For column chromatography, Sephadex LH-20 (GE Healthcare, Chicago, IL, USA) was used. Middle pressure liquid chromatography (MPLC) for isolation of compounds, HPLC pump SSC-3160 (Senshu Science, Tokyo, Japan), RI-UV detector YRU-883 (Yamazen, Osaka, Japan), glass column Ultra pack ODS-S-50B (26 × 300 mm, Yamazen), and chromatographic conditions as follows: flow rate: 8 mL/min, detected by RI-UV detector. HPLC for isolation of compounds, HPLC pump LC-20AT (Shimadzu, Kyoto, Japan), UV detector SPD-20A (Shimadzu), column oven CO-965 (JASCO, Tokyo, Japan) and HPLC column Inertsustain C18 (5 μM, 10 × 250 mm), and chromatographic conditions as follows: flow rate: 2 mL/min, detected wavelength: 210 nm. NMR spectra were measured by ECAII 600 spectrometer (JEOL, Tokyo, Japan). The chemical shifts (δ) are given in ppm, and the coupling constants (J) in Hz. Optical rotation measured by DIP-10000 polarimeter (JASCO). Microplate reader was used as MPR-A100 (AS ONE, Osaka, Japan). ESI-MS were measured at negative mode on T-100LP mass spectrometer (JASCO).

The strain Commelinaceomycesaneilematis MAFF246963 isolated from Murdanniakeisak in Japan [14]. This fungus was identified based on morphology and a sequence analysis of the 28S, 18S, TEF, RPB1 and RPB2 regions (GenBank accession No. LC474620, LC474617, LC474623, LC474626 and LC474629). The fungus was maintained on potato dextrose agar medium before use.

Human hepatoma cells (HepG2) were obtained from the National Institutes of Bio-medical Innovation, Health, and Nutrition (Osaka, Japan). Human promyelocytic leukemia cells (HL60), human pancreatic carcinoma cells (PANC-1), and human glioblastoma multiforme tumor cells (T98G) were obtained from Riken BioResource Research Center (Ibaraki, Japan). Human colon adenocarcinoma cells (HT29) were obtained from KAC Co., Ltd. (Kyoto, Japan). Cells were cultured in Dulbecco’s modified Eagle medium (DMEM: Nacalai Tesque, Inc., Kyoto, Japan) or RPMI 1640 (Nacalai Tesque, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (Merck, Darmstadt, Germany) and 1% Antibiotic-Antimycotic Mixed Stock Solution (Nacalai Tesque, Inc.) at 37˚C in a humid incubator containing ambient air supplemented with 5% CO₂.

C.aneilematis was first cultivated on potato dextrose agar (PDA) medium at 25˚C for two weeks to prepare the inoculum. Rice medium (1800 g, with 150 g rice aliquoted into 12 Roux flasks) was pre-sterilized (120˚C, 20 min), and mycelial mats of the fungal inoculum were cut into 1 cm² squares with a sterile spatula, with about two pieces placed into each flask, and then statically cultured at 25˚C for three weeks. After cultivation, extraction with methanol was performed, yielding 40.3 g of methanol extract.

The obtained methanol extract was suspended in 300 mL of water and extracted three times with an equal volume of ethyl acetate. The resulting ethyl acetate extract was obtained by evaporation under vacuum, suspended in 300 mL of n-hexane, and extracted three times with an equal volume of acetonitrile (CH₃CN), yielding 2.96 g of hexane extract and 2.33 g of CH₃CN extract. Additionally, the H₂O fraction was partitioned with 1-BuOH and H₂O, yielding 12.50 g of 1-BuOH fraction and 28.56 g of H₂O fraction. All cultivation and extraction were performed at once.

The obtained MeOH extract and solvent fractions were used for cytotoxicity screening.

The cytotoxicity of the crude MeOH extract and solvent fractions was initially evaluated against human pancreatic carcinoma PANC-1 cells using five independent biological experiments by the WST-1 assay. Data are presented as mean ± SD (n = 5). Each sample was tested at 50 and 100 μg/mL. For crude fraction cytotoxicity tests, cells (T98G 0.2 × 10⁴ cell/well, PANC-1 2.0 × 10⁴ cell/well, HT-29 2.0 × 10⁴ cell/well, HL60 1.0 × 10⁴ cell/well, HepG2 1.0 × 10⁴ cell/well) were seeded at 100 µL per well into 96-well plates and incubated for 48 hours in a CO₂ incubator. For cells other than HL60, various test samples prepared at 50.0 µg/mL and 100 µg/mL were added by replacing the medium, and cells were cultured for 24 hours. For HL60, 5 µL of medium containing various test samples prepared at 1000 µg/mL and 2000 µg/mL were added, and cells were cultured for 24 hours. After 24 hours, 10 µL of Cell Proliferation Reagent WST-1 was added, and absorbance was measured after 4 hours (main wavelength: 450 nm, sub-wavelength: 660 nm) using a microplate reader. The number of samples was 5, and Dunnett’s multiple comparisons test was performed to determine statistical significance. 5-Fluorouracil (5-FU) was used as a positive control.

