The Counterion (SO₄²⁻ and NO₃⁻) Effect on Crystallographic, Quantum-Chemical, Protein-, and DNA-Binding Properties of Two Novel Copper(II)–Pyridoxal-Aminoguanidine Complexes

Abstract

New Cu(II) complexes with pyridoxal-aminoguanidine (PLAG) ligands and different counterions (SO₄²⁻ and NO₃⁻) were prepared and their crystal structures were solved by the X-ray crystallography. The geometries of the obtained complexes significantly depended on the counterions, leading to the square-pyramidal structure of [Cu(PLAG)NO₃H₂O]NO₃ (complex 1) and square-planar structure of [Cu(PLAG)H₂O]SO₄ (complex 2). The intermolecular interactions were examined using the Hirshfeld surface analysis. The theoretical structures of these complexes were obtained by optimization at the B3LYP/6-311++G(d,p)(H,C,N,O,S)/LanL2DZ(Cu) level of theory. The Quantum Theory of Atoms in Molecules (QTAIM) was applied to assess the strength and type of the intramolecular interactions and the overall stability of the structures. The interactions between the complexes and transport proteins (human serum albumin (HSA)) and calf thymus DNA (CT-DNA) were examined by spectrofluorometric/spectrophotometric titration and molecular docking. The binding mechanism to DNA was assessed by potassium iodide quenching experiments. The importance of counterions for binding was shown by comparing the experimental and theoretical results and the examination of binding at the molecular level.

1. Introduction

The introduction of cisplatin in the 1960s presents the starting point for the use of metal-containing compounds against different cancer types. Since then, many new transition metals and ligand systems have been introduced, as cisplatin is often associated with low selectivity, toxicity, resistance, and damage to different organs [1,2]. Copper is one of the most commonly investigated transition metals for preparing bioactive compounds due to its natural occurrence in the human body [3,4,5]. The cytotoxicity of copper complexes depends on the oxidation state of copper, the presence of various donor chelating atoms, geometry, dipole moment, and coordination number, as shown in a review by Ghorbanpour and coworkers [6]. Several groups of copper transition metal complexes have been developed, although it has been shown that the mechanism of action of these compounds is remarkably different from that of platinum drugs [7]. Some of the mechanisms include the activation of apoptosis through the effects of free radical species and the inhibition of angiogenesis [8]. The coordination of copper(II) with bioactive compounds also led to complexes with potent anticancer activity [9].

Pyridoxal-aminoguanidine (PLAG) ligand is obtained by the condensation between aminoguanidine and pyridoxal [10]. Both of these compounds are biologically active. Aminoguanidine (AG) prevents complications due to diabetes [11,12], as it suppresses the formation of advanced glycation end products [13]. On the other hand, pyridoxal represents one of the compounds from the vitamin B6 group that forms different transition metal complexes, especially after its modification to the Schiff base [14,15,16]. PLAG is more potent in treating diabetes than AG as it presents the source of vitamin B and allows better control of diabetes nephropathy [17]. Chen and coworkers showed that PLAG is a free radical scavenger due to chelation to transition metal ions [10]. The formation of PLAG in vivo was presented in reference [18]. Different crystal structures of PLAG with additional counterions are known in the literature [19].

PLAG is a tridentate ligand with an ONN donor system. Different complexation modes depend on the protonation of pyridine nitrogen, hydrazine nitrogen, and aromatic oxygen, as shown in Figure 1 [20]. Various transition metal complexes of PLAG with copper, iron, vanadium, and zinc are described in the literature [19,20,21,22,23,24]. Leovac and coworkers reported the synthesis and crystallographic structure of [CuCl₂PLAG] complex in a distorted pseudo-square-pyramidal geometry [21]. The photoluminescence properties of several copper(II) complexes with PLAG were examined in [25]. The DFT theoretical analysis of copper(II)-PLAG complexes allowed the elucidation of the ligand and counterion effects in [19]. The representative of deprotonated PLAG complexes is [Cu(PLAG-H)N₃], and its photoluminescent properties are described in the work by Jelić and coworkers [22]. Nitrate ions coordinated to copper(II)-PLAG were found in structure [Cu(PLAG)py(NO₃)]NO₃ [23].

The assessment of the biological activity of compounds usually includes interactions with transport proteins, for example, bovine serum albumin (BSA) and human serum albumin (HSA), DNA, and cytotoxicity screening towards cancer models [26,27,28,29,30]. Chakraborty and coworkers examined the DNA binding of [Cu(PLAG)Cl]₂(OAc)₂∙4H₂O and [Cu(PLAG)Cl(NCS)]∙H₂O, and it was concluded that both compounds bound to DNA with the binding constants comparable to some standard compounds [31]. Both compounds recognized L-histidine selectively. Coumarin-derived thiosemicarbazone copper(II) complexes showed anticancer potency towards MCF-7, MDA-MB-231, and multidrug-resistant MES-SA/Dx5 cells [32,33].

This study aims to present the results of the synthesis and X-ray crystal structure determination of two Cu(II) complexes with pyridoxal-aminoguanidine ligands and two different counterions. The structural differences due to the presence of these counterions are carefully examined. The theoretical structural analysis included the Density Functional Theory (DFT) and Quantum Theory of Atoms in Molecules (QTAIM), particularly emphasizing stabilization interactions. The binding affinity of the obtained complexes to important biomolecules (transport protein and calf thymus DNA (CT-DNA)) was followed by spectrofluorimetric titration at three temperatures, spectrophotometry, and the thermochemical parameters of the process were calculated. The molecular docking simulations were performed to elucidate the binding mechanism at the atomic level.

2. Results and Discussion

2.1. Crystal Structures

New copper(II) complexes with pyridoxal-aminoguanidine (PLAG) ligands were synthesized according to the procedure presented in the methodology section. Although the synthesis was performed with different initial simple salts (nitrates and sulfates), the chemical structures of the obtained complexes are surprisingly similar (Figure 2). In both complexes, the PLAG ligand is coordinated in a neutral form, although an additional coordination bond is present in complex 1 between the nitrate anion and central metal ion. The difference is that complex 2 had no coordination bond between the central copper(II) ion and the sulfate group from CuSO₄ (primary salts in the synthesis). This led to different coordination geometry: square-pyramidal for complex 1 and square-planar for complex 2. The reason for this may be the bulkiness of the sulfate group, which is much larger than the nitrate group. The color of the obtained complexes is highly dark green. Also, the syntheses were performed in ethanol, but the EtOH molecule did not enter the inner and outer coordination spheres. Crystalline water from simple copper salts (Cu(NO₃)₂∙3H₂O and CuSO₄∙5H₂O) is the one that coordinated in both newly synthesized complexes.

The square-pyramidal structure of complex 1 (Figure 2 and Figure 3) is interesting because the axially coordinated NO₃⁻ anion is only weakly bound to the central copper ion with a very long Cu-O distance (Cu(1)-O(9) 2.8248 (19) Å). For related Cu–Schiff base complexes, analogous distances are in the range of 2.530 (4)–2.551 (2) Å [34]. In both these and the current structure, the axial NO₃⁻ anion additionally forms several strong hydrogen bond interactions (see Figure 2), which may impact the structure. In comparison, the structure of [Cu(phen)(sal)(NO₃)] (sal = salicylaldehyde) in which there are no hydrogen bond interactions, the analogous Cu-O distance is 2.3666 (12) Å [35].

As already explained, the sulfate group is present in the outer sphere of complex 2. Figure 1 shows its structure, while Figure 4 shows the arrangement of the hydrogen bonds within a unit cell and among several cells.

Analyzing the square-planar geometry bond around copper(II) ion in both complexes leads to the conclusion that the value of the angles and the bond length are very similar. The selected bond lengths and angles that support this statement are listed in

Table 1. Characteristic bond lengths (in Å) and angles (in °).

