New Alkaloids and Polyketides from the Marine Sponge-Derived Fungus Penicillium sp. SCSIO41015

The sponge-derived fungus Penicillium sp. SCSIO41015 cultured on solid rice medium yielded twenty-one compounds (1–21), including two new alkaloids (1 and 2) and one new pyrone derivative (3). Their structures were elucidated by analysis of 1D/2D NMR data and HR–ESI–MS. Their absolute configurations were established by single-crystal X-ray diffraction analysis and comparison of the experimental with reported specific rotation values. Compound 16 exhibited selective cytotoxic activity against the human gastric cancer cells MGC803, with IC50 value of 5.19 μM. Compounds 9 and 18 showed weak antibacterial activity against Staphylococcus aureus and Acinetobacter baumannii, respectively, both with MIC values of 57 μg/mL. Furthermore, compound 16 displayed potent antibacterial activity against S. aureus with an MIC value of 3.75 μg/mL.


Introduction
A considerable number of structurally unique secondary metabolites with biological and pharmacological activities have been isolated from the marine-derived fungi in recent years [1,2]. Marine-derived Penicillium and Aspergillus are reported to be the most ubiquitous genera [1]. Penicillium has become one of the most attractive fungal genera to search for bioactive molecules, since the advent of penicillin [3]. More than 400 new marine natural products, including half the number of polyketides and a quarter the number of nitrogen compounds have been isolated from marine-derived Penicillium fungi, and 58% of the new products have displayed bioactivities such as anticancer, antibacterial, anti-HIV, and anti-inflammatory properties [4]. As part of our continued searching for bioactive secondary metabolites from sponge-derived fungi [5][6][7], the fungus Penicillium sp. SCSIO41015 was investigated. Three new compounds (1-3) and eighteen known ones (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21) were isolated from the rice solid culture extract of the fungus Penicillium sp. SCSIO41015 shown in Figure 1. Nearly all compounds were evaluated on their cytotoxic activities against six human cancer cells and their antibacterial activities against five pathogenic bacteria. Herein, we described the isolation, structure elucidation, and bioactivity evaluation of the twenty-one compounds, as well as the potential biosynthetic pathway of these alkaloids.

