Synthesis of Bio-Inspired 1,3-Diarylpropene Derivatives via Heck Cross-Coupling and Cytotoxic Evaluation on Breast Cancer Cells

The Heck cross-coupling reaction is a well-established chemical tool for the synthesis of unsaturated compounds by formation of a new C-C bond. In this study, 1,3-diarylpropene derivatives, designed as structural analogues of stilbenoids and dihydrostilbenoids, were synthesised by the palladium-catalysed reactions of 2-amidoiodobenzene derivatives with either estragole or eugenol. The products were obtained with high (E) stereoselectivity but as two regioisomers. The ratios of isomers were found to be dependent on the nature of the allylbenzene partner and were rationalised by electronic effects exercising a determining influence in the β-hydride elimination step. In addition, the cytotoxic effects of all the Heck reaction products were evaluated against MCF-7 and MDA-MB-231 human breast cancer cells, with unpromising results. Among all, compound 7d exhibited weak cytotoxic activity towards MCF-7 cell lines with IC50 values of 47.92 µM in comparison with tamoxifen and was considered to have general toxicity (SI value < 2).


Introduction
The Heck reaction has perpetuated its significance in synthetic organic chemistry for over half a century. Indeed, the introduction of this carbon-carbon bond-making crosscoupling reaction by Heck and Mizoroki has revolutionised synthetic methodology, thanks to its simplicity and effectiveness [1][2][3][4][5]. Since the first detailed mechanism was proposed by Heck and Nolley in 1972 [2], the elementary steps, reaction parameters and catalytic systems have counted among the major aspects attracting the interest of many researchers [4,6]. One crucial part of the Heck catalytic cycle, that sometimes receives little attention by synthetic chemists, is the β-hydride elimination elementary step. This process plays a key role in enantioselective versions of the reaction, for which it must proceed with suitable regioselectivity [7][8][9][10]. Control is also essential in so-called reductive Heck reactions for which, conversely, β-hydride elimination is to be avoided [11]. In relation to this context, the synthesis of 1,3-diarylpropene derivatives by the Heck reaction is especially worthy of study. Indeed, when the two aromatic groups differ, two possible regioisomers can be produced (Scheme 1). Scheme 1. Synthesis of 1,3-diarylpropene derivatives by the Heck re structures of the products with those of stilbenes and dihydrostilbenes.
Moreover, such compounds can be viewed as highe diphenylethylene derivatives (stilbenes), as well as of dihydrosti is worth pointing out that hydroxylated stilbenes (stilbeno dihydrostilbenes are of special biological importance. Indeed, stilb various plants in response to stressors such as mechanical inju attack by herbivores, or infection by pathogenic agents [12 phytoalexin molecules are derived from trans-resveratrol, the pre vines has been thought to be involved in the "French paradox" [ other naturally occurring (E)-stilbene derivatives, such as piceat isorhapontigenin, have attracted much attention and have been studies. Various interesting pharmacological properties have be antioxidant, anticancer, anti-inflammatory, cardioprotective, anti activities [14,15,16,17,18,19]. Interestingly, dihydrostilbenoids exhibit a similar range of biological effects [20]. Some year preparation of stilbenoid structural analogues having an amido position of one of the two aromatic parts, using the Heck reactio molecules were found to possess significant anticancer propert DU-145, MCF-7 and BxPC-3 cancer cell lines, or chemopreventiv In our continuing efforts directed towards the discovery of derivatives, we decided to investigate the application of the Heck of new compounds having a 1,3-diarylpropene skeleton, thus mo a C6-C3-C6 structure, but retaining the pharmacophores, i.e., an a of the phenyl rings and oxygen-based substituents on the Moreover, such compounds can be viewed as higher homologues of 1,2-diphenylethylene derivatives (stilbenes), as well as of dihydrostilbenes. In this respect, it is worth pointing out that hydroxylated stilbenes (stilbenoids) and hydroxylated dihydrostilbenes are of special biological importance. Indeed, stilbenoids are produced by various plants in response to stressors such as mechanical injury, e.g., resulting from attack by herbivores, or infection by pathogenic agents [12][13][14]. Most of these phytoalexin molecules are derived from trans-resveratrol, the presence of which in grape vines has been thought to be involved in the "French paradox" [15]. (E)-Resveratrol and other naturally occurring (E)-stilbene derivatives, such as piceatannol, oxyresveratrol or isorhapontigenin, have attracted much attention and have been the subject of extensive studies. Various interesting pharmacological properties have been disclosed, including antioxidant, anticancer, anti-inflammatory, cardioprotective, antifungal and antibacterial activities [14][15][16][17][18][19]. Interestingly, dihydrostilbenoids have been reported to exhibit a similar range of biological effects [20]. Some years ago, we studied the preparation of stilbenoid structural analogues having an amido substituent at the ortho position of one of the two aromatic parts, using the Heck reaction [21,22]. Some of these molecules were found to possess significant anticancer properties against HT-29, P388, DU-145, MCF-7 and BxPC-3 cancer cell lines, or chemopreventive action [21,22].
In our continuing efforts directed towards the discovery of novel biologically active derivatives, we decided to investigate the application of the Heck reaction to the synthesis of new compounds having a 1,3-diarylpropene skeleton, thus moving from a C6-C2-C6 to a C6-C3-C6 structure, but retaining the pharmacophores, i.e., an amido substituent on one of the phenyl rings and oxygen-based substituents on the other (Scheme 2). It is noteworthy that these new target molecules can be viewed not only as structural analogues of stilbenoids and dihydrostilbenoids, but, in fact, as hybrid compounds having characteristics of both types of molecules.

