Isolation and Characterization of Anti-Mycobacterial Natural Products from a Petrosia sp. Marine Sponge

Tuberculosis (TB) is a dreadful infectious disease and a leading cause of mortality and morbidity worldwide, second in 2020 only to severe acute respiratory syndrome 2 (SARS-Cov-2). With limited therapeutic options available and a rise in multidrug-resistant tuberculosis cases, it is critical to develop antibiotic drugs that display novel mechanisms of action. Bioactivity-guided fractionation employing an Alamar blue assay for Mycobacterium tuberculosis strain H37Rv led to the isolation of duryne (13) from a marine sponge Petrosia sp. sampled in the Solomon Islands. Additionally, five new strongylophorine meroditerpene analogues (1–5) along with six known strongylophorines (6–12) were isolated from the bioactive fraction and characterized using MS and NMR spectroscopy, although only 13 exhibited antitubercular activity.

T uberculosis (TB), a contagious bacterial disease caused by Mycobacterium tuberculosis (Mtb), resulted in 1.3 million mortalities in 2020 alone. 1 More than a billion people have succumbed to TB over the past 200 years, a death toll that exceeds the combined deaths from malaria, plague, influenza, HIV/AIDS, cholera, and smallpox. 2 While Bacillus Calmette-Gueŕin (BCG) is the only licensed vaccine against TB and has limited efficacy with vastly variable results (0−80% effectiveness), treatment for the active state of TB consists of an extended (6-month) regimen of first-line drugs: isoniazid, rifampicin, ethambutol, and pyrazinamide (successful in 85% of cases, data for 2019). 1 On the other hand treatment of multidrug-resistant TB (MDR-TB) requires a second line of drugs that are comparatively less effective, have lengthy treatment regimens (up to 24 months), are more expensive, and have severe side-effects. 3 Additionally, treatment success using second-line drugs decreases substantially for patients who are older, are further in their disease progression, are also infected with HIV, or have undergone previous TB treatment. These as well as other factors explain why MDR-TB has a low treatment success rate, posing serious risks for individual and public health. 4 Terrestrial and marine organisms harbor a wide variety of bioactive molecules that have been the source of numerous FDA-approved drugs. 5,6 Almost 56% of all antibacterial medications approved between 1981 and 2019 were either natural products or their derivatives. In fact, major anti-TB drugs including rifampin (belonging to the ansamycin class of polyketides) and the aminoglycosides were derived from soil bacteria Amycolatopsis rifamycinica and Streptomyces griseus, respectively. 7−9 Several natural products have shown promising anti-TB activity and, with their mechanisms of action already elucidated, hold great potential for further development as viable drug candidates. 10−12 Although substantial research has been devoted to development of new anti-TB agents, it is concerning that over the last three decades only two new drugs (approval of bedaquiline in 2012 by the U.S. Food and Drug Administration and delamanid in 2014 by the European Union) have been added to combat MDR-TB. Considering the promising anti-TB activity and mechanisms of action that have been revealed for several natural products, a greater interest and investment in natural products-based lead optimization may help address the health crisis caused by MDR-TB. In the current study, we explored natural products from a Solomon Islands collection of the marine sponge Petrosia sp., 13−16 leading to identification of new and known natural products with promising anti-TB profiles.

