Chemistry and Bioactivity of the Deep-Water Antarctic Octocoral Alcyonium sp.

Chemical investigation of an Antarctic deep-water octocoral has led to the isolation of four new compounds, including three illudalane sesquiterpenoids (1–3) related to the alcyopterosins, a highly oxidized steroid, alcyosterone (5), and five known alcyopterosins (4, 6–9). The structures were established by extensive 1D and 2D NMR analyses, while 9 was verified by XRD. Alcyopterosins are unusual for their nitrate ester functionalization and have been characterized with cytotoxicity related to their DNA binding properties. Alcyopterosins V (3) and E (4) demonstrated single-digit micromolar activity against Clostridium difficile, an intestinal bacterium capable of causing severe diarrhea that is increasingly associated with drug resistance. Alcyosterone (5) and several alcyopterosins were similarly potent against the protist Leishmania donovani, the causative agent of leishmaniasis, a disfiguring disease that can be fatal if not treated. While the alcyopterosin family of sesquiterpenes is known for mild cytotoxicity, the observed activity against C. difficile and L. donovani is selective for the infectious agents.


Introduction
Corals are encountered from the tropics to the polar seas, found on seamounts or geological formations up to 6000 m below the ocean's surface [1,2]. In the south, corals are separated by the Antarctic Circumpolar Current from the contiguous oceans resulting in an ecological niche [3,4]. Biochemical knowledge of deep-water corals from Antarctica is impeded by the remoteness and extreme conditions required for access [5,6], leading to great interest in coral natural products for ecological and biomedical studies [7][8][9]. Past research suggests that deep-water coral species offer potential drug discovery resources from the terpenoids class, ranging from mono-to triterpenes [10][11][12]. Various cold-water terpenoids from deep-sea soft corals include the paesslerins [13], ainigmaptilones [14], and keikipukalides [6], many of which exhibit moderate cytotoxicity toward either human cancer cell lines or microbial pathogens [8].

Results and Discussion
Coral specimens were collected during a 2013 cruise to the Scotia Arc in the So Ocean near Shag Rocks, at a depth of between 126 and 130 m. Phylogenetic analy conducted on one specimen (WAM Z97931) using the msh1 sequence. The coral cl with other known Alcyonium spp. from the Southern Ocean region, but was di from those species ( Figure S1), leading to its current identification as Alcyonium sp The dichloromethane/methanol (1:1) extract of the freeze-dried coral was part between ethyl acetate and water, and the lipophilic partition was separated usin dient normal-phase medium pressure liquid chromatography (MPLC) system, y eight fractions. Several MPLC fractions were chosen for HPLC purification based characteristics of their 1 H NMR spectra. In particular, the mid-and late-polar fr displayed 1 H NMR signals characteristic of the previously reported alcyopterosins in particular the aromatic singlet (H-8) and a midfield oxymethylene (H2-4). Frac G, and H, eluting roughly between 60-90% ethyl acetate in hexane, were found to R 1 = H, R 2 = ONO 2 , R 3 = OAc: Alcyopterosin T (1) R 1 = C-10 ketone, R 2 = ONO 2 , R 3 = OAc: Alcyopterosin U (2) R 1 = C-10 ketone, R 2 = ONO 2 , R 3 = H: Alcyopterosin C (6) R 1 = H, R 2 = ONO 2 , R 3 = OH: Alcyopterosin G (7) R 1 = H, R 2 = OAc, R 3

