Stability constants and complex formation equilibria between iron, calcium, and zinc metal ions with vitamin B9 and glycine

Abstract

The complexation equilibria of folic acid (FA) and glycine (Gly) were studied in aqueous solutions at room temperature (ca. 298.15 K) and in fixed ionic strength (0.15 mol.dm⁻³ NaNO₃), by means of potentiometry technique. The ferric (Fe³⁺), ferrous (Fe²⁺), calcium (Ca²⁺), and zinc (Zn²⁺) complex-forming capacities of FA and Gly, and their overall stability constants in aqueous solutions were obtained by applying the potentiometric data in Hyperquad 2008 computer program. From the determined stability constants of the metal complex species, the concentration distribution of the various metal ion complex species involving FA and Gly in solutions was estimated using HySS 2009 software. The complex species distribution diagrams were explained and discussed. Supplementary, the Gibbs free energies and the molecular structures of the formed complexes were evaluated and predicted using Gaussian 09 software molecular modeling and calculations.

GRAPHICAL ABSTRACT

Keywords:

• Folic acid
• Calcium
• Zinc
• Iron
• Complexation
• Potentiometry
• Molecular modeling
• Cytotoxic
• Antioxidant

Introduction

Vitamin B9 (folic acid, pteroyl-l-monoglutamic acid, FA), a water-soluble compound of the vitamin B9 group [1], is added to many food products to prevent folate deficiency in individuals [2]. Supplementation with FA is particularly important in pregnant women, as insufficient FA can cause neural tube defects in the developing fetus [3]. Moreover, folate deficiency is the most common cause of anemia after iron deficiency and is thought to increase the likelihood of heart attacks and strokes [4].

Glycine (Gly) is a non-essential amino acid that can be produced by body at a rate of 125 mg kg⁻¹ body weight with adequate total protein intake. Gly is generated from serine by glycine hydroxymethyl transferase, and threonine by L-threonine dehydrogenase and 2-amino-3-ketobutyrate coenzyme A ligase. For dietary source, Gly can be found in all foods, particularly beef (55 mg g⁻¹ protein) and pork (57 mg g⁻¹). Daily oxidation of Gly is about 90 mg kg⁻¹ body weight with an energy yield of 2.011 kcal g⁻¹ [5]. As an amino acid, Gly is the building blocks of protein and nucleotide (purine). As an antioxidant, Gly is one of the amino acids that make vital antioxidant peptide glutathione. Moreover, Gly is a major inhibitory neurotransmitter in brain [6].

Vitamin B9 (FA) is one of the most biologically important compounds, and information about their metal ions complexation properties are rare in the literature. However, the interaction between some transition metal ions and vitamin B9 resulted in the formation of a number of binary transition metals–FA complexes in which number of drugs can interfere FA supplement [7–14]. The present study was done to determine the global stoichiometric stability constants of calcium/zinc/iron metal ions, FA, and Gly binary and mixed ligand complexes using potentiometric technique, and by applying the row data on Hyperquad 2008 computer program. The elucidation of the plausible molecular structures of the vitamin B9–Gly complexes in solutions and the possible binding sites of FA to bind calcium, zinc, and iron metal ions, and the Gibbs free energy calculations of the complex species were done using Gaussian 09 molecular modeling software. In addition, calcium, zinc, and iron metal speciation studies were performed using Hyperquad Simulation and Speciation (HySS 2008) program, in order to evaluate the suitability of supplementation of vitamin B9 in pregnant women by studying the interaction of FA with iron, zinc, and calcium metal ions in the presence of Gly, the simplest amino acid.

Experimental

Materials and solutions

All the chemical materials used are of analytical reagent grade and were used without further purifications. FA and Gly were purchased from Sigma Aldrich (USA). A nitric acid (Panreac, Spain) solution was prepared and used after being standardized. Carbonate-free sodium hydroxide (NaOH) from Across Organics (USA) was standardized with potassium hydrogen phthalate (Sigma Aldrich, USA). All metal salts [calcium chloride dihydrate (CaCl₂.2H₂O) from Sigma Aldrich (USA), ferric chloride hexahydrate (FeCl₃.6H₂O), ferrous sulfate (Fe(SO₄)₂), and zinc nitrate hexahydrate (Zn(NO₃)₂.6H₂O) from Across Organics (USA)] were weighted accurately before preparing the solutions. To maintain the ionic strength in the solution, sodium nitrate (NaNO₃) from Across Organics (USA) was used. All solutions used throughout the experiments were prepared freshly in ultrapure water obtained from a NANO-pure ultrapure water system in which water was distilled and deionized with a resistance of 18.3 MΩ cm⁻¹.

