2025 年 73 巻 8 号 p. 707-712
Prealnumycin B is an aromatic polyketide characterized as a reduced form of prealnumycin with a 6S-alcohol. As an extension of our earlier work on the synthesis of prealnumycin, the C7–C8 double bond was hydrogenated using Wilkinson’s catalyst, and the C6 carbonyl group was stereoselectively reduced via asymmetric transfer hydrogenation with the (S,S)-Ts-DENEB catalyst, where the use of methanol as a solvent was crucial for achieving high regioselectivity. By comparing its spectroscopic data with those of the natural isolate, the absolute configuration of prealnumycin B was determined to be 1R,6S. In addition, the antibacterial activities of prealnumycin B and related compounds were evaluated.
Aromatic polyketides serve as a versatile platform for drug discovery, featuring diverse structural motifs such as glycosides.1) Among them, alnumycin possesses a benzoisochromanequinone skeleton containing a structurally rearranged d-ribose attached through a carbon–carbon bond2,3) (Fig. 1). In connection with our synthetic study on alnumycin-class antibiotics, we recently reported the stereoselective synthesis of the alnumycin aglycon,4) prealnumycin (1).5) A key feature of this synthesis was the highly diastereoselective introduction of an n-propyl group at the C1 position, utilizing a stereochemically prearranged tricyclic lactone. In 2023, a structurally related aromatic polyketide, prealnumycin B (2), was newly isolated from the marine-derived Streptomyces sundarbansensis.6,7) The absolute configuration of compound 2 was determined by comparison with its calculated electronic circular dichroism (ECD) spectrum, which suggested 1R and 6S, as depicted in the structure. To confirm the stereostructure and evaluate the antibacterial activity, we aimed to synthesize compound 2 and its C6-epimer (3) from a known intermediate reported in our previous synthesis of 1, using an asymmetric transfer hydrogenation approach.
The retrosynthetic analysis of prealnumycin B (2) is illustrated in Chart 1. To construct the C6-stereogenic center, a regioselective asymmetric transfer hydrogenation of diketone 4 would be performed at the final stage of the synthesis. The C3–C4 double bond would be introduced by dehydration of compound 5. Hydrogenation of the C7–C8 double bond would be carried out at the stage of naphthoquinone 6. This tricyclic intermediate can be assembled from phthalide 7 and optically active enone 8 through an anionic annulation.8–10) As an alternative route, diketone 4 can be directly obtained by reduction of prealnumycin (1).
Our synthesis began with the preparation of the common intermediate 4 from either the known compound 6 or prealnumycin (1) (Chart 2). Hydrogenation of 6 using Wilkinson’s catalyst afforded 1,4-diketone 5 in 92% yield (toluene, room temperature [r.t.]).11) Subsequent treatment with a large excess of trifluoroacetic acid (TFA) led to the stepwise removal of both the benzyl and the methoxymethyl groups, furnishing diol 9 in 71% yield (dichloromethane [CH2Cl2], 0°C → r.t.).12) Dehydration of 9 with trifluoromethanesulfonic anhydride (Tf2O) and pyridine afforded compound 4 in moderate yield (CH2Cl2, 0°C).13) During the course of the reaction, 4 was found to be prone to oxidation to the corresponding quinone via double enolization of the two carbonyl groups, followed by air oxidation. As a result, prolonged reaction times led to a considerable loss of product.
On the other hand, hydrogenation of prealnumycin (1) under the same conditions also proceeded smoothly, selectively reducing the C7–C8 double bond to afford compound 4 in 96% yield (Chart 3).
