvent content, the ΔtrG of neutral form protonation is determined by solvation difference between protonated pyridoxine and H+. The

ð10Þ

where kB is for Boltzmann constant, T is for temperature, η is for dynamic viscosity coefficient of solvent, r is for particle radius. The pyridoxine molecule radius can be found from quantum chemical calculations (r ~ 3.5 Ǻ - data of present paper) or molecular volume (V = 154.10 Ǻ3 as it was computed using Molinspiration Cheminformatics software [30], hence, r ~ 3.3 Ǻ). The dynamic viscosity coefficient of hexane (0.2949 mPa s) is taken from [31]. Taking these values to Eq. (10) gives Dm = 2.2 · 10− 9 m2 s−1. Pyridoxine diffusion coefficient in water or aqueous DMSO would be lesser than that in hexane due to the higher viscosity of these solvents (see [32]). Permeability coefficients for pyridoxine calculated using Eq. (9) in water and binary solvent are given (Table 4). Given values are probably overestimated taking into account the fact that permeability coefficient of such order are more specific for small molecules (HCl, H2S, CO2 [29]). It is a consequence of overestimated

Scheme 2. Pyridoxine transport from aqueous dimethyl sulfoxide (XDMSO = 0.1) through model membrane into aqueous medium of cell. Pyridoxine concentration in the left corresponds to the physiological PLP level in blood serum of men of 19–50 years old [33].

462

G.A. Gamov et al. / Journal of Molecular Liquids 221 (2016) 457–462

decreasing of zwitter-ionic species solvation effectiveness at XDMSO N 0.2 m.f. provides growth of the protonation constant. Basing on the results of immiscible phases distribution experiment, the possibility to transport directionally the pyridoxine from aqueous dimethyl sulfoxide (0.05 or 0.1 m.f. of DMSO) to water through model membrane was shown. It may mimic the transport process of B6 vitamin from aqueous DMSO to aqueous medium of living cell. Acknowledgements Study was carried out in the framework of State Task of the Ministry of Education and Science of the Russian Federation and with financial support of Russian Foundation for Basic Research (project 16-3360017) and Council on Grants of the President of the Russian Federation under grant 14.Z56.16.5118-MK. Aleksandr Levantovskii, MSc, the developer of MagicPlot software, is greatly acknowledged. References [1] D. Dolphin, R. Poulson, O. Avramovic, Vitamin B6: Pyridoxal Phosphate Volume 1, Part B: Coenzymes and Cofactors, Wiley Interscience, New York, 1986. [2] R. Percudani, A. Peracchi, EMBO Rep. 4 (9) (2003) 850–854, http://dx.doi.org/10. 1038/sj.embor.embor914. [3] D.B. McCormick, Biochemistry of coenzymes, in: R.A. Meyers (Ed.), Encyclopedia of Molecular Biology and Molecular Medicine, Vol. 1, VCH, Weinheim, Germany 1996, pp. 396–406. [4] J.E. Leklem, Vitamin B-6, in: L. Machlin (Ed.), Handbook of Vitamins, Marcel Decker Inc., New York 1991, pp. 341–378. [5] M. Chan-Huot, A. Dos, R. Zander, S. Sharif, P.M. Tolstoy, S. Compton, E. Fogle, M.D. Toney, I. Shenderovich, G.S. Denisov, H.-H. Limbach, J. Am. Chem. Soc. 135 (48) (2013) 18160–18175, http://dx.doi.org/10.1021/ja408988z. [6] H.-H. Limbach, M. Chan-Huot, S. Sharif, P.M. Tolstoy, I.G. Shenderovich, G.S. Denisov, M.D. Toney, Biochim. Biophys. Acta, Proteins Proteomics 1814 (11) (2011) 1426–1437, http://dx.doi.org/10.1016/j.bbapap.2011.06.004. [7] M. Chan-Huot, S. Sharif, P.M. Tolstoy, M.D. Toney, H.-H. Limbach, Biochemistry 49 (51) (2010) 10818–10830, http://dx.doi.org/10.1021/bi101061m. [8] S. Sharif, D. Schagen, M.D. Toney, H.-H. Limbach, J. Am. Chem. Soc. 129 (14) (2007) 4440–4455, http://dx.doi.org/10.1021/ja066240h. [9] J. Llor, S.B. Asensio, J. Solut. Chem. 24 (12) (1995) 1293–1305, http://dx.doi.org/10. 1007/BF00972834.

