Stereocalpin B, a New Cyclic Depsipeptide from the Antarctic Lichen Ramalina terebrata

Stereocalpin B, a new cyclic depsipeptide (1), and a new dibenzofuran derivative (3), were isolated from the Antarctic lichen, Ramalina terebrata (Ramalinaceae), along with a known cyclic depsipeptide (2). The structures of new compounds were characterized by comprehensive spectrometric analyses; high-resolution fast atom bombardment mass spectrometry (HR-FABMS) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). Stereocalpin B (1) existed in a rotameric equilibrium, which was confirmed using nuclear Overhauser effect spectroscopy (NOESY)/exchange spectroscopy (EXSY) spectrum. Absolute configurations of the amino acid units in 1 were assigned using the advanced Marfey’s method and subsequent NOESY analysis of the 5-hydroxy-2,4-dimethyl-3-oxo-decanoic acid residue confirmed the complete stereochemistry of 1. Compounds 1-3 exhibited moderate antimicrobial activities against E. coli, with the IC50 values ranging from 18–30 μg/mL. Compound 2 exhibited cell growth inhibition against HCT116 cell lines, with the IC50 value of 20 ± 1.20 μM, and compounds 1 and 2 also showed potent anti-inflammatory activities against lipopolysaccharide (LPS)-induced RAW264.7 macrophages with the IC50 values ranging from 5–7 μM.


Introduction
Lichens, which are also referred to as lichenized fungi, are mutualistic symbionts involving microalga and/or cyanobacterium and fungus [1]. They are known to possess a broad variety of secondary metabolites with diverse biological activities, which are reported to be useful for the treatment of heart diseases, bronchitis, bleeding pile, vomiting, asthma, inflammation, and stomach disorders [1,2]. However, only a limited number of study has been carried out for the discovery of medicinal use and therapeutic potential of lichen-derived substances [3]. Lichens are well known for their composition of depsides and dibenzofurans as major secondary metabolites, where depsides are reported to have antioxidant and antifungal activities [2,4], and dibenzofurans are known to exhibit anticancer, antifungal, and antineoplastic properties [5][6][7]. The Antarctic lichen, Ramalina terebrata, is also known to possess various secondary metabolites such as usnic acid derivatives, ramalin, and anthraquinones [1,[8][9][10]. Derived from R. terebrata, usnic acid, and usimines A-C were reported to have antibacterial activity against B. subtilis [11], ramalin to have 2 of 10 anti-inflammatory, anticancer, and antioxidant activity [9,12], and parietin was shown to serve as both anti-aggregative and antioxidant agents in tauopathies [10,13].
In our continuous search for novel bioactive secondary metabolites from the Antarctic lichens, the acetone extract of R. terebrata was subjected to chemical investigation. Here, the identification of the unknown compound 1, stereocalpin B, is reported along with compound 3, which has been isolated as a natural product for the first time, together with a known cyclic depsipeptide (2). The structure elucidation of the new compounds (1 and 3) and the biological activities of all isolated compounds (1-3) are discussed below.

