Biocatalytic diversification of abietic acid in Streptomyces Authors: Caitlin A McCadden, Tyler A Alsup, Ion Ghiviriga, Jeffrey D Rudolf Published: 2025 ABSTRACT Abstract Biocatalysis provides access to synthetically challenging molecules and commercially and pharmaceutically relevant natural product analogs while adhering to principles of green chemistry. Cytochromes P450 (P450s) are among the most superlative and versatile oxidative enzymes found in nature and are desired regio- and stereoselective biocatalysts, particularly for structurally complex hydrocarbon skeletons. We used 10 genome-sequenced Streptomyces strains, selected based on their preponderance of P450s, to biotransform the bioactive diterpenoid abietic acid. We isolated and structurally characterized seven oxidized abietic acid derivatives from three different strains, including four products that are new bacterial biotransformants or enzymatic products. Oxidations (hydroxylation, dehydrogenation, and aromatization) were seen on both the B and C rings of abietic acid and five products had multiple modifications. Notable conversions observed in the study were that of abietic acid to 15-hydroxy-7-oxo-8,11,13-abietatrien-18-oic acid, 7, which involves multiple hydroxylation reactions and dehydrogenation. The findings from this study will lead to identifying P450s or other enzymes that may act as general biocatalysts to modify abietanes and other labdane-type diterpenoid skeletons.One-Sentence SummaryGenome-guided biotransformation of the bioactive diterpenoid abietic acid in Streptomyces yielded seven oxidized derivatives including four that have not been previously seen from bacteria. Introduction The identification of new bioactive scaffolds and the enzymes that make or modify them is of growing interest to pharmaceutical, bioremediation, and biotechnology companies. In the past three decades, there has been a paradigm shift in the chemical industry, which prioritizes the elimination of waste, utilization of raw materials, and avoidance of hazardous substances (Bell et al., 2021). Adhering to 10 of the 12 principles of green chemistry, biocatalysis is one of the most attractive fields for sustainability (Bell et al., 2021; Nett et al., 2009; Wei et al., 2018). Compared with conventional synthetic approaches, enzymatic reactions typically obviate the need for additional reagents for functional group activation, protection, and deprotection, resulting in more step economy in reaction schemes and generating less hazardous material waste (Sheldon et al., 2020). In addition, the regio- and stereoselectivity of enzymes have contributed to biocatalysis becoming more common in the synthesis of drugs and pharmaceutical intermediates (France et al., 2023; Romero et al., 2021). Biocatalysis is generally executed using in vitro or whole-cell strategies. Whole-cell biocatalysis may be preferred to increase the product profile due to the presence of multiple enzymes with native cofactor regeneration and increased protein stability (Li, Jiang, et al., 2020; Romero et al., 2021). With recent advances in genomic technologies, microbial enzymes have been at the forefront of new developments in biocatalysis. Cytochromes P450 (P450s) are a superfamily of heme-dependent oxidative enzymes, which functionalize a multitude of substrates via radical mechanisms (Rudolf et al., 2017). The diversity of the reactions they catalyze distinguishes P450s from most other enzymes and supports their use as “green” catalysts (Di Nardo & Gilardi, 2020; Hoebenreich et al., 2017; Jung et al., 2011; Urlacher et al., 2019; Wei et al., 2018). Streptomyces, encoding 20 or more P450s in each strain, possess a vast reservoir of these potential biocatalysts, even though the functional capabilities and natural roles of most of these enzymes are unknown (Rudolf et al., 2017). Despite this gap in knowledge, there is high potential for their application in the production of desirable compounds. For example, CYP105A3 from Streptomyces sp. Y-110 is currently implemented in the industrial bioconversion of