Complex 1	Complex 2
Cu1–O1   1.8745 (17)	Cu1–O1   1.886 (3)
Cu1–N5    1.915 (2)	Cu1–N5    1.908 (3)
Cu1–O3    1.9418 (18)	Cu1–O3    1.937 (4)
Cu1–N2    1.959 (2)	Cu1–N2    1.974 (4)
Cu1–O9    2.8248 (19)
O1–Cu1–N5   173.83 (8)	O1–Cu1–N5   173.4 (2)
O1–Cu1–O3   91.18 (7)	O1–Cu1–O3   90.01 (18)
N5–Cu1–O3   94.96 (8)	N5–Cu1–O3   96.42 (15)
O1–Cu1–N2   92.16 (8)	O1–Cu1–N2   91.56 (18)
N5–Cu1–N2   81.83 (8)	N5–Cu1–N2   82.00 (15)
O3–Cu1–N2   171.31 (8)	O3–Cu1–N2   178.38 (13)
O1–Cu1–O9   83.44 (7)	O1–Cu1–O1   91.33 (10)
N5–Cu1–O9   96.40 (7)	N5–Cu1–O1   89.34 (9)
O3–Cu1–O9   86.15 (7)	O3–Cu1–O1   98.50 (16)
N2–Cu1–O9   102.20 (7)	N2–Cu1–O1  81.88 (13)

, while the complete list of the experimental parameters is presented in Tables S1–S4.

Hydrogens on the pyridine N1 and hydrazine N3 nitrogens prove that the PLAG ligand is coordinated in a neutral form in both cases. The square-planar (pyramidal) geometry is formed by the coordination of the tridentate (ONN) PLAG ligand with the central Cu(II) ion and oxygen O3 from coordinated water.

In both complexes, the shortest bond in the square plane is with oxygen O1 (phenolic oxygen), while the longest is with nitrogen N2. The pyridine nitrogen N1 with the adjacent C atoms in the aromatic ring overlapping angles less than 125° is additional evidence that PLAG ligand in neutral form was coordinated (C2–N1–C3 124.1 (2)—complex 1 and C2–N1–C3 123.7 (3)—complex 2). The importance of the counterion on the overall geometry surrounding the central metal ion is shown through these values. Additional theoretical investigation is undertaken to examine these effects in detail.

2.2. Hirshfeld Surface Analysis of Crystal Structures

The effects of the counterions on the overall stabilization of the crystal structure were examined by the Hirshfeld surface analysis, starting from the experimental structure. The Hirshfeld surfaces are presented in Figure 5 for both complexes, while the fingerprint plots are shown in the Supplementary Materials (Figures S1 and S2). In this discussion, particular emphasis was placed on the interactions involving counterions. It should be noted that complex 1 is considered square-pyramidal, with one NO₃⁻ in the coordination sphere, and this ion is included in the Hirshfeld surface.

The planar geometry of complex 2 allows the formation of interactions between two central metal ions, denoted as Cu∙∙∙Cu (0.1%), that were not found in 1. Other interactions including copper(II) are Cu∙∙∙H (0.9%) and Cu∙∙∙C (1.3%) in complex 1 and Cu∙∙∙O (4.4%) and Cu∙∙∙C (0.7%) in complex 2. These higher percentages in the case of 2 are a consequence of the overall geometry. The overall geometry and central metal ion profoundly affect the relative abundance of different interactions [36,37,38]. The most numerous interactions are formed between oxygen and hydrogen atoms in a structure of 1 (52.1%). This result is explained by the presence of oxygen-rich nitrate ions and relative hydrogen abundance in the structure of PLAG. The Hirshfeld surface of 1 (Figure 5) shows that protonated nitrogen atoms of the pyridine ring, hydrazine, and amino moieties behave as hydrogen atom donors towards neighboring species. The oxygen atoms in the structure are hydrogen atom acceptors. A similar can be found in the structure of 2. However, the relative contribution of these interactions is lowered to 34.8%, which is the essential difference between these stabilization interactions in 1 and 2. Weak H∙∙∙H interactions account for 23.6% (1) and 34.9% (2), similar to the previously described complex compounds with this type of ligand [20]. These percentages are a consequence of the protonation of nitrogen-containing groups.

It is essential to notice that interactions involving hydrogen atoms and nitrogen/carbon atoms are almost the same in the case of both compounds, around 5.0% (H∙∙∙N) and 4.0% (H∙∙∙C). Except for weak hydrogen bonds, interactions between hydrogen and carbon atoms include those formed between positively charged hydrogen atoms and negatively charged aromatic rings [39]. The planarity of complex 2 also allows the formation of interactions between carbon and nitrogen atoms in a higher percentage (8.9 vs. 4.2%). Other types of interactions contribute much less. It is important to note that the sulfur atom from the sulfate group is not included in any interaction as it is surrounded by oxygen atoms, which limits these interactions.

2.3. Theoretical Analysis of Structure and Interactions with Counterions

The structures of the obtained complexes were optimized at B3LYP/6-311++G(d,p)(H,C,N,O,S)/def2-TZVP(Cu) and B3LYP/6-311++G(d,p)(H,C,N,O,S)/LanL2DZ(Cu) levels of theory starting from the crystal structures. These theoretical models were selected as the common ones for describing spectra, stability, and magnetic properties of copper-containing complex compounds [40,41,42,43]. The comparison between the experimental and theoretical structures was performed by calculating the correlation coefficient (R) and mean absolute error (MAE), as explained in references [44,45]. The optimized structures containing counterions are depicted in Figure 6, while the bond lengths and angles are enlisted in Tables S1–S4.

The performance of two different basis sets for optimizing obtained compounds was examined by comparing the R and MAE parameters for the bond lengths and angles. When complex 1 is concerned, the R and MAE values for the bond lengths are 0.96 and 0.040 Å for both theoretical models. Slight differences exist when the MAE values for bond angles are concerned (1.77 (LanL2DZ) and 1.79° (def2-TZVP)), while the R values are 0.99 in both cases. The optimization of complex 2 also gave identical R (0.99) and MAE (0.049 Å) values for both basis sets. The correlation coefficients for the bond angles in the case of complex 2 are lower (0.88 and 0.89) with the MAE values of 4.28 (LanL2DZ) and 4.13° (def2-TZVP) due to the relaxation of the system, as explained in the following paragraphs. It can be concluded that both theoretical models reproduce the experimental structural parameters well. As similarities in the calculated values exist, only the structures optimized at the B3LYP/6-311++G(d,p)(H,C,N,O)/LanL2DZ(Cu) level of theory are described and used throughout the article.

Upon the optimization of complex 1, significant changes in bond lengths occur due to the absence of other compounds that interact with the polar groups of ligands, as shown in Figure 3. The shortest distance is between the phenyl oxygen atom and copper(II) (1.874 in the crystallographic and 1.966 Å in the theoretical structure). The distance between the amino nitrogen atom and the copper(II) ion is shorter than that with the hydrazine nitrogen atom (1.915 vs. 1.959 Å in the experimental structure). The water molecule is coordinated through O3-Cu1 interaction characterized by the bond length of 1.942 in the crystallographic and 2.046 Å in the optimized structure (Table S1). The longest distance between donor atoms and central metal ion includes nitrate ion oxygen (2.825/2.310 Å, Table S1). The presence of this counterion leads to a change in the geometry of the complex. However, the bond length is shortened after optimization, as the only interaction of this anion is with a central metal ion. In contrast, in the crystal structure, interactions with other units are present (Figure 3). When the isolated compound in a vacuum was optimized, the changes in bond angles were present, especially among the donor atoms and copper(II) ions. The distorted square-pyramidal geometry is equilibrated through slight changes in the angle values. The angles between phenyl oxygen and amino/hydrazine nitrogen atoms are 173.83/92.96° in the crystallographic and 164.48/87.79° in the optimized structure (Table S2). The weakly bound nitrate ion is also very flexible, leading to the difference between the angles in the experimental and theoretical structures (83.44 vs. 85.92° for the O1-Cu1-O9 angle). These differences were among the highest obtained when two data sets were compared. The bond angles within the optimized PLAG structure are much more similar to the experimental values due to the extended delocalization. The position of the second nitrate ion is changed in the optimized structure because of the interaction formation with the protonated nitrogen, hydroxymethyl, and amino groups of PLAG. These interactions lead to the loss of symmetry within the nitrate ions in the theoretical structure. As explained in the previous section, additional species are essential for stabilizing the experimental structure. The obtained parameters prove that the selected level of theory is appropriate for the description of the system.

