IN SILICO STUDIES OF SOPHORAFLAVANONE G: QUANTUM CHARACTERIZATION AND ADMET SILICO STUDIES OF SOPHORAFLAVANONE G: QUANTUM CHARACTERIZATION AND ADMET.”

Currently, the search for new drugs with greater therapeutic potential and less side effects has been fostered by the advancement of the use of molecular modeling drugs, which in addition to supporting the full characterization of the molecule, allow simple algorithms to predict pharmacokinetic. In this context the present work aimed to perform the electronic / structural characterization, to evaluate the pharmacokinetic properties and to perform a virtual screening of the possible biological targets of Sophoraflavonone G, a promising flavonone, which presents several pharmacological properties. Sophoraflavanone G was geometrically optimized by semiempirical quantum calculations, plot the MESP, identifying the nucleophilic sites. Using the boundary orbitals, it was possible to identify a greater tendency for electron donation in relation to Naringeni, with lower ionization potential, higher hardness and less softness. With respect to pharmacokinetics Sophoraflavonone G confirmed the safety of the compound for oral administration with good skin permeability, which allows applications in topical formulations. Presents indications for gastro intestinal absorption, as for possible interactions with biological targets, interaction with the estrogen receptor alpha, sodium / glucose co-transporter 2, beta-secretase 1, cyclooxygenase-1.The data obtained from an early stage for a comparative analysis between its analogues and fundamental for future studies of relationships between the three-dimensional structure of Sophoraflavanone G and its biological activities.


Introduction
Flavonoids are a class of secondary metabolites found in plants that perform a variety of functions. They are known as plant pigments mainly present in flower petals to attract pollinators and for their antioxidant activities, providing consumers with some hope for medicinal uses, potentially cancer treatment. [1].
The dried roots of the Sophora flavescens plant belonging to the Leguminosae family, is used in traditional Chinese medicine and has antibacterial [2], antiviral, antiprotozoal, anti-inflammatory [3], antipyretic [3], healing, anti-hemorrhagic, antimalarial [4], phospholipase-Cγ1 enzyme inhibiting activity [5], antiarrhythmic effects [6] used in the treatment of skin ulcers and also as an insecticide. In addition to flavonoids with regular prenyl side chains, S. flavescens also produces several flavanones with lavasulil chains, irregular monoterpenoid groups such as kurarinone and sophoraflavanone G [7], [8] noting that studies have shown the importance of the chain. side of Lavandulila for antitumor activity [9] Flavonoids isolated from this plant have phospholipase-Cγ1 inhibition activity [5].
Sophoraflavanone G, one of the major flavanones isolated from the dried roots of S. flavescens, is known to possess antimalarial, antimicrobial, antiviral and antioxidant activities and inhibits the production of nitric oxide and prostaglandin E2 in lipopolysaccharide treated RAW cells. [10]. These biological activities are indicative of the pharmacological potential of Sophoraflavanone G.
Thus, with the development of chemistry, especially organic synthesis, many substances that have some kind of biological activity have been synthesized. As a result, drugs of synthetic origin became about 80% of the drugs on the market. Thus, many of these substances come from prototype structures found in natural products. Often, organic synthesis contributes not only to a greater production of the natural substance that has biological activity, but also to the structural modification of this substance to obtain pharmacologically superior products [11].
With the increase in computer processing potential and the development of more accurate mathematical algorithms, molecular modeling has been increasing the development of new drugs, reducing research costs through electronic and structural characterization, pharmacokinetic, toxicological simulations and receptor drug interactions , which make it possible with great precision to analyze the potential drug, predicting its potential, indicating the continuity or not of tests in vivo and in vitro [12][13] [14] [15].
Given the need for the development of new drugs, the aim of this study was to perform the electronic / structural characterization, evaluate the pharmacokinetic properties and perform a virtual screening of the possible biological targets of sophoraflavonone G. This study is a fundamental step for DRUG studies. design, molecular docking and molecular dynamics.

Computational Details
All the computations simulations were performed on personal computer with intel® Core TM i7 7700HQ processor, 16  Microsoft Windows 10® as operating system. All softwares codes used are free license for academic use.

Structural Optimization
Initially, after the literature review, the two-dimensional molecular structure of Sophoraflavanone G from the ChemSpider® virtual repository (http://www.chemspider.com) ( Figure 1A) was used to obtain the following data: nomenclatures, physicochemical properties and mechanisms of action. After obtaining the data in its two-dimensional form, the molecule was geometrically optimized by semi-empirical quantum calculations, implemented in the code ArgusLab® [16] configured with Hamiltonian Parametric Method 3 (PM3) parameters [17]. As Sophoraflavanone G is an analogue derivative of anagerin ( Figure 1B), for comparative analysis, optimization data were used under the same conditions.

