Skip to main content
NCBI home page
Microbial Biotechnology logoLink to Microbial Biotechnology
. 2022 Sep 13;15(12):2890–2904. doi: 10.1111/1751-7915.14145

Salinomycin biosynthesis reversely regulates the β‐oxidation pathway in Streptomyces albus by carrying a 3‐hydroxyacyl‐CoA dehydrogenase gene in its biosynthetic gene cluster

Jiaxiu Wei 1,2, Binbin Chen 3, Jianxin Dong 1,2, Xueyu Wang 1,2, Yongquan Li 1,2, Yingchun Liu 4,✉, Wenjun Guan 1,2,✉
PMCID: PMC9733648  PMID: 36099515

Abstract

Streptomyces is well known for synthesis of many biologically active secondary metabolites, such as polyketides and non‐ribosomal peptides. Understanding the coupling mechanisms of primary and secondary metabolism can help develop strategies to improve secondary metabolite production in Streptomyces. In this work, Streptomyces albus ZD11, an oil‐preferring industrial Streptomyces strain, was proved to have a remarkable capability to generate abundant acyl‐CoA precursors for salinomycin biosynthesis with the aid of its enhanced β‐oxidation pathway. It was found that the salinomycin biosynthetic gene cluster contains a predicted 3‐hydroxyacyl‐CoA dehydrogenase (FadB3), which is the third enzyme of β‐oxidation cycle. Deletion of fadB3 significantly reduced the production of salinomycin. A variety of experimental evidences showed that FadB3 was mainly involved in the β‐oxidation pathway rather than ethylmalonyl‐CoA biosynthesis and played a very important role in regulating the rate of β‐oxidation in S. albus ZD11. Our findings elucidate an interesting coupling mechanism by which a PKS biosynthetic gene cluster could regulate the β‐oxidation pathway by carrying β‐oxidation genes, enabling Streptomyces to efficiently synthesize target polyketides and economically utilize environmental nutrients.

Comprehensive understanding the relationship between primary metabolism and secondary metabolism is important for high‐efficiency production of antibiotics in Streptomyces species. In contrast to the well‐known pattern that primary metabolic processes regulate secondary metabolism, this study reports that a secondary biosynthetic pathway can reversely regulate primary metabolic process by carrying a key gene of the primary metabolic pathway.

graphic file with name MBT2-15-2890-g004.jpg

INTRODUCTION

Streptomycetes have a great potential to produce many biologically active secondary metabolites (also known as natural products) (Berdy, 2005; Hwang et al., 2014). The production of secondary metabolites is highly dependent on the availability of biosynthetic precursors, most of which are products of primary metabolism (Olano et al., 2008; van Keulen et al., 2011). Changes in primary metabolism can result in knock‐on effects on secondary metabolism. For example, although all Streptomyces strains harbour the similar primary metabolic pathways, their abilities to yield a same secondary metabolite are quite different (Kallifidas et al., 2018; Peng et al., 2018; Wang et al., 2013). In Streptomyces, to achieve efficient synthesis of a target secondary metabolite, it is necessary to understand intrinsic link between its biosynthetic pathway and primary metabolism, so as to optimize the related primary metabolic pathways and provide sufficient precursors for the biosynthesis of target secondary metabolite.

Polyketides are a large family of structurally diverse secondary metabolites exhibiting a vast array of biological and pharmacological activities such as antibacterial, antifungal, anticholesterol, antiparasitic, anticancer and immunosuppressive properties (Staunton & Weissman, 2001). Polyketide backbones are always assembled from starter and extender units derived from carboxylated acyl‐coenzyme A (acyl‐CoA) precursors, such as acetyl‐CoA, malonyl‐CoA (M‐CoA), (2S)‐methylmalonyl‐CoA (MM‐CoA), (2S)‐ethylmalonyl‐CoA (EM‐CoA) and propionyl‐CoA (P‐CoA), via sequential Claisen‐like condensation reactions catalysed by polyketide synthases (PKSs) (Chan et al., 2009). These precursors are mainly derived directly or indirectly from acetyl‐CoA generated by glycolysis and fatty acid oxidation pathways.

