In vitro α-glucosidase, docking and density functional theory studies on novel azide metal complexes

Abstract

Aim: The goal of this study is to synthesize new metal complexes containing N-methyl-1-(pyridin-2-yl)methanimine and azide ligands as α-glucosidase inhibitors for Type 2 diabetes. Materials & methods: The target complexes (12–16) were synthesized by reacting N-methyl-1-(pyridin-2-yl)methanimine (L₁) with sodium azide in the presence of corresponding metal salts. The investigation of target protein interactions, vibrational, electronic and nonlinear optical properties for these complexes was performed by molecular docking and density functional theory studies. Results: Among these complexes, complex 13 (IC₅₀ = 0.2802 ± 0.62 μM) containing Hg ion showed the highest α-glucosidase inhibitory property. On the other hand, significant results were detected for complexes containing Cu and Ag ions. Conclusion: Complex 13 may be an alternate anti-diabetic inhibitor according to in vitro/docking results.

GRAPHICAL ABSTRACT

Plain language summary

Summary points.

• In this study, new effective α-glucosidase inhibitor series of Schiff base-azide metal complexes {[Cu₂(L₁)₂(N₃)₃], (12), [Hg(L₁)₂(N₃)·H₂O] (13), [Cd(L₁)₂(N₃)], (14), [Ag(L₁)(N₃)₂·2H₂O], (15), [Zn₂(L₁)₂(N₃)₂Cl₂·4H₂O], (16); L₁: N-methyl-1-(pyridin-2-yl)methanimine} were synthesized. The electronic spectral features for these complexes characterized by ¹H and ¹³C NMR (except for complex 12 including Cu ion), mass (LC-HRMS) and FT–IR spectra were examined by UV–Vis spectra.

• In complexes 12–16, being α-glucosidase inhibitory potential, the IC₅₀ values of 0.2802 ± 0.62 and 420.88 ± 1.42 μM range were obtained.

• Considering the CAM-B3LYP and ωB97XD/6–311+G(d,p)//LanL2DZ levels of density functional theory (DFT), the optimal ground state geometries and vibrational wavenumbers of the complexes 12–16 were computed, and the obtained results were compared with the corresponding experimental wavenumbers.

• The corresponding theoretical wavelengths, oscillator strengths as well as major contributions to the electronic transitions were calculated at TD-DFT/ωB97XD and TD-DFT/CAM-B3LYP levels. Additionally, the molecular orbitals and atoms/functional groups that make important contributions to electronic transitions were determined with the SWizard and Chemissian programs.

• Moreover, to survey nonlinear optical parameters of the complexes 12–16, the CAM-B3LYP and ωB97XD/6–311+G(d,p)//LanL2DZ levels of density functional theory were used.

• Demonstration of target protein interactions with complexes 12–16 has been fulfilled by molecular docking studies.

• It is deduced that while IC₅₀ results of the 15 and 12 against α-glucosidase are obtained remarkably, the β and γ results that complexes 15 and 12 are highly efficient in terms of microscopic second- and third-nonlinear optics properties, respectively.

• The experimental vibrational wavenumbers and absorption wavelength results were compared with corresponding theoretical ones, it was observed that the results were quite compatible.

1. Background

Many studies of azide metal complexes, in which azide ions (N₃^-) coordinate with various metal ions, such as Cu(II) [1–4], Zn(II) [5], Co(II) [5], Mn(II) [5], etc. to form stable complexes, are notable in coordination chemistry. The structural, spectral, optical, magnetic [6–8] and pharmaceutical [9,10] properties of these metal complexes have attracted great attention due to their different reactivity and potential applications in various fields. The azide ion serves as a versatile ligand with its linear trigonal planar geometry that can bind to metal centers via nitrogen atoms. Furthermore, this ligand has a high degree of coordination flexibility, which can also participate in the formation of monodentate, bidentate or polydentate complexes. Depending on the metal ion and coordination environment, azide complexes exist a wide variety of geometries, including distorted trigonal bipyramidal [3], distorted square pyramidal [5] and distorted octahedral [5,7] and bridge structures [1,2,4,11]. It was reported that synthesis of metal complexes including azide and/or heterocyclic ligands with various transition metals, such as cobalt [4,6–8], copper [1–4], zinc [5], iron [8], erbium [12] and platinum [13]. The utilization of these complexes underscores their role in advancing our understanding and utilization of these intriguing materials. Additionally, azide/azide derivative metal complexes, as well as compounds with heterocyclic rings including nitrogen, have been investigated for many biological activities. These include antimicrobial and antifungal properties [3], as well as anticancer potential [14,15]. Their capacity to interact with biological targets and demonstrate selective cytotoxicity underscores their significance in biomedical research [16–19]. In this context, studying azide metal complexes having various application potentials in the fields of physics, chemistry, material science and biology is valuable in terms of the synthesis of new azide metal complexes, revealing new results by investigating their different properties, as well as contributing to the development of new technologies.

α-Glucosidase enzyme (EC 3.2.1.20), having a significant impact on the digestion and absorption of carbohydrates, is found in the small intestines of the digestive system and is known to provide the digestive process by breaking down complex carbohydrates into simple sugars [20–24]. With the help of this enzyme, the 1,4-glycosidic bond in polysaccharides is hydrolyzed, resulting in the creation of α-D-glucose. As a result, during absorption, α-D-glucose is formed and enters the circulation, which raises postprandial hyperglycemia [23,25]. The activity of the α-glucosidase enzyme is a process regulated by using α-glucosidase inhibitors, which bind to the α-glucosidase enzyme and reduce or stop its activity, as well as slowing down the digestive process and allowing carbohydrates to be absorbed more slowly. It is noteworthy that α-glucosidase inhibitors are widely used in the treatment of Type 2 diabetes and that Type 2 diabetes is a common metabolic disease worldwide. Acarbose, miglitol and voglibose, which function as α-glucosidase inhibitors, are currently used as clinically prescribed drugs in the treatment of Type 2 diabetes [22]. Although some side effects of these inhibitors cause digestive system problems such as abdominal pain, diarrhea and bloating, there are already many studies using α-glucosidase inhibitor [26–40]. Therefore, scientists are studying new enzyme inhibitors [41–45] associated with various diseases. They aim to develop options and clinical treatments [46] that are both safer and more effective against specific targets. The synthesis of more effective new inhibitors taking these effects into consideration continues to be an important research topic today.

In the development of science and technology, it is extremely important to correlate quantum chemical results with experimental results and calculate quantities that can be obtained directly from experimental measurements. Many studies conducted in recent years show that theoretical and experimental research are handled together and density functional theory (DFT) methods are used extensively. The electronic structure and characteristics of molecular systems (organic, inorganic, etc.) are extensively studied. This is achieved using the DFT levels [47–50], providing valuable insights into their properties and behavior. Additionally, further advancements in understanding are facilitated through the application of more sophisticated computational methods [51–53]. The ligands used in these systems allow examining the electronic structure, geometry, spectral and optical properties of coordination compounds consisting of metal ions/atoms and obtaining detailed information about the structure. In addition to comparing the values calculated using DFT methods with experimental data, it is important to shed light on the stability, electronic transitions, optical properties and magnetic behavior of metal complexes and to enable the design of new materials for various applications. The development of improved exchange-correlation functionals and the inclusion of solvent effects have further enhanced the predictive power of DFT in studying metal complex behavior. It seems that the selection of functionals suitable for the complex structure from the developed exchange-correlation functionals and the inclusion of solvent effects further increase the predictive power of DFT in examining metal complex behavior. CAM-B3LYP (a long-range corrected version of B3LYP using the Coulomb attenuation method from Handy and co-workers) [54] and ωB97XD (the last functional including empirical dispersion and long-range corrections suggested by Head-Gordon and co-workers) [55,56] functionals are known to reliably predict the geometries of metal complexes, energies and spectroscopic properties, making them a popular choice for a wide range of applications.

