Complexes of Copper and Iron with Pyridoxamine, Ascorbic Acid, and a Model Amadori Compound: Exploring Pyridoxamine’s Secondary Antioxidant Activity

The thermodynamic stability of 11 complexes of Cu(II) and 26 complexes of Fe(III) is studied, comprising the ligands pyridoxamine (PM), ascorbic acid (ASC), and a model Amadori compound (AMD). In addition, the secondary antioxidant activity of PM is analyzed when chelating both Cu(II) and Fe(III), relative to the rate constant of the first step of the Haber-Weiss cycle, in the presence of the superoxide radical anion (O2•−) or ascorbate (ASC−). Calculations are performed at the M05(SMD)/6-311+G(d,p) level of theory. The aqueous environment is modeled by making use of the SMD solvation method in all calculations. This level of theory accurately reproduces the experimental data available. When put in perspective with the stability of various complexes of aminoguanidine (AG) (which we have previously studied), the following stability trends can be found for the Cu(II) and Fe(III) complexes, respectively: ASC < AG < AMD < PM and AG < ASC < AMD < PM. The most stable complex of Cu(II) with PM (with two bidentate ligands) presents a ΔGf0 value of −35.8 kcal/mol, whereas the Fe(III) complex with the highest stability (with three bidentate ligands) possesses a ΔGf0 of −58.9 kcal/mol. These complexes can significantly reduce the rate constant of the first step of the Haber-Weiss cycle with both O2•− and ASC−. In the case of the copper-containing reaction, the rates are reduced up to 9.70 × 103 and 4.09 × 1013 times, respectively. With iron, the rates become 1.78 × 103 and 4.45 × 1015 times smaller, respectively. Thus, PM presents significant secondary antioxidant activity since it is able to inhibit the production of ·OH radicals. This work concludes a series of studies on secondary antioxidant activity and allows potentially new glycation inhibitors to be investigated and compared relative to both PM and AG.


Electronic Supplementary Information
(58 pages) Contents: Table S1. Absolute enthalpies and Gibbs free energies of the different species considered in this study at the M05(SMD)/6-311+G(d,p) level of theory in water at 298.15 K. Table S2. 〈̂2〉 values for the calculated open-shell copper and iron complexes before and after annihilation of the first spin contaminant. Table S3. Standard formation Gibbs free energy change (∆°, kcal/mol) and formation constant ( , ) for the calculated complexes of Cu(II) and Fe(III) with ASCwith unusually low coordination numbers in aqueous solution at 298.15 K. Table S4. Standard Gibbs free energy of reaction (∆G°, kcal/mol) and activation (∆G ≠ , kcal/mol), various rate constants (k, kD and kapp, M -1 s -1 ) and the rate constant ratio (using kapp for the reduction of [ ( 2 ) 4 ] 2+ as reference) for the initial reaction of the Haber-Weiss cycle (with and without iron complexation with PM) with 2 •− in aqueous solution at 298.15 K. Table S5. Standard Gibbs free energy of reaction (∆G°, kcal/mol) and activation (∆G ≠ , kcal/mol), various rate constants (k, kD and kapp, M -1 s -1 ) and the rate constant ratio (using kapp for the reduction of [ ( 2 ) 4 ] 2+ as reference) for the initial reaction of the Haber-Weiss cycle (with and without iron complexation with PM) with ascorbate ( − ) in aqueous solution at 298.15 K. Table S6. Standard Gibbs free energy of reaction (∆G°, kcal/mol) and activation (∆G ≠ , kcal/mol), various rate constants (k, kD and kapp, M -1 s -1 ) and the rate constant ratio (using kapp for the reduction of [ ( 2 ) 6 ] 3+ as reference) for the initial reaction of the Haber-Weiss cycle (with and without iron complexation with PM) with 2 •− in aqueous solution at 298.15 K. Table S7. Standard Gibbs free energy of reaction (∆G°, kcal/mol) and activation (∆G ≠ , kcal/mol), various rate constants (k, kD and kapp, M -1 s -1 ) and the rate constant ratio (using kapp for the reduction of [ ( 2 ) 6 ] 3+ as reference) for the initial reaction of the Haber-Weiss cycle (with and without iron complexation with PM) with ascorbate ( − ) in aqueous solution at 298.15 K.     Using eq S1, the rate constant (k) was calculated following conventional transition state theory. The standard Gibbs free energy of activation (∆ ≠ ) was estimated applying Marcus theory [34,35]. ∆ ≠ is also known as the single electron-transfer activation barrier (∆ ≠ ), which is calculated using eq S2. In this equation, ∆ 0 is the standard Gibbs free energy of the reaction and λ is the reorganization energy, which can be calculated with eq S3. In this formula, is the nonadiabatic difference between the single-point energy calculations of reactants and vertical products.
