New Derivatives of 3,4-Dihydroisoquinoline-3-carboxylic Acid with Free-Radical Scavenging, d-Amino Acid Oxidase, Acetylcholinesterase and Butyrylcholinesterase Inhibitory Activity

A series of 3,4-dihydroisoquinoline-3-carboxylic acid derivatives were synthesised and tested for their free-radical scavenging activity using 2,2-diphenyl-1-picrylhydrazyl radical (DPPH·), 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) radical (ABTS·+), superoxide anion radical (O2·−) and nitric oxide radical (·NO) assays. We also studied d-amino acid oxidase (DAAO), acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) inhibitory activity. Almost each of newly synthesised compounds exhibited radical scavenging capabilities. Moreover, several compounds showed moderate inhibitory activities against DAAO, AChE and BuChE. Compounds with significant free-radical scavenging activity may be potential candidates for therapeutics used in oxidative-stress-related diseases.

Our main interests are concerned with the determination of antioxidative properties due to the fact that phenolic compounds often present potent free radical scavenging activities. Several known, potent free-radical scavengers such as flavonols (e.g., quercetin, kaempferol), flavones (e.g., luteolin, apigenin), flavanols (e.g., catechin) and isoflavones (e.g., genistein) contain phenols and polyphenolic moiety [12]. Previous studies have shown that polyphenolic structures are more potent free-radical scavengers than those with only one hydroxyl group [13]. Also, the ortho location of two hydroxyl groups increases the antioxidative properties [13,14]. The free-radical scavenging potential of polyphenolic compounds is associated with their reducing (hydrogen-or electron-donating) activities [12,15].
Oxidative stress is a process that arises when free radicals and oxidants are produced in excess and cells cannot destroy them properly. In other words, an imbalance between the formation and neutralisation of free radicals can occur, owing to a depletion of antioxidants or the accumulation of reactive oxygen species (ROS), and can result in the appearance of oxidative stress [16][17][18][19]. The latter may lead to the change of cell components such as proteins, lipids, lipoproteins, nucleic acids and so forth [18][19][20].
Oxidative stress plays a crucial role in the development of various diseases such as cardiovascular disorders; atherosclerosis; hypertension or heart failure; diabetes mellitus; cancer; rheumatoid arthritis; ocular diseases; neurodegenerative diseases such as Parkinson's and Alzheimer's disease (AD), as well as psychic impairments such as schizophrenia and is linked to aging [15][16][17][18].

Chemistry
Derivatives of the isoquinoline 5 have been obtained in a multistep synthesis using the Bischler-Napieralski reaction [21] as a cyclization step (Scheme 1, Table 1). Benzyl alcohols 1 can be easily obtained by known methods from commercially available hydroxy-or methoxy-substituted benzaldehydes. We have used benzyl protection for hydroxyl groups. Some other protective groups like allyl, carbonate or acetal groups were not stable in the cyclization step. Reduction of the carbonyl group and chlorination with thionyl chloride gave benzyl chlorides 2. Alkylation of dibenzyl formamidomalonate with 2 gave Bischler-Napieralski reaction substrates 3. Cyclization was conducted under standard conditions using POCl3 in acetonitrile at 70-95 °C. Higher reaction temperatures resulted in extensive formation of polymers and lowered the yield. Two isomers can form in the case of 2,6-unsubstituted derivatives 3b, d, g and h (R 1 = H). Usually only one product was obtained. For example cyclization of compound 3b, gives predominantly 8-benzyloxy-7-bromo-3,4-dihydroisoquinoline instead of the corresponding 6-benzyloxy-7-bromo derivative. Apparently the C2 carbon atom in formamidomalonate 3b is more nucleophilic than C6. The stereochemistry of 4d, g and h has been determined on the basis of HMBC correlation spectra, and derivative 4b showed a vicinal coupling constant between C5-H and C6-H. In the last step of the synthesis, all benzyl groups were removed with BBr3 [22] and decarboxylation took place gently to give acid 5 (Table 2). We have also used hydrogenation on Pd/C with 1,4-cyclohexadiene as hydrogen donor. This is very mild method which selectively removes benzyl groups in the presence of C=N bond. However formation of isoquinoline derivatives due to aromatization of the ring was noticed. More difficult was removal of methyl groups with BBr3. In the case of trimethoxy-substituted dihydroisoquinoline we have obtained mixture of isomers, where 5i was the major one.    (17) Cl OH H OH 5l (8) Cl OH OH H a Reference [9].
The most problematic was purification of polyhydroxylated derivatives of 3,4-dihydroisoquinoline-3-carboxylic acid 5. The products are sensitive to air and bases giving aromatic derivatives of isoquinolines. Due to the very low solubility in organic solvents and high polarity of the products, we had to use chromatography on ion exchange and polyamide columns. In a similar way, we obtained 4,5-dihydro-3H-2-benzazepin-3-carboxylic acid 9 from 4-(2-iodoethyl)-1,2-dimethoxybenzene (Scheme 2). Optically active 3-methyl derivative of 3,4-dihydroisoquinoline-3-carboxylic acid 13 was obtained using L-methyl-DOPA by formylation with formyloxyacetonitrile and Bischler-Napieralski cyclization. We also synthesised a β-lactam derivative of dihydroisoquinoline 15 by the Bose method using [2+2] cycloaddition with chloroketene generated in situ from chloroacetyl chloride [23]. This type of reaction is usually highly stereoselective. Examination of the coupling constant of C1-H (1.7 Hz) suggests that we have obtained trans stereoisomer. The benzyl groups in 14 were removed by hydrogenolysis to give β-lactam 15 (Scheme 3).

