Optical rotations were measured on a PerkinElmer MPC 500 (Waltham) polarimeter. The UV, IR, and CD spectra were recorded on a Shimadzu UV-2600 PC spectrometer (Shimadzu), an IR Affinity-1 spectrometer (Shimadzu), and a Chirascan circular dichroism spectrometer (Applied Photophysics), respectively. NMR spectra were recorded on a Bruker Avance spectrometer (Bruker) operating at 500 and 700 MHz for ¹H NMR and 125 and 175 MHz for ¹³C NMR that used tetramethylsilane as an internal standard. High resolution electrospray ionization mass spectroscopy (HRESIMS) spectra were acquired on a Bruker miXis TOF-QII mass spectrometer (Bruker). Column chromatography was performed over silica gel (200–300 mesh) (Qingdao Marine Chemical Factory). Spots were detected on TLC (Qingdao Marine Chemical Factory) under 254-nm UV light. All solvents employed were analytical grade (Tianjin Fuyu Chemical and Industry Factory). Semipreparative HPLC was performed using an octadecylsilyl (ODS) column (YMC-pack ODS-A, YMC Co., Ltd., 10 mm × 250 mm, 5 μm). Artificial sea salt was a commercial product (Guangzhou Haili Aquarium Technology Company).

The fungal strain Penicillium ludwigii SCSIO 41408 was isolated from a mangrove sediment sample, collected in the Hongsha River estuary to South China Sea, in Sanya city, Hainan Island. The fungus was identified according to the internally transcribed spacer (ITS) region sequence data of the rDNA, and the sequence was deposited in GenBank with the accession number OL823084. The strain was stored on Muller Hinton broth agar (malt extract, 15 g; sea salt, 10 g; agar, 15 g; H₂O, 1 L; pH 7.4–7.8) at 4°C and deposited in the CAS Key Laboratory of Tropical Marine Bioresources and Ecology, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China.

The strain Penicillium ludwigii SCSIO 41408 was cultured in the seed medium (malt extract, 15 g; sea salt, 10 g; H₂O, 1 L; pH 7.4–7.8) for 4 days at 28°C on a rotating shaker (180 rpm). A large scale of fermentation was incubated statically at 26°C for 60 days in 1 L × 65 conical flasks with a rice medium (each flack contains 180 g of rice, 3 g of sea salt, 200 ml of H₂O). The fermented cultures were overlaid and extracted with EtOAc three times to afford a brown extract (346.6 g).

The EtOAc crude extract was chromatographed over a silica gel column eluted with PE/CH₂Cl₂ (0–100%, v/v) and CH₂Cl₂/MeOH (0–100%, v/v) in a gradient to yield fourteen fractions (Frs. 1–14). Fr. 2 (0.9 g) was separated by semipreparative HPLC (65% MeCN/H₂O, 2 ml/min) to afford 9 (3.4 mg, t_R = 17.0 min) and Fr. 2-1. Fr. 2-1 was separated again by semipreparative HPLC (40% MeCN/H₂O, 2.7 ml/min) to afford 7 (6.7 mg, t_R = 16.1 min) and 8 (7.5 mg, 14.5 min). Fr. 4 (6.4 g) was separated by semipreparative HPLC (48% MeCN/H₂O, 3 ml/min) to afford 6 (2.1 mg, t_R = 17.3 min). Fr. 8 (8.5 g) was purified by semipreparative HPLC (65% MeOH/H₂O, 2 ml/min) to afford 2 (5.2 mg, t_R = 10.1 min). Fr.10 was divided into four subfractions by ODS silica gel chromatography eluting with MeOH/H₂O (10–100%). Fr. 10 was divided into four subfractions by Sephadex LH-20. Fr.10-1 was purified by semipreparative HPLC (35% MeCN/H₂O, 2 ml/min) to afford 11 (10.2 mg, t_R = 8.1 min) and 1 (6.0 mg, t_R = 27.1 min). Fr. 10-4 was purified by semipreparative HPLC (55% MeOH/H₂O, 3 ml/min) to afford 3 (6.0 mg, t_R = 7.5 min). Fr. 13 was divided into four subfractions by ODS silica gel chromatography eluting with MeOH/H₂O (10–100%). Fr. 13-1 was further purified by semipreparative HPLC (45% MeOH/H₂O, 2.7 ml/min) to afford 4 (26.0 mg, t_R = 12.4 min), 5 (5.7 mg, t_R = 14.1 min), and 10 (16.4 mg, t_R = 21.9 min).

