Fungi produce a wide variety of structurally unique and biologically active secondary metabolites.12 Of the fungal natural products, meroterpenoids, which contain fragments from both terpenoid and polyketide precursors, are widely distributed in several species, in particular, Penicillium and Aspergillus spp.3 The remarkable structural diversity and significant bioactivity of these compounds have attracted significant chemical and biomedical interests.4 One particularly interesting group of fungal meroterpenoids is the tetra- or penta-cyclic austin class. Since austin, the first example, was identified from Aspergillus ustus in late 1970s,5 various compounds of this family of highly oxygenated meroterpenoids have been isolated from both Aspergillus and Penicillium spp.6 Recently, studies on these compounds have been focused on their biosynthesis in which not only a full biosynthetic pathway but also key enzymes with fascinating activities have been defined.789 These studies have significantly contributed to the structural diversity of fungal meroterpenoids and the potential of expression systems in the biosynthesis of fungal natural products.

In our search for fungi-derived novel compounds, a strain (strain number FCH061) of Penicillium sp. was isolated from sediment samples collected offshore of Chuja-do, Korea. Based upon the results of LC-ESI-MS analysis of the crude culture extract, the presence of various secondary metabolites prompted us to thoroughly investigate this strain. The large scale cultivation, followed by extraction and separation using a variety of chromatographic methods yielded fourteen compounds (1 – 14) including two new compounds, preaustinoids E (1) and F (2) (Fig. 1). Herein we report the structure determination of these compounds as new austin-type meroterpenoids by a combination of spectroscopic analyses.

The molecular formula of preaustinoid E (1) was deduced as C₂₅H₃₂O₆ by HRFABMS analysis. The ¹³C NMR data of this compound showed signals of three carbonyl groups at δ_C 213.9, 171.7 and 167.6 which were indicative of a ketone and two ester carbonyls, respectively (Table 1). Signals of four olefinic carbons were also found at δ_C 155.3 (CH), 146.6 (C), 120.9 (CH), and 107.6 (CH₂). Corresponding proton signals were found at δ_H 6.22 (1H, d, J = 12.2 Hz), 5.83 (1H, d, J = 12.2 Hz), 5.14 (1H, s), and 5.12 (1H, s) in the ¹H NMR spectrum. The remaining signals in the ¹³C NMR data were attributed to six unprotonated, three methine, three methylene, and six methyl carbons. From the ten degrees of unsaturation inherent in the mass data, the ¹³C NMR data suggested 1 is a pentacyclic compound.

Given this information, the structure of compound 1 was determined by a combination of 2-D NMR experiments. First, all of the protons and their attached carbons were precisely matched by an HSQC experiment. Since the COSY data indicated only a small number of proton spin systems, the structure determination was mainly accomplished by the HMBC experiments (Fig. 2). That is, the significant differentiation of the COSY signals corresponding to the double bond (δ_C 155.3, δ_H 6.22; δ_C 120.9, δ_H 5.83) suggested it was in conjugation with an electron-withdrawing group, which was confirmed by its long-range coupling with a carbonyl carbon (δ_C 167.6) in the HMBC data (C-1-C-3). In addition, the HMBC correlations of two methyl groups (δ_C 32.3, δ_H 1.40; δ_C 26.6, δ_H 1.42) with an unprotonated carbon (δ_C 85.5) readily defined an isopropyl group (C-4, C-14 and C-15). Additional HMBC correlations of these methyl groups suggested an adjacent methine carbon (C-5; δ_C 56.0). Then, the connection of this moiety with the pre-defined conjugated carbonyl group via an unprotonated carbon (C-10; δ_C 43.7) was also constructed by a series of HMBC correlations including a methyl proton (H₃-13; δ_H 1.16); H-1/C-5, H-2/C-10, and H₃-13/C-1, C-5, C-10. The diagnostic shielding (δ_C 85.5) of the C-4 carbon suggested an ester bridge with the C-3 carbonyl carbon constructing a 7-membered lactone (ring A; C-1-C-5, C-10, and C-13), which was confirmed by interpretation of the mass data (discussed later).

The COSY data revealed a direct connection between the C-5 methine and the ethylene group at C-6 and C-7, which was extended by a HMBC correlation with an isolated methyl group (C-12; δ_C 18.2, δ_H 1.29). Subsequently, the HMBC correlations of H₃-12 and H₃-13 with a common methine (C-9; δ_C 47.8, δ_H 2.15) constructed a 6-membered ring with two methyl substituents (ring B; C-5-C-10, C-12 and C-13). Similarly, another 6-membered ring (ring C; C-8, C-9, C-11, C-2′, C-3′, and C-7′) as well as attachment of an exomethylene (C-1′) and a methyl group (C-9′) to this ring was constructed by a COSY-derived proton spin system (H-9-H₂-11) coupled with the HMBC correlations of key protons with neighboring carbons: H₃-12/C-7′; H₂-1′/C-3′ and C-7′; and H₃-9′/C-11, C-2′ and C-3′. An additional HMBC correlation with H₃-9′ placed a ketone carbon at the neighboring position (C-4′; δ_C 213.9), and this fragment was extended by long-range couplings of a hydroxyl proton (δ_H 3.05) with this ketone and an unprotonated carbon (C-6′; δ_C 90.6). The latter carbon was also connected to a COSY-derived methyl methine moiety (C-5′ and C-10′) by the HMBC correlations of H-5′/C-6′ and H₃-10′/C-6′.

