Substituted l-tryptophan-l-phenyllactic acid conjugates produced by an endophytic fungus Aspergillus aculeatus using an OSMAC approach

The endophytic fungus Aspergillus aculeatus isolated from leaves of the papaya plant Carica papaya was fermented on solid rice medium, yielding a new l-tryptophan-l-phenyllactic acid conjugate (1) and thirteen known compounds (11, 14–25). In addition, an OSMAC approach was employed by adding eight different sodium or ammonium salts to the rice medium. Addition of 3.5% NaNO3 caused a significant change of the metabolite pattern of the fungus as indicated by HPLC analysis. Subsequent isolation yielded several new substituted l-tryptophan-l-phenyllactic acid conjugates (1–10) in addition to three known compounds (11–13), among which compounds 2–10, 12–13 were not detected in the rice control culture. All structures were unambiguously elucidated by one and two dimensional NMR spectroscopy and by mass spectrometry. The absolute configuration of the new compounds was determined by Marfey's reaction and X-ray single crystal diffraction. Compounds 19–22 showed cytotoxicity against the L5178Y mouse lymphoma cell line with IC50 values of 3.4, 1.4, 7.3 and 23.7 μM, respectively.


Introduction
Endophytic fungi thrive widely in different healthy tissues of living plants, and have signicant inuence on the growth of their hosts. 1,2 Endophytes also comprise a large reservoir of structurally diverse secondary metabolites including alkaloids, steroids, terpenoids, xanthones, peptides and quinones, which exhibit a variety of biological activities including anticancer, antibacterial, antifungal, anti-inammatory and antidepressant activity. 2,3 For example, 14-membered macrolides isolated from the endophytic fungus Pestalotiopsis microspora showed significant cytotoxicity against the murine lymphoma cell line L5178Y while fusaric acid derivatives produced by the fungal endophyte Fusarium oxysporum showed signicant phytotoxicity to leaves of barley. 4,5 However, the high rediscovery rate of known compounds from fungi is a severe obstacle for the search for new drug leads. 6 One of the approaches to increase the diversity of metabolites from fungi involves the application of the OSMAC (One Strain Many Compounds) method which is based on systematic variations of the cultivation parameters (media type and composition, pH value, temperature etc.). [7][8][9] For instance, the fermentation of the fungus Gymnascella dankaliensis on solid rice medium following addition of 3.5% NaBr led to the isolation of ten brominated tyrosine-derived alkaloids, which showed cytotoxicity against the L5178Y mouse lymphoma cell line. 10,11 In addition, cultivation of Fusarium tricinctum on fruit and vegetable juice-supplemented solid rice media that led to the production of the new fusarielins K and L. 12 The fungus Aspergillus aculeatus has been reported to produce several bioactive secondary metabolites, such as secalonic acids D and F, aculeatusquinone B and D, aculeacins A-G and aspergillusol A. [13][14][15][16][17] In this study, A. aculeatus was isolated from leaves of Carica papaya collected in Awka in Nigeria. The fungus was fermented on solid rice medium, yielding a new Ltryptophan-L-phenyllactic acid conjugate (1) together with thirteen known compounds including N-[(2S)-2-hydroxy-1-oxo-3-phenylpropyl]-L-tryptophan methyl ester (11), 18 okaramine A (14), 19 oxaline (15), 20 emindole SB (16), 21 16-keto-aspergillimide (17), 22 JBIR 75 (18), 23 secalonic acids D and F (19 and 20), 13 asperdichrome (21), 24 RF 3192C (22), 25 pannorin (23), 26 altechromone A (24), 27 and variecolactone (25). 28 Due to the pronounced chemical diversity of this fungus we decided to subject A. aculeatus to an OSMAC approach by adding either 3.5% NaCl, 3.5% NaBr, 3.5% NaI, 1% NaF, 3.5% NaNO 3 , 3.5% NH 4 Cl, 3.5% (NH 4 ) 2 SO 4 or 3.5% NH 4 OAc to the rice medium. The selection of these salts for the OSMAC study was based on previous experiments with other fungi that had indicated the usefulness of these chemical stimuli for the accumulation of cryptic metabolites. 10,11,29 The fungus did not grow on rice medium containing 1% NaF or 3.5% NH 4 OAc. Addition of 3.5% NaNO 3 , however, caused a signicant change of the metabolite pattern as indicated by HPLC analysis. Addition of the remaining salts had no inuence. Subsequent workup of the extract resulting from addition of 3.5% NaNO 3 yielded several new substituted L-tryptophan-L-phenyllactic acid conjugates (1)(2)(3)(4)(5)(6)(7)(8)(9)(10) in addition to three known analogues N-[(2S)-2-hydroxy-1oxo-3-phenylpropyl]-L-tryptophan methyl ester (11), N-[(2S)-2hydroxy-1-oxo-3-phenylpropyl]-L-tryptophan (12) and acudioxomorpholine (13). 18 Compounds 2-10, 12-13 were not detected in the rice control culture. On the other hand, from all compounds isolated from the rice control culture, only oxaline (15), JBIR 75 (18), secalonic acid F (20) and RF 3192C (22) could be detected in the fungal culture aer addition of 3.5% NaNO 3 . In this study we report the structure elucidation and the biological activities of the isolated compounds.