Apparatus and procedure

The pH-potentiometric titrations were performed using a Metrohm 836 Titrando with a model 801 magnetic stirrer, coupled with a Dosino burette model 803. The electrode response can be read to the third decimal place in terms of pH units with a precision of ±0.001 and the potential with a precision of ±0.1 mV. The titroprocessor was coupled to a personal computer and the titration software TINET version 2.4 was used to control the titration and data acquisition [15]. The pH titrations were carried out in a 150 cm³ commercial glass vessel. The ionic strength of the solutions was maintained at a constant level by using the desired concentration of NaNO₃ solution as supporting electrolyte. The pH meter was calibrated with standard buffer solutions (pH 4.0 and 7.0) before and after each series of pH measurements. The results of strong acid versus alkali titrations were analyzed using a computer program (GLEE, glass electrode evaluation) that has been used for the calibration of glass electrodes by means of a strong acid–strong base titration [16]. The program GLEE has been developed as part of the Hyperquad suite of programs for stability constant determination [16]. This program provides an estimate of the carbonate concentration of the base, the pseudo-Nernstian standard potential and slope of the electrode, and optionally the concentration of the base and the values of pK_w (p at 25°C).

For the protonation constants estimation of FA and Gly, the following solutions were prepared (total volume 50 cm³) and titrated potentiometrically against a standard carbonate-free NaOH (0.1 mol dm⁻³) solution: (a) 0.003 mol dm⁻³ HNO₃+0.15 mol dm⁻³ NaNO₃; (b) Solution (a) +0.001 mol dm⁻³ FA; and (c) Solution (a) +0.001 mol dm⁻³ Gly. For the determination of the binary metal complexes, solutions containing FA, Gly, and metal ions (calcium, zinc, and iron) were titrated at 1:1, 1:2, and 1:3 metal ion to ligand mole ratios to fulfill the maximum coordination number of the metal ion, and for ternary systems, the 1:1:1 metal ion/FA/Gly ratio was used (table 1). The concentration of metal ion and ligand solutions in the titrated samples was always the same and varied in the range 0.00004–0.0001 mol·dm⁻³ (table 1). Each solution was thermostated at 298.15 K, where the solutions were left to stand for several minutes before titration. A magnetic stirrer was used during all titrations. Each titration was repeated at least 3 times under carefully controlled experimental conditions. Typically, more than 80 pH readings (points of potentiometric measurements) were collected and taken into account for each titration.

Table 1.  Metal and ligand concentrations that were used to determine the overall formation constants of the complexes at 298.15 K and in 0.15 mol·dm⁻³ NaNO₃ solutions.

	Ratios
Systems		TM (mol dm^−3)
L^a		1·10^−3
	1:1	1·10^−3
	1:2	5·10^−4
	1:3	4·10^−4
	1:1:1	1·10^−3

Note: FA, folic acid; Gly, glycine.

^aOnly in this system, T_M represents the ligand total concentration in the solution.

Data analysis

The Hyperquad2008 program was used to obtain protonation and complex stability constants from potentiometric data [18,19]. For this purpose, a fitting criterion based on the minimization of the nonlinear least-squares sum defined by the difference between the calculated and the experimental data of the titration curves was used (equation (1)):

These constants were presented as the overall formation constant (β_pqrs) with stoichiometric coefficients p, q, r, and s that represent metal ions, FA, Gly, and hydrogen atom, respectively. The formation constant is a function of complex species concentration and the free reactant concentrations [M], [FA], [Gly], and [H] are expressed by the following equations:

Hyperquad 2008 program permits the determination of formation constants of different complex species that can be formed in the aqueous solution simultaneously [17,18]. For each experimental titration point, the concentrations of the species in solutions are defined by the following mass balance equations:

where T_M is the total concentration of metal ions. This value was obtained from the initial amount of metal in solution n_M, metal concentration in the burette a_M, initial volume of solution v₀, and added volume of base v. Thus, the mass balance equations (5)–(8) are described as follows:

Various models with possible composition of formed complex were proposed in the program and the model that gave the best statistical fit and chemically sensible values was chosen. First, estimation was made for the stability constants of all formed complex species. Then, these estimations were refined by the Newton–Raphson iteration. Besides the formation constant value (logβ), the result of stability constant refinement may include the goodness of fitting, standard deviation of parameters, the concentrations of all species in the model at all data points, and number of calculated data points and its residuals, illustrating the difference between experimental and calculated values. Besides Hyperquad 2008, HySS 2009 program was also used for providing speciation diagrams to present the distributions of various complex species that were formed in electrolyte solutions at the selected pH ranges [18]. The graphic representation of the complex species concentration curves is given by the distribution diagram produced by the HySS modeling program [18,19], which furnishes a variety of data presentations, including tables of concentrations of all species present in solutions in the selected pH ranges. Besides the Hyperquad 2008 and HySS 2009 programs, results from Gaussian09 simulation were also used to support our experimental results [18,19].

In this work, Gaussian 09 program was used to predict the binding site of the ligands that contributed to complex formation and the structure of the complexes [19]. The model structure of these complexes were optimized using density functional theory (DFT) with the Becke's three-parameter hybrid method [20] in correlation with the Lee–Yang–Parr (LYP) (B3LYP) method, and the 6-31G basis set was used with diffuse and polarization +d for hydrogen and heavy atoms [21–25]. Along with geometry optimization, frequency analysis was done to obtain the thermo-chemical properties of the complexes. The B3LYP/6−31+G(d) basis set was chosen since it has been shown to be one of the suitable basis sets to approach the molecular orbital of metal complexes [22–26]. The validity of the structural optimization was checked by using normal-mode frequency analysis, in which the real minimum structure must exhibit positive value for all frequencies. The modeling was performed by employing a number of assumptions as simplification, such as the use of fully deprotonated ligands, only considering the complex formation of one metal ion and one molecule of ligand, not considering the usage of salt to maintain the solution's ionic strength [22–25].

Results and discussions

From the refinement of the pH-potentiometric titration data by Hyperquad 2008 program, the protonation constants as overall formation constants (logβ) were obtained according to the following equations:

These constants are then presented as the minus logarithm of a stepwise acid dissociation constant (pK_a) by the following equations:

The relationship between stepwise acid dissociation constants and overall association constants is expressed as in the following equations:

The acid dissociation constant (pK_a) values of Gly were found to be 2.32±0.01 and 9.62±0.01. By comparing our calculated values of this work with the literature values [26], it was found that there are apparent differences which are explained in regard to variations in experimental conditions. The acid dissociation constant values can give an idea about the pH at which the ligand is 50% ionized. Once the pK_a value is determined, the degree of deprotonation at any pH can be easily calculated and observed by the following equation:

The pH-potentiometric curves of Gly along with their speciation diagram are presented in figure 1. As shown in this figure, the first proton dissociation of Gly started at very low pH (approximately below pH 2) and the second proton dissociation started at its isoelectric point (pH ∼6).

Figure 1. Potentiometric titration and speciation diagrams of Gly for protonation constants determination at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃.

Four protonation constants as overall equilibrium constants [log β₁, log β₂, log β₃, and log (related to α–COOH), 12.1547 (related to β–COOH), 16.3104 (related to NH₂ amine group – site 3), and 19.9647 (related NH/CO (N-3 site) fragment of the pteridine ring)] of FA were determined by fitting the pH-potentiometric experimental data of FA on Hyperquad 2008 program (figure 2 and scheme 1). Another two low protonation constant values (pK_a<−1.5, 0.20) of FA investigated can be associated with (N-5 site) and (N-10 site), respectively [26]. As shown, these two values are very low and not significant in the physiological pH region. Thus, these values are not used in the pH-potentiometric data analysis measured in the range of 2.00≤ pH ≤ 12.5. The graphical distribution species diagrams in figure 2 showed that the protonated FA species FAH₄ started to form in very low acidic conditions (pH <2.5). While the FA species FAH₃ was formed in acidic conditions (pH ∼2.5–5), conversely, the species FAH₂ started to form at normal acidic conditions and extremely formed at neutral conditions (pH ∼3–7). The folate species FAH started to form in normal acidic conditions (pH ∼4) and completely formed in neutral and alkaline conditions (pH ∼5.5–9.5).

