Micromonohalimanes A and B: Antibacterial Halimane-Type Diterpenoids from a Marine Micromonospora Species

Abstract

Despite the fact that actinomycetes harbor the genetic potential to produce terpenes, terpenoid natural products tend to be a rare occurrence in fermentation broths. Here we report two new halimane-type diterpenoids, micromonohalimanes A (1) and B (2), that were isolated from a Micromonospora sp. cultivated from the marine ascidian Symplegma brakenhielmi. This is the first report of the halimane-type diterpenoids from Micromonospora. The structures were determined using spectroscopic methods including X-ray crystallography to establish the absolute configuration. Micromonohalimane B demonstrated moderate antibacterial activity against methicillin-resistant Staphylococcus aureus.

TOC Image

Traditional methods used for bacterial natural products drug discovery have become time- and labor-intensive, in part, due to high rediscovery rates. In order to overcome this historical weakness that has led to high rates of rediscovery, our group has used LCMS-based metabolomics.^1–4

We and others have demonstrated that LCMS-based metabolomics provides a high-throughput platform to prioritize bacterial strains and drive discovery of novel natural products. Using LCMS–principal component analysis (PCA), we have identified strains with unique metabolite profiles, and in our experience those metabolic outliers tend to yield novel natural products. One of the strengths of PCA versus other statistical methods is that common background ions have little effect on the analysis. In other words, background subtraction, which can be challenging, is not necessary. Therefore, PCA provides an analysis option that is robust to small variations within the data sets and does not rely on fragmentation. The strain identified in this report had a unique metabolic profile as determined by PCA, but the molecules presented here were not the drivers of the statistical variance within the analyses. Overall, this is not surprising since many metabolites detected by mass spectrometry are below isolation thresholds. In our experience, unique molecular signatures often identify a bacterium as having potential for production of novel natural products.

Strain WMMC-218 was identified as having unique metabolite markers by PCA in an analysis of 29 actinomycetes (Supporting Information, Figure S1). Our current discovery workflow involves identification of unique bacteria by LCMS–PCA followed by extract production from three distinct media and subsequent screening for biological activity. WMMC-218 was prioritized based on broad-spectrum antibiotic activity. Although the most potent compound had antibacterial activity against methicillin-resistant Staphylococcus aureus (MRSA), we also discovered two new halimane-type diterpenoids. Unfortunately, the compound responsible for the potent antibacterial activity was not stable and decomposed.

WMMC-218 was cultivated from the marine ascidian Symplegma brakenhielmi (Shirley Parker-Nance, 2013). Micromonospora spp. are a well-known group of Gram-positive, spore-forming microbes, and many classes of compounds are known to be produced by Micromonospora.⁵ Several Micromonospora spp. have previously been cultivated from ascidians.^6–8 However, this is the first report of halimane-type diterpenoids from Micromonospora. The most closely related diterpenoid was from a marine invertebrate gorgonian, Echinomuricea sp.⁹

Micromonospora sp. WMMC-218 was prioritized due to unique m/z signatures identified via LCMS-based metabolomics in conjunction with bioactivity (PCA in Supporting Information, Figure S1). Following fractionation and screening, some of the active fractions appeared to contain terpenoid compounds. Because terpenes tend to be rather rare in actinomycete extracts and activity was observed in the terpene-containing fractions, the terpenes were targeted for isolation and structure determination.

Micromonohalimane A (1) was obtained as colorless crystals and gave a molecular formula of C₂₂H₃₄O₆ with six degrees of unsaturation, as determined by HR-ESIMS at m/z 417.2247 [M + Na]⁺. The IR spectrum showed absorptions typical of hydroxy (3395 cm⁻¹) and ester carbonyl (1787 and 1708 cm⁻¹) groups.