Results and discussion
Compound 1 had the molecular formula C 22 H 24 O 4 N 2 as established by HRESIMS exhibiting 12 degrees of unsaturation. The 13 C NMR spectrum of 1 (Table 1) included 22 signals, corresponding to two methyls, two methylenes, two alphatic methines and ten aromatic methines as well as six quaternary carbons. The 1 H NMR spectrum of 1 (Table 2) showed the presence of a monosubstituted benzene ring and an indole moiety as indicated by signals between d H 6 and 8, two -CH 2 -CHunits and two methyl groups (d H 3.70 and 3.62). These data were similar to those of the co-isolated known compound N-[(2S)-2-hydroxy-1oxo-3-phenylpropyl]-L-tryptophan methyl ester (11). 18 However, the detection of an additional methyl substituent at d C 32.7 and d H 3.70 and the HMBC correlation from the protons of this methyl group to C-10 (d C 138.5) and C-12 (d C 129.0) indicated its attachment to N-11. Detailed analysis of the 2D NMR spectra of 1 (Fig. 2) revealed that its remaining substructure was identical to that of 11. The absolute conguration of 1 was determined by Xray single crystal analysis as 2S, 2 0 S (Fig. 3), being identical to that of 11. Thus, the structure of 1, for which the trial name aculeatine A is proposed, was elucidated as shown in Fig. 1.
Compound 2 possessed the molecular formula C 21 H 22 O 4 N 2 as determined by HRESIMS, which indicated the loss of a methyl group compared to 1. This was conrmed by the absence of the 1 H and 13 C signals of the methoxy group attached to C-1 in the NMR spectra of 2 (Tables 1 and 2). The structure of 2 was conrmed by detailed analysis of the 2D NMR spectra. Its absolute conguration was determined as 2S, 2 0 S by X-ray single crystal analysis (Fig. 3).
The molecular formula of compound 3 was determined as C 27 H 32  2), C-14 (d C 41.9) and C-15 (d C 149.1) and from Me-13 (d H 3.72) to C-12 and C-10 (d C 139.0) indicated the attachment of a 1,1-dimethylprop-2-en-1-yl side chain at C-12. Detailed analysis of the 2D NMR spectra of 3 ( Fig. 2) revealed that its remaining substructure was identical to that of 1.
Compound 4 had the same molecular formula as 3. The 1 H and 13 C NMR data were also similar to those of 3. The appearance of a broad triplet proton (d H 5.11, H-15) and one methylene group (d H 3.48 and 3.44, H 2 -14) together with two olenic methyl groups (d H 1.74 and 1.82, Me-17 and 18) in the 1 H NMR spectrum of 4 suggested the existence of a 3-methyl-but-2-en-1yl group at C-12, which was conrmed by the COSY correlations between H-15 and H 2 -14 together with the HMBC correlations from Me-17 and 18 to C-15 (d C 122.4) and C-16 (d C 133.9), and from H 2 -14 to C-4 (d C 106.0) and C-12 (d C 138.8) (Fig. 2).
Compound 5 was isolated as white powder. Its molecular formula was established as C 26 H 30 O 4 N 2 by HRESIMS. The 1 H and 13 C NMR spectra of 5 were similar to those of 4 except for to C-15 (d C 121.5) and C-16 (d C 137.1) indicated the presence of a 3-methyl-but-2-en-1-yl residue and its attachment to N-11 of compound 5 (Fig. 4). The molecular formula of 6 was determined as C 25 H 28 O 4 N 2 , indicating the lack of a methyl group compared to 5, which was conrmed by the absence of signals of the methoxy group at C-1 compared to 5 (Tables 3 and 4). The remaining substructure of 6 was unambiguously elucidated to be identical to that of 5 by detailed analysis of 2D NMR spectra.
Compound 7 possessed the molecular formula C 27 H 32 O 4 N 2 as deduced from the HRESIMS data. The 1 H and 13 C NMR spectra of 7 were similar to those of 4, revealing the presence of one monosubstituted benzene ring, an indole moiety and a 3methyl-but-2-en-1-yl side chain. However, the appearance of an aromatic proton signal at d H 6.82 (H-7), which was split to a doublet, suggested that the 3-methyl-but-2-en1-yl group was substituted on the benzene ring of the tryptophan moiety. The HMBC correlation from H 2 -14 (d H 3.73) to C-5 (d C 126.8), C-6 (d C 135.4) and C-7 (d C 120.7), and from H-7 to C-5, C-9 (d C 108.5) and C-14 (d C 33.3) indicated the 3-methyl-but-2-enyl group to be located at C-6. The structure of 7 was determined by analysis of the 2D NMR spectra as shown in Fig. 4.
