In silico screening of brazilian semiarid compounds to identify potential drugs with glucocorticoid receptor interaction

The glucocorticoid receptor regulates the anti-inflammatory response, and prevents transcription of anti-inflammatory substances such as nuclear factor kB and lipocortin-1, IL-2, IL-6, TNF and prostaglandins. Thus, a search for new molecules with potential interaction with the glucocorticoid receptor is an interesting strategy for the treatment of inflammatory diseases. Virtual screening has proven to be a viable tool for discovering new products at lower cost and practicality. Thus, the aim of this study is to identify and evaluate brazilian semiarid compounds with anti-inflammatory potential with glucocorticoid receptor action through molecular coupling. Protein selection was performed by searching the 3D structure database, Protein Data Bank. A total of 382 semi-arid molecules available in the ZINC database of State University of Feira de Santana (UEFS) were used. Molecular docking was performed using Autodock Vina and as interaction clouds analyzed by the Discovery Studio Visualizer program. Mometasone furoate shows a binding energy of -12.7 Kcal.mol-1. A ZINC 69481862 molecule fits Lipinski and Veber rules, however, the best interaction was the ZINC 69482012 molecule, evidenced by the binding energy -11.2 Kcal.mol-1. Analyses of intermolecular interactions have shown that Van der Waals interactions and electrostatic bonds are crucial for the binding of the molecule at the receptor's active site. It is necessary to test in vitro to verify the viability and toxicity of the potential drug. Discovery Studio Visualizer. El de Kcal.mol-1. La molécula ZINC la molécula Los análisis de las interacciones intermoleculares las interacciones de Waals y los electrostáticos cruciales la de la molécula en el sitio activo del receptor. Son necesarias pruebas in vitro para verificar la viabilidad y la toxicidad del fármaco potencial.


Introduction
Inflammation is a local or systemic physiological process, responsible for eliminating stimuli that cause injury, repair tissue and promote immune memory (Fullerton & Gilroy, 2016). Inflammatory criticisms are complex and mediated by the immune system, through macrophages and dendritic cells, with a production of soluble mediators, such as system complement, cytokines and chemokines, which cause the influx of neutrophils and monocytes in the information site. However, acute and systemic inflammation can result in organ failure and tissue death, and if it persists for a long period of time, it can cause chronic inflammatory diseases, including autoimmunity and câncer (Fullerton & Gilroy, 2016;Kumar, 2019).
Among the receptors involved in the inflammatory process, are the nuclear receptors of the glucocorticoid type. Glucocoirticoid receptors are a product of the NR3C1 gene located on chromosome 5q31-32, and are present throughout the body with different sensitivities.
These factors act of transcription in response to a hormone, modulating the expression of target genes (Faria & Longui, 2006;Kadmiel & Cidlowski, 2013).
The formation of the glucocorticoid receptor complex regulates the anti-inflammatory response by two engines: genetic transrepression, which is the recruitment of the histone deacetylase enzyme induces DNA condensation, thus preventing the transcription of proinflammatory substances, such as the factor nuclear kB and lipocortin-1; and genetic transaction, in which there is an increase in acetylation of genes encoding anti-inflammatory proteins, such as IL-2, IL-6, TNF and prostaglandins (Torres, Insuela & Carvalho, 2012). Research, Society and Development, v. 9, n. 9, e734997865, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7865 No drug development process, a virtual study or screening in silicone, is demonstrated as a viable strategy due to its lower cost and practicality (Shityakov & Foerster, 2014;Martins et al., 2014). In view of high cost of drug development, computational approaches are used to screen a large number of compounds and then select a restricted number of drug uses (Kazmi et al., 2019). Virtual screening uses molecular documentation, which identifies compounds with chemical characteristics and the different ligand distances used to determine the molecular target. Thus, it allows the selection of molecules that present a set of favorable intermolecular interactions (Rodrigues et al., 2012).
The virtual screening process depends on a database with a wide molecular diversity, which provides the compounds for the tests to be carried out. Among the virtual libraries, the ZINC database has a rich collection with more than 20 million compounds, available for free (Irwin & Shoichet, 2005). The Feira de Santana State University provides a collection of hundreds of cataloged molecules from the Brazilian semiarid region.
According to Ministry of the Environment (2020), the brazilian semiarid region occupies approximately 11.5% of the national territory; and is an estimated eight thousand plant species, of which 318 species are endemic to the caatinga. In these plant species, there is great potential for pharmacological use (Barreiro & Bolzani 2009). Studies by Costa et al., 2008;Costa et al., 2010 demonstrated that the Brazilian semiarid region has plants with immunomodulatory and antibacterial activities.
The glucocorticoid receptor is a potential pharmacological target in the treatment of inflammatory and autoimmune disorders (Kadmiel & Cidlowski, 2013). Thus, the search for new molecules that interact with the glucocorticoid receptor is important to aggregate the therapeutic collection of inflammatory diseases. There is a lack of related studies, so the objective is to identify and evaluate compounds from the Brazilian semiarid region with antiinflammatory potential with action on the glucocorticoid receptor, through molecular coupling.

