Antibacterial and Antitubercular Activities of Cinnamylideneacetophenones

Cinnamaldehyde is a natural product with broad spectrum of antibacterial activity. In this work, it was used as a template for design and synthesis of a series of 17 cinnamylideneacetophenones. Phenolic compounds 3 and 4 exhibited MIC (minimum inhibitory concentration) and MBC (minimum bactericidal concentration) values of 77.9 to 312 µM against Staphylococcus aureus, Streptococcus mutans, and Streptococcus sanguinis. Compounds 2, 7, 10, and 18 presented potent effects against Mycobacterium tuberculosis (57.2 µM ≤ MIC ≤ 70.9 µM). Hydrophilic effects caused by substituents on ring B increased antibacterial activity against Gram-positive species. Thus, log Po/w were calculated by using high-performance liquid chromatography-photodiode array detection (HPLC-PDA) analyses, and cinnamylideneacetophenones presented values ranging from 2.5 to 4.1. In addition, the effects of 3 and 4 were evaluated on pulmonary cells, indicating their moderate toxicity (46.3 µM ≤ IC50 ≤ 96.7 µM) when compared with doxorubicin. Bioactive compounds were subjected to in silico prediction of pharmacokinetic properties, and did not violate Lipinski’s and Veber’s rules, corroborating their potential bioavailability by an oral route.


Introduction
Cinnamaldehyde (1) is the main component in cinnamon bark oil (50-70%), and it is responsible for its spicy and sweet taste [1,2]. Structurally, cinnamaldehyde is a natural simple phenylpropanoid, bearing an α,β-unsaturated aromatic aldehyde. The synthetic applications of 1 are vast and related to its abundance and low economic costs, and have led to a collection of compounds with high chemodiversity [3].
In vitro and in vivo toxicity of 1 have been clearly elucidated due to its application as a food additive [16,17]. In this context, 1 was classified as GRAS (generally recognized as safe) by the FDA (Food and Drug Administration) [18]. On the other hand, cinnamaldehyde possess an aliphatic aldoxyl group, which has undesirable functionality due to its high reactivity [19]. Thus, we designed a series of cinnamylideneacetophenones with the replacement of the aldoxyl by an acetophenone moiety. The resonance of an additional benzene ring leads to compounds that are chemically more stable than cinnamaldehyde, allowing insertion of versatile substituents to derive preliminary structure-activity relationships (Scheme 1).

Scheme 2. Synthesis of cinnamylideneacetophenones.
Reactions between cinnamaldehyde and corresponding acetophenone were performed under basic catalysis, at room temperature, resulting in yields of 33-98% (Table 1). Cinnamylideneacetophenones exhibited substitutions at para-and meta-positions on ring B by electron-donating groups (OH, NH2, OMe, OCH2O, and Me) and electron-withdrawing groups (F, Cl, Br, NO2, and CF3). In general, a clear and unambiguous relationship between the stereoelectronic effect of substituents on ring B and reaction yields cannot be established. In our ongoing search for new antibacterials from natural product-based synthetic compounds, we synthesized a series of 17 cinnamylideneacetophenones (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18). Their antibacterial and antitubercular activities were evaluated against Gram-positive and Gram-negative species, as well as M. tuberculosis. The two most active compounds (3 and 4) were investigated regarding their selective toxicity on eukaryotic cells by using two human pulmonary cell lineages (MRC-5 and A549). In addition, hydrophilicity of all compounds was measured by HPLC-PDA experiments, which enabled calculation of partition coefficients (log P o/w ). In silico investigations using the Molinspiration online property calculation toolkit contributed to the prediction of absorption, distribution, metabolism and excretion (ADME) properties of selected bioactive cinnamylideneacetophenones.
In vitro and in vivo toxicity of 1 have been clearly elucidated due to its application as a food additive [16,17]. In this context, 1 was classified as GRAS (generally recognized as safe) by the FDA (Food and Drug Administration) [18]. On the other hand, cinnamaldehyde possess an aliphatic aldoxyl group, which has undesirable functionality due to its high reactivity [19]. Thus, we designed a series of cinnamylideneacetophenones with the replacement of the aldoxyl by an acetophenone moiety. The resonance of an additional benzene ring leads to compounds that are chemically more stable than cinnamaldehyde, allowing insertion of versatile substituents to derive preliminary structure-activity relationships (Scheme 1).