Compound 8 possessed the molecular formula C 26 H 30 O 4 N 2 as established by HRESIMS, thus lacking a methyl group compared to 7. The 1 H and 13 C NMR spectra were almost identical to those of 7 except for the absence of signals of the methoxy group at C-1, which was further conrmed by detailed analysis of the 2D NMR data of 8. In addition, the HMBC correlation from H 2 -14 to C-6 (d C 118.9), C-7 (d C 134.4) and C-8 (d C 123.5), from H-6 (d H 7.36) to C-4 (d C 109.4), C-8, C-10 (d C 137.1) and C-14 (d C 32.0), and from H-8 (d H 7.02) to C-6, C-10 and C-14 indicated the 3hydroxy-3-methylbutyl moiety to be attached at C-7 (Fig. 4).
Compound 10 shared the same molecular formula as 9. The 1 H NMR data of 10 resembled those of compound 9 except for the chemical shis of the aromatic protons of the indole moiety. The HMBC correlation from H 2 -14 (d H 2.78) to C-7 (d C 121.0), C-8 (d C 137.7) and C-9 (d C 109.4), from H-7 (d H 6.91) to C-5 (d C 127.7), C-9 and C-14 (d C 32.3), and from H-9 (d H 7.11) to C-5, C-7 and C-14 conrmed that the 3-hydroxy-3-methylbutyl moiety was located at the C-8 position. The remaining substructure of 10 was the same as that of 9 as conrmed by detailed analysis of 2D NMR spectra of 10 (Fig. 4).
Compounds 3-10 share the same core structure with compounds 1 and 2, for which the absolute conguration had been assigned through X-ray analysis. Hence it is concluded on biogenetic terms that the absolute conguration of the former compounds is also 2S, 2 0 S. Compounds 11 and 12 were determined to have 2S absolute conguration by Marfey's reaction. Based on the close biogenetic similarity, the absolute conguration at C-2 0 of the latter two compounds is assumed to be S. A plausible biosynthetic pathway of the substituted L-tryptophan-L-phenyllactic acid conjugates obtained from A. aculeatus is proposed to start from L-tryptophan and phenylpyruvate. The important intermediate 12 is suggested to be formed by a condensation reaction between L-tryptophan and L-phenyllactic acid. The indole metabolites isolated in this study could be produced by further methylation of 12 and prenylation of the indole nucleus of tryptophan as shown in Fig. 5. The results of the OSMAC experiments indicated that substituted L-tryptophan-L-phenyllactic acid conjugates were induced by nitrate but not by sodium in the medium. In the absence of nitrogen sources favoured by the fungus including ammonium and glutamine, fungi are able to use secondary nitrogen sources such as nitrate, purines, urea, amines and amides etc., which is commonly known as nitrogen metabolite repression (NMR). 30 In fungi, the activity of nitrogen regulators for derepression of NMR genes, which also affect secondary metabolites formation, is regulated by the intracellular nitrogen status and extracellular nitrogen availability. 31,32 For example, the main GATA transcriptional regulator of nitrogen   metabolism AreA accumulates in the nucleus in Fusarium graminearum with nitrate as sole nitrogen source, which is required for activation of the nitrate assimilation system including the nitrate reductase genes. 33 Thus, the biosynthesis of fungal secondary metabolites can be affected by the quality and quantity of the nitrogen sources. For example, 67% of secondary metabolites silent gene clusters of Fusarium fujikuroi were expressed based on the modication of nitrogen sources. 34 Furthermore, the natural product beauvericin was accumulated by Fusarium oxysporum by utilizing nitrate as sole nitrogen source. 35 In this study, the production of substituted L-tryptophan-L-phenyllactic acid conjugates was stimulated by the activation of the nitrate assimilation system in A. aculeatus due to the presence of sodium nitrate in medium.
The substituted L-tryptophan-L-phenyllactic acid conjugates identied in this study showed no cytotoxic or antibacterial activity. However, the two known compounds (11 and 12) were claimed as plant growth regulators in a patent and showed pronounced rooting promoting effect. 36 It may hence be hypothesized that a high concentration of nitrate in host plants, as simulated in this study by addition of sodium nitrate to solid rice medium, may induce the production of plant growth stimulating indole metabolites of the endophytic fungus A. aculeatus. 37 This could lead to an increased growth and production of biomass by the host plant. In return, the fungus could receive nutrients, water, minerals and nitrogen from its host. Further studies will be necessary to evaluate this hypothesis.

