New Sesterterpenes from the Antarctic Sponge Suberites sp.

1. Introduction

With an expansive range and biogeographic isolation, the waters of the Southern Ocean surrounding Antarctica host rich and diverse marine benthic invertebrate communities [1,2]. Marine sponges often act as the dominant taxa in many benthic habitats in the region, and these organisms often produce chemical defences to survive in communities structured by biotic influences like predation and competition for limited resources [2]. Sponges from Antarctica have been found to produce a wide variety of chemical compounds, including terpenes [3]. Naturally occurring terpenes have interesting chemical structures that exhibit bioactivities such as antimicrobial, anticancer, and anti-inflammatory properties that may have pharmaceutical applications in the field of drug discovery [4,5,6]. The isolation of these compounds reveals auspicious results for bioprospecting efforts in Antarctica [4].

Suberites is a genus of the family Suberitidae. These organisms can be found across various oceanic habitats including coral reefs, rocky intertidal zones, and deep sea environments [2]. Suberites spp. play an important role in the ecosystems they inhabit through their water filtration and benthic shaping qualities [2]. The genus Suberites belongs to the largest and most diverse class of sponges, Demospongiae, which leaves room for vast research to be conducted on the natural product compounds they possess. Our research group has previously studied Antarctic Suberites sp. and discovered a variety of unreported sesterterpenes, including neosuberitenone, which was characterized by a new carbon scaffold, as well as six suberitenone derivatives, an ansellane-type terpenoid, and a highly degraded sesterterpene [7]. Some of these compounds were reported with antiviral activity [7].

Herein, we report on our continuing investigation of the Antarctic sponge Suberites sp.. New specimens of the sponge were subject to ¹H NMR-guided purification using reverse-phase high-performance liquid chromatography (HPLC), resulting in the isolation of five new suberitane derivatives, three being derivatives of suberiphenol [8]. Additional supplies of our previously published suberitenones were isolated together with the two major metabolites suberitenone A (6) and B (7), and the minor metabolites suberitenone C (8), D (9), and oxaspirosuberitenone (10) [7,8,9,10]. We report the structure elucidation of the previously unreported compounds and their RSV antiviral activity.

2. Results and Discussion

The molecular formula of suberitenone K (1) was determined to be C₂₇H₃₈O₅ based on HRESIMS data ([M + Na]⁺ m/z 465.2622) and corroborated by the ¹H and ¹³C NMR spectra (Table 1). The ¹³C NMR spectrum indicated 1 to have four olefinic carbons, including two methines and two olefinic quaternary signals, as well as three carbonyls. These account for five degrees of unsaturation, indicating 1 to have a four ring system. The remaining 20 carbons are accounted for by three quaternary carbons, three aliphatic methines, two oxygen-bearing methines, six methyl groups, and six methylene groups. The ¹H NMR data (Table 1) showed two oxygen-bearing methines at δ_H 4.25 (H-1) and δ_H 5.56 (H-13), two olefinic methines at δ_H 6.80 (H-2) and δ_H 6.40 (H-22), three aliphatic methines at δ_H 3.35 (H-6), δ_H 1.96 (H-10), and δ_H 1.23 (H-14), and six singlet methyl groups. These NMR data indicated this compound to be structurally related to other reported compounds from Suberites sp. (Figure 1), known as the suberitane carbon skeleton [9].

The ring A partial structure could be established by a combination of COSY and HMBC correlations (Figure 2). Geminal methyl groups H₃-20 and H₃-25 (δ_H 0.94, 1.06) anchor the spin system, both displaying HMBC correlations to C-14, C-18, and C-19 (δ_C 57.3, 45.3, and 35.1, respectively). The methine H-14 (δ_H 1.23) displayed HMBC correlations to methyl groups C-20 (δ_C 33.3) and C-24 (δ_C 17.4) and the quaternary carbons C-15 (δ_C 38.2) and C-19. H₃-24 (δ_H 1.34) had HMBC correlations to C-10 (δ_C 54.9), C-14, C-15, and methylene C-16 (δ_C 42.2). The COSY data further connects H₂-17 (δ_H 1.81/1.51) to H₂-16 (δ_H 1.69/0.98) and H₂-18 (δ_H 1.41/1.26), resulting in ring A.