Free-Radical Scavenging Activity
All the synthesised derivatives were assayed in vitro for their antioxidant and free-radical scavenging activity. In these assays, the compounds were tested for their ability to scavenge DPPH · , ABTS ·+ , O2 ·− and · NO. The EC50 values are summarized in Table 3. In general, almost all of the described compounds were potent scavengers on at least one free radical used in our study. Lower EC50 value indicates higher radical-scavenging activity. Free-radical scavenging activities of synthesised derivatives were compared to the properties shown by naturally obtained compound 5e.
Compound 5e, was the most active scavenger on ABTS ·+ . The EC50 value of the above-mentioned compound was approximately one quarter of that of ascorbic acid. Also, 5l, 15, 5g, 9, 5f, 5c, 13, 5h, 5d, 5i and 5k showed very potent activities. The presence of two hydroxyl groups at the 6-and 7-positions (5e, 5l, 5g, 13, 5h, 5d) and respectively at 7-and 8-positions in 15 and 9, resulted in significant ABTS ·+ scavenging activity of the compounds. Halogen atoms (chlorine or bromine) at the 5-and 8-positions did not significantly influence the activity in comparison to 5e. ABTS ·+ was scavenged by compounds 5e, 5l, 15, 5g, 9, 5f, 5c, 13 and ascorbic acid in a concentration-dependent manner ( Figure 2).  We obtained nine compounds (15, 5d, 5l, 9, 5f, 5g, 5i, 5e, 5k) with superoxide anion radical (O2 ·− ) scavenging activities better than the standard (ascorbic acid). Compounds 15, 5d, 5l, 9, and 5f presented very high superoxide anion radical scavenging activities. Their EC50 values were 22-times lower (15 and 5d) and more than 10-times lower (5l, 9, 5f) than the EC50 value for ascorbic acid. The chloroazetidinone moiety at [2,1a]-position in compound 15 and methyl group at the 4-position improved the activity in comparison to compound 5e. Also, the presence of two hydroxyl groups at the 6-and 7-positions and iodine at the 8-position (5d) or chlorine at the 5-position (5l) caused significant O2 ·− scavenging activity. Ortho position of two hydroxyl groups in the tested compounds contributed to significant superoxide anion radical scavenging activity. A bromine atom at the 8-position (compound 5g) enhanced O2 ·− scavenging activity of the compound, in comparison to a chlorine atom at the same position (compound 5h). The EC50 value for compound 5g was about sevenfold lower than the EC50 value for ascorbic acid. This data suggests that substitution of phenolic compounds with a halogen atom (I > Br > Cl) stabilizes the superoxide anion radical. Compound 9, with a dihydrobenzoazepine ring, showed 6-times greater O2 ·− scavenging activity in comparison to the similar compound, 5e, with a dihydropyridine ring. Tested compounds were able to scavenge superoxide anion radicals in a concentration-dependent manner ( Figure 3). Compound 13 was the most potent · NO scavenger. Its EC50 value was lower than the standard (Trolox). All of the other tested compounds showed rather moderate scavenging activities on the nitric oxide radical. Comparing the results of all above assays, in general, compounds with two hydroxyl groups (5c-l, 9, 13, 15) were more potent free-radical scavengers than phenolic compounds (5a, 5b). Also, an ortho position of hydroxyl groups determined more significant free radical scavenging activity than the meta position.