Adametizine C (1): colorless needles; [α]25 D, −128.5 (c 0.1, MeOH); UV (MeOH) λ_max (log ε), 205 (3.52) nm; ECD (0.36 mM, MeOH) λ_max (Δε), 203 (−10.98), 242 (−9.06), and 293(+1.33); IRν_max, 3,352, 2,945, 2,835, 1,670, 1,506, 1,431, 1,250, 1,204, 1,096, 1,018, 831, 795, 677, 601, and 557 cm^–1; ¹H and ¹³C NMR, data see Table 1; HRESIMS at m/z 563.0381 [M + H]⁺ (calculated for C₂₁H₂₄ClN₂O₈S₃, 563.0378).

2-methyl-3-(5-oxohexyl) maleic acid (7): colorless oil; UV (MeOH) λ_max (log ε), 250 (3.95) and 207 (3.69) nm; IRν_max, 2,938, 2,866, 1,759, 1,712, 1,362, 1,273, 916, and 735 cm^–1; ¹H and ¹³C NMR data, see Tables 2, 3; HRESIMS at m/z 227.0927 [M-H]^– (calculated for C₁₁H₁₅O₅, 227.0925).

2-(4-hydroxyhexyl)-3-methylmaleic acid (8): colorless oil; [α]25 D, −0.7 (c 0.1, MeOH); ECD (2.46 mM, MeOH) λ_max (Δε), 203 (−2.49) and 213 (+2.27); UV (MeOH) λ_max (log ε), 250 (3.25), 206 (3.65) nm; IRν_max, 3,390, 2,943, 1,763, 1,682, 1,435, 1,283, 1,202, 1,136, 1,026, 800, and 721 cm^–1; ¹H and ¹³C NMR data, see Tables 2, 3; HRESIMS at m/z 229.1082 [M-H]^– (calculated for C₁₁H₁₇O₅, 229.1081).

3-(ethoxycarbonyl)-2-methylenenonanoic acid (9): brown oil; [α]25 D, +5.9 (c 0.1, MeOH); ECD (3.83 mM, MeOH) λ_max (Δε) 200 (−10.62), 214 (+0.68), and 227 (−2.64); UV (MeOH) λ_max (log ε), 204 (3.86) nm; IRν_max, 2,955, 2,928, 2,857, 1,732, 1,715, 1,202, 1,153, 1,113, 1,036, 953, and 835 cm^–1; ¹H and ¹³C NMR data, see Tables 2, 3; HRESIMS at m/z 243.1594 [M + H]⁺ (calculated for C₁₃H₂₃O₄, 243.1591).

7-hydroxy-3-(methoxycarbonyl)-2-methylenenonanoic acid (10): colorless oil; [α]D25, +1.9 (c 0.1, MeOH); ECD (2.05 mM, MeOH) λ_max (Δε), 200 (−3.34), 201 (+4.57), 205 (+2.35), and 211 (−0.98); UV (MeOH) λ_max (log ε), 205 (2.93) nm; IRν_max, 3,447, 2,936, 2,866, 1,717, 1,628, 1,456, 1,435, 1,204, 1,155, 1,024, 9,54.8, and 824 cm^–1; ¹H and ¹³C NMR data, see Tables 2, 3; HRESIMS at m/z 245.1388 [M + H]⁺ (calculated for C₁₂H₂₁O₅, 245.1384).

