Cyclopeptide Derivatives from the Sponge-Derived Fungus Acremonium persicinum F10

Cyclopeptides usually play a pivotal role, either in the viability or virulence of fungi. Two types of cyclopeptides, six new hydroxamate siderophore cyclohexapeptides (1–6), including acremonpeptides E and F, and their complexes with aluminum and ferric ions; one new cyclic pentapeptolide, aselacin D (9); together with a known compound, aselacin C (10), were isolated and characterized from the sponge-derived fungus Acremonium persicinum F10. In addition, two new siderophore analogues chelating gallium ions (Ga3+), Ga (III)-acremonpeptide E (7) and Ga (III)-acremonpeptide F (8), using isolated acremonpeptides E and F, were prepared. The planar structures of 1–10 were elucidated by HRESIMS and (1D and 2D) NMR. The absolute configurations of amino acids were determined by means of the advanced Marfey’s method and X-ray single-crystal diffraction analysis. X-ray fluorescence (XRF) spectrometer was performed to disclose the elements of compound 1, indicating the existence of aluminum (Al). Al (III)-acremonpeptides E (1), Ga (III)-acremonpeptides E (5), Al (III)-acremonpeptide F (7), and Ga (III)-acremonpeptide F (8) displayed high in vitro anti-fungal activities, which are comparable to amphotericin B, against Aspergillus fumigatus and Aspergillus niger.


Introduction
Siderophores from bacteria, fungi, and plants [1,2], assembled by non-ribosomal peptide synthetases (NRPSs), are high-affinity iron chelators responsible for iron acquisition and storage. Siderophores usually provide a "Trojan horse" strategy for humans to exploit new antibiotics using a combination a transporter and antibacterial or bactericidal moieties, such as cefiderocol, the recent Food and Drug Administration (FDA)-approved siderophoreconjugated antibiotic [3]. Hydroxamate-based peptidyl siderophores are the major type in fungi, including the extracellular siderophore, the depsipeptides and coprogen family, and the intracellular siderophore, the ferrichrome family, which enable the uptake of iron from surrounding environment [4]. Among them, ASP2397, a previously reported hydroxamate-based peptidyl siderophore, is a novel natural compound from Acremonium persicinum MF-347833, exhibiting potent fungicidal activity against the invasive Aspergillus genus with unique bacteriostatic mechanism [5,6]. Eight naturally occurring derivatives of ASP2397 have been isolated from Acremonium persicinum ( Figure S1), some of which exhibit antifungal and antiviral activities, indicating their potentials as drug leads [5,7].
Marine-derived fungi are valuable resources for the exploration of structurally novel and bioactive compounds and drug leads [8][9][10][11]. In particular, chemical studies of marine sponge-derived fungi have afforded a variety of bioactive secondary metabolites [12][13][14]. Here, we report seven new cyclopeptide compounds from Acremonium persicinum F10 derived from marine sponge Phakellia fusca in the South China Sea.
Originally, sponge-derived fungal strains were screened by adopting the OSMAC (one strain many compounds) approach; as a result, we found that Acremonium persicinum F10 Here, we report seven new cyclopeptide compounds from Acremonium persicinum F10 d rived from marine sponge Phakellia fusca in the South China Sea.
Originally, sponge-derived fungal strains were screened by adopting the OSMA (one strain many compounds) approach; as a result, we found that Acremonium persicinu F10 displayed the richest metabolites in the HPLC chemical profiles when statically cu tured on rice medium.. Consequently, three new cyclopeptides (2,4,9), four Al 3+ and Fe complexes of 2 and 4, and a known compound 10 were isolated by scale-up fermentatio of A. persicinum F10. In addition, two new chelates (compounds 7 and 8) with Ga 3+ we synthesized in vitro by two new ligands (compounds 2 and 4) (Figure 1). Compounds 5, 7, and 8 exhibited excellent anti-fungal activities, comparable with positive control am photericin B (Minimum Inhibitory Concentration, MIC at 1 μM). Meanwhile, all the compounds showed no cytotoxicity to normal cell (human embryonic lung fibroblas MRC-5) at the concentration of 30 μM.