DAAO Inhibitory Activity
Similarly, all newly synthesised compounds were screened for DAAO inhibitory activity. Compound 9 was the most active inhibitor of DAAO (Table 3). It showed a fourfold lower IC50 value than the standard benzoic acid. Other compounds, including 5l, 5c, 5h, 5g, 15 and 5f, showed comparable or better DAAO inhibitory activities than the standard. The presence of two hydroxyl groups at the meta position in the benzene ring (5c) as well as the ortho location of hydroxyl groups with additional chlorine or bromine atom at 5-or 8-positions (5g, 5h, 5l, 5f) exhibited good activity. Comparison of the compound with two hydroxyl groups at 6-and 7-positions (5e) and compounds with an additional halogen atom at the 8-position (5g, 5h, 5d) showed that halogen atoms enhanced DAAO inhibition. Moreover, the more electronegative the substituent (Cl > Br > I) at the 8-position, the stronger the resulting DAAO inhibitor. Compounds 5a-5l and 9 are obtained as racemic mixtures. Biological activities of separated enantiomers may be higher (even two times). However, the activity would remain in milimolar concentration range. Synthesised derivatives inhibited DAAO in a concentration-dependent manner ( Figure 4).

AChE Inhibitory Activity
The effect of the synthesised compounds on AChE activity was evaluated, using a selective AChE inhibitor, galanthamine hydrobromide as a standard. Only a few of the synthesised compounds (5j, 5f, 5e, 5k) showed rather weak activity against AChE. None of the tested compounds were more potent than the standard. The results are presented in Table 3.
NMR spectra were recorded using Varian 400 MR (400 MHz) spectrometer. Chemical shifts of aqueous solutions were referenced against solvent signal (4.75 ppm). Most of the 13 C-NMR chemical shifts of deprotected phenols 5 could not be determined due to low solubility and deuteriation in the ring. IR spectra were recorded on Shimadzu IRAffinity-1S spectrophotometer using single reflection ATR (ZnSe). Mass spectra were obtained using LTQ Orbitrap Velos (Thermo Scientific) and QToF Premier (Waters), HR and LR respectively. The radical scavenging activity and enzymes inhibitory activity were assessed by UV-VIS spectrophotometer (FLUOstar Omega BMG LABTECH microplate reader). Copies of 1 H and 13 C-NMR spectra are available in the Supplementary Information file.

General Procedure for the Synthesis of Benzyl Chlorides 2
Pyridine (0.36 mL) and thionyl chloride (0.37 mL) were added dropwise to a solution of protected benzyl alcohol 1 (4.78 mmol) in CH2Cl2 (15 mL). The resultant mixture was then stirred at rt for 2.5 h. The reaction was quenched with 10 mL of water and stirred for 0.5 h to destroy excess of thionyl chloride. The organic layer was removed, and the aqueous layer was extracted twice with CH2Cl2 (3 × 15 mL). The combined organic layers were dried over MgSO4 and evaporated. The products were used in the next step without purification.

General Procedure for Alkylation of Dibenzyl Formamidomalonate 3a-3l
Dibenzyl formamidomalonate (315 mg, 0.96 mmol) followed by K2CO3 (1.76 g, 12.75 mmol) and KI (478 mg, 2.88 mmol) were added to a solution of respective benzyl chlorides (0.96 mmol) in acetone (20 mL). The reaction mixture was stirred under reflux (oil bath, 60 °C) for 20 h. When complete, the reaction mixture was cooled and filtered through Celite. The solvent was evaporated to give a solid.   POCl3 (0.23 mL) was added to a stirred solution of the respective dibenzyl formamidomalonate 3a-3l (0.83 mmol) in acetonitrile (10 mL) and the mixture was heated under various conditions depending on the compound. The solvent was evaporated and residue was made alkaline with aqueous NaHCO3 (5 mL). The mixture was extracted with CH2Cl2 (3 × 10 mL). The combined organic layers were dried with MgSO4 and concentrated to give the product.

