Amycolachromones A–F, Isolated from a Streptomycin-Resistant Strain of the Deep-Sea Marine Actinomycete Amycolatopsis sp. WP1

In this study, a detailed chemical investigation of a streptomycin-resistant strain of the deep-sea marine, actinomycete Amycolatopsis sp. WP1, yielded six novel amycolachromones A–F (1–6), together with five known analogues (7–11). Amycolachromones A–B (1–2) possessed unique dimer skeletons. The structures and relative configurations of compounds 1–11 were elucidated by extensive spectroscopic data analyses combined with X-ray crystal diffraction analysis. Plausible biogenetic pathways of amycolachromones A–F were also proposed.


Introduction
Marine microbial natural products, especially those derived from marine actinomycetes, have become an important source of novel bioactive compounds [1][2][3]. However, traditional screening strategies generally do not provide access to the full array of secondary metabolites encoded within actinomycete genomes [4]. For example, Streptomyces coelicolor initially produces four classes of metabolites using laboratory fermentation, despite genome sequencing revealing the capacity to produce >30 families of metabolites [5,6]. To solve this problem, various strategies have been proposed to activate the expression of otherwise silent biosynthetic gene clusters, including the 'one strain many compounds' (OSMAC) approach [7], co-cultivation with other microorganisms [8] and chemical epigenetics [9]. Recently, a ribosome engineering approach that targets ribosomal proteins or RNA polymerase (RNAP) has shown promise for expression of cryptic gene clusters. This method selects for mutants that are resistant to antibiotics that target the bacterial ribosome, presumably activating the expression of bacterial cryptic genes by resistant mutants [10,11]. Shima and co-workers demonstrated this method in actinomycetes by activating the biosynthetic pathway for actinorhodin in mutant Streptomyces that developed resistance to streptomycin [12]. Recent adoptions of this approach demonstrated the ability of streptomycin-resistant mutants to enhance production of actinolactomycin [13], fredericamycin A and chlorinated alkaloids, inducamides A-C [14,15].
According to the HMBC correlations from H-1 to C-6, C-5 and C-7, the sulfur atom present in 4 was shown to be attached at C-1 and C-3 , indicating that C-1 was attached at C-6, Further confirmation was found for HMBC correlations of CH 3 O-7 to C7; H-3 to CH 3 -2, C-4a, C-2; H-8 to C-4a, C-6, C-8a, C-7, C-4, a methoxy group could be located at C-7 and a methyl group could be located at C-2. Selected key correlations in the observed NMR spectrum are shown in Figure 3. On the basis of these results, the structure of compound 4 was established as shown.
According to the HMBC correlations from H-1 to C-6, C-5 and C-7, H-3 to C-1 , the sulfur atom present in 5 was shown to be attached at C-1 and C-3 , indicating that C-1 was attached at C-6, Further confirmation was found for HMBC correlations of CH 3 O-7 to C7; H-3 to CH 3 -2, C-4a, C-2; H-8 to C-4a, C-6, C-8a, C-7 and C-4, a methoxy group could be located at C-7, a methyl group could be located at C-2. Selected key correlations in the observed NMR spectrum are shown in Figure 3. On the basis of these results, the structure of compound 5 was established as shown.