The second complex is square-planar, with the bond lengths between the oxygen atoms and copper(II) ion being 1.886/1.937 in the crystal and 2.05/2.300 Å in the optimized structure (Table S3). The theoretical bond distances between the central metal ion and nitrogen atoms are 2.000 and 2.238 Å. Although differences between the experimental and theoretical bond lengths with the central metal ions are between 0.2 and 0.4 Å, the overall MAE value was 0.049 Å with a correlation coefficient of 0.99. After the optimization, the bond lengths between the copper(II) and donor atoms increased due to the system’s relaxation. This change is especially pronounced in bond length with a water molecule bound to a sulfate ion and a complex molecule in the crystal structure, as shown in Figure 4. A specific incline of water molecule towards phenolic oxygen atom and the formation of additional hydrogen bonds in an optimized structure that has not been experimentally observed. The changes in bond angles can be observed for the bonds around the central metal ion, leading to the distorted square-planar geometry. Some of the angles change for more than 20° upon optimization, which can be explained by the absence of the other stabilization interactions. Except for these angles, other significant differences were not found due to the elongated delocalization within the ligand structure, as previously mentioned. A more detailed overview of the stabilization interactions is presented in the following paragraphs.

The interactions between donor atoms and the central metal ion were examined by the QTAIM approach. This approach allows the investigation of the nature and strength of bonding in molecular systems based on the Bader theory of atoms in molecules [46] and the topological properties of electron density and Laplacian in Bond Critical Points (BCPs). Within this section, other weak interactions important for overall stability were also included. According to this approach, BCP is characterized by the electron density (ρ(r)), Laplacian (∇²ρ(r)), Lagrangian kinetic electron density (G(r)), potential electron density (V(r)), density of total electron energy (H(r) = G(r) + V(r)), and interatomic bond energy (E_bond = V(r)/2), as presented in [47]. Based on the classification by Bader and Essen, shared (covalent) interactions have high electron density (>0.1 a.u.), while closed-shell interactions (ionic bonds, van der Waals interactions, and hydrogen bonds) commonly have electron density of around 0.01 a.u [48]. The obtained values are shown in Table S5, with the graphical representation of BCPs in Figure S3.

According to the calculated parameters, two types of interactions within the examined structures exist. The highest values of electron density, between 0.076 and 0.082 a.u., and Laplacian, between 0.422 and 0.516 a.u., were obtained for the interactions between the donor atoms and copper(II) ion in complex 1. The values of electron density (0.035 a.u.) and Laplacian (0.185 a.u.) are much lower for the interaction between copper(II) and nitrate oxygen atom, as expected due to the longer interatomic distance. Although all the mentioned interactions have electron density lower than 0.1 a.u., the interactions between nitrogen atoms and copper(II) ions have partial covalent character, determined by the negative H(r) values and −G(r)/V(r) value being between 0.5 and 1 (Table S5). Additionally, both interactions have an absolute value of bond energy higher than 160 kJ mol⁻¹. The other three interactions with oxygen atoms have the ratio −G(r)/V(r) close to 1, with bond energies between −47.3 and 166.7 kJ mol⁻¹. This is consistent with previous findings that Cu-N bonds are slightly more covalent than Cu-O [49,50]. Due to its weak character, it can be expected that the coordination bond between nitrate ion and copper(II) might be broken under physiological conditions, as suggested in the following section. Several hydrogen bonds are present within the optimized structure with a wide range of electron densities (0.006–0.064 a.u.). The strongest hydrogen bond is present between the water molecule and the nitrate ion, coordinated to copper(II), a bond energy of −86.6 kJ mol⁻¹. This interaction shows a partial covalent character with a negative value of total electron energy, −36.8 kJ mol⁻¹. The position of the second nitrate ion is slightly distorted in the theoretical structure due to several interactions. The strongest interaction is present between the amino group of PLAG and the nitrate ion, with an interaction energy of 77.8 kJ mol⁻¹, negative H(r), and value of −G(r)/V(r) 0.81, thus making this interaction partially covalent [49]. The second interaction is formed between the hydroxymethyl group and the nitrate ion, with a bond energy of −27.0 kJ mol⁻¹, as found in the crystal structure. The aliphatic chain’s end amino group and methine group interact with nitrate ions through weaker, closed-shell interactions (Table S5).

In the case of complex 2, the strength of interactions changed in the absence of the additional coordination of counterions. The electron densities for the interactions with central metal ions are characterized by much lower electron densities, between 0.035 and 0.080 a.u., although these Cu-N and Cu-O bonds can be classified as closed-shell [49]. The increased interatomic distance between copper(II) ion and water molecule leads to much lower values of electron density (0.035 a.u.) and Laplacian (0.195 a.u.). The only interaction that shows partial covalent character is Cu1-N5, with a total electron energy of −18.4 kJ mol⁻¹ and bond energy of −173.3 kJ mol⁻¹, similar to the other square-planar copper(II) complex with trans-N₂O₂ ligands [51]. The rest of the interactions have much lower bond energies. As previously mentioned, the water molecule coordinated to the central metal ion position is significantly influenced by the formation of the hydrogen bond between the water molecule and the phenolic oxygen atom, denoted as O3-H∙∙∙O1, with a bond energy of −28.8 kJ mol⁻¹. This interaction explains specific differences between the experimental and theoretical bond lengths and angles. The presence of sulfate ion additionally stabilizes the overall structure through hydrogen bonds with the hydroxymethyl group (O2-H∙∙∙O4). The protonated pyridine nitrogen interacts with sulfate oxygen through N1-H∙∙∙O5 interaction, with a bond energy of −28.8 kJ mol⁻¹. The same oxygen atom also forms a weak hydrogen bond with the neighboring CH group (−10.5 kJ mol⁻¹).

2.4. Experimental and Theoretical HSA-Binding Affinity

The affinity towards HSA transport protein was examined by the titration of HSA solution by different concentrations of the complexes. The HSA solution was irritated by the excitation wavelength of 280 nm, which is sufficient for the excitation of tryptophan and tyrosine residues. The measured fluorescence intensities were corrected for the inner filter by measuring the absorbance of the complexes at the excitation and emission wavelength according to Equation (5) (methodology section). The position of the HSA fluorescence emission maximum at 340 nm was only slightly dependent on the concentration of quenchers, which proved that the polarity of the amino acid’s chemical environment was not changed [52,53]. The measurements were performed at three temperatures for a successive addition of complexes 1 and 2, as shown in Figure 7 and Figure S4.

The graphs in Figure 7 show that the fluorescence emission intensity of tryptophan residues decreased concentration-dependently. Two different quenching mechanisms could explain this change in intensity. The first one is dynamic quenching, which includes the collision between the excited fluorescent molecule and a quencher, and the excess energy is released by increasing the kinetic energy of the participating compounds. The second mechanism is static quenching characterized by the complex formation between the fluorescent molecule and a quencher. The quenching mechanism was determined by analyzing the Stern–Volmer (K_SV) constants obtained for the measurements at three different temperatures [53,54]. This equation gives a linear relationship between the fluorescence intensity of pure compound (F₀) divided by the intensity (F) in the presence of various concentrations of quenchers ([complex]). The slope of this dependency is the Stern–Volmer constant. This constant can be represented as the product between the bimolecular quenching rate constant (k_q) and the fluorophore lifetime (τ₀).

$${\displaystyle \frac{{\mathrm{F}}_0}{F}}=1+K_{SV}\left[complex\right]=1+{\tau }_0{k}_q[complex]$$

(1)

The dependency of K_SV on temperature is different for the two mentioned mechanisms. If the K_SV increases with an increase in temperature, the mechanism is dynamic quenching. Conversely, if this parameter decreases, the mechanism includes complex formation between quencher and fluorophore [53]. In dynamic quenching, the K_SV is dependent on diffusion, leading to its temperature increase. K_SV also gives information about the bimolecular quenching constant rate if time-resolved measurements are available. The K_SV values for each of the measurements are presented in

Table 2. The binding process’s parameters for the interaction between the obtained complexes and HSA.