Characterisation of the 1,3-Diarylpropene Heck Products
Compounds 6a-j-9a-j were characterised by FT-IR, 1D-and 2D-NMR spectroscopies as well as an HRMS analysis. Compounds 8b,9b were taken as an example to describe the obtained data. The FT-IR spectrum of the 8b/9b mixture exhibited characteristic absorptions at 3398 (broad, w), 3291 (broad, m), 2970-2873 (w), 1654 (s), 1509 (s) and 1371 (m) cm −1 , corresponding to O-H stretching, N-H stretching, Csp 3 -H stretching, saturated amide C=O stretching, C=C aromatic stretching and CH 3 bending, respectively. An out-of-plane C-H bending vibration at 753 cm −1 (s) indicated the presence of orthosubstituted aromatic rings in compound 8b/9b. On the other hand, a CH 2 bending vibration was observed at 1446 cm −1 (m), indicating the presence of a methylene group (CH 2 ). The para-substituted aromatic ring displayed out-of-plane C-H bending vibration bands at 812 (w) and 848 (m) cm −1 .
Typically, the comprehensive description of all the 1 H and 13 C NMR signals of mixtures of isomers such as 8b/9b was not straightforward. However, it could be achieved by first carefully assigning the signals of the few Heck products for which we had managed to obtain samples of pure isomers, including 9b. This was done on the basis of 1D and 2D NMR experiments: COSY, DEPT135, HSQC and HMBC, as well as by comparison with reported NMR data of simpler compounds sharing structural subunits with our molecules, in particular 1- [26], N-(2allylphenyl)acetamide [27] and N-[2-[(E)-cinnamyl]phenyl]prop-2-enamide [28]. Once this operation was completed, the corresponding descriptions were used as reference data for the remaining analyses.
The 1 H NMR spectrum of the 8b/9b mixture exhibited important characteristic signals within the range δ H 1.20-7.95 ppm. A downfield broad singlet for the amide proton (NH) of the major isomer 9b was observed at δ H 7.35 ppm. The seven aromatic protons were distributed in two regions: 7.06−7.95 ppm and 6.72−6.85 ppm. The first area was assigned to the protons attached to the nitrogen-substituted aromatic ring. The doublet of the proton at the ortho position relative to the nitrogen atom was characteristic, with a particularly high chemical shift value: δ H 7.95 ppm (J 8.0 Hz) for 9b and δ H 7.84 ppm (J 7.6 Hz) for 8b. The somewhat lower value for the latter was a general observation in all the 6/7 and 8/9 series. The aromatic signals lying in the higher field region were assigned to the methoxyphenol part. They appeared as two sets of overlapping signals: a singlet, accounting for the proton located ortho to the methoxyl substituent; and an AB system for the other two protons, with a J AB coupling constant of 9.0 Hz. These signals were observed at a markedly lower field in the case of 9b (δ H 6.81−6.87 ppm) than in the case of 8b (δ H 6.72−6.74 ppm), which was a general trend for all 8/9 mixtures.
For each isomer and as expected, the propene unit bridging the two aromatic rings appeared as a doublet (one proton, J 16.0 Hz) for the vinylic CH directly attached to one of the aromatic rings (δ H 6.38 ppm for 9b and 6.42 ppm for 8b), a doublet of triplets (one proton, J 16.0 and 6.0 Hz) for the central vinylic CH (δ H 6.16 ppm for 9b and 6.27 ppm for 8b) and a doublet (two protons, J 6.0 Hz) for the allylic protons, in the aliphatic region (δ H 3.53 ppm for 9b and 3.50 ppm for 8b). The high 3 J coupling constant of 16.0 Hz confirmed the E configuration of the CH=CH double bond. The hydroxyl (OH) and methoxyl (OCH 3 ) protons appeared as two singlets at δ H 5.73 and 3.87 ppm, respectively, in the case of 9b. The OH proton of 8b came out at a somewhat higher field (δ H 5.61 ppm), which was again a general trend in the 8/9 series.
The isopropyl group of 9b was characterised by an upfield doublet at δ H 1.21 ppm (J 7.0 Hz), accounting for the magnetically equivalent six protons of the methyl groups. A septet (J 7.0 Hz) at δ H 2.47 ppm was assigned to the neighbouring CH. It is worth mentioning that in the 1 H NMR spectrum of the 8b/9b mixture, as well as in several other cases, traces of other products could be detected. Their isolation and characterisation were not attempted, but by analogy with related results reported in the literature [24][25][26], these were likely to be minor by-products of the Heck reaction, namely Z isomers of 8b and 9b and/or the vinylidene product resulting from alternative regioselectivity of the alkene 1,2-insertion step.
The 13 C NMR spectrum of 8b/9b exhibited important characteristic signals within the range δ c 19.6-175.1 ppm. A total of two sets of 19 carbon signals were observed, with different intensities, with the highest peaks accounting for the major isomer 9b. The presence of the carbonyl carbon was evidenced by a downfield signal at δ c 175.1 ppm. The signals of the twelve aromatic ring carbon nuclei of 8b/9b were observed in the range δ c 107.9−146.7 ppm. The peak of the CH located at the ortho position relative to the methoxyl substituent was especially characteristic, with a comparatively low chemical shift value due to the +M electron-donating effect of the substituents of this ring (δ c 107.9 ppm for 9b and 111.2 ppm for 8b). The signals of the propene subunit also exhibited marked differences depending on the isomers. The CH 2 group was clearly identifiable, with a chemical shift around 39 ppm for all compounds 8a-j (δ c 39.2 ppm for 8b) and a value around 36 ppm for 9a-j (δ c 36.1 ppm for 9b). The chemical shifts of the vinylic carbon atoms were also good markers. For instance, the central vinylic CH appeared at around 134 ppm in the 8 series (δ c 134.2 ppm for 8b) and at around 125 ppm in the 9 series (δ c 125.1 ppm for 9b). The methoxyl substituent (OMe) gave a signal at δ c 55.9 ppm for 9b and 56.0 ppm for 8b; the isopropyl group was identified by two peaks at δ c 36.7 (CH) and 19.6 ppm (Me) for both isomers.

