Synthesis and Anticancer Activity of 3,4-Diaryl-1,2-dihydro- and 1,2,3,4-Tetrahydroquinolines

Tetrahydroquinolines are key structures in a variety of natural products with diverse pharmacological utilities and other applications. A series of 3,4-diaryl-5,7-dimethoxy-1,2,3,4-tetrahydroquinolines were synthesized in good yield by reacting 3-aryl-5,7-dimethoxy-2,3-dihydroquinolin-4-ones with different Grignard reagents followed by the dehydration of the intermediate phenolic compounds. Subsequent reduction and deprotection were carried out to achieve the desired tetrahydroquinolone moiety. The lead compound 3c showed low micromolar inhibition of various cancer cell lines. Demethylation under different reaction conditions was also investigated to afford the corresponding monohydroxy analogues.


Introduction
Quinolones and their derivatives occur in numerous natural products, and many of them display interesting biological activities.In particular, hydrogenated quinoline moieties are important structures in various natural products that exhibit a broad range of biological properties and potential pharmaceutical applications [1].
Rapid development in the chemistry of tetrahydroquinolines has been observed in recent years, because they are core structures in many important pharmacological agents [2] and drug molecules, such as antiarrhythmic and cardiovascular agents, anticancer drugs, and immunosuppressants, and ligands for 5-HTTA and NMDA receptors [3].Besides their pharmaceutical applications, tetrahydroquinoline derivatives are useful as pesticides [4,5], antioxidants [6][7][8], and corrosion inhibitors [9] and are active components of various dyes [10,11].In addition, they also have found application in modern recording technologies [12,13].
Several compounds of this class are found to occur in nature.For example, discorhabdin C, a polycyclic system based on tetrahydroquinoline, is a marine alkaloid [2,26,27] and dynemicin A, a natural antitumor antibiotic, has a complex structure built on the tetrahydroquinoline nucleus [28,29].A 2,4,6-trisubstituted tetrahydroquinoline, isolated from Martinella iquitosensis, exhibits activity as a bradykinin antagonist against α-adrenergic and histaminergic receptors [30].
2-Substituted tetrahydroquinolines have also recently been discovered as a class of natural products, that includes angustureine, cuspareine, oxamniquine, and galipeine.Some members of this family exhibit interesting pharmacological properties [31].
The dihydroquinolone structure can be found in a variety of natural products [32,33] and in a large number of compounds, which display biological activity.For example, 2,2,4-trisubstituted-1,2-dihydroquinolines have been used to produce potent compounds that possess antibacterial, antidiabetic, and anti-inflammatory activities [34].Compounds possessing this motif have also been shown to act as lipid peroxidation inhibitors, 4 HMG-CoA reductase inhibitors, ileal bile acid transporter inhibitors, and progesterone agonists and antagonists [34].
Due to the diverse biological activities exhibited by this class of compounds, we were interested in synthesizing dihydrogenated and tetrahydrogenated derivatives of quinolines.It was envisaged that these hydrogenated compounds would show promising biological activities, due to the pharmacological applications shown by the isoflavone class of compounds.The isoflavone analogue triphendiol 1 (Figure 1) has recently been granted orphan drug status by the FDA for pancreatic and bile duct cancers and latestage melanoma.Inspired by the success of triphendiol 1 it was decided to introduce an aryl group at the C4 position of the 3-substituted quinolone ring system and keep the oxygenation pattern similar to that of natural isoflavones to generate azaisoflavene 2 and azaisoflavan 3 structures.Furthermore, many biologically active isoflavones suffer from poor oral bioavailability and rapid CYP450 metabolism [35,36].Our objective in this study was to synthesize biologically active tetrahydroquinolines with improved solubility and metabolic stability compared to current isoflavone analogues.The biological activity of the tetrahydroquinolines was investigated through a single concentration screen in different cancer cell lines.
2-Substituted tetrahydroquinolines have also recently been discovered as a class natural products, that includes angustureine, cuspareine, oxamniquine, and galipei Some members of this family exhibit interesting pharmacological properties [31].
The dihydroquinolone structure can be found in a variety of natural products [32, and in a large number of compounds, which display biological activity.For example, 2,2 trisubstituted-1,2-dihydroquinolines have been used to produce potent compounds th possess antibacterial, antidiabetic, and anti-inflammatory activities [34].Compounds p sessing this motif have also been shown to act as lipid peroxidation inhibitors, 4 HM CoA reductase inhibitors, ileal bile acid transporter inhibitors, and progesterone agoni and antagonists [34].
Due to the diverse biological activities exhibited by this class of compounds, we w interested in synthesizing dihydrogenated and tetrahydrogenated derivatives of quin lines.It was envisaged that these hydrogenated compounds would show promising b logical activities, due to the pharmacological applications shown by the isoflavone cl of compounds.The isoflavone analogue triphendiol 1 (Figure 1) has recently been grant orphan drug status by the FDA for pancreatic and bile duct cancers and late-stage me noma.Inspired by the success of triphendiol 1 it was decided to introduce an aryl gro at the C4 position of the 3-substituted quinolone ring system and keep the oxygenati pa ern similar to that of natural isoflavones to generate azaisoflavene 2 and azaisoflav 3 structures.Furthermore, many biologically active isoflavones suffer from poor oral b availability and rapid CYP450 metabolism [35,36].Our objective in this study was to sy thesize biologically active tetrahydroquinolines with improved solubility and metabo stability compared to current isoflavone analogues.The biological activity of the tetrah droquinolines was investigated through a single concentration screen in different can cell lines.