Results and Discussion
Coral specimens were collected during a 2013 cruise to the Scotia Arc in the Southern Ocean near Shag Rocks, at a depth of between 126 and 130 m. Phylogenetic analysis was conducted on one specimen (WAM Z97931) using the msh1 sequence. The coral clustered with other known Alcyonium spp. from the Southern Ocean region, but was divergent from those species ( Figure S1), leading to its current identification as Alcyonium sp. indet.
The dichloromethane/methanol (1:1) extract of the freeze-dried coral was partitioned between ethyl acetate and water, and the lipophilic partition was separated using a gradient normal-phase medium pressure liquid chromatography (MPLC) system, yielding eight fractions. Several MPLC fractions were chosen for HPLC purification based on the characteristics of their 1 H NMR spectra. In particular, the mid-and late-polar fractions displayed 1 H NMR signals characteristic of the previously reported alcyopterosins [16,17], in particular the aromatic singlet (H-8) and a midfield oxymethylene (H 2 -4). Fractions F, G, and H, eluting roughly between 60-90% ethyl acetate in hexane, were found to harbor alcyopterosins E (4), C (6), G (7), 4,12-bis(acetyl)alcyopterosin O (8), and alcyopterosin L (9) (Table S1). Two new alcyopterosins (1, 2) were found in the earlier eluting MPLC fractions, D and E, and fraction H was found to contain the previously undescribed hydrolysis product (3) of alcyopterosin E (4), along with 4.
The spectral data for alcyopterosin U (2) were very similar to those of 1 and again reminiscent of the alcyopterosin family of metabolites. The HRESIMS ([M + H] + : m/z 336.1429; calcd for C 17 H 22 NO 6 : 336.1442) found that 2 has one additional oxygen and two protons fewer than 1. The IR spectrum displayed the same sharp bands at 1640 and 1280 cm −1 supportive of the nitrate ester moiety, along with the absorptions at 1700 and 1750 cm −1 typical of ketone and ester functions, respectively [16]. The most obvious difference between the 1 H NMR spectra of 1 and 2 was the absence of one methylene and the shift of the aromatic proton H-8, from δ H 7.06 in 1 to 7.64 in 2. The HMBC spectrum demonstrated a correlation between the gem-dimethyl protons (H 3 -14/15, δ H 1.24) and a carbon signal at δ C 211.4, reflecting a departure in 2 from the oxidation state of 1. Taken with the missing methylene group in 2, the ketone must be at C-1 or C-10. A methylene signal at δ H 3.03 (H 2 -1) also correlated in the HMBC spectrum to the ketone, as well as δ C 151.5 and 135.4. Because H 2 -12 (δ H 5.25) also had an HMBC correlation to δ C 151.5, but not to δ C 135.4, then δ C 151.5 must be C-2 and δ C 135.4 must be C-9. An HMBC correlation between H 2 -1 and C-3 secured the position of the carbonyl at C-10. Further 1 H and 13 C shifts as well as HMBC correlations ( Figure S10) supported the remaining substitution on the aromatic ring of 2 mirroring that observed for 1.
The 1 H NMR spectrum of alcyopterosin V (3) displayed a new pattern relative to those from 1 and 2, though certain resemblances remained. Lacking an acetoxy signal found in 1 and 2, the molecular formula of 3 was established as C 15  The aromatic ring was established to be very much like that for 1: from the HMBC, a significantly deshielded/aromatic proton at δ H 7.24 (H-8) correlated with δ C 141.1 (C-2 or C-9) and 142.4 (C-6), the latter of which also had HMBC correlation from highly deshielded/aromatic methyl at δ H 2.37 (H 3 -13). The aromatic methyl showed further HMBC correlations to δ C 130.0 (C-7) and 131.8 (C-8). With the observation of HMBC correlation of δ H 5.55 (H-5) to C-6 and δ C 122.5 (C-3), only C-2 and C-9 (δ C 146.7 and 141.1) remained to secure as part of the aromatic ring. H-8, H 2 -1 (δ H 3.04), and H 2 -10 (δ H 2.74), the only hydrogen-bearing carbons near C-2 and C-9, are all 2 or 3 bonds apart and thus cannot assist in the assignment. Instead, we have assigned C-2 and C-9 based on their shift comparisons to similar carbons in 1 and 2, but we note that they may be interchanged.
Substitution on the aromatic ring of 3 was completed by considering the HMBC correlations of the remaining protons and carbons. H 2 -1 and H 2 -10 were noted above as correlated in the HMBC with both C-2 and C-9, locating them on the ring relative to already established H 3 -13 and H-8; H 2 -10 was distinguished from H 2 -1 by HMBC correlation to C-8, disambiguating their relative positions. They also both correlated with C-11 (δ C 40.9) and C14/15 (δ C 28.8), completing the fused cyclopentane ring found on all the alcyopterosins. The final feature of alcyopterosin V was established by observation of the HMBC correlation of H-5 to both an oxymethylene (C-4, δ C 63.2) and an ester-type carbonyl at δ C 170.8 (C-12). As the protons of the oxymethylene (H-4a, δ H 4.25; H-4b, δ H 3.81) were COSY coupled to H-5, which was already affixed to the aromatic ring at C-6 as described above, the ester carbonyl must be located at C-3, completing a lactone ring. Insufficient material for optical spectra prevented comparison of the configuration of C-5 in 3 and alcyopterosin E (4), but 3 represents the nitrate ester hydrolysis product of 4, due to which we suggest the two will share a common configuration. Additional support for the assigned configuration comes from an analysis of the coupling constants for the chiral proton H-5 of 3, which match those of 4 in magnitude (3: 3 J 4a-5 = 2.5 Hz, 3 J 4b-5 = 6.1 Hz; 4: 3 J 4a-5 = 2.3 Hz, 3 J 4b-5 = 6.6 Hz).
Further work was done to bring forward additional alcyopterosins, and a subsequent extraction was conducted and similarly fractionated. Alcyosterone (5) eluted late in the silica gradient (hexanes to ethyl acetate), suggesting a moderately polar metabolite. Upon analysis, it was determined to have the molecular formula C 33 H 50 O 8 based on HRESIMS data that was corroborated by proton and carbon counts from their NMR spectra (Table 2). From the HRESIMS, the [M + H] + was observed at m/z 575.3555, and [M − HOAc] + was observed at m/z 515.3364. Analysis of the 13 C NMR spectrum supported the 33 carbons accounted for by the MS and further indicated a ketone (C-1, δ C 203.9), three ester-type carbons (C-1 , δ C 169.4; C-3 , δ C 170.4; C-5 , δ C 169.9), two olefinic carbons (C-2, δ C 128.4; C-3, δ C 142.5), and four carbon signals in the oxygen-bearing region (C-6, δ C 69.7; C-11, δ C 70.4; C-15, δ C 70.5; C-16, δ C 73.0). The HSQC established the two olefinic carbons and all four of the oxygen-bearing carbons as methines and further indicated five aliphatic methines, six aliphatic methylenes, and eight methyl carbons. The 1 H NMR spectrum provided few additional insights into this overview of alcyosterone other than to suggest that three of the methyl carbons were associated with acetate esters, based on their chemical shifts (H 3 -2 , δ H 1.93; H 3 -4 , δ H 2.06; H 3 -6 , δ H 2.02) and HMBC correlation to their respective ester carbonyl. The chemical shift of H-3 (δ H 6.58) and its associated carbon (C-3, δ C 142.5) supported the presence of a conjugated system, which must be an α,β-unsaturated ketone. The HMBC strengthened that assignment as both H-2 (δ H 5.83) and H-3 correlated with C-1 (δ C 203.9) (Figure 3). H-3 was further correlated in the HMBC with methine C-5 (δ C 46.6), while H-2 correlates to the quaternary C-10 (δ C 47.7) and the methylene C-4 (δ C 28.4). With correlations of H-4a (δ H 2.79) and H-4b (δ H 2.11) to C-2, C-3, C-5, and C-10, a six-membered ring was established bearing the aforementioned α,β-unsaturated ketone.
