Tryptophan triggers the production of anthranilamide and quinazolinone derivatives in Streptomyces sp. PC4-3 Authors: Tam Pooprasert, Mark Aldren Feliciano, Sanya Sureram, Thammarat Aree, Nontaphat Leerach, Nopporn Chutiwitoonchai, Somboon Tanasupawat, Yongpeng Yao, Gang Liu, Chulabhorn Mahidol, Somsak Ruchirawat, Prasat Kittakoop Published: 2026 ABSTRACT Streptomycetes are a valuable source of bioactive compounds, yet their biosynthetic diversity is often limited in laboratory conditions. The one strain-many compounds (OSMAC) strategy has been used to trigger the production of diverse microbial metabolites. In the present work, we observed an increased production of anthranilamide derivatives in a culture of Streptomyces sp. PC4-3 when supplemented with l-tryptophan. Three new natural products—streptotryptophin A (1), trans-streptotryptophin B (2), and cis-streptotryptophin B (3), together with three known compounds including 2,3-dihydro-2,2-dimethylquinazolin-4(1H)-one (4), nocardamine (5), and geldanamycin (6), were isolated from the tryptophan-supplemented Streptomyces culture. Structure elucidation of the isolated compounds was performed by analysis of NMR and MS data, as well as by single-crystal X-ray analysis. Compounds 2 and 3 are new natural products, and the trans and cis configurations of 2 and 3 have never been reported. We conclusively established the trans and cis configurations of 2 and 3 by single-crystal X-ray analysis, and proposed a biosynthetic pathway for compounds 1–3. A large amount (389.8 mg) of the trans isomer 2 was obtained when compared to the cis isomer 3 (15.2 mg). To our knowledge, this is the first report on the production of compounds 1–3 in nature, which is stimulated by an amino acid tryptophan. The isolated compounds did not exhibit antiviral activity against SARS-CoV-2, however, nocardamine (5) and geldanamycin (6) had antibacterial activity. Introduction With the emergence of antibiotic resistance and the COVID-19 pandemic, the discovery of new antimicrobial and antiviral compounds is an urgent need.1Streptomyces, the largest genus of actinobacteria with over 800 species, is a major source of bioactive compounds, providing 70–80% of natural products used in pharmacological and agrochemical applications.2 These bacteria possess large genomes (generally 8–10 Mbp) that contain various biosynthetic gene clusters, enabling them to synthesize multiple classes of secondary metabolites, including polyketides, macrolactams, macrolides, glycosides, and terpenoids.3Streptomyces often produces a limited range of secondary metabolites in laboratory cultures due to silent biosynthetic genes, hindering the discovery of new bioactive compounds.4 In 2002, Bode and colleagues proposed an inexpensive method to activate the silent biosynthetic genes, revealing more diverse microbial metabolic profiles without requiring high-cost genetic manipulation. This method was termed the “One Strain Many Compounds” (OSMAC) approach, which relies on altering cultivation parameters, such as media composition, pH, temperature, and addition of chemical elicitors and biosynthetic precursors.4–6 Despite the extensive discovery of compounds derived from Streptomyces over years, genomic studies indicate that many biosynthetic gene clusters in these bacteria remain largely unexplored.7,8 Therefore, the OSMAC approach can be employed to expand the diversity of secondary metabolites produced by Streptomyces. In addition to being a key building block of proteins, tryptophan serves as a precursor for various secondary metabolites. It is commonly found as a residue in calcium-dependent lipopeptide antibiotics, such as daptomycin (Fig. 1).9 The phenoxazinone chromophore in the antibiotic actinomycin D is biosynthesized from tryptophan.10 Additionally, tryptophan is a precursor for Streptomyces indole alkaloids, including the antitumor staurosporine11 and the antifungal legonimide (Fig. 1).12 Examples of tryptophan-derived compounds from Streptomyces. The red parts indicate core structures from tryptophan. The integrated transcriptomic–proteomic study by Palazzotto and colleagues revealed that Streptomyces coelicolor cultured in a tryptophan-supplemented medium exhibited various biochemical changes, including differentiation, sporulation, and increased lipopeptide