Compound	T [K]	KSV [M−1]	Kb [M−1]	n	ΔHb [kJ mol−1]	ΔSb [J mol−1 K−1]	ΔGb [kJ mol−1]
1	300	9.10 × 105	6.82 × 105	1.17	121.2	515.2	−33.4
	305	8.81 × 105	1.29 × 106	1.22			−36.0
	310	8.60 × 105	3.27 × 106	1.31			−38.5
2	300	2.60 × 105	1.02 × 104	0.92	206.7	765.8	−23.1
	305	2.34 × 105	4.17 × 104	1.04			−26.9
	310	2.15 × 105	1.47 × 105	1.16			−30.7

. It should be noted that for these calculations, the intercept of the curve was set to one.

The dependency of the relative decrease in fluorescence emission intensity with concentration was linear for both complexes and at three temperatures. This proved that a single fluorophore in the protein was quenched by the presence of the obtained complexes. The K_SV values range between 9.10 and 8.60 × 10⁵ M⁻¹ for complex 1 and between 2.60 and 2.15 × 10⁵ M⁻¹ for complex 2. The order of magnitude of K_SV showed strong binding between the complexes and HSA [55]. The values of the constants decreased with the increase in temperature, which led to the conclusion that the quenching mechanism is static quenching [53]. In the paper by Hu and coworkers, the HSA fluorophore lifetime was taken to be 4.43 ns [56], which leads to the bimolecular quenching constants of the order 10¹³ M⁻¹ s⁻¹. These values are more than 1000 times higher than the maximum collisional quenching constant (2.0 × 10¹⁰ M⁻¹ s⁻¹) [56]. This is additional proof that the quenching of HSA fluorescence occurs through the static quenching mechanism. The isosbestic point in spectra is due to the significant absorption of complexes at the excitation wavelength of HSA. The increase in the second peak intensity is because of the increase in the concentration of complexes. Further examination of the static quenching mechanism is possible by plotting the double-logarithm regression curve with the modified Stern–Volmer equation (Equation (2)) [57]:

$$\log \left({\displaystyle \frac{{\mathrm{F}}_0-F}{F}}\right)=\log K_b+n\log \left[complex\right]$$

(2)

In the previous equation, F₀ and F are the fluorescence emission intensities of HSA without and with added metal complexes, K_b denotes a binding constant, n is the number of binding positions within the protein structure, and the concentration of metal complex is denoted as [complex].

The decrease in fluorescence emission intensity with the increased concentration of quencher, depicted in Figure 7, followed the mentioned equation with correlation coefficients of around 0.98. The number of binding positions was between 1.17 and 1.31 for 1 and between 0.92 and 1.16 for 2, which led to the conclusion that the complexes and HSA interact in 1:1 ratios (Table 2). The binding constants were one order of magnitude higher in the case of 1 (between 6.82 × 10⁵ and 3.27 × 10⁶ M⁻¹) when compared to 2 (between 1.02 × 10⁴ and 1.47 × 10⁵ M⁻¹). The thermodynamic parameters from measurements at three temperatures allow for obtaining additional information about the complex formation and driving forces [53]:

$$\ln K_b=-{\displaystyle \frac{∆H_b}{RT}}+{\displaystyle \frac{∆S_b}{R}}$$

(3)

The changes in the enthalpy and entropy of binding were positive for both compounds. Higher values were obtained for the binding of 1 (ΔH_b = 121.2 kJ mol⁻¹ and ΔS_b = 515.2 J mol⁻¹ K⁻¹). The mentioned values resemble those obtained for the [Cu(PLAG)(NCS)₂] complex in reference [20]. Based on these values, it can be concluded that non-specific interactions, mainly Van der Waals forces and hydrogen bonds, are responsible for the formation of complexes [58]. The entropic term (TΔS_b) is higher than enthalpic, leading to the entropy-driven processes [53] reflected in the loss of rotational and translational motion within the protein structure [20]. The change in Gibbs free energy of binding was between −33.4 (27 °C) and −38.5 kJ mol⁻¹ (37 °C) for complex 1. The spontaneity of the binding process was lower in the case of 2, between −23.1 and −30.7 kJ mol⁻¹. These results support the assumption that the presence of the second nitrate ion is important for the stabilization interactions within the active pocket.

Molecular docking simulations are a complementary approach for studying the interactions between compounds and biomolecules, allowing for the determination of the binding mechanisms [59]. Human serum albumin (HSA) contains only one tryptophan residue, Trp214, located in a secluded region of the FA8 binding site adjacent to the FA7 binding site in the IIA subdomain. This region of the FA8 site is enclosed by two helices at a sharp angle, making the area around Trp214 significantly less spacious than the FA1 binding site. Consequently, the area around Trp214 is less accessible for complex binding. Despite this, the literature reports indicate the binding of metal complexes in the FA8 binding site near the Trp214 residue, including Cu(II) complexes in structures 7Y2D [60] and 6L4K [61], a Pt-complex in structure 8ISM [62], a Pd-complex in structure 8J8E [63], a Fe-complex in structure 5GIX [61], and a Ru-complex in structure 5GIY [61].

Table 3. Change in Gibbs free energy of binding of the Cu(II) complex and counterions (SO₄²⁻ and NO₃⁻) with HSA obtained from spectrofluorimetric experiments at 300 K and molecular docking calculations at 298 K.

	Spectrofluorimetry at 300 K		Molecular Docking Calculations at 298 K
Target	Ligand	ΔGb [kJ mol−1]	Target *	Ligand	ΔGb [kJ mol−1]	Site
	Experiments were not conducted for separate species		HSA	Cu-PLAG	−28.0	FA1
				SO42−	−17.6	FA3
				NO3−	−14.6	IB, above Trp(214)
HSA	Complex 1	−33.4	HSA-NO3	Cu-PLAG	−29.7	FA7
	Complex 2	−23.1	HSA-SO4	Cu-PLAG	−27.6	FA8

* the targets can be the HSA structure alone, the HSA structure with a bound SO₄²⁻ ion (HSA-SO4), or the HSA structure with a bound NO₃⁻ ion (HSA-NO₃). Cu-PLAG refers to the copper complex. The abbreviation “Site” refers to the binding site denoted as FA, followed by a number from 1 to 9. The abbreviation FA means fatty acid, while the number following FA denotes the specific binding site.

presents the changes in Gibbs free energy of binding for the species included in the experiments.

The great advantage of molecular docking calculations is the ability to investigate the separate binding of Cu-PLAG ligands and counterions in a sequential manner. In solution, the binding of the ligand and a counterion co-occur. The binding energy values in Table 3 reveal that the binding of the Cu-PLAG has a significant advantage compared to the binding of counterions. While the binding energies for counters range from −14.0 and −18.0 kJ mol⁻¹, the binding energy of the Cu(II) complex is −28.0 kJ mol⁻¹ at the FA1 binding site, distant from the Trp214 amino acid residue. Interestingly, in the presence of counterions, the binding site of Cu-PLAG shifts near the Trp214 residue, and the change in Gibbs free binding energy remains close to −30.0 kJ mol⁻¹. The binding orientations of the Cu-PLAG and counterions, as determined by the molecular docking calculations, are presented in Figure 8 of the main text and Figures S5 and S6 in the Supplementary Materials. Among the binding positions of a counterion, one typically stands out due to its binding free energy and the number of orientations. Only the best binding position is depicted, and the others are discarded, as only the HSA adduct with the best-bound counterion was used for further molecular docking calculations. Figure 8 of the HSA adducts also includes the best binding orientation of Cu-PLAG. The other binding positions of Cu-PLAG are depicted in Supplementary Materials in Figures S5 and S6. For clarity, not all binding orientations are shown. If more than one binding orientation of Cu-PLAG is found at a single binding position, only the one with the highest change in Gibbs free energy of binding is shown. However, all the binding orientations are listed in Table S6, along with the information about the change in the Gibbs free energy of binding and the corresponding subdomain. In Figure 8 and Figure S5, it is evident that the nitrate ion is situated at the edge of the IB subdomain, above the Trp214. Ten ligand orientations are observed around nitrate ions. Since the nitrate ion is in proximity to Trp214, six out of ten ligand orientations are near Trp214. The sulfate ion is located in subdomain IIIA on the right half of HSA in Figure 8, and on the left half of HSA in Figure S6, while all the Cu-PLAG orientations are situated on the opposite half of the HSA molecule, suggesting that the sulfate anion repels away the complexes. Additionally, the lower ΔG_b value indicates that the sulfate ion weakens the binding of Cu-PLAG. Fortunately, this repulsion simultaneously shifts Cu-PLAG from the FA1 to the FA8 binding site near Trp214 in five out of ten orientations. In the presence of the nitrate ion, the binding energy increases from −28.0 to −29.7 kJ mol⁻¹. The nitrate ion not only enhances the binding of Cu-PLAG but also draws it into the vicinity of the Trp214 residue, improving the overall binding energetics and spatial arrangement.