DPPH Radical Scavenging Assay
The radical-scavenging activity of tested compounds against DPPH · was determined by Williams' method [25]. Tested compounds dissolved in water or methanol (30 μL) were added at various concentrations (to final concentration of 2,4,8,16,32,64,128,256, 512 μg/mL) to freshly prepared 0.22 mM DPPH · in methanol (120 μL). The absorbance was measured against the control in the microplate reader at 517 nm after 30 min of incubation at room temperature (rt). The control contained all reagents without the tested compound. The corresponding blank reading was also taken and the remaining DPPH · was calculated. Each sample was replicated three times. Ascorbic acid was used as a standard.

ABTS Radical Scavenging Assay
The radical-scavenging activity of tested compounds against ABTS ·+ and the generation of the ABTS ·+ radical was performed according to methods described previously [26,27]. Briefly, an ABTS ·+ solution was prepared by mixing 2 mM ABTS (50 mL) with 70 mM potassium persulphate (200 μL), both in deionized water, and used after 2 h of being kept in the dark at rt. The ABTS ·+ solution (40 μL) was added to 0.1 M phosphate buffer pH 7.4 (60 μL) and tested compounds (25 μL) were dissolved in water or methanol at various concentrations (to final concentration of 2,4,8,16,32,64,128,256, 512 μg/mL). The absorbance was measured against the control in the microplate reader at 734 nm after 30 min of incubation at rt. The control contained all reagents without the tested compound. The corresponding blank reading was also taken and the remaining ABTS ·+ was calculated. Each sample was replicated three times. Ascorbic acid was used as a standard.

Superoxide Anion Radical Scavenging Assay
Superoxide scavenging assay was based on inhibition of the formazan dye formation. O2 ·− was generated in alkaline DMSO by the addition of sodium hydroxide to air-saturated pure DMSO [28]. The generated superoxide reacts with NBT to give coloured diformazan. To the reaction mixture containing NBT (10 μL; 1 mg/mL solution in DMSO) and the various concentration of scavenger (30 μL; to final concentration of 2, 4, 8, 16, 32, 64, 128, 256, 512 μg/mL), alkaline DMSO (100 μL) was added to give a final volume of 140 μL per well in a microplate. The absorbance was measured in the microplate reader at 560 nm after 30 min of incubation at rt [29]. The same procedure was repeated for the control in which DMSO was added instead of scavenger solution. Each sample was replicated three times and the corresponding blank reading was also taken. Ascorbic acid was used as a standard.

Nitric Oxide Radical Scavenging Assay
Nitric oxide was generated from sodium nitroprusside and measured by the Griess reaction [26]. Scavengers of · NO compete with oxygen leading to reduced production of nitric oxide. 10 mM sodium nitroprusside in 0.1 M phosphate buffer pH 7.4 (30 μL) was mixed with tested compounds dissolved in water or methanol (25 μL) at various concentrations (to final concentration of 2, 4, 8, 16, 32, 64, 128, 256, 512 μg/mL) and incubated at 25 °C for 5 h. After incubation, Griess reagent (50 μL) was added. The absorbance was measured in the microplate reader at 546 nm against a control (without tested compounds) treated in the same way with Griess reagent. The corresponding blank reading was also taken. Each sample was replicated three times. Trolox was used as a standard.

Effective Concentration of Scavengers (EC50)
The percentage of radical scavenging was calculated using Equation (1) (DPPH · , ABTS ·+ and · NO scavenging assays). The absorbance for the control is Ac and the absorbance in the presence of the compounds or other scavenger is As: The percentage of O2 ·− scavenging was calculated using Equation (2). The absorbance for the control is Ac and the absorbance in the presence of the compounds or other scavenger is As: The amount of tested compound needed to scavenge free radicals by 50%, EC50 (effective concentration), was calculated by the linear regression between scavenger concentration and percentage of scavenging and expressed as μM.