General procedures
A Jasco P-2000 polarimeter was used to measure the optical rotation. 1D and 2D NMR spectra were recorded on Bruker Avance DMX 600 or 700 NMR spectrometers. Chemical shis were referenced to the solvent residual peaks. Mass spectra were recorded with a LC-MS HP1100 Agilent Finnigan LCQ Deca XP

Fungal material and cultivation
A. aculeatus was isolated from leaves of Carica papaya collected in Awka in Nigeria and was identied by DNA amplication, sequencing of ITS region and by comparing with GenBank data (GeneBank accession no. KX137846) following standard procedures. 38 The fungal strain was grown on solid rice medium (100 g rice and 100 mL distilled water autoclaved) in ten Erlenmeyer asks (1 L each) at 22 C under static conditions for 14 days. The OSMAC experiments were performed on rice medium containing either 3.5% NaCl, 3.5% NaBr, 3.5% NaI, 1% NaF, 3.5% NaNO 3 , 3.5% NH 4 Cl, 3.5% (NH 4 ) 2 SO 4 or 3.5% NH 4 OAc under static conditions until they reached their stationary phase of growth (16 days except for rice media spiked with 1% NaF or 3.5% NH 4 OAc where the fungus failed to grow).

Extraction and isolation
Fungal cultures were extracted with EtOAc followed by evaporation under reduced pressure. Initial purication of the EtOAc extract (12.5 g) of the fungal culture fermented on rice was performed by partitioning between n-hexane and 90% aqueous MeOH. The 90% aqueous MeOH phase (9.2 g) was centrifuged and then fractionated by vacuum liquid chromatography on reversed-phase silica gel using a gradient elution of H 2 O-MeOH (10 : 90 -0 : 100) to give 10 fractions (Fr.1 to Fr.10).
The fungal cultures from the OSMAC experiments that were grown on rice medium containing different salts were extracted with EtOAc (2 Â 500 mL) followed by solvent evaporation under reduced pressure. The obtained crude extracts were analyzed by HPLC. The EtOAc extracts (28.6 g) of fungal cultures that had been grown on rice medium (30 asks) aer adding 3.5% NaNO 3 were dissolved in MeOH and then subjected to vacuum ltering. The obtained MeOH solution was evaporated under reduced pressure and fractionated by vacuum liquid chromatography on silica gel using a gradient elution of n-hexane-EtOAc to give 24 fractions (Fr.N1 to Fr.N24). Fr.N19 (221 mg) was separated by a Sephadex LH-20 column with MeOH as mobile phase followed by semi-preparative HPLC to give 3 (3.5 mg) and 4 (15.0 mg). Fr.N22 (465 mg) was separated by a Sephadex LH-20 column with MeOH as mobile phase to give 3 fractions. Fr.N22-2 (54 mg) was further puried by semipreparative HPLC to give 1 (10.1 mg), 2 (10.4 mg), 11 (3.5 mg), 12 (3.4 mg) and 13 (2.5 mg). Fr.N22-3 (380 mg) was further puried by a silica gel column with DCM-MeOH as mobile phase followed by semi-preparative HPLC to give 5 (2.4 mg), 6 (0.9 mg), 7 (4.8 mg), 8 (4.7 mg), 9 (1.2 mg) and 10 (1.2 mg).