Structure Elucidation of Compounds 1-10
Al (Ⅲ)-acremonpeptide E (1) was isolated as a white needle-like crystal. The mole ular formulation was unable to be deduced by element composition of C, H, O, N, or from the HRESIMS data m/z 872.4123 [M + H] + . X-ray fluorescence (XRF) spectrometr was performed to check the elements of compound 1, and the existence of aluminum (A in 1 was confirmed by comparing with the control ( Figure S2).

Structure Elucidation of Compounds 1-10
Al (III)-acremonpeptide E (1) was isolated as a white needle-like crystal. The molecular formulation was unable to be deduced by element composition of C, H, O, N, or S from the HRESIMS data m/z 872.4123 [M + H] + . X-ray fluorescence (XRF) spectrometry was performed to check the elements of compound 1, and the existence of aluminum (Al) in 1 was confirmed by comparing with the control ( Figure S2) A detailed analysis of the total correlation spectroscopy (TOCSY), heteronuclear single quantum coherence (HSQC), and heteronuclear multiple-bond correlation (HMBC) data allowed for the construction of an alanine (Ala), a leucine (Leu), a phenylalanine (Phe), and three ornithine (Orn) residues ( Figure 2). HMBC correlations of three methyl signals with the remaining three carbonyls (δ H 2.05/δ C 161.5, δ H 2.09/δ C 161.7, and δ H 2.09/δ C 161.3), together with HMBC correlations of the methylenes in the Orn to these three carbonyls (δ H 3.69/δ C 161.5, δ H 4.02/δ C 161.7, and δ H 3.41/δ C 161.3) indicated that each Orn was modified by one acetyl group via an amide bond. The remaining three oxygens and aluminum were deduced to be located in 5-N atom of Orn to form N 5 -acetyl-N 5 -hydroxyornithine, then chelating with aluminum by detailed review of the literature concerning related structures [15]. Given that 12 of the 13 degrees of unsaturation were attributed to nine carbonyl carbons and a benzene ring, the absence degree of unsaturation suggest that 1 might be a cyclopeptide. The amino acid sequence in 1 of AcN(OH)Orn 1 -AcN(OH)Orn 2 -AcN(OH)Orn 3 -Ala-Leu-Phe was established by the HMBC correlations of Phe-NH (δ H 8.95)/Leu-CO (δ C 172. 3 (Figure 2 and Figure S10). The planar structure of 1 was further confirmed by X-ray diffraction analysis with Cu Kα radiation accompanied by the Flack parameter of -0.10 (5), which also determined the absolute configuration assignments of L -Ala, L -Leu, D -Phe, L -Orn 1 , L -Orn 2 , and L -Orn 3 in 1. Thus, compound 1 was identified and named Al (III)-acremonpeptide E.  (Figures 2 and S10). The planar structure of 1 was further confirmed by X-ray diffraction analysis with Cu Kα radiation accompanied by the Flack parameter of -0.10 (5), which also determined the absolute configuration assignments of L-Ala, L-Leu, D-Phe, L-Orn 1 , L-Orn 2 , and L-Orn 3 in 1. Thus, compound 1 was identified and named Al (Ⅲ)-acremonpeptide E. Acremonpeptide E (2) was obtained as a faint yellow solid with a molecular formula of C39H61N9O12, on the basis of its high resolution electrospray ionization mass spectroscopy (HRESIMS) data, which suggested 14 degrees of unsaturation. Comparison of the 1 H NMR and 13 C NMR data of compound 2 ( Table 1) with those of 1 suggested that both compounds shared similar structural features, except obvious shifts of the three acetyl group carbonyls (from δC 161.5, 161.7, and 161.3 in 1 to 170.8 in 2) and methyls (from δC 15.9, 16.2, and 15.3 in 1 to 20.8 in 2) in AcN(OH)Orn residues, forming overlapped carbons. These similar shifts were also observed in compound Al(Ⅲ)-acremonpeptide D and acremonpeptide D or compound ASP2397 and AS2488059, indicating compound 2 was the ligand compound of 1 by deleting Al (Ⅲ). As expected, detailed analysis of its TOCSY correlations allowed for the construction of an alanine (Ala), a leucine (Leu), a phenylalanine (Phe), and three N 5 -acetyl-N 5 -hydroxyornithines in 2 ( Figure 2). Further, the amino Acremonpeptide E (2) was obtained as a faint yellow solid with a molecular formula of C 39 H 61 N 9 O 12 , on the basis of its high resolution electrospray ionization mass spectroscopy (HRESIMS) data, which suggested 14 degrees of unsaturation. Comparison of the 1 H NMR and 13 C NMR data of compound 2 ( Table 1) with those of 1 suggested that both compounds shared similar structural features, except obvious shifts of the three acetyl group carbonyls (from δ C 161.5, 161.7, and 161.3 in 1 to 170.8 in 2) and methyls (from δ C 15.9, 16.2, and 15.3 in 1 to 20.8 in 2) in AcN(OH)Orn residues, forming overlapped carbons. These similar shifts were also observed in compound Al(III)-acremonpeptide D and acremonpeptide D or compound ASP2397 and AS2488059, indicating compound 2 was the ligand compound of 1 by deleting Al (III). As expected, detailed analysis of its TOCSY correlations allowed for the construction of an alanine (Ala), a leucine (Leu), a phenylalanine (Phe), and three N 5 -acetyl-N 5 -hydroxyornithines in 2 ( Figure 2). Further, the amino acid sequence in 2  (Figure 3 and Figure S21). The absolute configuration of the amino acid residues in 2 was determined using advanced Marfey's method. The hydrolysates of 2 were derivatized with 1-fluoro-2,4-dinitrophenyl-5-L -leucinamide ( L -FDLA) and analyzed by ultra performance liquid chromatography mass spectrometric (UPLC-MS). By comparing L -FDLA derivatives of amino acid from compound 2 with standard amino acid, we determined the L -Ala, L -Leu, and D -Phe residues in 2 ( Figure S22), identical to compound 1 ( Figure S11). The absolute configuration of AcN(OH)Orn units of compound 2 was deduced as the same as that of compound 1 with L -Orn residues, owing to a shared biosynthetic pathway. On the basis of these data, we established the structure of compound 2, which was the ligand compound of 1, acremonpeptide E.  (Figures 3 and S21). The absolute configuration of the amino acid residues in 2 was determined using advanced Marfey's method. The hydrolysates of 2 were derivatized with 1-fluoro-2,4-dinitrophenyl-5-L-leucinamide (L-FDLA) and analyzed by ultra performance liquid chromatography mass spectrometric (UPLC-MS). By comparing L-FDLA derivatives of amino acid from compound 2 with standard amino acid, we determined the L-Ala, L-Leu, and D-Phe residues in 2 ( Figure S22), identical to compound 1 ( Figure S11). The absolute configuration of AcN(OH)Orn units of compound 2 was deduced as the same as that of compound 1 with L-Orn residues, owing to a shared biosynthetic pathway. On the basis of these data, we established the structure of compound 2, which was the ligand compound of 1, acremonpeptide E.  (8)), suggesting the planar structure and absolute configuration of compound 3 were identical to 1, except the substitution of metal ion Al (Ⅲ) in 1 by Fe (Ⅲ) in 3. Accordingly, the structure of compound 3, which was named Fe (Ⅲ)-acremonpeptide E, was corroborated. Table 1. 13 C NMR (150 MHz, DMSO-d6) data and 1 H NMR (600 MHz, DMSO-d6) data for Al (Ⅲ)-acremonpeptide E (1), acremonpeptide E (2), and Ga (Ⅲ)-acremonpeptide E (7).   (8)), suggesting the planar structure and absolute configuration of compound 3 were identical to 1, except the substitution of metal ion Al (III) in 1 by Fe (III) in 3. Accordingly, the structure of compound 3, which was named Fe (III)-acremonpeptide E, was corroborated. ), as designated by the HRESIMS and 13 C NMR data. The 1D NMR and 2D NMR spectra indicated that compound 5 possessed a similar structure to compound 4, except that the overlapped carbon signals in 4 were clearly separated in 5 from the 13 C NMR, which was also observed between compounds 1 and 2. Therefore, we speculated that 5 was the siderophore-metal (Ⅲ) complex of 4. The crystal of compound 5 was obtained by repeated recrystallization from n-hexane-chloroform (1:1) and subjected to single-crystal X-ray diffraction analysis with Cu Kα (Figure 4) (Flack parameter = 0.02 (6)), further confirming its planar structure and absolute configuration as cyclo-(L-AcN(OH)Orn 1 -L-AcN(OH)Orn 2 -L-AcN(OH)Orn 3 -L-Ser-L-Leu-D-Phe) chelating with Al (Ⅲ). Fe (Ⅲ)-acremonpeptide F (6) was obtained as an orange-red solid. NMR spectra indicated the possible presence of Fe (Ⅲ) in compound 6 as compound 3. The HRESIMS data of compound 6 displayed an [M + H] + ion at m/z 917.3520, corresponding to a molecular formula of C39H59N9O12Fe, which indicated that compound 6 may be the complex of   (Table 2) was similar to that of compound 2 (Table 2). Comparing the 1D NMR data of compound 4 with those of 2, we found that the presence of an oxygen methylene group (δ H 3.55 and 3.58, δ C 61.0) and the absence of a methyl group (δ H 1.24, δ C 18.1) in 4 was the major difference. Detailed analysis of TOCSY and HMBC correlations demonstrated