In the binding of the Cu(II) complex, hydrogen bonds play a predominant role (Figure 9). The hydroxymethyl substituent of a ligand is predominantly responsible for hydrogen bonding. In the presence of different counterions, the Cu-PLAG binds to different amino acid residues and subdomains. The methyl group attached to the aromatic ring of PLAG is involved in the hydrophobic alkyl–alkyl interactions. The sulfate ions direct the Cu(II) complex to interact with amino acids from the IB subdomain, while with the nitrate ions involved, the Cu(II) complex interacts with amino acids from the IIA subdomain. Amino acids from the inter-subdomain regions are also involved.

2.5. Ethidium Bromide Displacement Studies

The interactions of the obtained compounds with DNA are usually examined as part of the biological assessment [64] of the leading anticancer regiments [65]. These interactions in the present research were investigated based on the removal of ethidium bromide (EB) to determine which of the two compounds formed stronger interactions. EB is commonly used as the indicator of intercalation [66]. The interactions between EB and CT-DNA lead to the formation of a stable fluorophore that is easily excited at 520 nm with an emission maximum of 600 nm. The fluorescence intensification results from the intercalation of the planar phenenthridinium ring between DNA base pairs on a double helix [67]. Upon the addition of compounds that interact with DNA, a partial decomposition of EB-CT-DNA structure occurs, which leads to fluorescence quenching [68]. EB in unbound form is quickly quenched by solvent molecules; therefore, measurable emission can not be observed [69]. As the DNA is an anionic polyelectrolyte with phosphate groups, the stable ionic strength of the solution is essential for weakening the electrostatic interactions between complexes and DNA. In these experiments, NaCl and KCl were used to control the ionic strength.

Figure 10 and Figure S7 present the emission spectra of EB-CT-DNA in the absence and presence of the investigated compounds. As can be seen, the emission maxima are located at 600 nm, and no visible change in this value was found throughout the experiment. The Stern–Volmer constants were calculated according to Equation (1), and their values are shown in

Table 4. Binding process’s parameters for the interaction between the obtained complexes and DNA.

Compound	T [K]	KSV [M−1]	Kb [M−1]	n	ΔHb [kJ mol−1]	ΔSb [J mol−1 K−1]	ΔGb [kJ mol−1]
1	300	1.50 × 105	7.34 × 104	1.16	−118.6	−300.5	−28.0
	305	1.29 × 105	3.62 × 104	1.09			−26.5
	310	7.70 × 104	1.59 × 104	1.07			−25.0
2	300	1.75 × 105	7.20 × 104	1.11	−58.4	−101.7	−27.9
	305	1.56 × 105	5.06 × 104	1.10			−27.4
	310	1.44 × 105	3.38 × 104	1.08			−26.9

. These values are in the range between 1.50 × 10⁵ (27 °C) and 7.70 × 10⁴ (37 °C) for complex 1. This parameter decreases with the increase in temperature, concluding that the quenching is static and interactions between the species lead to the deactivation of the fluorophore. For the second complex, the range of values is narrower, between 1.75 × 10⁵ and 1.44 × 10⁵ M⁻¹, with the same dependency on temperature. As already explained, the decrease in the K_SV values with temperature is an indication of static quenching and the formation of a complex between the fluorophore and quencher. Similar K_SV values were obtained for copper(II) complexes with 2-oxo-1,2-dihydroquinoline-3-carbaldehyde semicarbazone ligands, and the authors discussed that the probable binding mechanism was intercalation [70]. Two copper complexes with nitrate ions in the coordination sphere from reference [71] had lower K_SV values, probably due to the absence of polar groups that form interactions with the surrounding amino acids.

The intensity of emission lowered with the addition of compounds, according to the double log Stern–Volmer plots, with high correlation coefficients (>0.98). The number of binding positions was around one, proving that one molecule of compounds 1 and 2 interacts with one DNA molecule. It is important to observe that the binding constants were of the same order of magnitude (10⁴ M⁻¹), and their values decreased with an increase in temperature, for example, between 7.34 × 10⁴ (27 °C) and 1.59 × 10⁴ M⁻¹ (37 °C) for the binding of complex 1. The changes in the values of the thermodynamic parameters were negative in both processes, as shown in Table 4, leading to decreased spontaneity with temperatures. The main contribution to the negative change in Gibbs free energy is from the enthalpic term (ΔH_b > TΔS_b, both changes in absolute values), emphasizing the importance of hydrogen bonds and van der Waals forces for the stability of the formed complexes [72]. Large and negative enthalpy changes are typical for the intercalation processes within DNA, as shown for three water-soluble copper complexes [73]. The range of changes in Gibbs free energy was −28.0 to −25.0 kJ mol⁻¹ for complex 1 and −27.9 to −26.9 kJ mol⁻¹ for complex 2. At two temperatures (32 and 37 °C), the values of the change in Gibbs energy were comparable for both complexes, signifying that the spontaneity of the process is probably a consequence of the presence of counterions with a plethora of oxygen atoms that form different interactions with the surrounding amino acids.

2.6. Experimental and Theoretical DNA-Binding Affinity

As the previous study led to the conclusion that both compounds bind similarly to the DNA-EB complex, only complex 1 was selected for the absorption spectral studies. This is one of the most common ways to examine compounds’ interactions with DNA without adding other substances. When complex compounds are included, the hypochromism of bands in the visible region is observed due to strong interactions’ formation in the intercalation process [70]. Intercalation is especially common for the planar compounds, such as those examined in this contribution. Figure 11 presents the absorption spectra of the first complex in the absence and presence of various concentrations of CT-DNA. The band intensity decreased concentration-dependently upon the addition, proving that the interactions between the two compounds were formed. Slight shifts of about 1–2 nm were observed. It is important to mention that the CT-DNA solutions showed no significant absorption in the examined wavelengths’ range. The relative decrease in absorbance in the given CT-DNA concentration range was 36%. It can be assumed that the interactions between complex 1 and CT-DNA include the coupling of the π and π* orbitals of the intercalated molecules and base pairs, leading to decreased energy of transition, as shown previously for similar copper(II) complexes [70].