Isolation of Compounds 1-3
Air-dried and powdered lichen, R. terebrata, was extracted with acetone which provided the resultant acetone extract. The extract was subjected to fractionation and isolation, where column chromatography and HPLC purification techniques were used, resulting in the isolation of a new cyclic depsipeptide, stereocalpin B (1), and a new dibenzofuran derivative, 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6-di(ethanoyl)dibenzofuran (3), along with a known cyclic depsipeptide (2) (Figure 1). Compound 2 was identified to be stereocalpin A (2) [14] by comparing its NMR spectroscopic and physical data with those previously reported. usimines A-C were reported to have antibacterial activity against B. subtilis [11], ramalin to have anti-inflammatory, anticancer, and antioxidant activity [9,12], and parietin was shown to serve as both anti-aggregative and antioxidant agents in tauopathies [10,13]. In our continuous search for novel bioactive secondary metabolites from the Antarctic lichens, the acetone extract of R. terebrata was subjected to chemical investigation. Here, the identification of the unknown compound 1, stereocalpin B, is reported along with compound 3, which has been isolated as a natural product for the first time, together with a known cyclic depsipeptide (2). The structure elucidation of the new compounds (1 and 3) and the biological activities of all isolated compounds (1-3) are discussed below.
The absolute configurations of the amino acid units were assigned using the advanced Marfey's method [16,17]. LC-MS analysis of the acid hydrolysate of 1 revealed that the amino acid residues were L-Phe and N-Me-L-Phe, by comparing with the amino acid standards treated with Marfey's reagent (1-fluoro-2-4-dinitrophenyl-5-L-alanine amide, FDAA) ( Figure S11). Further study to assign the absolute configuration of the rest of the structure, 5-hydroxy-2,4-dimethyl-3-oxo-decanoic acid residue, was proposed by NOESY analysis. The nuclear Overhauser effects (NOEs) were observed between NH/H-13β, NH/H-21, H-21/H 3 -25, and H 3 -22/H-24, as reported for stereocalpin A by Seo et al. in 2008 [14]. The reported absolute configuration at C-21 for naturally occurring stereocalpin A is R; however, stereocalpin A possesses a potentially labile α-methyl β-keto carboxamide, where C-22 methyl functionality is capable of experiencing configurational change and converts into the 21-epi form via the enol (enolate) forms E1 and/or E2 during purification processes ( Figure S12), clarifying the misleading NOE correlations of H-21 and H 3 -22 [18]. This also implies the directions of C-22 and C-25 methyl groups toward the opposite sides, from the macrocycle ring [18]. NOEs were also observed between H-24/H-26, H 3 -25/H-26, and H 3 -25/H 2 -27 posing difficulty in corroborating the stereochemistry at C-24 and C-26. Instead, vicinal J coupling constants between H-24 and H-26 were indicative of cis configuration ( 3 J H-24/H-26 = 3.6 Hz in 1a and 2.7 Hz in 1b), suggesting close proximity of the two protons with a small dihedral angle [18]. Due to the conformational flexibility of the macrocyclic nature of 1, we could not obtain suitable ECD calculations, as flexible natural products are known to have a large number of stable conformers [19]. The conformational differences of the rotamers were observable in the NOESY spectrum (Figure 3), and careful analysis of the spectrum revealed that rotamers 1a and 1b share the same configurations of the cyclic depsipeptide backbone, but significant differences were observed due to conformational changes. Prominent differences were the observation of NOEs between H-2/H-12 and H-3a/H 3 -10 for the conformational isomer 1a, while they were not observed in case of 1b. Instead, correlations between H-2/H 3 -10 and H 3 -10/H-12 were observed in 1b ( Figure 3). Based on these data, an estimation could be made that free rotation around the C-2-C-3 and C-3-C-4 bonds caused conformational changes affecting the interproton distances between neighboring protons. We expected that principal distinctions that would have been made due to this rotation would be distances between H-3/H-2 and H-3/H 3 -10, where calibrated interproton distances between H-3a/H-2 and H-3a/H 3 -10, calculated using NOE-based interproton distance measurement technique via the peak amplitude normalization for improved cross-relaxation (PANIC) method [20][21][22][23][24], were 2.48 Å and 2.50 Å in 1a, respectively (Figures S13-S17). To predict the corresponding interproton distances in 1a and 1b, molecular mechanics (MM) and quantum mechanics (QM)-optimized 3D structures of each rotamer were generated, where the predicted values were 2.83 Å and 3.06 Å for H-3a/H-2, and 3.59 Å and 4.36 Å for H-3a/H 3 -10 in 1a and 1b, respectively. Although there was a discrepancy in calculated/predicted interproton distance between H-3a/H 3 -10, the calculated values were closer to that of 1a, and the difference in predicted values of each rotamer also confirmed the occurrence of free rotation around the C-2-C-3 and C-3-C-4 bonds. The calculated interproton distances of H-2/H 3 -10 and H 3 -10/H-12 were 2.33 Å and 2.18 Å in 1b, respectively (Figures S16 and S17), implying medium-strong NOEs [21]. δC 67.3, respectively. This appeared to be caused by an anisotropic shielding effect of the benzene ring in 1b, generating upfield shift for H-2 as compared with the corresponding proton chemical shift in 1a ( Figure 3) [25,26], as free rotation occurs around the C-2-C-3 and C-3-C-4 bonds. N-methyl group also seemed to be exposed to the shielding effect, where 1 H and 13 C chemical shifts were δH 3.06 and δC 29.5 in 1a, and δH 2.56 and δC 39.4 in 1b (Figure 3). Based on the above intensive structural analysis, the complete structure of 1 was determined as shown in Figure 1 and named stereocalpin B.  (Table 2) showed signals for two deshielded methyl groups at δH 1.99 (3H, s) and 2.66 (3H, s), and a chelated hydroxyl group at δH 13.37 (1H, s). The 13 C NMR data ( Table 2) showed only 9 carbon signals comprised of two methyl groups, a carbonyl group, oxygenated and quaternary aromatic carbons, which suggested that compound 3 is symmetric.  Additionally, a noticeable disparity in the chemical shifts was observed, where chemical shifts for H-2 were δ H 4.90 in 1a, δ H 3.33 in 1b, and 13 C chemical shifts were δ C 59.5 and δ C 67.3, respectively. This appeared to be caused by an anisotropic shielding effect of the benzene ring in 1b, generating upfield shift for H-2 as compared with the corresponding proton chemical shift in 1a ( Figure 3) [25,26], as free rotation occurs around the C-2-C-3 and C-3-C-4 bonds. N-methyl group also seemed to be exposed to the shielding effect, where 1 H and 13 C chemical shifts were δ H 3.06 and δ C 29.5 in 1a, and δ H 2.56 and δ C 39.4 in 1b (Figure 3). Based on the above intensive structural analysis, the complete structure of 1 was determined as shown in Figure 1 and named stereocalpin B.  (Table 2) showed signals for two deshielded methyl groups at δ H 1.99 (3H, s) and 2.66 (3H, s), and a chelated hydroxyl group at δ H 13.37 (1H, s). The 13 C NMR data ( Table 2) showed only 9 carbon signals comprised of two methyl groups, a carbonyl group, oxygenated and quaternary aromatic carbons, which suggested that compound 3 is symmetric. The HMBC correlations of H 3 -2 /C-1 (δ C 200.8), H 3 -2 /C-4 (δ C 101.0), 3-OH/C-4, 3-OH/C-2 (δ C 106.9), CH 3 -2/C-2, CH 3 -2/C-1 (δ C 157.0) and CH 3 -2/C-3 completed the partial structure of compound 3 (Figure 4). Based on the evidence that 3 had a symmetric structure, its spectroscopic data were compared with those previously reported, and it was found that compound 3 was structurally similar with 1,3,7,9-tetrahydroxy-2,8dimethyl-4,6-di(2-methylbutanoyl)dibenzofuran and 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6di(2-methylpropionyl)dibenzofuran [27]. The only difference was the side chains at C-4 and C-6, where the acetyl group was present in the case of 3. The hydroxyl groups attached at C-3 and C-7 (δ H 13.37) appeared to participate in a strong intermolecular hydrogen bond, each with OC-1 and OC-1 ' groups, respectively. The cross-peaks between 3-OH/H 3 -2 , and 3-OH/2-CH 3 in NOESY spectrum further supported the structure (Figure 4), deducing 3 to be 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6-di(ethanoyl)dibenzofuran. Compound 3 was registered in SciFinder with CAS registry number 674786-23-3; however, it has not been isolated as a natural product, and this is the first study to describe 3 in terms of its chemical nature. The HMBC correlations of H3-2'/C-1' (δC 200.8), H3-2'/C-4 (δC 101.0), 3-OH/C-4, 3-OH/C-2 (δC 106.9), CH3-2/C-2, CH3-2/C-1 (δC 157.0) and CH3-2/C-3 completed the partial structure of compound 3 (Figure 4). Based on the evidence that 3 had a symmetric structure, its spectroscopic data were compared with those previously reported, and it was found that compound 3 was structurally similar with 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6-di(2-methylbutanoyl)dibenzofuran and 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6-di(2methylpropionyl)dibenzofuran [27]. The only difference was the side chains at C-4 and C-6, where the acetyl group was present in the case of 3. The hydroxyl groups attached at C-3 and C-7 (δH 13.37) appeared to participate in a strong intermolecular hydrogen bond, each with OC-1' and OC-1'' groups, respectively. The cross-peaks between 3-OH/H3-2', and 3-OH/2-CH3 in NOESY spectrum further supported the structure (Figure 4), deducing 3 to be 1,3,7,9-tetrahydroxy-2,8-dimethyl-4,6-di(ethanoyl)dibenzofuran. Compound 3 was registered in SciFinder with CAS registry number 674786-23-3; however, it has not been isolated as a natural product, and this is the first study to describe 3 in terms of its chemical nature.