Chemistry
Synthesis of cinnamylideneacetophenones 2-18 was achieved by Claisen-Schmidt condensation, using protocol reported by Weldon and co-authors, with minor modifications [20][21][22] (Scheme 2). Reactions between cinnamaldehyde and corresponding acetophenone were performed under basic catalysis, at room temperature, resulting in yields of 33-98% (Table 1). Cinnamylideneacetophenones exhibited substitutions at para-and meta-positions on ring B by electron-donating groups (OH, NH2, OMe, OCH2O, and Me) and electron-withdrawing groups (F, Cl, Br, NO2, and CF3). In general, a clear and unambiguous relationship between the stereoelectronic effect of substituents on ring B and reaction yields cannot be established. Reactions between cinnamaldehyde and corresponding acetophenone were performed under basic catalysis, at room temperature, resulting in yields of 33-98% (Table 1). Cinnamylideneacetophenones exhibited substitutions at paraand meta-positions on ring B by electron-donating groups (OH, NH 2 , OMe, OCH 2 O, and Me) and electron-withdrawing groups (F, Cl, Br, NO 2 , and CF 3 ). In general, a clear and unambiguous relationship between the stereoelectronic effect of substituents on ring B and reaction yields cannot be established.  The structure of compounds was identified by their melting point, UV-Vis, 1 H-NMR, and 13 C-NMR spectra analyses and was compared with previous data [20][21][22][23]. Out of 17 compounds, compound 9 was identified as a new chemical entity. UV-Vis spectra performed on HPLC-PDA experiments indicated λ max values ranging from 343 to 361 nm, corroborating the presence of an α,β,γ,δ-unsaturated ketone system in cinnamylideneacetophenone structures. For all compounds, NMR data, including chemical shifts, integrations, multiplicities, and coupling constants, corresponded with the proposed structures. Cinnamylideneacetophenones were analyzed by HPLC-PDA and their purity was measured through peak area values, which ranged from 96.1 to 99.8%. Spectra and chromatograms were presented as supplementary material.

Antibacterial, Antitubercular Activity, and Structure-Activity Relationship (SAR)
For antibacterial evaluations, three Gram-positive and two Gram-negative species were used, including Staphylococcus aureus (ATCC 14458), Streptococcus mutans (ATCC 25175), Streptococcus sanguinis (ATCC 10557), Pseudomonas aeruginosa (ATCC 15442) and Escherichia coli (ATCC 10536). The antibacterial potency was demonstrated by MIC (minimum inhibitory concentration) and MBC (minimum bactericidal concentration) values. MIC and MBC were determined by broth microdilution method, in 96-well microtiter plates, following the criteria previously described by the National Committee for Clinical Laboratory Standards [24]. Penicillin, vancomycin, and chlorhexidine were used as positive controls. All cinnamylideneacetophenones were assayed in concentrations ranging from 3.9 to 500 µM.
For antitubercular assays, Mycobacterium tuberculosis H37Rv (ATCC 27294) was used as a reference strain. Antitubercular potency was expressed as MIC values, which were established by protocol reported by Palomino and co-author, with minor modifications [25,26]. Isoniazid was used as the reference antitubercular drug. Compounds 1-18 were evaluated in concentrations ranging from 1 to 200 µM (Table 2).