The CH₃CN extract, which showed the strongest cytotoxicity against PANC-1 cells, was further evaluated against additional human cancer cell lines, including T98G, HT-29, HL-60, and HepG2 cells. This screening was performed to identify the active fraction containing cytotoxic metabolites.

Because the CH₃CN extract exhibited strong cytotoxicity, this extract was subjected to further chemical investigation. The CH₃CN extract was separated by repeated chromatographic procedures, including Sephadex LH-20 column chromatography, ODS-MPLC, and preparative HPLC.

Component isolation from the cytotoxic CH₃CN extract was performed using multi-stage chromatography. First, the CH₃CN extract (2.33 g) was chromatographed on LH-20 column chromatography with the mobile phase using 200 mL of n-hexane-chloroform (1:4), 200 mL of chloroform-acetone (3:2), 200 mL of chloroform-acetone (1:4), 200 mL of acetone and 500 mL of methanol to obtain 13 fractions.

The ninth fraction (168 mg) obtained from LH20 column chromatography was purified by ODS-MPLC (mobile phase: 60% acetonitrile) and silicagel-HPLC (mobile phase: chloroform-methanol 20:1) to get 3 (12.0 mg).

For the isolation of isochaetochromin B₂ (1), Fr. 7, obtained from LH20 column chromatography, was separated by ODS-MPLC using 80% acetonitrile with 0.1% HCOOH as the mobile phase. Fr. 6 from this ODS-MPLC was then separated and purified by ODS-HPLC using 70% acetonitrile with 0.1% HCOOH as the mobile phase, yielding 9.7 mg of isochaetochromin B₂ (1) [16][17].

For the isolation of ustilaginoidin D (2), Fr. 6 obtained from LH20 column chromatography, was separated by ODS-MPLC using 60% acetonitrile as the mobile phase, followed by ODS-HPLC using 80% acetonitrile with 0.02% trifluoroacetic acid (TFA) as the mobile phase, yielding 4.2 mg of ustilaginoidin D (2) [16][18].

White powder: [α]^D + 12.9˚ (c 0.485, MeOH). HR-ESI-MS (positive mode) m/z: 381.1212, calcd. 381.1186 for C₁₈H₂₁O₉, [M+H]⁺. The ¹H- and ¹³C-NMR data are shown in Section 3.4.

The cytotoxicity of compounds 1 - 3 was evaluated against PANC-1 and HT-29 cells, because these cell lines were sensitive to the CH₃CN extract. These cytotoxic assays were performed using four independent biological experiments by the WST-1 assay. Data are presented as mean ± SD (n = 4). The compounds were tested at several concentrations, and IC₅₀ values were calculated. Doxorubicin was used as a positive control.

For isolated compound cytotoxicity tests, PANC-1 5.0 × 10⁴ cell/well and HT-29 5.0 × 10⁴ cell/well were seeded at 100 µL per well into 96-well plates and incubated for 48 hours in a CO₂ incubator. After 48 hours, various compounds prepared at 3.125 µM, 12.50 µM, 25.00 µM, 50.00 µM, and 100 µM were added by replacing the medium, and cells were cultured for 24 hours. After 24 hours, the medium was replaced, and 10 µL of WST-1 reagent was added. Subsequently, absorbance was measured after 4 hours. Four independent biological experiments were performed. For each biological replicate, IC₅₀ values were independently estimated by linear regression analysis using the linear portion of the dose-response curve. The reported IC₅₀ values represent the mean ± standard deviation (SD) of four independent biological experiments.

Statistical analysis was performed using GraphPad Prism version 7.0e (GraphPad Software, San Diego, CA, USA) and multcomp package in R, version 4.1.0 [19]. Comparisons with the respective negative controls were conducted using Dunnett’s multiple comparisons test in GraphPad Prism. IC₅₀ values were independently estimated for each biological replicate by linear regression analysis using the linear portion of the dose-response curve and are presented as the mean ± SD. Values of p < 0.05 were considered statistically significant.

To investigate whether Commelinaceomycesaneilematis produces cytotoxic secondary metabolites, the MeOH extract obtained from rice cultures was partitioned into four solvent extractions (MeOH, Hexane, CH₃CN and 1-BuOH extract) (Figure 1(a)), and each extract was evaluated for cytotoxicity against PANC-1 cells.

