Syntheses and antibacterial activities of 4 linear nonphenolic diarylheptanoids

Four linear nonphenolic diarylheptanoids were synthesized and their antibacterial activities were studied. ( S )-2-Me-CBS-catalysed reduction of alnustone with BH3SMe2 gave ( R )(-)(4 E ,6 E )-1,7-diphenylhepta-4,6-dien-3-ol, a natural product. Reduction of alnustone with Na in t -BuOH at –15 °C under N3 atm gave (E)-1,7-diphenylhept-5- en-3-one as a Birch-type reduction product. t-BuOK catalysed condensation of benzalacetone with propionyl chloride gave (4 Z ,6 E )-5-hydroxy-1,7-diphenylhepta-4,6-dien-3-one, a natural product. (1 E ,4 Z ,6 E )-5-Hydroxy-4-phenethyl-1,7-diphenylhepta-1,4,6-trien-3-one, a curcuminoid, was synthesized starting from pentan-2,4-dione in 3 steps. The synthesized chemical compounds were applied against 2 gram-positive bacteria ( Bacillus cereus and Arthrobacter agilis ), 4 gram-negative bacteria ( Pseudomonas aeruginosa , Xanthomonas campestris , Klebsiella oxytoca , and Helicobacter pylori ), and 1 yeast (Candida albicans) by the disc diffusion method. All of the synthesized compound exhibited different degrees of antimicrobial activity at concentrations between 20–100 μg/disc against the test organisms.

Alnustone (1) itself was synthesized from several starting materials using different synthetic methods [18][19][20][21]. In our previous studies we described an efficient method for the preparation of alnustone (1) itself [22] and alnustone-like compounds [23], and their antitumor activities against estrogen receptor alpha-positive human breast cancer.
To the best of our knowledge, the enantioselective synthesis of R -(2) and total synthesis of compound 3 are not known in the literature. In this study we report the first enantioselective synthesis of compound R -(2), its enantiomer S -(2), and 3 via chemoselective reduction of alnustone (1). Moreover, we report an efficient one-pot synthesis of compound 4 and an attempted synthesis of compound 5 ( Figure 1).
As mentioned above, efficient synthesis of alnustone (1) was previously reported by our research group [22,23].
If one considers the relationship between alnustone (1) and ( R) -2 or ( S) -2, it can be readily seen that ( R) -2 and ( S) -2 could be obtained from enantioselective reduction of the carbonyl group of alnustone (1). Chiral oxazaborolidines have been used as reagents or catalysts (with borane sources) for enantioselective reduction of prochiral ketones [26]. In this manner, Corey et al. [27] efficiently used ( S) -2-methyl-CBS-oxazaborolidine as a catalyst with BH 3 in the enantioselective reduction of aryl ketones to give the corresponding chiral alcohols with predominant R configuration (ee~84-96.7%). We proposed that alnustone (1) itself may be successively reduced to give the corresponding alcohol ( R)-2 by following this methodology. Indeed, while ( S)-2-Me-CBS-catalysed reduction of alnustone (1) gave alcohol ( R) -2 (ee 90%), ( R) -2-Me-CBS-catalysed reduction gave alcohol ( S) -2 (ee 90%) in good yield (80%) where the ee values were determined by chiral HPLC. The absolute configuration of alcohol ( R) -2 was determined by agreement of its specific rotation with data given in the literature for natural ( R) -2 [8]. The opposite specific rotation obtained for alcohol ( S)-2 also confirmed its structure. The NaBH 4 reduction [25] of alnustone (1) also gave racemic-2 in 80% yield ( Figure 2).  confirmed the reduction of the carbonyl group. Both 1 H-NMR and 13 C-NMR spectra of ( S)-2 and ( R)-2 were in agreement with the data given for ( R) -2 in the literature [5,9]. Thus, we easily synthesized enantioselectively the natural compound ( R)-2 and its enantiomer ( S) -2 ( Figure 2).
Mechanistically, the enantioselective formation of ( R) -2 and ( S) -2 is in good agreement with the results reported by Corey et al. [27]. As seen in Figure 3, both the ethyl phenyl ketone and alnustone (1) are reduced by hydride transformation from si faces.
