Synthesis of Indolostilbenes via FeCl3-promoted Oxidative Cyclisation and their Biological Effects on NG108-15 Cell Viability and H2O2-induced Cytotoxicity

A convenient and simple radical cation cyclisation of 3,5-dimethoxystilbene was developed using the commercially available FeCl3 under mild condition. It enabled the construction of a new class of indolostilbenes (i.e., indole-stilbene hybrid). Various parameters were investigated to obtain better yields (more than 42%) compared with the previously reported. The synthesised indolostilbenes were characterised, and their mechanism of formation was discussed. The synthesised compounds were submitted for biological assay on NG108-15 cell viability and H2O2-induced cytotoxicity. The result showed that two indolostilbenes have promising protective activity against H2O2.


INTRODUCTION
Neurodegenerative diseases such as dementia (Parkinson's, Huntington's and Alzheimer's diseases) and stroke result in significant disabilities and dependency among the elders worldwide. 1 The progression of these neurodegenerative diseases is attributed to various factors. For instance, these diseases can be initiated by oxidative stress-induced cell damage corresponding to the overproduction of reactive oxygen species (ROS) such as hydrogen peroxide (H 2 O 2 ). H 2 O 2 acts as the precursor of ROS, which, in excess, induces oxidative stress that leads to neuronal cell apoptosis. 2,3 The overproduction of ROS promotes mitochondrial dysfunction leading to protein, lipid and DNA damage. 2,3 Based on these observations, it may be beneficial to study the capability of newer therapeutic agents in mitigating H 2 O 2induced apoptosis in the treatment of neurodegenerative diseases. 2,3 Therefore, chemists have been actively synthesising new classes of active compounds that possess H 2 O 2 protective properties to combat neurodegenerative diseases.
Stilbenes are the family of bioactive compounds found in plants. 4 Transresveratrol, o-carboxamido stilbene, piceid, piceatannol, astringin, pterostilbene and oligomers (viniferin and hopeaphenol) are the groups of compounds from this family ( Figure 1). 4 Stilbenes have received considerable attention because of their potent health-promoting properties, such as preventing cancers, cardiovascular and neurodegenerative diseases. [5][6][7][8][9][10][11][12][13] Oligostilbenes such as pallidol, balanocarpol and (+)-α-viniferin also show a variety of biological activities ( Figure 1). 14-17 Indoles have gained much attention from medical and synthetic chemists due to its ability to act as a free radical scavenger and antioxidant property. This resulted in the synthesis of a new class of oligostilbenes, namely, indolostilbene. 18 Indolostilbene is the polymeric form of stilbene that has the indole ring. 19 Previous study has shown that the indole ring in the molecule is the reactive site for oxidation. This is due to its highly stable resonance and relatively low activation energy gap for free radical reactions. 20,21 Thus, in this paper, we discussed the synthesis of a new group of indolostilbenes and their protective activity against H 2 O 2 -induced cell death in NG108-15 cells. To our knowledge, this is the first report on optimising and synthesising these new indolostilbenes and their biological activity.

General
Unless otherwise noted, materials were purchased from Sigma-Aldrich Co., Acros Organics and Merck Chemical Co., and used without purification. Dimethylformamide (DMF) was dried over 4Å molecular sieves (Merck) before use. Column chromatography was carried out using silica gel from Merck (40-63 µm). For thin-layer chromatography, Merck TLC aluminium sheets (silica gel 60 F 254 ) were used. All reactions were carried out in heat-dried glassware under a dry nitrogen atmosphere unless otherwise stated. All liquid transfer was conducted using standard syringe or cannula techniques. All spectral data were obtained on the following instruments: infrared spectra were recorded on a Perkin Elmer FTIR Spectrum RX-1 spectrometer at wavenumber from 4000-400 cm -1 . Nuclear magnetic resonance spectra were obtained on a Bruker AVN 500 MHz spectrometer from Bruker Bioscience, Billerica, MA, United States) and the spectra were reported in ppm units on the δ scale, relative to CDCl 3 . The proton, 1 H resonance was recorded at 500 MHz and 13 C at 125 MHz. The coupling constants are given in Hz. The mass spectra were measured using Agilent 6530 Accurate-Mass QTOF LC/MS system. The HPLC analysis was done using Waters auto purification system (Waters 2767 and 2996).

Procedure for synthesis of styrene 2, protected amide 4 and stilbene 5
The procedure for the preparation of styrene 2, protected amide 4 and stilbene 5 was reported previously by our group. 6 The detailed procedure and spectroscopic data are attached in the supplementary information.

HPLC Analysis of Crude Reaction
The HPLC analysis was done using Waters auto purification system equipped with a sample manager (Waters 2767), a binary pump, an automatic injector and a photodiode array detector (190-600 nm, Waters 2996). The separation was carried out on a Waters ® X-Bridge C18 column (250 × 4.6 mm, 5.0 μm). An isocratic solvent system of 25% H 2 O: 75% acetonitrile was used with the flow rate of 1.0 ml min -1 .