Fronteir Orbitals, MESP and Descriptores Reactivity
To start the reactivity study we need to obtain the spatial conformation (3D) of the molecule, because from it we can calculate the energies of the orbitals. In this context, the LCAO-MO Model used in the theory of molecular orbitals, which uses the linear combination of atomic orbitals, to generate the molecular orbitals, keeping the same number of orbitals and fundamental for rendering and identification of the orbitals. In terms of reactivity analysis we can highlight the homo and the lunar orbitals, which are called boundary orbitals, since from the energy difference among them we can identify the electron receptor or donor bias of the molecule [18][15] [19].
Thus it was possible to obtain the lowest potential energy conformation of the molecule, characterizing HOMO (Higher Energy Busy Molecular Orbital) (HOMO) LUMO (Low Energy Unoccupied Molecular Orbital) (LUMO) [20] [21] [22] and the reactivity descriptors, GAP (Amount of energy required for the electron to transition), electron affinity (A), ionization potential (I), electronegativity (χ), hardness (η), Chemical softness (S), chemical potential (µ) and electrophilicity index (Ω) (   The properties of Absorption, Distribution, Metabolization, Excretion and Toxicity (ADMET) including solubility, blood brain barrier (BBB), plasma protein binding, CYP2D6 binding, gastrointestinal absorption and hepatotoxicity were predicted using the Swissadme code (http: // www. swissadme.ch/), an online software written in JavaScript, HTML, and PHP5 with Python 2.7 encoded computing backend that allows you to accurately analyze 2D structures of drug candidates against ADMET descriptors [33][34] [35]. lipophilicity was predicted according to Wildman et al [36],violations of oral bioavailability were assessed according to the Lipinski rule [37].At this stage we also examined the possible inhibitory character of Sophoraflavonone of Cytochrome Family (P450 CYP) isoforms such as CYP1A2 and CYP2D6, in addition to other pharmacokinetic predictions (gastrointestinal absorption, P-glycoprotein and blood brain barrier), as Ghose and Veber Rules and bioavailability [38] and Brain Or IntestineL EstimateD permeation method (BOILED-Egg) to predict gastrointestinal absorption predict gastrointestinal [39].

2D Structure Modeling
The 2D Structural Models were designed in MarvinSketch code (ChemAxon Software) [40]and coded for SMILES in the SMILES online translator and generator structure file found in OpenBabel software [41].

Virtual Screening for Target Classes
To identify possible target proteins, Swiss Target Prediction [42] [43] was used, an online script that uses the reverse trace scope to identify through structural similarities between molecules to similarly predict possible receptors (biological targets). Homo sapiens receiver search identifier [44], the Sophoraflavonone G two-dimensional structure was used, which was converted to 3D format, using the Tanimoto coefficient as an adjustment factor and similarity, vectorized by the Manhattan distance, defined by by the equation using the vectors (X and Y) and equation 2 to calculate the final value 3D similarity between molecules of i and j, where d ij is the shortest Manhattan distance between the 20 × 20 calculated distances over all possible conformations of each molecule [42].