Evaluation of Cytotoxicity and Determination of Selectivity Index
The two most potent compounds (3 and 4) were investigated for their selective toxicity against eukaryotic cells. Human lung fibroblast (MRC-5, ATCC, CCL-171) and human lung adenocarcinoma cells (A549, ATCC, CCL-185) were used for cytotoxicity evaluations. The choice for pulmonary cells was based on their importance as host site infection of S. aureus and M. tuberculosis [26,31]. The cytotoxic assays were performed using protocol described by our group [32]. Compounds 3 and 4 displayed similar effects against both human cells, demonstrating values of IC50 (concentration able to inhibit 50% of cell growth) ranging from 46.3 to 96.7 µM (Table 3). These compounds were moderately toxic against pulmonary cells when compared with doxorubicin which exhibited IC50 values of 0.7 and 1.8 µM against MRC-5 and A549, respectively. In order to compare the effect towards prokaryotic cells versus eukaryotic cells, the selectivity indexes (SI) of 3 and 4 were calculated as the ratio of cytotoxicity to antimicrobial activities (SI = IC50 eukaryotic cell/MIC prokaryotic cells) ( Table 3). The effectiveness of a compound is indicated by SI ≥ 10. However, SI ≤ 1 indicates toxicity of that compound for eukaryotic cells and low selectivity for the prokaryotic cells. Despite 3 and 4 demonstrating promising antibacterial and antitubercular activities, these compounds showed SI values ranging from 0.5 to 1.2.
Antimicrobial agents with in vitro toxicity against human cells should not be ignored. Although SI values are relevant as a general orientation, this ratio cannot be used as the unique parameter to decide if a compound can continue to preclinical trials [33,34]. In this context, in vivo toxicity trials are highly indicated to capture the potential of 3 and 4 as valuable molecules. m-OMe Cinnamaldehyde (1) displayed MIC > 500 µM against S. aureus, S. mutans, S. sanguinis, P. aeuruginosa, and E. coli. These results corroborated other investigations, which have described MIC values higher than 500 µM. Domadia and collaborators reported MIC values of 1000 mg/L (7566 µM) against E. coli [14]. Ferro and co-authors described MIC values of 0.25 mg/mL (1891 µM) against the S. aureus strains [27]. On the other hand, our result against M. tuberculosis (MIC > 200 µM) was opposite to that described by Ochoa and colleagues, who reported MIC value 3.12 µg/mL (23 µM) [10].
We used unsubstituted cinnamylideneacetophenone (2) as a framework to investigate the relationship between substitutions on ring B and antibacterial/antitubercular activities. Firstly, 2 was compared with phenolic compounds 3 and 4, which suggested the hydroxyl group improved antibacterial activity, mainly against Gram-positive species. Compound 3 demonstrated MIC and MBC values of 77.9 and 156 µM against S. aureus, S. mutans and S. sanguinis, respectively. Compound 4 displayed similar antibacterial potency with MIC and MBC values ranging from 77.9 to 312 µM. On the other hand, the hydroxyl group reduced antitubercular activity, because unsubstituted compound 2 (MIC = 57.2 µM) was more potent than phenolics 3 (MIC = 88.8 µM) and 4 (MIC = 81.9 µM). Also, aminoderivative 5 exhibited activity against S. mutans (MIC = MBC = 313 µM) and was not able to act against S. aureus, S. sanguinis, nor M. tuberculosis. These results indicate that electron-donating and hydrophilic groups interfere with bioactivity, and weak acidic compounds 3 and 4 were more active than basic compound 5.
Third, comparisons were made to establish if hydrophilicity or electron donation is pivotal for bioactivity. Cinnamylideneacetophenones carrying hydrophobic and electron-withdrawing substituents (F, Cl, Br, CF 3 , and NO 2 ) (10-16) were inactive against S. aureus, S. mutans, and S. sanguinis. In addition, methylated compounds 17 and 18, which are compounds substituted by hydrophobic and electron-donating groups, were also inactive. Thus, the electronic effect was not crucial for antibacterial activity, but hydrophilicity increased it. In contrast, antitubercular assays indicated para-fluorinated and para-methylated Interestingly, cinnamylideneacetophenones that were bioactive against Gram-positive species were not able to inhibit Gram-negative species growth, such as P. aeruginosa and E. coli, therefore, they exhibited reduced spectrum. Gram negative bacteria demonstrate a complex barrier system, which makes them less susceptible to action of antibacterial agents. As part of this system, the cell wall is mainly responsible for impermeability due to the presence of an external envelope, composed of lipopolysaccharides, phospholipids, and proteins [28,29].