that the only discrepancy in compound 4 was a Ser residue for the Ala moiety substitution relative to 2, which was supported by , as designated by the HRESIMS and 13 C NMR data. The 1D NMR and 2D NMR spectra indicated that compound 5 possessed a similar structure to compound 4, except that the overlapped carbon signals in 4 were clearly separated in 5 from the 13 C NMR, which was also observed between compounds 1 and 2. Therefore, we speculated that 5 was the siderophore-metal (III) complex of 4. The crystal of compound 5 was obtained by repeated recrystallization from n-hexane-chloroform (1:1) and subjected to single-crystal X-ray diffraction analysis with Cu Kα (Figure 4) (Flack parameter = 0.02 (6) Ga (III)-acremonpeptide E (7) and Ga (III)-acremonpeptide F (8) were obtained as white powders prepared from acremonpeptide E (2) and F (4) with Ga 2 (SO 4 ) 3 ·H 2 O, respectively. The structures of compounds 7 and 8 were further identified under the guidance of HRESIMS and NMR data ( Figures S54-S60 for 7 and Figures S63-S69 for 8).
Aselacin D (9) was isolated as a faint yellow solid, and its molecular formula was assigned as C 46 H 66 N 8 O 10 on the basis of HRESIMS. The 13 C NMR and distortionless enhancement by polarization transfer (DEPT) spectra revealed 46 resonances, including 3 methyl, 17 methylene, 14 methine, 3 quaternary, and 9 carbonyl carbons. The presence of signals in the amide NH and α-amino acid protons in 1 H NMR spectrum and carbonyl groups of its 13 C NMR data (Table 3) indicated the peptidic nature of this molecule. Interpretation of the TOCSY and HMBC correlations suggested the existence of Gly, Ala, Trp, Thr, and Gln residues, five common amino acid residues. In addition, two methylene groups (δ H 2.54, δ H 2.28, and δ C 34.2/δ H 3.47, δ H 3.01, and δ C 36.5) coupled to the carbonyl carbon δ C 171.7 revealed the presence of a β-Ala residue. Whereafter, detailed 1D and 2D NMR data revealed that the remaining proton and carbon signals accounted for an aliphatic chain containing a diene and two carbonyl groups ( Table 3). The HMBC correlations of Gly-NH (δ H 7.81)/Ala-CO (δ C 172.8), Ala-NH (δ H 8.97)/Trp-CO (δ C 173.6), Trp-NH (δ H 7.91)/ β-Ala -CO (δ C 171.7), β-Ala-NH (δ H 7.34)/Thr-CO (δ C 168.4), Thr-NH (δ H 8.40)/Gln-CO (δ C 173.0), and Gln-NH (δ H 8.45)/fatty acid-CO (δ C 173.8) confirmed the sequence of Gly-Ala-Trp-βAla-Thr-Gln in this compound and the connection of aliphatic chain with amino group via the Gln residue. Furthermore, a cyclic depsipeptide, formed between Thr and Gly residues, was supported by the key correlations of Thr-3 (δ H 5.40)/Gly (δ C 168.2) and Thr-4 (δ H 1.05)/Gly (δ C 168.2). The absolute configuration of each amino acid residues was confirmed using advanced Marfey's method. LC-MS analysis of the hydrolysate's derivatives of each amino acid residue and comparison with the retention times of the standards assigned the L -Ala, L -Thr, D -Trp, and D -Gln (detected as D -Glu) in 9 ( Figure S79). Finally, the configuration of E,E-diene in aliphatic chain was established from their proton-proton coupling constants of 15.6 Hz.
Aselacin C (10) was obtained as a faint yellow solid, and its molecular formula C 46 H 66 N 8 O 11 was suggested by the HRESIMS data at m/z 907.4929 [M + H] + . 1 H and 13 C NMR data indicated that compound 10 shared high similar structural features with compound 9, except for an oxygen methylene group (δ C 60.4) instead of the methyl (δ C 16.2) at Ala, suggesting that the Ala in 9 was displaced by Ser residue in 10. A literature survey indicated 10 to be identical to the known compound aselacin C [16]. The application of advanced Marfey's analysis supported the absolute configuration as depicted.
Invasive aspergillosis usually leads to a severe life-threatening infection, especially for immunocompromised patients [17,18]. Recently, ASP2397 (also known as VL-2397 under clinical trial), a hydroxamate-containing siderophore isolated from fungus Acremonium persicinum, exhibited potent antifungal activities [6,19]. Hydroxamic acids usually possess a formula RC(O)N(OH)R' and can be regard as a type of N-hydroxy amides [20]. Here, elucidation of acremonpeptide E; acremonpeptide F; and their complexes with Al 3+ , Fe 3+ , and Ga 3+ (1-8) further enrich the chemical structural diversity of the hydroxamate siderophore family.