Biological Activities of the Isolated Compounds (1-3)
All isolated compounds were evaluated for antimicrobial activities against Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Candida albicans, and Mycobacterium smegmatis. The tested compounds (1-3) only showed significant activity against E. coli, where compounds 1-3 exhibited moderate inhibition with the IC50 values of 30, 28, and 18 μg/mL, respectively, compared to the positive control (Apramycin, IC50 = 2 μg/mL) (Table  S1). Compounds 1-3 were also tested for their cytotoxicity and anti-inflammatory activity. Compound 2 exhibited cytotoxicity against HCT116 cells, human colorectal carcinoma cell lines, with the IC50 value of 20 ± 1.20 μM (Table S2). Compounds 1and 2 showed strong inhibition on NO production in LPS-induced RAW264.7 macrophages, in which the IC50 values were 7 ± 0.004 μM and 5 ± 0.006 μM, respectively. Stereocalpins are a unique class of depsipeptides derived from the Antarctic lichens, having only stereocalpin A being previously reported. Only small groups of dibenzofurans have been reported thus far, and the most studied type is usnic acid, adding uniqueness to the identification of compound 3 as another type of dibenzofuran isolated from the Antarctic lichen. Based on these findings, it was rationally concluded that stereocalpins and dibenzofuran derivatives

