C-demethylation and 1, 2-amino shift in (E)-2-(1-(3-aminophenyl) ethylidene)hydrazinecarboxamide to (E)-2-(2-aminobenzylidene)hydrazinecarboxamide and their applications

A Novel (E)-2-(1-(3-aminophenyl)ethylidene)hydrazinecarboxamide 1 was synthesized by traditional method and converted to (E)-2-(2-aminobenzylidene)hydrazinecarboxamide 2 by single step in DMSO at room temperature. Synthesized compound 1 was analysed by spectroscopy (NMR and LC–MS) techniques and molecule 2 was characterized using single crystal X-ray diffraction and spectroscopy (NMR and GC–MS) techniques. These analytical technique results revealed that, C-demethylation and 1, 2 amino shift in phenyl ring of compound 1 gives molecule 2. DNA binding studies of compounds 1 and 2 was carried out by electronic absorption spectroscopy. This result revealed that, compounds 1 and 2 showed hyperchromism with bathochromic shift. Anticancer activity of compounds 1 and 2 is carried out by molecular docking with five receptors.Computer aided virtual screening demonstrated that the synthesized molecules possess ideal drug likeliness, pharmacokinetics features, toxicity profile for structure based drug discovery. The molecular docking studies revealed that the synthesized molecules are significant binding with the five selected cancer receptors with minimum binding energy (kcal/mol), number of hydrogen bonds, weak interaction, docking score and cluster RMS. The docking studies also suggested that the molecules showed interactions with DNA and the theoretical values of the binding are comparable with that of the experimental values. Hirshfeld surface analysis was used to analyze and quantify the intermolecular interactions in the crystal structure of compound 2.

widely used drugs in treatment of various cancers. Their remarkable success has been marred somehow because of several side effects such as nephrotoxicity, adverse effects on the peripheral nervous system as well as liver damage, severe emesis and drug resistant tumors etc. 28 . Limitations of platinum based drugs arouse to design novel organic molecules as anticancer drugs. DNA binding mechanism of molecules has been studied in order to develop anticancer drugs. Organic molecules interact with DNA helix structure through an intercalation/ non-intercalation process intervene with the cell cycle such as translation, transcription and replication causing cell death 29 . In the light of above, detailed analyses of structure of novel synthesized semicarbazones are completed by different analytical techniques. The interaction mechanism of molecules with nucleic acid by electronic absorption spectroscopy were carried out. Anticancer activity of compounds 1 and 2 is carried out by molecular docking with five receptors and ideal drug likeliness, pharmacokinetics features, toxicity profile for structure based drug discovery of compounds 1 and 2 were evaluated using computer aided virtual screening. An estimation of close intermolecular interactions in the compound 2 is also reported using Hirshfeld surface analysis.

Results and discussion
The compound was kept for crystallization in polar solvent DMSO; the polar solvent attacked the carbonyl group of compound 1 gives carbanion as intermediate. The carbanion lost methyl group along with bonding electrons as methyl anion and also solvent regaining its original state given carbocation. The positive ion moves towards the phenyl ring and also shifting of proton from phenyl ring to azomethine carbon takes place. So that it produces positive charge on phenyl ring carbon, it is ortho position to amine group. However, 1, 2 shift of amino group in a phenyl ring and addition of hydride ion from methyl anion gave stable compound 2. The intermediates of this conversion carbanion, carbocations and nitronium ion are stabilized by resonance, since molecule 1 has double bonds and heteroatoms in conjugation throughout the molecule. Conversion mechanism is given in Scheme 1.