Biological Evaluation of These Compounds
Hydroxamate-containing compounds 1-8 were evaluated for antifungal activities against Aspergillus fumigatus and Aspergillus niger. We found that compounds 1, 5, 7, and 8 showed obvious antifungal activities against A. fumigatus and A. niger with MIC values ranging from 1 to 3 µM, which is comparable to the positive control amphotericin B (Table S1).
In line with the previous report, the free acremonpeptides or Fe (III)-acremonpeptides failed to show antifungal activities in biological evaluation [6]. We speculate that the tested fungi may use the free acremonpeptides or Fe (III)-acremonpeptides as a vector to take in Fe 3+ for survival, but this could be blocked by acremonpeptides chelating other ions, which will occupy the transport receptor of absorption.
The cytotoxic assays indicated that compounds 1-10 were inactive against non-small cell lung cancer cell line A549, small cell lung cancer cell lines H446 and H1688, and human embryonic lung fibroblast cell MRC-5 at concentrations up to 30 µM. This result is consistent with the deductive mechanism of action for this class of compounds, targeting Sit1 of Aspergillus genus, which is lacking in mammalian cells, indicating the potential druggability of these compounds [21]. Aselacins D (9) and C (10) were not tested for the antifungal activity for the ullage of samples.

Fungal Material and Fermentation
The fungal strain Acremonium persicinum F10 was isolated from the fresh inner issue of marine sponge Phakellia fusca collected at a depth of 10-20 m near the Yongxin Island (112 • 20 E, 16 • 50 N) in the South China Sea in June 2013. This fungus was identified as Acremonium persicinum on the basis of morphological characteristics and sequence analysis of the ITS region (GenBank, accession no. MH882418). A voucher specimen was preserved at Marine Biotechnology Laboratory, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai, China. The fungus A. persicinum F10 was cultured on potato dextrose agar (PDA) for 7 days. The spores of A. persicinum F10 were inoculated into 50 mL of seed medium (PDB) in a 250 mL Erlenmeyer flask and incubated on a rotary shaker at 25 • C (150 rpm) for 48 h. Then, 5% (v/v) seed cultures were transferred into 200 mL production medium (rice 80 g/L, peptone 6 g/L in artificial seawater [22] at pH 7.0) in 60 × 1 L Erlenmeyer flasks under static conditions at room temperature.