Mechanistic Interpretation of the Results of the Heck Cross-Coupling Reactions
An examination of the literature reveals several examples of related Heck reactions giving mixtures of isomers [29][30][31][32][33][34][35][36]. In several cases, the ratios of isomers were not given and generally, the origin of the selectivities observed was not discussed. In a few instances, the observation of two main products having the same atom connections was interpreted as the formation of E and Z isomers [31,33], whereas those compounds are perhaps more likely to be regioisomers, as in the present work.
The mechanism of the Heck reaction and the exact nature of the intermediate palladium complexes involved have been extensively studied and discussed [3][4][5][6]. A simplified view is displayed in Scheme 3. A pre-activation step generated palladium(0), which then underwent the oxidative addition of the aryl iodide substrate 3. This produced the intermediate complex 10, where the palladium atom was attached to the ortho amido-substituted aryl group. Coordination of the allylbenzene reactant 4 or 5, followed by syn 1,2-insertion of the C=C bond, gave rise to the alkylpalladium species 11. The next step was a syn β-hydride elimination which, after de-coordination, delivered the product: 6 or 7 from 4 or 8 or 9 from 5, along with a palladium(II) hydride species. Finally, the palladium(0) catalyst was regenerated in a reductive elimination process.
The crucial elementary step with respect to the stereo-and regioselectivity of the formation of the products was the syn β-hydride elimination event taking place from intermediate 11. Indeed, the outcome was fully determined by which of the four neighbouring hydrogen atoms that are located at the β position relative to the metal centre participates in this process. In our reactions, little regioselectivity was observed when 1-allyl-4-methoxybenzene (estragole) 4 was employed, whereas with 1-allyl-3-methoxy-4-hydroxybenzene (eugenol) 5, selectivity in favour of product 9 was substantial, ranging from 62:38 to 81:19. In all cases, the (E) isomers were produced with high selectivity. The latter stereoselectivity is commonly observed in Heck reactions and it is generally considered that this is the result of kinetic control governed by the Curtin-Hammett principle [3][4][5]. The crucial elementary step with respect to the stereo-and reg formation of the products was the syn β-hydride elimination event intermediate 11. Indeed, the outcome was fully determined by neighbouring hydrogen atoms that are located at the β position relative participates in this process. In our reactions, little regioselectivity wa allyl-4-methoxybenzene (estragole) 4 was employed, whereas with 1 hydroxybenzene (eugenol) 5, selectivity in favour of product 9 was s from 62:38 to 81:19. In all cases, the (E) isomers were produced with h latter stereoselectivity is commonly observed in Heck reactions a considered that this is the result of kinetic control governed by th principle [3][4][5].
Similarly, it appears that in the present examples, the observe cannot be explained by thermodynamic control. Products 7 and 9 w stabilised by conjugation of the C=C alkene fragment with the aromatic group, whereas in 6 and in 8, the presence of the ortho amido the conformation of the neighbouring alkene out of the plane of the aro compromising a similar stabilising conjugation effect. Semi-empirica performed with R 1 = Me or iPr, support this analysis: 7a, 7b, 9a and 9b be more stable than their respective regioisomers 6a, 6b, 8a and kcal.mol −1 . Using estragole 4, thermodynamic control would thus resu selectivity in favour of 7, in contrast to what was experimental Similarly, it appears that in the present examples, the observed regioselectivities cannot be explained by thermodynamic control. Products 7 and 9 were expected to be stabilised by conjugation of the C=C alkene fragment with the oxygen-substituted aromatic group, whereas in 6 and in 8, the presence of the ortho amido substituent forced the conformation of the neighbouring alkene out of the plane of the aromatic ring, thereby compromising a similar stabilising conjugation effect. Semi-empirical PM7 calculations, performed with R 1 = Me or iPr, support this analysis: 7a, 7b, 9a and 9b were estimated to be more stable than their respective regioisomers 6a, 6b, 8a and 8b by 0.87 to 1.83 kcal.mol −1 . Using estragole 4, thermodynamic control would thus result in significant 6/7 selectivity in favour of 7, in contrast to what was experimentally observed. More importantly, the energy difference between 8a and 9a was found to be lower than between 6a and 7a: 0.87 and 1.39 kcal.mol −1 , respectively. The same situation was calculated in the case of 8b/9b vs. 6b/7b, with respective energy differences of 1.37 and 1.83 kcal.mol −1 . Pure thermodynamic control would thus lead to lower selectivities with eugenol 5 than with estragole 4, which is the opposite of what is actually observed. Therefore, regioselectivity is most plausibly under kinetic control. In the reaction of intermediate 11, four possible transition states can be drawn, among which two, 12 and 13, account for the formation of the observed major (E) isomers (Scheme 4).