Single crystal X-ray diffraction studies. The structure of molecule 2 with labeling of atoms, hydrogen bonds and packing diagram is given in Figs. 1, 2 and 3. Selected bond angles, bond lengths and hydrogen bonding data are listed in Tables 1 and 2. The C2-N3 bond length is 1.281(3) Å indicates that it is a characteristic of azomethine group 30 . Phenyl group in a molecule is delocalized bonding structure, confirms C-C bond lengths of ring in the ranging from 1.371(5) to 1.401(4) Å 31 . The bond length of C1-O1 is 1.250(3) Å; it is a characteristic of amide carbonyl group of semicarbazone 32     DNA binding studies. The electronic spectral data of compounds 1 and 2 were recorded in the absence and presence of rising amounts of CT-DNA (25 µL) (Fig. 4). The molecules were showed hyperchromism 34 together with bathochromic shift. Hyperchromism with bathochromic shift could be the results of formation of covalent bond between the molecules and DNA; hence the combined forms are having more π-bonds and heteroatoms rather than molecules.    Table 4. From the table, it is clear that, the two molecules showed almost similar pharmacokinetics features. Blood brain barrier permeability, Buffer solubility, human epithelial colorectal adenocarcinoma cell permeability, CYP inhibition, human intestinal absorption, plasma protein binding, pure water solubility and skin permeability models showed various values, signifies that most the pharmacokinetic features are in the acceptable range for ideal drug. The toxicity features of the selected lead molecules are shown in Table 5. The study indicated that, both molecules showed positive for carcinogenicity in mouse models and negative in rat models. Similarly, the compounds were predicted to be mutagens by Ames test. The predicted toxicity in algae and fish such as daphina, medaka and minnow were displayed varying results however, these results can be considered as features for ideal lead molecules. The in vitro Ames test results in TA100 strain were predicted to be negative for both the compounds. The in vitro hERG inhibition of compound 1 and compound 2 were predicted to be low risk and medium risk respectively. Thus, virtual screening and chemoinformatics features of the two compounds showed that these compounds probably possess ideal drug likeliness, pharmacokinetics and toxicity profile required to be considered the lead for structure based drug discovery.
Study of binding potential of compounds with various cancer targets. Each of the three dimensional structures of the selected cancer protein targets (Table 6) were docked with the compounds 1 and 2. The binding potential of these compounds towards the prioritized targets are shown in Table 7. From the dock-  www.nature.com/scientificreports/ ing studies, it is clear that each of the synthesized compounds possessed good binding potential towards five selected cancer targets with minimum binding energy (kcal/mol), cluster root mean square deviation, number of hydrogen bond stabilization and maximum number of interacting residues. The molecular docking studies clearly showed that the synthesized compound 1 is demonstrated good binding potential to all the five selected cancer targets. The compound showed good binding to β-catenin (PDB ID: 1JDH) with binding energy of − 5.9 kcal/mol. The main residues present in the binding cavity were found to be Ser106, Pro192, Glu228, Ala230 and Lys233 (Fig. 5a). The compound demonstrated the potential binding with epidermal growth factor receptor (EGFR) (PDB ID: 4R3P) with binding energy of − 5.8 kcal/mol. Arg836, Lys860, Tyr869, Ala871 and Gly873 were identified to be the main residues present in the binding cavity of the receptor. The interaction also stabilized by two hydrogen bond formation with the residues such as Arg836 and Tyr869 (Fig. 5b). Similarly, the compound exhibited good binding potential to kinase domain of human HER2 (ERBB2) (PDB ID: 3PP0) with binding energy of − 6.5 kcal/mol. The major residues present in the binding pose were identified to be Leu755, Gly865 and Phe731; however, no hydrogen bond involved in the interaction (Fig. 5c). The compound also showed potential binding to cyclin D1-cyclin-dependent kinase 4 (PDB ID: 2W96) with a binding energy of − 6.7 kcal/mol. The main residues present in the binding cavity were found to be Pro183, Leu186, Glu184, Thr190, Trp238, Phe278 and Pro280. The interaction also stabilized by two hydrogen bonds with the residues Pro183 and Thr190 (Fig. 5d). Similarly, compound 1 also demonstrated good binding with RAC-beta serine/ Table 6. Selection of probable drug targets from various types of cancers for structure based drug screening. The drug targets were screened based on the virulent function in the metabolic pathways of each type of cancer. The structural coordinates of these drug targets were retrieved from PDB.  Table 7. The binding potential of (E)-2-(1-(3-aminophenyl) ethylidene) hydrazinecarboxamide and (E)-2-(2aminobenzylidene) hydrazinecarboxamide towards selected drug targets of various types of cancers predicted by molecular docking studies by AutoDock Vina.