The change in absorbance upon CT-DNA can be used to calculate the binding constant (K_b) according to the Benesi–Hilderbrand equation [74]:

$${\displaystyle \frac{A_0}{A_0-A}}={\displaystyle \frac{{\mathsf{\epsilon }}_G}{{\mathsf{\epsilon }}_{H-G}-{\mathsf{\epsilon }}_G}}+{\displaystyle \frac{{\mathsf{\epsilon }}_G}{{\mathsf{\epsilon }}_{H-G}-{\mathsf{\epsilon }}_G}}{\displaystyle \frac{1}{K_b\left[DNA\right]}}$$

(4)

where Kb is the binding constant, A0 and A are the absorbances of the pure complex and complex mixed with CT-DNA, and ε_G and ε_H-G are the absorption coefficients of the complex and species formed through the interaction of the complex and CT-DNA. When the relative decrease in absorbance is plotted against reciprocal CT-DNA concentration, the binding constant can be determined as the intercept–slope ratio. When this plot was prepared for the measurements, the values of slope and intercept were −(4.01 ± 0.09)×10⁵ M and −(2.44 ± 0.04), respectively, leading to the binding constant of 6.08 × 10⁴ M⁻¹. The change in Gibbs free energy obtained from this value at 298 K is −27.3 kJ mol⁻¹. The value of Gibbs free energy is higher than that calculated at 300 K from the fluorescence measurements. This leads to the conclusion that the EB present in the solution influences the binding process between the complex and CT-DNA. The obtained value of the binding constant is similar to those calculated by Raja and coworkers for similar compounds [70], and they support the assumption of the intercalation binding. Parsekar and coworkers determined the Kb of Cu(II) complex containing a Schiff base ligand to be 1.12 × 105 M⁻¹, although that complex contained a much larger ligand system that allowed a higher number of interactions [75].

The possible binding mechanism was accessed through potassium iodide quenching experiments. Iodide ions are highly negatively charged quenchers of the intrinsic fluorescence of compounds, and the Stern–Volmer constant of this process depends on the binding mechanism between DNA and the complexes. If the complex is intercalated, then the K_SV of the complex’s fluorescence quenching process should decrease as iodide ions are repelled by the negatively charged phosphate backbone. If the values of K_SV are similar, then groove binding is a more probable mechanism [76]. The Stern–Volmer constants of the complex’s fluorescence by iodide ions were determined in the absence and presence of DNA molecules by Equation (1). The fluorescence emission spectra of complex 1 are presented in Figure 12. Adding iodide ions led to a decrease in emission intensity in a concentration-dependent manner. When DNA molecules were present, the starting fluorescence intensity was lower due to the interaction between DNA and complex 1. The values of K_SV in the absence and presence of DNA were 53.8 and 19.2 M⁻¹, respectively. It is apparent that the value of K_SV decreased upon the addition of DNA, which indicated that complex 1 has intercalated between the base pairs of DNA, similar to other planar copper(II) complexes [73].

From the diminishing absorbance signal observed upon adding CT-DNA to the solution of complexes in these experiments, it is evident that the Cu(II) complex intercalates with the DNA. Regarding molecular docking calculations, the most interesting DNA structures are those in the physiological B-form that include a gap (PDB ID: 1XRW). The other two DNA forms analyzed in the calculations include the plain B-form without any gaps or irregularities (PDB ID: 1BNA) and the Z-form of DNA that occurs under extreme conditions (PDB ID: 2ACJ).

The initial molecular docking calculations were conducted for all three forms of DNA structures. The first step was to compare the changes in the Gibbs free energy values and binding positions with the experimental results (Figure 12 and Figure S8). These values for binding to the B-DNA and Z-DNA forms were around −35 kJ mol⁻¹, almost 10 kJ mol⁻¹ higher than the experimentally obtained values, which led to the exclusion of these structures. For the DNA structure with PDB ID: 1XRW, the Cu-PLAG complex is located around the sugar–phosphate backbone or intercalated inside the gap of the DNA (Figure 13 and

Table 5. Gibbs free energy of the binding of the Cu(II) complex and counterions (SO₄²⁻ and NO₃⁻) with DNA obtained from spectrofluorimetric experiments at 300 K and molecular docking calculations at 298 K.

	Spectrophotometry at 298 K		Molecular Docking at 298 K
Target	Ligand	ΔGb [kJ mol−1]	Target *	Ligand	ΔGb [kJ mol−1]	Site
	Experiments were not conducted for separate species		DNA	Cu-PLAG	−27.1	Out (1)
					−26.9	Gap (6)
				SO42−	−10.8	Gap
				NO3−	−10.5	Out
					−10.8	Gap
DNA	Complex 1	−27.3	DNA-(NO3)2	Cu-PLAG	−27.3	Gap
	Complex 2	/	DNA-SO4	Cu-PLAG	−27.2	Out (2)
					−26.6	Gap (8)

* he targets can be the DNA structure alone, the DNA structure with a bound SO₄²⁻ ion (DNA-SO₄), or the DNA structure with two bound NO₃⁻ ions (DNA-(NO₃)₂). Cu-PLAG refers to the copper complex. Upon docking, one NO3− ion is intercalated inside the gap, and the other is bound in the minor groove away from the gap. The abbreviation “Orient. No.” indicates the number of orientations of the Cu(II) complex at a particular binding site. The abbreviation “Site” refers to the binding site, which can be located either inside the gap (gap) or outside the gap (out).

). The complex near the sugar–phosphate backbone possesses the best binding energy value of −27.1 kJ mol⁻¹. At the intercalation site, the Cu(II) complex reaches a similar binding energy of −26.9 kJ mol⁻¹. This is the preferred binding site, as it accommodates six Cu(II) complex orientations in total and it coincides with the potassium iodide quenching experiments. The calculated value of −26.9 kJ mol⁻¹ is similar to that obtained experimentally. Further calculations were continued with the 1XRW structure containing the gap since its initial binding energy and Cu(II) complex binding site match the spectrofluorimetric results. The other binding positions are depicted in Figure S9 with the corresponding binding energies in Table S7.

In the next set of calculations, the counter anions (SO₄²⁻ and NO₃⁻), present in the experimental solutions, were included to check whether the counterions interfere with the intercalation site. The optimal docking site for the SO₄²⁻ ion is the gap with a binding energy of −10.8 kJ mol⁻¹. The NO₃⁻ ion binds both into the gap and at the minor groove, with the binding energies of −10.5 kJ mol⁻¹ and −10.8 kJ mol⁻¹, respectively. In the initial calculation, which included only the Cu-PLAG, the binding energy inside the intercalation site was 2.5 times higher (−26.9 kJ mol⁻¹) than for the intercalated sulfate or nitrate ions. The binding energy for the site outside the gap was slightly higher (−27.1 kJ mol⁻¹). However, the statistics favor gap-binding (−26.9 kJ mol⁻¹) over the site outside the gap, with a ratio of 6:4 structures. Since the SO₄²⁻ and NO₃⁻ ions are more mobile than the heavier Cu(II) complex, it is expected that SO₄²⁻ and NO₃⁻ ions would attach to the DNA before the Cu(II) complex reaches the gap.

The last set of calculations was conducted for the Cu(II) complex on the adducts between the DNA structure and the sulfate or nitrate counterions. The presence of the SO₄²⁻ ion in the gap did not change the priority of the Cu(II) complex binding to the DNA. The best binding energy recorded for the binding outside the gap is barely increased (−27.2 kJ mol⁻¹). The second-ranked energy value is for the Cu(II) complex placed inside the gap (−26.6 kJ mol⁻¹), which is slightly lower (by 0.3 kJ mol⁻¹) compared to the case when counterions were excluded. In the presence of the SO₄²⁻ counterion, gap-binding is less favored than in the absence of the same anion. However, the intercalation of the Cu(II) complex remains the better option. Statistically, there are eight intercalated orientations of the Cu(II) complex compared to two orientations at the binding site outside the gap with the highest binding energy. The docking of the Cu(II) complex was repeated in the presence of two NO₃⁻ ions. Notably, the NO₃⁻ ion intercalates and simultaneously binds at the minor groove with similar probability. The NO₃⁻ ion improves the binding of the Cu(II) complex in the gap region, with the binding energy value increased to −27.3 kJ mol⁻¹. The calculated value corresponds well to the experimental binding of complex 1 and DNA, shown in the experimental section. Thus, it can be stated with certainty that the Cu(II) complex binds in the gap region together with a sulfate or nitrate counterion. The presence of counterions is important for stability, although only slight changes exist if nitrate or sulfate ions are included.