Determination of Partition Coefficient Log P o/w
Preliminary structure-activity relationship (SAR) indicated the essentiality of the hydrophilicity for antibacterial and antitubercular activities. Thus, we calculated the partition coefficient (n-octanol/water) (log P o/w ) by using the HPLC-PDA method suggested by Organisation for Economic Co-operation and Development (OECD) protocols [30]. Log P o/w values of the cinnamylideneacetophenones 2-18 are summarized in Table 2. The reference compounds, aniline, phenol, acetophenone, benzoic acid, benzene, toluene, benzophenone, and thymol, were used to construct a log K versus log P o/w curve. The capacity factor (log K) of the cinnamylideneacetophenones was determined from retention times. Anti-Staphylococcus and anti-Streptococcus compounds exhibited log P o/w values ranging from 2.5 to 3.3. Among these, compounds 3 and 4 were the most potent, with log P o/w values of 2.7 and 2.6, respectively. The four most active anti-Mycobacterium compounds (MIC ď 70.9 µM) presented log P o/w values ranging from 3.2 to 3.5. Highly hydrophobic compounds displayed log P o/w > 3.5 (11-14 and 17) and were not able to act against S. aureus, S. mutans, S. sanguinis, and M. tuberculosis.

Evaluation of Cytotoxicity and Determination of Selectivity Index
The two most potent compounds (3 and 4) were investigated for their selective toxicity against eukaryotic cells. Human lung fibroblast (MRC-5, ATCC, CCL-171) and human lung adenocarcinoma cells (A549, ATCC, CCL-185) were used for cytotoxicity evaluations. The choice for pulmonary cells was based on their importance as host site infection of S. aureus and M. tuberculosis [26,31]. The cytotoxic assays were performed using protocol described by our group [32]. Compounds 3 and 4 displayed similar effects against both human cells, demonstrating values of IC 50 (concentration able to inhibit 50% of cell growth) ranging from 46.3 to 96.7 µM (Table 3). These compounds were moderately toxic against pulmonary cells when compared with doxorubicin which exhibited IC 50 values of 0.7 and 1.8 µM against MRC-5 and A549, respectively. In order to compare the effect towards prokaryotic cells versus eukaryotic cells, the selectivity indexes (SI) of 3 and 4 were calculated as the ratio of cytotoxicity to antimicrobial activities (SI = IC 50 eukaryotic cell/MIC prokaryotic cells) ( Table 3). The effectiveness of a compound is indicated by SI ě 10. However, SI ď 1 indicates toxicity of that compound for eukaryotic cells and low selectivity for the prokaryotic cells. Despite 3 and 4 demonstrating promising antibacterial and antitubercular activities, these compounds showed SI values ranging from 0.5 to 1.2. Antimicrobial agents with in vitro toxicity against human cells should not be ignored. Although SI values are relevant as a general orientation, this ratio cannot be used as the unique parameter to decide if a compound can continue to preclinical trials [33,34]. In this context, in vivo toxicity trials are highly indicated to capture the potential of 3 and 4 as valuable molecules.
The low selectivity of the cinnamylideneacetophenones 3 and 4 can be correlated with their phenolic nature. Studies report the ability of phenolic compounds to act on widespread radical processes shared by eukaryotic and prokaryotic cells [35]. We hypothesized this behavior may be responsible by simultaneous activity toward bacteria and pulmonary cells. Therefore, we evaluated the radical scavenging ability of 3 and 4 using DPPH (2,2-diphenyl-1-picrylhydrazyl) protocol [36,37]. Compounds 3 and 4 were evaluated at 1600, 800, 400, and 200 µM and exhibited all percentages of DPPH scavenging ability below 8% (Figure 1), suggesting their low efficacy on radical processes. Altogether, low selectivity of 3 and 4 cannot be strongly related to radical processes. Quercetin and gallic acid were used as positive controls due to their phenolic nature and potent radical scavenging ability [38,39]. These compounds were tested in the same concentrations used for compounds 3 and 4, demonstrating percentages of DPPH scavenging ability ě95%.  The low selectivity of the cinnamylideneacetophenones 3 and 4 can be correlated with their phenolic nature. Studies report the ability of phenolic compounds to act on widespread radical processes shared by eukaryotic and prokaryotic cells [35]. We hypothesized this behavior may be responsible by simultaneous activity toward bacteria and pulmonary cells. Therefore, we evaluated the radical scavenging ability of 3 and 4 using DPPH (2,2-diphenyl-1-picrylhydrazyl) protocol [36,37]. Compounds 3 and 4 were evaluated at 1600, 800, 400, and 200 µM and exhibited all percentages of DPPH scavenging ability below 8% (Figure 1), suggesting their low efficacy on radical processes. Altogether, low selectivity of 3 and 4 cannot be strongly related to radical processes. Quercetin and gallic acid were used as positive controls due to their phenolic nature and potent radical scavenging ability [38,39]. These compounds were tested in the same concentrations used for compounds 3 and 4, demonstrating percentages of DPPH scavenging ability ≥95%.