Biological Activities of the Isolated Compounds (1-3)
All isolated compounds were evaluated for antimicrobial activities against Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Candida albicans, and Mycobacterium smegmatis. The tested compounds (1-3) only showed significant activity against E. coli, where compounds 1-3 exhibited moderate inhibition with the IC 50 values of 30, 28, and 18 µg/mL, respectively, compared to the positive control (Apramycin, IC 50 = 2 µg/mL) (Table S1). Compounds 1-3 were also tested for their cytotoxicity and anti-inflammatory activity. Compound 2 exhibited cytotoxicity against HCT116 cells, human colorectal carcinoma cell lines, with the IC 50 value of 20 ± 1.20 µM (Table S2). Compounds 1 and 2 showed strong inhibition on NO production in LPS-induced RAW264.7 macrophages, in which the IC 50 values were 7 ± 0.004 µM and 5 ± 0.006 µM, respectively. Stereocalpins are a unique class of depsipeptides derived from the Antarctic lichens, having only stereocalpin A being previously reported. Only small groups of dibenzofurans have been reported thus far, and the most studied type is usnic acid, adding uniqueness to the identification of compound 3 as another type of dibenzofuran isolated from the Antarctic lichen. Based on these findings, it was rationally concluded that stereocalpins and dibenzofuran derivatives from the Antarctic lichen R. terebrata might represent new structural classes of antimicrobial, anti-proliferative, and anti-inflammatory agents.

Lichen Material
The specimens of lichen, R. terebrata, were collected from King George Island, Antarctica, (62 • 12 53.69 S; 58 • 55 23.87 W) in January 2021, and identified by Dr. Ji Hee Kim. A voucher specimen (An-L16) was deposited at the Natural Product Chemistry Laboratory of the Korea Polar Research Institute.