Results and Discussion
The project mainly involved the synthesis of a series of 3,4-diaryl-5,7-dimethoxy 1 dihydro and 1,2,3,4-tetrahydroquinolines.Various demethylating conditions were inv tigated in order to obtain the corresponding hydroxyl analogues, which could possib show greater biological activity.

Results and Discussion
The project mainly involved the synthesis of a series of 3,4-diaryl-5,7-dimethoxy 1,2-dihydro and 1,2,3,4-tetrahydroquinolines.Various demethylating conditions were investigated in order to obtain the corresponding hydroxyl analogues, which could possibly show greater biological activity.
The acid-catalyzed dehydration of 5 did not result in the formation of the desired 3,4-diaryl-1,2-dihydroquinoline 2 but was found to preferentially generate the more highly stabilized 3,4-diaryl-5,7-dimethoxyquinoline 6 in 54-66% yield instead (Scheme 1).The formation of the quinoline moiety can be attributed to further dehydrogenation presumably due to aerial oxidation, occurring together with the dehydration reaction.The quinoline structure was confirmed by 1 H NMR spectroscopy, which showed the disappearance of the aliphatic protons H2 and H3 present in 5.The presence of a significantly downfield shifted singlet at δ 8.75 ppm in 6 correlating to H2 was the key resonance signal for the identification of compound 6.
multiplet at δ 3.46 ppm.The OH group was seen as a singlet at δ 3.79 ppm and the phenyl ring protons appeared as a multiplet at δ 7.11-7.20 ppm.
The acid-catalyzed dehydration of 5 did not result in the formation of the desired 3,4diaryl-1,2-dihydroquinoline 2 but was found to preferentially generate the more highly stabilized 3,4-diaryl-5,7-dimethoxyquinoline 6 in 54-66% yield instead (Scheme 1).The formation of the quinoline moiety can be a ributed to further dehydrogenation presumably due to aerial oxidation, occurring together with the dehydration reaction.The quinoline structure was confirmed by 1 H NMR spectroscopy, which showed the disappearance of the aliphatic protons H2 and H3 present in 5.The presence of a significantly downfield shifted singlet at δ 8.75 ppm in 6 correlating to H2 was the key resonance signal for the identification of compound 6.Scheme 1. Grignard reaction on 2,3-dihydroquinolin-4-ones.
This result was found to be consistent with literature reports [37,38], which state that 1,2-dihydroquinolines that are unsubstituted at the nitrogen atom and have at least one hydrogen at C2 are unstable.These dihydroquinolines are rapidly oxidized in air to the quinoline or undergo disproportionation by trace acids to give a mixture of the quinoline and tetrahydroquinoline.
To overcome this problem, it was decided to protect the NH group and then perform the Grignard reaction on the N-protected dihydroquinolin-4-ones to give the corresponding N-protected azaisoflavan-4-ol, followed by dehydration and deprotection to give the desired 3,4-diaryl-azaisoflavenes 2.
Hence, compound 7 was heated at reflux for 4 h with aryl magnesium bromide.This reaction gave a mixture of 3,4-diaryl-5,7-dimethoxy-1,2,3,4-tetrahydroquinolin-4-ol 5 and ethyl-3,4-diaryl-4-hydroxy-5,7-dimethoxy-3,4-dihydroquinoline-1(2H)-carboxylate 8.The ratio of 5 to 8 was found to be dependent upon the reaction conditions.The use of three equivalents of arylmagnesium bromide afforded 5 and 8 in 30% and 50% yield, respectively, whereas the use of two equivalents of arylmagnesium bromide gave only 8 in 78-88% yield.The structure of compound 8 was confirmed by 1 H NMR spectroscopy data.Compound 8a showed the presence of the COOEt group as a triplet and a quartet at δ 1.19 ppm (J = 6.0 Hz) and 4.15 ppm (J = 6.0 Hz) for the CH3 and CH2 groups, respectively.The OH proton was present as a singlet at δ 5.04 ppm and the phenyl ring protons as a multiplet at δ 7.07-7.14ppm.The 1 H NMR data indicate that 8 was formed as a single diastereomer.Due to the presence of an aryl substituent at the C3 position, we assume that the This result was found to be consistent with literature reports [37,38], which state that 1,2-dihydroquinolines that are unsubstituted at the nitrogen atom and have at least one hydrogen at C2 are unstable.These dihydroquinolines are rapidly oxidized in air to the quinoline or undergo disproportionation by trace acids to give a mixture of the quinoline and tetrahydroquinoline.