Azomethine and azide groups are regarded as highly potent ligands in medicinal applications of coordination chemistry [5,17,20,22]. Because these ligands have a high degree of coordination flexibility, which can also participate in the formation of monodentate, bidentate or multidentate complexes. Moreover, it is well-known that they play important roles in potential biological activities. Therefore, these ligands were chosen in the synthesis of target complexes. So as to design and create more powerful α-glucosidase inhibitors, novel five metal complexes (12–16) including N-methyl-1-(pyridin-2-yl)methanimine (L₁) and azide (N₃^-) ligands were synthesized as a result of our ongoing studies. The structures of these complexes (12–16) were determined by ¹H and ¹³C NMR (except for complex 12 including Cu ion), mass (LC-HRMS) and FTIR spectra. The IC₅₀ values against α-glucosidase were obtained. In vitro α-glucosidase study on synthesized complexes and examination of their electronic properties by experimental and DFT methods play an important role in determining the structure–activity relationship. In addition, molecular docking study was conducted to determine the interactions of the target protein with the ligands and to interpret the activity results. Furthermore, the unique properties and capabilities of the DFT/ωB97XD and DFT/CAM-B3LYP functionalities allow these methods to obtain deeper insights into the behavior of metal complexes in structure–property relationships.

2. Experimental & computational details

2.1. Spectroscopic implements

The NMR spectrometer (Varian Infinity Plus, Oxford, UK) was utilized to define the structure of the synthesized the N-methyl-1-(pyridin-2-yl)methanimine (L₁) ligand and complexes 13–16. The synthesized compounds were dissolved in 500 μl of DMSO-d₆ containing tetramethylsilane (0.03 vol. %) as an internal standard. Temperature of the samples during the data acquisition was at 29°C. The ¹H and ¹³C NMR measurements were fulfilled at 300 and 75 MHz, respectively. The ¹H and ¹³C NMR spectra for each sample were obtained with 32 and 5000 accumulations, and pulse angles were 45 and 20 degrees, respectively. The structures of novel five metal complexes (12–16) were determined by using the SHIMADZU LCMS-9030 (OR, USA) model triple quadrupole liquid chromatograph mass spectrometer. The concentrations of samples were prepared at 50 ppm in methanol. LC-MS spectra were recorded at the following conditions: Mobile phase A: acetonitrile; Mobile phase B: water; pump system A:B concentration: 50:50%; flow: 0.5000 ml/min; pump A and B pressures: 30 bar; oven temperature: 28.7°C; interface type: electrospray ionisation; interface voltage (+)/(-): +4.50/-3.50 kV. The FT-IR spectra in the range of 4000–400 cm^-1 for complexes 12–16 in powder form were recorded at the Perkin Elmer UATR-TWO (Perkin Elmer Spectrum-two equipped with ATR) spectrophotometer (CT, USA). The methanolic solutions of samples were prepared at 10^-5 M and electronic absorption wavelengths (UV-Vis spectra) in the 900-200 nm range were obtained with the SHIMADZU UV-2600 UV-Vis spectrophotometer. Using the CHNS-932 (Leco, MI, USA) elemental analyzer, elemental studies of the synthesized complexes 12–16 in powder form were carried out.

2.2. The synthesis of the N-methyl-1-(pyridin-2-yl)methanimine (L₁) & complexes 12–16 {[Cu₂(L₁)₂(N₃)₃], (12), [Hg(L₁)₂(N₃)·H₂O], (13), [Cd(L₁)₂(N₃)], (14), [Ag(L₁)(N₃)₂·2H₂O], (15), [Zn₂(L₁)₂(N₃)₂Cl₂·4H₂O], (16)}

2-Pyridinecarboxaldehyde (99%), methylamine solution (33 wt.% in absolute ethanol), sodium azide (≥99.5%), copper(II) acetate monohydrate (≥99.0%, Cu(OAc)₂·H₂O), mercury(II) acetate (≥98.0%, Hg(OAc)₂), cadmium(II) acetate (99.995%, Cd(OAc)₂), silver nitrate (≥99.0%, AgNO₃), zinc chloride (≥98%, ZnCl₂) reagents, magnesium sulfate (anhydrous, ≥99.5%), triethylamine (for synthesis), ethanol (absolute for analysis), dimethyl sulfoxide-D6 (DMSO-d₆, ≥99.8%) purchased from Sigma-Aldrich (Sternheim, Germany) and Merck (Darmstadt, Germany) were utilized without any special reprocessing. Figure 1 depicts the synthesis methods of complexes 12–16.

Figure 1.

Synthesis of complexes 12–16.

rt: Room temperature.

The synthesis and NMR spectra of the N-methyl-1-(pyridin-2-yl)methanimine is presented in Supplementary Material (Supplementary Figure S1).