synthesis.13 Moreover, actinorhodin levels were increased, although tryptophan is not a direct precursor of actinorhodin.13 In this work, with the aim of discovering new bioactive compounds using OSMAC approach, tryptophan was fed to the Streptomyces sp. PC4-3 culture to trigger the production of distinct secondary metabolites. The bacterial strain PC4-3 was isolated from shore soil in Samed island, Thailand, and exhibited good antimicrobial activity.14 Upon the bacterial culture extraction and purification, we report herein the isolation of six compounds from the tryptophan-supplemented culture of the strain PC4-3. Among these, compounds 1–3 are anthranilamide derivatives that have not been obtained as natural products from living organisms. Results and discussion Effects of tryptophan for the production of anthranilamide derivatives by Streptomyces sp. PC4-3 The Streptomyces sp. PC4-3 was previously reported to produce the ansamycin antibiotics, geldanamycin and 17-O-demethylgeldanamycin, in ISP2 medium.14 16S rDNA sequencing revealed 99.1% similarity of this bacterium to Streptomyces sporoclivatus LMG 20312.14 In this study, we observed that culturing PC4-3 in tryptophan-supplemented ISP2 medium resulted in a different metabolic profile compared to the normal ISP2 medium (Fig. 2A and B). NMR spectra of ethyl acetate extracts from the PC4-3 cultures revealed that when the bacterium was supplied with tryptophan, it produced a major metabolite with a 1,2-disubstituted benzene structure (δH 7.58, 7.15, 6.74, 6.52) (Fig. 2B), which showed similar 1H NMR chemical shifts to those of anthranilamide derivatives.15 The present work used anthranilamide as a standard compound for 1H NMR comparison (Fig. 2C). The bacterial fermentation in tryptophan-supplemented ISP2 was subsequently scaled up to 9 liters, followed by extraction, purification, and spectroscopic analyses. This process led to the isolation of four anthranilamide derivatives (compounds 1–4), nocardamine (5), and geldanamycin (6), as shown in Fig. 3. Compound 1 was identified as a new natural product, while 2 and 3 have not been reported as originating from living organisms. 1H NMR spectra (400 MHz, acetone-d6) of EtOAc extract of PC4-3 cultured in ISP2 medium (A) and tryptophan-supplemented ISP2 medium (B), compared with anthranilamide standard (C). Chemical structures of the isolated compounds 1–6. Structure elucidation of new compounds Compound 1 was obtained as a yellow solid. The molecular formula of this compound was defined as C20H18N4O2 by HRESIMS, which showed a quasi-molecular ion [M + H]+ at m/z 347.1503, calculated for C20H19N4O2 with theoretical m/z 347.1502. The 1H and 13C NMR spectra of compound 1 showed only 9 carbon signals and 8 proton signals (Table 1), corresponding to 10 carbons (two carbons overlapping at δC 122.3 for the positions 9, 10 or 9′, 10′) and 9 protons (two protons overlapping at δH 7.14 for the positions 9, 10 or 9′, 10′), which are half of the number of carbons and protons determined by HRESIMS, suggesting that this compound is a symmetrical molecule. The following structure elucidation describes NMR data and 2D NMR correlations of a half molecule of compound 1. Analysis of DEPT 90 and DEPT 135 spectra revealed the presence of 5 methine and 4 quaternary carbon signals. 1H NMR data displayed 5 methine proton signals (δH 7.69, 7.29, 7.16, 7.14, and 6.73) at the aromatic region and 3 exchangeable proton signals (δH 9.96, 8.03, and 7.41). The 1H–1H COSY correlations indicated that 4 methine protons belonged to the same spin system of a 1,2-disubstituted benzene, with chemical shifts at δH 7.69, 7.29, 7.16, and 6.73. The broad NH signals at δH 8.03 and 7.41 were assigned to the NH2 group of the amide at C-7 or C-7′ (δC 171.5) because both protons showed exchange correlations with one another in the NOESY data and had NOESY correlations to H-6 or H-6′ (δH 7.6). The HMBC correlations of the NH at δH 9.96 (2,2′-NH) to C-1 (δC 116.4), C-3 (δC 114.0), C-8 (δC 136.1), and C-9 or C-10 (δC 122.3) confirmed the connection between anthranilamide and 1,4-para-disubstituted benzene ring, as illustrated in Fig. 4. In addition, four methine protons (H-9, H-10, H-9′, H-10′) of the 1,4-para-disubstituted benzene ring appearing as a singlet at δH 7.14 had a doubled peak area relative to the other