In general, most Streptomyces strains preferentially utilize sugar as carbon source for cell growth and polyketide biosynthesis. This process mainly relies on the glycolytic pathway. Previous studies had found that oils, which are cheaper than other carbon sources, could also be employed as the main carbon source to produce polyketide antibiotics in several industrial Streptomyces strains (Efthimiou et al., 2008; Li et al., 2009; Wang, Liu, et al., 2017; Wang, Shan, et al., 2017). As the main component of oil, exogenous triglycerides (TAGs) can be hydrolysed into glycerol and free fatty acids (FFAs) via a series of secreted lipases (Alvarez & Steinbuchel, 2002). The majority of absorbed FFAs are finally broken down into acetyl‐CoA through β‐oxidation pathway, which always consists of five enzymatic steps (Figure S1) (Schulz, 1991). In Escherichia coli, FFAs are activated by thioesterification with CoA derivatives and subsequently undergo oxidation by an acyl‐CoA synthetase (FadD, EC 6.2.1.3) (Black et al., 1992). After that, a four‐step cycle begins with the generation of an unsaturation at the α‐β position by acyl‐CoA dehydrogenase (FadE, EC 1.3.8.7/8) (Cox et al., 2019; Fujita et al., 2007; Ghisla & Thorpe, 2004). The next step is carried out by enoyl‐CoA hydratase (ECH, EC 4.2.1.17), which hydroxylates the carbon chain at β‐position to produce β‐hydroxyacyl‐CoA. Its β‐hydroxy group is subsequently oxidized by 3‐hydroxyacyl‐CoA dehydrogenase (3HCDH, EC1.1.1.35) to yield the β‐keto group using NAD+ as a cofactor (Cox et al., 2019; DiRusso, 1990; Elena Volodina1, 2014; Fujita et al., 2007). Interestingly, single 3‐hydroxyacyl‐CoA dehydrogenase can not only function as a dimer but also associate with the enoyl‐CoA hydratase to form a complex to catalyse the reaction. Moreover, a 3HCDH domain and an ECH domain can form a multifunctional protein (FadB) to catalyse two‐step reaction in β‐oxidation cycle (Elena Volodina1, 2014; Menendez‐Bravo et al., 2017; Yang et al., 1988). Previous studies showed that a purified 3‐hydroxyacyl‐CoA dehydrogenase (FadB2) in Mycobacterium tuberculosis catalysed not only the dehydration of (3S)‐hydroxybutyryl‐CoA to acetoacetyl‐CoA at pH 10 but also the reverse reaction, converting acetoacetyl‐CoA to (3S)‐hydroxybutyryl‐CoA in a lower pH range of 5.5–6.5 (Taylor et al., 2010). This type of protein often has conserved catalytic residues, such as Ser, His and Asn (Cox et al., 2019; Kim et al., 2014). The final step of β‐oxidation cycle, catalysed by β‐ketoacyl‐CoA thiolase (FadA, EC 2.3.1.9/16), is the cleavage of β‐ketoacyl‐CoA to an acetyl‐CoA and an acyl‐CoA shortened by two carbon atoms (Yang et al., 1990). The shortened acyl‐CoA re‐enters the β‐oxidation cycle, whereas acetyl‐CoA enters the Krebs cycle and the fatty acid synthetic pathway or is transformed into the acyl‐CoA precursors for polyketides biosynthesis (Chan et al., 2009).

Since the β‐oxidation pathway is closely related to the supply of polyketide precursors, previous studies proposed several strategies to control the synthesis of downstream polyketides by manipulating the content of TAGs pool (Banchio & Gramajo, 1997; Hobbs et al., 1997; Wang et al., 2020). For example, the overexpression of fatty acyl‐CoA synthetase in Streptomyces coelicolor could guide more acyl‐CoAs into the β‐oxidation cycle, resulting in more production of acetyl‐CoA. The genomic analysis of S. coelicolor revealed many potential β‐oxidation gene homologues, including three fadAB gene clusters (Menendez‐Bravo et al., 2017). Deletion of these fadAB homologues could significantly impair the β‐oxidation process and increase intracellular TAGs accumulation, which then reduced the production of acetyl‐CoA.