The stage was set for the investigation of asymmetric transfer hydrogenation (Chart 4). Treatment of diketone 4 with an azeotropic mixture of formic acid and triethylamine as the reductant, in the presence of (S,S)-Ts-DENEB as the catalyst,14) yielded regioisomeric alcohols 2 and 10.15) We found that the choice of solvent was crucial for achieving high regioselectivity (Table 1). When the reaction was carried out in acetonitrile (MeCN), the alcohols were obtained in 62% combined yield with poor selectivity (entry 1).16) The use of tetrahydrofuran provided moderate regioselectivity (entry 2), while ethyl acetate (EtOAc) improved the combined yield without compromising selectivity (entry 3). In contrast, N,N-dimethylformamide resulted in almost no regioselectivity (entry 4). Finally, methanol as a protic solvent proved to be the most effective, affording the alcohols in 95% combined yield with high regioselectivity (1 : 0.11), where the non-hydrogen-bonded C6 carbonyl group was selectively reduced (entry 5). After separation of the alcohols by silica gel preparative TLC eluted with hexane–acetone (7 : 3), the absolute stereochemistry of alcohol 2 was determined by the Mosher–Kusumi method,17) confirming the 6S configuration.18,19) Note that prealnumycin B (2) was obtained as a single diastereomer, as confirmed by 13C-NMR analysis in comparison with compound 3, as described later.20)
| Entry | Solvent | Yield (%)b) | Ratio (2 : 10)c) |
|---|---|---|---|
| 1 | MeCN | 62 | 1 : 0.77 |
| 2 | THF | 63 | 1 : 0.54 |
| 3 | EtOAc | 88 | 1 : 0.57 |
| 4 | DMF | 78 | 1 : 1.09 |
| 5 | MeOHd) | 95 | 1 : 0.11 |
a) Reaction conditions: 4 (0.05 mmol), (S,S)-Ts-DENEB (10 mol%), HCO2H–NEt3 (280 μL), and solvent (2 mL) at 0°C. b) Combined yield of 2 and 10. c) The ratios for 2 : 10 were determined by 1H-NMR. d) The reaction was performed at room temperature for 2 h.
With the optimized conditions in hand, we also synthesized C6-epi-prealnumycin B (3) for spectroscopic and bioactivity comparison (Chart 5). Upon treatment of 4 with a catalytic amount of (R,R)-Ts-DENEB, the transfer hydrogenation proceeded, giving alcohols 3 and 11 in 98% combined yield with 1 : 0.14 selectivity. Each compound was characterized after separation by silica gel preparative TLC.
The UV–Vis absorption and ECD spectra of compounds 2 and 3, measured in MeOH (26 μg/mL), are shown in Fig. 2. The absorption spectra of both compounds showed no significant differences, exhibiting three maximum peaks at approximately 221, 258, and 353 nm (Fig. 2A). In contrast, the ECD spectra were clearly distinguishable in the range of 200–250 nm, although a positive and a negative Cotton effect were commonly observed around 353 and 258 nm, respectively (Fig. 2B). In comparison with the ECD spectra of the natural sample,6) the absolute stereochemistry of prealnumycin B (2) was confirmed as 1R,6S as originally reported.
The antimicrobial activities of prealnumycin-related compounds 1–3, 10, and 11 were evaluated against both Gram-positive and Gram-negative bacterial strains using the microdilution method (Table 2). Overall, the compounds exhibited moderate antibacterial activity against Gram-positive bacteria.21) In particular, compounds 2 and 3 showed relatively higher activity, suggesting that the hydroxy group at the C6 position plays an important role in their antibacterial effect. None of the compounds showed significant activity against Gram-negative strains, including Escherichia coli, both normal and drug-resistant strains of Pseudomonas aeruginosa, and Klebsiella pneumoniae. Notably, compound 2 exhibited measurable activity against both methicillin-resistant Staphylococcus aureus (MRSA) and multidrug-resistant (MDR) Acinetobacter baumannii, which are major clinical concerns, although the overall potency remains moderate.
| MIC (μg/mL) | |||||
|---|---|---|---|---|---|
| Compounds | B. subtilis | S. aureus | MRSA | M. smegmatis | MDR A. baumannii |
| 1 | 12.5 | 100 | — | 100 | — |
| 2 | 25 | 25 | 50 | 25 | 25 |
| 3 | 25 | 25 | 100 | 25 | 50 |
| 10 | 50 | 50 | 100 | 50 | 50 |
| 11 | 100 | 100 | — | 50 | 100 |
Blank indicates inactive.