Cell Culture
The NG108-15 cells were cultured in DMEM supplemented with 10% heatinactivated FBS, 2% penicillin/streptomycin, 1% amphotericin B and HAT as a complete medium. The cells were cultured at 37°C in 5% CO 2 atmosphere with 95% humidity and checked routinely under an inverted microscope (Motic) for any contamination.

Assessment of biological Activity
The biological activity of indolostilbenes was evaluated in NG108-15 by using MTT cell viability assay. 22 Metabolically active cells containing mitochondrial dehydrogenase converted the tetrazolium salt and MTT into insoluble purple formazan crystals at a rate proportional to the cell viability. The NG108-15 cells were rinsed with PBS, harvested with accutase, plated at a total density of 5.0 × 10 3 cells/well in a 96-well plate and left to adhere for 24 h. Cells were preincubated for 2 h with indolostilbenes 8 and 9 (3.125-100 µM) before H 2 O 2 (400 μM) exposure for subsequent 24 h. EGCG (1-50 µM) was used as a positive control. A 20 µl MTT solution (5 mg ml -1 ) was added and incubated at 37°C for another 4 h. The medium was removed, and DMSO was added to dissolve the formazan crystals. The absorbance reading was measured directly using a microplate reader (ASYS UVM340) at 570 nm (with a reference wavelength of 650 nm). Cell viability was calculated based on the formula below: % of cell viability = absorbance of treated cells (As)/ absorbance of control cells (AC) × 100 %

Statistical Analysis
The biological assays were conducted in triplicate, and the data was presented in means ± standard error (SE). Student's t-test and Mann's Whitney were used to determining the statistical significance (p values < 0.05) and differences between the control, untreated and treated groups were discussed. Statistical analyses between treated groups were determined using one-way ANOVA and Dunnett's test (significant # p values < 0.05).

Synthesis of Indolostilbenes 9 and 10
Initially, the building blocks were synthesised under an optimised condition with various catalyst, time and reaction conditions (Tables 1 and 2). The styrene 2 was obtained via our optimised Wittig reaction in 91% yield from the corresponding aldehyde 1 (Scheme 1) as shown in Table 1.
Notes: a Isolated yields after column chromatography; b data from the previous report; c the reactions were carried out in dry solvent and under nitrogen gas for 24 h; d the reactions were carried out in non-dry solvent and under nitrogen gas for 24 h.
The Heck coupling proceeded smoothly according to our protocol, as shown in Table 2 and Scheme 1. The Heck reaction was performed by heating at 120°C to the corresponding styrene 2 with acetamide 4 in the presence of Pd(OAc) 2 in dry DMF for 24 h. The corresponding stilbene 5 was generated in yields exceeding 76% (Table 2). Notes: a Isolated yields after column chromatography; b data from the previous report The 3,5-dimethoxystilbene 5 was subjected to anhydrous FeCl 3 oxidation (1:5 molar ratio, Scheme 2). Four products were isolated, i.e., an orthoacetamidobenzaldehyde 6, dichlorostilbene 7 and two dimers 8 and 9. Compounds 8 and 9 contained an intact stilbene olefinic bond (Scheme 2). Orthoacetamidobenzaldehyde 6 was identified by the presence of an aldehyde peak at δ H 9.90 and δ C 195.7 (in the 1 H and 13 C NMR, respectively) resulting from cleavage of the stilbene double bond. Apart from the mass spectrometric evidence for incorporating two chlorines into the molecule, the 13 C chemical shift for C-2 and C-6 of dichlorostilbene 7 was observed at δ C 114.0. The HRESIMS of dimers 8 and 9 showed a pseudo-molecular ion peak at m/z 648 and m/z 625, respectively, corresponding to the molecular formula of C 36  Previously, Kartini et al. isolated these indolostilbenes 8 and 9 at a low yield, i.e., 14% and 10%, respectively. 19 To obtain further insights into the oxidative cyclisation, control experiments (the solvent volume, the temperature of reaction and the proportion of Lewis acid) were examined using FeCl 3 as the catalyst (Table 3). From these reactions and careful examination of the variety of reaction condition, 5.0 equiv. of FeCl 3 in 50 ml dichloromethane at room temperature leads to a drastic improvement in yield of the indolostilbenes 8 and 9 at 43% and 42%, respectively. Trace amounts of ortho-acetamidobenzaldehyde 6 and dichlorostilbene 7 were also detected. Use of low temperature did not result in the formation of the indolostilbenes, while high temperatures resulted in low product yield. The use of dilute concentration of reactants improved the yield of the product in the radical cation reaction. 19,23,24 In this radical reaction, the optimum ratio between stilbene and the oxidant is 1:5. Further increase in the oxidant will drastically decrease the target molecules, resulting in the stilbene decomposition. Dichloromethane was found to be an excellent solvent for this reaction, aiding in the electrophilic chlorination step in the formation of chlorophenylindole 15, which was an intermediate in the formation of indolostilbenes 8 and 9 (Scheme 3). The use of acetone as the solvent did not result in the formation of indolostilbenes 8 and 9 (Table 3).