methines, indicating that this molecule contained two anthranilamide rings and one 1,4-para-disubstituted benzene ring. Based upon these spectroscopic data, the structure of compound 1 was established, and a trivial name, streptotryptophin A, was given for this compound. Fortunately, crystals of streptotryptophin A (1) were obtained after recrystallization from a mixture of tetrahydrofuran and cyclohexane; its structure was definitively confirmed by single-crystal X-ray diffraction analysis. ORTEP plot of streptotryptophin A (1) is shown in Fig. 5. Table 1 1H (400 MHz) and 13C NMR (100 MHz) data of compounds 1–3 in DMSO-d6PositionStreptotryptophin A (1) trans-Streptotryptophin B (2) cis-Streptotryptophin B (3) δ C, type δ H, mult. (J in Hz) δ C, type δ H, mult. (J in Hz) δ C, type δ H, mult. (J in Hz)1,1′116.4, C114.8, C114.8, C2,2′146.1, C146.6, C146.6, C3,3′114.0, CH7.16, d (8.4)114.9, CH6.79, d (8.1)115.0, CH6.76, d (8.1)4,4′132.3, CH7.29, t (7.7)133.4, CH7.27, td (7.7, 1.6)133.6, CH7.26, td (7.7, 1.6)5,5′117.0, CH6.73, t (7.5)117.1, CH6.67, t (7.4)117.2, CH6.67, t (7.5)6,6′129.4, CH7.69, d (7.9)127.3, CH7.59, dd (7.7, 1.3)127.5, CH7.59, dd (7.7, 1.5)7,7′171.5, C163.3, C163.5, C8,8′136.1, C66.8, C66.9, C9,10,9′,10′122.3, CH7.14, s32.1, CH21.87, s32.2, CH21.8–2.0, m2,2′-NH9.96, s6.55, br s6.47, br s–CONHa8.03, br s7.67, br s7.81, br s–CONHb7.41, br s 1H (400 MHz) and 13C NMR (100 MHz) data of compounds 1–3 in DMSO-d6 Key 2D NMR correlations of streptotryptophin A (1) and trans-streptotryptophin B (2). ORTEP diagrams of 1–3 (30% probability level; atom colors: C, cyan; O, red; N, purple; H, white). The inversion center at the cyclohexane-ring centroid of the trans form of 2 is indicated by a blue dot. The cis form of 3 is stabilized by intermolecular OH⋯O hydrogen bonds with a lattice hydration water molecule (magenta line). Compounds 2 and 3 shared the same chemical formula, C20H20N4O2, as analyzed by HRESIMS, which resulted in quasi-molecular ions [M + H]+ at m/z 349.1654, calculated for C20H21N4O2 with theoretical m/z 349.1659 for both compounds. However, 1H and 13C NMR spectra of compounds 2 and 3 showed signals of one half of a molecule, e.g., signals of 10 carbons and 10 protons, indicating that the molecules of 2 and 3 are symmetrical. The following structure elucidation of compounds 2 and 3 is based on NMR data and 2D NMR correlations of their half molecule. In the case of 2, analysis of 13C and DEPT NMR data revealed the presence of 4 methine, one methylene, and 4 quaternary carbon signals. 1H–1H COSY data confirmed a spin system of 1,2-disubstituted benzene featuring 4 methine protons (H-3 to H-6, δH 6.6–7.6). HMBC correlations of both NH protons (δH 6.55 for 2,2′-NH and 7.67 for –CONH) to C-8 or C-8′ (δC 66.8) confirmed the closed ring structure of anthranilamide, forming a quinazolinone skeleton. The methylene protons at δH 1.87 (H-9, H-10, H-9′, H-10′) had a four-fold 1H peak area compared to the other methine signals and also showed HMBC correlation to C-8, suggesting that C-8 was connected to two equivalent methylene carbons, C-9 and 10 (δC 32.1). Therefore, 2 was a spirocyclic compound, composed of a 1,1,4,4 tetrasubstituted cyclohexane ring with two units of quinazolin-4(1H)-one (Fig. 3 and 4). The structure of 2 was also confirmed by single-crystal X-ray diffraction analysis, which revealed the trans configuration (ORTEP plot of 2 is shown in Fig. 5). Upon these spectroscopic data, the structure of compound 2 was secured, and a trivial name, trans-streptotryptophin B, was given for this compound. Compound 3 shared a great deal of similarity of 1H and 13C NMR data to that of trans-streptotryptophin B (2), suggesting that it was an isomer of 2. Analysis of 1D and 2D NMR data indicated that compound 3 had the same planar structure as that of 2, except the multiplicity of methylene protons at δH 1.8–2.0, which appeared as a singlet in 2 but as a multiplet in 3. Again, crystals of compound 3 were obtained, and single-crystal X-ray diffraction analysis established the structure of 3 as a cis isomer of trans-streptotryptophin B (2). ORTEP plot of compound 3 is displayed in Fig. 5. Because of compound 3 being a cis-isomer of trans-streptotryptophin B (2), a trivial name, cis-streptotryptophin B, was therefore given for compound 3. It is worth mentioning