The ¹H NMR spectrum of micromonohalimane A (1) (Table 1) revealed five methyls [δ_H 0.66 (3H, s), 0.83 (3H, d, J = 6.7 Hz), 0.95 (3H, s), 1.05 (3H, s), 2.13 (3H, s)], one olefinic proton [δ_H 5.40 (1H, m)], and three (of five) resolved methine protons [(δ_H 4.46 (1H, s), 5.31 (1H, d, J = 10.3 Hz), 2.04 (1H, m)], along with a resolved diastereotopic methylene [(δ_H 4.27 (1H, d, J = 10.1 Hz), 4.13 (1H, d, J = 10.1 Hz)]. The ¹³C NMR and HSQC spectra of compound 1 (Table 1) revealed the presence of two ester carbonyls (δ_C 175.3 and 170.9), two olefinic carbons (δ_C 145.5 and 115.8), three sp³ quaternary carbons (δ_C 36.0, 37.2, and 78.5), four sp³ methines (δ_C 33.4, 40.4, 71.8, and 71.8), six sp³ methylenes (δ_C 22.3, 27.5, 31.6, 34.6, 40.8, and 71.8), and five methyls (δ_C 15.5, 16.4, 21.2, 28.8, and 29.6) in the structure. The NMR spectroscopic features (Table 1) were characteristic of an acetoxy ester containing a halimane-type diterpene containing an acetoxy ester.¹⁰ COSY and HSQC analyses revealed three isolated spin systems: (a) C(10)H–C(1)H₂–C(2)H₂–C(3)H₂, (b) C(6)H–C(7)H₂–C(8)H–C(17)H₃, and (c) C(11)H₂–C(12)H. The HMBC correlations from H₃-18/H₃-19 to C-3/C-4/C-5, H₃-18 to C-19, H-6 to C-4/C-5/C-10/C-7/C-8, H₃-20 to C-8/C-9/C-10/C-11, and H₃-17 to C-7/C-8/C-9, together with the spin systems (a) and (b) deduced above, led to the assignment of rings A and B.

Table 1.

¹H NMR (600 MHz) and ¹³C NMR (125 MHz) Spectral Data of Compounds 1 and 2 in CDCl₃

	1		2
position	δC, type	δH (J in Hz)	δC, type	δH (J in Hz)
1	27.5, CH2	1.85, m 1.03, m	27.5, CH2	1.74, m 0.99, m
2	22.3, CH2	1.62, m 1.45, m	21.9, CH2	1.56, m 1.18, m
3	40.8, CH2	1.40, m 1.21, m	40.7, CH2	1.41, d (12.0) 1.19, m
4	36.0, C		36.1, C
5	145.5, C		146.2, C
6	115.8, CH	5.40, m	115.5, CH	5.45, d (2.2)
7	31.6, CH2	1.93, m 1.77, m	31.5, CH2	1.91, m 1.81, m
8	33.4, CH	1.58, m	33.4, CH	1.64, m
9	37.2, C		37.2, C
10	40.4, CH	2.04, m	40.8, CH	2.35, d (13.0)
11	34.6, CH2	2.08, m 1.28, m	35.1, CH2	2.06, dd (16.1, 1.1) 1.72, dd (16.1, 10.3)
12	71.8, CH	5.31, d (10.3)	79.3, CH	4.63, dd (10.3, 1.1)
13	78.5, C		77.6, C
14	71.8, CH	4.46, s	73.6, CH	5.60, s
15	175.3, C		168.6, C
16	71.8, CH2	4.27, d (10.1) 4.13, d (10.1)	46.6, CH2	3.75, d (11.9) 3.67, d (11.9)
17	15.5, CH3	0.83, d (6.7)	15.2, CH3	0.83, d (6.7)
18	28.8, CH3	0.95, s	29.1, CH3	1.03, s
19	29.6, CH3	1.05, s	29.7, CH3	1.07, s
20	16.4, CH3	0.66, s	16.3, CH3	0.70, s
21	170.9, C		170.0, C
22	21.2, CH3	2.13, s	20.5, CH3	2.24, s
–OH				3.20, s

The HMBC correlations from H-12 to C-9/C-11/C-14/C-16, from H-16 to C-14/C-15, and from H-11 to C-13, as well as the spin system (c) C(11)H₂–C(12)H, allowed for the assignment of ring C. Additionally, the HMBC correlations from H-12 to C-21 suggested that the ester carbonyl group was connected to C-12.