compactin to pravastatin, a therapeutic treatment for hypercholesterolemia (Julsing et al., 2008). Using a one-compound-many-strains approach to biocatalysis, we sought to derivatize the bioactive diterpenoid abietic acid (AA) using biotransformation in Streptomyces. AA is best known as a natural product from conifer tree resin (France et al., 2023), although AA has been detected in cyanobacteria (Romero et al., 2021) and very recently an abietadiene synthase was identified in Streptomyces (Chen et al., 2024). AA and other abietane-type diterpenoids have a wide spectrum of biological properties including antimicrobial, antifungal, antitumor, antiviral, antiulcer, cardiovascular, antioxidant, and anti-inflammatory activities (Gáborová et al., 2022; González, 2015; Puente-Villegas et al., 2024; Yang et al., 2013). Materials and Methods Bacterial Strains and Chemicals Strains used in this study are listed in Supplementary Table S1. Abietic acid was purchased from Thermo Scientific. All other media components, chemicals, and solvents were purchased at high quality from various commercial vendors. General Experimental Procedures All 1D (1H and 13C) and 2D (1H-1H COSY, 1H-13C HSQC, and 1H-13C HMBC) NMR experiments were performed in CDCl3 on either a Bruker AVANCE III Ultrashield 600 or a Bruker AVANCE III HD 600 equipped with a 5 mm TXI CryoProbe. All NMR chemical shifts in CDCl3 were referenced to residual CHCl3 at 7.26 ppm for 1H or 77.00 ppm for 13C. For liquid chromatography-mass spectrometry (LC-MS) analysis, an Agilent 1290 infinity II single quadrupole LC/MSD equipped with a GL Sciences C18 column (150 mm × 4.6 mm, 5 µm) was used. Analytical high-performance liquid chromatography (HPLC) was performed on an Agilent 1260 Infinity II system equipped with an Agilent HC-C18(2) column (150 × 4.6 mm, 5 μm). For preparative reversed-phase chromatography, an Agilent 1260 Infinity LC system equipped with an Agilent Eclipse XDB-C18 column (250 mm × 21.2 mm, 7 µm) was used. Genome Mining for P450-Abundant Streptomyces Predicted bacterial P450s were identified through utilization of the UniProt, InterPro, and NCBI databases. Proteins belonging to the cytochrome P450 superfamily (InterPro Entry: IPR036396) were identified in Streptomyces. The preliminary search resulted in >26,800 proteins. Entries were cross-referenced with an in-house actinomycetes strain library and priority was given to strains with high-quality genome sequences containing high numbers of P450 genes. Terpene-related proteins were identified using Pfam or InterPro entries for cyclases (PF19086, IPR034686, IPR032696, IPR032697, and IPR044878), polyprenyl synthases (PF00348 and IPR000092), or prenyltransferases (PF00432). Small-Scale Screening and LC-MS Analysis For initial screening of Streptomyces strains (see Supplementary Table S1), each strain was grown on ISP2 or ISP4 solid medium. Spores were harvested for seed cultures after 10–14 days of growth and directly used to inoculate seed cultures. Seed cultures containing 10 mL of sterile tryptic soy broth (TSB) medium were grown for 48 hr at 28°C with shaking at 250 rpm. In triplicate, baffled flasks (250 mL) containing 50 mL of TSB and ∼30 glass beads (Ø = 6 mm) were inoculated with 4% (v/v) of the 48-hr-old seed cultures. Small-scale fermentations were maintained at 28°C with shaking at 250 rpm for 7–10 days. Abietic acid, dissolved in dimethyl sulfoxide (DMSO), was added on day 2 of the large-scale fermentation under sterile conditions to a final concentration of 20 mM. These fermentations were completed in parallel with triplicate control flasks adding the same volume of DMSO. From each flask, a 5-mL aliquot was taken, mixed with 5 mL of EtOAc with vigorous vortexing, and centrifuged at 3,900 rpm for 15 min. The organic layer was dried under air before being reconstituted in 500 µL of CH3OH; 10 µL of sample was analyzed by LC-MS. Liquid chromatography was carried out at 0.5 mL min−1 starting from 5% CH3CN in H2O for 5 min, followed by a 20-min solvent gradient from 5–98% CH3CN in H2O, and ending with a 15-min hold