that Streptomyces sp. PC4-3 produces large amount of the trans isomer 2 (389.8 mg) when compared to the cis isomer 3 (15.2 mg). The exact match for streptotryptophin A (1) was not found from chemical databases. However, trans-streptotryptophin B (2) and cis-streptotryptophin B (3) were previously obtained from chemical syntheses using anthranilamide and 1,4-cyclohexanedione as starting materials. Several studies have reported the reaction conditions for this chemical transformation, including methods utilizing iodine,16,17 glyceroborate ionic liquid,18p-toluene sulfonic acid,19 and acid-functionalized nanoparticles.20,21 A similarity search for streptotryptophin A (1) resulted in N-substituted analogs featuring the same N,N′-diarylbenzene-1,4-diamine core.17 Wang and colleagues found that the condensation of anthranilamide derivatives and 1,4-cyclohexanedione generally produced dispiroquinazolinones (e.g., derivatives of 2 or 3) when using iodine as a catalyst.16,17 Interestingly, in some cases, the use of N-substituted anthranilamides as starting materials could yield either dispiroquinazolinone or N,N′-diarylbenzene-1,4-diamines (e.g., derivatives of 1). The synthesized compound, 1″,3″-dihydrospiro[1′,3′-dihydrospiro[cyclohexane-1,2′-quinazolin]-4′-one-4,2″-quinazolin]-4″-one, previously reported by Wang and colleagues16,17 had NMR spectra identical to those of trans-streptotryptophin B (2). However, NMR data that matched with streptotryptophin A (1) was not reported in Wang and colleagues’ work.16,17 Although trans-streptotryptophin B (2) and cis-streptotryptophin B (3) have been previously reported as chemically synthetic compounds, their stereochemistry at the 1,1,4,4 tetra-substituted cyclohexane ring has never been differentiated before.16–21 As evidenced by our analyses of NMR and X-ray diffraction data, the previously reported 2 and 3 should be revised to their correct trans and cis forms, respectively. We found that the NMR spectra of previously synthesized products from three research works16,19,20 were identical to those of trans-streptotryptophin B (2), featuring a singlet methylene proton around the δH of 1.8–2.0. The green synthesis by Yadev and colleagues21 provided a compound with NMR data showing a multiplet at δH 1.8–2.0, matching with cis-streptotryptophin B (3). In 2013, Safaei and colleagues developed a glyceroborate ionic liquid as a sustainable medium for the synthesis of quinazolinones.18 The structure of one of the synthesized compounds was displayed as that of cis-streptotryptophin B (3); however, its 1H NMR data showed two methylene signals at δH 1.09 and 1.47, which is inconsistent with our finding. Therefore, we recommend that the structure of that compound reported by Safaei's group18 should be revised to the correct structure. Streptotryptophin A (1) and streptotryptophin B (2 and 3) have never been isolated from any living organisms. However, some natural products containing anthranilamide or quinazolin-4(1H)-one in their structures have previously been isolated from Streptomyces bacteria, as illustrated in Fig. 6. A natural compound that closely resembles 2 and 3 is SIPI-886-C2 (Fig. 6), which was isolated from the Streptomyces strain OUCMDZ-5511 under fluoride stress by Liu and colleagues in 2024.22 Machushynets and colleagues employed an OSMAC approach and molecular networking to identify new derivatives of quinazolin-4(1H)-one, including quinazolinone A, quinazolinone B, and actinomycin L, derived from Streptomyces sp. MBT27 cultures.23,24 Quinazolinone A and B (Fig. 6) were isolated in 2019 by supplementing the MBT27 culture with glycerol.23 The discovery of novel actinomycins L in 2022 showed the potential of OSMAC approach to produce new compounds with a well-known core structure;24 these antibiotics feature the actinomycin moiety, of which the proline residue is connected to a quinazolin-4(1H)-one moiety (Fig. 6). In substrate-feeding experiments, they showed that anthranilamide, a substrate for actinomycin L biosynthesis, was biosynthesized from anthranilic acid.24 Isofuranonaphthoquinones (IFQs) and isoindolequinones (IIQs) were isolated from Streptomyces sp. CB01883 by Guo and colleagues in 2017.25 Among the eleven reported compounds, three contain either an anthranilamide or quinazolin-4(1H)-one moiety in their