5. Physical pharmacy in the description of molecular mechanism of flavonoids

As already discussed in the sections before, the stabilization of the complex is made by non-covalent interactions, such as hydrogen bonds, hydrophobic interactions, van der Waals and electrostatic forces, salt bridge, π stacking, and cation-π interactions. Consequently, the affinity and the thermodynamic parameters of each system are results of these interactions.

The complete characterization of a complex involves the description of these interactions, the affinity, and the thermodynamics. That is the reason why the use of techniques based on physical approaches in the pharmacy area has been growing.

Several authors characterized the complex formed by flavonoids and biological targets based on fluorescence spectroscopy data, as showed in Section 5.1. Some others used molecular docking to describe the complexes, as shown in Section 5.2.

Nevertheless, the most accurate description of the complex is reached by the association of data from fluorescence and docking, as one can see in Section 5.3.

### 5.1 The use of steady-state fluorescence spectroscopy

Based on the theory of fluorescence quenching described in Section 2, several authors investigated the molecular mechanism of interaction between flavonoids and proteins. The most basic method used to estimate the affinity of the ligand toward the protein is the Stern-Volmer plot, where one can calculate the Stern-Volmer constant (Ksv) and associate the magnitude of this constant with the binding affinity.

Papadopoulou and co-workers [42] used Ksv to evaluate the interaction of two flavonoids, quercetin and rutin, and the bovine serum albumin (BSA). The authors verified that quercetin demonstrated stronger affinity toward BSA compared to rutin. Such difference was explained based on the structure of both flavonoids, once rutin is a glucoside of quercetin. The authors concluded that hydrophobicity was an important factor in the change of affinity, once the incorporated disaccharide rutinose made the flavonoid less hydrophobic. Besides that, another important factor was the steric hindrance caused by the incorporated disaccharide.

With a similar methodology, Cao and co-workers [43] investigated the influence of glycosylation of quercetin and baicalein in the affinity toward BSA. The authors used the Stern-Volmer plot to classify the quenching mechanism as static or dynamic. Then they used the double-logarithm plot to calculate the binding affinity of the ligands studied. They verified that the glycosylation made the interaction weaker and suggested that the decrease in binding constant was an effect of steric hindrance caused by glycoside groups.

The influence of glycosylation in the binding affinity observed by Papadopoulou and co-workers [42] and Cao and co-workers [43] was attributed to steric hindrance. The steric hindrance suggested by the authors could be verified if they had used molecular docking to model the system.

The influence of ions Cu2+, Al3+, Mg2+, and Zn2+ in the interaction of flavonoids and human serum albumin (HSA) was investigated by Bi and co-workers [23]. The authors calculated the binding constant by double-logarithm plot (Eq. (6)), and they concluded that binding constants were 14.2–99.6% of the ones without these metal ions. Therefore, the ion concentration would shorten the storage time of the compounds in blood plasma and enhance the effectiveness of the flavonoids. The influence of ions in the binding constant was also observed by Hu and co-workers [44] during experiments with morin and BSA. In this case, the authors observed that the constant decreased in the presence of Ba2+ and Hg2+, but with the addition of the ions K<sup>+</sup> , Li<sup>+</sup> , Mn2+, Fe2+ and Sn2+, the constant increased.

Another experimental methodology has been used to characterize the interaction between protein and small ligands. As described in Section 2, thermodynamic parameters are calculated from fluorescence quenching data by using the van't Hoff equation.

Li and co-workers [45] used fluorescence experiments to compare the inhibition ability of three flavonoids (quercetin, isoquercetin, and rutin) toward α-glucosidase activity. They concluded that the complexes formed by the protein and the three flavonoids were spontaneous processes, driven mainly by hydrophobic forces, once ΔS and ΔH are both positive numbers. Based on the results obtained, the authors suggested that the flavonoids studied would be useful as an inhibitor of the enzyme and would help in the treatment of hyperglycemia and obesity.

Xi and co-workers [46] studied the interaction between hemoglobin and the flavonoids quercetin and rutin by using double-logarithm and Van't Hoff plots. The authors observed that for both flavonoids, the values of ΔH and ΔS were

negative and concluded that the acting forces are hydrogen bonding and van der Waals. As one can see, the results obtained by [45, 46] reveal that molecular forces that drive quercetin and rutin into α-glucosidase and hemoglobin are different. This difference is noticed because the same molecules interact in cavities with different characteristics.

Another molecular target for flavonoids in biological systems is nucleic acid. The study of the interaction between flavonoids and DNA using fluorescence spectroscopy is made by following fluorescence signal from flavonoids once DNA exhibits a very weak intrinsic emission.

The article published by Jana and co-workers [47] described the affinity of 3 hydroxyflavone toward DNA, by following the flavonoid fluorescence signal. The same methodology was used by Sengupta and co-workers [48] to describe the affinity of fisetin toward DNA. The interaction of polyphenolic compounds with DNA has a protective role, once flavonoids exhibit antioxidant properties.

DNA is also a target for luteolin, as shown in the article published by Chowdhury and co-workers [49]; the interaction between luteolin and nucleic acid is one of the causes of the decrease in affinity between DNA and topoisomerase I.

#### 5.2 The use of molecular docking

The influence of flavonoid structure, the addition of organic group as well as the main interaction between the small molecules and biological targets can be described through the molecular docking technique.

The potential of some flavonoids as inhibitors of α-amylase verified by experimental assays was explained in the work published by Lo Piparo and co-workers [50, 51] using molecular docking. The authors observed that flavonols and flavones have the same scaffolds (Figure 2), and both subclasses possess a carbonyl group in the position C4 of the pyrone ring. The C2-C3 double bond is conjugated to the 4-keto group and is responsible for electron delocalization between the ring C and ring A. As a consequence, flavones and flavonol form a highly conjugated π-system that confers better stability of the protein-ligand complex. The authors also described a specific pattern of OH group, which interacts with the catalytic residues and promotes the inhibition of the enzyme.

The inhibitory effect of some flavonoids was also observed for β-secretase; the work published by Shimmyo and co-workers [52] showed that OH groups in myricetin, quercetin, kaempherol, morin, and apigenin stabilized the binding poses of flavonoids against the β-secretase active center by hydrogen bonds. In many cases, the OH directly interacted with the Asp catalytic residue and enhanced the inhibitory activity of polyphenolic compounds.

Si and co-workers [53] showed that some flavonoids inhibiting CYP2C9 activity may increase the risk of toxicity from coadministered drugs that are CYP2C9 substrates. The authors used molecular docking to give details about the interaction of CYP2C9–6-hydroxyflavone complex, showing that the 6-hydroxyflavone is bound by a π-π stacking interaction with the phenyl group of Phe 100 and by two hydrogen bonds with Leu102 and Phe100.

The examples above used molecular docking to describe the interaction between flavonoids and molecular targets in order to explain the results obtained from experimental assays. However, one can also use molecular docking in order to select the best candidates and then submit the chosen molecules to the experimental assays.

In the work published by Salam and co-workers [54], the authors built a library of 200 natural compounds including some flavonoids. With virtual screening, they selected the 29 best candidates to bind to the peroxisome proliferator-activated