We proposed that synthesis of compound 3 from alnustone (1) could be performed by a Birch reaction, which reduces 1,3-butadienes to 2-butenes. However, classical Birch reduction of alnustone (1) with metallic lithium in liquid NH 3 yielded a mixture with many undesirable reduction products. Therefore, instead of the classical Birch reduction conditions, the modification by Menzek et al. [28,29] was applied to obtain 3. This modification successfully allowed reduction to the conjugated 1,3-diene moiety of aromatic compounds using metallic lithium or sodium as the reducing agent in Et 2 O and t -BuOH under NH 3 (gas). Thus, reduction of alnustone (1) with 2.5 molar eq. Na as a reducing agent afforded compound 3 in a yield of 85% via Menzek's method ( Figure 4).  Hz, J 4,5 = 6.3 Hz, J 6,7 = 6.2 Hz). In this context, H 2 C(4) at δ 3.17 and H 2 C(7) at δ 3.43 also resonated as doublets. The spectral values belonging to H 2 C(4) and H 2 C(7) of compound 3 are in good agreement with allylbenzene and 4-penten-2-one, 2 similar structures. In the structures, the CH 2 of 4-penten-2-one resonates at δ 3.19 ppm with J = 6.9 Hz [30], and the CH 2 of allylbenzene resonates at δ 3.33 ppm with J = 6.7 Hz [31] ( Figure 5). In double-resonance experiments, irradiation of both olefinic protons together, HC(5) and HC(6), converted doublets of H 2 C(4) and H 2 C(7) to singlets, which supports the CH 2 -CH =CH-CH 2 structure.
H 2 C(1) at δ 2.96 and H 2 C(2) at δ 2.81 also resonated as triplets. The HR-ESI-MS spectrum of compound 3 was also in good agreement with the structure of compound 3 (C 19 H 20 NaO: 287.1418). HMBC correlations of compound 3 are given in Figure 6.
However, the 1 H-NMR data of the heptenone part of compound 3 were in full disagreement with the reported data of Zhang et al. [12]. H/C assignments of the heptenone skeleton for synthetic compound 3 compared with data of the isolated 3 are given in Table 1. For example, they reported 1 H-NMR data for  H 2 C(4) at δ 2.74 (d, J = 2.9 Hz, 2H) and H 2 C(7) at δ 2.77 (d, J = 2.9 Hz), which are also in disagreement with 4-penten-2-one and allylbenzene.
Nonetheless, for compound 3 isolated by Zhang et al. [12], the presence of 2 phenyl rings, 1 carbonyl, 4  Thus, we synthesized compound 3 starting from alnustone (1) in a 1-step reaction, which is the first reported synthetic method for compound 3.
For the simple synthesis of compound 4, we modelled a reaction described by Krishnamurty and Ghosh [32] for synthesis of dihydrocurcumin, in which they synthesized dihydrocurcumin via C-acylation of O-protected vanillylidene acetone with O-protected dihydroferuloyl chloride. We applied this methodology with a few modifications by treatment of benzalacetone (6) with KOBu t at 0°C and reacting with 3-phenylpropanoyl chloride to give compound 4 in a yield of 60% (Figure 7). In the 1 H-NMR spectrum of compound 4 the C 7 chain has an AB system belonging to HC(7) at δ 7.60 and HC(6) at δ 6.46, a singlet belonging to HC(4) at δ 5.64, and an A 2 B 2 system belonging to H 2 C(2) at δ 2.74 and H 2 C(1) at δ 3.00. The 1 H-NMR and 13 C-NMR spectra are in agreement with the data given for natural product 4 [13].