Scheme 4.
Proposed transition states of the β-hydride elimination step from 11, leading either to regioisomers (E)-6/8 or (E)-7/9. The higher selectivity observed using eugenol 5 indicates that the transition state 13 became comparatively more energetically accessible than 12 when the oxygen-substituted aromatic ring (displayed in blue) became more electron-rich. This was consistent with a partial positive charge being developed at the carbon atom of the C−H bond that was being broken in the process. This positive charge was best accommodated by the most electron-rich adjacent aromatic ring. This provides new support for a similar explanation that had been proposed by the group of Sigman to explain the regioselectivities of Heck cross-coupling reactions of allylbenzene with aryldiazonium salts having various electronic properties [34].

Cytotoxicity Activity
The MTT assay was used to evaluate the toxicity of N-(2-cinnamylphenyl)amides on the viability of two types of breast cancer cell lines, i.e., MCF-7 (with a receptor) and MDA-MB-231 (without a receptor). In this study, a normal breast cell line (MCF-10A) was used. MCF-7 is an ER-positive breast cancer cell, whereas MDA-MB-231 is an ER-negative breast cancer cell. These cancer cells have receptor sites that bind to oestrogen, promoting their growth and spread. Tamoxifen is a first-line anti-oestrogen therapy that is used to prevent the production or action of oestrogen.
For the cytotoxicity assays, only pure isomers were submitted for evaluations. The cell lines were treated with different concentrations of the synthesized compounds, i.e., between 10 and 100 µM and incubated for 24−72 h. Table 2   The higher selectivity observed using eugenol 5 indicates that the transition state 13 became comparatively more energetically accessible than 12 when the oxygen-substituted aromatic ring (displayed in blue) became more electron-rich. This was consistent with a partial positive charge being developed at the carbon atom of the C−H bond that was being broken in the process. This positive charge was best accommodated by the most electron-rich adjacent aromatic ring. This provides new support for a similar explanation that had been proposed by the group of Sigman to explain the regioselectivities of Heck cross-coupling reactions of allylbenzene with aryldiazonium salts having various electronic properties [34].

