Merosesquiterpene Congeners from the Australian Sponge Hyrtios digitatus as Potential Drug Leads for Atherosclerosis Disease

A study of the chemical constituents from the Australian Sponge Hyrtios digitatus has provided a perspective on the connection between the chemistry and biology of the puupehenones, a unique and unusual class of merosesquiterpenes. In this study, a new tetracyclic merosesquiterpene, 19-methoxy-9,15-ene-puupehenol (1) was isolated from the marine sponge Hyrtios digitatus along with the known 20-methoxy-9,15-ene-puupehenol (2). Their structures were elucidated on the basis of spectroscopic data (1H and 13C NMR) in combination with experimental electronic circular dichroism (ECD) data. Compounds 1 and 2 are active at 1.78 μM and 3.05 μM, respectively, on Scavenger Receptor-Class B Type 1 HepG2 (SR-B1 HepG2) stable cell lines, targeting atherosclerosis disease.


Introduction
Marine diversity contributes a plethora of natural metabolites often without structural precedent elsewhere in the natural world [1,2]. During the past 30 years, more than 20,000 marine natural product compounds have been identified and marine derived compounds have been approved as therapeutics agents [3]. According to a recent review, 1378 new marine natural products were reported in the year 2014, indicating the marine environment as a huge source of chemical diversity and interestingly, 283 new compounds were identified from the phylum of Porifera [4]. Over the past 30 years, the FDA approved eight marine drugs which are Halaven ® , Lovaza ® , Adcetris ® , Prialt ® , Yondelis ® , Cytosar-U ® , Vira-A ® , and Carragelose ® [5]. Currently, three marine compounds (plitidepsin, plinabulin, and tetrodotoxin) are in phase III clinical trials. The clinical trial pipeline for marine drugs is promising [6]. In the course of ongoing investigations aimed at the identification of compounds that can modulate atherosclerosis disease, we identified the Australian sponge Hyrtios digitatus, which showed anti-atherosclerotic activity in our initial screening studies, using a Scavenger Receptor-Class B Type 1 (SR-B1) reporter gene assay.
Herein, we report the isolation and structural elucidation of a new 19-methoxy-9,15-enepuupehenol (1) and the known 20-methoxy-9,15-ene-puupehenol (2) (Figure 1) from an active fraction of the Australian sponge Hyrtios digitatus. Compound 2 was previously found from an Indo-Pacific Hyrtios sponge. The molecular structures of 1 and 2 were established on the basis of 1D and 2D NMR, UV, and HRESIMS data. The stereochemistry assignment of 1 and 2 using experimental and calculated electronic circular dichroism (ECD) is also described. Compounds 1 and 2 were evaluated for their up-regulatory activity on the anti-atherosclerotic SR-B1 HepG2 stable reporter cell line with the value of half maximal effective concentration (EC50) of 1.78 μM and 3.05 μM for 1 and 2, respectively. Trichostatin A (TSA) was used as positive control displaying an EC50 of 5.25 μM.
The establishment of ring C to ring D was confirmed by gHMBC correlations between the olefinic proton H-15 to C-16 (δ C 115.0 ppm), C-17 (δ C 145.9 ppm), and C-21 (δ C 110.9 ppm). The methoxy proton signal H-22 (δ H 3.68 ppm) was correlated with C-19 (δ C 141.9 ppm), while a hydroxyl proton signal (δ H 8.98 ppm) was correlated with C-18, C-19, and C-20. gHMBC correlations also observed from an aromatic proton signal, H-18 (δ H 6.22 ppm) to C-17 (δ C 145.9 ppm) and C-16 (δ C 115.0 ppm), supported the C/D ring system. The relative stereochemistry of 1 was deduced on the basis of nuclear overhauser effect spectroscopy (NOESY) interactions ( Figure 3). NOESY analysis disclosed interactions between two proton singlets, H-12 (δ H 0.92 ppm) and H-14 (δ H 1.14 ppm), which revealed their co-facial relationship and assignment as β-oriented. It indicated a trans fusion of rings A/B. Meanwhile, the NOESY correlation between a methine proton signal, H-5 (δ H 1.37 ppm) and a methyl proton, H-13 (δ H 1.25 ppm) implied a cis fusion of rings B/C. Taken together, the relative configuration of 1 was assigned as 5S*,8S*,10S* as shown in Figure 3.
The experimental ECD spectrum (Figure 4a), matched with the ECD spectrum of (5S,8S,10S)-1 ( Figure 5) with 98.8% relative populations. In the simulated ECD spectra of the four possible diastereomers of 1 as shown in Figure 5, the ECD spectrum of (5S,8S,10S)-1 exhibited a clear designated ECD curve, with a negative peak at 214 nm and a positive peak at 238 nm, 278 nm, and 320 nm. The UV-Vis spectrum of 1 exhibited absorption maxima at 217, 235, 279, and 325 nm (Figure 4b), which agreed with the experimental ECD spectrum as in Figure 4a. Four bands at 217 nm for n→π* transition, 235 nm for π→π* transition, 279 nm for π→π* transition, and 325 nm for π→π* transition, were indeed easily observed in the UV-Vis spectrum of 1. Thus, after considering the close similarities and supported by the ECD spectrum, the assignment of the absolute configuration of 1 as (5S,8S,10S) was confirmed ( Figure 6). The spectrum of 1 was virtually identical to that of compound 2 (Figure 4a). Hence, compound 2 possesses the same absolute configuration for all stereogenic centers as compound 1, (5S,8S,10S).