We have accomplished the first total synthesis of prealnumycin B (1) and evaluated its antimicrobial activity. The key step in the synthesis involved an asymmetric transfer hydrogenation performed in a regio- and stereoselective manner. Further investigations toward the synthesis of alnumycin-class natural products22) are currently underway in our laboratory.
All experiments dealing with air- and moisture-sensitive compounds were conducted under an atmosphere of dry argon. Toluene, CH2Cl2, MeCN, and pyridine (dehydrated; Kanto Chemical Co., Inc., Tokyo, Japan) were used as received. For TLC analysis, Merck pre-coated plates, Darmstadt, Germany (silica gel 60 F254, Art 5715, 0.25 mm) were used. Flash column chromatography was performed using silica gel 60 N (spherical, 63–210 μm, Kanto Chemical Co., Inc.) and preparative TLC was performed on Merck PLC silica gel 60 F254 pre-coated plates (1 mm). Optical rotations were measured with a JASCO P-2300 polarimeter. IR spectra were recorded using a JASCO FT/IR-4200 spectrophotometer, JASCO, Tokyo, Japan. Circular dichroism (CD) spectra were recorded on a JASCO J-1100 spectrometer. UV–Vis spectra were measured on a JASCO V-650 spectrophotometer. 1H-NMR and 13C-NMR spectra were measured on a Bruker (Billerica, MA, U.S.A.) AVANCE III HD-500 (500 MHz/125 MHz). High-resolution MS (HRMS) were recorded with a JEOL JMS-700 (electron ionization), or Waters SYNAPT G2-Si HDMS (electrospray ionization [ESI]).
Diketone 5A mixture of quinone 6 (420 mg, 0.968 mmol) and RhCl(PPh3)3 (46.0 mg, 0.0497 mmol) in toluene (12 mL) was stirred under a hydrogen atmosphere (balloon) at r.t. for 12 h. After filtration through a Celite pad (CH2Cl2), followed by concentration, the residue was purified by silica-gel column chromatography (hexane/acetone = 9/1) to give diketone 5 (389 mg, 92%) as a colorless oil.
Rf 0.37 (hexane/EtOAc = 7/3);
To a solution of diketone 5 (440 mg, 1.00 mmol) in CH2Cl2 (4.0 mL) was slowly added TFA (2.0 mL) at 0°C. After stirring for 10 h at r.t., the mixture was diluted with H2O and CH2Cl2. The products were extracted with CH2Cl2 (×3), washed with saturated aqueous NaHCO3 and brine, and then dried over Na2SO4. After concentration, the residue was purified by silica-gel column chromatography (hexane/EtOAc = 8/2) to give diol 9 (215 mg, 71%) as a yellow oil.
Rf 0.29 (hexane/EtOAc = 7/3);
To a solution of diol 9 (158 mg, 0.518 mmol) in CH2Cl2 (17 mL) were successively added pyridine (188 μL, 2.33 mmol) and Tf2O (170 μL, 1.04 mmol) at 0°C. After stirring for 1.5 h at 0°C, the reaction was quenched with saturated aqueous NaHCO3, and the products were extracted with CH2Cl2 (×3). The combined extracts were washed with brine and dried over Na2SO4. After concentration, the residue was purified by silica-gel column chromatography (hexane/EtOAc = 95/5 → 9/1) to give diketone 4 (88.1 mg, 59%) as a red amorphous solid.
Rf 0.28 (hexane/EtOAc = 9/1);
A mixture of quinone 1 (24.9 mg, 0.0875 mmol) and RhCl(PPh3)3 (4.7 mg, 0.0051 mmol) in toluene (4 mL) was stirred under a hydrogen atmosphere (balloon) at r.t. for 18 h. After filtration through a Celite pad (CH2Cl2), followed by concentration, the residue was purified by silica-gel column chromatography (hexane/EtOAc = 95/5 → 9/1) to give diketone 4 (24.0 mg, 96%) as a red amorphous solid.