H2-12 to C-11, C-18 (δC 15.8)m and C-13 (δC 43.7), and H3-18 (δH 1.22) to C-13 and C-14 (δC 56.6), establishes a third ring fused to the previously established decalin. A fourth ring the five-membered ring of a steroid ring system, was established by observation of a COSY correlation between H-14 (δH 1.31) and H-15 (δH 5.34), between H-15 and H-16 (δH 5.51), and between H-16 and H-17 (δH 1.34), all of which were HMBC correlated with C 13.  Figure S22). These relationships were confirmed by XRD, which also provided the abso lute stereochemistry ( Figure 5).  Figure S22). These relationships were confirmed by XRD, which also provided the absolute stereochemistry ( Figure 5).  Alcyopterosins are known to be mildly cytotoxic toward human tumor cell lines [16,19] but little attention has been focused on their infectious disease (ID) activity. Metabolites from Alcyonium sp. indet. isolated in this study in sufficient quantity were therefore screened in three ID assays. Alcyopterosins V (3), E (4), and alcyosterone (5) were inactive against the ESKAPE panel of bacterial pathogens, but both 3 and 4 demonstrated potent activity against Clostridium difficile, a difficult-to-treat intestinal bacterium which afflicts up to half a million people annually and caused 30,000 deaths in 2015 [20]. Alcyopterosin E (MIC 6.9 μM) was slightly more active against C. difficile than alcyopterosin V (MIC 8.1 μM). Cytotoxicity against host cell lines HEK293T and HepG2 also found 4 less toxic (CC50 570 and 331 μM, respectively) than 3 (CC50 220 and 288 μM, respectively). Vancomycin as a control displays an MIC of 0.34 μM against C. difficile and was non-toxic to the host cells at the same concentrations alcyopterosins were assayed.
Alcyopterosin C, E (4), L, 4,12-bis(acetyl)alcyopterosin O, V (3), and alcyosterone (5) were screened against Leishmania donovani and found with roughly equal, single digit μM, activity [21]. Leishmania, the disease caused by L. donovani, is often disfiguring and can lead to death if not properly treated, though current treatment regimes can be expensive and toxic, and are considered ineffective [22]. The highest potency was displayed by 4,12-Bis(acetyl)alcyopterosin O (IC50 1.2 μM), though alcyosterone (IC50 1.5 μM), alcyopterosin L (IC50 2.4 μM), and alcyopterosin E (IC50 3.1 μM) were largely indistinguishable. Alcyopterosin V (IC50 7.0 μM) and alcyopterosin C (IC50 13 μM) were only slightly less potent than the control, miltefosine (IC50 6.2 μM). Only 3 and 4 were available in sufficient quantity to assay against the Leishmania host cell line, J774.A1 macrophages, which showed alcyopterosin E, though low in toxicity, was twice as toxic (IC50 62 μM) as alcyopterosin V (IC50 110 μM) to the mammalian cells.   Alcyopterosins are known to be mildly cytotoxic toward human tumor cell lines [16,19] but little attention has been focused on their infectious disease (ID) activity. Metabolites from Alcyonium sp. indet. isolated in this study in sufficient quantity were therefore screened in three ID assays. Alcyopterosins V (3), E (4), and alcyosterone (5) were inactive against the ESKAPE panel of bacterial pathogens, but both 3 and 4 demonstrated potent activity against Clostridium difficile, a difficult-to-treat intestinal bacterium which afflicts up to half a million people annually and caused 30,000 deaths in 2015 [20]. Alcyopterosin E (MIC 6.9 μM) was slightly more active against C. difficile than alcyopterosin V (MIC 8.1 μM). Cytotoxicity against host cell lines HEK293T and HepG2 also found 4 less toxic (CC50 570 and 331 μM, respectively) than 3 (CC50 220 and 288 μM, respectively). Vancomycin as a control displays an MIC of 0.34 μM against C. difficile and was non-toxic to the host cells at the same concentrations alcyopterosins were assayed.
Alcyopterosin C, E (4), L, 4,12-bis(acetyl)alcyopterosin O, V (3), and alcyosterone (5) were screened against Leishmania donovani and found with roughly equal, single digit μM, activity [21]. Leishmania, the disease caused by L. donovani, is often disfiguring and can lead to death if not properly treated, though current treatment regimes can be expensive and toxic, and are considered ineffective [22]. The highest potency was displayed by 4,12-Bis(acetyl)alcyopterosin O (IC50 1.2 μM), though alcyosterone (IC50 1.5 μM), alcyopterosin L (IC50 2.4 μM), and alcyopterosin E (IC50 3.1 μM) were largely indistinguishable. Alcyopterosin V (IC50 7.0 μM) and alcyopterosin C (IC50 13 μM) were only slightly less potent than the control, miltefosine (IC50 6.2 μM). Only 3 and 4 were available in sufficient quantity to assay against the Leishmania host cell line, J774.A1 macrophages, which showed alcyopterosin E, though low in toxicity, was twice as toxic (IC50 62 μM) as alcyopterosin V (IC50 110 μM) to the mammalian cells. Alcyopterosins are known to be mildly cytotoxic toward human tumor cell lines [16,19] but little attention has been focused on their infectious disease (ID) activity. Metabolites from Alcyonium sp. indet. isolated in this study in sufficient quantity were therefore screened in three ID assays. Alcyopterosins V (3), E (4), and alcyosterone (5) were inactive against the ESKAPE panel of bacterial pathogens, but both 3 and 4 demonstrated potent activity against Clostridium difficile, a difficult-to-treat intestinal bacterium which afflicts up to half a million people annually and caused 30,000 deaths in 2015 [20]. Alcyopterosin E (MIC 6.9 µM) was slightly more active against C. difficile than alcyopterosin V (MIC 8.1 µM). Cytotoxicity against host cell lines HEK293T and HepG2 also found 4 less toxic (CC 50 570 and 331 µM, respectively) than 3 (CC 50 220 and 288 µM, respectively). Vancomycin as a control displays an MIC of 0.34 µM against C. difficile and was non-toxic to the host cells at the same concentrations alcyopterosins were assayed.