When the value of k was above 10 8 M -1 s -1 (in the diffuse-limited regime), eq S4 was employed to calculate kapp (the apparent rate constant), following the Kimball-Collins theory [36].
In eq S5, R is the reaction distance (taken as the sum of the radii of the two reactants assuming a spherical shape), NA is Avogadro's constant, and DAB is the mutual diffusion coefficient of the reactants, computed as the sum of DA and DB. The Stokes-Einstein approach can be employed to determine these values, as shown in eq S6 [38,39]. In this equation, kB is the Boltzmann constant; T is the absolute temperature; η, the viscosity of the solvent (8.91 x 10 -4 Pa s for water); and αA or B is the radius of the solute (A or B), assuming it is spherical. = 6 (S6) S13 Appendix 2. Additional explanation on the pKa calculation for the neutral (zwitterion) model Amadori compound.
In a previous publication [46], Brown and Mora-Diez showed that the pKa of a chemical species can be found with greater accuracy relative to a reference acid of similar structure whose pKa value is known. They used this methodology to determine the pKa of various protonated benzimidazoles (HBz + ), employing eq S7, in which HBz1 + is the reference acid of known pKa: Using this approach, L-alanine (pKa (HAla) = 9.87) was selected as reference acid, given its structural similarity to the model Amadori compound studied, following the equilibrium shown in eq S8.
HAMD + Ala - AMD -+ HAla (S8) HAMD and AMDrefer to the neutral and deprotonated Amadori compound respectively. HAMD, the most stable form at physiological pH, is a zwitterion, and therefore the proton is present in the amine group, and not the carboxylic group. Optimizing the structures of both the zwitterion and the neutral molecule in aqueous solution showed that the zwitterion is much more stable, thus confirming this. AMDis the species that forms the most stable complexes with Cu(II) and Fe(III). Finally, HAla is the zwitterionic L-alanine, whereas Alais the deprotonated molecule. Consequently, the proton is migrating from one amine group to the other.
According to this approach, the unknown pKa can be obtained using eq S9, where (calculated using eq S10) refers to the equilibrium constant of the system shown in eq S8. This value is closely related to the experimental aqueous pKa value of L-alanine, which is structurally very similar to the model used to study the Amadori compound. S14 Appendix 3. Additional explanation on the pK calculation for the equilibrium between protonated pyridoxamine H2PM + and the anionic form PM -.
At physiological pH, the most stable form of PM is the protonated zwitterion (labelled H2PM(±) + in Scheme S11, which was taken from the paper by Casasnovas et al. [9]), but the species that forms the most stable complexes with Cu(II) and Fe(III) is the anionic form (labelled PM(-)in Scheme S11, and PMin our paper (see Figure 1)). To calculate the pK for the equilibrium between H2PM(±) + and PM(-) -, needed to estimate the Gibbs free energy cost in forming PM(-)at physiological pH, the information of Scheme S11 was used. The structures of the tautomers H2PM + and HPM were optimized considering all the possible conformations. It was found that H2PM(±) + and HPM(0) are the most stable tautomers at the M05(SMD)/6-311+G(d,p) level of theory. Thus, we followed the path that goes from H2PM(±) + to PM(-)through HPM(0). In order to be consistent, we made use of the experimental pKa values reported by Vilanova et al (indicated as the superscript a in Scheme S11, being b calculated pKa values taken from another reference) [52].
Note that the value shown to go from H3PM(+) 2+ to H2PM(±) + , 3.51, is the reverse of what is needed (-3.51). The resulting path, which includes the aqueous pKa values used for each acid dissociation, is represented in Scheme S12.