If we consider the similarity between compounds 4 and 5, we readily see that 5 is the alkylated derivative of compound 4. Therefore, we proposed that we could readily obtain 5 by alkylating 4 with 2phenylethylbromide. However, alkylation of 4 to give 5 with bases such as LDA, NaH, and KOBu t was not successful. In another attempted synthesis, we proposed that an alkylation of 6 with PhCH 2 CH 2 Br and then acylation with 3-phenylpropanoyl chloride could give compound 5. However, the first step of this attempted synthesis afforded styrene, an elimination product, instead of the expected ( E)-1,6-diphenylhex-1-en-3-one. Therefore, we decided to apply a different strategy, where 2,4-pentanedione was first condensed with 2phenylacetaldehyde in the presence of AcOH and pyrrolidine to give 8 as a sole compound (Figure 8). Catalytic hydrogenation of 8 gave 9 as a keto-enol tautomer. Nichols et al. [33] developed a practical methodology for solvent-free microwave-assisted synthesis of curcumin analogues based on B 2 O 3 -mediated condensation of 2,4-pentanedione with benzaldehydes in the presence of morpholine and AcOH. Applying this methodology to 9 with benzaldehyde gave curcuminoid 10.
Curcuminoid 10 and natural product 5 differ from each other with a hydrogenated double bond. Therefore, we applied selective reduction of the double bonds of compound 10. Changtam et al. [34] reported an easy procedure for preparing dihydrocurcumin by reduction of curcumin with Zn in AcOH. However, applying this methodology to curcuminoid 10 gave cyclic compound 11. We proposed that this cyclization might occur with Zn(OAc) 2 formed in situ in the reaction medium. Indeed, an independent reaction performed with Zn(OAc) 2 supported this hypothesis with the formation of 11 from 10 in good yield (70%) (Figure 8). However, attempted selective reduction of compound 10 with Na or Mg to give 5, failed. by disc diffusion method. The antimicrobial activity of the chemicals was variable, as seen in Table 2. All the synthesized chemicals showed different degrees of antimicrobial activity at concentrations of 30-90 µg/disc against the test organisms. Each value is expressed as mean (n = 3). Inhibition zone was greater than 7 mm. These evaluations were carried out by diffusion disc tests.
Compound 3 was highly antibacterial against K. oxytoca, A. agilis, B. cereus, and H. pylori. Additionally, 3 and 1 were observed to have stronger inhibitory effects against C. albicans. Among the compounds, 4 was determined as the least effective against the microorganisms. Ofloxacin and nystatin are well known as antibacterial and antifungal substances, respectively. The average inhibition zones of ofloxacin and nystatin were found to be 22 and 20 mm, respectively.
The results of this study indicate that the synthesized chemicals showed dose-dependent inhibition against the tested microorganisms. This result may suggest that alnustone (1) and dihydroalnustone (3) structures can be developed for new antibiotic drugs.

Experimental
THF was used by distillation over Na. 1 H-and 13 C-NMR spectra were recorded with 400 (100) MHz Bruker and Varian instruments. Interchangeable hydrogens or carbons were shown with the same letters. Elemental analyses were performed with a LECO CHNS-932. HRMS spectra were recorded with an Agilent 6530 LC-MS QTOF. Enantiomeric excesses were determined by HPLC analysis using a chiral column eluting n-hexane-i-PrOH (90:10), and detection was performed at 210-254 nm. Optical rotations were measured on a Bellingham Stanley ADP220 589-nm spectropolarimeter. All percent yields were calculated from isolated compounds.
A plastic balloon was connected to 1 neck of the flask and the other neck was fastened to a supply of NH 3 (gas). The solution was deoxygenated by the passing of 15 L of NH 3 (gas) and kept under NH 3 atmosphere by the filled balloon. The mixture was cooled to -15°C, and then metallic Na (20 mg, 0.95 mmol) was added.
At the same temperature, the reaction mixture was stirred for approximately 4 min, and then t-BuOH (70 mg, 0.95 mmol) was added and the reaction mixture was stirred for 16 min. At this stage, the colour of the reaction mixture changed to dark green as monitored by TLC and the completion of the reaction was observed. EtOH (15 mL) was added to quench the reaction mixture and then all solvents were removed under reduced pressure.
The residue was dissolved in H 2 O (5 mL) and extracted with AcOEt (4 ×15 mL). The combined organic phase was dried on Na 2 SO 4 and filtered. Evaporation of the solvent gave a crude product, which was then
The 1 H-NMR and 13 C-NMR data are in agreement with the data given for isolated natural product 4 [13].