2-(4-hydroxypentyl)-4-methyl-5-oxo-2,5-dihydrofuran-3-carbo xylic acid (11): brown oil; [α]D25, +2.6 (c 0.1, MeOH); ECD (4.39 mM, MeOH) λ_max (Δε), 200 (−10.62), 214 (+0.68), and 227 (−2.64); UV (MeOH) λ_max (log ε), 225 (3.85) nm; IRν_max, 3,414, 2,932, 1,744, 1,715, 1,337, 1,231, 1,117, 1,024, 947, 764, and 721 cm^–1; HRESIMS at m/z 229.1076 [M + H]⁺ (calculated for C₁₁H₁₇O₅, 229.1072).

The crystallographic data of compound 1 obtained in MeOH were collected with a Rigaku XtaLAB PRO single-crystal diffractometer using Cu Kα radiation (λ = 1.54184). Briefly, their X-ray crystal structure was solved using SHELXS97, expanded by difference Fourier techniques, and refined by full-matrix least-squares calculation finally. The non-hydrogen atoms were refined anisotropically, and hydrogen atoms were fixed at calculated positions. The crystallographic data of compound 1 have been deposited in the Cambridge Crystallographic Data Centre.

Crystal Data for adametizine C (1): C₂₁H₃₁ClN₂O₁₂S₃, M_r = 635.11, crystal size 0.1 mm × 0.08 mm × 0.06 mm, orthorhombic, a = 9.16490 (10) Å, b = 11.25800 (10) Å, c = 27.3339 (2) Å, α = β = γ = 90°, V = 2407.9 (4) ų, Z = 4, T = 100.00 (10) K, Space group P2₁2₁2₁, μ = 3.837 mm^–1, ρ_calc = 1.496 g/cm³, 14,343 reflections measured (6.468° ≤ 2Θ ≤ 148.482°), 5,546 unique (R_int = 0.0275, R_sigma = 0.0325). The final R₁ values were 0.0287 [I > 2σ(I)]. The final wR (F²) values were 0.0758 [I > 2σ(I)]. The final R₁ values were 0.0299 (all data). The final wR (F²) values were 0.0764 (all data). The goodness of fit on F² was 1.049. The Flack parameter is 0.002 (6) (CDCC 2130918).

The antimicrobial activities against five bacteria (Erysipelothrix rhusiopathiae WH13013, Streptococcus suis SC19, Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and Staphylococcus aureus ATCC 25923) and four fungi (Botrytis cinerea, Septoria nodorum Berk., Fusarium graminearum Schw., and Rhizoctonia solani Kühn) were evaluated using the methods described previously (Wan et al., 2014). Cephalosporin and cycloheximide were used as positive controls against bacteria and fungi, respectively.

The inhibitory activities of LPS-induced NF-κB activation in RAW264.7 cells were evaluated as detected by luciferase reporter gene assay as described previously (Tan et al., 2020). In brief, the RAW264.7 cells stably transfected with a luciferase reporter gene were plated in 96-well plates and then pretreated with tested compounds (20 μM) and BAY11-7082 (NF-κB inhibitor as positive control, 5 μM, Sigma-Aldrich) for 30 min, followed by LPS stimulation (5 μg/ml) for 8 h. Cells were harvested, and luciferase activities of the triplicate tests were measured by the luciferase assay system (Promega, Madison, WI, United States). For further study of compounds 1, 5, and 9 on osteoclastogenesis, bone marrow macrophage cells (BMMCs) were added with macrophage-stimulating factor (50 ng/ml) and RANKL (100 ng/ml) stimulation at 5 μM concentrations for 3 days. Then, the cells were fixed and stained for TRAP activity, and the images were photographed by using an inverted microscope (Nikon, Japan). Data are expressed as the mean ± SD and analyzed using GraphPad Prism 7.0 software (San Diego, CA, United States). Statistical differences among groups were performed using one-way analysis of variance with Bonferroni post hoc test. A p-value of < 0.05 was considered statistically significant.

Cell viability was analyzed by 3-(4,5)-dimethylthiahiazo (-z-y1)-3,5-di-phenytetrazoliumromide (MTT) assay as previous described (Wang et al., 2021). In brief, cells were seeded in 96-well plate at a density of 5 × 10³ per well overnight and treated with compounds for demand time. OD₅₇₀ values were detected using a Hybrid Multi-Mode Reader (Synergy H1, BioTek). The experiment was repeated three times independently.