Extending the spin system, COSY correlations were observed between H-14 and oxymethine H-13 (δ_H 5.59), which further correlated with H₂-12a/12b (δ_H 2.06/1.66). H₂-12 had HMBC correlations to C-10, C-11 (δ_C 37.7), C-13 (δ_C 72.0), and C-14, closing the six-membered ring B with quaternary carbon C-15 between C-10 and C-14 based on the previously described HMBC correlation between H₃-24 and C-10. The singlet methyl H₃-23 (δ_H 1.29) had HMBC correlations to C-10, C-11, C-12, and the olefinic methine C-22 (δ_C 160.7), placing the methyl on quaternary carbon C-11. H₂-9 (δ_H 2.52/2.50) showed HMBC correlations to carbonyl C-8 (δ_C 202.1), C-10, and C-11. HMBC correlations from H-22 to C-8 and the olefinic quaternary carbon C-7 (δ_C 134.2) connects carbonyl C-8 and olefinic methine C-22 with C-7, completing ring C.

A separate spin system comprising ring D was constructed based on COSY correlations from oxymethine H-1 (δ_H 4.25) to aliphatic methine H-6 (δ_H 3.35) and olefinic methine H-2 (δ_H 6.80). Further, H-2 showed HMBC correlations to C-1 (δ_C 65.2), C-4 (δ_C 202.2), C-6 (δ_C 38.5), and C-21 (δ_C 15.7). The singlet methyl group H₃-21 (δ_H 15.7) displayed HMBC correlations to C-2 (δ_C 145.3), C-3 (δ_C 137.2), and C-4, placing it on the quaternary carbon C-3 with the olefinic methine (C-2) and ketone (C-4) on each side. The non-equivalent methylene H₂-5a/5b (δ_H 2.75/2.15) had HMBC correlations with C-1, C-4, and C-6, closing ring D. This was supported by COSY correlations from methine H-6 to H-1 and H₂-5a/5b.

The two spin systems were joined based on HMBC correlation between H-5a and H-6 with olefinic quaternary carbon C-7 (δ_C 134.2), establishing the suberitane backbone. The acetoxy group can be placed on oxygen-bearing C-13 based on the deshielded shift of both its carbon (δ_C 72.0) and proton (δ_H 5.59) and HMBC correlations from H-13 to C-26 (δ_C 172.1), completing the planar structure of suberitenone K (1).

The relative stereochemistry was determined by key NOE correlations between H₃-20, H-13, and H-14 and further between H-14 and H-10. H₃-23 has an NOE correlation to the acetoxy methyl, placing those on the same face of the ring. H₃-25 has an NOE correlation to H₃-24 placing those on the same face as the acetoxy, as H₃-25 is opposite to H₃-20. The relationship of H-1 and H-6 were determined to be syn to each other based on their coupling constants of J= 4.0. NOE correlations and coupling constants were used to determine 1 to share stereocenters of the same orientation as the known compound suberitenone A (6), suggesting they have the same absolute configuration of 1R,6R,10S,11S,13R,14S,15R as previously determined by Shin et al. [9] for suberitenone A using a modified Mosher’s method and CD analysis as well as our analysis of suberitenone E, which was confirmed by X-ray diffraction [7].