The final step is to discover the types of interactions (Figure 14) governing the established binding architecture. Common interactions involve the substituted pyridine ring of the PLAG ligand. The methyl substituent establishes alkyl-π interactions with cytosine and guanine nucleic acids from chain B and guanine from chain A. In supramolecular interactions, the hydroxymethyl substituent participates with both groups. Classic hydrogen bonds are characteristic of the OH group, while weaker carbon–hydrogen bonds are formed by the methyl groups.

There are additional interactions in the presence of the SO₄²⁻ counterion. All three hydrogen atoms of the methyl substituent on the pyridine ring participate in C-H∙∙∙O interactions (weak hydrogen bonds) with the SO₄²⁻ anion, while both hydrogen atoms of the CH₂ group of the hydroxymethyl moiety also establish C-H∙∙∙O interactions with a phosphate group in the nucleotide backbone. The coordinated water molecule participates in hydrogen bonding with phosphate groups or the aromatic nitrogen from the six-membered ring of the guanine base in chain A. The only interaction of the ending amino group of the Cu-PLAG with the intercalation site is a classic hydrogen bond between the N-H group from the aromatic ring and the oxygen from the phosphate group (N-H∙∙∙O).

There are notable differences in the binding of the Cu(II) complex in the presence of NO₃⁻ ions that explain its slightly increased binding energy. In the DNA-NO₃ adduct, the supramolecular interactions with DNA are evenly distributed between the pyridine and ending amino groups of the Cu(II) complex. The N-H amino group forms classic hydrogen bonds with the oxygen atoms from the sugar ring and the phosphate group of the sugar–phosphate backbone. Both the hydrogen atoms of the secondary amino group on the five-membered ring are engaged in classic hydrogen bonds with a phosphate group. There are no interactions involving the water molecule from the Cu(II) complex.

To briefly recapitulate the results of binding the Cu(II) complex together with the counterions to the intercalation site, Table 5, with the data obtained from the experiments and molecular docking calculations, was designed. Overall, intercalation inside the DNA is characteristic of the Cu(II) complex. This explains the experimentally observed quenching of the fluorescent signal. The absence or presence of SO₄²⁻ and NO₃⁻ ions does not alter the trend; it only fine-tunes the thermochemical parameters of binding. The Cu(II) complex intercalates together with the counterions. The SO₄²⁻ ion slightly decreases the affinity for intercalation, while the NO₃⁻ ion improves binding in the gap region. The A chain (Gua5) and the B chain (Cyt4 and Gua5) are involved in interactions with the Cu(II) complex. The calculated binding energy values are consistent with the experimentally obtained values and are similar in the presence of sulfate and nitrate ions. The slight advantage for binding in the presence of nitrate anions can be attributed to the evenly distributed interactions between the five- and six-membered rings of the Cu(II) complex and the DNA structure.

3. Materials and Methods

3.1. Chemicals

All the chemicals were obtained from Merck (Darmstadt, Germany) and used without further purification.

3.2. Synthesis of Ligand Pyridoxal-Aminoguanidine (PLAG)

A total of 0.70 g of aminoguanidine-hydrocarbonate (AG∙H₂CO₃) (5 mmol) was dissolved in 15 cm³ of H₂O. The dissolution was enhanced by heating up to the boiling point, and a warm solution of 1.0 g (5 mmol) PL·HCl in 5 cm³ H₂O was added. Additionally, 0.75 g (2.5 mmol) Na₂CO₃·10H₂O, previously dissolved in 10 cm³ H₂O and heated, was added to the abovementioned mixture. The mixture was then left at room temperature. After two hours, the yellow crystals were washed with EtOH and Et₂O. Yield: 1.15 g (92%). The obtained yellow powder was soluble in chloroform, dichloromethane, dimethylformamide, acetone, and acetonitrile; moderately soluble in methanol, ethanol, and water; and insoluble in diethyl ether and toluene.

3.3. Synthesis of [Cu(PLAG)(NO₃)(H₂O)]NO₃ (Complex 1)

Ligand PLAG (0.15 g—0.5 mmol) was dissolved in 15 cm³ of EtOH with gradual heating. When the ligand was dissolved entirely and the temperature was close to boiling, Cu(NO₃)₂∙3H₂O (0.12 g—0.5 mmol) was added to the solution. The mixture was heated for about five minutes. After a week, large dark green crystals were formed. The crystals were washed with EtOH and Et₂O. Yield: 0.15 g (82%). The dark green crystals were soluble in dimethylformamide, acetone, and acetonitrile; moderately soluble in methanol, ethanol, and a physiological solution of phosphate buffer and water; and insoluble in toluene and diethyl ether.

3.4. Synthesis of [Cu(PLAG)(H₂O)]SO₄ (Complex 2)

The ligand PLAG (0.15 g—0.5 mmol) was dissolved in 15 cm³ of EtOH with gradual heating. When the ligand was dissolved entirely and the temperature was close to boiling, CuSO₄·5H₂O (0.09 g—0.5 mmol) was added to the solution. Heating continued until the salt and ligand were completely dissolved. After filtration, a completely clear solution was obtained, and the filtrate was left at room temperature to crystallize. After three days, large dark green crystals formed. The crystals were washed with EtOH and Et₂O. Yield: 0.20 g (89%). The dark green crystals were soluble in dimethylformamide, acetone, and acetonitrile; moderately soluble in methanol, phosphate-buffered saline, ethanol, and water; and insoluble in toluene and diethyl ether.

3.5. X-ray Structural Analysis

A representative pale green thin plate crystals with the dimensions 0.072 × 0.047 × 0.010 mm (complex 1) and 0.144 × 0.072 × 0.011 mm (complex 2) were selected and positioned on a nylon cryoloop. Diffraction data were collected at a temperature of 123 K using CuKa radiation (λ = 1.54184 Å) on a Rigaku Synergy S diffractometer fitted with a HYPIX 6000 hybrid photon counting detector. Data were processed, including an empirical (multi-scan) absorption correction, with the CrysAlisPro software [77]. The structure refinement by standard methods was performed using the SHELX software suite in conjunction with the Olex2 graphical interface [78,79]. The non-hydrogen atoms were positioned with anisotropic displacement ellipsoids, and the hydrogen atoms attached to carbon were placed in calculated positions using a riding model. The positions of the hydrogen atoms attached to oxygen and nitrogen were apparent in the difference Fourier map and were refined with restrained d(O-H) = 0.88 (2)Å or d(N-H) = 0.91 (2)Å (DFIX).

Crystallographic data were deposited in the Cambridge Crystallographic Data Centre (CCDC, 12 Union Road, Cambridge CB2 IEZ, UK; e-mail: [email protected]) with the following CCDC number 2344024 for [Cu(PLAG)(NO₃)(H₂O)]NO₃ (1) and 2344013 for [Cu(PLAG)(H₂O)]SO₄ (2). The crystal data and structure refinement details are listed in

Table 6. Crystal data of the newly obtained complexes.

Empirical formula	Complex 1 [Cu(PLAG)(NO3)(H2O)]NO3 C9H15CuN7O9	Complex 2 [Cu(PLAG)(H2O)]SO4 C9H15CuN5O7S
Formula weight	428.82	400.86
Temperature (K)	123	123
Wavelength (Å)	1.54184	1.54184
Crystal system	monoclinic	Orthorhombic
Space group	P21/c	Pna21
Volume (Å3)	1544.83 (6)	1365.26 (3)
Unit cell dimension (Å/°)	a = 15.9478 (3) b = 9.0095 (2) c = 11.0931 (3) β = 104.251 (2)	a = 6.6590 (1) b = 22.1886 (3) c = 9.2401 (1) β = 108.125 (2)
Z	4	4

.

3.6. Hirshfeld Surface Analysis

The crystallographic structures of the investigated compounds were stabilized by interactions between units and examined by the Hirshfeld surface analysis. These interactions were accessed by the CrystalExplorer program [80] based on the experimental structures. Within this approach, surface analysis is a graph characterized by two distances, one between the two nearest nuclei (d_e) and the other connecting nuclei with the external surface (d_i) [81,82,83]. The distances are colored after their comparison with the van der Waals separations between atoms. The colors red, white, and blue are used if the separation is shorter, equal, or longer than the sum of the van der Waals radii between connecting atoms. In this contribution, the normalized distances are between −0.7428 (red) and 1.1252 (blue). The specific interactions between the atoms are also represented by the fingerprint plots that determine the relative percentages of these contacts to the total number. The fingerprint plots are included in the Supplementary Materials.