Synthesis of complexes 12–16 (Figure 1): 1 mmol of corresponding metal salts ((Cu(OAc)₂·H₂O/Hg(OAc)₂/Cd(OAc)₂/AgNO₃/ZnCl₂) were added to N-methyl-1-(pyridin-2-yl)methanimine (L₁) (2 mmol) dissolved in 15 ml of absolute ethanol. After 20 min, 2 mmol NaN₃ dissolved in 5 ml water was added to this solution. The mixture was stirred at ambient temperature for overnight (complexes 12–14) or for 3 h (complexes 15,16), then allowed to evaporate at ambient temperature. It took around 7 to 10 days to acquire complex compounds in powder form. The structures of these products (12–16) were confirmed by ¹H and ¹³C NMR (except for complex 12 including Cu ion), and mass spectroscopy. Complex 12: Anal. Calc. for C₁₄H₁₆Cu₂N₁₃: C, 34.08; H, 3.27; N, 36.90; found: C, 34.42; H, 3.45; N, 36.63. Calculated exact mass (m/z): 492.0244 (C₁₄H₁₆Cu₂N₁₃); found: LC-HRMS (m/z): 492.9582 ([M]⁺). Complex 13: ¹H NMR (DMSO-d₆, 300 MHz) δ/ppm: 8.66 (1H, d, J = 4.6 Hz), 8.62 (1H, s), 7.99–8.05 (1H, m), 7.91–7.94 (1H, m), 7.60–7.66 (1H, m), 3.51 (3H, s). ¹³C NMR (DMSO-d₆, 75 MHz) δ/ppm: 162.65, 162.60, 150.70, 150.65, 147.55, 140.98, 140.90, 129.24, 129.22, 129.20, 129.16, 128.30, 46.93, 46.84. Anal. Calc. for C₁₄H₁₈HgN₇O: C, 33.57; H, 3.62; N, 19.57; found: C, 33.86; H, 3.84; N, 19.02. Calculated exact mass (m/z): 502.1279 (C₁₄H₁₈HgN₇O); found: LC-HRMS (m/z): 502.7797 ([M]⁺). Complex 14: ¹H NMR (DMSO-d₆, 300 MHz) δ/ppm: 8.72 (1H, d, J = 4.7 Hz), 8.51 (1H, d, J = 1.5 Hz), 8.05–8.12 (1H, m), 7.90–7.93 (1H, m), 7.63–7.68 (1H, m), 3.41 (3H, s). ¹³C NMR (DMSO-d₆, 75 MHz) δ/ppm: 161.89, 150.06, 150.02, 140.70, 128.49, 127.45, 46.36. Anal. Calc. for C₁₄H₁₆CdN₇: C, 42.60; H, 4.09; N, 24.84; found: C, 42.12; H, 3.86; N, 25.27. Calculated exact mass (m/z): 396.0501 (C₁₄H₁₆CdN₇); found: LC-HRMS (m/z): 399.0392 ([M+2H]⁺). Complex 15: ¹H NMR (DMSO-d₆, 300 MHz) δ/ppm: 8.64 (1H, d, J = 4.2 Hz), 8.52 (1H, d, J = 1.5 Hz), 7.99–8.05 (1H, m), 7.86–7.88 (1H, m), 7.58–7.62 (1H, m), 3.56 (3H, s). ¹³C NMR (DMSO-d₆, 75 MHz) δ/ppm: 163.83, 151.61, 150.98, 139.06, 127.43, 125.50, 48.66. Anal. Calc. for C₇H₁₂AgN₈O₂: C, 24.15; H, 3.47; N, 32.19; found: C, 24.84; H, 3.75; N, 31.82. Calculated exact mass (m/z): 347.0134 (C₇H₁₂AgN₈O₂); found: LC-HRMS (m/z): 348.7087 ([M]⁺). Complex 16: ¹H NMR (DMSO-d₆, 300 MHz) δ/ppm: 8.65 (1H, s, br), 8.58 (1H, s), 8.16 (1H, s, br), 8.00 (1H, s, br), 7.73 (1H, s, br), 3.37 (3H, s). ¹³C NMR (DMSO-d₆, 75 MHz) δ/ppm: 162.43, 162.39, 149.76, 149.66, 140.64, 140.61, 128.22, 128.14, 128.08, 126.09, 126.02, 45.62, 45.54. Anal. Calc. for C₁₄H₂₄Cl₂N₁₀O₄Zn₂: C, 28.12; H, 4.04; N, 23.42; found: C, 28.65; H, 4.48; N, 22.94. Calculated exact mass (m/z): 593.9942 (C₁₄H₂₄Cl₂N₁₀O₄Zn₂); found: LC-HRMS (m/z): 597.1465 ([M+2H]⁺). The ¹H and ¹³C NMR (except for complex 12 including Cu ion), and mass (LC-HRMS) spectra were presented in Supplementary Material (Supplementary Figures S2–S10).

2.3. α-Glucosidase inhibitor activity

The IC₅₀ values against α-glucosidase of complexes 12–16 were defined taking into considering reported previously [30–32,36]. The detailed technique of α-glucosidase assay was provided in Supplementary Material. Taking into account A_c (the absorbance of the control) and A_s (the absorbance of the samples) in the equation below, the IC₅₀ values against α-glucosidase for complexes 12–16 were obtained.

$$\text{Glucosidase} \text{inhibition}\%=\left[\frac{({\text{A}}_{\text{c}}-{\text{A}}_{\text{s}})}{{\text{A}}_{\text{c}}}\right]\times 100$$

The calculations of IC₅₀ for complexes 12–16 were used the Graphpad Software.

2.4. Computational details

To investigate the features of structural, vibrational, electronic and nonlinear optics (NLO), the Gaussian 16, Revision C.01 [57] and GaussView 6 [58] programs were utilized. The CAM-B3LYP [54] and ωB97XD methods [55,56] with 6–311+G(d,p)//LanL2DZ [59,60] basis set in the ground state were applied to obtain the optimum geometrical structures, vibrational frequencies and natural bond orbital (NBO) results for complexes 12 and 16. For complexes 13–15, the same properties were examined by considering DFT/CAM-B3LYP and ωB97XD/LanL2DZ levels. Time-dependent DFT (TD–DFT) levels [61] with the conductor-like polarizable continuum model (CPCM) [62] were used to investigate electronic absorption wavelengths (λ) and oscillator strength of complexes 12–16 in MeOH. The SWizard [63] and Chemissian [64] programs were applied to define major contributions in electronic transitions. By using the CAM-B3LYP and ωB97XD levels in gas phase, the NLO properties (the parameters of microscopic polarizabilities (α and Δα), first- and second-order hyperpolarizabilities (β and γ) via equations presented in Supplementary Material were examined. Finally, the DFT/CAM-B3LYP level was utilized to survey the molecular electrostatic potential (MEP) surfaces.

2.5. Docking procedure

AutoDock4 program [65] and AutoDockTools (ADT1.5.7) [66] was utilized to examine the interactions between the Saccharomyces cerevisiae isomaltase (Protein Data Bank [PDB] ID: 3A4A) template structure provided from the Protein Data Bank and synthesized complexes 12–16. By considering the optimized structures of complexes 12–16 and genistein obtained at the DFT/CAM-B3LYP method, the inhibition constants (K_i) and binding energies of these complexes were calculated from docking findings. The Discovery Studio 4.0 program [67] provides 2D and 3D views of interactions between amino acid residues in enzyme structure and complexes 12, 13 and 15. The docking details are also given in Supplementary Material.

3. Results & discussion

3.1. The chemical descriptions of the produced complexes

The N-methyl-1-(pyridin-2-yl)methanimine (L₁) was characterized by ¹H and ¹³C NMR spectra. The structures of synthesized complexes 12–16 were verified by using ¹H and ¹³C NMR (except for complex 12 including Cu ion), mass spectra (Supplementary Figures S2–S10), as well as FTIR spectra. The aromatic proton signals for L₁ appeared at the range of 7.44 and 7.93 ppm, aromatic CH neighbor pyridine-N and CH protons belonging to imine were observed at 8.61 and 8.33 ppm, respectively (in ¹H NMR spectra given in Supplementary Figure S1). In ¹H NMR spectra of complexes 13–16, the aromatic protons, aromatic CH neighbor pyridine-N and CH protons belonging to imine shifting the peak downfield were obtained at the range from 7.58 to 8.16 ppm, from 8.64 to 8.72 ppm, and from 8.51 to 8.62 ppm, respectively (Supplementary Figures S3, S5, S7 & S9). The aromatic carbon signals for the L₁ ligand emerged at the range of 120.89 and 163.84 ppm, while these carbon signals for complexes 13–16 were detected at the range from 125.50 to 163.83 ppm (in ¹³C NMR spectra given in Supplementary Figures S1, S3, S5, S7 & S9). The carbon signal belonging to imine in L₁ ligand was observed at 154.81 ppm, whereas this signal shifting the peak upfield was obtained at the range of 149.76 and 151.61 ppm in complexes 13–16 (Supplementary Figures S3, S5, S7 & S9). Considering these chemical shifts in upfield or downfield in NMR spectra of L₁ ligand and complexes 13–16, the forming of complexes was supported. Besides, the structures of complexes 12–16 verified by mass spectra (Supplementary Figures S2, S4, S6, S8 & S10).

