For the creation of sigma-factor mutant strains, Escherichia coli DH5αMCR (Grant et al., 1990) was used as the cloning host, E. coli ET12567 (pUZ8002) was used as the donor strain for conjugation into Actinoplanes sp. SE50/110, and E. coli ER2925 was used as the methylase-free plasmid replicant strain for transformation into Actinoplanes sp. SE50/110.

Actinoplanes strains were grown on a solid soy flour medium (SFM: 20 g L⁻¹ soy flour, 20 g L⁻¹ mannitol, and 20 g L⁻¹ agar, pH 8 adjusted with NaOH and tap water). For the pre-culture, 30 mL NBS (11 g ·L⁻¹ glucose × 1H₂O, 4 g ·L⁻¹ peptone, 4 g ·L⁻¹ yeast extract, 1 g ·L⁻¹ MgSO₄·7H₂O, 2 g ·L⁻¹ KH₂PO₄, and 4 g ·L⁻¹ K₂HPO₄) was inoculated in a plate and cultivated for 72 h at 28°C and 140 rpm in shake flasks. Cultivation for dereplication was performed using 350-mL baffled cell culture flasks for 70 mL (wild-type) or 50 mL (mutants) culture at 28°C and 140 rpm in a maltose minimal medium [(1) 72.06 g ·L⁻¹ maltose·1H₂O, (2) 5 g ·L⁻¹ (NH₄)2SO₄, 0.184 g ·L⁻¹ FeCl₃·4H₂O, 5.7 g ·L⁻¹ Na₃C₆H₅O₇·2H₂O, 1 g ·L⁻¹, 200 µL trace elements (stock concentration: 1 μM CuCl₂, 50 μM ZnCl₂, and 7.5 μM MnCl₂), (3) MgCl₂.6H₂O, 2 g ·L⁻¹ CaCl₂·2H₂O, and (4) 5 g ·L⁻¹ each K ₂HPO₄ and KH₂PO₄]. Inoculation was performed using 1 mL pre-culture, which was harvested by centrifugation at 5,000 rpm for 5 min and washed twice with the maltose minimal medium. Cells were dissolved in 1 mL maltose minimal medium, of which 200 µL was used to inoculate 80 mL maltose minimal medium. The cultivation was performed for 7 days. For the extraction of compounds of interest, 5.5 L bacterial cell culture was cultivated in 250-mL baffled cell culture flasks (Corning Inc., Corning, NY, United States) at 28°C and 140 rpm for 7 days.

General procedures: pH₂O was produced using the in-house Milli-Q system.

For UPLC-MS2 analysis and dereplication, 70 mL (wild-type) and 50 mL (mutants) cultivation broths were frozen in separate glass beakers at −80°C for 20 min, before freeze-drying for 2 days. The dried cultures were then wetted with pH₂O and extracted with 70 mL MeOH (v/v; HiPerSolv, VWR, Radnor, Penns., United States). After stirring for 4 h using a magnetic stirrer, the extraction broth was vacuum-filtered using a Büchner funnel and filter paper (Whatman No. 3, 6 µm pore size). To exclude medium components in the analysis, the extraction process was also performed for the medium. The extracts were then stored at −20°C until dry under pressure and 40°C.

For isolation of compounds, Actinoplanes sp. SE50/110 was cultivated in 5.5 L maltose minimal medium for 7 days with 50 mL in each 250-mL baffled cell culture flasks (Corning Inc., Corning, NY, United States) and frozen at −20°C until extraction. Therefore, Diaion HP-20 resin (Supelco, Bellefonte, PA, United States) was activated by incubation with methanol for 30 min and subsequent washing with pH₂O for 15 min, before the addition of 60 g·L⁻¹ activated resin to 1.4 L defrosted bacterial culture and further incubated for 4 days at room temperature and under movement. The beads were harvested by vacuum filtration through a cheesecloth mesh (X001LCT5PF, Dueco, Wolfsburg, Germany) using a Büchner funnel and washed with 150 mL pH₂O. The molecules extracted from the resin were eluted two times in 200 mL 100% methanol for 30 min under shaking. Vacuum filtration was used to separate the resin from the methanol and a final filtration through a filter paper (Whatman No. 3, 6 µm pore size, Sigma-Aldrich). The crude extract was dried under pressure at 40°C, and 11.558 g of the dried extract was received and stored at −20°C until further use.