3.7. Theoretical Analysis

The structures of the investigated complexes were optimized in the Gaussian 09 Program Package (Gaussian 09, Revision C 01) [84], starting from the crystal structures. These optimizations were performed without any geometrical constraints using B3LYP functional [85] in conjunction with the 6-311++G(d,p) basis set [86] for non-metallic atoms and the def2-TZVP and LanL2DZ basis sets for copper(II) [87,88,89]. The solvent molecules were removed except those directly included in complex ion formation. The stability of the structure was shown through the absence of imaginary frequencies. The interactions between the donor atoms and central metal ions were further examined by the Quantum Theory of Atoms in Molecules (QTAIM), as proposed by Bader [90,91]. This analysis is based on determining electron density and its Laplacian, along with other parameters, within the Bond Critical Point (BCP) [91,92]. AIMAll program package [93] was used to calculate the QTAIM parameters from the .wfx file obtained in the Gaussian 09.

3.8. Spectrofluorometric Titration

The interactions between the obtained complexes and transport proteins (HSA) were investigated by spectrofluorimetric titration. The used instrument was the Cary Eclipse MY2048CH03 instrument (Agilent Technologies, Santa Clara, USA). The scan rate was set to 600 nm min⁻¹ with both slits of 5 nm. The excitation wavelength was set to 280 nm, as this energy is sufficient to excite the tryptophan residues within the protein structure. The concentration of proteins was held constant at 5 × 10⁻⁶ M in 1 M phosphate saline (pH = 7.4). The concentration of the complexes was between 1 and 10 × 10⁻⁶ M. As explained for similar compounds, the emission spectra were recorded two minutes after adding quenchers [20,38]. The fluorescence intensities were corrected for the absorption of excited light and the re-emission of the emitted light. The following equation was used [94,95]:

$$F_c=F_m{e}^{{\displaystyle \frac{A_1+A_2}{2}}}$$

(5)

In this equation, F_c and F_m represent the corrected and measured fluorescence, while A1 and A2 are the absorbances of the complexes at excitation and emission wavelengths [94]. All the reported fluorescence intensities are the corrected ones.

Spectrofluorimetric titration studies were also employed to determine the binding affinity of the obtained compounds towards DNA. In these experiments, ethidium bromide (EB) replacement occurs in the presence of copper(II) complexes. CT-DNA and ethidium bromide concentrations were held constant at 50 μM and 5 μM in phosphate-buffered saline (PBS) (pH = 7.4) with concentrations of NaCl and KCl of 137 and 2.7 mM. The concentration of the complexes varied between 1 and 54 μM. The excitation wavelength was set to 520 nm, as this energy was sufficient to excite the CT-DNA-EB complex. The amount of bound EB lowers after adding quenchers, which is reflected in the decrease in fluorescence emission intensity. The emission spectra were obtained between 540 and 650 nm, and both slits were set to 10 nm. The relative reduction in fluorescence emission intensity followed the previously mentioned double-log Stern–Volmer dependency. The measurements were repeated at three temperatures (27, 32, and 37 °C), and the thermodynamic parameters of the process were calculated.

The binding of complexes with CT-DNA was also examined by the spectrophotometric titration on a Thermo Scientific Evolution 220 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The concentration of complexes was held constant at 50 μM, and the concentration of CT-DNA changed from 10 to 110 μM. Both solutions were prepared in PBS with concentrations of NaCl and KCl of 137 and 2.7 mM. The previously described CT-DNA solution was used for these measurements. Upon the addition of CT-DNA, the resulting solution was left to equilibrate for 5 min, after which the spectra between 320 and 480 nm were recorded.

The iodide quenching experiments were performed by keeping the concentrations of complex 1 and CT-DNA constant at 10 and 20 μM, respectively. The concentration of KI increased from 0 to 9 mM, with the step of 1 mM. The excitation wavelength of complex 1 was determined from the absorption spectra (385 nm), and the emission spectra were recorded between 415 and 600 nm. Both slits were set to 20 nm.

3.9. Molecular Docking

The investigation of binding to HSA and DNA continued at the molecular level using docking calculations in AutoDock 4.2.6 with the assistance of the AMDock program, version 1.5.2 [96]. The optimized structures at the B3LYP/6-311++G(d,p)/def2-TZVP(Cu) level of theory were used as ligands for these simulations. The target molecules included HSA (PDB ID: 6R7S [97], 7WLF [98]) and DNA (PDB ID: 1BNA [99], 1XRW [100], 2ACJ [101]). Ambient temperature (298 K) and physiological pH (7.4) were set before the calculations. The area included in the molecular docking calculations was the entire volume of target molecules. Molecular docking calculations are in contrast to the experimental measurements from solution in the ability to investigate the binding of ligands and counterions separately. This enables the investigation of the influence of counterions by comparing the Cu-PLAG binding in the presence or absence of counterions. To account for the presence of counterions, the binding of a counterion was conducted first. Subsequently, the binding of Cu-PLAG to the best adduct between HSA and counterion from the previous calculations was performed. The best binding positions are those with the highest absoluite values of Gibbs free binding energies. Exceptionally, when two different binding positions achieve similar binding energy values, the better position is the one with the highest number of ligand orientations. The binding of the complex on the surface of HSA is not interesting for two reasons. First, binding at sites that neither directly nor indirectly interfere with Trp fluorescence emission cannot be detected by spectrofluorimetry. Second, and more important, as soon as serum albumins enter the bloodstream, the Cu(II) complex or any other molecule not covalently bound to the surface will be immediately detached. The surface of serum albumins is not suitable for transport within the bloodstream.

4. Conclusions

Two novel copper(II)–pyridoxal-aminoguanidine complexes were synthesized and characterized by X-ray crystallography. The first complex is square-pyramidal with the formula [Cu(PLAG)NO₃H₂O]NO₃ (complex 1), while the second is a square-planar structure with the formula [Cu(PLAG)H₂O]SO₄ (complex 2). In both complexes, PLAG is a tridentate ligand, while the first one of the counterions is coordinated to the central metal ion. This led to different intramolecular interactions within crystal packaging, as determined by the Hirshfeld surface analysis. The main difference is in the interactions that include the central metal ion which are more numerous in the case of complex 2. The structures optimized at the B3LYP/6-311++G(d,p)(H,C,N,O,S)/LanL2DZ(Cu) level of theory showed better resemblance with the experimental structure when the experimental and theoretical bond lengths and angles were compared. The high correlation coefficients and low mean absolute error proved that the chosen level of theory was appropriate for the investigated compounds. The QTAIM parameters showed that bonds between the central metal ion and nitrogen had partial covalent character, although all of the coordination bonds were classified as closed-shell. The presence of nitrate ions in the coordination sphere induced the formation of additional stabilization interactions with the surrounding groups. The Cu(II) complex tends to bind near the tryptophan residues within HSA, while it intercalates inside DNA structures. This preference of the Cu(II) complex for Trp residues and intercalation facilitates the simple experimental detection of its binding to serum albumins and DNA by quenching the fluorescence signal. The calculated trend concerning the values of the Gibbs free energy of binding aligns well with the trend determined by the spectrofluorimetric measurements. In the presence of nitrate ions, the Cu(II) complex binds more effectively. In the DNA molecule, the influence of sulfate and nitrate ions is similar. The main binding mechanism to DNA was determined by the fluorescence quenching with KI to be intercalation, which was later proven by the molecular docking simulations. Overall, the Cu(II) complex has potential as a remedy since it binds to the FA7 binding site in HSA, both of which are suitable for transport in the bloodstream. The intercalation of the Cu(II) complex into the DNA molecule can potentially influence the function of the DNA molecule and the entire cell.