Supplementary Figure S11 displays the optimized structures (obtained at DFT/CAM-B3LYP) of complexes 12–16 confirmed by ¹H and ¹³C NMR (except for complex 12 including Cu ion), and mass spectra (Supplementary Figures S2–S10). DFT/ωB97XD and DFT/CAM-B3LYP levels were applied to obtain the geometrical parameters (bond lengths and angles and dihedral angles) given in Supplementary Table S1. The azide ions (N₃^-), can both enable the assembly of binuclear complexes and bind to transition metal atoms with different coordination modes [68].

Complex 12 [(Cu)₂(L₁)₂(N₃)₃] consists of the coordination of two L₁ ligands and one azide ligand with Cu(II) ion by forming a bridge with two Cu(II) ions over the N atom of the azide ligand. It was understood that a complex structure with distorted tetrahedral geometry was formed by the coordination of the L₁ ligand and azide ligands to Cu(II) ions via N atoms.

Complex 13 [Hg(L₁)₂(N₃)·H₂O] and complex 14 [Cd(L₁)₂(N₃)] consist of the coordination of two L₁ ligands and an azide ligand with the Hg(II) and Cd(II) ions, respectively. It was understood that both complex structures with distorted trigonal bipyrimidal geometry was formed when two L₁ ligands and an azide ligand coordinated to the Hg(II) and Cd(II) ions via N atoms. Besides, complex 13 includes the uncoordinated water molecule.

Complex 15 [Ag(L₁)(N₃)₂·2H₂O] is comprised of an L₁ ligand and the coordination of two azide ligands with an Ag(I) ion. It was understood that a complex structure with distorted tetrahedral geometry was formed when an L₁ ligands and two azide ligands coordinated to the Ag(I) ion via N atoms. Complex 15 contains also the uncoordinated two water molecules.

Complex 16 [(Zn)₂(L₁)₂(N₃)₂Cl₂·4H₂O] consists of the coordination of two L₁ and azide ligands and a Cl atom with Zn(II) ion by forming a tetrahedral bridge with two Zn(II) ions over the N atoms of two azide ligands. It was understood that a complex structure with distorted trigonal bipyramidal geometry was formed by the coordination of an L₁ and two azide ligands and a Cl atom to Zn(II) ion via N atoms.

These results are consistent with the coordination geometries of Cu(II), Zn(II) and Ag(I) coordinated with different ligands [1,2,5,6,69,70].

Supplementary Table S1 contains theoretical some geometrical parameters for complexes 12–16. These complexes are comprised of five-membered chelate rings occurred by L₁ ligand with metal ions. For complexes 12–16, the M(metal)–N2 bond lengths between the metal ion and pyridine N atom of the complexes were obtained at 2.143, 2.350, 2.418, 2.253, 2.280 Å by using DFT/CAM-B3LYP level. The similar bond lengths between the metal ion and imine N atom of the complexes were calculated at 2.086, 2.509, 2.263, 2.355, 2.135 Å by using the same DFT level, respectively. In addition, in all complexes, the theoretical bond lengths obtained by the DFT/ωB97XD level are similar trends. For complexes 12–16, the N2–M–N5 bond angles among imine N, metal ion and azide N in the coordination environment were found to be 78.09, 71.93, 68.82, 73.12, 74.87° with DFT/CAM-B3LYP level, respectively. On the other hand, the N–M–N bond angles among imine/pyridine N, metal ion and azide N in the coordination environment were observed the values greater or less than 90°, regardless of coordination geometry. These results indicate the presence of a distorted coordination geometry in the synthesized complexes 12–16. Despite the coordination environment/geometry differences, the theoretical results of the complexes are compatible with the previously reported ones [1,2,4,5,69–71].

Studying conjugative interactions, charge transfer in molecular systems and intra- and inter-molecular bonding are accomplished successfully with NBO analysis [72–74]. By considering the second-order perturbation approach [75,76], the hyperconjugative interactions were obtained. In NBO analysis, the large E⁽²⁾ energy value is an indication that the interactions between electron acceptors and electron donors are more intense and intramolecular and intermolecular charge transitions are high. The natural bond orbital calculation outcomes of the synthesized complexes 12–16 are given in Supplementary Table S2. The interactions between LP(n) nitrogen orbital pairs and LP*(n) metal(II)/(I) (Cu(II), Hg(II), Cd(II), Ag(I), Zn(II)) displayed the n→n* interactions indicate the coordination environment around the metal ions in the complex structures. The E⁽²⁾ values belonging to these interactions were obtained at the range of 6.01 and 47.25 kcal/mol at the DFT/CAM-B3LYP level (Supplementary Table S2). Additional significant interactions in the complexes indicate stabilization within the L₁ ligand. It is concluded that the bonding and antibonding orbitals exhibit charge transfer (CT) interactions, which cause the system to stabilize.

The FTIR spectra for the synthesized complexes 12–16 demonstrate the wavenumbers in the range of 4000–400 cm^-1. The corresponding theoretical vibrational frequencies computed at the DFT/ωB97XD and DFT/CAM-B3LYP levels to get closer to the experimental IR spectra were multiplied by 0.96. The experimental and theoretical characteristic vibrational wavenumbers were presented in Supplementary Figure S12, Supplementary Tables S1 & S4. The ν_CH(ring), ν_CH(imine) and ν_CH(methyl) stretching modes appeared at the range from 3095 to 2926 cm^-1 for complex 12, from 3137 to 2931 cm^-1 for complex 13, from 3089 to 2919 cm^-1 for complex 14, from 3070 to 2917 cm^-1 for complex 15, from 3070 to 2910 cm^-1 for complex 16. These results are coherent with previously reported for the Schiff base metal complexes [3,18,21,28,29,31,68]. The corresponding ones for complexes 12–16 obtained at DFT/ωB97XD were found at 3093–2904 cm^-1 range, 3150–2895 cm^-1 range, 3089–2897 cm^-1 range, 3117–2927 cm^-1 range and 3073–2910 cm^-1 range, respectively (Supplementary Tables S3 & S4). The vibrational bands emerging at the range of 2032 and 2060 cm^-1 were assigned as ν_NN asymmetric stretching modes belonging to the azide ligand. These findings are compatible with the modes reported in different complexes [2,4,5]. Theoretical these bands were calculated at 1885 and 2178 cm^-1, and 1844 and 2185 cm^-1 ranges, in order of the DFT/ωB97XD and DFT/CAM-B3LYP levels. Moreover, the ν_C=N(imine) stretching modes of the C=N bond (imine) of the L₁ for complexes 12–16 observed at 1654, 1653, 1658, 1648 and 1655 cm^-1 were obtained theoretically at 1662, 1671, 1687, 1692 and 1701 cm^-1 (with DFT/ωB97XD level), respectively (Supplementary Tables S3 & S4). Besides, the ν_NC(imine) stretching modes of the N–CH₃ bond of the L₁ for complexes 12–16 emerged at the range of 1015 and 1028 cm^-1 range, and the corresponding values computed at DFT/ωB97XD and DFT/CAM-B3LYP levels were obtained at the range of 1011 and 1025 cm^-1, 1006 and 1023 cm^-1, respectively. The ν_CC(ring) stretching modes attributed to ring C=C bond emerged within the ranges 1588–1623 cm^-1 in FTIR spectra, 1589–1614 cm^-1 in DFT/ωB97XD level, and 1588–1615 cm^-1 in DFT/CAM-B3LYP level, as can be seen in Supplementary Tables S3 & S4. These results are consistent with similar group frequency values in different structures in previous reports [5,20,22,26–28,71]. The vibration bands appearing at 1341, 1304, 1330, 1333 and 1340 cm^-1, which are assigned as the ν_NN(azide) symmetric stretching vibration, are attributed to the coordination of the metal ion with the N atom of the azide. The corresponding vibrational modes obtained at DFT/ωB97XD and DFT/CAM-B3LYP levels were found at the range from 1291 to 1362, and from 1283 to 1359 cm^-1, respectively. The β_HCC(ring) in-plane bending modes appeared at the range of 1039 and 1573 cm^-1 in FTIR spectra were found at the range 1039 and 1591 cm^-1 in DFT/ωB97XD level, and range 1036 and 1590 cm^-1 DFT/CAM-B3LYP level. Furthermore, the detailed stretching, in-plane and out-of-plane bending, torsion vibrational modes belonging to ring, azide, imine and methyl groups for complexes 12–16 are tabulated in Supplementary Table S4.