The relative and absolute configurations of compound 1 were determined by X-ray crystallographic analysis (Figure 1). The X-ray analysis allowed for assignment of the absolute configurations of the stereogenic carbons to be 8S, 9R, 10S, 12R, 13R, and 14R. Therefore, compound 1 was named (8S,9R,10S,12R,13R,14R)-micromonohalimane A.

Figure 1.

Molecular drawing of the X-ray crystal structure of 1.

Compound 2 had the molecular formula C₂₂H₃₃ClO₅, as established by HR-ESIMS, with six degrees of unsaturation. A comparison of NMR (Table 1) and HR-ESIMS data with compound 1 suggested that the structures were similar in rings A and B, but differed in ring C. The HMBC correlations from H-12 to C-9/C-16/C-14/C-15/C-13, from H-16 to C-14/C-12/C-13, from H-11 to C-13, from H-14 to C-15, and from –OH to C-12/C13/C-14/C-16, as well as the spin system (c) C(11)H₂–C(12)H, allowed for the assignment of ring C, which connected with ring B through C-11. Additionally, the HMBC correlations from H-14 to C-21 suggested that the ester carbonyl group was connected to C-14 of ring C. The placement of the Cl was determined largely based on the downfield shift of the diastereotopic methylene protons (H-16, δ 3.67, 3.75), in conjunction with carbon resonance for C-16 at δ_C 46.6.

The relative configuration of compound 2 was investigated by ROESY, 1D-DPFGSE-NOE, and DFT studies. Correlations of H-10 with H₃-18/H-8/H-11a, the absence of ROESY correlation between H-10 and H₃-20, and the same chemical shifts of rings A and B as compound 1 confirmed the relative configurations of rings A and B were the same as for compound 1. Although a ROESY correlation was observed between H-12 and H-16, molecular modeling indicated that these protons would be sufficiently close for NOE regardless of configuration, in part due to the flat nature of the lactone. From a biosynthesis perspective, the configuration of compound 2 around the lactone ring should be identical to that found in 1 since compound 2 appears to simply be a regioisomer in terms of the lactone. To help confirm the biosynthetic hypothesis, we used molecular modeling and DFT shielding tensor calculations to compare carbon chemical shifts around the lactone ring for all eight of the possible diastereomers. Spartan’14¹¹ was used to find an equilibrium conformer using molecular mechanics, and Gaussian 09¹² was used for further geometry optimization and NMR calculations (B3LYP/6-31G(d,p)).¹³ NMR shifts were referenced to tetramethylsilane (TMS) and benzene using the multistandard approach.¹⁴ After referencing, the calculated chemical shifts were compared to those experimentally observed for all atoms in the lactone ring. The structure shown fit best with the experimental data. Furthermore, the best fit from the DFT studies possessed stereochemical configurations identical to those found in compound 1, as would be predicted. We also evaluated the conformational flexibility with 2 to ensure that only one conformation would be observed for the lactone ring. To that end, we examined 10 000 conformers via molecular mechanics and selected the 16 that were below 40 kJ/mol followed by subsequent local geometry optimization and DFT energy calculations to determine a Boltzmann distribution. Although molecular mechanics provides easy access to conformers, molecular mechanics does not provide accurate energy calculations relative to DFT. Hence, our final Boltzmann distribution was based on energies determined by DFT calculations. Only two relevant conformations were found. Both closely resembled the equilibrium conformer used for the DFT NMR calculations (Supporting Information, Figure S15). Therefore, compound 2 was named (8S*,9R*,10S*,12R*,13R*,14R*)-micromonohalimane B.

