Synthesis, Molecular Docking, Druglikeness Analysis, and ADMET Prediction of the Chlorinated Ethanoanthracene Derivatives as Possible Antidepressant Agents

: Ethanoanthracene cycloadducts (5–7) anti , (5–7) syn , and (5–7) dec have been synthesized from the Diels–Alder (DA) reaction of diene 1,8-dichloroanthracene 2 , with the dienophiles; acrylonitrile 3 , 1-cynavinyl acetate 4, and phenyl vinyl sulfone 5, individually. The steric e ﬀ ect of dienophile substituents were more favorable toward the anti -isomer formation as deduced from 1 H-NMR spectrum. The cheminformatics prediction for (5–7) anti and (5–7) syn was investigated. The in silico anticipated anti-depression activity of the (5–7) anti and (5–7) syn compounds were investigated and compared to maprotiline 9 as reference anti-depressant drug. The study showed that steric interactions play a crucial role in the binding a ﬃ nity of these compounds to the representative models; 4xnx, 2QJU, and 3GWU. The pharmacokinetic and drug-like properties of (5–7) anti and (5–7) syn exhibited that these compounds could be represented as potential candidates for further development into antidepressant-like agents.

Diels-Alder (DA) reaction [20], for which German chemists Diels and Alder were awarded the Nobel Prize in chemistry in 1950, is one of the most useful synthetic approaches for building ethanoanthracenes [21][22][23]. For example, maprotiline is a selective norepinephrine reuptake

PASS Online
For all compounds and maprotiline, the structure was drawn using a ChemDraw software; ChemDraw ® Ultra, version 12, 1986-2009, Cambridge Soft Corp, Akron, OH, USA. The structure was checked and cleaned up through pressing structure icon and saved as MDL Mol files (*.mol) to be ready for uploading into the PASS prediction website http://195.178.207.233/PASS/index.html (accessed on 15 April 2020). Then, the icon "predict activity" was pressed to give the predicted activities and their corresponding Pa values.

Preparation of the Compounds and Maprotiline
The preparation of the compounds (5-7) anti, (5-7) syn, (5-7) dec, and maprotiline 9 for docking studies is carried out as follows; the compound structure was drawn using ChemDraw Ultra 12/ Chem14 Pro 12 program, the structure was checked and cleaned up through pressing structure icon. The ACS document 1996 setting was selected by pressing file icon. The structure was copied and pasted into Chem14 Pro 12 window in order to obtain the 14 structures of the compound and to pre-optimize the compound with its function MM force field then saved as MOL File (.mol) format and thus being prepared for docking.

Preparation of the Protein Models
The crystal structures of hNET homology models with access protein data bank (PDB) codes 4xnx, 2QJU, and 3GWU were imported from the protein data bank (PDB; http://www.rcsb.org/pdb (accessed on 15 April 2020)) into the Molegro Virtual Docker (MVD 2013.6.0.0 (win32)) program individually. The structure was prepared through the pressing icon, and the errors in the amino acids, which appeared as a warning, were corrected. The protein surface and binding cavity were identified, respectively. After resetting the view from the view icon, the ligand that already binds in the binding site of the protein was removed, and the other ligands were converted to cofactors. The protein was ready for docking.

Preparation of the Compounds and Maprotiline
The preparation of the compounds (5-7) anti, (5-7) syn, (5-7) dec, and maprotiline 9 for docking studies is carried out as follows; the compound structure was drawn using ChemDraw Ultra 12/ Chem14 Pro 12 program, the structure was checked and cleaned up through pressing structure icon. The ACS document 1996 setting was selected by pressing file icon. The structure was copied and pasted into Chem14 Pro 12 window in order to obtain the 14 structures of the compound and to preoptimize the compound with its function MM force field then saved as MOL File (.mol) format and thus being prepared for docking.

Preparation of the Protein Models
The crystal structures of hNET homology models with access protein data bank (PDB) codes 4xnx, 2QJU, and 3GWU were imported from the protein data bank (PDB; http://www.rcsb.org/pdb) into the Molegro Virtual Docker (MVD 2013.6.0.0 (win32)) program individually. The structure was prepared through the pressing icon, and the errors in the amino acids, which appeared as a warning, were corrected. The protein surface and binding cavity were identified, respectively. After resetting the view from the view icon, the ligand that already binds in the binding site of the protein was removed, and the other ligands were converted to cofactors. The protein was ready for docking.