3.2. Analysis of the electronic absorption spectra & electronic transitions

The electronic absorption spectra and the plot of experimental E_g (optical band gap energy) in methanol for synthesized complexes 12–16 are given in Figure 2. The corresponding theoretical wavelengths, oscillator strengths as well as major contributions to the electronic transitions were computed at TD-DFT/ωB97XD and TD-DFT/CAM-B3LYP levels. Additionally, the molecular orbitals and atoms/functional groups that make important contributions to electronic transitions were determined with the SWizard [63] and Chemissian [64] programs. According to Figure 2A, three absorption peaks appeared in bridge-structured complexes 12 and 16, while two maximum absorption peaks emerged in other complexes 13–15. In complex 12, the absorption peaks appeared at 371.1, 264.5 and 203.5 nm, while in complex 16, the absorption peaks emerged at 280.6, 231.5 and 213.8 nm. The absorption peaks of 251.1 and 207.3 nm for complex 13, 264.5 and 244.5 nm for complex 14, and 258.4 and 204.6 nm for complex 15 were detected (Supplementary Table S5). These peaks were attributed to intra-ligand π→π*, ligand-to-metal/metal-to-ligand and n→π* transitions. Figure 3 displays ligand-ligand charge transfer in the high energy region as charge transitions within the ligand.

Figure 2.

Optical spectra. (A) The UV-Vis spectra and (B) the plot of experimental E_g (optical band gap energy) in methanol for complexes 12–16.

Figure 3.

Frontier molecular orbitals and their energies that contribute the most to the electronic transition obtained by the DFT/CAM-B3LYP method in methanol for complexes 12–16.

DFT/CAM-B3LYP: Density functional theory/Coulomb attenuation method- Becke, 3–parameter, Lee–Yang–Parr; FMO: Frontier molecular orbital: HOMO: Highest occupied molecular orbital; LUMO: Lowest occupied molecular orbital.

It has been reported that large contributions from metal orbitals to the lowest occupied molecular orbital (LUMO) in complex structures can increase the π-acceptor property of the ligand [77]. The metal orbital contributions to the highest occupied molecular orbital (HOMO)/HOMOs and LUMO/LUMOs of complexes 12–16 were obtained at the range from 2 to 37% (HOMO/HOMOs) and from 3 to 37% (LUMO/LUMOs), by using TD-CAM-B3LYP level. The LUMO/LUMOs contributions of the L₁ including 2-pyridyl and imine groups, and azide (N₃) ligands for complexes 12–16 vary between 20 and 100% (in L₁), 0 and 45% (in N₃), respectively (Supplementary Table S5). Due to the obtained larger LUMO/LUMOs contributions for L₁, the L₁ ligand has the more electron acceptor property. It is clear from Figure 3 that the electron density for complex 12 is localized in the Cu ion, L₁, and azide on one side for the HOMO α-level, localized in the L₁ and azide on the other side for LUMO α-level. In complexes 13 and 14, HOMO and LUMO α-levels are localized on the L₁ ligand including pyridyl and imine group in opposite parts. For the HOMO α-levels of complexes 15 and 16, the electron density is mostly localized on the metal ions and azide ligand, while in the LUMO α-levels, it is predominantly located on the L₁ ligand. (Figure 3). Supplementary Table S5 shows the remarkable metal-to-ligand charge transfer (MLCT), ligand-to-ligand charge transfer (LLCT) and ligand-to-metal charge transfer (LMCT) transitions. The 54% contribution for HOMO→LUMO α spin for complex 12 depicts the MLCT transition obtained at the contributions of Cu (20%), L₁ (30%) containing pyridyl (22%) and imine (8%) groups and N₃ (50%) ligand in HOMO, and L₁ (61%) and N₃ (36%) in LUMO. In complex 13, the absorption wavelength obtained at 536.2 nm (f = 0.0007) displays the contribution of 99% HOMO→LUMO changing from pyridyl (45%), imine (51%), H₂O (2%) to pyridyl (60%), imine (38%), N₃ (2%) and attributed to the LLCT transition. The LMCT transition for complex 15 was calculated at the contribution of 43% H-1→L β spin coming from pyridyl (7%), imine (2%), N₃ (90%) to Ag (37%), pyridyl (11%), imine (9%), N₃ (43%). The detailed percentage contributions in HOMOs→LUMOs transitions and assignments for HOMOs and LUMOs were presented in Supplementary Table S5, considering the metal ion and its environments. The frontier molecular orbital contributions in electronic transitions obtained DFT/ωB97XD level are presented in Supplementary Table S6.

MEP surfaces are used to understand the structure–activity relationship (SAR) for many molecular systems. These surfaces are related to electron density. The red and yellow areas on the MEP surface illustrate the negatively charged region with high electron density while the blue areas are spaces with positive charges. Green areas depict spaces where there is no potential and it is neutral [78]. MEP surfaces of complexes 12–16 were described by DFT/CAM-B3LYP level. The regions of complexes 12–16 with the highest electron density are shown in Supplementary Figure S13, where lone pairs of electronegative atoms are linked to the strong red electron clouds around the azide ligand. Areas showing nucleophilic reactivity were observed around the -CH groups in the L₁ ligand, which has the highest positive potential. In this way, the possible hydrogen bond interactions and reactive parts can be revealed in these complexes.