In Silico Prediction of ADME of Cinnamylideneacetophenones
Several reasons contribute to the attrition of drug candidates, including pharmacokinetic difficulties, insufficient efficacy, and adverse reactions in animal and humans. In the development phase of oral drugs, nearly 30% of drug candidates fail due to poor ADME properties [40]. Bioactive cinnamylideneacetophenones were also subjected to pharmacokinetic filters. An in silico study was performed for prediction of ADME properties using the Molinspiration property explorer toolkits [41]. In Table 4, we summarize the computed molecular properties of the selected compounds, such as absorption (%ABS), topological polar surface area (TPSA), lipophilicity (milog P), molecular weight (MW), number of hydrogen bond donors (HBD), number of hydrogen bond acceptors (HBA),

In Silico Prediction of ADME of Cinnamylideneacetophenones
Several reasons contribute to the attrition of drug candidates, including pharmacokinetic difficulties, insufficient efficacy, and adverse reactions in animal and humans. In the development phase of oral drugs, nearly 30% of drug candidates fail due to poor ADME properties [40]. Bioactive cinnamylideneacetophenones were also subjected to pharmacokinetic filters. An in silico study was performed for prediction of ADME properties using the Molinspiration property explorer toolkits [41]. In Table 4, we summarize the computed molecular properties of the selected compounds, such as absorption (%ABS), topological polar surface area (TPSA), lipophilicity (milog P), molecular weight (MW), number of hydrogen bond donors (HBD), number of hydrogen bond acceptors (HBA), violations of Lipinski's rule of five (Ro5), and number of rotatable bonds (NROTB). TPSA has been used as descriptor for characterizing absorption and passive transportation properties through biological membranes, allowing a good prediction of transport of candidate drugs in the intestines and through the blood-brain barrier (BBB). TPSA is sum of van der Waals surface areas of electronegative atoms (oxygen and nitrogen). Erlt and co-authors described TPSA values useful for prediction of drug transport properties [42]. Compounds with TPSA values in the range ď150 Å 2 have good intestinal absorption and ď90 Å have BBB penetration. All selected compounds exhibited TPSA values ď46.53 Å 2 . The four more potent antitubercular cinnamylideneacetophenones (2, 7, 10, and 18) displayed TPSA values ranging from 17.07 to 26.30 Å 2 , suggesting their efficient BBB penetration, which is a crucial property to combat tuberculous meningitis [43]. TPSA was used to calculate the percentage of absorption (%ABS) according to the equation: %ABS = 109´0.345ˆTPSA. All selected compounds demonstrated high %ABS, ranging from 92.9 to 100. The three most potent anti-Mycobacterium compounds (2, 10, and 18) exhibited a %ABS value of 100, which is relevant for their oral absorption, as well as for penetration through the highly hydrophobic mycolic acid layer of M. tuberculosis [44].
In order to achieve the optimum properties intended for oral administration, Lipinski and collaborators established a series of rules for drug candidates such as: (a) a log P less than 5, (b) a molecular weight less than 500 Da, (c) less than five hydrogen bond donors (sum of OH and NH groups), and (d) less than 10 hydrogen bond acceptors (sum of oxygen and nitrogen) [45]. Selected cinnamylideneacetophenones did not violate Lipinski's parameters, corroborating their oral drug likeness properties.
Among pharmacokinetic difficulties to drug candidates, low oral bioavailability has been considered a central reason for interrupting further drug development phases. In this context, Veber and collaborators suggested a filter for drug candidates, containing two essential criteria for oral bioavailability in rats: (a) TPSA equal or less than 140 Å 2 and (b) ten or fewer rotatable bonds [46]. All selected compounds exhibited NROTB ď5 and TPSA values ď46.53 Å 2 , making them potentially bioavailable by an oral route.