Salinomycin produced by Streptomyces albus is a polyether antibiotic, which has been widely used as a food additive to prevent intestinal coccidiosis in poultry (Fuchs et al., 2010; Gumila et al., 1997; Kuo et al., 2012). Using a typical type I PKS (Figure S2A), the biosynthesis of salinomycin polyketide backbone requires one acetyl‐CoA, five M‐CoAs, six MM‐CoAs and three EM‐CoAs molecules as the biosynthetic precursors (Jiang et al., 2012). Most genes in the salinomycin biosynthetic gene cluster (BGC) are positively regulated by SlnR, which is a pathway‐specific regulator (Zhu et al., 2017) (Figure S2B). Streptomyces albus ZD11 was derived from an industrial strain which preferred to use soybean oil as carbon source and could produce up to 8 g/L salinomycin in shake flasks (Zhu et al., 2017). Our previous study revealed the remarkable enrichment of genes related to TAGs metabolism in S. albus ZD11 according to comparative genomic analysis (Li et al., 2020). For example, there are up to 101 predicted β‐oxidation genes in S. albus ZD11, including 24 acyl‐CoA synthetase genes, 39 acyl‐CoA dehydrogenase genes, 17 enoyl‐CoA hydratase genes, seven 3‐hydroxyacyl‐CoA dehydrogenase genes, 12 β‐ketoacyl‐CoA thiolase genes and also two genes encoding ECH‐3HCDH multifunctional protein. In addition, the transcriptomic profiling showed that the expression levels of genes involved in fatty acid catabolic and synthetic pathways were significantly increased when using soybean oil as a sole carbon source in S. albus ZD11. It supports an inference that high efficiency of soybean oil utilization could provide abundant precursors for cell growth and high yield of salinomycin (Li et al., 2020). However, the way how these mentioned genes work together to make the β‐oxidation pathway work efficiently is still unclear.

To clarify the relationship between the working efficiency of β‐oxidation pathway and the high yield of salinomycin in S. albus ZD11, we focused on the 3‐hydroxyacyl‐CoA dehydrogenases in this study. By deletion of eight predicted 3‐hydroxyacyl‐CoA dehydrogenase genes in S. albus ZD11 (Table S1), it was confirmed that FadB1 (DUI70_0391), FadB2 (DUI70_0803) and FadB3 (DUI70_0275, also known as SlnP) were important for salinomycin production. Among them, fadB3 locating inside the salinomycin BGC plays the most important role. The evidences from various experiments in this study suggest that the contribution of FadB3 to the high yield of salinomycin is mainly realized by enhancing fatty acid β‐oxidation in S. albus ZD11. Our work showed for the first time that the Streptomyces strain could efficiently synthesize the polyketide with the help of additional β‐oxidation genes carried by its PKS BGC.

EXPERIMENTAL PROCEDURES

Media and cultivation conditions

All experiments were performed with the strains and vectors listed in Table S2. Different vectors were propagated in E. coli cultured in Luria‐Bertani (LB) medium at 37°C. To prepare spores, S. albus ZD11 and its derivative strains were cultivated on ISP4 agar plates (BD) for 10 days at 30°C. The Streptomyces strains were cultivated in tryptic soy broth (TSB) medium (3%) for 24 h growth at 30°C and 220 rpm to obtain high‐quality genomic DNA for whole‐genome sequencing and genotype confirmation. The industrial seed medium (soybean powder, 3%; glucose, 4%; yeast extract, 1%; CaCO3, 0.1%) was used for seed cultivation for 29 h at 30°C and 220 rpm. Then, the seed cultures (10%) were inoculated into ionic medium (Li et al., 2020; Zhang et al., 2017) (NaCl, 0.5%; KCl, 0.5%; [NH4]2SO4, 2%; MgSO4, 0.02%; K2HPO4, 0.02%; CaCl2, 0.01%; CaCO3, 2%) with soybean oil (oil medium) or glucose (glucose medium) containing an equimolar amount of carbon (2.7 M) for 5‐day growth at 32°C and 180 rpm. When appropriate, apramycin, spectinomycin, kanamycin or chloramphenicol was added to the medium at a final concentration of 50, 50, 50 or 25 μg/ml, respectively.