To overcome this problem, it was decided to protect the NH group and then perform the Grignard reaction on the N-protected dihydroquinolin-4-ones to give the corresponding N-protected azaisoflavan-4-ol, followed by dehydration and deprotection to give the desired 3,4-diaryl-azaisoflavenes 2.
Hence, compound 7 was heated at reflux for 4 h with aryl magnesium bromide.This reaction gave a mixture of 3,4-diaryl-5,7-dimethoxy-1,2,3,4-tetrahydroquinolin-4-ol 5 and ethyl-3,4-diaryl-4-hydroxy-5,7-dimethoxy-3,4-dihydroquinoline-1(2H)-carboxylate 8.The ratio of 5 to 8 was found to be dependent upon the reaction conditions.The use of three equivalents of arylmagnesium bromide afforded 5 and 8 in 30% and 50% yield, respectively, whereas the use of two equivalents of arylmagnesium bromide gave only 8 in 78-88% yield.The structure of compound 8 was confirmed by 1 H NMR spectroscopy data.Compound 8a showed the presence of the COOEt group as a triplet and a quartet at δ 1.19 ppm (J = 6.0 Hz) and 4.15 ppm (J = 6.0 Hz) for the CH 3 and CH 2 groups, respectively.The OH proton was present as a singlet at δ 5.04 ppm and the phenyl ring protons as a multiplet at δ 7.07-7.14ppm.The 1 H NMR data indicate that 8 was formed as a single diastereomer.Due to the presence of an aryl substituent at the C3 position, we assume that the Grignard reagent attacks the planar carbonyl group from the opposite side to the aromatic ring.
The acid-catalyzed dehydration of 8 with BF 3 .OEt 2 in DCM resulted in the formation of ethyl 3,4-diaryl-5,7-dimethoxy-quinoline-1(2H)-carboxylate 9 in 52-58% yield.The structure of compound 9 was also confirmed by 1 H NMR spectroscopy data.Compound 9a indicated the disappearance of the H3 proton and the appearance of a singlet at δ 4.54 ppm integrating for two protons, correlating to H2 and a triplet at δ 1.29 ppm (J = 9.0 Hz) and quartet at δ 4.24 ppm (J = 9.0 Hz) indicating the presence of the protecting group, -COOEt.
Subsequent deprotection of 9 was investigated using both acidic and basic conditions under an inert atmosphere.When HBr in AcOH [39] was utilized, 9 underwent dehydrogenation in addition to deprotection, giving the more stable quinoline moiety 6.Alternatively, heating 9 at reflux with 10% NaOH in EtOH for 5 h generated a mixture of two compounds by TLC (Scheme 2).Separation of these two compounds was attempted using various methods, such as column chromatography, crystallization, and preparative TLC, but was ultimately unsuccessful.When the crude mixture was analyzed using 1 H NMR spectroscopy, it was found to contain compounds 2 and 6 in an approximate ratio of 30:70.The presence of compound 2 was confirmed by the DEPT-135 NMR experiment, which showed the presence of a CH 2 at δ 47.5 ppm.
Grignard reagent a acks the planar carbonyl group from the opposite side to the aromatic ring.
The acid-catalyzed dehydration of 8 with BF3.OEt2 in DCM resulted in the formation of ethyl 3,4-diaryl-5,7-dimethoxy-quinoline-1(2H)-carboxylate 9 in 52-58% yield.The structure of compound 9 was also confirmed by 1 H NMR spectroscopy data.Compound 9a indicated the disappearance of the H3 proton and the appearance of a singlet at δ 4.54 ppm integrating for two protons, correlating to H2 and a triplet at δ 1.29 ppm (J = 9.0 Hz) and quartet at δ 4.24 ppm (J = 9.0 Hz) indicating the presence of the protecting group, -COOEt.
Subsequent deprotection of 9 was investigated using both acidic and basic conditions under an inert atmosphere.When HBr in AcOH [39] was utilized, 9 underwent dehydrogenation in addition to deprotection, giving the more stable quinoline moiety 6.Alternatively, heating 9 at reflux with 10% NaOH in EtOH for 5 h generated a mixture of two compounds by TLC (Scheme 2).Separation of these two compounds was a empted using various methods, such as column chromatography, crystallization, and preparative TLC, but was ultimately unsuccessful.When the crude mixture was analyzed using 1 H NMR spectroscopy, it was found to contain compounds 2 and 6 in an approximate ratio of 30:70.The presence of compound 2 was confirmed by the DEPT-135 NMR experiment, which showed the presence of a CH2 at δ 47.5 ppm.Therefore, the proposed synthesis of 3,4-diaryl-5,7-dimethoxy-1,2-dihydroquinoline 2 could not be accomplished via the methodologies used in this project.Instead, the more stable quinoline derivatives 6 were prepared in similar yields using the two strategies outlined in Schemes 1 and 2.