Docking of the Compounds into Protein Models
The previously optimized

ADMET Prediction
For all compounds and maprotiline, the structure was drawn using ChemDraw software; ChemDraw ® Ultra, version 12, 1986-2009, Cambridge Soft Corp., USA. The structure was checked and cleaned up through pressing the structure icon, saved as an MDL Mol file (*.mol), and were ready for uploading into the SwissADME server; http://www.swissadme.ch/. In the SwissADME server,

ADMET Prediction
For all compounds and maprotiline, the structure was drawn using ChemDraw software; ChemDraw ® Ultra, version 12, 1986-2009, Cambridge Soft Corp., USA. The structure was checked and cleaned up through pressing the structure icon, saved as an MDL Mol file (*.mol), and were ready for uploading into the SwissADME server; http://www.swissadme.ch/. In the SwissADME server,

ADMET Prediction
For all compounds and maprotiline, the structure was drawn using ChemDraw software; ChemDraw ® Ultra, version 12, 1986-2009, Cambridge Soft Corp., USA. The structure was checked and cleaned up through pressing the structure icon, saved as an MDL Mol file (*.mol), and were ready for uploading into the SwissADME server; http://www.swissadme.ch/ (accessed on 15 April 2020). In the SwissADME server, the import icon in the molecular sketcher was clicked and a new window opened to select the prepared structure, presented on the molecular sketcher, then transferred this structure to the SMILES format. After that, the icon "Run" was pressed to give the ADMET parameters and related values.

Synthesis
The starting material 1,8-dichloroanthracene 1 was prepared in a good yield from commercially available 1,8-dichlroroanthraquinone, according to the House et al. procedure [47]. A literature survey revealed that 1,8-dichloroanthracene 1 is not reactive in DA reaction as anthracene itself. For example, del Rosario Benites et al. [48] reported the DA reaction between 1,8-dichloroanthracene 1 and phenyl vinyl sulfoxide; the reaction was under N 2 at chlorobenzene reflux (130 • C) and gave the cycloadduct in 30% yield after an 8-day reaction time, while Paquette et al. reported the successful DA reaction in 83% yield between the same dienophile and anthracene [34]. With the line of our interest on the DA reaction, 1,8-dichloroanthracene 1 reacted with acrylonitrile, 1-cynavinyl acetate, and phenyl vinyl sulfone to investigate the substituents effect in terms of regioselectivity. Initially, the reaction of phenyl vinyl sulfone-with 1,8-dichloroanthracene 1, either in dichloromethane or toluene, under reflux conditions for 10 h-did not occur. Obviously, the reaction of 1,8-dichloroanthracene 1 with dienophiles is a challenge, and to get successful may be the reaction should be run under harsher conditions. An alternative way is to use the microwave-assisted organic synthesis technique because it has advantages compared to the conventional method, including being faster, safer, greener, more selective, and more economic [49][50][51]. This technique became an efficient tool in all organic reactions, including DA reaction [52,53]. We successfully synthesized the ethanoanthracenes via DA reaction with the microwave-assisted synthesis approach (Scheme 1).
Appl. Sci. 2020, 10, 7727 18 of 22 the import icon in the molecular sketcher was clicked and a new window opened to select the prepared structure, presented on the molecular sketcher, then transferred this structure to the SMILES format. After that, the icon "Run" was pressed to give the ADMET parameters and related values.

