6-Deoxy- and 11-Hydroxytolypodiols: Meroterpenoids from the Cyanobacterium HT-58–2

Abstract

Chemical investigation of cyanobacterial strain HT-58–2, which most closely aligns with the genus Brasilomena, has led to the isolation of two compounds related to tolypodiol. The structures and absolute configuration of 6-deoxytolypodiol (1) and 11-hydroxytolypodiol (2) were elucidated by spectroscopic and spectrometric analysis. While tolypodiol previously showed anti-inflammatory activity in a mouse ear edema assay, only 2 reduced in vitro thromboxane B₂ (TXB₂) and superoxide anion (O₂⁻) generation from Escherichia coli lipopolysaccharide (LPS)-activated rat neonatal microglia to any appreciable degree.

Graphical Abstract

Cyanobacteria are well-known producers of secondary metabolites.¹ The vast majority of these compounds are produced via a combination of polyketide synthase and non-ribosomal peptide synthetases.¹ Deeper investigations into the biosynthesis of many of the peptidic compounds from cyanobacteria revealed a large number of those compounds were actually ribosomally synthesized and post-translationally modified peptides (RiPPs).¹ Taken together, these trends highlight the unusual nature of cultured cyanobacterial strain HT-58–2, originally classified as Tolypothrix nodosa based on morphology, but subsequent 16S rRNA analysis suggested it more closely resembles the genus Brasilonema.² While the genome contains putative clusters for several typical cyanobacterial metabolites, the two main classes of secondary metabolites expressed under a variety of culture conditions belong to neither of these structural classes. Instead, the first set of compounds, the tolyporphins, are a unique series of tetrapyrroles remarkable for their structure² and biological provenance that display potent activity as photodynamic therapy agents,³ while the second class of compounds is exemplified by the meroterpenoid tolypodiols.⁴ To date, only one naturally-occurring and one synthetic derivative of this latter class of anti-inflammatory compounds have been disclosed. Reported here are results of further investigations of strain HT-58–2, which was driven primarily by our interest in exploring the tolyporphins’ potential as photodynamic therapy agents given they are more potent in vivo than the FDA-approved drug photofrin.^{5, 6} These investigation have led to the isolation and structure elucidation of two tolypodiol analogs, 6-deoxytolypodiol (1) and 11-hydroxytolypodiol (2) from the tolyporphin-containing fraction, establishment of the absolute configurations of all members of the series (1–3) and biological evaluation of 1–3 in a model of pro-inflammatory mediator inhibition with relevance to neurological disorders such as Alzheimer’s Disease.

From the extracts of HT-58–2 cultures, two new analogs of tolypodiol were discovered. Briefly, lyophilized biomass from HT-58–2 (43.35 g) was exhaustively extracted in 1:1 CH₂Cl₂:i-PrOH to produce an organic extract (2.33 g) after removal of solvent in vacuo. The residue was partitioned using a modified Kupchan partition,⁷ and the resulting CH₂Cl₂ extract separated using RP-flash chromatography and HPLC to yield 1 (0.5 mg), 2 (0.6 mg) and tolypodiols (3;8.6 mg). The tolypodiols were minor metabolites in the extract. with tolyporphins being the most abundant metabolites.

The planar structure of 1 was determined based on conventional spectroscopic and spectrometric approaches such as LC-MS, 1D, and 2D NMR-based experiments. The HRESIMS of 1 produced a protonated molecule at m/z 441.3000 [M+H]⁺ and established a neutral molecular formula of C₂₈H₄₀O₄. The similarity of 1 to tolypodiol, also isolated from this strain, was deduced based on resonances found in the ¹H NMR spectrum consistent with five methyl signals, a methoxy group, and a 1,2,4-substituted aromatic ring. A full spectroscopic analysis of 1 revealed the lack of a hydroxy group at C-6 (δ_C 18.2) relative to tolypodiol; a modification in agreement with the molecular formula differences which established the planar structure as depicted. Further evidence of this assessment was observed in a change of multiplicity of H-5 in tolypodiol from a doublet (δ_H 0.76, J = 2.2 Hz) to a doublet of doublets in 1 (δ_H 0.78, J = 12.1, 2.6 Hz).