Mar. Drugs 2017, 15, 6 5 of 10 diastereomers of 1 as shown in Figure 5, the ECD spectrum of (5S,8S,10S)-1 exhibited a clear designated ECD curve, with a negative peak at 214 nm and a positive peak at 238 nm, 278 nm, and 320 nm. The UV-Vis spectrum of 1 exhibited absorption maxima at 217, 235, 279, and 325 nm ( Figure  4b), which agreed with the experimental ECD spectrum as in Figure 4a. Four bands at 217 nm for n→π* transition, 235 nm for π→π* transition, 279 nm for π→π* transition, and 325 nm for π→π* transition, were indeed easily observed in the UV-Vis spectrum of 1. Thus, after considering the close similarities and supported by the ECD spectrum, the assignment of the absolute configuration of 1 as (5S,8S,10S) was confirmed ( Figure 6). The spectrum of 1 was virtually identical to that of compound 2 (Figure 4a). Hence, compound 2 possesses the same absolute configuration for all stereogenic centers as compound 1, (5S,8S,10S).    diastereomers of 1 as shown in Figure 5, the ECD spectrum of (5S,8S,10S)-1 exhibited a clear designated ECD curve, with a negative peak at 214 nm and a positive peak at 238 nm, 278 nm, and 320 nm. The UV-Vis spectrum of 1 exhibited absorption maxima at 217, 235, 279, and 325 nm ( Figure  4b), which agreed with the experimental ECD spectrum as in Figure 4a. Four bands at 217 nm for n→π* transition, 235 nm for π→π* transition, 279 nm for π→π* transition, and 325 nm for π→π* transition, were indeed easily observed in the UV-Vis spectrum of 1. Thus, after considering the close similarities and supported by the ECD spectrum, the assignment of the absolute configuration of 1 as (5S,8S,10S) was confirmed ( Figure 6). The spectrum of 1 was virtually identical to that of compound 2 (Figure 4a). Hence, compound 2 possesses the same absolute configuration for all stereogenic centers as compound 1, (5S,8S,10S).     (Figure 7). The efficacies of 1 and 2, compared with TSA (100%), were estimated to be 130% and 121%, respectively (Table 2). Hence, compounds 1 and 2 are full agonists. It has been widely reported that SR-B1, the high-density lipoproteins (HDL) receptor, plays an important role in the development of atherosclerosis [26][27][28]. Zhang and his coworkers demonstrated an increase of 86% in mean atherosclerotic lesion of the proximal aorta in SR-B1 −/− as compared to SR-B1 +/+ apolipoprotein E-deficient mice that strongly suggested the antiatherogenic nature of SR-B1 in an in vivo model [29]. Therefore, this study indicates the potential role of the merosesquiterpene class of compounds in reducing the progression of atherosclerosis.  Drug-likeness is a property that is of importance for compounds in the drug discovery effort. There are various rules to evaluate the drug-likeness properties, such as Lipinski's rule and Veber's  (Figure 7). The efficacies of 1 and 2, compared with TSA (100%), were estimated to be 130% and 121%, respectively (Table 2). Hence, compounds 1 and 2 are full agonists. It has been widely reported that SR-B1, the high-density lipoproteins (HDL) receptor, plays an important role in the development of atherosclerosis [26][27][28]. Zhang and his co-workers demonstrated an increase of 86% in mean atherosclerotic lesion of the proximal aorta in SR-B1 −/− as compared to SR-B1 +/+ apolipoprotein E-deficient mice that strongly suggested the antiatherogenic nature of SR-B1 in an in vivo model [29]. Therefore, this study indicates the potential role of the merosesquiterpene class of compounds in reducing the progression of atherosclerosis.  (Figure 7). The efficacies of 1 and 2, compared with TSA (100%), were estimated to be 130% and 121%, respectively (Table 2). Hence, compounds 1 and 2 are full agonists. It has been widely reported that SR-B1, the high-density lipoproteins (HDL) receptor, plays an important role in the development of atherosclerosis [26][27][28]. Zhang and his coworkers demonstrated an increase of 86% in mean atherosclerotic lesion of the proximal aorta in SR-B1 −/− as compared to SR-B1 +/+ apolipoprotein E-deficient mice that strongly suggested the antiatherogenic nature of SR-B1 in an in vivo model [29]. Therefore, this study indicates the potential role of the merosesquiterpene class of compounds in reducing the progression of atherosclerosis.  Drug-likeness is a property that is of importance for compounds in the drug discovery effort. There are various rules to evaluate the drug-likeness properties, such as Lipinski's rule and Veber's rule. Lipinski's rule criteria states that molecular weight (MW) ≤500, partition coefficient (log P) ≤5, hydrogen bond acceptors (HBA) ≤10, and hydrogen bond donors (HBD) ≤5 [30]. In Veber's rule, the other two parameters considered are that polar surface area (PSA) ≤140 and the number of rotatable bonds (NROT) ≤10 [31]. Thus, compound 1 and 2 were further investigated for drug-likeness evaluation using Lipinski's rule and Veber's rule.
The four Lipinski properties of compounds 1 and 2 were evaluated using the Instant J-Chem 5.12.0 software ( Table 3). The results of the calculated MW, HBA, and HBD for compounds 1 and 2 comply with properties of Lipinski's rule. Applying Veber's rule (Instant J-Chem 5.12.0 software), compounds 1 and 2 demonstrated a PSA value of 38.69 and NROT as only 1 unit, satisfying two more requirements for drug-like molecules. Referring to the calculated value of Lipinski's and Veber's rule, compounds 1 and 2 have physicochemical properties consistent with predicted oral bioavailability [32] ( Table 3).