It is concluded from Figure 2B that the experimental optical band gap energy (Eg) values of complexes 12–16 were obtained at 4.01, 4.37, 4.09, 4.48 and 4.16 eV, in order. The HOMO-LUMO band gap results obtained from DFT/CAM-B3LYP calculations in α-spin level were obtained at 5.172, 3.797, 2.952, 5.672 and 6.422 eV, respectively. It seems that there are consistent results between experimental and theoretical approach errors. Molecular parameters χ (electronegativity), η (chemical hardness), S (chemical softness), ω (electrophilicity), φ (nucleophilicity) were calculated using HOMO and LUMO energy values [79] and their results are given in Supplementary Table S7. Chemical hardness (η) is known as the resistance of chemical systems to changing their electronic distribution. The η, χ, S parameters are indicators of chemical reactivity and stability of systems. The χ, η, S, ω and φ parameters of complexes 12–16 in α spin obtained at the DFT/CAM-B3LYP level were found at the range from 2.957 to 4.645 eV, from 1.476 to 3.211 eV, from 0.311 to 0.678 1/eV, from 2.302 to 3.360 eV, from 0.298 to 0.434 1/eV, respectively, seen in Supplementary Table S7. The low η parameters obtained in these synthesized complexes indicate that intra-ligand charge transfer (ILCT) occurs. These results are coherent with previously reported ones in different structures [77,80].

3.3. The analysis of nonlinear optics parameters

The synthesis and related computational research of NLO materials have been intensively studied due to their potential applications in optoelectronics and telecommunications technology, such as data storage and processing [81–87]. Theoretical NLO studies are taken into consideration in order to calculate the response of molecules to these properties if doing an experimental investigation is not feasible on the NLO properties of newly synthesized molecules. The findings obtained from these studies have the potential to shed light on electronic devices that are considered to be developed or designed. In this context, the $\overline{\alpha }$ and Δα indicating the linear optical (LO), β/γ demonstrating the second-/third-order NLO parameters of complexes 12–16 were obtained at DFT/ωB97XD and DFT/CAM-B3LYP levels in gas phase by using equations (7)-(10) [81,83] given in Supplementary Material, as can be seen in Supplementary Table S8.

To compare/evaluate the obtained theoretical results, p-nitroaniline (pNA) [88] and urea [89,90] were referenced as prototype compounds. According to the DFT/CAM-B3LYP level, the mean polarizability ($\overline{\alpha }$) values of complexes 12–16 in the gas phase obtained at 29.78 × 10^-24 and 49.75 × 10^-24 esu ranges exhibited higher than that of pNA (17 × 10^-24 esu) [88]. It has been observed that the polarization tendency increases in the large molecules. On the other hand, the β parameters of complexes 12–16 in the gas phase were obtained at the range of 2.28 × 10^-30 and 18.21 × 10^-30 esu, 2.33 × 10^-30 and 17.05 × 10^-30 esu in order of the DFT/ωB97XD and /CAM-B3LYP levels (Supplementary Table S8). While these results were observed to be greater than the β results of urea (0.32 × 10^-30 esu [89] and 0.130 × 10^-30 esu [90]), lower values than those of pNA (9.2 × 10^-30 esu [88]) were obtained, except for complexes 13,16. Based on the γ values for complexes 12–16 in the gas phase in the same levels, the highest values of 85.71 × 10^-36 and 96.93 × 10^-36 esu for complex 12 were calculated while the smallest values of 1.66 × 10^-36 and 15.32 × 10^-36 esu for complex 15. It is observed that the γ values of complex 12 are about 5.7/12.2 and 6.5/13.8-times higher than those of pNA (15 × 10^-36 esu) [88]/urea (7 × 10^-36 esu) [89]. It is concluded from the β and γ results that complexes 15 and 12 are high efficiency regarding the microscopic second- and third-NLO features, respectively. Depending on the electronic structure of the complexes, the effects of the metal ion and its coordination environment in the structure were also observed in the differences in the calculated NLO parameters.

3.4. α-Glucosidase inhibitory activity

The IC₅₀ values for complexes 12–16 were obtained at the range from 0.2802 ± 0.62 and 420.88 ± 1.42 μM, seen in Table 1. Among these complexes against α-glucosidase, the complex 13 (IC₅₀ = 0.2802 ± 0.62 μM) exhibits the strongest inhibitory activity, whereas the complex 14 (IC₅₀ = 420.88 ± 1.42 μM) displays the lowest inhibition effect. Complex 13 (IC₅₀ = 0.2802 ± 0.62 μM) was observed at 48.39-times more effective compared with genistein (IC₅₀ = 13.56 ± 1.21 μM), utilized as a standard in this study.

Table 1. In vitro inhibition IC₅₀ values (μM) for α-glucosidase of complexes 12–16 and ligands.

Ligand/complex	IC50 (μM)†
N-methyl-1-(pyridin-2-yl)methanimine (L1)	Not active
Sodium azide (NaN3)	Not active
Complex 12 [Cu2(L1)2(N3)3]	1.562 ± 0.83
Complex 13 [Hg(L1)2(N3)·H2O]	0.2802 ± 0.62
Complex 14 [Cd(L1)2(N3)]	420.88 ± 1.42
Complex 15 [Ag(L1)(N3)2·2H2O]	1.274 ± 0.32
Complex 16 [Zn2(L1)2(N3)2Cl2·4H2O]	171.5 ± 1.72
Genistein	13.56 ± 1.21

^†IC₅₀ values represent the means ± S.E.M. of three parallel measurements (p < 0.05).

For the SAR evaluation based on Table 1, the following results are deduced:

• The coordination of complex 13 [Hg(L₁)₂(N₃)·H₂O] and complex 14 [Cd(L₁)₂(N₃)] were defined as a distorted trigonal bipyrimidal geometry. Complex 13 (IC₅₀ = 0.2802 ± 0.62 μM), having the larger atomic diameter of Hg, has better inhibitory activity than complex 14 (IC₅₀ = 420.88 ± 1.42 μM).

• The coordination of complex 12 [(Cu)₂(L₁)₂(N₃)₃] and 16 [(Zn)₂(L₁)₂(N₃)₂Cl₂·4H₂O] have a bridge structure with azide ligand. However, complex 12 was defined as a distorted tetrahedral geometry formed by the coordination of the L₁ ligand and azide ligands to Cu(II) ions via N atoms while complex 16 was described as a distorted trigonal bipyramidal geometry formed by the coordination of an L₁ and two azide ligands, and a Cl atom to Zn(II) ion via N atoms. Besides the differences in coordination geometry and environment between complexes 12 and 16, complex 12 (IC₅₀ = 1.562 ± 0.83 μM) with a larger atomic diameter showed a higher inhibition effect than complex 16 (IC₅₀ = 171.5 ± 1.72 μM).

• Moreover, the α-glucosidase inhibition values of complexes 12 and 15 obtained at 1.562 ± 0.83 μM and 1.274 ± 0.32 μM, respectively. It can be stated that the small inhibition difference between complexes 15 and 12 is due to the difference in coordination in the structures as well as the fact that the atomic diameter of Cu metal is smaller than that of Ag.

• As a result, it is clear that there are differences in IC₅₀ values depending on the role of metal ions and the coordination environment.

The different organic compounds were reported as α-glucosidase inhibitors with the IC₅₀ values ranging from 4.51 ± 0.09 μM to 45.7 ± 0.23 μM [37–39]. The IC₅₀ value of the Zn complex including N₂O₂ was observed at 0.23 mM [20]. In addition, the IC₅₀ values of Cr, Mn, Co and Ni complexes including pyridine derivatives could not be observed but were obtained between 0.22 and 0.92 mM in other complexes [30,91,92]. Moreover, IC₅₀ values for transition metal complexes between 6-methylpicolinic acid/Schiff-base and isothiocyanate were obtained at the range of 0.2376 ± 0.82 and >600 μM [30,71].