Synthesis
The starting material 1,8-dichloroanthracene 1 was prepared in a good yield from commercially available 1,8-dichlroroanthraquinone, according to the House et al. procedure [47]. A literature survey revealed that 1,8-dichloroanthracene 1 is not reactive in DA reaction as anthracene itself. For example, del Rosario Benites et al. [48] reported the DA reaction between 1,8-dichloroanthracene 1 and phenyl vinyl sulfoxide; the reaction was under N2 at chlorobenzene reflux (130 °C) and gave the cycloadduct in 30% yield after an 8-day reaction time, while Paquette et al. reported the successful DA reaction in 83% yield between the same dienophile and anthracene [34]. With the line of our interest on the DA reaction, 1,8-dichloroanthracene 1 reacted with acrylonitrile, 1-cynavinyl acetate, and phenyl vinyl sulfone to investigate the substituents effect in terms of regioselectivity. Initially, the reaction of phenyl vinyl sulfone-with 1,8-dichloroanthracene 1, either in dichloromethane or toluene, under reflux conditions for 10 h-did not occur. Obviously, the reaction of 1,8dichloroanthracene 1 with dienophiles is a challenge, and to get successful may be the reaction should be run under harsher conditions. An alternative way is to use the microwave-assisted organic synthesis technique because it has advantages compared to the conventional method, including being faster, safer, greener, more selective, and more economic [49][50][51]. This technique became an efficient tool in all organic reactions, including DA reaction [52,53]. We successfully synthesized the ethanoanthracenes via DA reaction with the microwave-assisted synthesis approach (Scheme 1). The DA reactions of anthracenes with dienophiles, which lead to the possible formation of more than products, in terms of regioselectivity and stereoselectivity, have attracted the attention of more researchers. The effects of the substituent nature on the reaction of 2-substituted anthracenes with malic anhydride, affording two isomeric adducts, have been reported [54]. The substituent on the 1-position of anthracenes exerted a significant steric effect on the reaction stereoselectivity [55]. The steric factor of the substituent on the 1-position of 1-succinimidoanthracene and 1-phthalimidoanthracene was found to favor anti adduct formation [56]. On the other hand, when the substituent located in the 2-position postulated to have no steric effect as well, no major electronic effect could be explained in terms of the mechanism of the reaction [54]. However, the determination ratio of the two possible adducts was initiated as an approach to investigate the reaction mechanism [57]. Khan et al. reported that the amino group has an electronic effect on the isomer ratio when it is located in the 1-or 2-position of anthracene moiety [58]. In our pervious wok, the steric effect was found to play a major role on the regioselectivity of the DA reaction of 10-allyl-1,8-dichloroanthracene with 2-chloroacrylonitrile, 1-cyanovinyl acetate, and phenyl vinyl sulfone [25]. Recently, we reported the DA reaction of 1,8-dichloroanthracene 2 and 2-chloroacrylonitrile under xylene reflux for 48 h affording two regioisomers in a 2:1 ratio [43]. The synthetic route towards target regioisomers (5-7) anti, (5-7) syn, and (5-7) dec is shown in Scheme 1. As a result of the reaction, 1,8-dichloroanthracene 1 with acrylonitrile 2, regioisomers 5 anti, 5 syn, and 5 dec were obtained as a mixture in a ratio 3.02:1.41:1, respectively (Supplementary Materials Figures S1 and S2). It sounds that the steric effect of the nitrile group, to chlorine atom, plays a crucial role in favor of regioisomer 5 anti.
The NMR signals assigned to the bridge-head protons are usually utilized for isomer identification, and regioselectivity [5,16,18,19,25], (Table 1, Figure S2). It is clear that isomer 5 anti was the major cycloadduct, because the steric effect of the nitrile group (CN). The bridge-head proton H-9 of the isomer anti 5 appeared at δ 5.46 ppm, which assigned, as triplet, with coupling constants J 2.4 Hz, on the other hand, the proton H-10 appeared at 4.60 ppm, as doublet, with coupling constants J 3.2 Hz, while the signal of the corresponding proton H-9 of the isomer syn 5 appeared at δ 4.44 ppm as triplet, with coupling constants J 2.8 Hz, and its signal assigned to the proton H-10 appeared at δ 5.71 ppm, as doublet, with coupling constants J 2.8 Hz. Moreover, the signals assigned to the proton H-9" of the dechlorinated isomer dec 5 appeared at upfield δ 4.39 ppm, as triplet, with coupling constants J 3.2 Hz, and its signal assigned to the proton H-10" appeared at δ 4.57 ppm, as doublet, with coupling constants J 2.4 Hz. The H-NMR signals of the unexpected dechlorination product 5 dec are well-matched to the signal of this compound, as reported in literature [59]. The DA reaction of 1,8-dichloroanthracene 1 with 1-cynavinyl acetate 3 afforded 6 anti, 6 syn, and 6 dec. The steric effect of acetoxy substituent, beside the nitrile group on the dienophile, slightly increases the formation of regioisomer 6 anti, since the ratio of 6 anti: 6 syn: 6 dec regioisomers is 2.7:1:2.5, respectively. In case of 10-allyl-1,8-dichloroanthracene, only anti isomer had been obtained when allowed 10-allyl-1,8-dichloroanthracene to react with the same dienophile 1-cyanovinyl acetate 3. This may be attributed to the strong steric effect of the ally substituent [25]. Interestingly, the tracing of these isomers on TLC using the solvent system EA/PE 1:5 showed that R f of 6 anti is the largest one and R f of 6 dec is larger than 6 syn ( Figure S3). After immersing this TLC in vanillin stain, the 6 anti and 6 syn exhibited yellow color, but 6 dec did not. However, the 6 dec spot was the most fluorescent under the UV lamp. Furthermore, the migration of these compounds on the TLC plate was consistent with their chemical nature, as the 6 anti and 6 syn had two chlorine atoms, but the 6 dec did not. The NMR signals assigned to the bridge-head protons are presented in the Table 2. It is clear that isomer 6 anti is the major, because the steric effect resulted from the acetoxy (OCOCH 3 ) and nitrile (CN) groups. The bridge-head proton H-9 of the isomer anti 6 appeared at δ 5.12 ppm and assigned as singlet, on the other hands, the proton H-10 appeared at 5.38 ppm as triplet with coupling constants J 2.8 Hz, while the signal of the corresponding proton H-9 of the isomer syn 6 appeared at δ 4.41 ppm, as triplet with coupling constants J 2.8Hz, indeed, the proton H-10 appeared at δ 6.09 ppm as broad singlet, as well, the proton H-9" of the dechlorinated isomer dec 7 appeared at upfield δ 4.36 ppm, as triplet, with coupling constants J 2.0Hz. Finally, the proton H-10" appeared at δ 5.07 ppm as singlet. The upfield shifting of the H-9" signals in the isomer dec 6, comparison to the isomer anti 7, may be attributed to the absence of chlorine atoms (Cl). Table 2. The distinguishing between the isomers 6 anti, 6 syn, and 6 dec using chemical shifts.