General Experimental Procedures
Optical rotations were measured on a Rudolph Research Analytical AUTOPOL IV digital polarimeter at 589 nm. UV absorptions were acquired with an Agilent Cary 60 UV-vis spectrophotometer. IR spectra were recorded with an Agilent Cary FTIR 630 spectrometer and PerkinElmer Spectrum Two equipped with a UATR (single reflection diamond) sample introduction system. NMR spectra were recorded on Varian Direct Drive 500 MHz and Varian Inova 500 MHz spectrometers. Chemical shifts are reported with the use of the residual CDCl 3 signals (δ H 7.27 ppm; δ C 77.0 ppm) as internal standards for 1 H and 13 C NMR spectra, respectively. COSY, HSQC, HMBC, and ROESY experiments corroborated the 1 H and 13 C NMR assignments. Analytical LC/MS with a Phenomenex Kinetex C18 column (50 × 2.1 mm, 2.6 µm) on an Agilent 6230 LC/TOF-MS with electrospray ionization detection provided the high-resolution masses. Semi-preparative and analytical HPLC separations were performed on a Shimadzu LC-20 AT system equipped with an ultraviolet (UV) detector using a Luna silica column (5 µm, 250 × 10 mm), and a YMC C-18 column (10 µm, 150 × 4 mm). MPLC was performed on a Teledyne Isco CombiFlash Rf 200i equipped with an evaporative light-scattering detector (ELSD) and a multiwavelength UV detector using a RediSep Rf silica 80 g flash column, and silica gel 230-400 mesh was used to load samples.