Cytotoxicity Activity
The MTT assay was used to evaluate the toxicity of N-(2-cinnamylphenyl)amides on the viability of two types of breast cancer cell lines, i.e., MCF-7 (with a receptor) and MDA-MB-231 (without a receptor). In this study, a normal breast cell line (MCF-10A) was used. MCF-7 is an ER-positive breast cancer cell, whereas MDA-MB-231 is an ER-negative breast cancer cell. These cancer cells have receptor sites that bind to oestrogen, promoting their growth and spread. Tamoxifen is a first-line anti-oestrogen therapy that is used to prevent the production or action of oestrogen.
For the cytotoxicity assays, only pure isomers were submitted for evaluations. The cell lines were treated with different concentrations of the synthesized compounds, i.e., between 10 and 100 µM and incubated for 24−72 h.  The SI value was obtained by dividing the IC 50 value for normal cell lines by the IC 50 for cancerous cell lines. All compounds with SI values less than 2 were considered to have general toxicity, suggesting that they can cause cytotoxicity in normal cells as well [37].
The procedure of intermediates and final compounds synthesis and their NMR; ATR-IR and HRMS spectra of the compounds are presented in Supplementary Materials.

General Procedure for the Preparation of the Ortho-Amido-Substituted Iodobenzene Precursors
A solution of 2-iodoaniline 1 (1.00 equiv.) and Et 3 N (1.00 equiv.) in dry THF (1.6-2.0 mL per mmol of 1) was cooled to 0-5 • C. The requisite acyl chloride 2 (1.00 equiv.) was then added dropwise, with stirring. The cold bath was removed, and the mixture was stirred vigorously overnight at room temperature. The white precipitate of Et 3 N.HCl was filtered off and rinsed with THF (3 × 5 mL). All the organic fractions were combined and concentrated under reduced pressure to afford the crude amide product 3, which was then purified by recrystallisation from n-hexane/chloroform. Full details and characterisation data for compounds 3a-3j are attached in the Supporting Information.

General Procedure for Preparing the Heck Cross-Coupling Reaction Products
In a dry three-necked flask equipped with a thermometer and a condenser, a solution of N-(2-iodophenyl) amide 3 (1.00 equiv.) in dry DMF (4.0 mL per mmol of substrate 3) was heated up to 120 • C and stirred for 20 min under nitrogen. Palladium(II) acetate (1.00% equiv.), triethylamine (3.50-5.00 equiv.) and the requisite allylbenzene derivative 4 or 5 (1.60 equiv.) were then added successively. The reaction mixture was stirred at 120 • C until the TLC analysis indicated complete consumption of 3 (3-6 h). After cooling, the saturated NH 4 Cl aqueous solution (6.0 mL per mmol of starting 3) was added and the mixture was extracted with EtOAc (three times, with portions of 2.0 mL per mmol of starting 3). The combined organic fractions were washed with H 2 O (6.0 mL per mmol of starting 3), dried over anhydrous sodium sulphate, filtered and concentrated under reduced pressure. The ratio of product isomers was determined by 1 H NMR spectroscopy of the crude product. Purification by gradient elution column chromatography (n-hexane/ethyl acetate, 95:5 to 50:50) then yielded the desired Heck reaction products. Some regioisomers could not be separated and reported as a mixture. Full details for compounds 6a-9j are attached in the Supporting Information.  A plot of % cell proliferation versus sample concentration was used to show the 50% inhibitory concentration (IC 50 ). The selectivity index (SI) was used to indicate the cytotoxic selectivity of the samples against cancer cells and normal cells. It was calculated from the IC 50 of the samples in normal cells versus cancer cells. An SI of less than 2 suggests the general toxicity of the drug/sample in cells, while an SI of more than 2 is an indication of the sample selectively targeting cancer cells [37]. All data were analysed and presented as means standard deviation (±SD). Comparisons between each concentration of 18 samples were evaluated using Student's t-test, where p < 0.05, p < 0.01 or p < 0.001 were considered as statistically significant relative to the untreated control cells.

Conflicts of Interest:
The authors declare no conflict of interest.