at 98% CH3CN in H2O. Mobile phases contained 0.1% (v/v) formic acid. Abietic acid-related products were detected using a combination of UV (in comparison to the spectra of AA) and detected m/z values (range of 300–375). Isolation and Structural Elucidation For isolation of biotransformed compounds, large-scale fermentations (12–36 L) were carried out in 2-L baffled flasks containing 500 mL of TSB and inoculated with 4% (v/v) of seed cultures prepared as described earlier. Abietic acid, dissolved in DMSO, was added on day 2 of the large-scale fermentation under sterile conditions to a final concentration of 20 mM. Fermentations were carried out for a total of 7–10 days at 28°C with shaking at 200 rpm. The broth and mycelia were then separated by centrifugation at 6,500 rpm for 30 min. The clarified culture broth was extracted with 1:1 culture broth: EtOAc three times and concentrated in vacuo to afford crude extract. The crude extract was adsorbed onto silica gel and fractionated via column chromatography over a gradient of hexanes–EtOAc (0:100–100:0). Individually, fractionated samples were subjected to preparative reversed-phase HPLC with a flow rate of 20 mL min−1 starting from 5% CH3CN in H2O for 5 min, followed by a 20-min solvent gradient from 5–98% CH3CN in H2O, and ending with a 15-min hold at 98% CH3CN in H2O. Fractions were analyzed for purity by analytical HPLC and all isolated compounds were analyzed via NMR spectroscopy. Isolable yields: 1 = 14.5 mg (2.4 mg L−1 from S. viridochromogenes); 1 = 12.8 mg (2.1 mg L−1 from S. melanosporofaciens); 2 = 7.6 mg (1.3 mg L−1 from S. ghanaensis); 2 = 3.2 mg (0.5 mg L−1 from S. melanosporofaciens); 3 = 6.5 mg (1.0 mg L−1 from S. viridochromogenes); 4 = 3.3 mg (0.6 mg L−1 from S. ghanaensis); 5 = 3.4 mg (0.6 mg L−1 from S. viridochromogenes); 6 = 4.2 mg (0.7 mg L−1 from S. viridochromogenes); and 7 = 2.4 mg (0.4 mg L−1 from S. viridochromogenes). Results and Discussion We first set out to identify Streptomyces hosts encoding large numbers of P450s for biotransformation. We hypothesized that P450s were the most likely culprits for the oxidation of AA as they are common in the biosynthesis of plant AA-derived natural products (Guo et al., 2013; Hamberger & Bohlmann, 2024; Helliwell et al., 2001; Ro & Bohlmann, 2006). P450s in Streptomyces were first identified by searching for any protein categorized with the P450 superfamily InterPro tag (IPR036396). We then selected 10 strains from our library of Streptomyces (see Supplementary Table S1), encompassing a total of 376 putative P450s, as well as countless other oxidative enzymes. Only two of these P450s have been previously experimentally characterized: CotB3 and CotB4 from the Streptomyces melanosporofaciens cyclooctatin biosynthetic gene cluster (BGC) (Kim et al., 2009). A sequence similarity network (SSN) (Gerlt et al., 2015) of these P450s, along with several characterized enzymes (Rudolf et al., 2017), highlights the sequence, and thus the expected functional, diversity of this small library (Fig. 1a). Additionally, many of these P450s are found co-clustered with other terpene BGCs, that is, any putative terpene-related gene (cyclases, prenyltransferases, and polyprenyl synthases) found within 10 open reading frames (ORFs) of the P450 (Fig. 1b). We hypothesized that Streptomyces P450s clustered near terpene biosynthetic genes may be more likely to accept AA as a substrate, given their possible predisposition for binding and catalyzing the oxidation of terpenoids. Genome mining to prioritize Streptomyces hosts for biotransformation. (a) A focused sequence similarity network of Streptomyces P450s showing all predicated P450s from the 10 Streptomyces strains (see legend for color code) used in this study and 121 selected P450s characterized in the literature (Rudolf et al., 2017). The sequence similarity network is shown at an E value cutoff = 10−83 (approximately equivalent to the CYP family threshold of 45% sequence identity over 400 residues). Characterized P450s that are known to oxidize terpene-related substrates are