Exophiala sp.

Cheikh-Ali, Zakaria, Glynou, Kyriaki, Ali, Tahir, Ploch, Sebastian, Kaiser, Marcel, Thines, Marco, Bode, Helge B. & Maciá-Vicente, Jose G., 2015, Diversity of exophillic acid derivatives in strains of an endophytic Exophiala sp., Phytochemistry 118, pp. 83-93 : 84-86

2.2. Isolation and identification of Exophiala sp. secondary metabolites

The ethyl acetate extract of strain P1860 was chromatographed by semi-preparative RP-HPLC to furnish exophillic acid (5, 2 mg /plate) and its corresponding monomeric form (19, 0.2 mg /plate). Compound 20 (1 mg /plate) was isolated from the ethyl acetate extract of strain P1095 by RP-HPLC. Compound 19 was obtained as a slightly violet oil, and its ESIMS showed a molecular ion [M—H] — at m / z 441.2136, in accordance with the molecular formula C 22 H 33 O 9. The IR spectrum showed the presence of a carboxyl (1735 cm —1) and aromatic (1655 cm —1) moieties. 1 H NMR analysis showed two doublets corresponding to aromatic meta-coupled protons (δ H 6.38, 6.56, J = 5.0 Hz) and a triplet (δ H 0.9, J = 10.0 Hz) corresponding to a terminal methyl group. The 13 C NMR spectrum obtained in deuterated methanol ( Table 1 ) exhibited resonances for 22 carbons, including an ester carbonyl at δ 172.6 (C- 10), a methyl at δ C 14.4 CH 3 (C- 160) and eight methylene groups between δ C 23.7 and δ C 34.9, characteristic of methylene groups present in a long alkyl chain. The aromatic ring was characterized by signals corresponding to quaternary carbons at δ C 118.2, 160.6, 157.5, and 144.3 (C- 20, C- 30, C- 50, and C- 70 respectively) and two CH groups at 102.2 (C- 40), 111.7 (C- 60) showing correlations with δ H 6.56 and δ H 6.38 ppm respectively in HSQC. The value of the coupling constant between the hexose protons (J = 7.5 Hz; Table 1 ) and the chemical shifts of C- 100 to C- 600 (δ C 103.6, 75.0, 77.7, 71.2, 78.3, and 62.5, respectively) were in good agreement with reported data for a β- D- glucopyranosyl moieties ( Fig. 4 ; Ondeyka et al., 2003; Waki et al., 2007).

Interpretation of 1 H, 13 C, and 2D NMR (COSY–HSQC, and HMBC; Figs. S2–S 6) spectra allowed the structure elucidation of compound 19 to be 4-hydroxy-2-nonyl-6-(((2 S,3 R,4 S,5 S,6 R)-3,4,5-trihydroxy- 6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)benzoic acid.

The corresponding dimeric form, exophillic acid (5), has been previously isolated ( Ondeyka et al., 2003). Interestingly, we isolated a novel dimeric structure (20) from another Exophiala sp. strain, P1095. Compound 20 displayed in ESIMS a molecular ion [M—H] — at m / z 541.3195, consistent with the molecular formula C 32 H 45 O 7. Thus the molecular weight of compound 20 was 162 Da less than exophillic acid, a shift that is in accordance with the loss of the β- D- glucopyranosyl moiety ( Fig. 5 ). Subsequently, the structure of compound 20 was established by 1D and 2D NMR (Figs. S7–S11) as 4-((2,4-dihydroxy-6-nonylbenzoyl)oxy)-2- hydroxy-6-nonylbenzoic acid.