Chemical Shift
Chemical Shift H-9 H-9 H-9" H-10 H-10 H-10" The DA reaction of 1,8-dichloroanthracene 1 with phenyl vinyl sulfone 4 gave the isomers; 7 anti, 7 syn, and 7 dec in a ratio 6.58:1.70:1, respectively, as deduced from NMR spectrum of their crude ( Table 3, Figures S4 and S5). It is clear that isomer 7 anti is the major, because the steric effect resulted from the bulk phenyl sulfone group. In case of DA reaction of phenyl vinyl sulfone 4 with 10-allyl-1,8-dichloroanthracene, anti and syn isomers were obtained in a ratio 2:1 [25]. The increasing ratio of 7 syn may due to the steric hindrance of the phenyl group, and its close position to the allyl substituent that enforces the formation of the syn isomer in a more quantity than in case of 1,8-dichloroanthracene 1, where no allyl substituent [25]. The signal that assigned to bridge-head protons H-9 and H-10 of the isomer anti 7 appeared at chemical shifts δ 5.31 ppm, as triplet, and δ 4.82 ppm, as doublet, with coupling constants J 2 and 2.6 Hz, respectively, while the signal of the corresponding proton H-9 of the isomer syn 7 appeared at chemical shift δ 4.31 ppm, as triplet, with coupling constants J 2.6, and its signal assigned to the proton H-10 appeared at δ 5.56 ppm as broad singlet. Moreover, the signals assigned to the protons H-9 ' and H-10" of the dechlorinated isomer dec 7 appeared at upfield chemical shifts δ 4.25 ppm, as triplet, and 4.71 ppm, as doublet, with coupling constants J 2.4 and 2 Hz, respectively. The upfield shifting of the H-9" and H-10" signals in the isomer dec 7, comparison to the isomers anti 7 and syn 7, may be attributed to the absence of chlorine atoms (Cl). The H-NMR signals assigned to the 7 dec is well-matched to the signals of this compound reported in the literature [26]. It is noteworthy to report, here, that the obtaining unexpected dechlorination products 5 dec, 6 dec, and 7 dec is establishing a new dechlorination method that could be utilized and developed in the field of removal of chlorinated aromatic pollutants. Table 3. The distinguishing between the isomers 7 anti, 7 syn, and 7 dec using chemical shifts.