Protein selection
Protein selection was carried out by searching the 3D structure database, Protein Data Bank -PDB (Berman et al., 2000). The selection filters were the experimental X-ray diffraction method, with a resolution and R value greater than 2.0 Å and 0.2, respectively. The Research, Society and Development, v. 9, n. 9, e734997865, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7865 6 presence of a ligand, mometasone bore and a nuclear activator in the receptor structure is considered.

Preparation of the protein
The AutodockTools 1.5.6.rc3 (Sanner 1999) program was used to prepare a threedimensional structure of the molecular target, or the glucocorticoid receptor. Through the preparation of the molecular target it is possible to carry out fitting studies. In this process, the binders and solvents present in the structure are removed. They are hydrogen gases and fillers to adapt the chemical structure. The spatial determination of the target site of the receptor was made through the position of the ligand and interaction with the receptor, with the determination of the active site being made. For this purpose, the AutodockTools 1.5.6.rc3 (Sanner 1999) program was also used to determine how to coordinate the search space of the active location of the receiver. Through the grid box, the coordinates of the research space were applied in the space of 1 Å.

Selection of semiarid molecules
A total of 382 semiarid molecules are available in the ZINC (Irwin & Shoichet, 2005) database provided by the State University of Feira de Santana. As molecules present in the .mol2 format, the databases were converted to the .pdbqt format using the Autodock Tools 1.5.6.rc3 (Sanner, 1999) program.

Docking and identification of interactions with the glucocorticoid receptor
The coupling or molecular coupling was performed using Autodock Vina (Trott & Olson, 2010) which is responsible for calculating the binding energy of the ligand interaction with the receptor. The results were observed from the computer's command prompt. Together with the identified binding energies, such as clouds of interactions and the present amino acid residues analyzed by the Discovery Studio Visualizer (BIOVIA, 2016) program. The parameters of Lipinski (Lipinski et al., 2001) and Veber (Veber et. al., 2002) for oral bioavailability were considered in the analysis of the molecules.