For fractionation of the extract, FLASH chromatography (Biotage SP4TM system, Uppsala, SE) was performed using a water–methanol–acetone step gradient with Diaion HP20-SS resin (Supelco, Bellefonte, PA, United States) as the stationary phase for the wild-type extract, pSETT4-0284 extract, and the media control. Then, 1.5 g extract was re-dissolved in 10 mL 90% MeOH (v/v), and 1.5 g resin was added. The mixture was dried under pressure at 40°C before adding to the column of pre-loaded 4 g activated resin in 5% MeOH in a flash cartridge (Biotage, Uppsala, Sweden). The extract was separated into 15 fractions under a pH₂O:MeOH gradient of 5%–100% MeOH (v/v; HiPerSolv, VWR, Radnor, Penns., United States) over 32 min with a flow rate of 12 mL ·min⁻¹ and in further three fractions with an MeOH:acetone gradient to 100% acetone for 18 min and a flow rate of 12 mL ·min⁻¹. The eight collected fractions were then dried under pressure at 45°C (Buchi Syncore^® Polyvap, Flawil, Switzerland), afterward solved in DMSO at 10–80 mg ·mL⁻¹, and stored at −20°C until further use.

The extracts were diluted to 1:3 (v/v) with 80% MeOH (v/v) for analysis using UHPLC-QToF-MS. The UHPLC system (Acquity I-class, Waters, Milford, MA, United States) was coupled to a PDA detector (Waters, Milford, MA, United States) and Vion IMS QToF (Waters, Milford, MA, United States). The Acquity UPLC BEH C18 column (1.7 µm, 2.1 × 1.5 mm) (Waters, Milford, MA, United States) was used for the separation over an 10%–100% gradient of acetonitrile (HiPerSolv, VWR, Radnor, Penns., United States) with pH₂O, both supplemented with 0.1% formic acid (v/v; 33015; Sigma-Aldrich), in 12 min with a flow rate of 0.2 mL ·min⁻¹. A 5-µL sample was injected at an infusion rate of 15 μL ·min⁻¹, and mass spectrometry was run with a column temperature of 40°C with positive electrospray ionization. Leucine-enkephalin (100 pg ·μL⁻¹) was used as an internal standard. As controls, the medium fractions and MeOH were analyzed as well.

Preparative reversed-phase HPLC was used for the isolation of compounds from flash fractions. The collection of the respective fractions for the first and second rounds of purification was triggered by the recorded mass signal throughout the chromatographic separation. The HPLC system used consisted of a Waters 600 HPLC -pump with an additional degasser and flow-splitter (all Waters), a Waters 515 HPLC pump functioning as a “make-up” pump, a Waters 2996 photoarray detector, a Waters 3100 Mass detector, and a Waters 2767 sample manager. For system control, MassLynx V4.1 (Waters) software was used. As stationary phases, a SunFire RP-18 preparative column (10 μm, 10 mm × 250 mm) and XSelect CSH preparative Fluoro-Phenyl column (5 μm, 10 mm × 250 mm) (both Waters) were used. As mobile phases A [pH₂O with 0.1% (v/v) formic acid] and B [acetonitrile with 0.1% (v/v) formic acid], different gradients at a flow rate of 6 mL·min⁻¹ were used for separation. Acetonitrile (Prepsolv^®, Merck KGaA, Darmstadt, Germany) and formic acid (33015, Sigma-Aldrich) were purchased in appropriate quality, and ddH₂O was produced using the in-house Milli-Q^® system. For the mass-triggered fractionation, 1% of the eluent was split and blended with 80% MeOH in pH₂O (v/v) acidified with 0.2% formic acid (Sigma-Aldrich) and directed to the ESI-quadrupole-MS. The fractions were collected by mass-triggered fraction collection controlled by MassLynx, and the respective fractions were reduced to dryness under reduced pressure and by vacuum centrifugation, both at 40°C.