PASS Online
Prediction of Activity Spectra for Substances (PASS) is a free, online server designed for predicting the anticipated biological activities of the drug-like compounds (http://www.way2drug.com/passonline). This prediction is based on the structure-activity relationships (SAR) analysis of a library containing more than 300,000 organic compounds. Thus, PASS can be employed to optimize and well-target the chemical synthesis and biological testing [36]. The antidepressant and anti-phobic activities of the regioisomers (5-7) anti, (5-7) syn and maprotiline 9, as reference drug, were predicted, and the results are presented in Table 4. It is clear that compounds (5-7) anti and (5-7) syn could be recruited for the treatment of phobic disorders, as they have high predictive activity and are comparable to the maprotiline 9 (Pa: 0,842-0,877 for compounds (5-7) anti and (5-7) syn versus 0.897 for maprotiline 9, but their antidepressant activity is lower than maprotiline 9 (Pa: 0,142-0,226 for compounds (5-7) anti and (5-7) syn versus 0.626 for maprotiline 9. The binding affinity of the compounds has been investigated and recoded as MolDock score (Table 5, Figure S8). The compound 7 syn has the highest MolDock score, −117.691, while compound 5 syn has the lowest MolDock, −89.0117, Table 5. In contrast to maprotiline, the compounds 5 anti, 5 syn, 6 syn, 7 anti, and 7 syn exhibit no hydrogen bond. Their binding with protein model is due to their steric interaction with the amino acids; Phe 325, Val 327, Asp 121, Val 120, Phe 43, Gly 425, Tyr 123, and Tyr 124. The compound 6 anti exhibits hydrogen bond with distance 3.09982 Å between its oxygen and nitrogen of Tyr 124, and maprotiline 9 exhibits hydrogen bond with distance 2.92323 Å between its nitrogen and amine oxygen of Ser 421, Figure 1.

PASS Online
Prediction of Activity Spectra for Substances (PASS) is a free, online server designed for predicting the anticipated biological activities of the drug-like compounds (http://www.way2drug.com/passonline). This prediction is based on the structure-activity relationships (SAR) analysis of a library containing more than 300,000 organic compounds. Thus, PASS can be employed to optimize and well-target the chemical synthesis and biological testing [36]. The antidepressant and anti-phobic activities of the regioisomers (5-7) anti, (5-7) syn and maprotiline 9, as reference drug, were predicted, and the results are presented in Table 4. It is clear that compounds (5-7) anti and (5-7) syn could be recruited for the treatment of phobic disorders, as they have high predictive activity and are comparable to the maprotiline 9 (Pa: 0.842-0.877 for compounds (5-7) anti and (5-7) syn versus 0.897 for maprotiline 9, but their antidepressant activity is lower than maprotiline 9 (Pa: 0.142-0.226 for compounds (5-7) anti and (5-7) syn versus 0.626 for maprotiline 9. The binding affinity of the compounds has been investigated and recoded as MolDock score (Table 5, Figure S8). The compound 7 syn has the highest MolDock score, −117.691, while compound 5 syn has the lowest MolDock, −89.0117, Table 5. In contrast to maprotiline, the compounds 5 anti, 5 syn, 6 syn, 7 anti, and 7 syn exhibit no hydrogen bond. Their binding with protein model is due to their steric interaction with the amino acids; Phe 325, Val 327, Asp 121, Val 120, Phe 43, Gly 425, Tyr 123, and Tyr 124. The compound 6 anti exhibits hydrogen bond with distance 3.09982 Å between its oxygen and nitrogen of Tyr 124, and maprotiline 9 exhibits hydrogen bond with distance 2.92323 Å between its nitrogen and amine oxygen of Ser 421, Figure 1. Docking of the Compounds (5-7) anti, (5-7) syn, and Maprotiline 9 into 2QJU Model The compound 7 anti has the highest MolDock score, −103.8, with hydrogen bond interaction (2.74666 Å) between its oxygen and the amine of Arg 30, while 5 syn has the lowest MolDock score, −81.0802, with no hydrogen bonding, Table 5. The amino acids Phe 320 and Arg 30 may play a significant role in the binding affinity, since they participate in the steric interactions of all docked compounds. Furthermore, Arg 30 participates in the hydrogen bonding with the two docked compounds; 6 anti and 7 syn. Maprotiline exhibits hydrogen bond with distance 2.75112 Å between its nitrogen and oxygen of Asp 404, Figure 2. The other amino acids that affect the binding of the compounds are listed in Table 5.
Docking of the Compounds (5-7) anti, (5-7) syn, and Maprotiline 9 into 3GWU Model Although compound 7 syn has the highest MolDock score, −104.221, it has no hydrogen bond, while 5 anti has the lowest MolDock score, −79.4448, with hydrogen bond (2.79245 Å), Figure 3. The other docked compounds have no hydrogen bonding, so the hydrogen bonding is not more significant for enhancing binding affinity, (Table 5, Figure S9). The amino acid Ala 319 may play a major and significant role in the binding affinity, since it participates in the steric interaction of all docked compounds. As well, the amino acids Phe 320 and Arg 30 may play a significant role in the binding affinity, since they participate in the steric interaction of the most docked compounds. The others amino acids that affect the binding affinity of the compounds are listed in the Table 5. It is worthy, here, to mention that most of the amino acids that participate in the steric interaction of the compounds with the 3GWU model are the same or similar to those of the 2QJU model.