Isolation and Structural Elucidation
Compound 1 was obtained as a colorless oil. Its molecular formula was established as C 14 3 -13), and two active hydrogen protons (δ H 8.59, dd, J = 9.1, 4.2 Hz, NH-11; 8.23, d, J = 7.7 Hz, NH-9). Accordingly, the 13 C NMR (DEPT) ( Table 1) data also displayed fourteen carbon signals, including four sp 2 methines, a sp 3 methylene (δ C 40.2, C-8), a sp 3 methine (δ C 47.9, C-9), three methyls (δ C 52.0, C-14; 29.0, C-11; 22.3, C-13), two sp 2 non-protonated carbons, and three carbonyls (δ C 198.1, C-7; 172.2, C-10; 169.2, C-12). Part of its NMR data were similar to N-acetyl-6-nitrotryptophanmethyl ester synthetized by R.R. King [8]. The obvious differences were that the signals of a nitro group and a double-bond were absentin compound 1, while an aromatic proton, and a methyl and a carbonyl group were observed. The changes indicated that the five-membered ring opened between C-2 and C-3 in N-acetyl-6-nitrotryptophanmethyl ester, which were confirmed by the 1 H-1 H correlation spectroscopy (COSY) of H-3/H-4/H-5/H-6, H 3 -11/NH-11, and the heteronuclear multiple bond correlation (HMBC) spectrum of H 2 -8, H-9, and H-6, to C-7 ( Figure 2). Thus, the planar structure of 1 was further ascertained by the HMBC spectrum ( Figure 2). The potential biosynthetic pathway of 1 was speculated that 1 was synthesized from l-tryptophan as shown in Figure S1.    Figure 2). Therefore, 2 was a N-methyl quinolone lactam, and the planar structure of 2 was determined as hydroxyl at C-3 reductive quinolactacin D, based on 2D NMR data ( Figure 2). The deduction above was supported by the Cu Kα single crystal X-ray diffraction analysis (ORTEP, oak ridge thermal ellipsoid plot, drawing in Figure 3), while the X-ray data showed that the crystal of 2 existed as a pair of enantiomers. The chiral phase of HPLC analysis showed that were two conjoint peaks that could not separate well, also 2 dissolved in a solvent that was unstable and then changed to quinolactacin D, as speculated in the reference [10]. Compound 2 appeared to be synthesized from kynurenine and leucine, which was similar to the speculation reported by Nakagawa et al. [11]; shown in Figure S1. Thereby, 2 was elucidated as a pair of enantiomers and named quinolactacin E. The HR-ESI-MS data showed the molecular formula of compound 3 to be C11H16O4, which indicated four degrees of unsaturation in this molecular structure. The 1 H NMR and 13 C NMR data ( Table 1) of 3 were quite similar with those of germicidin L [12], except for the absence of an aromatic proton signal, which was replaced by a non-protonated carbon (δC 110.0, C-5) and amore methyl group (δH 2.01, s, δC 10.0, CH3-8). The changes were confirmed by the HMBC correlations of H3-8 to C-4 (δC 168.0), C-5, C-6 (δC 161.6), indicating 3 was methyl-substituted germicidin L at position of C-5. The proposed structure was further verified by X-ray diffraction analysis of the single crystal of compound 3 (ORTEP drawing in Figure 3). Its single-crystal X-ray diffraction experiment with Cu Kα radiation and the value of the Flack parameter, 0.07 (7), allowed the Compound 2 was isolated as a light-yellow powder that returned a HR-ESI-MS pseudomolecularion (m/z 293.1263, [M + Na] + ), corresponding to a molecular formula (C 16 H 18 N 2 O 2 ) requiring nine degrees of unsaturation. Analysis of its NMR data (Table 1) revealed that the structural features of 2 closely resembled those of quinolactacin D [10]. The only difference was a methine (δ H 4.86, brd, J = 8.4 Hz, δ C 52.5, CH-3) in 2 instead of the oxygenated non-protonated carbon in quinolactacin D, which was confirmed by the correlations of H-3 to C-1 (δ C 167.9), C-3a (δ C 165.6), C-10 (δ C 42.3), C-11 (δ C 24.7) in the HMBC spectrum and the COSY cross-peaks of H-3/H-10a/H-10b ( Figure 2). Therefore, 2 was a N-methyl quinolone lactam, and the planar structure of 2 was determined as hydroxyl at C-3 reductive quinolactacin D, based on 2D NMR data ( Figure 2). The deduction above was supported by the Cu Kα single crystal X-ray diffraction analysis (ORTEP, oak ridge thermal ellipsoid plot, drawing in Figure 3), while the X-ray data showed that the crystal of 2 existed as a pair of enantiomers. The chiral phase of HPLC analysis showed that were two conjoint peaks that could not separate well, also 2 dissolved in a solvent that was unstable and then changed to quinolactacin D, as speculated in the reference [10]. Compound 2 appeared to be synthesized from kynurenine and leucine, which was similar to the speculation reported by Nakagawa et al. [11]; shown in Figure S1. Thereby, 2 was elucidated as a pair of enantiomers and named quinolactacin E.  Figure 2). Therefore, 2 was a N-methyl quinolone lactam, and the planar structure of 2 was determined as hydroxyl at C-3 reductive quinolactacin D, based on 2D NMR data ( Figure 2). The deduction above was supported by the Cu Kα single crystal X-ray diffraction analysis (ORTEP, oak ridge thermal ellipsoid plot, drawing in Figure 3), while the X-ray data showed that the crystal of 2 existed as a pair of enantiomers. The chiral phase of HPLC analysis showed that were two conjoint peaks that could not separate well, also 2 dissolved in a solvent that was unstable and then changed to quinolactacin D, as speculated in the reference [10]. Compound 2 appeared to be synthesized from kynurenine and leucine, which was similar to the speculation reported by Nakagawa et al. [11]; shown in Figure S1. Thereby, 2 was elucidated as a pair of enantiomers and named quinolactacin E. The HR-ESI-MS data showed the molecular formula of compound 3 to be C11H16O4, which indicated four degrees of unsaturation in this molecular structure. The 1 H NMR and 13 C NMR data ( Table 1) of 3 were quite similar with those of germicidin L [12], except for the absence of an aromatic proton signal, which was replaced by a non-protonated carbon (δC 110.0, C-5) and amore methyl group (δH 2.01, s, δC 10.0, CH3-8). The changes were confirmed by the HMBC correlations of H3-8 to C-4 (δC 168.0), C-5, C-6 (δC 161.6), indicating 3 was methyl-substituted germicidin L at position of C-5. The proposed structure was further verified by X-ray diffraction analysis of the single crystal of compound 3 (ORTEP drawing in Figure 3). Its single-crystal X-ray diffraction experiment with Cu Kα radiation and the value of the Flack parameter, 0.07 (7), allowed the The HR-ESI-MS data showed the molecular formula of compound 3 to be C 11 H 16 O 4 , which indicated four degrees of unsaturation in this molecular structure. The 1 H NMR and 13 C NMR data (Table 1) of 3 were quite similar with those of germicidin L [12], except for the absence of an aromatic proton signal, which was replaced by a non-protonated carbon (δ C 110.0, C-5) and amore methyl group (δ H 2.01, s, δ C 10.0, CH 3 -8). The changes were confirmed by the HMBC correlations of H 3 -8 to C-4 (δ C 168.0), C-5, C-6 (δ C 161.6), indicating 3 was methyl-substituted germicidin L at position of C-5. The proposed structure was further verified by X-ray diffraction analysis of the single crystal of compound 3 (ORTEP drawing in Figure 3). Its single-crystal X-ray diffraction experiment with Cu Kα radiation and the value of the Flack parameter, 0.07 (7), allowed the assignment of the absolute configuration of 3 as 9R, 10R. Thus, compound 3 was established as 4-hydroxy-6-((2R,3R)3-hydroxybutan-2-yl)-3,5-dimethyl-2H-pyran-2-one and named germicidin O. Polyketides and alkaloids are two major secondary metabolites from marine-derived fungi, and the same polyketides or alkaloids were often isolated from different fungi. According to the detailed spectroscopic analysis and comparison to the reported literature data, the known compounds (4-21) were identified as quinolactacin B (4) [10], quinolonimide (5) [10], quinolonic acid (6) [10], 4-hydroxy-3-methyl-2(1H)-quinolinone (7) [13], sydowinin A (8) [14], pinselin (9) [15], β-diversonolic ester (10) [16], coniochaetone J (11) [17], dihydrocitrinone (12) [18], stoloniferol A (13) [18], 6,8-dihydroxy-3,4,5-trimethylisochroman (14) [19], moniliphenone (15) [20], frangula-emodin (16) [21], methyl-2-(2-acetyl-3,5-dihydroxy-4,6-dimethylphenyl) acetate (17) [22], phenol A (18) [23], latifolicinin C (19) [24], penicitrinone A (20) [25], and 22-acetylisocyclocitrinol A (21) [26], respectively.