Amycolachromone  (Table 3) revealed sixteen carbon signals: three carbonyl group C-10 (δ C 198.6), C-12 (δ C 191.8) and C-14 (δ C 168.5), three aromatic carbon C-3 (δ C 138.7), C-2 (δ C 109.5), and C-4 (δ C 107.4), a nonoxygenated quaternary aromatic carbons at C-13 (δ C 106.5), two oxygenated quaternary aromatic carbons at C-1 (δ C 161.9) and C-5 (δ C 158.9), two sp 3 -quaternary carbon C-11 (δ C 90.0) and C-6 (δ C 73.0), a methoxy group C-15 (δ C 52.7), a methyl group C-16 (δ C 18.6), an oxygenated methine C-7 (δ C 71.8), a methine C-8 (δ C 31.1) and a methylene C-9 group (δ C 43.1). Analysis of the 1 H and 13 C NMR data of 6 revealed the presence of the same 5-hydroxy-4H-chromen-4-one moiety as found in xanthones [18,19]. In the 1 H-1 H COSY spectrum, the correlations from H-7 to H-8 and OH-7, from H-8 to H-9 and H-16. Further confirmation was found for HMBC correlations of H-7 to C-16, C-6, C-11 and C-9; H 3 -16 to C-7, C-8 and C-9, indicated that C-16 was attached to C-8, and OH-7 was located at C-7. HMBC correlations from the O-methyl proton signal H 3 -15 to the carboxylic carbon C-14 confirmed that the O-methyl group was located at C-14. HMBC correlations from OH-11 to C-11, C-6 and C-10, OH-1 to C-2, C-13 and C-1 indicated that OH-1 and OH-11 were attached to C-1 and C-11, respectively [22,23]. Selected key correlations in the observed NMR spectrum are shown in Figure 3. Thus, the planar structure of 6 was established. Moreover, the relative configuration of 6 was established to be 6R*, 7S*, 8R* and 11R* by X-ray crystallography using Mo Ka radiation (Figure 4). C-11, C-6 and C-10, OH-1 to C-2, C-13 and C-1 indicated that OH-1 and OH-11 were attached to C-1 and C-11, respectively [22,23]. Selected key correlations in the observed NMR spectrum are shown in Figure 3. Thus, the planar structure of 6 was established. Moreover, the relative configuration of 6 was established to be 6R*, 7S*, 8R* and 11R* by X-ray crystallography using Mo Ka radiation (Figure 4).   Further analysis of the structures allowed us to raise a plausible biosynthetic pathway of compounds 1-6. As outlined in the Scheme 1, compounds 1-5 were structurally related to the known metabolite 6-methoxymethyleugenin, which was derived from the widely existing 5,7-dihydroxy-2-methylchromone via the hydroxymethylation with formaldehyde and the methylation with SAM (S-adenosyl methionine). The compound 1 was the dimerization of 6-methoxymethyleugenin, and the sequential methylation with SAM could afford the related compound 2. For compound 3-5, we proposed that the sulfur in these structures was from L-cysteine. Thus, the Michael addition of L-cysteine to the ortho-quinone methide intermediate from 6-methoxymethyleugenin gave the compound I. Then, transamination, decarboxylation and reduction sequence of I furnished the 2-sulfo-ethanol II occurred. An oxidation of sulfur in II gave the compound III. Finally, compound 3 was obtained through the double oxidation of sulfur in II. The oxidation of the hydroxyl group in III to the corresponding carboxylic acid occurred and followed with a decarboxylation afforded for compound 5. Furthermore, compound 4 was the oxidation product of 5 [24,25]. In addition, compound 6 was the oxidation product of the known natural product blennolide B, which was proposed by Franck to be a derivative of emodin (Scheme 2) [26].
The AlkB family of DNA repair enzymes utilize an α-ketoglutarate/Fe(II)-dependent mechanism to oxidize the aberrant alkyl groups, finally repairing alkyl DNA bases [27,28]. Compounds 1-11 were evaluated for their in vitro ABH2 inhibitory activities. Compounds 1-11 exhibited weak inhibitory activity against the ABH2 enzyme. However, in 2019, a paper was published that tested emodin (10). It exhibited strong inhibitory activity for the ALKH3 enzyme with IC50 of 8.8 μM [29]. This hinted that these compounds might inhibit other members of the AlkB family of enzymes. In conclusion, the chemical investigation of a streptomycin-resistant strain of the deep-sea marine actinomycete, Amycolatopsis sp. WP1, led to the isolation and identification of six novel compounds, amycolachromones A-F (1-6) and five known analogues (7)(8)(9)(10)(11). Among them, amycolachromones A-B (1-2) represents an unusual fused skeleton be-Scheme 1. Proposed hypothetical biosynthesis pathway of 1-5. ported values and are described here for the first time as produced by Amycolatopsis sp.
The AlkB family of DNA repair enzymes utilize an α-ketoglutarate/Fe(II)-dependent mechanism to oxidize the aberrant alkyl groups, finally repairing alkyl DNA bases [27,28]. Compounds 1-11 were evaluated for their in vitro ABH2 inhibitory activities. Compounds 1-11 exhibited weak inhibitory activity against the ABH2 enzyme. However, in 2019, a paper was published that tested emodin (10). It exhibited strong inhibitory activity for the ALKH3 enzyme with IC50 of 8.8 μM [29]. This hinted that these compounds might inhibit other members of the AlkB family of enzymes. Scheme 1. Proposed hypothetical biosynthesis pathway of 1-5.
The structures of five known compounds were identified as 6-ethoxymethyleugenin (7), 6-methoxymethyleugenin (8), 6-hydroxymethyleugenin (9), emodin (10) and the ascomycete metabolite chaetoquadrin D (11) by comparison of spectroscopic data with reported values and are described here for the first time as produced by Amycolatopsis sp.