Structural and Conformational Characterization
Using Sophoraflavanone G's two-dimensional structure obtained from the Chewspider.com virtual repository (http://www.chemspider.com/Chemical-Structure.65766.html?rid=ee036cea-6045-4fad-82ca-0b7579f6a86f) (ID 65766) its name could be identified according to the International Union of Pure and Applied Chemistry IUPAC (2S) -2-(2,4-Dihydroxyphenyl) -5,7-dihydroxy-8 -[(2R) -2-isopropenyl-5-methyl-4-hexen-1-yl] -2,3-dihydro-4H-chromen-4-one and its molecular formula C25H28O6. In addition, physical and chemical properties could be obtained [ Table 1], linked to the structural composition of the molecule, from which we can highlight its density (1.3 ± 0.1 g cm-3), its surface tension (55.9 ± 3.0 dyne cm-1). , molar volume (335.3 ± 3.0 cm3) and their ability to form hydrogen bonds by determining atoms with the potential to receive or donate electrons in hydrogen bonds among other properties.  The two-dimensional structure of Sophoraflavanone G ( Figure 1B), available from the ChemSpider® repository, was in its ground state, presenting only the molecular formula (C25H28O6) and the connectivity of atoms, with an easy to see initial conformation but with different potential energy. of the molecule in its native form. because the two-dimensional coordinates of a molecule available in an online repository are not necessarily in the most stable conformation [45]. Thus, to obtain more accurate calculations about the molecule and its more stable final conformation, we need to perform a geometric optimization using the energy minimization process [46] . This geometric optimization was performed using the semi-empirical quantum method formalism QM_PM3 (Parametric Method 3), performing uninterrupted cycles of 200 interactions, repeated until the structure reaches a point of lowest possible minimum energy, thus each of the atoms that the structure occupied its place of lowest energy in the system and the three-dimensional structure reached the most stable calculable point (Figure 2), presenting a spatial distribution that allows the lowest potential energy possible [47] [48], making the potential energy of the molecule assumed a value of (-118893.6623kcal / mol), no longer varying, reaching a stationary point of the energy surface [4], the formation heat (-206.3487kcal / mol) and dipole moment (μ) in length. (4.95590298 debye) characterizing the molecule as polar. Some other properties of the structure are directly linked to the dipolar moment (μ), such as melting and boiling points and their solubility in water. By obtaining the molecule in a theoretically more stable structure, it was possible to calculate the formal and partial charges of all atoms as well as their valence [49] [50]. All atoms showed zero formal charge and significant variations in their partial charges were found, such as Hydrogen from 0.027 to 0.292, Carbon from -0.099 to 0.174 and Oxygen from -0.507 to -0.292. These data (table 2), especially valence, correspond to the literature, which serves to validate the results obtained. Despite the neutrality through optimization, it is possible to observe in the results obtained the largest and smallest (residual) partial loads. It was noticed that atoms 59, 58,57,50 (H) and 30 (O) presented larger and smaller partial charge respectively, charges from electrons being closer or farther from one of the bond atoms, taking with them charge.  addition, the links (16, 19 and 44) between carbons are noted for their rotability (Table 3).
Complementing the conformational analysis, the largest and smallest angles between joints were identified, angles 18 (CCC) and 44 (HCH) with 124.7844 ° and 105.4585 ° respectively; and the largest and smallest torsion angles, the 147 (CCOH) and 67 (ACC) systems with 179,999 ° and -179,978 ° respectively.