Prealnumycin B (2)To a solution of diketone 4 (14.5 mg, 0.0506 mmol) in MeOH (2.1 mL) were successively added (S,S)-Ts-DENEB (3.5 mg, 0.0054 mmol) and a HCO2H–triethylamine azeotropic mixture (280 μL) at 0°C. After stirring for 2 h at r.t., the mixture was diluted with H2O and EtOAc. The products were extracted with EtOAc (×3). The combined extracts were washed with brine and dried over Na2SO4. After concentration, the residue was purified by silica-gel column chromatography (hexane/acetone = 8/2) to give a mixture of 2 and 10 (13.9 mg, 95% combined yield, 1 : 0.11 selectivity). Each analytically pure sample was collected using silica-gel preparative TLC (hexane/acetone = 75/25).
Orange solid; Rf 0.23 (hexane/EtOAc = 7/3);
Orange solid; Rf 0.19 (hexane/EtOAc = 7/3);
To a solution of diketone 4 (35.2 mg, 0.123 mmol) in MeOH (5.2 mL) were successively added (R,R)-Ts-DENEB (5.4 mg, 0.0083 mmol) and a HCO2H–triethylamine azeotropic mixture (670 μL) at 0°C. After stirring for 2 h at r.t., the mixture was diluted with H2O and EtOAc. The products were extracted with EtOAc (×3). The combined extracts were washed with brine and dried over Na2SO4. After concentration, the residue was purified by silica-gel column chromatography (hexane/acetone = 8/2) to give a mixture of 3 and 11 (34.9 mg, 98% combined yield, 1 : 0.14 selectivity). Each analytically pure sample was collected using silica-gel preparative TLC (hexane/acetone = 75/25).
Orange solid; Rf 0.23 (hexane/EtOAc = 7/3);
Orange solid; Rf 0.19 (hexane/EtOAc = 7/3);
Bacillus subtilis NBRC 13719, S. aureus NBRC 100910, Mycobacterium smegmatis NBRC 13167, E. coli NBRC 102203, P. aeruginosa NBRC 106052, and K. pneumoniae NBRC 3512 were purchased from the National Institute of Technology and Evaluation, Biological Resource Center (Chiba, Japan). MRSA ssp. aureus derived from ATCC 33592 was purchased from Microbiologics (MN, U.S.A.). Drug-resistant P. aeruginosa BAA 2108 and multidrug-resistant A. baumannii BAA 1796 were obtained from the American Type Culture Collection (VA, U.S.A.). Each strain was maintained on Mueller Hinton Broth plates (Becton, Dickinson and Co., NJ, U.S.A.), at 35°C. Each strain was inoculated into Mueller Hinton Broth II and incubated for 12 h, and the bacterial density was adjusted to 0.5–1 McFarland standard. The adjusted microbial culture was further diluted 500 times and added to a 96-well plate (100 μL/well). Stock solutions of the test samples were prepared at 10 mg/mL in dimethyl sulfoxide and were then diluted to each desired concentration in 96-well plates containing microbial culture. After 24 h of incubation, the antibacterial activity of the compounds was determined by measuring the turbidity of the medium. All assays were performed in triplicate. Ampicillin (Nacalai Tesque, Kyoto, Japan), oxacillin, ciprofloxacin, and amikacin (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) were used as positive controls. The MIC values of each positive control against the bacterial strains were as follows: B. subtilis (ampicillin 0.16 μg/mL), S. aureus (ampicillin 0.16 μg/mL), MRSA (oxacillin 32 μg/mL), M. smegmatis (ampicillin 128 μg/mL), E. coli (ampicillin 4 μg/mL), P. aeruginosa (ampicillin 256 μg/mL), multi-drug resistant P. aeruginosa (MDRP) (ciprofloxacin 1 μg/mL), K. pneumoniae (ampicillin 128 μg/mL), and MDR-A. baumannii (amikacin 16 μg/mL).
This research was supported by a Grant-in-Aid for Scientific Research (C) (20K06955) from JSPS and a SUNBOR Grant from the Suntory Foundation for Life Sciences.
The authors declare no conflict of interest.
This article contains supplementary materials.