Suberitandiol (2) was determined to have the molecular formula C₂₇H₄₀O₄ by the negative ion formate adduct ([M + CHOO]^- 473.2908 m/z). The NMR data of 2 (Table 2) suggest ring A and B to be as found in 1. Similarly to 1, 2 has six methyl groups, one being part of the acetoxy group on C-13 (δ_C 70.6). The H₃-23 (δ_H 1.43) methyl has HMBC correlations to C-10 (δ_C 58.9), C-11 (δ_C 34.8), C-12 (δ_C 46.7), and C-22 (δ_C 57.2), being placed on the quaternary carbon C-11, connecting ring B and C as shown for 1. C-22 for 1 was an olefinic methine, but for 2 was found as a non-equivalent methylene. Ring C was further verified and connected through H-22a (δ_H 1.57) with HMBC correlations to C-8 (δ_C 41.1), C-10, C-11, C-23 (δ_C 22.7), and quaternary carbon C-7 (δ_C 74.4) as shown in Figure 3. Methylene H-9a (δ_H 1.86) displays COSY correlations to H₂-8 (δ_H 1.90) and H-10 (δ_H 1.06), establishing much of ring C. Ring C was completed between C-8 and C-9 by H-9b (δ_H 1.69), which demonstrated HMBC correlations to C-7, C-10, and C-11. With ring D left, three olefinic/aromatic methines, one deshielded methyl group, three quaternary carbons and two oxygens remain left to account. The three olefinic/aromatic methines and three quaternary carbons make up an aromatic ring with three substituents. The methine H-2 (δ_H 7.07 (d)) had COSY correlations to methine H-1 (δ_H 6.90), creating an olefinic bond and suggesting the loss of the oxygen-bearing methine previously observed in 1. Further, H-2 had HMBC correlations to the quaternary carbons C-4 (δ_C 152.6) and C-6 (δ_C 149.5) and the methyl C-21 (δ_C 15.3), as observed for 1. The third methine (H-5, δ_H 6.95) is a singlet with HMBC correlations to C-1 (δ_C 116.6) and C-3 (δ_C 122.0). The H₃-21 methyl (δ_H 2.23) had HMBC correlations to C-2 (δ_C 130.7), C-3 (δ_C 122.0), and C-4, placing the methyl on the quaternary carbon C-3. The aromatic methines H-1 and H-5 both have HMBC correlations to C-7, suggesting a bridge between C-6 and C-7, connecting ring C and D. This leaves an open valence on C-4 and C-7, with two oxygens and two protons left to account for, indicating an hydroxyl group on each to complete the planar structure of 2. These assignments are supported by their deshielded shifts. The 3D structure was solved by comparing NOE and coupling constants to 1. The tertiary alcohol at C-7 was deduced by NOE correlations in DMSO-d₆ from OH-7 (δ_H 4.38) to H₃-23 (δ_H 1.34). The absolute configuration is proposed as 7R,10S,11S,13R,14S,15R based on similarly to the known compound suberitenone B [9].

Abeosuberitandiol (3) was found to have the formula C₂₇H₃₈O₅ based on its ¹H and ¹³C NMR data (Table 2) and the formate anion adduct at 487.2697 m/z. The three aromatic methines and its 2D data (Figure S27-29) indicate 3 to have the rings A, B, and D as found for 2. Figure 4 highlights the HMBC and COSY correlations to establish the differences found in ring C. H₃-23 (δ_H 1.16) has HMBC correlations to C-10 (δ_C 56.4), C-12 (δ_C 42.6), and C-22 (δ_C 83.5) as shown for the other structures, however, differently from 1 and 2, H-22 (δ_H 4.02) has the shift of an oxygen-bearing methine. H-22 has COSY correlation to methine H-7 (δ_H 3.58), which further shows COSY correlations to the non-equivalent methylene H₂-9a/9b (δ_H 2.09/1.66). H₂-9a/9b has COSY correlation to H-10 (δ_H 1.30), completing ring C as a five-membered ring. H-9a has HMBC correlations to C-7 (δ_C 50.1), C-10 (δ_C 56.4), and the carbonyl C-8 (δ_C 202.1). H-22 and H-1 (δ_H 7.43) display HMBC correlations to C-8, as shown in Figure 4, suggesting the ketone acts as a bridge between ring C and D. The absolute configuration was deduced as earlier, showing the relative orientation of chiral centers 10R*,11R*,13R*,14S*,15R*. NOE correlations between H-7/H₃-23 and H-10/H-22 places H-7 and H-22 anti to one another, suggesting the absolute configuration of 3 to be 7S,10R,11R,13R,14S,15R,22R.