Synthesis of Cinnamylideneacetophenones
All starting reagents were purchased from Sigma-Aldrich. Compounds were synthesized by Claisen-Schmidt condensation, according to the method reported by Weldon and co-authors, with minor modifications [20][21][22]. Cinnamaldehyde (6 mmol) was added to solution of respective acetophenone (5 mmol) in ethanol (5 mL). The solution was stirred at room temperature for 30 min, followed by dropwise addition of 5 mL ethanolic solution of NaOH (1.0 M). The progress of reaction was checked by thin-layer chromatography (TLC) analyses. Reaction medium was poured into crushed ice. Precipitated crude products were filtered, washed with cold water, and dried at room temperature. Soluble crude products were partitioned with ethyl acetate (3ˆ25 mL). Organic phase was concentrated under reduced pressure. Compounds 2, 7, 8 and 10-7 were recrystallized in ethanol. Compounds 3-6, 9 and 18 were recrystallized in mixtures of hexane and chloroform.

Antibacterial Activity of Cinnamylideneacetophenones
Microorganisms used in this study were Staphylococcus aureus (ATCC 14458), Streptococcus mutans (ATCC 25175), Streptococcus sanguinis (ATCC 10557), Pseudomonas aeruginosa (ATCC 15442), and Escherichia coli (ATCC 10536). MIC and MBC values were determined by broth microdilution method, in 96-well microtiter plates, following the criteria previously described by the National Committee for Clinical Laboratory Standards [24]. Bacterial cell cultures in Brain Heart Infusion broth (BHI) at the mid-log phase were harvested by centrifugation for 10 min at 3000ˆg. Supernatant was discarded and the pellet resuspended in Mueller-Hinton broth (Difco Laboratories, Detroit, MI, USA). The final bacterial suspension was 5-10ˆ10 5 CFU (colony forming units) per mL. Compounds were diluted in sterile deionized water in concentrations ranging from 500 to 3.9 µM. Bacterial suspension was inoculated in each well. The plates were incubated at 37˝C for 24 h in 5% CO 2 . Thus, 15 µL of 0.01% resazurin staining was added to each well and incubated for 4 h to determine cell viability. Wells corresponding to MIC and at least three previous wells were homogenized, serially diluted, and plated on Mueller-Hinton agar to determine MBC. The plates were incubated at 37˝C for 24 h in 5% CO 2 . Number of colonies forming units per mL (CFU/mL) was determined. The MBC was considered when the compounds killed more than 99% of the tested bacterial culture. Penicillin G, vancomycin, chlorhexidine, and culture medium were used as positive and negative controls, respectively. Three independent assays were performed.

Antitubercular Activity of Cinnamylideneacetophenones
The antitubercular activity of compounds was determined by the REMA (Resazurin Microtiter Assay) method [25,26]. Stock solutions of compounds were prepared in DMSO (dimethyl sulfoxide) and diluted in Middlebrook 7H9 broth (Difco) supplemented with oleic acid, albumin, dextrose, and catalase (OADC), to obtain the final concentration range from 100 to 1 µM. Suspension of Mycobacterium tuberculosis H37Rv strain (ATCC 27294) was cultured in Middlebrook 7H9 broth supplemented with OADC. When the culture obtained a turbidity of McFarland standard No. 1, it was adjusted by 3-5ˆ10 5 CFU per mL. Thus, 100 µL of inoculum were added to each well in a 96-well microplate with 100 µL of the compound solution. Plates were incubated for 7 days at 37˝C and 5% CO 2 . Resazurin was added (30 µL of 0.01%). The fluorescence of the wells was measured after 24 h on Cytation 3 (BioTek, Winooski, VT, USA) equipment. MIC value was defined as the lowest concentration resulting in 90% inhibition of growth of M. tuberculosis. Isoniazid and culture medium were used as positive and negative controls, respectively. Three independent assays were performed.