Biological Activity
Compounds 1-21, excepting 2 (since it was unstable in solvent and its quality was poor after the second preparation) were evaluated on their cytotoxic activities against human gastric cancer cells MGC803, human breast cancer cells MDA-MB-231, human prostate cancer cells C4-2B, human osteosarcoma 143B, and human hung cancer cells A549, using the CCK-8 method. Compound 16 selectively exhibited cytotoxic activity against MGC803 with IC 50 value of 5.19 µM. The antibacterial activities of all compounds, except 2, were estimated against A.baumannii (ATCC 19606), Klebsiella pneumonia (ATCC 13883), Escherichia coli (ATCC 25922), S.aureus (ATCC 29213) and Enterococcus faecalis (ATCC 29212). Compounds 9, 16, 18, and 21 with 100 µg/disc showed an inhibition zone against S. aureus with a diameter of about 9, 12, 21, and 7 mm, respectively. Compounds 18 and 13 with 100 µg/disc displayed an inhibition zone against A. baumannii both with a diameter of about 8 mm ( Figure S2). Further, their minimum inhibitory concentrations (MIC) were tested. The MIC values of compounds 9 and 16 against S. aureus were 57 and 3.75 µg/mL, respectively, and the MIC value of compound 18 against A. baumannii was 57 µg/mL, while the other compounds showed more than 100 µg/mL. Ampicillin and gentamicin were used as positive control against S. aureus and A. baumannii, respectively, with the MIC values of 0.45 and 0.9 µg/mL (Table S1).

