Antifungal Depsidone Metabolites from Cordyceps dipterigena, an Endophytic Fungus Antagonistic to the Phytopathogen Gibberella fujikuroi

Abstract

Among thirty four endophytic fungal strains screened for in vitro antagonism, the endophytic fungus Cordyceps dipterigena was found to strongly inhibit mycelial growth of the plant pathogenic fungus Gibberella fujikuroi. Two new depsidone metabolites, cordycepsidone A (1) and cordycepsidone B (2), were isolated from the PDA culture extract of C. dipterigena and identified as being responsible for the antifungal activity. Elucidation of their chemical structures was carried out using 1D and 2D NMR spectroscopy in combination with IR and MS spectroscopic data. Cordycepsidone A displayed strong and dose-dependent antifungal activity against the plant pathogenic fungus Gibberella fujikuroi. The isolates were inactive in bioassays for malaria (Plasmodium falciparum), leishmaniasis (Leishmania donovani), Chagas’s disease (Trypanosoma cruzi), and cytotoxicity at 10 μg/mL. The compounds were also found to be inactive against several bacterial strains at 50 μg/mL.

The International Cooperative Biodiversity Group program in Panama (ICBG-Panama) has been investigating cytotoxic, anti-parasitic and anti-microbial agents from various natural sources such as plants,¹ marine organisms,² and more recently, endophytic fungi.^3,4 From a total of 3582 endophytic fungal extracts, a number of active materials have been identified in our cytotoxic or anti-parasitic drug screening. The endophytic strains were screened for identification of antifungal metabolites using antagonism assay methods⁵ in which the endophytic strains were grown together with a fast growing phytopathogen. Thirty-four endophytic strains were screened against the well known phytopathogen Gibberella fujikuroi. This phytopathogen devastates rice crops by causing ‘bakanae’ disease (from the Japanese ‘foolish seedling’),⁶ a condition which results from over-production of the plant growth hormone gibberellic acid.

Unfortunately, some higher yielding strains of rice in Asia, Africa and North America are very susceptible to infection by this pathogen. After a series of antagonistic screens on potato dextrose agar media of endophytic fungi, the most potent strain was found to be Cordyceps dipterigena⁷ (strain F0307) with an inhibition zone diameter of 13 mm after seven days. The crude extract from the fungus inhibited the phytopathgen at 20 μg in a disk agar diffusion assay. On the basis of this activity the fungus was grown on a larger scale to identify its active secondary metabolites. Herein, we describe the isolation⁸ and structural elucidation of two new antifungal depsidone compounds, cordycepsidone A (1) and cordycepsidone B (2), as the active metabolites from the extract. The new compounds were also evaluated for their anti-parasitic, anti-bacterial and cytotoxic activities.

Compound 1 was obtained as a white amorphous powder, mp 197 °C. The molecular formula C₁₈H₁₂O₈ was determined by APCI-HR-TOFMS (m/z 357.0606 [M + H]⁺). The ¹H NMR spectrum of 1 showed a resonance at 10.5 for an aldehyde proton, one aromatic proton at 6.85, one oxymethylene group at 5.27, and two aromatic methyl groups at 2.42 and 2.13. The ¹³C NMR spectrum possessed 18 carbon signals with their number of attached protons determined from a DEPT spectrum. These were present as two methyls [δ_C 11.3 and 21.7], one methylene [δ_C 68.7], two methines [δ_C 118.1 and 193.5] and thirteen quaternary carbons. Two long range correlations were observed by COSY between the methyl singlet at 2.42 and the aromatic proton and the methyl singlet at 2.13 and the oxymethylene protons. These suggested that the further downshielded methyl group was ortho to the aromatic proton whereas the second methyl group was ortho to the oxymethylene. HMBC correlations observed from an aromatic proton at _H 6.85 (H-8) to C-16 (δ_C 193.5), C-7 (δ_C 165.3), C-10 (δ_C 112.6), C-11 (δ_C 160.7), and C-17 (δ_C 21.7) allowed identification of a substructure corresponding to aromatic ring A. The positions of the substituent groups about ring A were confirmed by observing HMBC correlations from _H 10.54 (H-16) and _H 2.42 (H-17) to various carbon resonances. Subsequently, a second substructure corresponding to aromatic ring C was identified by observing HMBC correlations from the oxymethylene at 5.27 (H-15) to C-1 (δ_C 168.6), C-2 (δ_C 109.6), C-3 (δ_C 146.4), C-14 (δ_C 145.1), C-13 (δ_C 114.7), and C-12 (δ_C 148.1) as well as from the methyl group at 2.13 (H-18) to carbons C-12 (δ_C 148.1), C-13 (δ_C 114.7), C-14 (δ_C 138.9) and C-4 (δ_C 138.9). This led to the hypothesis that aromatic ring A in 1 was similar to the lichen metabolite, diffractione A,⁹ whereas ring C was comparable to an endophytic fungus metabolite known as excelsione.¹⁰ A remarkable number of weak ⁿJ_CH couplings (n>3), such as between H15/C5 (⁷J_CH) and H15/C4 (⁵J_CH), were observed in the HMBC spectra of 1. These correlations were entirely consistent with the observations made for these two related metabolites.^{9, 10} Therefore, structure based on the known metabolites was elucidated as shown in Fig. 1, named as cordycepsidone A.¹¹