Micromonohalimanes A (1) and B (2) were tested for antibacterial activity against methicillin-resistant Staphylococcus aureus, and the minimum inhibitory concentration (MIC) was determined for each compound. Micromonohalimane A (1) was determined to have an MIC of greater than 200 μg/mL against MRSA. Micromonohalimane B (2) was determined to have an MIC of 40 μg/mL against MRSA. In order to identify if micromonohalimane B (2) was bactericidal or bacteriostatic, a sterile swab was dipped into each well that showed inhibition of bacterial growth and was inoculated on an LB plate. A bactericidal agent would show no growth on the LB plate, while a bacteriostatic agent would show bacterial growth.¹⁵ Bacterial growth of each sample on the LB plate was seen, which suggests that micromonohalimane B (2) is a bacteriostatic agent.

In conclusion, two diterpenes were isolated from a Micromonospora sp. Although the isolation of terpenes from actinomyctes has been rare, genomics suggests that most actinomycetes have the potential to biosynthesize terpenes. Nonetheless, this is the first report of a halimine-type diterpene from an actinomycete.

EXPERIMENTAL SECTION

General Experimental Procedures

Melting points were measured on a MEL-TEMP micromelting point apparatus and were uncorrected. Optical rotations were measured on a PerkinElmer 241 polarimeter. UV spectra were recorded using an Aminco/OLIS UV–vis spectrophotometer. IR spectra were measured with a Bruker Equinox 55/S FT-IR spectrophotometer. NMR spectra were obtained in CDCl₃ with a Bruker Avance 600 MHz spectrometer equipped with a ¹H{¹³C/¹⁵N} cryoprobe and a Bruker Avance 500 MHz spectrometer equipped with a ¹³C/¹⁵N{¹H} cryoprobe. HRMS data were acquired with a Bruker MaXis 4G QTOF mass spectrometer. Normal-phase flash chromatography was performed using a Combi Flash Rf 200 HPLC system and a 4 g gold silica gel column (Teledyne Isco).

Biological Material

Strain WMMC-218 was isolated from the ascidian Symplegma brakenhielmi collected on August 7, 2013, in Florida at Stanblum State Park (27.479326, −80.311687). A voucher specimen for the ascidian was identified by one of the authors (S.P.-N.).

WMMC-218 displayed over 99% identity to Micromonospora sp. HK160111. The 16S sequence for WMMC-218 was deposited in GenBank (accession number KU183007).

Fermentation, Extraction, and Isolation

Strain WMMC-218 was fermented in 25 × 150 mm culture tubes (4 × 10 mL) in medium ASW-A (20 g soluble starch, 10 g glucose, 5 g peptone, 5 g yeast extract, 5 g CaCO₃ per liter of artificial seawater) for 1 week at 28 °C. Baffled flasks (250 mL, 12 × 50 mL) containing ASW-A were inoculated with 2 mL from the culture tube and shaken at 200 rpm at 28 °C for 7 days. Flasks (4 L, 8 × 500 mL) containing medium ASW-A with Diaion HP20 (4% by weight) were inoculated with 25 mL and shaken at 200 rpm at 28 °C for 7 days. Filtered HP20 was washed with H₂O and extracted with acetone. The acetone extract (11 g) was subjected to a liquid–liquid partitioning using 10% aqueous MeOH and hexane (1:1), which then increased to 30% aqueous methanol, and partitioned using CHCl₃ (1:1). The CHCl₃-soluble partition (3.1 g) was subjected to ENV⁺ column chromatography (ISOLUTE, 3 × 500 mg) with H₂O and MeOH (10%, 50%, 75%, 100%). Fraction 100% (0.88 g) was chromatographed through a Sephadex LH-20 column with MeOH/H₂O (7:3) and was further purified by 4 g HP silica column chromatography (Teledyne Isco) with a linear gradient solvent system consisting of solvent A (hexane) and solvent B (2-propanol) at room temperature as follows: 0–5 min, 0% B; 5–25 min, 0% to 30% B; 25–30 min, 30% B, yielding 2 (3.0 mg, t_R 9.1 min) and 1 (1.6 mg, t_R 10.5 min), respectively. We cultured strain WMMC-218 in three media (ASW-A, RAM₂, GOT), analyzed the extract of each media, and then selected the media ASW-A.