Synthesis of 3,4-Diaryl-1,2,3,4-Tetrahydroquinolines
The literature indicates that direct reduction of quinolines is the most efficient method of preparing tetrahydroquinolines [40].Direct reduction of the 2,3-dihydroquinolin-4-ones is also possible [41][42][43], but this would lead to a C4-unsubstituted product.Hence, a empts were made to reduce the quinoline 6 to 3,4-diaryl-tetrahydroquinoline 3 directly by applying two different reduction conditions.The first one involved heating the quinoline 6 with LAH at reflux and the second involved catalytic hydrogenation with hydrogen over palladium on charcoal.However, both reaction conditions gave multiple spots by TLC, which could not be separated to allow successful identification of the compounds formed.Therefore, the proposed synthesis of 3,4-diaryl-5,7-dimethoxy-1,2-dihydroquinoline 2 could not be accomplished via the methodologies used in this project.Instead, the more stable quinoline derivatives 6 were prepared in similar yields using the two strategies outlined in Schemes 1 and 2.

Synthesis of 3,4-Diaryl-1,2,3,4-tetrahydroquinolines
The literature indicates that direct reduction of quinolines is the most efficient method of preparing tetrahydroquinolines [40].Direct reduction of the 2,3-dihydroquinolin-4ones is also possible [41][42][43], but this would lead to a C4-unsubstituted product.Hence, attempts were made to reduce the quinoline 6 to 3,4-diaryl-tetrahydroquinoline 3 directly by applying two different reduction conditions.The first one involved heating the quinoline 6 with LAH at reflux and the second involved catalytic hydrogenation with hydrogen over palladium on charcoal.However, both reaction conditions gave multiple spots by TLC, which could not be separated to allow successful identification of the compounds formed.
For instance, ethyl-3,4-diaryl-5,7-dimethoxy-quinoline-1(2H)-carboxylate 9 was hydrogenated in THF using Pd/C and hydrogen gas overnight to give ethyl-3,4-diaryl-5,7dimethoxy-3,4-dihydroquinoline-1(2H)-carboxylate 10 in 91-95% yield (Scheme 3).The structure of compound 10 was confirmed by 1 H NMR spectroscopy data.Compound 10 showed a doublet of a triplet for the H3 proton at δ 3.36 ppm (J = 3.0, 12.0 Hz) and a doublet for the H4 proton at δ 4.45 ppm (J = 3.0 Hz) and triplet at δ 3.67 ppm (J = 12.0 Hz) and a doublet of doublet of doublet at δ 4.09 ppm (J = 3.0, 6.0, 12.0 Hz) for two H2 protons.The deprotection of compound 10 was carried out by heating at reflux in 10% NaOH in EtOH for 6 h to give the desired 5,7-dimethoxy-3,4-diaryl-1,2,3,4-tetrahydroquinolines 3 in 82-90% yield.The structure of 3 was confirmed by the lack of protons correlating to the -COOEt group in the 1 H NMR spectrum and the presence of a NH proton as a singlet at δ 4.14 ppm.The coupling constant between H3 and H4 protons (J = 3.0 Hz) indicated the cis configuration had been retained, and this was further confirmed by X-ray crystallographic analysis.The ORTEP diagram of compound 3c is shown in Figure 3. Packing of molecules was dominated by CH3-π and slipped π-π interactions (Figure S1).The expected cis stereochemistry of the compound 10 was established on the basis of the coupling constant of J = 3.0 between H3 and H4.Moreover, NOE correlations between the H3 and H4 protons were evident in the 2D NMR experiment, further confirming the cis configuration (Figures 2 and S11).
Initially, tetrahydroquinoline 3c was heated at reflux with aluminum chloride in chlorobenzene for 1 h.Two compounds were obtained after aqueous work up and column chromatography.However, only one compound was pure enough to be characterized as 12 in 13% yield (Scheme 4).The reaction with AlCl 3 therefore resulted in subsequent aromatization and dearylation in addition to demethylation.After synthesizing the dimethoxy-substituted tetrahydroquinolines, aim was to investigate suitable demethylation conditions to generate the logues.