Additionally, it can be stated that complex 13 (IC₅₀ = 0.2802 ± 0.62) may be a good candidate for α-glucosidase inhibitor when the inhibition result is compared with previous studies involving similar metal ions [31,32,40,71,93,94]. The IC₅₀ values of complexes 12 (1.562 ± 0.83) and 15 (1.274 ± 0.32) were also found to be significant. It is concluded that some metal complexes containing Schiff base and azide ligands tend to increase α-glucosidase inhibition.

Since the structures in the synthesized complexes are not of similar geometry and coordination, a direct correlation is not expected between density functional theory calculations and the reactivity and α-glucosidase inhibitor activity of the compounds. In addition, the reactivities of the compounds depend on many different parameters (steric effect, reaction medium, structures of other components and nucleophilic and electrophilic behaviors, etc.). For these reasons, a direct relationship between the chemical reactivity parameter results calculated using DFT methods for these complexes and their inhibitor activities could not be obtained. However, an inverse correlation was observed between electric dipole moment and inhibitor activity of synthesized complexes. Complex 13, the strongest inhibitor (IC₅₀ = 0.280 μM) in this study, has the lowest dipole moment (μ = 5.87 Debye), while complex 14 and 16, the weakest inhibitors (IC₅₀ = 420.88 and 171.50 μM, respectively), have the highest dipole moment (μ = 12.25 and 11.77 Debye, respectively). On the other hand, a linear increase was seen between LUMO energies with rising inhibitory activity. Complex 13 has the highest E_LUMO energy, -1.058 eV, whereas E_LUMO energy of complex 14 is -1.538 eV.

3.5. Molecular docking & inhibition mechanism

Using the AutoDockTools (ADT1.5.7) [66] graphical user interface with the AutoDock4 software, the ligand–protein interactions of the complexes 12–16 were determined, and S. cerevisiae isomaltase (PDB ID: 3A4A) as target protein was considered as rigid. The binding energies and estimated inhibition constants (K_i) were defined by docking results. Discovery Studio 4.0 program [67] was utilized to display the 2D and 3D structures of the interactions between ligands and amino acid residues (Figures 4–6 & Supplementary Figures S14 & S15).

Figure 4.

3D- and 2D-structures displaying interactions between amino acid residues and ligands of the most active complex 12.

Figure 5.

3D- and 2D-structures displaying interactions between amino acid residues and ligands of the most active complex 13.

Figure 6.

3D- and 2D-structures displaying interactions between amino acid residues and ligands of the most active complex 15.

The binding energy values of the genistein taken as a standard inhibitor [71] and complexes 12–16 were found to be -6.35, -4.77, -4.97, -4.43, -4.93 and -4.70 kcal/mol, respectively. The K_i values related to these energies were calculated at 22.06 μM, 316.95 μM, 226.71 μM, 567.97 μM, 241.40 μM and 358.10 μM, respectively, seen in Table 2. The best inhibition values against α-glucosidase among complexes were obtained for complex 13 (IC₅₀ = 0.2802 ± 0.62 μM), at same time the lowest K_i value (26.71 μM) and binding energy (-4.97 kcal/mol) were calculated by considering docking results. The in vitro results were supported by docking study in order of complex from lowest to highest inhibitory activity.

Table 2. Protein–ligand interactions and distance values for complexes 12–16 and genistein.

Substrate	Receptor	Interaction	Distance (Å)	Ki (μM)	Binding energy (kcal/mol)	Ref.
Complex 12				316.95	-4.77
Azide (N)	Glu322 (OE1)	Charge–Charge	5.14
Azide (N)	Phe321(O)	Conventional H-Bond	3.03
Azide (N)	Phe321	Pi–Cation	4.92
L1	Leu323 (CD1)	Pi–Sigma	3.34
Complex 13				226.71	-4.97
Azide (N)	Lys523 (O)	Conventional H-Bond	2.53
Imine-CH3 (C)	Gly361 (O)	Carbon H-Bond	3.36
Imine (C)	Phe321	Pi–Cation	4.35
L1	Leu323 (C)	Pi–Sigma	3.43
L1	Lys524	Pi–Alkyl	5.37
Complex 14				567.97	-4.43
L1 (N)	Phe321	Pi–Cation	4.60
L1	Leu323	Pi–Alkyl	4.98
L1	Lys524	Pi–Alkyl	5.38
Complex 15				241.40	-4.93
Azide (N)	Glu411 (OE2)	Charge–Charge	4.91
Azide (N)	Asp242 (OD2)	Charge–Charge	4.03
Azide (N)	Asp307 (OD1)	Charge–Charge	5.09
Azide (N)	Asp307 (OD1)	Charge–Charge	4.08
Azide (N)	Pro312 (O)	Conventional H-Bond	2.82
L1 (C)	His280	Pi–Cation	4.88
L1	Tyr158	Pi–Pi Stacked	4.65
L1	Lys156	Pi–Alkyl	5.34
Complex 16				358.10	-4.70
Azide (N)	Val404 (O)	Conventional H-Bond	3.13
Azide (N)	Tyr407 (O)	Conventional H-Bond	3.03
Imine (N)	Tyr416	Pi–Cation	3.75
Imine (C)	Tyr416	Pi–Cation	3.33
Imine-CH3 (C)	Tyr416	Pi–Cation	3.50
L1	Glu405 (OE2)	Pi–Anion	4.06
L1	Val404 (C)/Glu405 (N)	Amide–Pi Stacked	5.57
L1	Val404	Pi–Alkyl	5.50
Genistein				22.06	-6.35	[71]
C=O (benzopyran-4-one)	Ile272	Conventional H-Bond	2.73
CO (5-hydroxy-benzopyran-4-one)	Ser298	Conventional H-Bond	2.45
CO (7-hydroxy-benzopyran-4-one)	Leu297	Conventional H-Bond	3.29
CO (7-hydroxy-benzopyran-4-one)	Ser291	Conventional H-Bond	2.88
CO (7-hydroxy-benzopyran-4-one)	Ala292	Conventional H-Bond	3.14
Phenyl	Arg263	Pi–Alkyl	4.19
Phenyl	Ile272	Pi–Alkyl	5.25

Furthermore, the direct/indirect interactions with the enzyme active site change attaching to the coordination environment of complexes 12–16. The interactions of some parts of complexes with amino acid residues are given in Table 2, Figures 4–6 & Supplementary Figures S14 & S15. These interactions can be listed as the conventional hydrogen bond, pi–cation, pi–sigma, pi–alkyl and charge–charge. By virtue of the docking results of complexes 12, 13 and 15 exhibiting the best inhibition values against α-glucosidase among complexes, the interaction distance ranges from 2.53 to 4.92 Å emerged at the conventional hydrogen bond/pi-cation interactions between azide (N)/imine(C)/L₁(C) and amino acid residues Phe321(O)/Phe321, Lys523(O)/Phe321 and Pro312(O)/His280 for complexes 12, 13 and 15, respectively (Table 2). Besides, the other remarkable interactions demonstrate the pi–sigma for complexes 12, 13, the pi–alkyl for complexes 13, 15, as well as the pi–pi stacked for complex 15. These interactions, having distance ranging from 3.34 to 5.37 Å, were observed between L₁ with amino acid residues Leu323(CD1)/Leu323(C) for pi–sigma, Lys524/Lys156 for pi–alkyl, and Tyr158 for pi–pi stacked. In comparison these results of the genistein, the conventional hydrogen bond and pi–alkyl interactions obtained at 2.73/2.45/3.29/2.88/3.14 and 4.19/5.25 Å distances are remarkable for Ile272/Ser298/Leu297/Ser291/Ala292 and Arg263/Ile272 residues with the C=O (benzopyran-4-one)/CO (5-hydroxy-benzopyran-4-one)/CO (7-hydroxy-benzopyran-4-one) and phenyl ring, seen in Table 2. Other than these interactions, having amide pi–pi stacked and pi–anion for complex 16 appeared at the distances of 5.57 and 4.06 Å, were depicted between Val404(C)/Glu405(N) and Glu405(OE2) residues and the N-methyl-1-(pyridin-2-yl)methanimine (L₁).