labeled. (b) Many P450s are found co-clustered with other putative terpene biosynthetic genes such as cyclases, polyprenyl synthases, or prenyltransferases. Ten P450s, including the CYP105A1 homolog D9XD69, from S. viridochromogenes are shown as a representative example. In the initial biotransformation screen, each of the 10 Streptomyces strains was grown in TSB medium containing 20 mM of AA. After 7–10 days, organic extracts of aliquots were analyzed by HPLC-UV and LC-MS analysis for disappearance of AA and the production of new metabolites with UV spectra reminiscent of AA and m/z values between 300 and 375 (see Supplementary Fig. S1). Three strains, S. viridochromogenes DSM 40736, S. ghanaensis ATCC 14672 (now referred to as S. viridosporus), and S. melanosporofaciens NRRL B-12234, successfully transformed AA into a series of seven products (Fig. 2). Each of the products was isolated using standard chromatographic techniques and their structures were determined based on 1D and 2D NMR experiments in comparison with data reported in the literature (Fig. 2, see Supplementary Tables S2–S8 and Supplementary Figs. S2–S36). Biotransformation of abietic acid in three Streptomyces strains. Seven abietic acid derivatives were isolated and structurally characterized. Absorbance shown at λ = 254 nm. The table summarizes the derivatives detected and their isolable yields and titers from the three strains. 15-Hydroxyabietic acid (1) was produced as the major product in the biotransformations of all three strains, S. viridochromogenes, S. melanosporofaciens, and S. ghanaensis biotransformations. This diterpenoid is known to have weak to moderate (2–50 μM) cytotoxic, anti-inflammatory, and antibacterial activities (Puente-Villegas et al., 2024). Additionally, 1 is a key intermediate in the semisynthesis of 15-hydroperoxyabietic acid, a major colophony allergen involved in common forms of allergic contact dermatitis that requires a time-consuming and expensive chemical process to synthesize for allergy tests (Haberkorn, 2001). Here, without optimization from 12 L, we produced 1 in both S. melanosporofaciens (12.8 mg isolated yield) and S. viridochromogenes (14.5 mg) using biotransformation. Previous investigations found that CYP105A1 from Streptomyces griseolus hydroxylates several substrates including AA and dehydroabietic acid at C15 (Janocha & Bernhardt, 2013; Janocha et al., 2013). Sequence analysis (Fig. 1, see Supplementary Table S1) of all predicted P450s in the genomes of these 10 strains revealed that seven strains, including the three AA metabolizing strains, have homologs of CYP105A1 (42–58% sequence identities); the three AA metabolizing strains have four homologs (46–58% identities). Given the known activity of CYP105A1 and that S. ghanaensis, S. viridochromogenes, and S. melanosporofaciens each have homologs, they are possible candidates for the oxidation of AA to 1. One of these homologs in S. viridochromogenes, D9XD69 (55.7% ID to CYP105A1), is clustered near a gene (eight ORFs away) encoding a characterized 7-epi-α-eudesmol synthase (Rabe & Dickschat, 2013). The methoxy derivative of 1, 15-methoxyabietic acid (2) was also found as a major product in S. ghanaensis and a minor product in S. melanosporofaciens. First isolated from pollen grains of the Himalayan cedar tree, 2 was recently shown to inhibit protein tyrosine phosphatase 1B (PTP1B) with an IC50 value of 6.8 μM (Hjortness et al., 2018; Ohmoto et al., 1987). Inhibitors of PTP1B are sought after for their potential in diabetes and cancer treatments (Coronell-Tovar et al., 2024; Zhao et al., 2025). The methyl group may have been attached by an endogenous methyltransferase present in S. melanosporofaciens and S. ghanaensis, but absent in S. viridochromogenes. 