The structures of thiazostatin D/E were elucidated by 1D and 2D NMR spectroscopy. NMR spectra were acquired in DMSO-d6 at 298 K, in 3-mm solvent-matched Shigemi tubes, on a Bruker Avance III HD spectrometer (Bruker, Billerica, MA, United States) at 600 MHz, equipped with an inverse TCI cryo-probe enhanced for 1H, 13C, and 2H. Spectroscopy was performed under standard pulse programs for proton, carbon, HSQC, HMBC, COSY, and ROESY.

Genetic engineering was planned, and oligonucleotides were designed using SnapGene 4.3 (GSL Biotech LLC, Chicago, IL, United States). AntiSMASH 7.0 (Blin et al., 2023) was used for comparative genome studies of secondary metabolite clusters, and UNIFI 1.9.4 Scientific Information System Software (Waters) was used for data processing and computational prediction of elemental composition. PubChem and ChemSpider were used for compound identification. Compound masses and isotopic distributions were calculated using ChemCalc (Patiny and Borel, 2013), and Collision Cross Section was determined using CCSbase (Ross et al., 2020). Visualization was created using BioRender.com.

Since Actinobacteria are known for their large genomic potential for natural product biosynthesis, the Actinoplanes sp. SE50/110 genome was investigated for putative secondary metabolite biosynthesis gene clusters. This offers the advantage to expand our knowledge not only on strongly produced metabolites but also on low expressed or silent secondary metabolite gene clusters of already established strains. Prior analysis showed 20 suspected BGCs of Actinoplanes sp. SE50/110, of which 3 metabolites were previously known: acarbose, eumelanin, and an unidentified carotenoid (Wolf et al., 2017). Recent analysis using antiSMASH 7.0 identified 19 putative secondary metabolite biosynthetic gene clusters.

The recent analysis also added redefined product classes, with NAPAA (non-alpha poly-amino acids like ε-Poly-L-lysine), LAP (linear azol (in)e-containing peptides), RRE-containing, and RiPP-like clusters. Two putative metabolite gene clusters were identified with a similarity of 86% and 100% to biosynthetic clusters of known metabolites. These clusters contain genes for the biosynthesis of NRPS, T1PKS, NRP-metallophore-type, and NAPPA-type molecules, similar to the watasemycin biosynthesis of Streptomyces venezuelae and the ε-Poly-L-lysine biosynthesis of Epichloe festucae. A low similarity of <80% was determined for most coded putative gene clusters to known metabolite clusters (Table 1), which underlines the potential for the identification of novel molecules Actinoplanes sp. SE50/110.

For the analysis of Actinoplanes sp. SE50/110 MeOH whole-cell extract, the focus was set on dominant peaks indicating a production in sufficient amounts for further compound isolation and structure elucidation. It also sets the focus on compounds which are produced in resource competition to the acarbose biosynthesis, and further knowledge about the metabolites can be beneficial for future strain engineering approaches. Several dominant compounds were found, and their elemental composition was computationally calculated using UNIFI 1.9.4 (Figure 1). It revealed seven sulfur-containing compounds (Table 2). These compounds were mostly present in fraction 5 but were also found in fractions 6 and 7 and partly in fractions 4 and 8. A molecular mass difference of 14 kDa and similar elemental compositions of compounds 1–3 (1: m/z 353.0988 [M + H]⁺ with C₁₆H₂₀N₂O₃S₂; 2: m/z 339.0836 [M + H]⁺ with C₁₅H₁₈N₂O₃S₂; 3: m/z 325.0656 [M + H]⁺ with C₁₄H₁₆N₂O₃S₂), compounds 4 and 5 (4: m/z 224.0743 [M + H]⁺ with C₁₁H₁₃N O₂S and 5: m/z 210.0583 [M + H]⁺ with C₁₀H₁₁N O₂S), and compounds 6 and 7 (6: m/z 369.0929 [M + H]⁺ with C₁₆H₂₀N₂O₄S₂ and 7: m/z 355.0779 [M + H]⁺ with C₁₅H₁₈N₂O₄S₂) leads to the assumption of a putative biosynthetic connection of compounds 1–7.