Biological Material
The soft coral was collected via trawling on the R/V Nathaniel B. Palmer vessel during the austral autumn in late April 2013. The specimens were collected between 126 and 130 m depth, frozen immediately upon collection, and maintained at −80 • C until extraction. The tissue of the frozen specimens was subsampled and preserved in 96% ethanol. Subsequent extraction was performed using a DNeasy blood and tissue kit (Qiagen) following manufacturer's protocols. Using primers ND42599F/mut3458R [23,24], a piece of the mitochondrial genome was amplified (msh1, a homolog of mutS). Cycling conditions included an initial 5× cycles at 45 • C annealing, followed by 39× cycles at 58 • C. Amplicons were sent to the Australian Genome Research Facility, Perth for purification and Sangar sequencing. The resulting bi-directional sequence was assembled and edited, primers removed, deposited in GenBank (OP429120), and aligned with other soft coral sequences from GenBank. A Maximum-Likelihood analysis using IQ-tree [25], implementing the evolutionary model VM+F+G4 selected with ModelFinder [26], was carried out. The nodes were tested with 1000 ultrafast bootstrap replicates.
Soxhlet extraction of an additional specimen in dichloromethane followed by a similar chromatographic profile described above resulted in seven fractions. Further purification of fraction E, via normal phase HPLC with a hexane-ethyl acetate (1:1) gradient, followed by reversed-phase HPLC using a water-acetonitrile (70% to 100%) gradient, led to alcyosterone (5)