Electronic Characterization and Reactivity Descriptors
After geometric optimization, the molecule a theoretically more stable structure, it was possible to calculate the atomic charges using the most traditional and known method among chemists: the Mulliken Population Analysis [51] This method divides the charge density between two atoms evenly without taking into account electronegativity; Atomic charges are useful in the correlational study between the structure of the molecule and its biological activity [52] [53].The calculations under the PM3 formalism how us that (TABLE 4) there was a large variation between the atomic charges of atoms of the same element, the oxygen with the highest load was O18 with -0.1916 while the lowest oxygen, O14 with -0.3783, varied 0.1867 charge; the carbons of the same form being the one with the highest load C21 with 0.4041 and the one with the lowest load C15 with -0.4321, varying 0.0280; for hydrogen, the atomic charges ranged from 0.1006 in hydrogen H44 to 0.2858 in hydrogen H59, ranging from 0.1543. To identify the areas of higher nucleophilicity and electrophilicity of the molecule, the electrostatic potential map (  Global reactivity descriptors act as a bridge between structural stability and global chemical reactivity [57]. This is also information about fundamental characteristics of chemical reactivity, such as ionization potential, electron affinity, electronegativity, chemical potential, global hardness and global softness (Table 3) [24]. The LCAO-MO Model used in the theory of molecular orbitals uses the linear combination of atomic orbitals to generate the molecular orbitals, keeping the same number of orbitals. frontier orbitals, because from the energy difference between them we can identify the electron receptor or donor tendency of the molecule. A good starting point for determining these parameters is the energetic values of the frontier molecular orbitals (HOMO and LUMO) that are used in reactivity [18],  [50].In addition, the GAP (amount of energy required for the electron to make a transition) is calculated on the energy difference between HOMO and LUMO (1).  [50].In addition, the GAP (amount of energy required for the electron to make a transition) is calculated on the energy difference between HOMO and LUMO (1). Comparing Sophoraflavanone with its reference compound Naringenin, it was possible to observe some differences that are fundamental in characterization. For example, Sophoraflavanone G has been shown to be a molecule that tends to donate electrons by presenting HOMO (-9,02585 eV) higher than Naringenin (-9,2525 eV) (figure 8). Table 5 shows the reactivity parameters of the compound Sophoraflavanone G and the compound Naringenin, which were obtained by calculations using the molecular orbital energies HOMO and LUMO. Regarding the energy required to remove an electron (ionization potential), Sophoraflavanone had a lower ionization potential (9,02585 eV) than Naringenin (9,25325 eV), indicating to be more stable. Having the energy of the frontier orbitals as a reference, the hardness (η) and softness (S) values in the study of Pearson's theory (HSAB, hard and soft acid and bases) [26][62] These are fundamental concepts for understanding the acid / base behavior of molecules, where we can simplify molecules with a donor / recipient center of small electron pairs (with a low tendency to be polarizable), are indicative of favoring electrostatic interactions being considered. hard acids (receptors) or bases (donors). Molecules that favor the interaction of the frontier orbitals, that is, with low energy value, are indicative of high polarizability, since in general they have high reactivity sites (donor / receptor sites of electron pairs) and are considered as bases (donors) or soft acids (receptors), thus having high reactivity [63]. That is, Interactions between for hard acids / bases, or soft acids / bases tend to form more stable complexes when compared to the interaction between soft acids and hard bases or hard acids and bases soft [64] [21]. Regarding hardness and softness, Sophoraflavanone tends to have difficulty being deformed, this is proven by observing that it had higher hardness and less softness compared to Naringenin.  The chemical structure of Sophoraflavanone G was subjected to ADMET in silico screening using SwissADME online software to predict overall absorption, distribution, metabolism, excretion and toxicity hazards. In the silico evaluation, we analyzed different descriptors that were calculated, partition coefficient (octanol / water), molecular weight (424.49 g / mol), number of H+ donor groups and acceptor groups that compounded moderate solubility. in water. Absorption or permeation of a molecule is more likely when the molecular weight is below the log P value is less than 5 and the value. Sophoraflavonone G has a maximum of 6 H+ donors and 4 H+ acceptor atoms, as shown in Figure 9. According to its pharmacokinetic properties, Sophoraflavanone G showed a high level of gastrointestinal adsorption which contributes to good oral bioavailability.
Sophoraflavanone G according to pharmacokinetic parameters evaluated in silico showed no inhibition of cytochrome P450 isomers and glycoprotein P with CYP2C9 isomer exemption. Prediction of similarity with other drugs was also performed according to Ghose and Veber rules and the molecule has characteristics of bioavailability [38].
Thus, the screening process with the Lipinski Rule showed that there were no compound violations, as the screening process with Ghose rules shows that the substance sophoraflavonone G can be accepted for not showing violations of drug similarity rules (figure 10).

Virtual Screening for Target Classes
An important data for the development of a new drug is the evaluation of interactions with possible biological targets, and may be a descriptor for the therapeutic improvement [72]. Regarding possible biological target classes, Sophoraflavanone G had a higher probability of interactions with enzymes (20%), hydrolases (13.3%) and nuclear receptor (13.3%) ( Figure 11). Regarding specific targets, it presented a higher probability (> 0.5) of interaction with Estrogen receptor alpha, Sodium / glucose cotransporter 2, Beta-secretase 1, Cyclooxygenase-1, which were ranked the best binging probability on the SwissTarget Prediction report (Table 6).

Conclusions
The molecular structure of Sophoraflavanone G was geometrically optimized by semi-empirical quantum calculations, obtaining the thermodynamically more stable structure. Through the electrostatic potential map it was possible to identify the areas where the nucleophilic bonds will occur that comprise the whole area of oxygen. Due to geometrization one can also obtain the atomic charges of Mulliken for the atoms of O, C, H highlighting the variation of charges between atoms of the same species.
Regarding the frontier orbitals, it was possible to identify a greater tendency to donate electrons compared to Naringeni, having a lower ionization potential, with greater hardness and less softness. Sophoraflavonone G drug confirmed the safety of the compound for oral administration. It also has good skin permeability which enables applications in topical formulations. However, it lacks the blood-brain barrier permeability. It is a highly selective structure that protects the Central Nervous System (CNS) from potentially neurotoxic substances present in the blood and is essential for normal metabolic function of the brain, yet has indications of intestinal gastro absorption.
Regarding possible interactions with biological targets, it was more likely to interact with Estrogen receptor alpha, Sodium / glucose cotransporter 2, Beta-secretase 1, Cyclooxygenase-1. The data obtained constitute an initial step for a comparative analysis between its analogues and fundamental for future studies of relationships between the three-dimensional structure of Sophoraflavanone G and its biological activities.