The relative configuration of the molecule was determined through analyses of 2D-ROESY correlations (Figure 1). The methine proton at C-1 was assigned an axial orientation in the ring system on the basis of an 11 Hz coupling to H-2, indicating the corresponding hydroxy group was in an equatorial orientation. Overall, the relative configuration for 1 is consistent with that reported for tolypodiol.

Figure 1.

Key correlations (double-headed arrows) used to determine relative configuration of 1.

The planar structure of 11-hydroxytolypodiol 2 was deduced in a similar manner. The HRESIMS of 2 produced a protonated molecule at m/z 473.2895 [M+H]⁺ and established a molecular formula of C₂₈H₄₀O₆. The presence of an additional hydroxy group relative to tolypodiol was evident from the signal at δ_H 4.87 in the ¹H NMR spectrum and was consistent with the proposed molecular formula. Analysis of the full 1D and 2D-NMR data revealed this hydroxy group was in an axial orientation at C-11 based on the three small 4.8, 2.7, and 2.1 Hz couplings of proton H-11 to protons at H-9 and H₂-12. Through a similar analysis of J_H-H coupling, the relative orientations of the carbinol protons at C-1 and C-6 were found to be the same as that observed in tolypodiol.

The absolute configurations of 1–3 were determined through ECD spectroscopy. Conformational analysis of 1R,5S,8R,9S,13S,14S-1, using Monte Carlo multiple minimization (MCMM) with the OPLS-2005 force field in MacroModel identified 10 conformers within 5 kcal/mol of the lowest energy conformer. A comparison with the experimental NMR data revealed no obvious contradictions in terms of expected coupling constants thus validating the conformational analysis. These optimized structure and the vibrational frequencies of these conformers were subsequently calculated in Gaussian09⁸ at the B3LYP⁹/DEF2SVP¹⁰ level. Time-dependent density functional theory (TDDFT)¹¹ calculations were performed using Gaussian09⁸ with six different functional and basis sets (Figure S15) to predict ECD spectra for the conformers of the 1R,5S,8R,9S,13S,14S enantiomer. The Boltzmann weighted ECD spectrum of these conformers, calculated using SpecDis,¹² was compared to the experimental data for 1 at all levels. These data were in good agreement with Δ_ESI values ranging from 0.875–0.924¹² and similarity factors of 0.908–0.926 and 0.0324–0.007 for the 1R,5S,8R,9S,13S,14S-configuration and its enantiomer, respectively (Figures 2A and S15). The ECD spectra of 1–3 (Figure 2B) displayed similar Cotton effects suggesting the same absolute configuration. Slightly different Δε values were obtained despite our efforts to normalize the concentrations of 1–2 by UV (260 nm) in comparison to the more abundant 3.

Figure 2.

Panel A) Comparison of BH&HLYP¹⁰/DEF2TZVPP¹³ calculated (red) and experimental (black) ECD curves for 1. Panel B) Experimental ECD curves for 6-deoxytolypodiol (1; black), 11-hydroxytolypodiol (2; purple) and tolypodiol (3; green).

There are relatively few cyanobacterial natural products that primarily use isoprene components as building blocks.¹ Noscomin¹⁴ and the comnostins¹⁵ isolated from the media extract of Nostoc commune are two other examples. The closest structural relatives to the tolypodiols are taondiol and its derivatives, isolated from marine brown algae, but their carbon skeletons differ with regards to substituents on the aromatic ring.¹⁶