Cancer receptors and PDB ID Ligand molecules Interacting residues Cluster RMS (Å)
No. of hydrogen bonds and interacting amino acids Binding energy (kcal/mol) Epidermal growth factor receptor (EGFR) (PDB ID: 4R3P) www.nature.com/scientificreports/ www.nature.com/scientificreports/ threonine-protein kinase B (PDB ID: 1GZK) with a binding energy of − 5.7 kcal/mol. The interaction is stabilized by a hydrogen bond with Thy231 and the major residues involved in the binding were found to be Thr231, Tyr177, Lys285 and Lys420 (Fig. 5e).
The molecular docking studies also demonstrated that the synthesized compound 2 exhibited potential binding to all the five selected cancer targets. The compound showed a good binding to β-catenin (PDB ID: 1JDH) with binding energy of − 5.9 kcal/mol. The main residues present in the binding cavity were found to be Phe21, Lys22, Glu36, Val349 and Lys354 (Fig. 6a). The compound 2 demonstrated the potential binding with epidermal growth factor receptor (EGFR) (PDB ID: 4R3P) with binding energy of − 5.7 kcal/mol. The interaction is stabilized by a hydrogen bond which was formed with Asp916 and the major residues present in the binding cavity were identified to be His393, Trp880, Lys879, Lys913, Gly917 and Asp916 (Fig. 6b). Similarly, the compound 2 exhibited good binding potential to kinase domain of human HER2 (ERBB2) (PDB ID: 3PP0) with binding energy of − 6.2 kcal/mol. The major residues present in the binding pose are Leu755, As758, Ala763, Glu766 and the interaction is also stabilized by a hydrogen bond at Asn758 (Fig. 6c). The compound 2 exhibited potential binding to cyclin D1-cyclin-dependent kinase 4 (PDB ID: 2W96) with a binding energy of − 6.0 kcal/mol. The main residues present in the binding cavity were found to be Glu184, Leu186, Gln188, Phe278 and Pro280 (Fig. 6d). Similarly, compound 2 also demonstrated good binding with RAC-beta serine/threonine-protein kinase B (PDB ID: 1GZK) with a binding energy of − 5.4 kcal/mol. The interaction is stabilized by a hydrogen bond with Asp440 and the major residues involved in the binding were found to be Glu236, Phe238, Glu279, Asp440 (Fig. 6e).
When the binding of two synthesized compounds was predicted against calf thymus (CT) DNA (PDB ID: 2DYW) by molecular docking studies, the modeling studies demonstrated that both the compounds showed profound binding towards the DNA. The compound 1 (1, 2-amino shift in (E)-2-(1-(3-aminophenyl) ethylidene) hydrazinecarboxamide) showed interaction with DNA at DG10, and the interaction stabilized with a hydrogen bond. The compound 1 interacted within the two strands of the DNA (Fig. 7a). The binding energy of the interaction is estimated to be − 6.71 kcal/mol. The ligand efficiency and intermolecular energy were found to be − 0.48 and − 7.24 kcal/mol respectively. The inhibition constant was found to be 12.13 µM. Similarly, when the binding of the compound 2 ((E)-2-(2-aminobenzylidene) hydrazinecarboxamide) towards the CT-DNA (PDB ID: 2DYW) were predicted by molecular docking studies showed significant binding with the DNA. The compound 2 showed profound binding with DNA with the binding energy of − 6.71 kcal/mol. The compound interacted within the two strands and the interaction stabilized with DG 6 and DG7 with two hydrogen bonds (Fig. 7b). The ligand efficiency and intermolecular energy were found to be − 0.47 and − 6.99 kcal/mol respectively. The inhibition constant was found to be 34.13 µM. From the interaction modeling by molecular docking, it is clear that both compound 1 and compound 2 interacted within both the strands of the DNA in almost the nearer binding sites. As shown in the experimental studies, the computational interaction modeling of the molecules and the DNA showed profound binding and the theoretical prediction is comparable with the experimental finding. The theoretical binding constants, inhibition constants, and binding energies are comparable with the experimental finding that showed a concurrent result. As shown in the experimental studies, computational modeling also suggested that the DNA binding potential of the compounds depends on the position of the amine group in the aromatic ring of the molecules.