Compound Preparation
After 40 days of fermentation, the rice medium was smashed and extracted by ethyl acetate to yield 50.1 g of ethyl acetate extract. Preparative medium-phased liquid chromatography (MPLC) was performed on a flash purification system (Bonna Agela Technologies Corporation, Tianjin, China). High-performance liquid chromatography (HPLC) was carried out on an Agilent 1200 liquid chromatography system equipped with a diode array detector (DAD) detector. 1-Fluoro-2,4-dinitrophenyl-5-L -/ D -leucinamide ( L -/ D -FDLA) and m-chloroperbenzoic acid were purchased from Sigma-Aldrich Chemical Corporation.

Spectrum Analysis
UV spectra were measured on a UV-VIS spectrophotometer (UV/EV300, Thermo scientic, Waltham, MA, USA). Optical rotations were recorded on a P-2000 digital polarimeter (Jasco, Japan) with a 1.0 mL cell. IR spectra were recorded on a KBr pellets using a Fourier transform infrared spectrometer (FT-IR) (Nicolet 6700, Thermo Nicolet Co., Waltham, MA, USA). XRF spectra were obtained by X-ray fluorescence spectrometer (XRF-1800, Shimadzu Inc., Tokyo, Japan). Mass spectra were measured on a positive ion mode using LC-HRMS with a Waters ACQUITY UPLC system (Waters Inc., Milford, MA, USA) coupled with a Waters Micromass Q-TOF Premier Mass Spectrometer, which was equipped with an electrospray interface. The NMR data were collected by a Bruker Avance III 600 MHz spectrometer (600 MHz, Bruker Co., Ltd., Karlsruhe, Germany) at 600 MHz for 1 H nuclei and 150 MHz for 13

In Vitro Chelate Compounds Synthesis
Here, in vitro chelation reactions of ligand compounds 2 and 4 with ferric, gallium, and aluminum ions were carried out. Firstly, 0.5 mL of each aqueous solution (0.01 mmol) of FeCl 3 ·6H 2 O, Ga 2 (SO 4 ) 3 ·H 2 O, and Al 2 (SO 4 ) 3 ·6H 2 O was stirred at 120 rpm. Then, equivalent volume of compounds 2 or 4 were dissolved in methanol and added in previous aqueous solution to continue stirring for 3 h. Finally, the reaction mixtures were analyzed by UPLC-MS.

X-ray Crystallographic Analysis
Colorless crystals of 1, 3, and 5 were obtained by diffusing n-hexane into a chloroform solution. Single-crystal X-ray diffraction data were collected on a Bruker D8 VENTURE diffractometer using graphite-monochromated Cu Kα radiation (λ = 1.54178 Å). The parameters in Common Intermediate Format (CIF) format for 1, 3,

Acid Hydrolysis of Compounds 1-6, and 9
Approximately 0.6 mg of each of compounds 1-6 was hydrolyzed with 6 N HCl (1 mL) for 16 h at 110 • C. For analysis of compound 9, 3.0% v/v triisopropylsilane was added, and the mixture was hydrolyzed with 6 N HCl (1 mL) for 2 h at 110 • C. After cooling to room temperature, the hydrolysate mixtures and traces of HCl were evaporated to dryness.