Results and Discussion
The protected protein in the PDB database -Protein Data Bank (Berman et al., 2000) for the -glucocorticoid receptor -code PDB 4P6W. The anchor box for selecting a spatial position of the connection site has 27,336, -51,637 and -36,965 X, Y and Z coordinates, respectively, and 10 x 14 x 14 Â dimensions.
The mometasone furoate ligand (Figure 1) was used as a prototype to relate to the results. The binding energy resulting from the interaction between a crystallized structure of the NR3C1 glucocorticoid receptor and mometanosa furoate was equal to -12.7 Kcal.mol -1 .
According to Derendorf and Meltzer (Derendorf & Meltzer, 2008) mometasone furoate is a potent corticosteroid that inhibits pro-inflammatory mediators by a downregulation mechanism, acting on the glucocorticoid receptor. In vitro assays, mometasone furoate stimulated the gene transactivation of the glucocorticoid receptor more potently (Smith & Kreutner, 1998). This molecule inhibited the production of TNF-α, interferon-y, leukotrienes and histamine (Kim et al., 2019).  Development, v. 9, n. 9, e734997865, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7865 8 was used as a ligand to attach to the glucocorticoid receptor. There is a high affinity between the ligand and the 17α position of the receptor, facilitated by hydrophobic interactions (Wang, Aslanian & Madison, 2008;Chen et al., 2005).  Intermolecular bonds between protein and ligand occur mainly by Van der Walls forces, electrostatics and hydrogen bonds . Notably the greater predominance of interactions is of the Van der Waals type, fundamental for the interaction of the ligand with the active site of the protein, and this interaction occurs with the amino acid residues: LEU566, MET604, GLY567, TRP600, MET601, LEU753, PHE623, ALA605, LEU608, PHE749, LEU732, ILE629, MET646, MET639, CYS643, GLN642, TYR735 and ILE747. To a lesser extent, there are also electrostatic interactions between the active site and the ligand, verified by the ARG611, LEU563, GLN570, ASN564, MET560, CYS736 and THR739 residues.
In Figure 3A, it can be seen that most of the interactions between the active site of the glycocoirticoid receptor and mometasone furoate are of the Van der Waals type, through the visualization of the hydrophobicity cloud, represented by the brownish color. Figure 3B Research, Society and Development, v. 9, n. 9, e734997865, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7865 9 shows the hydrogen acceptor and donor regions represented by the greenish and pink colors, respectively.  In the analysis of the hydrophobicity clouds and hydrogen bonds ( Figure 5) of the ZINC 69482012 molecule, it is noted that it is composed of hydrophobic groups and hydrogen acceptors, represented by the brownish and greenish colors, respectively. In accordance with the structure and chemical groups of mometasone furoate. Among the 382 molecules tested in the Bahian semiarid region with the glucocorticoid receptor NR3C1, the most favorable interaction showed a binding energy equivalent to -11.2 Kcal.mol -1 , while the least favorable interaction was equal to +21.1 Kcal.mol -1 .  12S,15S,16R,19S,21R)3,7,7,11,16,20,20heptamethylpentacyclo[13.8.0.03,12.06,11.01,21]tric os-1(23)-ene-8,19-diol (Figure 6), obtained the best binding energy with the active site of the glucocorticoid receptor (-11.2 Kcal.mol-1). The ZINC 69482012 molecule does not fit the rules of Lipinski (Lipinski et al., 2001) and Veber (Veber et. al., 2002), therefore, its pharmaceutical presentation should be non-oral.
Source: ZINC Database Irwin & Shoichet, (2005) The parameters established by Lipinski (Lipinski et al., 2001) and Veber (Veber et. al., 2002) are observed for the development of a possible oral drug. The oral bioavailability of the molecules is predicted by descriptors, analyze that parameters: Log P, number of electron donor groups, number of electron acceptor groups, molecular weight, number of rotatable bonds and polar surface area to determine the solubility and permeability of drugs before the membrane.
Thus, although the ZINC 69481862 molecule does not have the best binding energy among the analyzed molecules, it has the best potential as an oral drug (Table 1). Source: ZINC Database (Irwin & Shoichet, 2005), (Lipinski et al., 2001;Veber et. al., 2002) * Values made available by ZINC Database (Irwin & Shoichet, 2005) ** Reference values according to (Lipinski et al., 2001;Veber et. al., 2002) The modeling studies show that the lower the resulting binding energy between the ligand and the target site, the better the interaction affinity. This anchorage uses the connection energy generated in the process, to validate and capture intermolecular interactions (Sivakumar, Sajeevan & Bright Singh, 2016;Ozawa et al., 2019).
Molecular coupling and the study of the chemical structure of molecules are successful strategies to obtain a configuration of new drugs (Ferreira et al., 2005). Zianna et. al (2019) demonstrated an agreement between in vitro and silica studies, when evaluating the bacterial activity of the complex formed by the compound Paladium (II) and the Schiff base. These models are used both in industry and academia, and from them it is possible to predict the conformation of the ligand at the target site and its viability (Drwal & Griffith et al., 2013).
The molecule discussed in this study, ZINC 69482012, has the potential to interact with the glucocorticoid receptor NR3C1, and may have anti-inflammatory activity, however, evaluation of toxicological and pharmacokinetic parameters is necessary to determine the safety and viability of the drug as an anti-inflammatory medicine.
In silico studies have been shown to be an efficient strategy for the discovery of new drugs, however it is important to note that the results may or may not be confirmed in vivo (Zheng et al., 2013).

Conclusion and Suggestions
The ligand ZINC 69481862 fit the rules of Lipinski and Veber, however, the best interaction was of the molecule ZINC 69482012, evidenced by the binding energy -11.2 Kcal.mol-1. Although it does not fit the parameters of oral bioavailability, the potential drug can be proposed by other routes of non-oral administration.
The analysis of the interactions involved in the ligand-receptor complex demonstrated that hydrophobic bonds and electrostatic interactions are essential for the coupling of the molecule to the target site. Such characteristics were also verified in the molecule with the best interaction, ZINC 69482012. The clouds of interaction corroborated the forces at work in the system, Van der Waals connections.
From the results obtained, in vitro tests are necessary to ascertain the feasibility and further investigation in relation to the ZINC 69482012 molecule, to determine the toxicity and safety of the potential drug.