Further detected promising compounds are compounds 8 and 9 with m/z 440.2768 [M + H]⁺ and m/z 454.2917 [M + H]⁺, for which the elemental compositions C₂₃H₃₃N₇O₂ and C₂₄H₃₅N₇O₂ were predicted. Despite comprehensive literature and database research, no similar molecules could be determined for both compounds, which are suspected as putative novel molecules. Unfortunately, their low-intensity peaks present a challenge, indicating limited amounts in the already concentrated fractions and even less in the cell extract. This poses a critical challenge for its isolation and further analysis as substantial quantities are required for structure elucidation. That is why no isolation and identification attempts were made.

Comprehensive literature and database research of the elemental composition and fragmentation pattern of compound 1 showed that it corresponds with watasemycin. Watasemycin (C₁₆H₂₀N₂O₃S₂) is a natural product of Streptomyces sp. TP-A0597, of which two isoforms are known so far (Sasaki et al., 2002). It is a natural product of S. venezuelae, where the watasemycin biosynthesis pathway has four further products: thiazostatin (C₁₅H₁₈N₂O₃S₂), isopyochelin (C₁₄H₁₆N₂O₃S₂), pulicatin (C₁₀H₁₁NO₂S), and aerugine (C₁₀H₁₁NO₂S) (Inahashi et al., 2017). Thiazostatin has antioxidant activity (Shindo et al., 1989), pyochelin and pulicatin are siderophores with iron-chelating activity (Cox et al., 1981; Lin et al., 2010), and aerugine has antifungal and anti-oomycete activity (Lee et al., 2003). Interestingly, these known products have an identical elemental composition to the predicted composition of compounds 2–5. Since the genomic analysis determined a cluster similarity of 86% of cluster 15 and the S. venezuelae watasemycin biosynthesis gene cluster, compounds 1–5 are suspected as watasemycin, thiazostatin, isopyochelin, pulicatin, and aerugine of a putative similar watasemycin biosynthesis pathway in Actinoplanes sp. SE50/110. Analysis of a watasemycin reference enabled the confirmation of this assumption. When compared to the extract fraction dataset, the synthetic watasemycin eluted at a slightly earlier retention time but the extracted mass chromatogram of an analyzed mixed sample confirmed that compound 1 is watasemycin (Figure 2). Consequently, compound 2 is assigned as thiazostatin, compound 3 as isopyochelin, compound 4 as pulicatin, and compound 5 as aerugine. This adds five further identified products to two already known products of Actinoplanes sp. SE50/110.

The biosynthesis of watasemycin, thiazostatin, isopyochelin, pulicatin, and aerugine was investigated in S. venezuelae in 2017 (Inahashi et al., 2017). Nine genes encode biosynthesis proteins, which are essential for the formation of 5 different products, and 15 genes are part of the BGC comprising sven0503 to sven0517. The five products are formed from chorismic acid by several reactions involving several of the so-called Pch proteins.