At the time of its original isolation, tolypodiol (3) was shown to reduce chemically-induced inflammation in an in vivo mouse ear edema assay⁴ (ED₅₀ 30 μg/ear for 3 vs 20 μg/ear for the positive control). As the yields of 1–2 precluded testing in the original assay system, 1–3 were assessed for their effects on in vitro activation and generation of proinflammatory mediators by rat neonatal microglia (Figure 3) using the standard inflammation models in our labs.^17,18 Only 2 significantly inhibited the generation of proinflammatory thromboxane B2 (TXB₂) (apparent IC₅₀= 0.1 μM). Lack of inhibition of superoxide anion (O₂⁻) generation suggests that 2 inhibits TXB₂ generation in rat microglia through a mechanism that may be cyclooxygenase dependent. Compound 2 is more potent in this assay system than acetylsalicylic acid (aspirin) (IC₅₀ 3.12–10 μM), a U.S. Food and Drug Administration-approved nonsteroidal anti-inflammatory drug used to treat pain, fever and inflammation. Further pharmacological and toxicological investigation of 2 in both in vitro and in vivo models of neuroinflammation thus appears to be warranted given modulation of TXB₂ and O₂⁻ has been proposed as a therapeutic approach for the treatment of several neurodegenerative diseases where microglia activation has been implicated.¹⁹

Figure 3.

Panel A) tolypodiol (3), B) 11-hydroxytolypodiol (2) and C) 6-dexoytolypodiol (1). PMA-stimulated O₂⁻ and TXB₂ generation for tolypodiol, 2 and 1 were determined as described.¹⁹ Data are expressed as percentage of untreated control O₂⁻ and TXB₂ release triggered by PMA (1 μM) for 70 min. The data shown are the mean ± SD of 1 independent experiment with two replicates.

In summary, two new tolypodiols analogs have been isolated and characterized from the original strain doubling the number of known members of this series. Compounds 1 and 2 differ from the previous reported analog, tolypodiol acetate, as they differ in the oxidation state at C-6 and C-11, which suggests these modifications may occur late in the biosynthesis after the majority of the structure has been assembled. While we could not directly compare the anti-inflamatory activity of the new compounds with tolypodiol using the original assay system due to the limited quantities available, the ability of all three compounds to activate and generate proinflammatory mediators by rat neonatal microglia was determined, with only 2 showing appreciable biological activity at an apparent IC₅₀= 0.1 μM. Compound 2 showed a similar level of TXB2 inhibition as the clinically approved NSAID flurbiprofen (apparent IC₅₀ = 100 nM)²⁰ and was an order of magnitude more potent than aspirin.

EXPERIMENTAL SECTION

General Experimental Procedures.

Optical rotations were measured on a Jasco-DIP-700 polarimeter at the sodium line (589 nm). UV spectra were obtained on a Hewlett-Packard 8453 spectrophotometer, and IR spectra were measured as a thin film on a CaF₂ disc using a Perkin Elmer 1600 series FTIR. All 1D and 2D NMR spectra with the exception of ¹³C NMR spectra were acquired on a Varian Unity Inova 500 MHz spectrometer operating at 500 (¹H) or 125 (¹³C) MHz using the residual solvent signals (δ_H 7.26 and δ_C.77.0) as an internal reference. ¹³C NMR spectra were acquired on a Varian Unity 600 MHz spectrometer. NMR samples were analyzed in 3 mm Shigemi NMR tubes. High-resolution mass spectrometry (HRMS) data were obtained on an Agilent 6545 LC-MS Q-ToF with ESI ionization in the positive mode. Gradient HPLC separations used a Shimadzu system consisting of LC-20AT Solvent Delivery Modules, an SPDM20A VP Diode Photodiode Array Detector, and an SCL-20A VP System Controller.

Cultivation of Cyanobacteria.

Cultures were revived from cryostorage and grown in BG-11 media.²¹ Phyogenetic characterization and genomic sequencing of this strain has been previously reported.² For large-scale harvests, cultures were grown in 20 L Pyrex carboys and aerated at a flow rate of 5 L/min while under continuous illumination of fluorescent light banks.²² Cell material was harvested after 45 days of growth via decantation and filtration to be freeze-dried prior to extraction.

Extraction and Isolation of Meroterpenes.