General Experimental Procedures
UV spectra were recorded on a UV-2600 UV-Vis spectrophotometer (Shimadzu, Japan). Optical rotations were measured using a MCP-500 polarimeter (Anton, Austria). 1D and 2D NMR spectra were measured on a Bruker Avance 500 MHz or 700 MHz NMR spectrometer (Fällanden, Switzerland) with TMS as an internal standard. HR-ESI-MS data were recorded on a maXis Q-TOF mass spectrometer in the positive ion mode (Bruker, Fällanden, Switzerland). X-ray diffraction intensity data were collected on XtalLAB PRO single-crystal diffractometer, using Cu Kα radiation (Rigaku, Japan). The data were corrected for absorption, using

Fungal Material
The fungal strain SCSIO41015 was obtained from a Callyspongia sp. sponge, which was collected from the sea area near Xuwen County, Guangdong Province, China. The strain SCSIO41015 was stored on MB agar (malt extract 15 g, agar 16 g, sea salt 10 g, water 1 L, pH 7.4-7.8) slants at 4 • C and deposited at the Key Laboratory of Tropical Marine Bio-Resources and Ecology, Chinese Academy of Science. The ITS1-5.8S-ITS2 sequence region (508 base pairs, GenBank accession No. MK813890) of strain SCSIO41015 was amplified by PCR, and DNA sequencing showed that it shared significant homology with several species of Penicillium. The 508 base pair ITS sequence had 99% sequence identity to that of the Penicillium citrinum strain SCSGAF0167 (GenBank accession No. JN851046.1). Then, it was designated as a member of Penicillium sp. and named Penicillium sp. SCSIO41015.

Fermentation and Extraction
The mass fermentation of this fungus was carried out at 25 • C for 30 days, using a rice medium in a 1 L flask (×45) and every flask contained 200 g of rice, 2.5 g of sea salt, and 200 mL of tap H 2 O. The flasks were cultivated statically at 25 • C, under normal day-night cycle. After 30 days, the cultures were soaked in acetone (500 mL/flask) and mashed into small pieces and sonicated for 20 min. Then, the acetone was evaporated under reduced pressure to obtain an aqueous solution, which was extracted with ethyl acetate (EtOAc), for three times. The rice residue was also extracted with EtOAc. Both of the EtOAc solutions were concentrated under reduced pressure to gain a crude extract. The crude extract was suspended in MeOH and then partitioned with an equal volume of petroleum ether to remove the oil. Lastly, the MeOH solution was concentrated under reduced pressure to afford a brown extract (77.0 g).

Antibacterial Activity Assay
Compounds 1-21, excepting 2, were tested for their antibacterial activities against five pathogenic bacteria, using the method of agar filter paper diffusion. Those compounds which had an inhibition zone were tested in 96-well plates using a modification of the broth microdilution method [5]. Ampicillin and gentamicin were used as a positive control for the gram-positive and gram-negative bacteria, respectively.

Conclusions
Two new alkaloids (1 and 2) and one new pyrone derivative (3), along with eighteen known compounds (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21) were isolated from the rice cultures of sponge-derived fungus Penicillium sp. SCSIO41015. The new structures were elucidated by analysis of their NMR data and HR-ESI-MS. Their absolute configurations were established by single-crystal X-ray diffraction analysis or a comparison of the experimental with the reported specific rotation values.The biosynthetic pathway of quinolactacin A has been studied, andspeculated that it is synthesized from tryptophan and isoleucine [11]. Based on the current investigation, possible biosynthetic pathways of alkaloids 1, 2, and 4-6 were speculated.Compound 2 appeared to be synthesized from tryptophan and leucine, whereas compound 4 was made from tryptophan and valine. Compounds 1-21, excepting 2, were evaluated for their cytotoxic activities againstsix humancancer cells and their antibacterial activities against five pathogenic bacteria. Only compound 16 exhibited selective inhibitory activity against the MGC803 cells with an IC 50 value of 5.19 µM. Other compounds showed no inhibitory activities at concentrations of 5 µM in the preliminary screening. While compounds 9 and 16 displayed an antibacterial ability against S. aureus with MIC values of 57 and 3.75 µg/mL, respectively. Compound 18 exhibited a weak antibacterial ability against A. baumannii, with an MIC value of 57 µg/mL.
Author Contributions: X.P., J.W. (Junfeng Wang), S.X., and Y.L. contributed to the conception and design of the study. X.P. performed experiments, analyzed data and wrote the manuscript, X.L. did the isolation and identification of the fungus. J.W. (Junjian Wang) and G.C. performed the cytotoxicity assay. All authors contributed to the manuscript revision, and read and approved the submitted version.

Conflicts of Interest:
The authors declare no conflict of interest.