12β-Hydroxyabietic acid (3) and 12-oxoabietic acid (4) were seen in all three strains, with 3 being a significant product in S. viridochromogenes. Biotransformation product 3 has been shown to have biological activities similar to those seen for 1 (Wang et al., 2016); 4 has weak (IC50 = ∼40 μM) cytotoxicity (Prinz et al., 2002). CYP106A2 from Bacillus megaterium is known to hydroxylate AA at C12, interestingly in a nonstereospecific manner (Bleif et al., 2011). Given that CYP106A2 is from B. megaterium (now referred to as Priestia megaterium), there are no clear CYP family (>45% sequence identity) homologs in any of these three Streptomyces strains (Fig. 1), making direct comparison difficult. In S. ghanaensis, the C12 and C15 oxidizing enzymes work in tandem to produce 15-hydroxy-12-oxoabietic acid (5). 15-Hydroxydehydroabietic acid (6) was also produced in all three biotransformations, most significantly in S. viridochromogenes. This well-studied compound has been shown to have weak activities (IC50 >30 μM) in numerous bioassays (Baosong et al., 2020; Gao et al., 2018; Lekphrom et al., 2010; Wu et al., 2016; Yan et al., 2019; Yang et al., 2010). There are several possibilities for the formation of 6. Originating from the same presumed 12,15-diol precursor to 5, a competing nonenzymatic dehydration at C12 would yield 6. Similarly, 3 may dehydrate to dehydroabietic acid, which then undergoes C15 hydroxylation by an enzyme such as CYP105A1 (Janocha et al. 2013), as described earlier. Synthetically, aromatization of AA derivatives often requires nonrenewable, precious metal catalysts such as palladium or iridium for the dehydrogenation reaction (Buhl et al., 1998; Esteves et al., 2001; Gandeepan et al., 2019; Li, Ali, et al., 2020; Littmann, 1938). Thus, an enzymatic or whole-cell alternative provides a green alternative to this transformation. The most functionalized AA derivative isolated in this study, with modifications to both the B and C rings of AA, was 15-hydroxy-7-oxo-deabietic acid (7, 15-hydroxy-7-oxo-8,11,13-abietatrien-18-oic acid). Seen in all three biotransformations but most significantly in S. viridochromogenes, oxidation to the ketone at C7 likely results after aromatization to 6; this is compatible with the highest product of 6 in S. viridochromogenes (Fig. 2). Although we propose that these oxidations are P450-catalyzed, other enzymes such as α-ketoglutarate-dependent dioxygenases and dehydrogenases may be responsible for these reactions, as seen for the sequential C7 hydroxylation and oxidation of a similar diterpene skeleton in the biosynthesis of platensimycin and platencin (Dong et al., 2019). Diterpenoid 7 has been widely investigated for its impressive biological applications including diabetes, obesity, and other metabolic diseases (Baosong et al., 2020; Puente-Villegas et al., 2024; Wang et al., 2013; Yang et al., 2010). It was proposed to be a competitive inhibitor of human PTP1B with an IC50 value of 10.3 μM (Hu et al., 2016), and displayed strong antagonist activity (IC50 = 1.3 μM) for the Farnesoid X receptor, a transcriptional regulator involved in diabetes and cholestasis (Diao et al., 2018). Biocatalysis is considered a fundamental pillar of catalysis, especially in the context of green chemistry and sustainable processes. Here, we used whole-cell biotransformation in Streptomyces to biosynthetically diversify the commercially important diterpenoid AA. We prioritized 10 bacterial hosts based on their abundance of P450s, oxidative enzymes known to modify terpenoids. Three strains converted AA into seven products with oxidations at three distinct positions. All seven products are known natural products from coniferous plants, and while 1, 3, and 6 have been seen as enzymatic products from bacterial P450s, 2, 4, 5, and 7 were not previously seen as bacterial biotransformation or enzymatic products. Several of these derivatives have known or proposed use in synthetic routes to relevant abietanes. Integration of biocatalysis with traditional synthetic techniques has emerged to be a powerful combination for the production of highly desired natural products, drugs, and commodities. Future efforts will focus on identifying the enzymes responsible for these transformations through continued utilization of genome mining tactics, biochemical techniques, and heterologous expression. Supplementary Material