The synthesized compounds 1 and 2 demonstrated good drug likeliness, pharmacokinetic features, ideal toxicity parameters and good binding with all the five selected cancer drug targets. β-catenin is involved in various types of cancers such as gastric cancer, colorectal cancer, endometrial cancer, thyroid cancer, hepatocellular carcinoma and studies suggested that it is one of the potential targets for drug development 40 . Both the compounds showed good binding with these targets. The binding energy of the interaction is the same and the interaction is stabilized by various amino acids. Epidermal growth factor receptor (EGFR) is one of the major cancer receptor, responsible for several types of cancer such as oral cancer, esophageal cancer, gastric cancer, bladder cancer, choriocarcinoma, cervical cancer, glioma and laryngeal cancer. The target can be prioritized as one of the potential cancer target 41 . Both the synthesized compounds were found to be equally potential as the binding energy is almost same. Kinase domain of human HER2 (ERBB2) is another important target has potential role in several types of cancer such as gastric cancer, pancreatic cancer, bladder cancer, endometrial cancer, ovarian cancer, choriocarcinoma, cervical cancer, breast cancer and cholangiocarcinoma and can be considered as one of the potential target for anti-cancerous compound development 42 . Cyclin D1-cyclin-dependent kinase 4 is one of the major types of the cancer receptor which involved in various types of cancer such as hairy cell leukemia, multiple myeloma, oral cancer, esophageal cancer, breast cancer and laryngeal cancer and can be considered as one of the promising drug targets 43 . Both the molecules demonstrated good binding potential with the cancer target. RAC-beta serine/threonine-protein kinase B is another important kind of cancer receptor considered as potential drug targets for lung cancer, neuroblastoma and gastric cancer. Both the compound found to be interacted with this cancer receptor and probably possess good inhibitory potential to the cancer target 44 . The present study suggested that compound 1 possesses better binding potential to most of the selected cancer receptor, thus the compound probably considered as potential inhibitors towards all the five cancer receptors. Compound 2 also showed comparable binding potential with stable interactions with minimum binding energy.
There are several studies suggested the anticancer properties of various synthesized derivatives of hydrazinecarboxamide [45][46][47][48] . To the best of our knowledge, this is the first study demonstrated the binding potential of the synthesized compounds 1 and 2 towards five selected targets which are considered as the major targets in various types of cancer. The anticancerous properties of the synthesized compounds were suggested based on computational virtual screening and molecular docking studies. Thus, molecular dynamic simulations studies can be performed to confirm the stability of the docked complexes of the synthesized compounds and prioritized cancer targets. Further, the efficiency and inhibitory potential of the synthesized compounds towards the prioritized targets needed to be tested at lower micro molar concentration by appropriate in vitro assays to validate Scientific Reports | (2020) 10:21913 | https://doi.org/10.1038/s41598-020-79027-1 www.nature.com/scientificreports/ Hirshfeld surfaces analysis. Hirshfeld surface analysis gives information on a crystal contained unique information on each molecule by dividing the crystal space in to non-overlapping molecular volume. Which allows us to excess intermolecular interactions. Here, dnorm is the parameter for normalized contact distance, which is based on (di), (de) and (r vdW) van der Waals radii of atom, which gives the equation as follows 49 , The Hirshfeld anslysis resolves and quantifies the intercontact between atoms in the crystal structure www.nature.com/scientificreports/ interactions 50 . The strong hydrogen bond spots as dark red region which is the result of short intercontact between the atoms showed on the dnorm surface and light red spots indicated the other intermolecular contact between the atoms. The electrostatic potential range is between plotted in − 0.030 a.u. to + 0.030 a.u on Hirshfeld surfaces based on Hartree-Fock theory and which is showed electrostatic potential over the Hirshfeld surfaces in Fig. 9 red region (hydrogen acceptor) negative. And blue region (hydrogen donor) positive electrostatic potentials 51 . The intercontacts with respect to di and de plotted for 2D fingerprint plots and quantified by using visualization of the Hirshfeld surfaces (Fig. 10). The intercontacts were found to be C…H (13.3%), H…H (47.6%), N…H (4.1%), O…H (9.1%). The intercontacts generated from the 2D Finger print plots are shown in Fig. 10. The major contributions of intercontacts Hirshfeld analysis are from H…H, C…H, O…H and S…H when compared to other intercontacts.

