Endophytes are the microorganisms that reside in the inner tissues of living plants during all or part of the microbial life cycle. In 1993, Strobel and coworkers reported that an endophytic fungus isolated from Taxus brevifolia can produce the taxol,1 which was a milestone event in natural product discovery history. After that initial report, the plant endophytes have attracted considerable attention of natural products chemists in the search for new natural products with novel structures and bioactivities. Examples for new bioactive natural products obtained from fungal endophytes include cytoskyrins A and B,2 antimalarial codinaeopsin,3 anticancer rvirgatolides A–C,4 tyrosine phosphatase B inhibitor asperterpenoid A,5 acetylcholine esterase inhibitor asperterpenols A and B,6 anti-TMV aspergillines A–E,7 and the antifungal compound varioxepine A.8

In the past two decades, the large number of natural products obtained from plant-associated endophytes showed an enormous increase. Nevertheless, more and more compounds with known structures are re-isolated reflecting the bottle neck of standard pure fermentation procedures that fail to activate all biogenetic gene clusters that are present in a given fungus.9 In the host plants, the endophytes are always part of complex microbial communities that influence each other metabolically leading to an enhanced production of bioactive natural products. These triggers are lacking in axenic cultures. In order to gain access to a wider variety of compounds, a new microbial culture method, called “black-box” co-culture, is proposed for the rapid discovery of new compounds exhibiting antimicrobial activity from plant-associated endophytes. This new method followed the “Acquired Immune Systems Theory”.10 It took advantage of the fact that all culturable endophytic microorganisms when cultured in the same culture vessel could influence each other and maximize the chemical productivity by inducing biogenetic gene clusters that remained silent when cultured axenically. For mimicking the situation in the host plant, the biomass of host plant is added to the fermentation. This dynamic cocultivation in a plant-like environment is expected to yield new bioactive compounds.

During our ongoing search for new bioactive compounds from endophytes from rare and endangered plants from the Shennongjia District (People's Republic of China), some metabolites such as tremulanes and 5, 6-secotremulane,11 penicitroamide,12 new glycosides13 and fusaodavinvin14 had been reported earlier by us. In this paper, the structural elucidation and antimicrobial activity of two new polyketides produced by endophytes from the Chinese plant Distylium chinense following our “black-box” approach are reported.

Chinoketide A (1) was obtained as colorless oil, ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$= −35 (c 0.1, CH₃OH). Its UV spectrum displayed maximum absorption peaks at 245.3 nm. Its molecular formula was determined as C₁₁H₁₆O₄ by the pseudomolecular ion peak at [M+H]⁺ m/e 213.1124 (calcd. 213.1127) in the HRESIMS spectrum. The ¹H and ¹³C NMR spectra of 1 showed seven groups of proton signals and eleven carbon resonance signals. In the ¹³C and DEPT NMR spectra, there were four quaternary carbon signals at δ 199.4, 151.9, 127.4 and 76.4, two methine groups at δ 71.4 and 68.8, three methylene groups at δ 63.1, 36.9 and 36.7, and two methyl groups at δ 21.0 and 18.1. In the 1H -1H COSY spectrum, the cross peak between δ 3.73 (m, 1H) (H-3) and 2.52 (m, 1H)/2.18 (m, 1H) (H-4) proved that H-3 was adjacent to H-4, and the proton signal at δ 3.52 (m, 1H) correlated to two proton signals at 1.16 (d, 6.1, 3H) (H-9) and 2.09 (m, 2H) (H-7), suggesting the existence of substructure of −CH2-CH(OR)-CH₃. Following comprehensive analysis of the HSQC and HMBC spectra of compound 1, the planar structure of compound 1 was built up (shown in Fig. 1). In the HMBC spectrum of compound 1, H-10 correlated with C-1, C-2 and C-3, whereas H-3 correlated with C-1, C-2, C-4, C-5 and C-10, on the other hand, H-4 correlated with C-2, C-3, C-5 and C-6 indicating the presence of a six-membered ring with an α,β-unsaturated ketone core in chinoketide A. The cross peaks between H-11 and C-4, C-5, C-6 and between H-4 and C-5, C-6 suggested that C-11 was attached to an α,β-unsaturated ketone core at C-5. And H-7 correlated with C-1, C-5 and C-6, H-8 correlated with C-6 proving that C-7 was connected to C-6 of the six-membered ring. Then, the correlations from H-7 to C-8, C-9, from H-8 to C-7, C-9, from H-9 to C-8, C-9 verified the relative positions of C-7, C-8 and C-9. Finally, the correlations from H-8 to C-11 and from H-11 to C-8 proved that C-8 and C-11 were linked together to form a second six-membered ring in compound 1 via an ether bond. Thus, the structure of chinoketide A was established as shown in Fig. 1.