Construction of gene disruption mutants and analysis of the phenotypes

A previously constructed fosmid library was used to generate the SGD and MGD mutants of S. albus ZD11 (Zhu et al., 2017). The fosmids listed in Table S3 were used to construct the gene disruption mutants using the PCR‐targeting system (Gust et al., 2003; Zhu et al., 2017). fadB2 and fadB9 were displaced by the spectinomycin and apramycin disruption cassette, respectively, as described previously (Wang, Liu, et al., 2017; Wang, Shan, et al., 2017). Other six fadBs were displaced by a 81 bp scar. The exconjugants were inoculated onto the ISP4 plates for two rounds of non‐selective growth before selection by replica plating for sensitive colonies. In addition, the ΔfadB2ΔfadB1 mutant was constructed by deleting the fadB1 in the ΔfadB2 mutant. fadB3 and fadB4 were knocked out in the ΔfadB2ΔfadB1 mutant to construct the ΔfadB2ΔfadB1ΔfadB3 mutant and the ΔfadB2ΔfadB1ΔfadB4 mutant, respectively. All the mutants were confirmed by PCR, RT‐qPCR and DNA‐sequencing. Salinomycin yield was measured according to the method previously reported (Li et al., 2020). Five ml of fermentation broth was harvested at the end of fermentation (120 h) and then centrifuged for 10 min to remove the supernatant. After washing with equal volume methanol, the collected mycelia were dried in an oven at 80°C for 24 h to determine the dry weight.

RNA‐seq analysis and quantitative real‐time RT‐PCR (RT‐qPCR) assay

For total RNA extraction, S. albus ZD11 and its related mutants were grown in an ionic medium with soybean oil or glucose (0.45 M) as the sole carbon source. Three replicates of mycelia were collected at 36 h from the media and stored at −80°C until RNA extraction. Total RNA isolation and RNA‐seq were performed according to a previous report (Wang, Liu, et al., 2017; Wang, Shan, et al., 2017). RT‐qPCR was performed on a Mastercycler ep gradient S instrument (Eppendorf, Hamburg, Germany) using ChamQ STBR qPCR Master Mix (Vazyme, China) according to the manufacturer's guidelines. The expression level of target gene was internally normalized to housekeeping gene hrdB (DUI70_1421) and quantified by the 2−△△CT method (Livak & Schmittgen, 2001). Technical triplicates of three biological repeats were performed per condition.

Phylogenetic analysis of 3‐hydroxyacyl‐CoA dehydrogenase orthologous

Firstly, domain compositions of nine 3‐hydroxyacyl‐CoA dehydrogenases were analysed based on Pfam database (https://pfam.xfam.org). All relative amino acids sequences of the proteins were obtained from GenBank in FASTA format. The alignment was then generated by CLUSTALW (Larkin et al., 2007) using manual adjustments. Neighbour‐joining tree was constructed using MEGA11 (Tamura et al., 2021) and then landscaped by the iTOL (Interactive Tree of Life) online software (https://itol.embl.de/) (Letunic & Bork, 2016). 272 PKS BGCs in Streptomyces were collected from the MIBiG database (Kautsar et al., 2020). The amino sequences of predicted 3‐hydroxyacyl‐CoA dehydrogenase orthologous within the corresponding BGCs were downloaded to construct the phylogenetic tree according to the above method.