Extraction and Isolation
The air-dried and powdered lichen, R. terebrata (14.0 g), was extracted in acetone by maceration (2 × 1 L) at room temperature. Extracts were filtered, and the filtrate was evaporated under reduced pressure with a rotavapor to obtain a crude extract (1.9 g). From the concentrated crude extract, a large portion of yellow powder (292.0 mg) was separated into a different fraction by scraping from the surface of the flask. It was then extracted with chloroform and subsequently subjected to preparative HPLC (MeOH/H

Computational Details
Initial conformational searches were performed at the MMFF94 force field using the MacroModel (Version 9.9, Schrödinger LLC) program with a mixed torsional/low-mode sampling method where the gas phase with a 20 kJ/mol energy window and 10,000 maximum iterations were employed. Quantum-mechanics geometry optimization of all the identified conformers within 10 kJ/mol of the relative MMFF94 energy level was performed utilizing the Tmolex 4.3.1 at the B3LYP/6-31 + G (d, p) theory level with the polarizable continuum model (PCM) (methanol) for solvation.

PANIC Analysis
The PANIC analysis was carried out according to the previous communication with slight modification. Slices of 2D NOESY spectrum of 1 were obtained using MestReNova software (version 14.1.2-25024). As a reference, the interproton distance between two aromatic protons with ortho-relationship, H-5 and H-6 (r reference ), was set as 2.5 Hz. The corresponding NOE intensity (NOE reference ) was obtained by integrating the signal of H-6 from the slice of 2D NOESY spectrum irradiated at H-5. The NOE intensities of H-3a in 1a (NOE unknown ), and H 3 -10 in 1b (NOE unknown ), were acquired by irradiation at H 3 -10 and H-2 of 1a and H-2 and H-12 of 1b and integration of H-3a and H 3 -10 signals, respectively. The calibrated interproton distances (r unknown ) were calculated using the equation below: (r unknown ) 6 (1)

Antimicrobial Assay
The antimicrobial activities were tested with 5 microorganisms in total in 96-well-plate; Escherichia coli KCTC 1682 (Korea Collection for Type Cultures) and Klebsiella pneumonia ATCC 13883 (American Type Culture Collection) representing Gram-negative bacteria, Staphylococcus aureus KCTC 3881 and Mycobacterium smegmatis ATCC 607 representing Gram-positive bacteria and Candida albicans KCTC 27242 representing fungi. Cell culture (95 µL) was diluted to 0.5 McFarland Standard scale in each well. Compounds dissolved in DMSO were added up to final concentrations (0.5, 1, 2, 5, 10, 20, 50, and 100 µg/mL). Total culture (100 µL) was incubated at 37 • C for 16 h. Cell inhibition was measured at 600 nm using MultiskanTM GO Microplate Spectrophotometer (Thermo Scientific, Waltham, MA, USA). The IC 50 value was calculated using an exponential trend line in Excel (Microsoft, Redmond, WA, USA). Apramycin and nystatin were used as a positive control against the bacterium and yeast, respectively. All measurements were performed in triplicates.

Cell Culture
RAW264.7 macrophages and HCT116 were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fatal bovine serum (FBS) and 1% penicillin at 37 • C in a humidified CO 2 incubator. In this study, macrophages were subjected in the presence or absence of different concentrations of the isolated compounds which were added 1 h prior to LPS (0.5 µg/mL) stimulation. HCT116 cells were seeded in a 96-well plate in triplicate at a density of 5 × 10 3 cells/well, and incubated in 5% CO 2 supplement at 37 • C.

MTS Assay
MTS assay was performed to determine the cytotoxic effects of the isolated compounds against HCT116 cells. The cells were seeded 2 × 10 5 cells/mL on a 96-well plate. After incubation for 24 h, 10% MTS solution was added to the cell culture medium, and it was incubated at 37 • C for 1 h. The concentration of the treated compounds was from 50 µM to 6.25 µM, using serial dilution. The absorbance was measured after 24 h using a microplate reader (Promega., Madison, WI, USA) at 490 nm.

Measurement of nitric oxide (NO) production
NO concentration in the RAW264.7 cell culture supernatant was measured using the Griess reagent. Briefly, 100 µL of the collected supernatant was mixed with equal amounts of Griess reagent (1% sulfanilamide in 5% phosphoric acid, 0.1% N-(1-naphthyl) ethylenediamine). The mixtures were incubated for 10 min at room temperature, and then the absorbance value of each well was determined at the wavelength of 540 nm using a microplate reader. Nitrite concentration was determined using a sodium nitrite calibration curve (0-100 µM).

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