The 1,4-glycosidic bond in polysaccharides is hydrolyzed via catalyzed by α-glucosidase, resulting in the creation of α-D-glucose. As a result, during absorption, α-D-glucose is formed and enters the circulation, which raises postprandial hyperglycemia [23,25]. α-glucosidase contains Asp, Glu, Gly, His, Phe, Thr, Tyr, Leu, Lys, etc. residues in the structure. Additionally, especially Arg, Asp, Glu, Trp and Tyr residues are located at the entrance of the active site of α-glucosidase and they have been played critical for the enzyme's catalytic activity [95–97]. As can be seen in Table 2, Figures 4–6 & Supplementary Figures S14 & S15, the synthesized complexes interacted with some of these residues by conventional hydrogen bond, pi–cation, pi–sigma, pi–alkyl and charge–charge interactions. It is considered that these interactions with the active pocket of glucosidase can hinder the substrate from entering, decrease the catalytic activity, and finally cause the inhibition of glucosidase activity. Moreover, hydrogen bond interactions can decrease the hydrophilicity of α-glucosidase and rise its hydrophobicity that supports to increase the stability of the inhibitor–enzyme complex.

A similar trend was obtained between docking K_i values and in vitro inhibition results. It is concluded that the donor and acceptor properties in molecular systems may give rise to similarities and differences in localization, or delocalization of electron density in the molecule, as well as interactions with amino acid residues in the enzyme structure. Furthermore, by considering the enzyme kinetic studies [26,28,98], the majority of the metal complexes are non-competitive α-glucosidase inhibitors. Considering the literature, it is predicted that 12–16 complexes may be non-competitive inhibitors.

3.6. Molecular dynamic simulation

Molecular dynamic (MD) simulation calculations were performed for complexes 12–16 in ligands with the receptor using the GROMACS simulation package [99] via WebGro server [100]. Root mean square deviation (RMSD) profiles for protein–ligand complexes were assessed. The protein–ligand topology was generated by using Discovery Studio 4.0 program (Systèmes D. Dassault Systèmes Biovia; CA, USA) [67]. The MD simulation details are also given in Supplementary Material.

Considering that the stability of the system is inversely proportional to the amplitude of the fluctuations, the decrease in the stability of the systems is expressed by a higher RMSD value and higher fluctuations. It was determined that the α-glucosidase enzyme has an average RMSD of 0.24 nm. It appears to remain stable throughout the MD simulation, except for some large fluctuations between 20 and 30 ns (Supplementary Figure S16). The average RMSD of the α-glucosidase–compound-13 system of 0.25 nm was obtained. While some fluctuations were observed during 15–25 ns, it was observed that the MD simulation remained stable until 50 ns.

The mean RMSD values of α-glucosidase–compound-12/14/15 were calculated as 0.23, 0.26 and 0.31 nm, respectively. It was observed that α-glucosidase-12/14/15 compounds were stable during the MD simulation, except for a few minor/major deviations at 10–15, 15–20 and 10–20 ns, respectively. (Supplementary Figure S16). On the other hand, the average RMSD value of α-glucosidase–compound-16 was obtained as 0.31 nm. It showed high stability throughout the entire 50 ns MD simulation, while some minor fluctuations were observed during 10–15 and 35–45 ns. Supplementary Figure S16 shows the RMSD plots for all complexes. It is thought that the amino acids in α-glucosidase cause fluctuations during the MD simulation because they have different structural regions (such as helices, β sheets and coil) [25].

4. Conclusion

In vitro α-glucosidase inhibitions and docking studies were performed on the N-methyl-1-(pyridin-2-yl)methanimine (L₁) and complexes 12–16 {[Cu₂(L₁)₂(N₃)₃], (12), [Hg(L₁)₂(N₃)·H₂O] (13), [Cd(L₁)₂(N₃)₂], (14), [Ag(L₁)₂(N₃)₂·2H₂O], (15), [Zn₂(L₁)₂(N₃)₂Cl₂·4H₂O], (16)} synthesized and characterized by ¹H and ¹³C NMR (except for complex 12 including Cu ion), mass (LC-HRMS) and FT–IR spectra. The electronic spectral behaviors were investigated by UV–Vis spectra. The IC₅₀ values of complexes 12–16 were obtained at the range from 0.2802 ± 0.62 to 420.88 ± 1.42 μM. The complex 13 demonstrates the best α-glucosidase inhibitory activity, while the complex 14 has the lowest inhibitory activity. Besides, the IC₅₀ values of complexes 12 (1.562 ± 0.83) and 15 (1.274 ± 0.32) were also found to be significant. These results were backed up by the docking ones presenting receptor–ligand interactions. Additionally, by considering a total of 50 ns MD simulations, the molecular dynamics study revealed that the compounds were stable during MD simulation. Furthermore, the detailed theoretical characteristic analyses were fulfilled using ωB97XD and CAM-B3LYP levels of density functional theory. The coordination geometries of complexes 12–16 were confirmed by NBO results appearing among lone pair electrons (n) of nitrogen atoms and metal ions. The high contribution to LUMO/LUMOs results shows that the L₁ ligand has the strongest electron acceptor properties than the N₃ ligand. It is deduced that while IC₅₀ values of the 15 and 12 against α-glucosidase are obtained remarkably, the β and γ results that complexes 15 and 12 are highly efficient in terms of microscopic second- and third-NLO properties, respectively. Among Schiff base–azide metal complexes, for Type-2 diabetes, complex 13 may be an alternate anti-diabetic inhibitor according to in vitro/docking results, as well as for NLO material, complexes 12 and 15 may be recommended as a candidate.

Funding Statement

The work was supported by the Scientific Research Projects Unit of Sakarya University (2021-7-25-103).

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/17568919.2024.2342650

Author contributions

D Avcı: methodology, software, formal analysis, writing-original draft, writing review and editing. Ö Özgea: data curation, formal analysis, visualization. F Sönmezc: formal analysis, investigation, writing, review and editing. Ö Tamera: investigation, software. A Başoğlua: methodology, software. Y Atalaya: methodology, software. BZ Kurt: formal analysis.

Financial disclosure

The work was supported by the Scientific Research Projects Unit of Sakarya University (2021-7-25-103). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.