structures (Fig. 6). Additionally, they proposed that anthranilamide could incorporate into the IFQ parent compounds through a non-enzymatic biosynthesis, as supported by their analyses of chemical transformations and biosynthetic gene clusters.25 Examples of natural products from Streptomyces, which contain anthranilamide or quinazolin-4(1H)-one in the structures. Proposed biosynthetic pathways of anthranilamide and quinazolinone derivatives Based on the structural features of streptotryptophin A (1) and streptotryptophin B (2 and 3), we propose that they are biogenetically derived from l-tryptophan as shown in Fig. 7. The biosynthesis begins with the conversion of l-tryptophan to anthranilic acid via the kynurenine pathway.26,27 Once formed, anthranilic acid undergoes amidation to form anthranilamide, or, it undergoes multiple transformation steps to form 1,4-cyclohexanedione.28–30 Subsequently, two molecules of anthranilamide condense with one molecule of 1,4-cyclohexanedione to create an imine intermediate, which can either be oxidized and tautomerized into 1 or cyclized into 2 or 3. In the present work, it was found that the production of the trans isomer 2 (389.8 mg) was 25.6 times more than the cis isomer 3 (15.2 mg), indicating that enzymatic cyclization is selectively toward the production of the trans isomer 2. To our knowledge, this is the first work demonstrating the production of compounds 1–3 by the addition of an amino acid tryptophan to the bacterium culture medium. Proposed biosynthetic pathways of streptotryptophin A (1) and streptotryptophin B (2 and 3). The structures of compounds 4, 5, and 6 were identified by comparing their spectroscopic data with those in the literature.31–33 Compound 4, known as 2,3-dihydro-2,2-dimethylquinazolin-4(1H)-one, was previously isolated from the endophytic and mangrove-derived Streptomyces, and it displayed potent anticancer activity.31,34 However, the biosynthesis of 4 was not clearly mentioned in the existing literature. Based on its chemical structure, we suggest that 4 can be derived from the condensation of anthranilamide and acetone, a common volatile metabolite produced among many Streptomyces species.35,36 Compounds 5 and 6 were identified as nocardamine and geldanamycin, respectively, both of which are typically produced by Streptomyces bacteria.32,37 A limitation of this work is that the proposed biosynthetic pathways of the isolated compounds are not experimentally validated. To definitively elucidate the enzymes governing this biosynthetic pathway, a comprehensive validation strategy integrating bioinformatic mining, in vitro biochemical characterization, and in vivo genetic manipulation is required. Initially, the genome of Streptomyces sp. PC4-3 must be sequenced and mined to identify the putative biosynthetic gene cluster, enabling the precise annotation of candidate genes responsible for key transformations. These targeted genes are subsequently cloned, heterologously expressed in suitable hosts (such E. coli), and isolated via affinity chromatography. The core functional validation is achieved through enzymatic reconstitution assays, wherein the purified recombinant enzymes are incubated with chemically synthesized intermediate substrates and essential cofactors. The resulting catalytic conversions are then rigorously analyzed using high-resolution LC-MS to verify that the chromatographic and spectral properties of the biocatalytic products perfectly align with those of authentic synthetic standards. Finally, to corroborate these findings within a physiological context, targeted gene knockouts can be generated in the native producing strain, coupled with comparative metabolite profiling to observe the anticipated accumulation of upstream precursors and the concomitant abolition of final product biosynthesis. To systematically enhance the production yield of compounds 1–3 in the strain PC4-3, a multifaceted optimization strategy integrating genetic, metabolic, and bioprocess engineering must be employed. At the genetic level, the target biosynthetic gene cluster can be heavily refactored by replacing native regulatory elements with strong, constitutive promoters (such as ermE* or kasO*) and by overexpressing pathway-specific positive transcriptional regulators to bypass natural biosynthetic