Figure 2. Potentiometric titration (a) and speciation diagrams (b) of FA for protonation constants determination at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃.

SCHEME 1. Proposed metal ion binary ligand complex formation equilibria involving FA.

The global stability constants (log β_pqrs) for the binary and mixed ligand systems involving metal ions (Fe³⁺, Fe²⁺, Ca⁺, and Zn⁺²) with FA and Gly as well as the pH intervals in which the data were collected are presented in tables 2 and 3. The complexation between metal ion (M) and FA and Gly can be described by the following general equilibrium equations:

where p, q, r, and s are the coefficients that indicate the stoichiometry associated with the possible equilibria in solution; p, q, r, and s are the coefficients for metal ion, ligand FA, ligand Gly, and H-atoms, respectively. The stoichiometries and stability constants of the complexes formed were determined by trying various possible composition models. The selected model was the one that gave the best statistical fit and seemed chemically sensible and consistent with the titration data, without producing any systematic drifts in the magnitudes of the various residuals.

Table 2.  Global stability constants (logβ_pqrs) of metal ion complexes involving FA and Gly at 298.15°K and in 0.15 mol·dm⁻³ NaNO₃ solutions.

	log
Species	Fe (III)	Fe(II)	Ca(I)	Zn(II)
acid
			–	–
			–	–

Table 3.  Overall formation constants of metal ions (M)–FA–Gly mixed ligand complexes at 298.15°K and in 0.15 mol·dm⁻³ NaNO₃ solutions.

	log
Species	Fe (III)	Fe(II)	Ca(I)	Zn(II)
[MFAGly]

In all titrations, the experimental pH titration curves recorded for the binary and ternary metal ion–FA–Gly systems, when compared with the pH titration curves of FA and Gly ligands alone, prove that complex formation between the metal ions studied (Fe³⁺, Fe²⁺, Ca⁺, and Zn⁺²) and the two ligands (FA and Gly) did occur. The ternary complex formation may proceed through either a stepwise or a simultaneous mechanism depending on the chelating potentials of FA and Gly. No precipitation occurred in all solutions during the titrations. However, slight turbidity was observed in the case of the Ca⁺–FA–Gly ternary system at pH ∼6.71–9.37.

The bioavailability of metals and their physiological and toxicological effects depend on the actual species present. The equilibrium distributions of various complex species formed in water solutions are shown as functions of pH (figures 3–6), and showed the following:

1. For the iron (III) systems involving FA and Gly, the protonated and deprotonated binary complexes, [, [], [FeGlyH₋₁], and [FeGly]⁺ start to form in very low acidic media and reach maximum of 92%, 95%, 5%, and 18% at pH values 3.7, 5.7, 4, and 7.3, respectively. The formed complexes [, [FeFA₂H₂], [FeGlyH₋₁], and [FeGly]⁺ disappeared at pH values 6.2, 7.9, 5.8, and 10, respectively. While the binary and mixed ligand complexes [FeFAH₋₁], [, [, [FeFAGlyH₄], and [ start to form in slightly acidic media and the two complex species [FeFAH₋₁] and [ achieved maximum formation (88% and 84%) at pH values 6.8 and 9, respectively. The two mixed ligand complex species [FeFAGlyH₄] and [ achieved maximum complex formation (18% and 13%) in neutral media and disappeared in alkali media.

2.  For the divalent iron complexes involving FA and Gly, the protonated and deprotonated binary complexes [, [], [, [, and [FeGly]⁺ also start to form in very low acidic media and reach maximum of 64%, 72%, 31%, 51%, and 67% at pH values 4.8, 4.9, 4.2, 7, and 7.7, respectively. While the binary and mixed ligand complex species [FeFAH₋₁], [, and [FeFAGlyH₄] start to form in slightly acidic media and achieved maximum formation (47%, 26%, and 18%) in neutral media, respectively. These complexes disappeared in slightly alkaline media.