Furanosuberitandiol (4) was found with the molecular formula C₂₇H₃₆O₆, determined by the ¹H and ¹³C NMR data (Table 2) and the (-)HRESIMS ([M + CHOO]^- 501.2489 m/z). The formula suggests the structure to have ten degrees of unsaturation. The NMR data indicate 4 to have an aromatic ring, as observed for 2 and 3, and two carbonyl groups. This leaves 4 with 5 rings. Figure 5 highlights the key HMBC and COSY correlations to establish the aromatic ring, and further differences in ring C. Its NMR data (Table S4) suggests 4 to have the same A/B ring system observed for 1. Ring D differs in having two aromatic methines: C-2 (δ_C 112.1) and C-5 (δ_C 109.7). For 2 and 3, C-1 was observed as a methine; however, it is found in 4 as a quaternary carbon with a deshielded shift (δ_C 155.3), suggesting oxygen substitution. H₃-23 (δ_H 1.00) had HMBC correlations to C-10 (δ_C 45.9) and C-22 (δ_C 97.4), suggesting C-22 to be oxygen-bearing, as shown in 3. H-22 (δ_H 4.19) displays HMBC correlations to C-1 (δ_C 155.3), C-7 (δ_C 83.3), C-8 (δ_C 210.8), C-10 (δ_C 45.9), C-11 (δ_C 38.3), and C-23 (δ_C 18.9), suggesting a bridge between C-1 and C-22 through an oxygen. The non-equivalent methylene H-9a (δ_H 2.48) had COSY correlation to H-10 (δ_H 1.88) and HMBC correlations to quaternary carbon C-7 (δ_C 83.3) and carbonyl C-8 (δ_C 210.8), establishing further connectivity of ring C. The H-9a and H-22 correlations to C-7 and C-8 suggests completion of ring C between C-7 and C-22. To complete the planar structure, a hydroxyl group is left to account for and one and two valences are open on C-6 and C-7, respectively. This suggests the hydroxy group to be placed on the oxygen-bearing C-7 and a bridge between C-6 and C-7, establishing the fifth ring system. NOE correlations and coupling constants as described earlier were used to propose the absolute configuration of 5. Key correlations between H₃-23 and H-22 determined a syn relationship for these protons, while the C-7 tertiary alcohol was placed on the same face as H-22 and H₃-23, as shown for 3 and suberitenone B [9]. This was corroborated by key NOE correlations in (CD₃)₂SO between OH-7/H-22 and a weak correlation between OH-7/H₃-23 (Figure S42), giving 5 the proposed absolute configuration of 7S,10R,11R,13R,14S,15R,22R.

Norsuberitenone B (5) was isolated as a white crystal with the molecular formula C₂₀H₃₂O₃ based on the m/z 321.2433 proton adduct of the (+)HRESIMS. Its formula indicates five degrees of unsaturation with two being carbonyl groups, suggesting a three-ring system. Its NMR data (Table S5) suggests 5 to have a similar ring system as shown for 1, but lacking ring D. HMBC correlations from the methyl groups H₃-15 (δ_H 0.97) and H₃-18 (δ_H 1.02) to C-9 (δ_C 56.5), C-13 (δ_C 44.0), and C-14 (δ_C 34.0) and from H₃-17 (δ_H 1.23) to C-5 (δ_C 56.6), C-10 (δ_C 37.7), and C-11 (δ_C 42.3) suggests 5 to have a similar ring system for ring A and B as shown for previous suberitane compounds. The H₃-20 (δ_H 2.05) and its HMBC correlation to carbonyl C-19 (δ_C 170.3) suggests the presence of the acetoxy group on oxygen-bearing C-8 (δ_H 70.0). H₃-16 (δ_H 1.06) had HMBC correlations to C-1 (δ_C 59.6), C-5, C-6 (δ_C 37.9), and C-7 (δ_C 45.7), suggesting its place on the quaternary carbon C-6. The methylene H-1b (δ_H 2.03) and H-3b (δ_H 2.29) have HMBC correlations to a ketone (C-2: δ_C 210.9), suggesting a ketone to have replaced the tertiary alcohol bridging to ring D in 2. The stereochemistry was determined by key NOE correlations from H₃-15 to H-8 (δ_H 5.52) and H-9 (δ_H 1.15) and further between H-9 and H-5 (δ_H 1.50). NOE correlations were observed between H₃-16 and the acetyl methyl group. X-ray diffraction (Figure 6) confirmed the structure and determined the absolute configuration to be 5S,6S,8R,9S,10R.