Among the extracts examined, the CH₃CN extract exhibited the strongest cytotoxic activity, significantly reducing cell viability at 50 μg/mL (Figure 1(b)). Because this extract showed the highest activity, it was selected for further investigation.

The CH₃CN extract was subsequently evaluated against additional human cancer cell lines (T98G, HT-29, HL-60, and HepG2). Cytotoxicity was observed against all tested cell lines, with particularly strong activity against HT-29 cells (Figure 1(c)). These results suggested that the CH₃CN extract contained one or more cytotoxic metabolites responsible for the observed activity.

To identify the constituents responsible for the cytotoxicity, the CH₃CN extract was subjected to repeated chromatographic separation using silica gel column chromatography, Sephadex LH-20, MPLC, and preparative HPLC.

This procedure afforded a new isocoumarin glycoside (3), together with two

Figure 1. Cytotoxic effect of crude extract of C.aneilematis as assessed by WST assay. 5-FU was employed as positive control. Data are shown as means ± SD (n = 5). Comparisons to the negative control (fr.: 0 μg/mL, 5-FU: 0 μM) were performed by Dunnett’s multiple comparisons test. Conditions without an asterisk exhibited no significant difference. **p< 0.01, ***p< 0.001. (a) Culture and separation method. (b) Cytotoxicity against PANC-1 cells. (1) MeOH fr. (2) Hexane fr. (3) CH₃CN fr. (4) 1-BuOH fr. (c) Cytotoxicity of CH₃CN fr. against T98G, HT-29, HL-60 and HepG2 cells.

known fungal metabolites, isochaetochromin B₂ (1) and ustilaginoidin D (2) (Figure 2).

Figure 2.Structure of 1 - 3 isolated from C.aneilematis.

Compounds 1 - 3 were evaluated for cytotoxicity against PANC-1 and HT-29 cells.

Compound 3 did not exhibit cytotoxicity toward either cell line (IC₅₀ > 100 μM). In contrast, isochaetochromin B₂ (1) and ustilaginoidin D (2) showed significant cytotoxicity. Compound 1 inhibited the growth of PANC-1 and HT-29 cells with IC₅₀ values of 26.0 ± 0.4 and 33.4 ± 2.2 μM, respectively, whereas 2 exhibited IC₅₀ values of 12.0 ± 0.7 and 30.4 ± 0.4 μM, respectively. These IC₅₀ values are comparable to or even stronger than those of the positive control, doxorubicin (PANC-1: 24.7 ± 1.3 µM, HT-29: 36.0 ± 19.2 µM). The results of the cytotoxicity tests for the isolated compounds are shown in Table 1. These results indicate that the cytotoxicity observed in the CH₃CN fraction is likely attributable to isochaetochromin B₂ and ustilaginoidin D. Furthermore, the isolation of ustilaginoidin D, one of the characteristic metabolites of Villosiclavavirens, supports the close phylogenetic relationship between C.aneilematis and V. virens from a chemotaxonomic perspective.

Table 1. Cytotoxic effect of 1 - 3 as assessed by WST assay. Doxorubicin was employed as positive control. Data are shown as means ± SD (n = 4).

Among the isolated metabolites, compound 3 was identified as a new natural product. Compound 3 was obtained as a white powder. Its molecular formula was determined to be C₁₈H₂₀O₉ [m/z: 381.1212, [M+H]⁺, calcd. 381.1186 for C₁₈H₂₁O₉], by HRESIMS, indicating nine degrees of unsaturation. The ¹H and ¹³C NMR spectra data as shown in Table 2, together with HSQC, COSY, and HMBC analyses, revealed the presence of an isocoumarin glycoside. The HMBC correlation from the anomeric proton (H-1’) to C-4 established the attachment of the sugar moiety to C-4 of the isocoumarin nucleus, and the planar structure of 3 was thus established (Figure 3).

Figure 3.Structure elucidation of 3 by detailed analysis of 2D-NMR.

The sugar moiety was assigned as ribose based on the NMR spectroscopic data and comparison with literature values. Although the absolute configuration of the sugar was not determined experimentally, it was tentatively assigned as D-ribose because naturally occurring ribosylated secondary metabolites almost exclusively contain D-ribose [20]-[27]. The stereochemistry of the aglycone side chain could not be unambiguously determined because of insufficient spectroscopic evidence. Therefore, only the planar structure of 3 is proposed in the present study, and the absolute configuration of the aglycone remains to be elucidated by further investigation.

Table 2.¹³C- and ¹H-NMR spectral data of 3. Measurement solvent: CD₃OD. () indicates coupling constants.