In chinoketide A (1), there are three chiral centers at C-2, C-3 and C-8, and C-2 and C-3 bear a diol function group. The relative configurations at the chiral carbons were elucidated by a thorough analysis of the NOESY spectrum of compound 1. The correlation from H-3 to H-10 suggested that the methyl group of H-10 and H-3 were oriented at the same side, and the two hydroxyl groups were oriented to the opposite side. The third chiral carbon C-8 was far away from C-2 and C-3. Hence, it was difficult to assign the relative configuration of C-8. However, a detailed analysis of the NOESY spectrum of compound 1, the correlations from H-9 to H-11a, then from H-11a to H-4b, and the cross peaks between H-8 and H-11b, and cross peaks between H-11b and H-4a, proved that H-8 was oriented to the same side as the two adjacent hydroxyl groups linking at C-2 and C-3. On the other hand, H-9 was oriented to the same side as H-10 and H-3, which was opposite to the two adjacent hydroxyl groups linking at C-2 and C-3. Finally, the relative configuration was established as shown in Fig. 1. Due to the presence of an α,β-unsaturated ketone chromophore in chinoketide A, the ECD of compound 1 was calculated to determine the absolute configurations of C-2,C-3 and C-8 using the TDDFT method. Comparing the theoretically calculated CD spectrum with the experimentally recorded CD spectrum (see Fig. 3), showed a significant similarity. Hence, the absolute configurations of C-2, C-3 and C-8 were determined as 2R, 3R and 8S. Thus, the structure of compound 1 was established as shown in Fig. 1. The new natural product of compound 1 was named as chinoketide A.

Chinoketide B (2) was obtained as obtained as a colorless oil, ${\left[\mathrm{\alpha }\right]}_{\mathrm{D}}^{20}$= +32 (c 0.1, CH₃OH). Its UV spectrum displayed maximum absorption peaks at 218.5, 231.3, 282.7 nm. Its molecular formula was determined as C₉H₈O₅ by the molecular ion peak at [M]⁺ m/e 196.0374 [M]⁺ (calcd. for C₉H₈O₅, 196.0372) in the HR-EI-MS spectrum. The ¹H and ¹³C NMR spectra were very simple, there were only five groups of proton signals in ¹H NMR spectrum, and only nine carbon signals in ¹³C NMR spectrum. There were three quaternary carbon signals at δ 199.8, 164.6, 147.8 and 108.1, four methine groups at δ 108.2, 101.1, 71.6 and 69.8, one methylene group at δ 43.1 (see Table 1).

Following inspection of the HSQC, HMBC and NOESY spectra, the structure of compound 2 was established as shown in Fig. 1. Similarly to compound 1, there was a diol group at C-5 and C-6. The only cross peak between H-5 and H-6 could be observed in the NOESY spectrum, which cannot appear that H-5 and H-6 lied at the same side only by NOESY spectrum. To determine the absolute configurations of C-5 and C-6, the calculation of four enantiomers' ECD spectra of compound 2 was carried out using the TDDFT method and comparing the calculated ECD with the experimentally recorded ECD spectrum. Only the 5R, 6S enantiomer's ECD spectrum showed significant similarity with the experimental CD spectrum of compound 2 (see Fig. 4). Thus the structure of compound 2 was established as shown in Fig. 1. and compound 2 was given the name chinoketide B.

The known compound 3 was determined its structure by comparing with their NMR data and X-ray diffraction to the reference compound xylarphthalide A,18 and we found that they were the same in NMR data and X-ray diffraction, Thus the compound 3 was identified as xylarphthalide A.

The antimicrobial activities of the isolated compounds were investigated using Erwinia carotovora sub sp. carotovora (Jones) Bersey et al as model organisms and employing the filter paper disc method with MIC value at 20.5, 30.4 and 10.2 µg/mL, amphotericin B was used as positive control with MIC value at 0.5 µg/mL.