Entry
Initially, tetrahydroquinoline 3c was heated at reflux with alumin chlorobenzene for 1 h.Two compounds were obtained after aqueous work chromatography.However, only one compound was pure enough to be c 12 in 13% yield (Scheme 4).The reaction with AlCl3 therefore resulted in s matization and dearylation in addition to demethylation.The second compound that was obtained from this reaction is hypothesized to be 11c, however, its purity was very poor and decomposed during purification.Meanwhile, the 1 H NMR spectrum of compound 12 lacked the phenyl group protons at C4 in addition to the methoxy protons.Three signals as a singlet corresponding to the hydroxyl protons were present at δ 9.40, 9.69, and 10.26 ppm, which confirmed that the demethylation had occurred.Doublets at δ 8.19 ppm and 8.75 ppm (J = 2.4 Hz) for H3 and H4 indicated that aromatization had occurred.The structure of compound 12 was further confirmed by HRMS, giving a molecular ion peak [M + H] + at 254.0805 corresponding to the molecular formula.
To improve the yield and the purity of the dihydroxy analogues, BBr 3 in DCM was examined as an alternate demethylating agent.According to the previous results, two equivalents of BBr 3 were required for the cleavage of one methoxy group.Hence, the same strategy was applied and tetrahydroquinoline 3e was stirred with seven equivalents of BBr 3 in DCM at room temperature for 1 h (Scheme 5).However, the product of this reaction could not be purified even after several attempts.This could possibly be due to the presence of three hydroxyl groups in compound 11 rendering the compound significantly unstable.Therefore, it was decided to demethylate tetrahydroquinoline 3f, which has only two methoxy groups, with five equivalents of BBr 3 in DCM at room temperature for 1 h (Scheme 5).However, purification of compound 11f proved to be similarly problematic and pure spectra of the product could not be obtained.tion could not be purified even after several a empts.This could possibly be due to the presence of three hydroxyl groups in compound 11 rendering the compound significantly unstable.Therefore, it was decided to demethylate tetrahydroquinoline 3f, which has only two methoxy groups, with five equivalents of BBr3 in DCM at room temperature for 1 h (Scheme 5).However, purification of compound 11f proved to be similarly problematic and pure spectra of the product could not be obtained.In an a empt to improve the purity of the hydroxyl compounds, the number of equivalents of BBr3 was reduced.Tetrahydroquinoline 3f was stirred with three equivalents of BBr3 in DCM for 1 h.Work up and purification with column chromatography resulted in a 38% yield of compound 13f (Scheme 6, Table 2).The 1 H NMR spectrum of compound 13f still showed the presence of a methoxy group at δ 3.69 ppm and only one hydroxyl group was evident at δ 7.78 ppm.To confirm the structure of this compound a NOE experiment was performed, which showed a correlation between the methoxy and the H6 and H8 protons while the hydroxy signal only correlated to the H6 proton.This indicated that demethylation had occurred at the C5 position, thus giving 5-hydroxy-7methoxytetrahydroquinoline 13f.Scheme 6. Demethylation of 3f using BBr3 (3 equiv.) in DCM.In an attempt to improve the purity of the hydroxyl compounds, the number of equivalents of BBr 3 was reduced.Tetrahydroquinoline 3f was stirred with three equivalents of BBr 3 in DCM for 1 h.Work up and purification with column chromatography resulted in a 38% yield of compound 13f (Scheme 6, Table 2).The 1 H NMR spectrum of compound 13f still showed the presence of a methoxy group at δ 3.69 ppm and only one hydroxyl group was evident at δ 7.78 ppm.To confirm the structure of this compound a NOE experiment was performed, which showed a correlation between the methoxy and the H6 and H8 protons while the hydroxy signal only correlated to the H6 proton.This indicated that demethylation had occurred at the C5 position, thus giving 5-hydroxy-7methoxytetrahydroquinoline 13f.
BBr3 in DCM at room temperature for 1 h (Scheme 5).However, the product of this reac-tion could not be purified even after several a empts.This could possibly be due to the presence of three hydroxyl groups in compound 11 rendering the compound significantly unstable.Therefore, it was decided to demethylate tetrahydroquinoline 3f, which has only two methoxy groups, with five equivalents of BBr3 in DCM at room temperature for 1 h (Scheme 5).However, purification of compound 11f proved to be similarly problematic and pure spectra of the product could not be obtained.In an a empt to improve the purity of the hydroxyl compounds, the number of equivalents of BBr3 was reduced.Tetrahydroquinoline 3f was stirred with three equivalents of BBr3 in DCM for 1 h.Work up and purification with column chromatography resulted in a 38% yield of compound 13f (Scheme 6, Table 2).The 1 H NMR spectrum of compound 13f still showed the presence of a methoxy group at δ 3.69 ppm and only one hydroxyl group was evident at δ 7.78 ppm.To confirm the structure of this compound a NOE experiment was performed, which showed a correlation between the methoxy and the H6 and H8 protons while the hydroxy signal only correlated to the H6 proton.This indicated that demethylation had occurred at the C5 position, thus giving 5-hydroxy-7methoxytetrahydroquinoline 13f.Scheme 6. Demethylation of 3f using BBr3 (3 equiv.) in DCM.By the application of this strategy three monohydroxyl analogues were synthesized (Table 2).