Figure 1.

Chemical Structures of Cordycepsidone A (1) and B (2).

Compound 2 was isolated as a white amorphous powder, mp 205 °C. The molecular formula of 2 was established as C₁₈H₁₂O₉ by APCI-HR-TOFMS (m/z 373.0563 [M + H]⁺), 16 mass units higher than compound 1, which suggested the addition of a single oxygen atom. The ¹H NMR spectrum was very similar to that of 1, but lacked an aldehyde proton at 10.54; rather, it exhibited a proton signal corresponding to a carboxylic acid ( 14.17). This change from an aldehyde at position 16 to a carboxylic acid was the only structural difference between the two compounds. Detailed assignments for the carbon and protons in cordycepsidone B¹² were accomplished by 2D NMR spectroscopic data and comparison with compound 1.

All extracts¹³ were subjected to spectroscopic analysis to identify any new metabolites arising from an antagonistic effect. The chemical composition of these extracts was very similar, indicating that antagonism does not have a marked effect on the secondary metabolite profile of F0307. In the paper disk assay, the extracts showed identical inhibition against phytopathogen, suggesting that the antifungal metabolite production was not induced in F0307 by the phytopathogen, but rather, is constitutively expressed. The extracts were also tested against the parasites⁴ Plasmodium falciparum (malaria), Leishmania donovani (leishmaniasis), and Trypanosoma cruzi (Chagas’disease), as well as MCF-7 cancer cell lines4, and found to be inactive at 10 μg/mL against all. Both compounds were inactive against several bacterial strains (Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa at 50 μg/mL).

1 and 2 belong to the depsidone class of compounds, which are more commonly associated with lichens than endophytic fungi.^{10, 15} It is probable that compound 2 is derived from 1 by oxidation enzymes present in the fungal strain F0307. Compound 1 showed moderate to good growth inhibitory activity against phytopathogens G. fujikuroi (MIC, 8.3 μg/mL) and Pythium ultimum (MIC, 1.2 μg/mL), but was less potent (>50) against the G. fujikuroi anamorph Fusarium subglutinans (Table 1). However, compound 2 showed a general reduction in activity in the antifungal assays, indicating the importance of the aldehyde functionality to the biological properties of compound 1. It is noteworthy that both new depsidones were highly selective against fungi and not the other assay organisms tested.

Table 1.

Antifungal Activity¹⁴ [MIC, μg/mL]^a of 1 and 2

Compound	Giberella fujikuroi	Pythium ultimum
1	8.3 ± 2.7	1.2 ± 0.3
2	>50	25.0 ± 0.1
Cycloheximide	0.39 ± 0.1	0.65 ± 0.2

^aThe experiments were performed in triplicate and the average values were expressed.

Overall, the results presented here suggest that depsidone 1, produced by the endophytic fungal strain C. diptergena, is responsible for the observed antifungal activity. Considering the importance of pathogen control in food production, the antagonistic screen could be a useful and cost effective method to identifying novel chemical entities as antifungal agents.