The lyophilized biomass from HT-58–2 (43.35 g) was exhaustively extracted in 1:1 CH₂Cl₂:i-PrOH to produce an extract (2.33 g) after removal of solvent in vacuo. The residue was partitioned using a modified Kupchan extraction into four fractions (hexanes; 607.8 mg, CH₂Cl₂; 534.9 mg, BuOH; 28.5 mg and aq. MeOH; 100 mg). The CH₂Cl₂ extract (534.9 mg) was further fractionated over C8 silica gel using a step gradient of increasing amounts of MeOH in H₂O (25%; 86.1 mg, 50%; 38.2 mg, 75%; 52.6 mg, 100% MeOH; 275.6 mg). Tolypodiol (3) (8.6 mg, 0.09% yield, >99% purity by UV at 265 nm, t_R 18.0 min) and 11-hydroxytolypodiol (2) (0.6 mg, 0.007% yield, 97% purity by UV at 265 nm, t_R 14.0 min) were isolated from the 75% MeOH fraction, which contained tolyporphins,^{6, 23} using a linear gradient (2.7 mL/min) of CH₃CN in H₂O modified with 0.1% formic acid from 70–100% CH₃CN over 30 min and then 100% CH₃CN over 15 min on a Luna C₁₈ (10 μ, 250 × 10 mm) semi-preparative column. Another portion of the 75% MeOH extract of the CH₂Cl₂ partition, was purified by preparative RP-HPLC (Phenomenex Luna C18, 150 × 21.2 mm, 5 μm 100 Å; with a linear gradient of CH₃CN in H₂O with 0.1% formic acid from 75% to 100% CH₃CN over 25 min, and held at 100% CH₃CN for an additional 10 min; flowrate 15 mL/min) with fractions automatically collected every minute. Fractions 25 and 26 were combined from this purification and subjected to a linear gradient (3.5 mL/min) of EtOAc in hexanes from 20–95% over 30 min and then 100% EtOAc over 15 min on a Luna 10 μ Silica (2) (100 Å 250 × 10 mm) semi-preparative column which led to the isolation of 6-deoxytolypodiol (1) (t_R 8.1 min, 0.5 mg, 0.001% yield, >90% purity by ¹H NMR).

6-Deoxytolypodiol (1):

white, amorphous solid; ${\left[\alpha \right]}_D^{22}$ −480 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 291 (3.67), 264 (4.18) nm; ECD (1.1 × 10⁻⁴ M, MeOH), λ_max (Δε) 262 (−3.5), 222 (1.2) nm IR (CaF₂ disc) ν_max 3408, 2928, 1732, 1653, 1558, 1263 cm⁻¹; NMR data Table 1; HRESIMS m/z 441.3000 [M+H]⁺ (calcd for C₂₈H₄₁O₄⁺, 441.2999).

Table 1:

NMR Spectroscopic Data for 1 and 2 (¹H 500 MHz, ¹³C 125 MHz, CDCl₃)