Conclusion
The C-demethylation and 1,2 amino shift in phenyl ring takes place simultaneously in compound 1 under mild conditions and became molecule 2. These kind of chemical changes was not reported in a compound to date. Furthermore, compound 2 cannot be made directly from 2-amino benzaldehyde, which has less stability; it is due to presence of aldehyde and amino group in the same molecule. Hence, in future this kind of easy chemical conversion can be used to make novel molecules and might be used in various fields. Nucleic acid interactions of compounds with CT-DNA was carried out, results revealed that they have displayed hyperchromism with bathochromic shift. Generally, this kind of results obtained from metal complexes has positively charged metal ions with nucleic acid gave electrostatic interaction. But these organic molecules 1 and 2 displayed excellent interaction with DNA and the molecular basis of the interaction was predicted by molecular docking studies. The computational virtual screening suggested that synthesized compounds possessed ideal drug likeliness, pharmacokinetic features and toxicity properties and demonstrated profound binding with selected cancer receptors are β-catenin, epidermal growth factor receptor, kinase domain of human HER2, cyclin D1-cyclin-dependent kinase 4 and RAC-beta serine/threonine-protein kinase B. Thus, biological studies of compounds provide ample foundation in the scale up of the applied approaches and nucleic acid interactions and computational prediction probably provides profound scope and application for the screening and designing of novel anticancer lead molecules. The Hirshfeld surface analysis of compound 2 was carried out and fingerprint plots were studied, results revealed that nature of molecular interactions and their contributions to the molecular surface.

Experimental section
Material and methods. All       Bioinformatics. Selection of probable drug targets of various types of cancer. Based on the extensive literature survey, the major targets involved in various types of cancer were identified. The description of the cancer drug targets selected in the present study is shown in Table 6. The selected cancer receptors are β-catenin (PDB: 1JDH), Epidermal growth factor receptor (PDB: 4R3P), Kinase domain of human HER2 (PDB: 3PP0), Cyclin D1-cyclin-dependent kinase 4 (PDB: 2W96) and RAC-beta serine/threonine-protein kinase B (PDB: 1GZK).
Preparation of structure of ligands. The 2D structure of these ligands (Scheme 1) was converted in to 3D structures and structure optimization was carried out by ACD 3D viewer 60 . The 3D structure of these ligands was saved in pdb format. The ligand structure was loaded into MGL Tools and the root atoms, number of torsion, rotatable and nonrotatable bonds were assigned and the file was saved in pdbqt format for docking studies.