■ RESULTS AND DISCUSSION
Bioassay of several hundred marine macro-organism extracts sampled in Fiji and the Solomon Islands under the National Institutes of Health (NIH) funded International Cooperative Biodiversity Groups (ICBG) program showed that the hexanes-and CH 2 Cl 2 -soluble fractions generated from extracts of a Solomon Island marine sponge Petrosia sp. were bioactive against Mtb strain H37Rv. The bioactive hexanes-and CH 2 Cl 2soluble fractions were subsequently chromatographed separately on a silica gel normal phase column. Followed by an anti-TB bioassay, two chromatographic fractions were prioritized and subjected to normal phase, followed with reversed phase, HPLC separation to furnish five new strongylophorine analogues (1−5), known strongylophorines (6−11), and a known linear acetylene, 13.  (2) were isolated as colorless solids and showed 1 H and 13 C NMR spectroscopic features that resembled strongylophorine meroditerpene compounds previously reported from Petrosia corticata and Strongylophora strongilata (Table 1). 17 These molecular fragments could then be integrated into a pentacyclic meroditerpenoid structure (resembling the strongylophorines) with key HMBC correlations as depicted in Figure 1. 17 Figure 1). 20-O-Methyl strongylophorine-15 (3) and 20-O-methyl strongylophorine-16 (4) were isolated together as a mixture ( Figure 2). While 3 and 4 were clearly separable in the HPLC time frame, upon drying, a 1:1 mixture was generated due to the spontaneous epimerization at C-26, as observed with 1 H NMR spectroscopic data ( Figure S15). Hence, it is likely that 1 and 2 are artifacts of the isolation process, generated from 3 and 4, where during the rapid epimerization at C-26, the incipient oxenium ion gets captured by traces of EtOH generated from the use of EtOAc as a solvent for HPLC purification. HRMS data (m/z 427.28413 [M + H] + accounting for a molecular formula of C 27 H 38 O 4 ) together with 1 H NMR spectroscopic data (sharp singlet at δ H 3.74 ppm with an integration of 3 H atoms) clearly indicated that 3 and 4 were 20-O-methylated analogues of strongylophorine-15 (11) and strongylophorine-16 (12), respectively. 19 The planar structure and the relative configurations at C-26 for 3 and 4 were deduced with a combination of COSY, HMBC, and 1D ROESY NMR spectroscopic data ( Figure 2, Tables S1 and S2).   Figure 2). Two distinctive aldehyde protons appeared as sharp singlets at 9.76 and 9.79 ppm for distrongylophorine A (5) and were suggestive of a new dialdehyde analogue of the strong-ylophorine dimer, distrongylophorine (6). 14 Further interpretation with HMBC NMR spectroscopic data clearly suggested that the two aldehyde functionalities were positioned at C-26 and C-26′ ( Figure 3). The relative configurations of the aldehyde functionalities were revealed with 1D ROESY     Table S3. Finally, the identities of known natural products 6−13 were confirmed by comparing their HRMS and 1 H and 13 C data with existing literature. [13][14][15]19 Assessment of pure 1, 3, 4, 7−10, and 13 in an Alamar blue assay (MABA) showed that 13 was potently active against the virulent strain (H37Rv) of Mtb with an MIC of 1.4 μM, comparable to the tuberculosis drug linezolid (1.0 μM, Table  2). 20 The strongylophorines did not show significant inhibition at a test concentration of 25 μg/mL. As MABA represents the rapidly growing state of Mtb and does not recapitulate the nonreplicating persistence physiological state of the bacteria, a low-oxygen recovery assay (LORA) was also employed. 21−23 The nonreplicating persistence physiological state of the bacteria is a critical issue for all tuberculosis medications, requiring treatments that last several months, and hence represents a crucial phenotype to be targeted in the development of new drugs. Interestingly, duryne (13) retained its promising bioactivity in a LORA assay as well and had a MIC of 1.0 μM (comparable with the MIC for linezolid and pretomanid, Table 2). However, 13 was not active against M. abscessus and M. avium (data not shown). Its IC 50 of 0.4 μM against cultured Vero cells was consistent with previous reports of cytotoxicity in P388 murine leukemia, colon (HCT-8), lung (A549), mammary (MCF7), and HeLa cervical cells (IC 50 ∼0.2 μM). 15,16 The similar potency of 13 in MABA and LORA assays (Table 2), combined with toxicity toward Vero cells, is consistent with the possibility that 13 could be targeting the electron transport chain or a conserved cell membrane component, rather than acting on cell wall synthesis. Activity of 13 against M. tuberculosis but not M. abscessus or M. avium is not predictive of mechanism of action because such MIC differences can be due to any number of factors including cell differences in penetration, efflux, metabolism, or activation of the target itself. It is worth noting that structure−activity relationship studies on molecules similar to duryne have shown that the lipid chain length, stereochemistry of the propargyl alcohol, and the presence of one or multiple acetylenic groups affect their cytotoxicity. 24,25 Structural analogues of 13 with additional hydroxyl or carboxylic acid moieties (e.g., pellynols, neopetroformynes, petroformynic acids) have been shown to retain their cytotoxicity, generating hope that issues can be addressed related to druggability and solubility of these hydrophobic molecules. 26 In conclusion, considering the low success rate on finding new molecules that inhibit Mtb, 13 and its analogues could be explored further to determine whether the antitubercular and cytotoxic effects of 13 can be decoupled through medicinal chemistry efforts. 27,28 ■ EXPERIMENTAL SECTION General Experimental Procedures. All NMR spectroscopic data (1D and 2D) were acquired on an 18.8 T (800 MHz for 1 H and 201 MHz for 13 C) Bruker Avance IIIHD instrument equipped with a 3 mm triple resonance cryoprobe or a 16.4 T (700 MHz for 1 H and 175 MHz for 13 C) Bruker Avance IIIHD instrument with a 5 mm indirect broadband cryoprobe. CDCl 3 or pyridine-d 5 was used as the NMR solvent for all samples and referenced to residual solvent peaks δ H 7.26, δ C 77.2 for CDCl 3 and δ H 7.22, δ C 123.9 for pyridine-d 5 . NMR spectroscopic data were processed and analyzed using MestReNova 11.0.4.
High-resolution mass spectrometric data were acquired in a Thermo Scientific Orbitrap ID-X instrument. Both normal and reversed phase HPLC chromatographic separations were performed using a Waters 1525 binary pump coupled with a Waters 2996 photodiode array detector and an Altech 800 evaporative light scattering detector.
Cytotoxicity Assay. Green monkey kidney cells (Vero cell CCL-81) were used to assess the cytotoxicity. Cultured Vero cells in Eagle's minimum essential medium (MEM) containing 10% fetal bovine serum (FBS) supplemented with penicillin and streptomycin antibiotic mix were used. The cells were microscopically quantified into a density of 2 × 10 5 cells/mL. A 100 μL portion of cell suspension was added to the test compounds and incubated for 3 days at 5% CO 2 and 37°C. At the end of the incubation period 20 μL of 0.6 mM resazurin was added, and the fluorescence readout was measured after 4 h of incubation to calculate IC 50 . 33 ■ ASSOCIATED CONTENT
Additional specimen collection and natural product characterization data (PDF)