bottlenecks. Concurrently, metabolic flux toward the target compound can be maximized through the targeted deletion of highly expressed, competing secondary metabolite clusters, coupled with the exogenous supplementation or endogenous genetic overproduction of essential early precursors like l-tryptophan and anthranilic acid. To further unlock the strain's intrinsic biosynthetic potential, global strain improvement techniques, including ribosome engineering via antibiotic resistance selection and chemical epigenetic elicitation, can be applied to induce widespread pleiotropic activation of secondary metabolism. Finally, these cellular interventions must be synergistically paired with rigorous macroscopic bioprocess optimization, utilizing optimized culture media, optimized controlled-bioreactor parameters (such as precise dissolved oxygen and carbon-to-nitrogen ratios), and potential interspecies co-cultivation strategies to establish the ideal physiological and environmental conditions for maximum metabolite biosynthesis. Antibacterial and antiviral activities of the isolated compounds We evaluated antibacterial activity and antiviral activity against SARS-CoV-2 of the isolated compounds using the methods previously described;38,39 the methods are in the Supplementary materials. Streptotryptophin A (1), trans-streptotryptophin B (2), cis-streptotryptophin B (3), and compound 4 did not show antibacterial and anti-SARS-CoV-2 activities. In the present work, nocardamine (5), a cyclic hydroxamate siderophore, showed potent antibacterial activity by inhibiting the growth of Staphylococcus epidermidis TISTR 2172 (Gram-positive) and Escherichia coli ATCC 27853 (Gram-negative) with respective inhibition zones of 30.3 ± 1.5 mm and 19.7 ± 0.6 mm. Geldanamycin (6) inhibited the growth of S. epidermidis TISTR 2172 and Staphylococcus aureus ATCC 25923 with inhibition zones of 15.7 ± 0.6 mm and 14.0 ± 0.0 mm, respectively. Ciprofloxacin, the standard drug, had the inhibition zones of 27.3 ± 1.5 mm, 33.7 ± 0.6 mm, and 37.0 ± 1.0 mm against S. aureus ATCC 25923, S. epidermidis TISTR 2172, and E. coli ATCC 27853, respectively. However, the antimicrobial activities of nocardamine (5) and geldanamycin (6) were previously reported in the literature.14,40–43 A limitation of this work is that the biological activities of this work are limited to antiviral and antibacterial activities. The isolated compounds should also be evaluated for other biological activities, e.g., immunomodulatory and neuroprotective activities. Moreover, network pharmacology and molecular docking of the isolated compounds should be conducted in order to predict the enzyme targets for these natural products. Conclusions In summary, our work highlights the potential of OSMAC approach through feeding an amino acid tryptophan to induce the production of diverse secondary metabolites in Streptomyces bacterium. From the tryptophan-supplemented culture, we isolated anthraniamide derivatives including streptotryptophin A (1), trans-streptotryptophin B (2), cis-streptotryptophin B (3), and 2,3-dihydro-2,2-dimethylquinazolin-4(1H)-one (4), as well as two well-known Streptomyces metabolites, norcardamine (5) and geldanamycin (6). We identified 1 as a new compound and two new natural products 2 and 3 (synthetically known). Moreover, the structures of trans and cis isomers (2 and 3) were unambiguously established by single-crystal X-ray diffraction analysis in this work. The plausible biosynthetic pathway of 1, 2, and 3 was proposed in the present work, featuring the transformation of tryptophan precursor into anthranilamide and 1,4-cylohexanedione, both of which subsequently condense into the final products. Note that the detailed mechanisms and the specific enzymes involved in biosynthesizing these unique secondary metabolites (1, 2, and 3) should be investigated in the future. Experimental General experimental procedures NMR experiments were conducted using a Bruker AVANCE III HD 400 MHz spectrometer, operating at 400 MHz for 1H NMR and 100 MHz for 13C NMR. Analyses of HRESIMS utilized a Q Exactive Focus mass spectrometer (Thermo Scientific). Single-crystal X-ray diffraction analysis was performed with Bruker APEXII and Bruker PROSPECTOR diffractometers. Cultivation and extraction of PC4-3 metabolites Streptomyces