Computer aided virtual screening of the ligand molecules. The drug likeliness features of (E)-2-(1-(3-aminophenyl) ethylidene) hydrazinecarboxamide and (E)-2-(2-aminobenzylidene) hydrazinecarboxamide were analyzed by the computational tools such as PreADMET 61 and SwissADME 62 . The main filters used for the drug likeliness are Lipinski's rule of five, CMC-like rule, Lead-like rule, MDDR-like rule and WDI-like rule available in PreADMET 63-66 and Ghose filter, Veber filter, Egan filter and Muegge filter were available in SwissADME. The molar refractivity, Log P, Topological polar surface area (TPSA) and bioavailability score were also predicted by SwissADME. Further, the adsorption, distribution, metabolism and excretion (ADME) features of these lead molecules by various statistical models available in PreADMET such as blood brain barrier (BBB) penetration, buffer solubility, human intestinal absorption (HIA), heterogeneous human epithelial colorectal adenocarcinoma (caco2) cell permeability, Madin Darby canine kidney (MDCK) cell permeability, plasma protein binding, CYP2C19 inhibition, CYP 2C9 inhibition, CYP2D6 inhibition, CYP3A4 inhibition, CYP3A4 substrate, Pgp inhibition, pure water solubility and skin permeability assays. The toxicity of the ligands was predicted by various options available in PreADMET such as acute algae toxicity, carcinogenicity in mouse rat models, in vitro hERG inhibition, acute fish toxicities in daphina, medaka and minnow models and in vitro Ames test.

Molecular docking studies.
Interaction modelling of the synthesized compounds and probable cancer targets. The binding potential of (E)-2-(1-(3-aminophenyl) ethylidene) hydrazinecarboxamide and (E)-2-(2-aminobenzylidene) hydrazinecarboxamide to the selected cancer receptors were studied by molecular docking by AutoDock Vina 67 . The grid box was assigned for each cancer target by setting the 3D coordinates for the binding pocket predicted by various binding pocket prediction server mentioned previously. The x, y, z coordinates for binding pockets in each target were assigned in the configuration file. The total grid data per map for each of www.nature.com/scientificreports/ the selected receptor was assigned by fixing the values for number of points in x, y, z-dimensions (size_x, size_y, size_z). Further, the parameter values for center grid box of each receptor (center_x, center_y, center_z) were assigned as per the standard protocol. The values for exhaustiveness in the docking simulation were assigned as per standard (Trott and Olson 2010). The molecular docking simulation was performed through command prompt. The output (log) files of the best nine confirmations were generated and conformations were ranked according to minimum binding energy (kcal/mol), docking score, cluster RMSD, number of hydrogen bond formed and the interacting residues present at the close proximity of 1.0 Å VWD scaling factor. Out of which, the first confirmations were selected as the best predicted model. The receptor ligand complex was visualized in MGL tools and PyMol 68 .
Interaction modeling of the synthesized compounds and CT-DNA. The binding potential of (E)-2-(1-(3-aminophenyl) ethylidene) hydrazinecarboxamide and (E)-2-(2-aminobenzylidene) hydrazinecarboxamide towards the CT-DNA was predicted by molecular docking studies. The 3D structure of the DNA (PDB ID: 2DYW) was retrieved from the PDB database. The structure was analysed thoroughly and the bound form of the ligand was removed and the targets were prepared for docking studies as per the standard protocol 68 . The 3D structure of each compound was loaded in AutoDock tool and the ligands were prepared by setting the root atom, the number of torsions, and other parameters for the ligand preparation as per the standard protocol 68 . The binding sites were analysed by AutoGrid program and the grid maps were generated. The molecular docking was performed by AutoDock program using Lamarckian Genetic algorithm and the dock parameter files were generated. The best-docked conformations were analysed based on the binding energy (kcal/mol), number of hydrogen bonds, and other weak interactions in MGL tool. The binding and inhibition constant (µM), and binding energy were also estimated, and the theoretical values were compared with that of the experimental binding values.
Hirshfeld surface analysis. The Hirshfeld surfaces computational program used to create the graphical representations and is used to perform and quantify the intermolecular interactions in terms of surface contribution by using the program Crystal Explorer 17.05 [69][70][71] and to generating electrostatic potential 72 with TONTO 49 . This program explores molecular packing and provides surface information on different types of intermolecular interactions of the crystal and can be identified and their contribution of individual contact by2D fingerprint plots (FP) shown in the colour plot.