Absolute Configurations of Amino Acids by the Advanced Marfey's Analysis
Each acid hydrolysate was resuspended in 100 µL of H 2 O. To each half portion (50 µL), we added 1 N NaHCO 3 (20 µL) and 100 µL of L -FDLA (10 mg/mL in acetone). Each mixture was heated to 50 • C for 1 h. The reaction was quenched with 20 µL of 1 N HCl and dried under nitrogen. The residue was redissolved in MeOH and measured by UPLC-MS using an Acquity UPLC BEH C18 column (2.1 × 50 mm, 1.7 µm, 0.5 mL/min). MeCN/H 2 O containing 0.1% formic acid was used as mobile phase with a linear gradient from 10% to 100% over 12 min. Through comparison of the retention times of the L -FDLA derivatives of amino acid standard and corresponding amino acids from each compound, we established the absolute configuration ( Figures S11, S22, S39, and S79).

Antifungal Activity Assay
Compounds 1-8 were tested for antifungal activities against Aspergillus fumigatus ATCC204305 and Aspergillus niger ATCC16404 using broth microdilution antifungal susceptibility testing, according to previously published protocol [23,24]. In the experiments, the RPMI-1640 medium containing 10.4 g/L of RPMI-1640 medium (R8755; Sigma, St. Louis, MO, USA), 6.7 g/L of Yeast Nitrogen Base (YNB; Becton, Dickinson and Company, Franklin Lakes, NJ, USA), 1.8% (w/v) glucose, and 40 mM HEPES (pH = 7.1) was used as culture medium. The inoculum per well for a 96-well microplate was 2 × 10 4 fresh conidia/mL (150 µL/well), which were incubated for 24 h at 37°C. The minimal inhibitory concentration (MIC) in the assay were performed using resazurin at 0.002% (w/v) of final concentration. Amphotericin B was used as the positive control, exhibiting MICs at 1 µg/mL to both tested fungal strains.

Cytotoxicity Assay
The cytotoxic activity of compounds 1-10 against A549, H466, H1688, and MRC-5 cells (all cell lines were obtained from the Cell Bank of the Chinese Academy of Sciences, Shanghai, China) were evaluated by the MTT assay as described previously [25]. Cisplatin was used as the positive control with IC 50 values of 3.8 µM, 1.9 µM, 3.0 µM, and 23.8 µM, respectively.

Conclusions
The organic extract of the sponge-derived fungus Acremonium persicinum F10 yielded two types of cyclopeptides, including six new hydroxamate siderophore cyclohexapeptides, acremonpeptides E, acremonpeptides F, and their chelates (1-6); a new cyclic pentapeptolide, aselacin D (9); and a known compound, aselacin C (10). Analyses of the spectroscopic data of NMR and HRESIMS elucidated the planar structures of these compounds, while the absolute configurations were confirmed using the advanced Marfey's method and X-ray single-crystal diffraction analysis.
The differences among the new isolated acremonpeptide analogues 2 and 4 and the previously reported findings were varying amino acid replacements or simply different stereocenters ( Figure S1), indicating the substrate tolerance of the biosynthesis pathway of the hydroxamate-based peptidyl siderophore from Acremonium persicinum.
The compouds 1, 5, 7, and 8 displayed high anti-fungal activities in vitro against Aspergillus fumigatus and Aspergillus niger, with MICs at 1 µg/mL. Meanwhile, all these compounds showed no cytotoxicity to normal cell (human embryonic lung fibroblast, MRC-5) at the concentration of 30 µM.
Author Contributions: Y.L. conducted the isolation and structural elucidation of compounds and the bioactive assay, and also wrote the draft manuscript; Z.L. designed and supervised the experiments and edited the manuscript. All authors have read and agreed to the published version of the manuscript.