Leishmania donovani Infected Macrophage Assay
The Leishmania donovani infected macrophage assay and cytotoxicity screen were conducted as previously described [27].

Clostridium difficile Susceptibility Screening
The screening against C. difficile was performed in two steps. In step 1, overnight culture of a hyper-virulent clinical strain C. difficile UK6 was inoculated into a fresh BHIS medium at a volume ratio of 1:1000. After pre-incubation at 37 • C under an anaerobic atmosphere for 2 h, the bacterial culture was divided into a sterile 96-well plate and each well contained 192 µL of bacterial culture. Then, 8 µL of each extract was added to each well of the plate, mixed thoroughly, and incubated at 37 • C in an anaerobic chamber for 48 h. Control groups of 200 µL of BHIS medium only, 200 µL of bacterial culture only, and 192 µL of bacterial culture in 8 µL of DMSO were also included in separate columns within each plate. Extracts that displayed initial antibacterial activity were further evaluated for their minimum inhibitory concentration (MIC) against C. difficile. Serial dilutions of each extract (400 µg/mL, 200 µg/mL, 100 µg/mL, 50 µg/mL, 20 µg/mL, 10 µg/mL, 5 µg/mL, and 2 µg/mL) were prepared in a fresh BHIS medium. Then, 100 µL of each extract dilution was added to 100 µL of bacterial culture (pretreated as described), mixed well, and incubated at 37 • C in an anaerobic chamber for 48 h. Control groups including wells containing fresh medium only and bacterial culture only were also included as described. Activity was determined as +/− (clear or turbid (OD 600 ) culture). The MICs of the three recommended antibiotics metronidazole, vancomycin, and fidaxomicin against C. difficile UK6 were also determined using broth microdilution methodology.

Determination of the Half Maximal Inhibitory Concentration (IC 50 ) toward Human Liver Cells and Kidney Cells
The cytotoxicity of the metabolites to human liver cells and kidney cells was determined using an MTT based-In Vitro Toxicology Assay Kit (Sigma-Aldrich, St. Louis, MO, USA) following the manufacturing instructions. The human kidney HEK293T cells and the human liver HEPGZ cells were used for the evaluation in this study. Both cell samples were maintained and suspended in Dulbecco's Modified Eagle Medium (DMEM with 4.5 g/L glucose, L-glutamine and sodium pyruvate, Corning, Manassas, VA, USA) containing 10% fetal bovine serum (Thermo Scientific) and 1% penicillin/streptomycin at 37 • C under 5% CO 2 atmosphere. The cells were plated on a 96-well plate with approximately 5 × 10 3 -1 × 10 4 cells in each well, and incubated at 37 • C overnight. After that, each of the selected extracts from the antimicrobial susceptibility test was added to the wells and incubated with the cells at a series of 2-fold diluted concentrations ranging from 128 µg/mL to 0.125 µg/mL. Following a 24 h of incubation, 10 µL of 1-(4,5-dimethylthiazol-2-yl)-3,5-diphenylformazan (MTT) stock solution (5 mg/mL) was added to each well of the cells, mixed well, and incubated at 37 • C for another 4 h. After that, the liquid in each well of the plate was removed carefully and thoroughly, then the cells in the wells were treated with 100 µL of DMSO, and incubated at 37 • C for 15 m. Optical density (OD) values were measured at a wavelength of 540 nm (OD 540 ) using a microplate reader (Synergy HTX; Bio Tek Instruments, Inc. Winooski VT). Cells treated with vancomycin, a common option for treating CDI in clinical settings, were also included in the MTT tests as a control. Cell survival and the IC 50 were calculated according to the method used in a previous publication [26]: Survival of cells (%) = Drug-treated group OD 540 /control group OD 540 × 100. The IC 50 value was calculated as follows: lgIC 50 = X m − I [P − (3 − P m − P n )/4], where X m was the log maximum dose, I was the log (maximum dose/adjacent dose), P was the sum of the positive response rate, P m was the maximum positive response rate, and P n was the minimum positive response rate. (5) XRD methodology was conducted as we have previously done [28]. Data and refinement conditions are shown in Table S2. CCDC Deposition Number 2205919.

Patents
US patent 10,898,460, Leishmania Inhibitors, based on portions of this work was awarded 26 January 2021.

Institutional Review Board Statement: Not applicable.
Data Availability Statement: The mutS gene sequence was deposited with Genebank (accession number OP429120). X-ray crystallographic data was deposited with the Cambridge Crystallographic Data Center (Deposition Number 2205919).