Drug-Likeness Prediction
Drug likeness is considered as a complex balance of molecular properties and structure merits that show the extent of similarity between certain compounds and known drugs. The drug likeness is guided by molecule properties, including hydrophobicity, electronic distribution, hydrogen bonding, molecule weight, pharmacophore entity, bioavailability, reactivity, toxicity, and metabolic stability [60]. Lipinski's rule is one of the tools most used to estimate the solubility and permeability of the compounds and, thus, to predict its qualification as a drug candidate. The rule states that: poor absorption or permeation is more likely when a compound violates Lipinski rule of 5; that it has more than 5 H-bond donors (the sum of NH and OH), molecular weight (MW t ) is over 500, Log P is over 5, and more than 10 H-bond acceptors (the sum of N and O) [61]. Table 6 depicts the drug likeness properties of the compounds (5-7) anti, (5-7) syn, and maprotiline 9 using a free web SwissADME server; http://www.swissadme.ch/ [62]. All tested compounds (5-7) anti, (5-7) syn, and maprotiline 9 as a reference drug do not violate Lipinski rule as their values fall in the normal range and the bioavailability shows that all compounds (5-7) anti and (5-7) syn are identical to the reference drug 9. Furthermore, these compounds (5-7) anti and (5-7) syn are in the optimal range of the physiochemical space, as shown by the colored zone of the bioavailability radar and, thus, these compounds could be considered as lead compounds.
The pharmacokinetic parameters revealed that the tested compounds 5 anti, 5 syn, 6 anti, and 6 syn are similar to the reference drug maprotiline 9, since they are highly absorbed through the gastrointestinal tract (GI) after oral intake, and able to penetrate through the blood-brain barrier (BBB), and could be effluxed by P-glycoprotein (P-gp), while compounds 7 anti and 7 syn are low absorbed through GI. This may be attributed to the molar refractivity (MR) of 7 anti and 7 syn, since it is higher than the MR of reference drug 9 (MR: 109.04 for 7 anti and 7 syn versus 88.32 for reference drug 9), and even higher than MR values of other tested compounds.
The Brenk and pan assay interference compounds (PAINS) structural alerts have been used in medicinal chemistry to predict unstable, reactive, toxic fragments present in the structure [63,64]. All compounds (5-7) anti and (5-7) syn have zero alerts in Brenk and PAINS descriptors, providing other promising indicators to be drug candidates.