DAAO Inhibitory Activity
Inhibition of DAAO was based on the determination of the α-keto acid obtained from the reaction between a DAAO and D-Ala according to the described method [30]. The method was adapted to microtitre plates and a total volume of 210 μL per well. Inhibitor solution in DMSO (15 μL), 5 mM D-Ala in 0.2 M Tris-HCl buffer pH 8.2 (121 μL) and 10 U/mL DAAO solution in 0.2 M Tris-HCl buffer pH 8.2 (5 μL) were mixed and incubated at 37 °C for 30 min. After incubation, 1 mM 2,4-dinitro phenylhydrazine dissolved in 1 M HCl (14 μL) was added to the well, mixed and further incubated at 37 °C for 10 min. Next, 1.5 M NaOH (55 μL) was added and incubated for 10 min. The absorbance was read at 445 nm against a blank sample consisting the same assay mixture but without the substrate (using Tris-HCl buffer instead of D-Ala). The positive control of the enzyme activity consisted of DMSO (15 μL) instead of inhibitor solution and the analogous blank contained Tris-HCl buffer instead of D-Ala. Sample, control and related blanks were carried out under the same condition. Benzoic acid was used as a standard. Eight different solutions of inhibitor were examined: 0.0005, 0.0025, 0.005, 0.00625, 0.0125, 0.025, 0.0375 and 0.05 M. For each of these inhibitor concentrations percentage of inhibition was calculated by using Equation (3). The absorbance for the control is Ac and the absorbance in the presence of the compounds or other inhibitor is Ai: The inhibitor concentration needed to inhibit DAAO by 50% (IC50) was calculated by the linear regression between inhibitor concentration and percentage of inhibition. Each sample was replicated three times.

AChE Inhibitory Activity
The assay was based on Ellman's method [31] with modifications. AChE activity was measured by the determination of a yellow colour produced from acetylthiocholine iodide when it reacted with the dithiobisnitrobenzoate ion. The assay was performed using microtitre plates and a total volume of 100 µL per well. In a 96-well plate, 0.1 M pH 7.4 phosphate buffer (32 µL), a serially diluted solution of tested compounds (5 µL), 0.59 U/mL AChE solution (4 µL) and 0.3 mM DTNB (42 µL) were mixed and pre-incubated at 25 °C for 2 min. Then, the substrate, 0.4mM acetylthiocholine iodide (17 µL), was added to the well and mixed. The enzymatic reaction was followed for 10 min, with a measurement every 12 s. Changes in the absorbance at 412 nm were detected in the microplate reader. The absorbance was measured against a blank sample containing the same assay mixture but without an enzyme (using phosphate buffer instead of AChE). The positive control of the enzyme activity consisted of DMSO (1.5 μL) instead of inhibitor solution. The sample, control and related blanks were measured under the same conditions. Galanthamine hydrobromide was used as a standard. Each sample was replicated three times. The concentration of the compounds that caused 50% inhibition of AChE activity (IC50) was calculated using linear regression analysis. At least three concentrations of tested compounds were assayed. For each of these inhibitor concentrations, the percentage of inhibition was calculated using Equation (4). The difference in absorbance for the inhibitor at t = 10 min and t = 0 is denoted ΔAi. The difference in absorbance for the control at t = 10 min and t = 0 is denoted ΔAc:

BuChE Inhibitory Activity
The assay was based on Ellman's method [31] with modifications. BuChE activity was measured by the determination of the yellow colour produced from butyrylthiocholine iodide when it reacted with the dithiobisnitrobenzoate ion. The assay was performed using microtitre plates and a total volume of 100 µL per well. In a 96-well plate, 0.1 M pH 7.4 phosphate buffer (35 µL), a serially diluted solution of tested compounds (5 µL), 1.485 U/mL BuChE solution (5 µL) and 0.3 mM DTNB (42 µL) were mixed and pre-incubated at 25 °C for 2 min. Then, the substrate, 0.6 mM butyrylthiocholine iodide (13 µL), was added to the well and mixed. The enzymatic reaction was followed for 10 min, with a measurement every 12 s. Changes in absorbance at 412 nm were detected in the microplate reader. The absorbance was measured against a blank sample containing the same assay mixture, but without an enzyme (using phosphate buffer instead of BuChE). The positive control for enzyme activity contained DMSO (1.5 μL) instead of inhibitor solution. The sample, control and related blanks were measured under the same conditions. Galanthamine hydrobromide was used as a standard. Each sample was replicated three times. The concentration of the compounds that caused 50% inhibition of BuChE activity (IC50) was calculated using linear regression analysis. At least three concentrations of the tested compounds were assayed. For each of these inhibitor concentrations, the percentage of inhibition was calculated using Equation (4).
In conclusion, the potent free-radical scavengers obtained in the study may be potential candidates for therapeutics used in oxidative-stress-related diseases. The synthesised inhibitors of evaluated enzymes may deserve attention in the design of new antipsychotic drugs. Despite rather low inhibitory activities against DAAO, AChE and BuChE, it seems that compounds with potent free-radical scavenging properties may play a supporting role in the treatment of neurodegenerative and psychiatric diseases.