In addition to S. venezuelae, aerugine and pyochelin are also naturally produced by Streptomyces scabies 87.22. However, watasemycin, thiazostatin, pulicatin, and isopyochelin are not produced by the strain (Liu J. et al., 2021). A cluster comparison of these three strains revealed a conserved genomic region of biosynthesis genes (Figure 3). S. scabies carries 13 genes (scab1361–1481) that are associated with aerugine and pyochelin production. Except for scab_1391, Actinoplanes sp. SE50/110 encodes all homologous genes of all essential S. scabies biosynthesis genes within BGC 15. Since S. scabies produces two of five products from a similar pathway in S. venezuelae, the smaller BGC is no surprise. S. scabiei misses sven0504 and sven0515 homologs, coding for Na⁺ H⁺ antiporter (PF00999) and an SAM methylase protein, which is essential for methylation of thiazostatin to watasemycin in S. venezuelae. Interestingly, S. scabiei does not carry any for the thiazoline-reducing pchK homologs. However, putative oxidoreductase coding scab_1461 is homologous to thiazolinyl reductase pchG, which performs the reducing step in the pyochelin biosynthesis of the first known pyochelin producer Pseudomonas aeruginosa (Reimmann et al., 2001). Thus, it is possible that S. scabiei pchG (scab_1461) takes over the essential function of pchK. S. venezuelae carries two pchK genes, sven0508 and sven0516, whereas only sven0516 is essential for thiazoline reduction catalysis within the biosynthesis (Inahashi et al., 2017). Actinoplanes sp. SE50/110 only carries the homolog (ACSP50_6120) of the functional sven0516. When the gene cluster is compared to S. venezuelae, it encodes homologs of all biosynthesis-associated genes, except a sven0503, which codes for an AMP-binding domain (Pfam:PF00501) containing protein of an unknown function.

The gene cluster is conserved among these three strains. However, Actinoplanes sp. SE50/110 homologs of essential biosynthesis genes pchF (ACSP50_6119), pchK (ACSP50_6120), and transcriptional regulator (ACSP50_6121) are coded in greater distance to all other suspected biosynthesis genes, with several genes in between. Regarding a high structural similarity of compounds 6 and 7 to the five known pathway products, these cluster arrangements could indicate an extended biosynthesis pathway. The genes, which are coded in between the putative gene cluster, code for polyketide synthases (PKSs) (ACSP50_6127, ACSP50_6128, ACSP50_6129), flavin reductase (NADP(H)) (ACSP50_6125), tetracycline 7-halogenase (ACSP50_6130), and Na+/H+ antiporter (ACSP50_6131, sven0504) proteins. These genes belong to a subcluster comprising ACSP50_6125–ACSP50_6132, with 50% similarity the neocarzilin A/B biosynthetic gene cluster (BGC0000111) of Streptomyces carzinostaticus.

Actinoplanes sp. SE50/110 carries with BGC 15 (Table 1) a cluster that has high similarity to the BGC responsible for the biosynthesis of the mentioned molecules in S. scabies and S. venezuelae. Nevertheless, the cluster has differences to the other BGCs, with the S. venezuelae pchK homolog and S. scabies pchG homolog missing. It was further shown that the BGC 15 contains further putative biosynthetic genes that are in between the otherwise compact homologous cluster. These genetic variations of the BGC suggest possible extensions or modifications in the biosynthetic pathway in Actinoplanes sp. SE50/110, where additional genes for polyketide synthases and other enzymes may contribute to producing structurally related compounds. Despite the high similarity of the BGC 15 to the watasemycin BGC of S. venezuelae and pyochelin BGC of S. scabies, it must be further investigated to confirm whether BGC 15 is responsible for the production of the identified molecules in Actinoplanes sp. SE50/110.

Beside the five identified products of the watasemycin biosynthesis, two similar compounds were detected: compound 6 with m/z 369.0929 [M + H]⁺ and compound 7 with m/z 355.0779 [M + H]⁺. Calculated elemental composition is C₁₆H₂₀N₂O₄S₂ for compound 6 and C₁₅H₁₈N₂O₄S₂ for compound 7. Since the molecular mass and elemental composition are similar to those of the prior mentioned molecules, an association to the watasemycin biosynthesis was suspected.

The compound was isolated as described above using the [M + H]⁺ signal m/z 35,518 (low-resolution MS). Within the second isolation round, compound 7 eluted in an overlapping peak. Longer gradients did not lead to a better separation of the two species of the compound. The fractions were pooled and reduced to dryness with the weight of 1.0 mg.