sp. PC4-3 was provided by Prof. Dr Somboon Tanasupawat, Chulalongkorn University, Thailand. The bacterium was regularly sub-cultured onto International Streptomyces Project-2 (ISP2) agar medium, which consists of 4 g of yeast extract, 10 g of malt extract, 4 g of dextrose, and 20 g of agar per liter of distilled water, to maintain viability throughout the experiment. To prepare the seed culture, a small piece of the PC4-3 agar culture was inoculated into 250 mL of ISP2 broth and incubated at 200 rpm and 30 °C for 5 days. Large-scale fermentation was performed using 30 one-liter flasks, each containing 300 mL of tryptophan-supplemented ISP2 (4 g L−1 of l-tryptophan in ISP2 broth), equivalent to a total of 9 L culture. Each flask was inoculated with approximately 5 mL of the seed culture, followed by incubation at 200 rpm and 30 °C for 10 days. The entire culture was extracted three times, each with an equal volume of ethyl acetate. Approximately 30 L of ethyl acetate solution was collected and evaporated to obtain 6.69 g of a crude EtOAc extract. Purification of PC4-3 metabolites A crude EtOAc extract was resuspended in an excess amount of methanol and then filtered. The resulting insoluble solid was washed extensively with methanol and acetone, yielding a white powder of 2 (389.8 mg). The methanol solution was evaporated and resuspended again in approximately 60 mL of methanol. The saturated methanol solution (MS) was drawn through a cotton-plugged pipette, leaving behind 736.4 mg of residual solid (RS). The clear MS was subjected to Sephadex LH-20 open-column chromatography (5 cm diameter and 47 cm height) eluting with methanol. Twenty fractions were collected, each containing 60–120 mL. Fraction 2 (215.6 mg), after being resuspended in methanol and left overnight, yielded a white precipitate of 5 (53.5 mg). Fraction 16 (86.8 mg) was resuspended in methanol and acetone, yielding insoluble yellow powder of 1 (56.2 mg). Fraction 4 (148.7 mg) was subsequently purified by a silica gel (Merck Millipore, 1.07729) column chromatography, gradually eluting with 1 to 10% methanol in dichloromethane to obtain compound 6 (28.4 mg) as a yellow powder. Fraction 5 (987.7 mg) was purified by a silica gel column chromatography under the same solvent condition to obtain compound 4 (273.2 mg) as a colorless crystalline solid. Finally, the residual solid (RS) was purified by a silica gel column chromatography under the same solvent condition to obtain compound 3 (15.2 mg). Compounds 1–3 were recrystallized prior to single-crystal X-ray analysis. A small vial of compound 1 (8.0 mg) dissolved in tetrahydrofuran (THF) was surrounded with cyclohexane in a closed container. The yellow crystal of 1 grew after 5 days as a result of vapor diffusion. Compound 2 was also recrystallized by vapor diffusion using DMSO as solvent and water as antisolvent. Compound 3 was dissolved in THF/toluene/methanol mixture and left standing on a bench to allow slow evaporation. X-ray diffraction analysis of compounds 1–3 Light yellow, plate-like single crystals of 1 belonged to the monoclinic space group P21/c, while colorless rod-like crystals of trans–cis isomers 2 and 3 were found in P21/c and the triclinic P1̄ space groups, respectively. X-ray diffraction data were collected at 296(2) K using two Bruker KAPPA CCD diffractometers: APEXII (sealed-tube MoKα radiation, λ = 0.71073 Å) for 1, and PROSPECTOR (IµS microfocus CuKα radiation, λ = 1.54178 Å) for 2 and 3. Data processing—including integration, reduction, absorption correction, and merging—was performed using the APEX2 suite (SAINT, SADABS, and XPREP), yielding good-quality data (Rint ∼5–7%; completeness ∼100%). Structures were solved via intrinsic phasing (SHELXT) and refined by full-matrix least-squares on F2 (SHELXL). Non-hydrogen atoms were refined anisotropically, while hydrogen atoms were treated using a riding model. For 3, disordered electron density at the (100) face center was treated as a solvent void using the SQUEEZE/PLATON routine,44 as it could not be modeled as discrete water molecules. To resolve short intermolecular contacts, a BUMP restraint was applied; however, the resulting D–H⋯H–D distance involving one lattice water proton and one NH group remained relatively short at 1.95 Å. Final refinements for 1–3 converged to R1(F2) ∼5% with