Biological Activity of 3,4-Diaryl-1,2,3,4-Tetrahydroquinolines
A selection of the synthesized methoxy tetrahydroquinoline compounds were screened for their anticancer activity (Figure 4).The compounds selected for this screen gave a valuable insight into the structureactivity relationship of functionality of both aryl groups of the tetrahydroquinolines (Fig ure 4, Table S1).In vitro growth inhibition assays were performed at a fixed concentration of 25 µM in a range of cancer cell lines, including H460 lung carcinoma, DU145 prostate carcinoma, A-431 skin (epidermoid) carcinoma, HT-29 colon adenocarcinoma, and MCF7 breast adenocarcinoma.
The potent activity of the 1,2,3,4-tetrahydroquinolines in Figure 5 shows that incor The compounds selected for this screen gave a valuable insight into the structure-activity relationship of functionality of both aryl groups of the tetrahydroquinolines (Figure 4, Table S1).In vitro growth inhibition assays were performed at a fixed concentration of The potent activity of the 1,2,3,4-tetrahydroquinolines in Figure 5 shows that incorporating an aryl group into position 4 of the quinoline structure dramatically increased the antiproliferative effect by up to 90% compared to the parent compound 4. Compound 3c showed the greatest antiproliferative effect with an unsubstituted phenyl ring at the 4 position.Adding a substituent to this ring led to a total loss of activity in the DU145 prostate carcinoma cell lines.However, compounds 3b and 3e maintained a fairly good level of activity in the H460 lung carcinoma cell line, (30.7% and 32.5%, respectively) and in the MCF7 breast adenocarcinoma cell lines (25.4% and 23.9%, respectively).It is important to note that all the tetrahydroquinoline analogues tested (3b, 3c, and 3e) were still more effective than their original quinoline structures 4.  S1).In vitro growth inhibition assays were performed at a fixed concentration of 25 µM in a range of cancer cell lines, including H460 lung carcinoma, DU145 prostate carcinoma, A-431 skin (epidermoid) carcinoma, HT-29 colon adenocarcinoma, and MCF7 breast adenocarcinoma.
The potent activity of the 1,2,3,4-tetrahydroquinolines in Figure 5 shows that incorporating an aryl group into position 4 of the quinoline structure dramatically increased the antiproliferative effect by up to 90% compared to the parent compound 4. Compound 3c showed the greatest antiproliferative effect with an unsubstituted phenyl ring at the 4 position.Adding a substituent to this ring led to a total loss of activity in the DU145 prostate carcinoma cell lines.However, compounds 3b and 3e maintained a fairly good level of activity in the H460 lung carcinoma cell line, (30.7% and 32.5%, respectively) and in the MCF7 breast adenocarcinoma cell lines (25.4% and 23.9%, respectively).It is important to note that all the tetrahydroquinoline analogues tested (3b, 3c, and 3e) were still more effective than their original quinoline structures 4. Due to its notable biological activity, the lead compound 3c was further tested across various concentrations in these cell lines.The compound demonstrated effective activity in the H460 lung carcinoma, A-431 skin carcinoma, and HT-29 colon adenocarcinoma cells, with IC 50 values of 4.9 ± 0.7, 2.0 ± 0.9, and 4.4 ± 1.3 µM, respectively.Therefore, compound 3c showed its most potent inhibition effects against skin carcinoma cells.Although the IC 50 values were marginally higher in the DU145 prostate carcinoma and MCF7 breast adenocarcinoma cells, they still indicated substantial levels of activity, with values at 12.0 ± 1.6 and 14.6 ± 3.9 µM, respectively.