Biosynthetically, the suberitane class differs from many other sesterterpenes due to the pendant ring D [9]. A proposed biosynthetic pathway of newly reported metabolites 1-5 is presented in Figure 7. The degraded suberitenone, norsuberitenone B (5), could be rationalized as the elimination product of ring D from suberitenone B. Compound 1 can be derived from the oxidation of a tetracyclic intermediate, as shown in Figure 7. The phenolic compounds 2-4 are perhaps derived from a protonation-catalyzed dehydration of the alcohol allylic to the α,β-unsaturated ketone.

Bracegirdle et al. reported RSV activity for the known compounds suberitenone A (6) and B (7) with IC₅₀ values of 7.8 μM and 3.2 μM, respectively, while the compounds suberitenone F, G, and H showed moderate activity [7]. Herein, we report weak RSV activity for suberitenone D (9) (Table 3), while the others displayed no activity. A decrease in activity is shown with the degraded sesterterpene 5 relative to suberitenone B, suggesting the presence of ring D to be important for further structure-activity relationship studies. Furthermore, no activity was shown for 2 and 3, suggesting the importance of the α,β-unsaturated ketone on ring D. Suberitenone D has a decrease in activity compared to suberitenone B, with the only structural difference being the acetoxy replacing the hydroxyl group on ring D. Suberitenone C showed no activity and 1 displayed moderate activity, suggesting the Δ⁷⁽²²⁾ olefin is favourable over the Δ⁸ olefin, but an α,β-unsaturated ketone on ring C will decrease the activity.

3. Materials and Methods

3.1. General Experimental Procedures

Optical rotations were measured using an AutoPol IV digital polarimeter at 589 nm with a 1 dm path length cell. UV/Vis spectra were extracted from HPLC chromatograms. NMR spectra were acquired using a Bruker Neo 400 MHz broadband spectrophotometer with a cryoprobe or a Bruker Neo 600 MHz broadband spectrophotometer. The residual solvent peaks were used as an internal chemical shift reference (CDCl₃: δ_C 77.0; δ_H 7.27, MeOD: δ_C 49.0; δ_H 3.31, (CD₃)₂SO : δ_C 39.5; δ_H 2.50). High-resolution mass spectrometry-liquid chromatography data were obtained on an Agilent 6540 LC-MS QTOF coupled to an Agilent Jet-stream electrospray ionization detector. H₂O (A) and 0.1% FA in CH₃CN (B) were used as mobile phases on a Phenomenex Kinetex C18 column (2.6 μm, 100 Å, 150 x 3 mm: 0.5 mL/min). Reverse-phase HPLC was performed on a Shimadzu LC20-AT system equipped with a photodiode array detector (M20A) using a preparatory Phenomenex C18 column (5 μm, 100 Å, 250 x 21.2mm: 10 mL/min) or a semi-preparatory Phenomenex C18 column (10 μm, 100 Å, 250 x 10 mm: 4 mL/min. The methanol and acetonitrile used for column chromatography were obtained from Fisher Co. and were HPLC grade (>99% purity) while the H₂O was distilled and filtered. Solvents mixtures are reported as % v/v.

3.2. Biological Materials, Extraction, and Isolation

Sponge specimens were collected from Palmer Station, Antarctica in 2018. The frozen sponge (1.7 g) was freeze-dried before extracted in MeOH twice overnight. An HP20 column was prepared by washing with (CH₃)₂CO and pre-equilibrated in H₂O. The extract was passed through the column before the eluent was diluted with equal volume of H₂O and passed back through the column. The eluent was diluted a second time and passed through the column again. The column was eluted with 750 mL of 1) 30% Me₂CO/H₂O, 2) 75% Me₂CO/H₂O and 3) 100% Me₂CO. A ¹H NMR-guided fractionation was completed on the second fraction until pure compounds were achieved. Initial fractionation was performed using a preparative C18 HPLC with a 70-100% ACN/H₂O gradient over 15 minutes, resulting in 12 fractions. ¹H NMR was performed on each fraction, prioritizing fractions with a methyl around 2 ppm and an oxygen-bearing methine around 4.5 ppm. Semipreparative C18 HPLC was performed on fraction 8 using a gradient of 50-100% MeOH/H₂O, resulting in norsuberitenone B (5: 4.9 mg). The same method was used to purify 1, 3, and 6 affording to suberitenone K (1: 2.0 mg), furanosuberitandiol (4: 0.9 mg), abeosuberitandiol (3: 1.2 mg), respectively. Fraction 11 was purified by semipreparative C18 HPLC using a gradient of 75-100% MeOH/H₂O, resulting in suberitandiol (2: 3.2 mg) and the known compound suberitenone C (8). Fraction 10 was purified using the same method as described for fraction 11, resulting in the known compounds suberitenone D (9) and oxaspirosuberitenone (10). Fraction 9 and 12 contained the known compounds suberitenone B (7) and A (6), respectively.