A variety of related structures were also identified by analytical HPLC/MS based on comparisons with the above described authentic compounds (5, 19, and 20). All additional secondary metabolites identified in the Exophiala strains are shown in Tables 2 and 3 (for a detailed description see Tables S14 and S15). All of the newly identified metabolites (21–29) possessing a β- D- glucopyranoside are present in strains with a slimy phenotype, namely P1860, P1910 and P2772 ( Table 2 ). Conversely, metabolites 30–38 identified in strains that present a dry phenotype with aerial mycelium, P1095 and P2854, lacked the β- D- glucopyranoside residue ( Table 3 ). In general derivatives with double bonds and hydroxyl groups in the alkyl side chains have been identified according to MS2 fragmentation data ( Tables 2 and 3 ). To confirm the presence of double bonds in the identified structures, crude extracts from P1860 and P1095 were derivatized using dimethyl disulfide (DMDS) in a iodine solution ( Francis and Veland, 1987), resulting in the detection of the corresponding dimethyl disulfide adduct in HPLC–MS analysis (identified by a mass increase of 94 Da; Fig. S12).

The positions of these double bonds were determined as follows: crude extracts were methanolyzed under acidic conditions followed by addition of DMDS as described previously ( Bode et al., 2006; Francis and Veland, 1987), and then were subjected to GC/MS analysis. Due to the low amount of unsaturated derivatives ( Tables 2 and 3 ) we were only able to determine the potential position of a double bond in compounds 22, 31, and 32. The specific fragments allowed the localization of the double bond between C-9 and C-10 (Fig. S13).

The position of the hydroxyl groups have not been determined as they are present only in trace amounts.

Compound 19: slightly violet oil; UV (MeOH) λ max 220, 290 nm; IR (film, MeOH) v max 3410, 2945, 1735, 1655, 1450, 1260, 1205, 1045, 820 cm —1; HRESIMS m / z 441.2136 (calcd for C 22 H 34 O 9 — H —, 441.2130, Δ = 1.36 ppm); t R (HPLC) 8.6 min.

Compound 20: yellow oil; UV (MeOH) λ max 215, 280, 310 nm; IR (film, MeOH) v max 3265, 1670, 1235, 1150, 860, 755; HRESIMS 541.3169 m / z (calcd for C 32 H 45 O 7 + H —, 541.3171, Δ = 0.37 ppm); t R (HPLC) 14.1 min.

2.3. Bioactivity of Exophiala sp. secondary metabolites

The major compounds 5, 19 and 20 were tested against the causative agents of tropical neglected diseases Trypanosoma brucei rhodesiense , Trypanosoma cruzi , Leishmania donovani and Plasmodium falciparum , and their cytotoxicity was tested against mammalian L6 cells. For T. brucei , 19 had the best activity with an IC 50 of 16 µg ml —1, whereas the dimeric compounds 5 and 20 showed activities above 35 µg ml —1. A similar activity of all compounds was observed against P. falciparum with 25, 22 and 26 µg ml —1 for 5, 20, and 19, but they only had a weak bioactivity against T. cruzi and L6 cells (30–60 µg ml —1). A clear difference between the glycosylated and non-glycosylated compounds was observed against L. donovani . Whereas 5 and 19 showed an IC 50 of 55 and 63 µg ml —1, 20 showed an IC 50 of 16 µg ml —1. However, compared to standard drugs against these parasites, the activity of all tested compounds was at least 100-fold lower.

Figures

Fig. 4. Selected 1H–1H COSY (bold lines) and HMBC (arrows) correlations of compounds 19 and 20.

Fig. 2. Maximum Likelihood ITS rDNA phylogeny of the endophytic Exophiala sp. strains in this study (in bold), in relation to their 25 BLAST closest matches (including substrate and country of origin; ‘?’ indicates uncertainty in the country of origin) and the 25 closest reference strains at NCBI GenBank (GenBank accession numbers before sequence descriptions; voucher numbers of reference strains after species names). Only a portion of the whole phylogenetic tree (gray rectangle in the inset) is shown to better depict the relationship of the strains with their closest sequences. For a full version of the ML tree see Fig. S1. Bootstrap support values>70% for the maximum likelihood and for a Neighbor-Joining analysis (ML/NJ) are shown next to each tree branch.

Fig. 5. Mass spectrometry spectra and major fragments of 5 and 20. For each compound, the upper and lower plots show the MS and MS2 spectra, respectively. Continuous arrows represent the fragmentation of each compound, and the dashed arrow represents the mass difference between compounds 20 and 5.