Although compound 7 was repeatedly isolated as a single peak by mass-guided preparative HPLC, the results of analyzing the NMR spectra of 7 showed that it consisted of two structurally closely related compounds. This was apparent as there were double signals for most of the carbon and hydrogen atoms. There was no indication of rotamers. Due to the similarity to the thiazostatins, which did not produce rotamers but thus existed as two stereoisomers in position 2″ (thiazostatin A and B), it was hypothesized that the two compounds are stereoisomers. Structure elucidation was conducted using the mixed sample, as the two constituents, now termed 7a and 7b, could not be separated. 1D (1H and 13C; Table 3; Supplementary Figures S9, S10) and 2D (HSQC, HMBC, COSY, ROESY, and H2BC; Figure 3B; Supplementary Figures S11–S15) data of the 7a/7b mixture confirmed their structural similarity to the previously reported compounds in the thiazostatin and watasemycin families (Supplementary Figure S8). The connectivities of the atoms of the B and C rings were determined through HMBC and COSY signals (Figures 1B, Supplementary Figure S4, S8) and by comparing NMR data with those recorded for the previously recorded thiazostatin and watasemycin variants (Shindo et al., 1989). This leaves a C₆H₅O₂ fragment to be assigned, indicative of the presence of a benzene ring substituted with two hydroxy groups. The signals of the benzene ring were indeed detected. A clear HMBC was visible from an aromatic carbon signal (H-6, δH 6.82–6.75 in both 7a and 7b) to the C-2’ (δC 170.5 in 7a, δC 171.0 in 7b) carbon atom, linking the benzene fragment to the rest of the molecule. In addition to CH-6 (δC 114.7 and 114.9), two distinct signals for aromatic CH-groups (CH-3, δC 117.4 in both 7a and 7b, CH-4, δC 121.0 in 7a, and δC 121.1 in 7b), a fully substituted aromatic carbon atom (C⁻¹, δC 115.3 in both 7a and 7b), and two distinct signals for hydroxylated aromatic carbon atoms (C-2, δC 151.3 and C-5, δC 149.5 in both 7a and 7b, Supplementary Figure S12) were visible, confirming the presence of a dihydroxybenzene fragment. The coupling patterns of the benzylic protons (Supplementary Figure S9) revealed the substitution pattern to be 2-OH and 5-OH based on that the assigned proton atom H-6 was a double with a small coupling, characteristic for an isolated proton with a weak meta-coupling, in this case to proton atom H-4. Proton atom H-4 was a doublet of a doublet with one strong 8 Hz ortho-coupling to proton atom H-3 and one weak meta-coupling to proton atom H-6. The lack of a methylation in the C-5 position makes the compounds more similar to the thiazostatins rather than the C-5 methylated watasemycins. The compounds were consequently given the names thiazostatin D (7a) and thiazostatin E (7b) since Liu et al. (2022) published a thiazostatin-like compound bearing one hydroxyl group at ring A in position 2 and O-methylation of the carboxylic acid on ring C. The difference between 7a and 7b is attributed to different configurations in the C-2″ position, as for thiazostatins A and B (Figure 4).

The isolated compounds were a mixture of the two closely related thiazostatins D and E, and since a further separation of the two isomers was not possible, the two compounds were tested for potential bioactivities. For the MIC assay, we selected S. aureus and E. coli as Gram-positive and Gram-negative pathogens, respectively, as well as the Gram-positive S. agalactiae since the latter one has shown the highest sensitivity when screening extracts, extract fractions, and compounds for anti-microbial activities in our experience (Kristoffersen et al., 2018; Schneider et al., 2019; Schneider et al., 2020). We have tested the compound mixture for potential bioactivity against the three bacteria at 100 μg·mL⁻¹ test concentration and did not observe an anti-microbial effect. For the anti-cancer assay, we have initially tested the compound mixture at 100 μg·mL⁻¹ against the three cell lines MCF7, MRC5, and A2058, where A2058 has shown 53% survival, MCF7 has shown 15% survival, and MRC5 has shown 101% survival; since the immortalized lung-fibroblast cell line MRC5 was seemingly not affected by the compounds and the MCF7 (breast cancer) cell line has shown just 15% survival, we repeated the anti-proliferative assay using five cell lines (the three initially tested cell lines plus Molm13 and MDA-MB-468) and concentrations of 50 μg·mL⁻¹ and 10 μg·mL⁻¹, the limitation to five cell lines and two concentrations was due to the available compound quantity. The results are given in Figure 5.