3.3. Spectroscopic Data for the Suberites (1-5)

Suberitenone K (1): white film; ${\left[\alpha \right]}_D^{22}$ -31.1 (c 0.14, MeOH); UV (ACN/H₂O) λ_max 229 nm; ¹H NMR (600 MHz) and ¹³C NMR (150 MHz), Table S1; HRESIMS m/z 465.2616 [M + Na]⁺ (calcd for C₂₇H₃₈O₅Na, 465.2622; Δ 1.38 ppm).

Suberitandiol (2): white film; ${\left[\alpha \right]}_D^{22}$ -3.4 (c 0.32, MeOH); UV (ACN/H₂O) λ_max 274 nm; ¹H NMR (400 MHz) and ¹³C NMR (100 MHz), Table S2; HRESIMS m/z 473.2908 [M + CHOO]^- (calcd for C₂₈H₄₁O₆, 473.2908; Δ 0.13 ppm).

Abeosuberitandiol (3): white film; ${\left[\alpha \right]}_D^{22}$ +8.0 (c 0.09, MeOH); UV (ACN/H₂O) λ_max 261, 310 (sh) nm; ¹H NMR (400 MHz) and ¹³C NMR (100 MHz), Table S3; HRESIMS m/z 487.2697 [M + CHOO]^- (calcd for C₂₈H₃₈O₇, 487.2701; Δ 0.88 ppm).

Furanosuberitandiol (4): white film; ${\left[\alpha \right]}_D^{22}$ -8.6 (c 0.09, MeOH); UV (ACN/H₂O) λ_max 261 nm; ¹H NMR (400 MHz) and ¹³C NMR (100 MHz), Table S4; HRESIMS m/z 501.2489 [M + CHOO]^- (calcd for C₂₈H₃₇O₈, 501.2494; Δ 0.98 ppm).

Norsuberitenone B (5): white film; ${\left[\alpha \right]}_D^{22}$ +3.0 (c 0.24, MeOH); UV (ACN/H₂O) λ_max 225 nm; ¹H NMR (400 MHz) and ¹³C NMR (100 MHz), Table S5; HRESIMS m/z 321.2433 [M + H]⁺ (calcd for C₂₀H₃₃O₃, 321.2435; Δ 0.68 ppm).

3.4. RSV Antiviral Assay

The procedure was performed as previously described [7].

3.5. X-ray Diffraction

X-ray diffraction data on norsuberitenone B (5) was measured on Bruker D8 Venture PHOTON II CMOS diffractometer equipped with a Cu Kα INCOATEC ImuS micro-focus source (λ = 1.54178 Å). Indexing was performed using APEX4 (Difference Vectors method) [11]. Data integration and reduction were performed using SaintPlus [12]. Absorption correction was performed by multi-scan method implemented in SADABS [13]. Space group was determined using XPREP implemented in APEX3 [14]. Structure was solved using SHELXT and refined using SHELXL-2019/1 (full-matrix least-squares on F2) through OLEX2 interface program [15,16]. The ellipsoid plot was made with Olex2 [16]. Hydrogen atoms of -OH group and H₂O molecules were freely refined.

4. Conclusions

Herein, we reported five unreported sesterterpenes during our ongoing investigation of the chemical diversity of Antarctica and the sponge Suberites sp.. Three of the metabolites displayed an isolated phenol ring (2-4) with abeosuberitandiol showing a ring contraction and furanosuberitandiol displayed a new dihydrofuran ring. The biological activity of the unreported metabolites were tested together with the known compounds suberitenone A-D and oxaspirosuberitenone. Suberitenone A, B, and D has displayed interesting activity against respiratory syncytial virus (RSV). This research highlights the importance of marine organisms and their potential in drug discovery and supports further investigation of the relatively unexplored polar habitats.