F2 > 2σ(F2). Crystal data and refinement details of 1–3 are summarized in Table S1. The ORTEP drawings of the asymmetric units of 1, 3, and the inversion-related molecular halves of 2 (the crystallographic inversion center at the centroid of the cyclohexane ring is highlighted by a blue dot) are displayed in Fig. 5. Crystal data of 1–3 (CIFs) are given in Supplementary information and have been deposited with the Cambridge Crystallographic Data Centre (CCDC nos. 2558767–2558769). Crystallographic data for 1 C20H18N4O2, M = 346.38 g mol−1, monoclinic, space group P21/c (no. 14), a = 11.9730(12) Å, b = 8.0702(8) Å, c = 18.2934(17) Å, α = 90°, β = 96.589(2)°, γ = 90°. V = 1755.9(3) Å3, Z = 4, ρcalcd = 1.310 g cm−3, F(000) = 720, reflections collected 25 188, independent reflections: 5369 [Rint = 0.0664, Rsig = 0.0597]. Final R1 = 0.0545, wR2 = 0.1183 for 2564 reflections with I > 2σ(I). Crystallographic data for 2 C20H20N4O2, M = 348.40 g mol−1, monoclinic, space group P21/c (no. 14), a = 10.9915(4) Å, b = 7.5690(3) Å, c = 10.4565(4) Å, α = 90°, β = 100.752(2)°, γ = 90°. V = 854.65(6) Å3, Z = 2, ρcalcd = 1.354 g cm−3, F(000) = 368, reflections collected 13 011, independent reflections: 1547 [Rint = 0.0632, Rsig = 0.0363]. Final R1 = 0.0536, wR2 = 0.1412 for 1381 reflections with I > 2σ(I). Crystallographic data for 3 C20H20N4O2·H2O, M = 366.41 g mol−1, triclinic, space group P1̄ (no. 2), a = 7.0767(1) Å, b = 11.9346(1) Å, c = 12.1333(1) Å, α = 105.166(1)°, β = 90.868(1)°, γ = 103.951(1)°. V = 956.53(2) Å3, Z = 2, ρcalcd = 1.272 g cm−3, F(000) = 388, reflections collected 19 972, independent reflections: 3471 [Rint = 0.0481, Rsig = 0.0342]. Final R1 = 0.0482, wR2 = 0.1308 for 3009 reflections with I > 2σ(I). Streptotryptophin A (1) Yellow powder; 1H and 13C NMR data in DMSO-d6 are reported in Table 1. HRESIMS m/z 347.1503 [M + H]+ (calcd for C20H19N4O2, 347.15025). Trans-streptotryptophin B (2) White powder; 1H and 13C NMR data in DMSO-d6 are reported in Table 1. HRESIMS m/z 349.1654 [M + H]+ (calcd for C20H21N4O2, 349.16590). Cis-streptotryptophin B (3) White powder; 1H and 13C NMR data in DMSO-d6 are reported in Table 1. HRESIMS m/z 349.1654 [M + H]+ (calcd for C20H21N4O2, 349.16590). 2,3-Dihydro-2,2-dimethylquinazolin-4(1H)-one (4) 31 Colorless crystalline solid; 1H NMR (400 MHz, methanol-d4): δH 7.68 (1H, dd, J = 7.8, 1.6 Hz), 7.27 (1H, td, J = 7.7, 1.6 Hz), 6.70 (1H, m, overlapped), 6.67 (1H, m, overlapped), 1.48 (6H, s); 13C NMR (100 MHz, methanol-d4): δC 166.8, 148.8, 135.2, 128.6, 118.5, 115.8, 114.8, 68.5, 29.1. HRESIMS m/z 177.1022 [M + H]+ (calcd for C10H13N2O, 177.10224). Nocardamine (5) 32 White solid; 1H NMR (400 MHz, DMSO-d6): δH 9.62 (3H, br s), 7.73 (3H, br t, J = 5.6 Hz), 3.46 (6H, t, J = 6.7 Hz), 2.99 (6H, q, J = 6.4 Hz), 2.58 (6H, t, J = 6.8 Hz), 2.28 (6H, t, J = 6.9 Hz), 1.49 (6H, p, J = 6.9 Hz), 1.37 (6H, p, J = 7.1 Hz), 1.19 (6H, p, J = 7.2 Hz); 13C NMR (100 MHz, DMSO-d6): δC 172.1, 171.6, 46.9, 38.4, 30.0, 28.6, 27.6, 25.9, 23.2. HRESIMS m/z 601.3557 [M + H]+ (calcd for C27H49N6O9, 601.35555). Geldanamycin (6) 33 Yellow solid; 1H NMR (400 MHz, CDCl3): δH 8.71 (1H, s), 7.29 (1H, s), 6.94 (1H, d, J = 11.6 Hz), 6.57 (1H, t, J = 11.4 Hz), 5.84 (1H, m), 5.19 (1H, s), 4.81 (2H, br s), 4.32 (1H, d, J = 9.3 Hz), 4.12 (3H, s), 3.53 (1H, m), 3.39 (1H, m), 3.36 (3H, s), 3.30 (3H, s), 3.03 (1H, d, J = 6.8 Hz), 2.78 (1H, m), 2.46 (2H, m), 2.03 (3H, s), 1.85–1.64 (6H, overlapped), 0.97 (6H, overlapped); 13C NMR (100 MHz, CDCl3): δC 185.0, 184.2, 168.3, 157.0, 156.0, 138.1, 136.4, 134.8, 133.3, 133.1, 127.6, 127.2, 126.3, 111.8, 81.7, 81.3, 81.0, 72.7, 61.7, 57.3, 56.7, 34.7, 32.8, 32.2, 27.9, 23.0, 12.9, 12.5, 12.4. HRESIMS m/z 583.2626 [M + Na]+ (calcd for C29H40N2O9Na, 583.26260). Author contributions T. P.: investigation, formal analysis, writing – original draft; M. A. F.: investigation; S. S.: formal analysis; T. A.: investigation, formal analysis, writing – original draft; N. L.: investigation, formal analysis; N. C.: investigation, formal analysis; S. T.: formal analysis; Y. Y.: formal analysis; G. L.: formal analysis; C. M.: supervision; S. R.: supervision; P. K.: conceptualization, funding acquisition, writing – original draft, writing – review & editing. Conflicts of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Supplementary Material