PC-3 cells were seeded in six-well plate at a density of 1,000 cell per well overnight, and then, cells were treated with dimethyl sulfoxide (DMSO) (0.1%, v/v), docetaxel (1 μM), and compound 3 (1.25, 2.5, 5, and 10 μM), respectively, for demand time. The cell clone colonies were formed after treating for 2 weeks, and cells were fixed with 4% formaldehyde for 30 min, washed with phosphate buffer saline (PBS) buffer, and then stained with crystal violet stain solution for 30 min. The dye solution was removed, and the cells were washed with PBS buffer again. Cell colonies were recorded and analyzed by the colony count analysis system (GelCount, Oxford Optronix). The experiment was repeated three times independently.

PC-3 cells were seeded in six-well plate at a density of 2.0 × 10⁵ cell per well overnight and treated with DMSO (0.1%, v/v), docetaxel (1 μM), and compound 3 (1.25, 2.5, 5, and 10 μM), respectively, for 48 h. Then, cells were collected and stained with annexin V–fluoresceine isothiocyanate (FITC) and propidium iodide (PI) solution following the manufacturer’s manual (BMS500FI-300, Thermo Fisher Scientific). The apoptotic rate of PC-3 cells was examined and analyzed by flow cytometer (NovoCyte, Agilent). Each experiment was repeated three times independently.

Compound 1 was obtained as colorless needles and had the molecular formula C₂₁H₂₃ClN₂O₈S₃ with 11 degrees of unsaturation as established by HRESIMS. The ¹H-NMR (Table 1) and HSQC experiment of 1 showed the typical pattern of an thiodiketopiperazines skeleton with three exchangeable protons, assigned to 14-OH (δ_H 9.64, s), 10-OH (δ_H 5.17, d, J = 6.6 Hz), and 6-OH (δ_H 6.11, d, J = 2.3 Hz), two aromatic methine protons [H-17 (δ_H 6.53, d, J = 8.8 Hz) and H-18 (δ_H 7.08, d, J = 8.8 Hz)], two olefinic protons [H-8 (δ_H 5.63, dt, J = 10.3, 2.5 Hz) and H-9 (δ_H 5.58, dt, J = 10.3, 2.0 Hz)], five methines [H-2 (δ_H 4.63, d, J = 1.1 Hz), H-7 (δ_H 4.88, m), H-10 (δ_H 4.43, m), H-11 (δ_H 4.10, dd, J = 7.3, 1.9 Hz), and H-12 (δ_H 5.39, s)], one methylene [H₂-5α (δ_H 2.54, dd, J = 14.8, 2.0 Hz) and H₂-5β (δ_H 2.09, d, J = 14.8 Hz)], one N-methyl [H₃-21 (δ_H 3.12, s)], and two O-methyl [H₃-19 (δ_H 3.69, s) and H₃-20 (δ_H 3.78, s)]. Besides the above 13 corresponding hydrogen-bearing carbons, eight non-protonated (with six sp² and two sp³) carbon atoms remained in the ¹³C NMR spectrum. Detailed analysis of the above NMR data and 2D NMR correlations (Figure 2 and Table 1) resulted in the elucidation of the planar structure of 1. Upon slow evaporation of the solvent MeOH, which was achieved by storing the sample in a refrigerator for 4 weeks, single crystals of adequate quality of 1 were obtained, making an X-ray diffraction study possible that could unequivocally confirm the chemical structure of 1. The absolute configuration was determined on the basis of measuring the anomalous dispersion effects by collecting Friedel pair reflections in the X-ray diffraction experiment (Figure 3).