ADMET Prediction
The ADMET properties of the compounds (5-7) anti, (5-7) syn, and the reference drug maprotiline 9, were predicted using ADMETlab server and the Data Warrior program, and the results were recorded in Tables 6 and 7. In term of absorption, all compounds (5-7) anti and (5-7) syn were found to have better membrane permeability (Caco-2) than reference drug 9. The permeability glycoprotein (P-gp) is the most controlling drug efflux transporter; it shuttles drugs outside the cells, to the gut lumen, to urine from kidney, and outside the brain. The predicting efflux values reveal that all of the tested compounds, which are considered as antidepressants, are passing the central nervous system (CNS) drug efflux problem that leads to decrease their bioavailability. In more detail; the compounds 5 anti, 7 anti, and 7 syn are neither substrates nor inhibitors for P-gp, while the compounds 5 syn, 6 anti, and 6 syn are inhibitors, but not substrates that may help as synergic drugs. On the other hand, the reference drug 9 is inhibitor and substrate. Interestingly, the compounds 6 anti and 6 syn had greater probability as P-gp inhibitors than reference drug 9 (0.814, 0.839, and 0.558, respectively). The Human Intestinal Absorption (HIA) scores are coming in the line of qualitative gastrointestinal (GI) absorption descriptors, as predicted by SwissADME server, that the HIA scores of compounds 7 anti and 7 syn are less than the reference drug 9 score. In terms of distribution, three descriptors were measured; Plasma Protein Binding (PPB), Volume Distribution (VD), and Blood-Brain Barrier (BBB). In fact, there are three types of plasma protein that bind with drugs; human serum albumin, αacid glycoprotein, and lipoprotein. The 6 anti is the most one bound to the plasma protein followed by its isomer 6 syn (PPP%: 93.528 and 93.448 respectively; versus, 87.029 of reference drug 9), that is may refer to their acetoxy group (OCOCH 3 ) that increase their topological polar surface area (TPSA). In the same trend, the volume distribution (VD) results predict that compounds 6 anti, and 6 syn are the most confined to blood components (VD: −0.331 and −0.323; respectively; versus, 1.603 of reference drug 9), as well, the isomers 7 anti and 7 syn are strongly confined to blood. On the other hand, the isomers 5 anti and 5 syn were found to evenly distributed between blood and tissue components (VD: 0.581 and 0.604, respectively). All of the compounds could be penetrated through BBB. The 6 anti and 6 syn have the least BBB values (BBB: 0.915 for both; versus 0.984 for reference drug 9). It is clear that acetoxy group (OCOCH 3 ) of the 6 anti and 6 syn raise up their TPSA and could participate in the hydrogen bonding that ultimately lead to decrease their penetration through BBB. This is in a good agreement with the Young et al. study where the excessive hydrogen bonding restricted the antihistamines penetration into the CNS [65]. In terms of metabolism, the compounds (5-7) anti and (5-7) syn could be metabolized by P450 CYP1A2 enzyme, so the lower affinity drugs to this enzyme are not preferred to intake with these compounds, as well, these compounds are not preferred to intake with reference drug 9, since it has the highest affinity. In addition, only the 5 anti and its isomer 5 syn could behave as reversible and/or irreversible inhibitors for the P450 CYP1A2 enzyme and, thus, lead to reduce and/or diminish its activity. In the same manner, the compounds (5-7) anti and (5-7) syn could be metabolized by P450 CYP74, the enzyme responsible for biotransformation for about 30% of prescribed drugs [66]. The reference drug 9 could behave either as substrate or inhibitor for P450 CYP74 enzyme. The compounds 5 anti, 5 syn, 6 anti, 6 syn, and reference drug 9 are considered neither substrates nor inhibitors for the P450 CYP119 enzyme, while the 7 anti and 7 syn could behave either as substrates or inhibitors. The compounds (5-7) anti and (5-7) syn have a moderate half-life (T 1/2 ) time and low clearance rate (CL). The 5 anti and its isomer 5 syn have higher T 1/2 and CL, but reference drug 9 has the highest (T 1/2 : 2.174, 2.34, and 3.401 h; CL: 1.424, 1.425, and 2.599 mL/min/kg, respectively). Table 6. Physicochemical, pharmacokinetics, and medicinal chemistry properties of the compounds (5-7) anti, (5-7) syn, and Maprotiline 9 using SwissADME server.  Drug toxicology is one of the essential areas in the preclinical investigation, since the toxicity is a major factor leading for drug attrition at discovery and development phases [67]. Thus the in silico reliable prediction of compound toxicity is important in order to reduce expenses and to save a lot of time during the drug discovery and development [68]. The toxicity risk of the compounds (5-7) anti and (5-7) syn, and maprotiline 9, was assessed quantitatively by ADMETlab server. Tables 7 and 8 and qualitatively by Data Warrior, a free tool from the marker of OSIRIS Property Explorer, Table 8. Table 8. Toxicity risk assessment of the compounds (5-7) anti, (5-7) syn, and maprotiline 9 using Data Warrior and ADMETlab server *.