• Analytical data
Melting points (uncorrected) were measured using a Mel-Temp melting point apparatus.The Microanalysis Unit of the University of Otago, New Zealand, performed microanalyses.Infrared spectra were recorded as Nujol mulls on a Perkin-Elmer 298 (Beaconsfield, UK) or a Perkin-Elmer 580B spectrometer.Ultraviolet-visible spectra were recorded in methanol (unless otherwise stated) on a Hitachi UV-3200 spectrometer (Hitachinaka, Japan). 1 H and 13 C NMR spectra were obtained for the designated solvents on a Bruker AC300F (300 MHz) spectrometer (Bruker Pty Ltd., Preston, NSW, Australia). 1 H NMR data were recorded as follows: chemical shift measured in parts per million (ppm) downfield from TMS (δ), multiplicity, observed coupling constant (J) in Hertz (Hz), proton count.Multiplicities are reported as singlet (s), broad singlet (bs), doublet (d), triplet (t), quartet (q), quintet (quin), and multiplet (m). 13C NMR chemical shifts are reported in ppm downfield from TMS and identifiable carbons are given.The EI and ES mass spectra were recorded on an AEI MS 12 mass spectrometer (Washington, D.C., USA) at 70 eV ionizing potential and 8000 V accelerating voltage with an ion source temperature of 210 • C. Kieselgel 60H (Merck, Rahway, NJ, USA, Art 7736) was employed for flash chromatography and thin layer chromatography (TLC) was performed on DC Aluminium Foil Kieselgel F 254 (Merck, Art 5554).Solvents and reagents were purified by literature methods.Petroleum ether refers to the hydrocarbon fraction of boiling range 60-80 • C. Compounds were detected by short and long ultraviolet light and with iodine vapor.

•
In vitro growth inhibition assays In vitro growth inhibition assays were performed in triplicate using the Alamar Blue assay.Briefly, cells in logarithmic phase growth were seeded onto 96-well plates at densities as follows: MCF A Multidrop 384 (Thermo Scientific, Waltham, MA, USA) was used and cells were allowed to adhere.After 24 h of incubation, test compounds, positive control (25 µM thonzonium bromide), and vehicle only (DMSO) controls were added to duplicate wells using a Hamilton Nimbus robotic platform.After 72 h of drug exposure, metabolic activity was detected by addition of Alamar Blue reagent and determined by measurement of fluorescence intensity (excitation 555 nm, emission 585 nm) using a SpectraMax M5 (Molecular Devices, San Jose, CA, USA) plate reader.Percentage of cell viability was determined at a fixed drug concentration of 25 µM.A value of 0% is indicative of total cell growth inhibition.
Compounds showing appreciable percentage growth inhibition underwent further dose response analysis allowing for the calculation of an IC 50 value.This value is the drug concentration at which cell growth is inhibited to its half maximal value.

General Procedures
To a solution of azaisoflavone 4 (2.7 mmol) in anhydrous THF (10 mL) was slowly added phenylmagnesium bromide (5.5 mmol) in an atmosphere of nitrogen.The reaction mixture was heated at reflux for 5 h, then quenched with NH 4 Cl solution (25 mL, 20%) and extracted with ethyl acetate (2 × 50 mL).The combined organic layers were washed with brine (25 mL), dried over anhydrous Na 2 SO 4 , and concentrated under vacuum.The crude product was then purified using column chromatography (8% ethyl acetate in n-hexane) to give 3,4-diaryl-azaisoflavan-4-ol 5 as a white solid.

Conclusions
Attempts to synthesize 3,4-disubstituted 1,2-dihydroquinolines by a Grignard reaction approach were unsuccessful and resulted in the formation of the more stable quinoline moiety.
However, hydrogenation of N-protected 1,2-dihydroquinolines resulted in the formation of the 1,2,3,4-tetrahydroquinolines with cis configuration.Demethylation of these 5,7-dimethoxy tetrahydroquinolines in the presence of two equivalents of BBr 3 for each methoxy group resulted in the formation of the corresponding hydroxyl compounds, though the compounds were seen to have a low stability.The use of one equivalent of BBr 3 for each methoxy group resulted in the formation of 5-hydroxy analogues in good yields.
The biological activity of compound 3c offers opportunities for further SAR studies on various C4 and aromatic substituents of our 1,2,3,4-tetrahydroquinoline scaffold, which could result in lead compounds with improved anticancer potency.Testing of the monodeprotected analogues 13 would also provide insight into the role of the methoxy groups in the activity of our compounds.