Partition Coefficient (log P o/w ) of Cinnamylideneacetophenones
Values of log P o/w were determined using the HPLC method according to OECD-Guidelines for the Testing of Chemicals [30], on Shimadzu (Kyoto, Japan) HPLC-PDA (model CBM 20-A) equipped with a UV-Vis detector (model SPD-20A), binary pumping system (model LC-20AD), solvent degasser (model DGU-20A 3r ), and a Phenomenex Gemini C-18 column (Torrance, CA, USA) (250 mmˆ4.6 mm; 5 µm; 110 Å). Analysis was performed on isocratic mode in MeOH:H 2 O (3:1) and 1.0 mL/min. Injected volume of sample was 20.0 µL and the wavelength in the detector was adjusted according to λ max of each compound. Reference compounds phenol, acetophenone, benzoic acid, toluene, and thymol were used to construct the curve log Kˆlog P o/w . Capacity factor of the cinnamylideneacetophenones was determined from their retention times and interpolated in linearity curve log Kˆlog P o/w . Experiments were performed in duplicate.

Toxic Effect on Lung Cells of Selected Cinnamylideneacetophenones
Cytotoxicity of compounds was evaluated according to Silva and colleagues [32]. The cell lineages used were MRC-5 (human lung fibroblast cell line-ATCC, CCL-171) and A549 (human lung adenocarcinoma epithelial cell line-ATCC, CCL-185). Cells were routinely maintained in complete medium supplemented with 10% fetal bovine serum (FBS) plus amphotericin B (2 mg/L) and gentamicin (50 mg/L) at 37˝C in a humidified 5% CO 2 atmosphere. For cytotoxicity assay, 2.5ˆ10 4 cells/mL were seeded into a 96-well cell culture plate with a total volume of 100 µL for 24 h. Cells were treated with compounds at concentrations ranging from 100 to 1 µM. After 24 h incubation, medium was removed, and 50 µL of resazurin (0.01% w/v) was added to each well, and the plates were incubated at 37˝C for 3 h. Fluorescence of wells was measured on a Synergy H1 microplate reader (BioTek, Winooski, VT, USA) in which were used excitation (530 nm) and emission (590 nm) filters. The IC 50 values were calculated from a calibration curve by regression analysis. Doxorubicin and untreated cells constituted the positive and negative controls, respectively. Three independents assays were performed.

DPPH Scavenging Ability of Selected Cinnamylideneacetophenones
Radical scavenging activity of compounds 3 and 4 was determined using 2,2-diphenyl-1-picrylhydrazyl (DPPH) [36,37]. Aliquots of 10 µL of 3, 4, and positive controls (quercetin and gallic acid) at concentrations 1600, 800, 400, and 200 µM (in EtOH) were individually added in 190 µL DPPH (100 µM). The mixtures were incubated in the dark at 25˝C for 30 min. Remaining DPPH was determined spectroscopically at 490 nm by blanking against absolute ethanol. Radical scavenging ability was expressed by using mean values obtained from triplicates as the percentage of radicals reduced by calculating the following equation: DPPH scavenging % = [1´(A sample /A)ˆ100], where A is test absorbance without sample (only EtOH and DPPH) and Asample is test absorbance with compounds.

Conclusions
In summary, a series of cinnamylideneacetophenones containing several substituents on ring B were designed and synthesized as part of our ongoing search for antibacterial and antitubercular compounds based on natural products. Five compounds (3-6 and 9) exhibited potent antibacterial activity against Gram-positive species. Seven compounds (2-4, 7, 9, 10 and 18) were able to inhibit M. tuberculosis growth. Preliminary SAR investigations suggested hydrophilicity is the central parameter for antibacterial and antitubercular activities. Furthermore, our results corroborated the potential of cinnamaldehyde as a privileged starting material and template for synthetic collection of hits; it did not violate Lipinski's and Veber's boundaries.