From a quantitative point of view, the four toxicity descriptors; cardiotoxicity, hepatotoxicity, Ames mutagenicity, and median lethal dose (LD 50 ) of acute toxicity were predicted. Cardiotoxicity is one of the side effects resulting from the off-target interactions between drugs and ion channels that play crucial roles in adjusting the cardiac action potential, as well as control the cardiac rhythm [69]. Among these ion channels, the human ether-a-go-go-related gene (hERG) a potassium ion channel, was found to be more related to severe cardiotoxicity [70,71], so the hERG-blockade is considered as a major retardant in providing the pharmaceutical market with safe drugs. Furthermore, the off-target interaction of hERG is the leading cause of drug withdrawal from the pharmaceutical market [72,73]. Thus, the identification of potential hERG blocker is important at early phases of drug discovery [69]. The assessment results of the hERG-based toxicity show that 5 anti and its isomer 5 syn have no cardiotoxicity and are the safest among other compounds. The highest cardiotoxicity value of the reference drugs 9 is, partially, at least, due to their containing basic amine on their alkyl side chain (hERG: ++ 0.879 for reference drugs 9 vs. -0.261 for 5 syn). Hepatotoxicity and cardiotoxicity are the major factors participating for drugs attrition. Unfortunately, a quarter of the marketed drugs are withdrawn due to their adverse hepatic effects [74,75]. The 5 syn and 7 anti are safe to the liver, in contrast, 5 anti, 6 anti, 6 syn, 7 syn are not safe. Furthermore, reference drug 9 is predicted to be the most hepatotoxic with probability ++ 0.878. The Ames test, referring to the inventor Ames, is an assay to detect the mutagenicity of the compound; the ability of the compound to stimulate the genetic damage and mutations [76,77]. Similar to reference drug 9, the 6 anti, 6 syn, 7 anti, and 7 syn exert no mutagenicity effects, while the 5 anti and its isomer 5 syn tend to induce genetic damage and mutations as predicted by Ames test. These results are not consistent with qualitative results obtained by the Data Warrior program, since 6 anti and 6 syn are highly mutagenic, while other compounds are not mutagenic. This may refer to the difference in the reference models used in both tests. Determination of an acute toxicity is another important task in the destination of drug discovery and development. The acute toxicity is expressed as median lethal dose (LD 50 ); the dose amount of a tested compound that kill fifty percent (50 %) of the treated animals within a given time [78]. The 7 anti, and its isomer 7 syn exert the highest acute toxicity, and indicate their lowest LD 50 values, while the reference drug 9 is safer due to the highest value (LD 50 (mg/Kg): 262.66, 249.11, and 912.27, respectively). The 5 anti and its isomer 5 syn exhibit no acute toxicity with high LD 50 802.40, and 811.69 (mg/Kg), respectively, but the 6 anti and its isomer 6 syn may exert low toxicity.
From a qualitative point of view, the compounds 5 anti, 5 syn, 7 anti, and 7 syn exhibit no mutagenicity, no tumorigenicity, no reproductive effect, and also have no irritability toxicity. The reference drug 9 shares with compounds 6 anti and 6 syn in their excreting reproductive toxicity. Furthermore, compounds 6 anti and 6 syn revealed mutagenicity and irritability toxicity

Conclusions
In conclusion, ethanoanthracene cycloadducts (5-7) anti, (5-7) syn, and (5-7) dec were synthesized, and the substituent effects on the regioselectivity of these cycloadducts were described. The antidepressant and anti-phobic activities of these cycloadducts (5-7) anti, (5-7) syn, and maprotiline 9 as a reference drug, were theoretically predicted through PASS and molecular docking. The PASS activity of the (5-7) anti and (5-7) syn against phobic disorders were high and comparable to maprotiline 9, but variable and much less than maprotiline 9 as antidepressant. The MolDock scores of (5-7) anti and (5-7) syn are in the line with PASS values. The pharmacokinetic and druglikeness properties of (5-7) anti and (5-7) syn are valuable and provide an informative, promising analysis, as they are possible antidepressant-like compounds. Further in vitro and in vivo studies would be needed for confirmation of the chemoinformatics investigation.