Compound 7 was isolated as colorless oil, and its molecular formula was determined as C₁₁H₁₆O₅ with four degrees of unsaturation, as a deprotonated ion peak at m/z 227.0927 [M-H]^– in the HRESIMS spectrum. The NMR data of 7 (Tables 2, 3) indicated the presence of a ketone group (δ_C 208.8), two carboxylic groups (δ_C 166.7, 166.9), one double bond (δ_C 143.7, 141.5), two methyl group (δ_C 30.2, 9.8), and four aliphatic methylene groups. Through inspection of the ¹H-¹H COSY spectrum, we easily established a long spin system that started from H₂-5 (δ_H 2.40, t) and terminated at H₂-8 (δ_H 2.44, t) (Figure 2). In the HMBC spectrum, the H₂-5 showed correlations to the double bond (C-2/C-3), as well as to carboxylic group (C-1, δ_C 166.7), indicating that C-5 and C-1 were attached to the allylic carbon (C-2). The HMBC correlations from H₃-11 to double bond (C-3/C-2), as well as to C-4, indicating that C-11 and C-4 were attached to the allylic carbon (C-2). C-9 connected to C-10 and C-8 was substantiated by the HMBC correlations from H₂-8 to C-9 and from H₃-10 to C-9 and C-8. Therefore, 7 was established as 2-methyl-3-(5-oxohexyl) maleic acid.

Compound 8 exhibited UV maximum absorption at 206 and 250 nm similar to that of 7, indicating that they shared a similar chromophore. The HRESIMS data of 8 determined the molecular formula C₁₁H₁₈O₅, with two hydrogen atoms more than 7. Compared to the NMR data, 8 contained one less carbonyl group at C-9, but one more hydroxyl group at C-8 than 7. Thus, 8 was determined as 2-(4-hydroxyhexyl)-3-methylmaleic acid.

Compound 9 showed a prominent peak at m/z 243.1594 [M + H]⁺ in the HRESIMS spectrum, corresponding to the molecular formula C₁₃H₂₃O₄. Analysis of the NMR data (Tables 2, 3) of 9 revealed that it was also structurally related to 7. Compound 9 established a long chain from C-5 to C-10 through COSY spectrum (Figure 2). The position of the double bond is at C-3/C-11 and an ethyl group attached to the ester group (C1). Thus, compound 9 was determined as 3-(ethoxycarbonyl)-2-methylenenonanoic acid.

Compound 10, a colorless oil, was found to have the molecular formula C₁₂H₂₁O₅ on the basis of HRESIMS data. In addition, 10 exhibited a similar UV maximum absorption as 9 at 206 nm, indicating that they have similar chromophores. The NMR spectrum of 10 (Tables 2, 3) is similar to that of 9, with one more oxygen atom and one methylene group. The ¹H-¹H COSY correlations (Figure 2) of H₂-7/H-8 and H-8/H₂-9, as well as HMBC correlations from H-8 to C-6 and from C-8 to H₃-10, H₂-9, and H₂-7 indicated that the position of the oxymethine (δ_H 3.28; δ_C 71.3). Thus, 10 was elucidated as 7-hydroxy-3-(methoxycarbonyl)-2-methylenenonanoic acid.

Compound 11 was obtained as brown oil. HREIMS ion peak at m/z 229.1076 [M + H]⁺ gave the molecular formula C₁₁H₁₇O₅, suggesting four degrees of unsaturation. The ¹³C NMR spectrum (Table 2) revealed two carbonyl carbons (δ_C 173.6 and 164.2), together with a fully substituted double bond (δ_C 133.9 and 151.5). The ¹H–¹H COSY spectrum (Figure 2) revealed a spin system consistent with an n-pentyl chain (from C-5 to C-9). An oxymethine of C-8 (δ_H 3.55; δ_C 66.0) had HMBC correlations extending to C-6 (δ_C 21.1). The HMBC correlations from δ_H 5.13 (1H, m, H-4) to C-2 (δ_C 133.9) and C-3 (δ_C 151.5) constructed an α, β-unsaturated five-member lactone ring. In addition, a methyl group substituted at C-2 was supported by the HMBC correlations of C-10 (δ_H 2.02, δ_C 10.7) to C-1, C-2, and C-3, and a carboxyl substituted at C-3 was supported by the correlations of H-10 to C-11. Finally, compound 11 was identified and named 2-(4-hydroxypentyl)-4-methyl-5-oxo-2,5-dihydrofuran-3-carboxylic acid.