X-ray crystallographic analysis of compounds 1 and 2 Crystallization conditions. Suitable single crystals of 1 and 2 were obtained by slow evaporation from methanol solution and selected under a polarized light microscope. Data collection: compounds 1 and 2 were measured on a Bruker Kappa APEX2 CCD diffractometer with micro focus tube using Cu-Ka radiation (l ¼ 1.54178Å). APEX2 was used for data collection, 39 SAINT for cell renement and data reduction, 39 and SADABS for experimental absorption correction. 40 SHELXT was used for the structure solution by intrinsic phasing, 41 SHELXL-2017 was used for renement by full-matrix least-squares on F 2 . 42 The hydrogen atoms were positioned geometrically (with C-H ¼ 0.95Å for aromatic CH, 1.00Å for tertiary CH, 0.99Å for CH 2 and 0.98Å for CH 3 ). The renement was carried out using riding models (AFIX 43, 13, 23, 137, respectively), with U iso (H) ¼ 1.2U eq (CH, CH 2 ) and 1.5U eq (CH 3 ). The hydrogen atoms in the hydroxy and amine groups were rened with U iso (H) ¼ 1.5U eq (O/ N). The hydrogen atoms in the solvent methanol molecule in 2 were rened with U iso (H) ¼ 1.5U eq (O).

Marfey's reaction for compounds 11 and 12
Compounds 11 and 12 (0.5 mg) were hydrolyzed with 2 mL 6 M HCl containing 0.4% b-mercaptoethanol at 110 C for 24 h. The hydrolysate was evaporated to dryness and treated separately with 4 M NaOH at room temperature for 4 h. 4 M HCl was used to adjust the pH to 4. The above resulting solutions were evaporated until complete elimination of HCl and then resuspended in 50 mL H 2 O. To 25 mL of each resulting solutions was added 50 mL FDAA (1% 1-uoro-2-4-dinitrophenyl-5-L-alanine amide in acetone) and 10 mL NaHCO 3 . The reaction tubes were covered with an aluminum paper and heated over a hot plate at 40 C for 1 h. Aer cooling to room temperature, 5 mL of 2 M HCl was added and then evaporated to dryness. The residue was dissolved in 500 mL MeOH. L-Tryptophan and D-tryptophan were treated separately with FDAA in the same manner. The analysis of FDAA derivatives were carried out using HPLC and LC-MS by comparison of the retention time and molecular weight.

Cytotoxicity assay
Cytotoxicity was tested against the L5178Y mouse lymphoma cell line (European Collection of Authenticated Cell Cultures, Catalogue no. 87111908) using the MTT method as described before. 11 Kahalalide F was used as positive control with a IC 50 value of 4.3 mM.

Conflicts of interest
There are no conicts to declare.