Protein expression and purification

The nucleotide sequence of fadB3 was amplified from the genomic DNA of S. albus ZD11 and cloned into pET28a to generate the recombinant vector pETFadB3. Escherichia coli BL21(DE3)/pKJE7 was chosen for heterologous expression of FadB3. The overnight culture of recombinant E. coli BL21(DE3)/pKJE7 was inoculated into 1000 ml liquid LB medium with appropriate antibiotics and grown to an OD600 of 0.5–0.6 at 37°C. Then the strain was induced 16 h at 16°C by the addition of isopropyl‐β‐D‐thiogalactopyranoside (IPTG) and L‐arabinose to final concentration of 0.1 and 2.5 mM, respectively. The cells were harvested by centrifugation and resuspended in the lysis buffer (0.5 M NaCl, 10 mM imidazole, 1 mM DTT, 40 mM Tris–HCl, pH 7.6) and followed by cell disruption by sonication. After the recovery of supernatant by centrifugation, soluble FadB3 with double His6 tag was purified using nickel‐nitrilotriacetic acid (Ni‐NTA) resin. The resulting resins were first washed with the washing buffer (300 mM NaCl, 150 mM imidazole, 1 mM DTT, 40 mM Tris–HCl, pH 7.6), and the target protein was then eluted with the elution buffer (300 mM NaCl, 500 mM imidazole, 1 mM DTT, 40 mM Tris–HCl, pH 7.6). The purified protein was confirmed by SDS‐PAGE analysis. The enzyme was then dialyzed against 40 mM Tris–HCl buffer, pH 7.6, containing 150 mM NaCl and 10% glycerol. Purified FadB3 was concentrated and stored at −80°C for further use.

In vitro biochemical assays and product analysis by HPLC

Enzyme activity was measured by monitoring the conversion between acetoacetyl‐CoA and (3S)‐hydroxybutyryl‐CoA using HPLC. When (3S)‐hydroxybutyryl‐CoA was used as the substrate, the reaction mixture (100 μl) contained 50 mM NAD+, 25 μM FadB3, 25 μM (3S)‐hydroxybutyryl‐CoA, 10 mM MgCl2 and 1 mM 1,4‐dithiothreitol (DTT) in 40 mM Tris–HCl buffer (pH 7.6). When acetoacetyl‐CoA was used as the substrate, the reaction mixture contained 50 mM NADH, 25 μM FadB3, 25 μM acetoacetyl‐CoA, 10 mM MgCl2 and 1 mM DTT in 40 mM Tris–HCl buffer (pH 7.6). All the above reaction mixtures were incubated at 30°C for different time and then quenched with two volumes of methanol. The supernatants were then subjected to HPLC analysis, which was carried out with an Agilent 1260 infinity system, using a ZORBAX Eclipse XDB‐C18 column (4.6 × 150 mm, 5 μm). For the detection of two substrates after reaction, elution was performed at 0.8 ml min−1 with a mobile‐phase mixture consisting of a linear gradient of 50 mM ammonium acetate (pH 4.7) and methanol ([v/v]: 70:30, 0–25 min; 20:80, 25–30 min; 95:5, 30–38 min) (detection wavelength: 259 and 340 nm).

Computational methods of protein docking

The predicted protein structure of FadB3 was obtained from Robetta Server (Kim et al., 2004) (http://robetta.bakerlab.org), using an improved deep learning based on modelling method RoseTTAFold (Figure S10). The model of FadB3 has high conformation reliability based on the Ramachandran plot (Figure S11). Then the substrates were docked into FadB3 as the initial simulation structures by a flexible protein docking program AutoDock Vina (Trott & Olson, 2010). All simulations were performed with GROMACS 5.0 (Abraham et al., 2015) using an amber99sb force field. The force field parameters of the substrates (AA‐CoA, HB‐CoA, HCO‐CoA and OCO‐CoA) were constructed by Gaussian (Bahram & El‐Shehawy, 2015) and ACPYPE (Vranken, 2012). The Berendsen thermostat and Parrinello‐Rahman pressure coupling were applied to the system to maintain the constant temperature of 310 K and the constant pressure of 1.0 bar. Periodic boundary conditions were used in all simulations. The long‐range electrostatic interaction with a cut‐off distance of 1.2 nm was calculated by the particle mesh Ewald summation for the separation of the direct and reciprocal space. The cut‐off of non‐bonded van der Waals interaction was set at a distance of 1.2 nm. Bond lengths were constrained by the linear constraint solver algorithm (Hess et al., 1997). A docked substrate–protein binding configuration was first placed in a 10 × 10 × 10 nm3 box. Then water molecules were added as the solvents. Sodium and chloride ions were also included in the system to maintain electrical neutrality. Simulations were carried out with a time step of 2 fs. The system was equilibrated by 50,000 steps of energy minimization, as well as 1 ns MD simulation in the NVT ensemble and 2 ns in the NPT ensemble. Stable configurations were obtained after 50 ns (or 100 ns in some systems) of MD balance. Configurations in this study were visualized by VMD (Humphrey et al., 1996) and Pymol (DeLano, 2002).