The optical rotations of compounds 8–11 were close to zero, and these compounds showed little cotton effect in CD spectroscopy, suggesting them to be racemic mixtures (Supplementary Material). The α-hydro-carbon and C-8 with hydroxyl groups led to chiral centers. It had been reported in the literature that the alkane derivatives were found as enantiomers (Akone et al., 2014).

In addition, the known thiodiketopiperazine derivatives were elucidated as adametizine A (2) (Liu et al., 2015a), DC1149B (3) (Yamazaki et al., 2015), outovirin B (4) (Kajula et al., 2016), pretrichodermamide E (5) (Yurchenko et al., 2016), and peniciadametizine A (6) (Liu et al., 2015b), respectively, by comparing their physicochemical properties and spectroscopic data with the reported literature values.

All obtained compounds were evaluated for their antibacterial and antifungal activities. Compounds 1–5 exhibited weak antibacterial activities against Erysipelothrix rhusiopathiae WH13013 and Streptococcus suis SC19, with the minimal inhibitory concentration (MIC) values of 50–100 μg/ml. Compound 2 also exhibited activity against fungi Botrytis cinerea and Septoria nodorum Berk., with the MIC values of 25 μg/ml. However, all isolated compounds showed no activities against the other three bacteria (Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and Staphylococcus aureus ATCC 25923) and two fungi (Fusarium graminearum Schw. and Rhizoctonia solani Kühn). Cephalosporin and cycloheximide were used as the positive controls in the antibacterial (MIC values of 0.78 μg/ml) and antifungal (MIC values of 6.25 μg/ml) tests, respectively (Table 4).

Two human prostate cancer cell lines, PC-3 (androgen receptor negative) and 22Rv1 (androgen receptor positive), were used in the cytotoxicity tests. Compounds 1–3 exhibited cytotoxicity against 22Rv1 cells with half maximal inhibitory concentration (IC₅₀) values of 13.9, 13.0, and 13.6 μM, respectively, whereas 1 and 3 showed activities against PC-3 cells with IC₅₀ values of 44.0 and 5.1 μM, respectively. Compound 3 was further evaluated for its anti-tumor effect by plate clone formation assay and flow cytometry on PC-3 cells. The results showed that 3 reduced PC-3 cells colony formation (Figures 4A,C) and induced cell apoptosis in a dose-dependent manner (Figures 4B,D).

Compounds 1–11 were screened for their inhibitory activities of LPS-induced NF-κB activation in RAW264.7 cells. Compounds 1, 2, 5, 9, and 11 exhibited obvious inhibitory activities against LPS-induced NF-κB with IC₅₀ values of 8.2, 15.1, 12.6, 10.7, and 21.5 μM, respectively. Moreover, in the further study for evaluation with their effects on RANKL-induced osteoclastogenesis, 1, 5, and 9 could suppress the RANKL-induced osteoclast differentiation in BMMCs obviously, with the concentration of 10 μM. The new trithiodiketopiperazine derivative adametizine C (1) showed the strongest activity, relatively (Figure 5). Consequently, it is revealed that these compounds could be the promising osteoclast differentiation inhibitors for the treatment of osteoclast-related diseases.

In addition, we also found that the thiodiketopiperazine derivatives exhibited various activities reported in the literatures. For example, pretrichodermide A was active against Mycobacterium tuberculosis (Seephonkai et al., 2006); Adametizine A (2) was found to be active against a variety of bacteria (Liu et al., 2015a); Outovirin C was active against fungus Botrytis cinerea (Kajula et al., 2016). Gliovirin showed inhibitory effects on the expression of cytokines [tumor necrosis factor (TNF)-α and interleukin-2 (IL-2)] and pro-inflammatory enzymes [cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase (INOS)] in T cells and monocytes/macrophages (Rether et al., 2007).