Supplementary Materials:
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/molecules29174273/s1, Figure S1: Packing of molecules of 3c looking down a-axis; Table S1: Cell viability of cancer cell lines after treatment with tetrahydroquinolines; Figure S2:

Scheme 3 .
Scheme 3. Synthesis of 3,4-diaryl-1,2,3,4-tetrahydroquinolines.The expected cis stereochemistry of the compound 10 was established on the basis of the coupling constant of J = 3.0 between H3 and H4.Moreover, NOE correlations between the H3 and H4 protons were evident in the 2D NMR experiment, further confirming the cis configuration (Figures 2 and S11) )

Scheme 3 .
Scheme 3. Synthesis of 3,4-diaryl-1,2,3,4-tetrahydroquinolines.The expected cis stereochemistry of the compound 10 was established on the basis of the coupling constant of J = 3.0 between H3 and H4.Moreover, NOE correlations between the H3 and H4 protons were evident in the 2D NMR experiment, further confirming the cis configuration (Figures 2 and S11) )

Figure 2 .
Figure 2. NOE correlations of compound 10c.The deprotection of compound 10 was carried out by heating at reflux in 10% NaOH in EtOH for 6 h to give the desired 5,7-dimethoxy-3,4-diaryl-1,2,3,4-tetrahydroquinolines 3 in 82-90% yield.The structure of 3 was confirmed by the lack of protons correlating to the -COOEt group in the 1 H NMR spectrum and the presence of a NH proton as a singlet at δ 4.14 ppm.The coupling constant between H3 and H4 protons (J = 3.0 Hz) indicated the cis configuration had been retained, and this was further confirmed by X-ray crystallographic analysis.The ORTEP diagram of compound 3c is shown in Figure3.Packing of molecules was dominated by CH3-π and slipped π-π interactions (FigureS1).

Figure 2 .Figure 3 .
Figure 2. NOE correlations of compound 10c.The deprotection of compound 10 was carried out by heating at reflux in 10% NaOH in EtOH for 6 h to give the desired 5,7-dimethoxy-3,4-diaryl-1,2,3,4-tetrahydroquinolines 3 in 82-90% yield.The structure of 3 was confirmed by the lack of protons correlating to the -COOEt group in the 1 H NMR spectrum and the presence of a NH proton as a singlet at δ 4.14 ppm.The coupling constant between H3 and H4 protons (J = 3.0 Hz) indicated the cis configuration had been retained, and this was further confirmed by X-ray crystallographic analysis.The ORTEP diagram of compound 3c is shown in Figure 3. Packing of molecules was dominated by CH 3 -π and slipped π-π interactions (Figure S1).Molecules 2024, 29, x FOR PEER REVIEW 6 of 23

Figure 4 .
Figure 4. Lead anticancer screen compounds.The compounds selected for this screen gave a valuable insight into the structureactivity relationship of functionality of both aryl groups of the tetrahydroquinolines (Figure 4, TableS1).In vitro growth inhibition assays were performed at a fixed concentration of 25 µM in a range of cancer cell lines, including H460 lung carcinoma, DU145 prostate carcinoma, A-431 skin (epidermoid) carcinoma, HT-29 colon adenocarcinoma, and MCF7 breast adenocarcinoma.The potent activity of the 1,2,3,4-tetrahydroquinolines in Figure5shows that incorporating an aryl group into position 4 of the quinoline structure dramatically increased the antiproliferative effect by up to 90% compared to the parent compound 4. Compound 3c showed the greatest antiproliferative effect with an unsubstituted phenyl ring at the 4 position.Adding a substituent to this ring led to a total loss of activity in the DU145 prostate carcinoma cell lines.However, compounds 3b and 3e maintained a fairly good level of activity in the H460 lung carcinoma cell line, (30.7% and 32.5%, respectively) and in the MCF7 breast adenocarcinoma cell lines (25.4% and 23.9%, respectively).It is important to note that all the tetrahydroquinoline analogues tested (3b, 3c, and 3e) were still more effective than their original quinoline structures 4.

Figure 5 .Figure 5 .
Figure 5. Bar graph representation of cancer cell line viability after treatment with tetrahydroquinolines.