HPLC Analysis
The crude extract of FeCl 3 oxidative cyclisation was injected into an HPLC system. The chromatogram (Figure 2) was recorded at 254 nm. Dichlorostilbene 7, indolostilbene 8 and 9 gave the sharp peaks at t R = 13.7 min, 26.5 min and 34.57 min, respectively.

Mechanism Study
The 3,5-dimethoxystilbene has produced aldehyde 6, dichlorostilbene 7, and the unprecedented indolostilbene dimers 8 and 9. The formation of all target compounds was already described in our previous publication. 19 For the formation of indolostilbenes 8 and 9, the oxidation of 3,5-dimethoxystilbene 5 generates the radical cation 10. This proceeds with the formation of indole 12 by the rapid deprotonation of 11. Electrophilic chlorination of 12 via indolenium species 13 will give chlorophenylindole 15 (Scheme 3).

Protectivity of Indolostilbene Against H 2 O 2 -induced Cytotoxicity
The concentration-dependence of H 2 O 2 that reduced the NG108-15 cell viability was first optimised. The concentration selected was 400 μM as the cell viability of H 2 O 2 -treated cell at this concentration was decreased to 50 ± 1% after 24 h exposure, as summarised in Table 4 and Figure 3(a-b). The concentration of H 2 O 2 was optimised in our previous study using H 2 O 2 concentration ranging from 0.1-2 mM at 24 h. 25 Based on the optimisation, 400 µM of H 2 O 2 significantly reduced the NG108-15 cell viability around 50%. Thus, this concentration was selected for the protective evaluation. Prior to the pretreatment with indolostilbenes, the NG108-15 was treated with H 2 O 2 at various concentration, similar to our previous study. It was reported that at 400 µM, H 2 O 2 reduced the NG108-15 cell viability to 50% compared to untreated control cells. Thus, this concentration was selected for the H 2 O 2 protective evaluation in the MTT assay.
Interestingly, indolostilbene 9 pre-treatment at 100 μM resulted in a lower protection value compared at 50 μM. This suggests the possibility of indolostilbene 9 exhibiting toxicity at a higher dose. Due to indolostilbene 9 toxicity effects, H 2 O 2 protective effect of indolostilbene 8 was more dose-dependent than indolostilbene 9. Even though both indolostilbene 8 and 9 showed lower H 2 O 2 protective activity when compared to EGCG (Table 4), it was significantly different ( * P < 0.05) compared to the H 2 O 2 -treated group.
From the structure-activity relationship (SAR), we can conclude that both compounds contain an indole and stilbene cores, 3,5-dimethoxy B ring and N-acetamide substituent. However, the coupling position between an indole and stilbene moieties may cause differences in the activity. For compound 9 which is coupling at ortho-para position show a good activity at 80±1% for NG108-15 cell viability at the minimal concentration (50 μM) compared to compound 8 which is coupling at ortho-ortho position (72±1% cell viability at 100 μM). Interestingly, the formation of intramolecular hydrogen bonding in compound 8 between N-H amide in stilbene moiety and C=O amide in the indole moiety, make the compound well packed, rigid and decrease the activity. Meanwhile, the intramolecular H-bonding for compound 9 is impossible, and the compound is free to rotate. This evidence has been supported by PM6/MOPAC2007 calculation in our previous publication. 19   26,27 The NG108-15 cell line is a suitable choice because it exhibits angiotensin II receptor. It can be further differentiated into a cholinergic phenotype cell line with voltagedependent membrane currents, acetylcholine release and the formation of cholinergic synapses. [28][29][30] Furthermore, NG108-15 has been extensively used as the in-vitro neuroprotective neuronal model using various phytochemicals 25,31,32 and neuritogenesis studies. [33][34][35][36] Considering its previous use, NG108-15 is a suitable neuronal model for the in vitro neuroprotective study reported herein.

CONCLUSION
In summary, we have developed a simple and efficient method to synthesise indolostilbene via ferric chloride oxidative cyclisation. The synthesised stilbenes were evaluated for H 2 O 2 protective activity. The optimum condition was achieved by using 5.0 equiv. of anhydrous FeCl 3 in dichloromethane as the solvent. It gives an excellent yield of up to 43% and 42% of indolostilbenes 8 and 9, respectively. It was observed that the dichloromethane was involved in the formation of chlorophenylindole 15 as an intermediate for the formation of indolostilbene 8 and 9. In addition, the indolostilbenes 8 and 9 showed potent protective activity against H 2 O 2 in NG108-15 cells. Pretreatment with both compounds significantly increased the NG108-15 cell viability to 70%±1% (at 100 μM) and 80%±1% (at 50 μM) respectively. These were significantly different ( * P < 0.05) compared to the H 2 O 2 -treated group (50%±1%). Although the H 2 O 2 protective effects of indolostilbenes 8 and 9 were not as potent as EGCG, their protective effect against H 2 O 2 was significantly higher than the H 2 O 2 -treated value, with an increase in cell viability between 21%-31%. However, further studies are needed to understand the mechanism of action involved in the H 2 O 2 -induced intracellular pathway leading to cell apoptosis.