3.  The calcium binary and mixed ligand complex species [, [CaGly]⁺, [CaFAH₋₁], [, [CaGlyH₋₁], and [CaFAGlyH₄] started to form in very low acidic media. The complex species [, [CaGly]⁺, and [CaFAH₋₁] achieved maximum complex formation (53%, 52%, and 42%) in slightly acidic media, and ended in slightly neutral and alkaline media. The mixed ligand complex formation [ started to form at pH ∼6.7 and reached maximum (54%) in neutral solution and disappeared in slightly alkaline media.

4.  The binary and mixed ligand complexes involving zinc metal ion [, [ZnFA₂H₂], [ZnFAH₋₁], [, [, and [ZnFAGlyH₄] started to formed in very low acidic media, and both [ and [ZnFA₂H₂] complex species achieved complex formation maximum (60% and 53%) in low acidic media, and disappeared at pH values ∼4 and 7, respectively. While the complexes [ZnFAH₋₁] and [ completely formed (47% and 51%) in slightly neutral media, the complex species [ZnGly]⁺, [, and [ZnGly]⁻ started to form in neutral and slightly alkaline media, and achieved their maximum complex formation in alkaline media.

Figure 3. Species distribution diagrams for the Fe (III)+FA+Gly binary and mixed systems at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃. Percentages are calculated with respect to the analytical concentration of metal ion.

Figure 4. Species distribution diagrams for the Fe (II)+FA+Gly binary and mixed systems at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃. Percentages are calculated with respect to the analytical concentration of metal ion.

Figure 5. Species distribution diagrams for the Ca (I)+FA+Gly binary and mixed systems at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃. Percentages are calculated with respect to the analytical concentration of metal ion.

Figure 6. Species distribution diagrams for the Zn (II)+FA+Gly binary and mixed systems at T=298.15°K and I=0.15 mol·dm⁻³ NaNO₃. Percentages are calculated with respect to the analytical concentration of metal ion.

The logarithms of the overall stability constants (log β_pqrs) of the binary and mixed ligand complexes investigated are given in tables 2 and 3. From these data, the following was observed:

1. The global stability constants data of the binary complexes involving FA, and Fe³⁺, Fe²⁺, Ca⁺, and Zn²⁺ metal ions are decreased in the order of Fe complexes.

2.  The stability constants of ferric (III), ferrous (II), calcium (I), and zinc (II) binary complexes of Gly agreed well with my previous data [26].

3.  By comparing the stability constants of binary and mixed ligand complexes, it was demonstrated that mixed ligand complexation is responsible for the stabilization of the mixed complexes.

The negative values indicate the absence of undeprotonated amide nitrogen contribution in the complexation of metal ions with Gly. Gly, the smallest and the simplest ligand, can form five-membered ring complexes with metal ions via deprotonated amine and carboxyl group of the amino acid during the ternary complex formation (scheme 2). The phenomena of amide nitrogen contribution in the complex stability were further supported by the optimization and frequency analysis using Gaussian 09. From the geometry optimization, the structure of the complexes that formed can be predicted (scheme 2), and while the thermochemistry properties such as Gibbs free energy (G) can be obtained from the frequency analysis [27]. In this program, the complex equilibrium was presented as the Gibbs free energy of reaction (Δ_rG) that was calculated using the program. According to the aforementioned correlation, the more negative the Δ_rG, the larger the equilibrium constant (K_c) value will be, and the complex was found to be more stable (table 4).

SCHEME 2. Proposed metal ion mixed ligand complex formation equilibria involving FA and Gly.

Table 4.  Complex species Gibbs free energy of reaction obtained from optimization and frequency calculation by using Gaussian 09 program.

Molecule	G (hartree/particle)^a	ΔrG (hartree/particle)
Metal ions
Fe^+3
Fe^+2
Ca^+
Zn^+2
Ligands
FA
G
Fe(III) complexes
Fe(II) complexes
Ca(I) complexes
Zn(II) complexes
		−0.19350333

^aThe calculation was done using the DFT-B3LYP method combined with 6−31+G(d) as a basis set, 1 Hartree kJ mol⁻¹.

Funding

This work was financially supported by the King Khalid University, Abha, Kingdom of Saudi Arabia, through the project number [KKU_S112_85].