To investigate the influence of Actinoplanes sp. SE50/110 σ factor and σ factor-related genes on metabolite production, alternative σ factor encoding sigH and anti-anti-σ factor encoding ACSP50_0284 (0284) were manipulated as examples. Schlüter et al. showed that sigH overexpression enhanced acarbose production Actinoplanes sp. SE50/110, making it an interesting gene target to investigate its influence on the production of further metabolites (Schlüter et al., 2024). ACSP50_0284 encodes a sulfate transporter and anti-sigma-factor antagonist (STAS) domain protein (Kormanec et al., 2016; Mingyar et al., 2014; Sevcikova et al., 2010; Sevcikova et al., 2021). Gene deletion and expression strains were created for both genes, and the production of the different compounds was investigated since preliminary data indicated a putative regulatory influence of these genes on the production of further metabolites.

SigH expression resulted in an increased peak intensity of watasemycin (1) and putative watasemycin-related compound 6 of m/z 369.0929 [M + H]⁺ when compared to the wildtype (Figure 6). While the peak intensity of pulicatin (4) was decreased, the peak intensity of the highly similar aerugine (5) was slightly increased. The previous sections identified that watasemycin pathway-related thiazostatin D (7a) and thiazostatin E (7b) could not be detected in the sample, indicating the prevention of its production due to the gene expression. Surprisingly, the sigH deletion strain (ΔsigH) did not show the complete opposite effect of gene overexpression but led to a decrease in the watasemycin peak intensity, without further influence on pulicatin, aerugine, and thiazostatin D/E peak intensity. The ΔsigH strain showed an opposite effect on the peak intensities of the 2-hydroxyphenylthiazoline family compounds. Under gene deletion, watasemycin (1) peak intensity is significantly reduced in relation to the other watasemycin pathway-derived compounds and the peak intensity of aerugine (5) and pulicatin (4) is increased. This indicates a putative regulatory influence of σH on this biosynthesis pathway. Since the peak intensity of thiazostatin D/E (7) is also affected, its origin from the same biosynthesis pathway is even more assumed.

Interestingly, ACSP50_0284 manipulation had a more severe effect on the watasemycin biosynthesis pathway and its putative products (Figure 7). While ACSP50_0284 deletion (Δ0284) resulted in a change in the peak intensity distribution of the 2-hydroxyphenylthiazoline family molecules with an increased peak intensity of thiazostatin (2) and reduced peak intensity of pulicatin (4), ACSP50_0284 expression prevented the production of all products. These were not detectable in the whole-cell extract of the ACSP50_0284 expression strain. Strikingly, we hereby observed the production of an unknown compound 10 of m/z 462.1643 [M + H]⁺ with calculated elemental composition C₁₉H₂₇NO₁₂. This highly polar compound was not detected in the Actinoplanes sp. SE50/110 wildtype and is exclusively produced by the ACSP50_0284 expression strain. Despite several purification attempts, sufficient purity of the isolate was not achieved and prevented structure elucidation.

The rendered product distribution of the identified compounds by sigma factor engineering is suspected to affect the strains’ siderophoric ability since aerugine and (iso)pyochelin have siderophore and iron-chelating ability (Inahashi et al., 2017) and thiazostatin and watasemycin are suspected of having siderophore activity according to the functional group-based prediction provided by SIDERITE (He et al., 2024). A colorimetric-based Chrome Azurol S (CAS) assay was performed with supernatant. The sigH expression and deletion strains had a similar siderophore activity than the wildtype (Figure 8). Its gene manipulation resulted in a shift of peak intensities between the pathway products, but no drastic overall change in siderophore activity, whereas the expression of ACPS50_0284 resulted in the complete production shutdown of the watasemycin pathway products and the significant reduction to 68% of siderophore activity within the Actinoplanes sp. SE50/110 supernatant. Interestingly, the deletion strain ∆0284 showed a slightly increased, but insignificant, siderophore activity of 117% wildtype activity.