		1		2
Position	δC, Type	δH (J in Hz)	δC, type	δH (J in Hz)
1	79.9, CH	3.39, dd (9.2, 6.4)	78.7, CH	3.46, dd (11.0, 5.0)
2a	30.0, CH2	1.62, m	29.0, CH2	1.79, m
2b		1.67, m		1.81, m
3a	39.7, CH2	(eq.) 1.37, m	41.4, CH2	1.35, m
3b		(ax.) 1.29, m		1.28, m
4	33.0, C	-	33.7, C	-
5	55.0, CH	0.78, dd (12.1, 2.6)	57.1, CH	0.85, d (2.5)
6a	18.2, CH2	(eq.) 1.58, m	69.1, CH	4.52, ddd (4.0, 2.4, 1.6)
6b		(ax.) 1.50, m	-	-
7a	40.8, CH2	(eq.) 1.78, dt (12.6, 3.2)	50.6, CH2	(eq.) 2.07, dd (14.2, 2.1)
7b		(ax.) 1.04, ddd (12.7, 12.7, 4.4)		(ax.) 1.31, dd (14.5, 4.1)
8	38.0, C	-	37.4, C	-
9	61.2, CH	1.16, dd (11.0, 2.2)	61.2, CH	1.01, d (2.7)
10	43.6, C		43.8, C	-
11a	21.3, CH2	(eq.) 2.68, ddd (14.2, 3.5, 3.5)	69.8, CH	4.87, ddd (4.8, 2.7, 2.1)
11b		(ax.) 1.47, m
12a	41.1, CH2	(eq.) 2.06, ddd (12.6, 3.4, 3.4)	47.3, CH2	1.90, dd (13.8, 4.4)
12b		(ax.) 1.71, ddd (13.1, 13.1, 4.2)		2.34, dd (14.0, 2.7)
13	77.9, C	-	77.0, CH	-
14	52.1, CH	1.63, m	52.7, CH	1.56, dd (13.2, 4.6)
15a	22.1, CH2	2.62, m	22.0, CH2	2.85, dd (16.3, 13.5)
15b				2.69, dd (16.6, 4.8)
16	122.1, C	-	122.2, C	-
17	157.6, C	-	157.1, C	-
18	116.9, CH	6.75, d (8.5)	116.8, CH	6.75, d (8.6)
19	128.9, CH	7.75, dd (8.5, 2.2)	129.0, CH	7.75, dd (8.5, 1.9)
20	121.3, C	-	121.2, C	-
21	132.0, CH	7.78, d (2.1)	131.8, CH	7.80, d (1.2)
22	32.8, CH3	0.84, s	32.4, CH3	0.96, s
23	21.0, CH3	0.81, s	23.5, CH3	1.24, s
24	12.3, CH3	0.92, s	15.5, CH3	1.65, s
25	16.2, CH3	0.91, s	17.9, CH3	1.57, s
26	20.7, CH3	1.19, s	21.6, CH3	1.42, s
27	167.2, C	-	167.2, C	-
28	51.7, CH3	3.86, s	51.8, CH3	3.87, s

11-Hydroxytolypodiol (2):

white, amorphous solid; ${\left[\alpha \right]}_D^{22}$ −160 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 290 (3.81), 264 (4.27) nm; ECD (1.1 × 10⁻⁴ M, MeOH), λ_max (Δε) 262 (−1.9), 222 (1.2) nm IR (CaF₂ disc) ν_max 3472, 2928, 2852, 1699, 1608, 1581, 1437, 1437, 1269 cm⁻¹; NMR data Table 1; HRESIMS m/z 473.2895 [M+H]⁺ (calcd for C₂₈H₄₁O₆⁺, 473.2898).

Tolypodiol (3):

ECD (1.1 × 10⁻⁴ M, MeOH), λ_max (Δε) 264 (−3.7), 223 (1.2) nm.

Biological Assay:

Rat neonatal microglia were isolated and the assay performed as previously described.¹⁸ Primary rat neonatal microglia (200,000 cells/well) were activated with LPS (0.3 ng/mL) for 17 h at 39.5 °C. Thereafter, each compound was added 20 min before microglia were stimulated with phorbol myristate acetate (PMA) at 1 μM for an additional 70 min. PMA-stimulated O2- and TXB2 generation were determined as described¹⁸ using immunoassay (Cayman Chemicals) according to the manufacturer’s protocol. Samples were run in duplicate.

Computational Analysis.

Conformers within 5 kcal/mol of the lowest energy conformer were searched using the Monte Carlo multiple minimum (MCMM) method and the OPLS-2005²⁴ force field in MacroModel²⁵ (Schrodinger Inc.). Each conformer within 5 kcal/mol of the lowest energy conformer was optimized in Gaussian09⁸ at the B3LYP²⁶/DEF2SVP¹⁰ level and the geometries of all conformers with similar energies were checked for redundancy. Density functional theory (DFT) was used to perform calculations, which were carried out in Gaussian 09. TDDFT¹¹ calculations at the several levels, including BH&HLYP¹⁰/DEF2TZVPP,¹³ were conducted to calculate the electronic excitation energies and rotational strengths in MeOH. Boltzmann weighted ECD spectra (Figure S15) were calculated using SpecDis¹² for comparison with the experimentally determined data recorded in MeOH and to calculate similarity factors between experimental and computational spectra.