The AlkB family of DNA repair enzymes utilize an α-ketoglutarate/Fe(II)-dependent mechanism to oxidize the aberrant alkyl groups, finally repairing alkyl DNA bases [27,28]. Compounds 1-11 were evaluated for their in vitro ABH2 inhibitory activities. Compounds 1-11 exhibited weak inhibitory activity against the ABH2 enzyme. However, in 2019, a paper was published that tested emodin (10). It exhibited strong inhibitory activity for the ALKH3 enzyme with IC 50 of 8.8 µM [29]. This hinted that these compounds might inhibit other members of the AlkB family of enzymes.
In conclusion, the chemical investigation of a streptomycin-resistant strain of the deepsea marine actinomycete, Amycolatopsis sp. WP1, led to the isolation and identification of six novel compounds, amycolachromones A-F (1-6) and five known analogues (7)(8)(9)(10)(11). Among them, amycolachromones A-B (1-2) represents an unusual fused skeleton between two 6-hydroxymethyleugenin, and the relative configuration of amycolachromones F (6) was determined by the signal-crystal X-ray diffraction. The discovery of amycolachromones A-F not only expanded the chemical diversity of natural products and inspire further synthetic studies, but also provided a template for the exploration of inhibitors of other members of the AlkB family of enzymes.

Materials and Methods
General experimental procedures. All chemical reagents and solvents were purchased from Sigma-Aldrich (Shanghai, China). UV spectra were acquired with a DU 800 UV/vis spectrophotometer (Beckman Coulter, Brea, CA, USA). IR spectra were acquired with a Nicolet 380 FT-IR (Thermo Electron Corporation, Beverly, MA, USA). NMR experiments were conducted using an Agilent NMR 500 MHz spectrometer (Santa Clara, CA, USA) and BRUKER NMR 600 MHz spectrometer (San Jose, CA, USA) with (CD 3 ) 2 SO as the solvent (referenced to residual DMSO at δ H 2.54 and δ C 39.5) at 25 • C. Electrospray ionization mass spectra (ESIMS) were acquired using an AB Sciex TripleTOF 4600 spectrometer (Boston, MA, USA) in the positive and negative ion mode. HPLC experiments were performed on a Hitachi Elite LaChrom system (Tokyo, Japan) equipped with a diode array detector model L-2450, pump L-2130 and autosampler L-2200. Semipreparative HPLC experiments were completed with a Waters XBridge Prep C 18 (Miflord, CO, USA) 5 µm, 10 mm × 250 mm column and Phenomenex Luna C 18 5 µm, 250 mm × 21.2 mm column.
Bacterial Strain and Culture Conditions. The WP1 strain (CGMCC No. 10738) was isolated from deep-sea sediments of the Southwest Indian Ocean and identified as Amycolatopsis sp. by 16S rRNA sequence comparison. WP1 was grown in ISP 2 medium consisting of 1.0% (w/v) malt extract, 0.4% (w/v) yeast extract, 0.4% (w/v) glucose and 3% (w/v) sea salt, the pH of medium was adjusted to 7.4 using 2 M HCl and 2 M NaOH.
Mutants of strain WP1. The WP1 strain suspensions were spread onto ISP 2 plates containing different concentrations (0, 10, 20, 30, 40, 50 and 60 mg/mL) of streptomycin. The plates were incubated at 37 • C for 7 days. Mutant colonies producing the white pigment different than the WP1 strain were selected, generating mutant strain L-30-6, which was obtained on the IPS 2 plate containing 30 mg/mL streptomycin.
Extraction and isolation. The mutant L-30-6 strain was inoculated into ISP 2 broth with 3% sea salt in 250 mL Erlenmeyer flasks, at 30 • C on a rotary shaker at 180 rpm for 2 days as seed culture. Each of the seed cultures (32 mL) was transferred into 1 L Erlenmeyer flasks containing 400 mL of ISP 2 supplemented with 3% sea salt. These flasks were incubated at 30 • C on a rotary shaker at 180 rpm for 6 days. The resulting cultures (60 L) were centrifuged to yield the supernatant and a mycelial pellet. The supernatant was adsorbed onto macroporous resin XAD16N (DOW, St. Louis, Missouri, CA, USA) and eluted with linear gradient of 0-100% EtOH in H 2 O to afford six fractions (A-F).