[10] J.M. Sanchez-Ruiz, J. Llor, M. Cortijo, J. Chem. Soc. Perkin Trans. 2 (1984) 2047–2051, http://dx.doi.org/10.1039/P29840002047. [11] J. Llor, L. Munoz, J. Org. Chem. 65 (2000) 2716–2722, http://dx.doi.org/10.1021/ jo991821t. [12] G.A. Gamov, A.Y. Khokhlova, A.S. Gushchina, K.V. Grazhdan, V.A. Sharnin, J. Chem. Thermodyn. 97 (2016) 322–330, http://dx.doi.org/10.1016/j.jct.2016.02.011. [13] E. Lindner, Y. Umezawa, Pure Appl. Chem. 80 (1) (2008) 85–104, http://dx.doi.org/ 10.1351/pac200880010085. [14] R.P. Buck, S. Rondinini, A.K. Covington, F.G.K. Baucke, C.M.A. Brett, M.F. Camoes, M.J.T. Milton, T. Mussini, R. Naumann, K.W. Pratt, P. Spitzer, G.S. Wilson, Pure Appl. Chem. 74 (11) (2002) 2169–2200, http://dx.doi.org/10.1351/ pac200274112169. [15] V.A. Borodin, V.P. Vasil'ev, E.V. Kozlovskii, Zh. Neorg. Khim. 31 (1) (1986) 10–16 (in Russian). [16] D.E. Metzler, C.M. Harris, R.J. Johnson, D.B. Siano, J.A. Thomson, Biochemistry 12 (26) (1973) 5377–5392, http://dx.doi.org/10.1021/bi00750a022. [17] A.K. Lunn, R.A. Morton, Analyst 77 (1952) 718–731, http://dx.doi.org/10.1039/ AN9527700718. [18] A. Walter, J. Gutknecht, J. Membr. Biol. 90 (3) (1986) 207–217, http://dx.doi.org/10. 1007/BF01870127. [19] M. Ristila, J.M. Matxain, A. Strid, L.A. Eriksson, J. Phys. Chem. B 110 (2006) 16774–16780, http://dx.doi.org/10.1021/jp062800n. [20] http://magicplot.com (Accessed 22 June 2015). [21] Y.Y. Fadeev, V.A. Sharnin, V.A. Shormanov, Zh. Neorg. Khim. 40 (12) (1995) 1981–1983 (in Russian). [22] V.A. Sharnin, S.V. Dushina, M.A. Zevakin, A.S. Gushchina, K.V. Grazhdan, Inorg. Chim. Acta 362 (2) (2009) 437–442, http://dx.doi.org/10.1016/j.ica.2008.04.053. [23] T.I. Bychkova, G.A. Boos, Koord. Khim. 12 (2) (1986) 180–183 (in Russian). [24] C. Kalidas, G. Hefter, Y. Marcus, Chem. Rev. 100 (2000) 819–852, http://dx.doi.org/ 10.1021/cr980144k. [25] Y. Koga, Y. Kasahara, K. Yoshino, K. Nishikawa, J. Solut. Chem. 30 (10) (2001) 885–893, http://dx.doi.org/10.1023/A:1012719530195. [26] G.I. Egorov, D.M. Makarov, Russ. J. Phys. Chem. A 83 (5) (2009) 693–698, http://dx. doi.org/10.1134/S003602440905001X. [27] E.H. Kerns, J. Pharm. Sci. 90 (2001) 1838–1858, http://dx.doi.org/10.1002/jps.1134. [28] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Adv. Drug Deliv. Rev. 46 (2001) 3–26, http://dx.doi.org/10.1016/S0169-409X(00)00129-0. [29] E. Cuevasanta, A. Denicola, B. Alvarez, M.N. Moller, PLoS ONE 7 (4) (2012), e34562http://dx.doi.org/10.1371/journal.pone.0034562. [30] http://www.molinspiration.com/cgi-bin/properties (Accessed 26 June 2015). [31] J.H. Dymond, H.A. Oye, J. Phys. Chem. Ref. Data 23 (1994) 41–53, http://dx.doi.org/ 10.1063/1.555943. [32] J. Catalan, C. Diaz, F. Garcia-Blanco, J. Org. Chem. 66 (2001) 5846–5852, http://dx. doi.org/10.1021/jo010415i. [33] L.T. Miller, J.E. Leklem, T.D. Shultz, J. Nutr. 115 (12) (1985) 1663–1672.