(8S,9S,10S,12R,13R,14R)-Micromonohalimane A (1)

Colorless needles [hexane/2-propanol/MeOH, (2:1:4)]; mp 163–164 °C; [α]²⁰_D +18 (c 0.03, MeOH); UV (MeOH) λ_max (log ε) 210 (3.59) nm; IR (ATR) ν_max 3395, 3314, 2927, 2854, 1787, 1708, 1462, 1373 cm⁻¹; ¹H NMR (600 MHz, CDCl₃) and ¹³C NMR (125 MHz, CDCl₃) data, Table 1; HR-ESIMS m/z 417.2246 [M + Na]⁺ (calcd for C₂₂H₃₄NaO₆, 417.2248).

X-ray Crystallographic Analysis of Compound 1.¹⁶

Crystallization from hexane/2-propanol/MeOH (2:1:4) using the vapor diffusion method yielded colorless crystals of compound 1. A colorless crystal with approximate dimensions 0.40 × 0.15 × 0.15 mm³ was selected under oil under ambient conditions and attached to the tip of a MiTeGen MicroMount. The crystal was mounted in a stream of cold nitrogen at 100(1) K and centered in the X-ray beam by using a video camera. The crystal evaluation and data collection were performed on a Bruker SMART APEXII diffractometer with Cu Kα (λ = 1.541 78 Å) radiation and a diffractometer to crystal distance of 4.03 cm. Crystal data for C₂₂H₃₆O₇ (M = 412.51 g/mol): monoclinic, space group P2₁ (no. 4), a = 9.2381(5) Å, b = 6.9763(6) Å, c = 16.8910(8) Å, β = 90.987(3)°, V = 1088.43(13) ų, Z = 2, T = 100.01 K, μ(Cu Kα) = 0.759 mm⁻¹, D_calc = 1.259 g/cm³, 23 887 reflections measured (5.232° ≤ 2θ ≤ 146.796°), 4036 unique (R_int = 0.0195, R_sigma = 0.0121), which were used in all calculations. The final R₁ was 0.0297 (I > 2σ(I)) and wR₂ was 0.0790 (all data).

(8S,9S,10S,12S,13R,14S)-Micromonohalimane B (2)

Colorless oil; [α]²⁰_D +110 (c 0.01, MeOH); UV (MeOH) λ_max (log ε) 210 (4.09) nm; IR (ATR) ν_max 3390, 3312, 2927, 2854, 1780, 1700, 1460, 1372 cm⁻¹; ¹H NMR (600 MHz, CDCl₃) and ¹³C NMR (125 MHz, CDCl₃) data, Table 1; HR-ESIMS m/z 435.1906 [M + Na]⁺ (calcd for C₂₂H₃₃ClNaO₅, 435.1909).

Molecular Modeling Calculations

Molecular modeling calculations were performed on a Dell Precision T5500 Linux workstation with a Xeon processor (3.3 GHz, 6-core). Low-energy conformers were obtained using Spartan’14 software (MMFF, 10000 conformers examined).¹¹ The low-energy conformer for each compound was analyzed using Gaussian 09 for geometry optimization and NMR calculations (B3LYP/6-31G(d,p)). NMR shifts were referenced to TMS and benzene using the multistandard approach.¹⁷ Molecules were modeled in the gas phase.

Antibacterial Assay

Micromonohalimanes A (1) and B (2) were tested for antibacterial activity against MRSA (ATCC #33591), and MICs were determined using a dilution antimicrobial susceptibility test for aerobic bacteria.¹⁸ Micromonohalimanes A (1) and B (2) were dissolved in DMSO, serially diluted to 10 concentrations (0.125–64 μg/mL), and tested in a 96-well plate. Vancomycin was used as a control and exhibited an MIC of 1 μg/mL against MRSA. Micromonohalimanes A (1) and B (2) and vancomycin were tested in triplicate. Six untreated media controls were included on each plate. The plates were incubated at 33 °C for 18 h. The MIC was determined as the lowest concentration that inhibited visible growth of bacteria.