The proton-proton and proton-carbon based 2-D NMR analyses indicated an ester carbonyl carbon (δ_C 171.7) and six open ends (C-3, C-4, C-5′, C-6′, and C-7′ (two ends)), and of these, the latter accounts for three of the degrees of unsaturation inherent in the mass data (Fig. 2). The significant shielding of the C-5′ methine (δ_C 75.6, δ_H 4.23) was indicative of an ester conjugated with a carbonyl group (C-8′). Therefore, the only plausible connections of these fragments in the eastern portion of 1 were those between C-6′ and C-7′ and between C-7′ and C-8′, establishing a ketone-bearing 5-membered ring (ring D; C-2′-C-4′, C-6′, and C-7′) and a 5-membered methyl lactone (ring E; C-5′-C-8′ and C-10′), respectively. Similarly, an ester bridge was also placed between C-3 and C-4 to form a 7-membered lactone (ring A). These interpretations were supported by a literature study, and the NMR data of these moieties were in good agreement with reported values.12131415

Preaustinoid E (1) possessed several stereocenters mostly at the ring junctions. Accordingly, the configurations at these positions were assigned by conspicuous cross-peaks among the bridgehead methyl protons and neighboring protons in the NOESY data (Fig. 3). That is, H₃-12, H₃-13, and H₃-15 were axially oriented and cofacial by their characteristic NOESY cross-peaks: H-6β/H₃-12, H-6β/H₃-13, H-11β/H₃-12, H₃-12/H₃-13, and H₃-13/H₃-15. In contrast, the H-5 and H-9 methines were located on the opposite face, indicating of trans A/B and B/C ring junctures, by the NOESY correlations of H-5/H-9 and H-5/H₃-14.

The configurations of the remaining portion of 1 were also assigned based on the NOESY data. That is, the chair type conformation of the C-ring as well as the equatorial orientation of the C-9′ methyl group was derived from the spatial proximity between the C-12 methyl and the C-1′ exomethylene group: H-11β/H₃-9′, H₃-12/H₂-1′ (δ_H 5.14), and H₃-9′/H₂-1′ (δ_H 5.12). In addition, the α- and axial orientations of the D/E ring juncture were defined by the cross-peaks of 6′-OH with those on the A-C ring plane: H-7α/6′-OH and H-9/6′-OH. Similarly, a cis orientation was found between 6′-OH and 10′-CH₃ on the E ring: 6′-OH/H₃-10′. These assignments were supported by the NOESY cross-peaks between H₂-1′ and H-5′, which could be due to not only an axial orientation of the groups at the D/E ring junction and β-orientation of H-5′ relative to the E-ring but also the β-orientation of the C-8′ carbonyl carbon at C-7′ (Fig. 3). Overall, the relative configuration of 1 was assigned as 5R^*, 8R^*, 9R^*, 10R^*, 3′S^*, 5′S^*, 6′S^*, and 7′S^*, which was in good accordance with that of similar compounds.131416 Thus, the structure of preaustinoid E (1) was determined to be a new austin-type meroterpenoid.

The molecular formula of preaustinoid F (2) was established to be C₂₅H₃₂O₆, the same as 1, by HRFABMS analysis. The spectroscopic data of this compound were very similar to those of 1, suggesting their structures were closely related. A combination of 1-D and 2-D NMR analyses confirmed their similarity and indicated 2 and 1 had the same planar structure. However, detailed examination of the ¹H and ¹³C NMR data revealed significant differences in the chemical shifts of the protons and carbons at the C-5′ methine, C-10′ methyl and 6′-OH groups, suggesting they were C-5′ and/or C-6′ epimers (Table 1). This interpretation was confirmed by the NOESY data of 2 in which opposite cross-peaks were found for H-5′ and H₃-10′: H₂-1′/H₃-10′ and H-5′/6′-OH (Fig. 3). Thus, the structure of preaustinoid F (2) was defined as the C-5′ epimer of preaustinoid E (1).

In addition to compounds 1 and 2, extensive separation of the moderately polar chromatographic fractions allowed us to isolate several austin-type meroterpenoids. A combined of spectroscopic analyses of these congeners identified twelve known compounds, preaustinoid A2 (3), preaustinoid D (4), dehydroaustinol (5), dehydroaustin (6), actoxydehydroaustin (7), isoaustinone (8), 11β-acetoxyisoaustinone (9), (5′R)-isoaustinone (10), austin (11), neoaustin (12), austinoneol A (13), and verruculogen (14), the only prenylated diketopiperazine (Fig. 1). All the spectroscopic data for these compounds were in good accordance with those in the literature.512131415161718192021

Austin-type meroterpenoids have attracted significant biosynthetic interests.789 Unfortunately, these compounds failed to exhibit significant bioactivities, and only insecticidal and weak bacteriostatic activities against E. coli have been reported.1622 In our measurement, these compounds were also inactive against a variety of human cancer cell-lines (IC₅₀ > 50 µM) and human pathogenic bacterial and fungal strains (MIC > 128 µM). In addition, none of these meroterpenoids exhibited any positive results (IC₅₀ > 100 µM) in tests of selected enzyme-inhibition (isocitrate lyase, Na⁺/K⁺-ATPase, and sortase A) activities.