Animal Material
Samples of Hyrtios digitatus were collected at the depth of −17 m, latitude −21.70, longitude 152.55, Turner Reef, W. side, Swain Reefs, Queensland, Australia. It was identified as Hyrtios digitatus (phylum Porifera, class Demospongiae, order Dyctyoceratida, family Thorectidae). A voucher specimen G305703 has been deposited at the Queensland Museum, South Brisbane, Queensland, Australia.

ECD Calculation
All molecular mechanics analyses and calculations were determined using Macromodel interfaced to the Maestro program [33,34]. The initial conformations search of 19-Methoxy-9,15-ene-puupehenol (1) and 20-Methoxy-9,15-ene-puupehenol (2) were optimized using the OPLS-2005 force field method applying a 21 kJ/mol energy window [25]. The optimized conformations were used for the ECD calculations, which were performed using the (b3lyp/631gd) basis set supported by Gaussian 09 software [35]. ECD spectra were generated using the program SpecDis Version 1.61 software (University of Wuerzburg, Wuerzburg, Germany) applying a Gaussian band shape with the width of 0.4 eV. Boltzmann distributions were estimated from the (b3lyp/631gd) free energies in the solvent model calculations.

Atherosclerosis Assays
Hepatocellular carcinoma cells (HepG2) stably transfected with SR-B1 were kindly provided by the Malaysian Institute of Pharmaceuticals and Nutraceuticals (IPharm), Penang, Malaysia. SR-B1 HepG2 stable cell lines were grown in MEM media (Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS). Cells were grown under 5% CO 2 in a humidified environment at 37 • C. Thirty microliters (30 µL) of media containing 10,000 cells were added to a 384 well microtitre white clear bottom plate (Perkin Elmer, Waltham, MA, USA, part number: 353963) containing 0.3 µL of a compound. Final compound concentration range tested was 100 µM to 0.1 µM (final DMSO concentration of 1.0%). Each concentration in the media was tested in triplicate and in two independent experiments. Cells and compounds were then incubated for 48 h at 37 • C, 5% CO 2 and 80% humidity. Cell proliferation was measured with the addition of 30 µL of a Luciferase reagent (Invitrogen, Carlsbad, CA, USA) solution to each well of the microtitre plate. The plates were incubated at room temperature for 10 min. The luminescence of each well was read on the VICTOR X Multilabel Plate Readers (Perkin Elmer). Nine-point concentration-response curves were then analyzed using non-linear regression and EC50 values were determined by using GraphPad Prism 5 (GraphPad Software Inc., La Jolla, CA, USA). Trichostatin-A (TSA) was used as a positive control compound.

Conclusions
Contributing to the chemical study on Hyrtios digitatus, a new (5S,8S,10S)-19-methoxy-9,15-enepuupehenol (1) and the known (5S,8S,10S)-20-methoxy-9,15-ene-puupehenol (2) belonging to the merosesquiterpene structure class were identified. Structures of these compounds were determined by NMR spectroscopy and confirmed by ECD analysis. These two compounds satisfy properties for drug-like molecules with one violation (c logP) for oral bioavailability. This is the first time that the merosesquiterpene structure class has been reported to increase the activity of SR-B1 in a HepG2 cell line, indicating the potential role of these compounds in therapeutic intervention against atherosclerosis.  Figure S11: HSQC Spectrum of 2 in DMSO-d 6 , Figure S12: HMBC Spectrum of 2 in DMSO-d 6 , Figure S13: NOESY Spectrum of 2 in DMSO-d 6 , Figure S14: ROESY Spectrum of 2 in DMSO-d 6 , Figure S15: Optimized structure of 1, Figure S16: Optimized structure of 2, Figure S17: Specimen Graphic, Table S1: Cartesian coordinate of 1 optimized: Standard orientation, Table S2: Cartesian coordinate of 2 optimized: Standard orientation.