Heterologous biosynthesis of daunorubicin (DNR) in S. albus ZD11

The 40.64 kb daunorubicin BGC consisting of 37 genes was directly cloned from the genomic DNA of S. coeruleorubidus into a linear p15A vector using RecET direct cloning and got the vector p15A‐DNR. After inserting the cassette for conjugation and PhiC31 site‐specific recombination (oriT‐phiC31) into the p15A using Redαβ recombination, the E. coli–Streptomyces shuttle vector pDNR1 was obtained (Wang et al., 2016). The vector pDNR2 was obtained by replacing chloramphenicol resistance gene with apramycin resistance gene (aac[3]IV) in pDNR1. Then the promoter (172 bp) of dnrI in pDNR2 was replaced by a strong constitutive promoter kasOp* (Wang et al., 2013) to generate the vector pDNR3. The shunt pathway genes, dnrH, dnrX and dnrU (Hutchinson, 1997; Shrestha et al., 2019), were deleted from pDNR3 to generate the vector pDNR4. The S2 mutant (ΔsalΔfadB3) was constructed based on the S1 mutant (Δsal), which was constructed in our previous study (Dong et al., 2021). pDNR4 was respectively transferred into the S1 and S2 mutant to construct the S3 (ΔsalΔfadB3::pDNR4) and S4 (Δsal::pDNR4) mutants.

About 100 μl of fermentation broth was extracted using 900 μl of methanol for 12 h and centrifuged at 12,000 rpm for 10 min. The supernatant was filtered through 0.22‐μm filters and the yield of daunorubicin was detected by HPLC, using a ZORBAX Eclipse XDB‐C18 column on the Shimadzu LC‐20AT system. The biomass was determined by dry weight measurement mentioned above. For the detection of DNR, elution was performed at 1 ml min−1 with a mobile‐phase mixture consisting of a linear gradient of 10 mM ammonium acetate in water (phase A) and acetonitrile (phase B). Elution was performed as follows: a linear gradient from 20% to 50% solvent B from 0 to 15 min, a linear gradient from 50% to 90% solvent B from 15 to 18 min and 90% solvent B from 18 to 22 min (detection wavelength: 254 nm).

RESULTS

Multiple analysis revealed the importance of 3‐hydroxyacyl‐CoA dehydrogenases in the β‐oxidation pathway of S. albus ZD11

To investigate the distinctiveness of β‐oxidation pathway in S. albus ZD11, we carried out a comparative genomic analysis between S. albus ZD11 and three commonly used Streptomyces strains (S. coelicolor A3[2], S. albidoflavus J1074 and S. lividans TK24), which can utilize both glucose and oil as the carbon sources (Kuo et al., 2012). According to the COG (Clusters of Orthologous Groups) database and the KEGG (Kyoto Encyclopedia of Genes and Genomes) database (Huerta‐Cepas et al., 2019; Kanehisa & Goto, 2000), it was found that not only the total number of β‐oxidation genes but also the amount of genes involved in every single step of β‐oxidation pathway were significantly higher in S. albus ZD11 than those in the other three strains (Figure 1A,B). Next, the result of transcriptomic analysis indicates that the genes belonging to the β‐oxidation pathway exhibited higher expression levels when S. albus ZD11 was cultured in the soybean oil medium compared to the glucose medium (Figure 1C). Interestingly, although the total number of 3‐hydroxyacyl‐CoA dehydrogenase genes (abbreviated to fadBs) is minimal among the five reaction steps of the β‐oxidation pathway in S. albus ZD11, the sum of FPKM values (Fragments Per Kilobase per Million) of these genes is the highest when the strain was cultured in the soybean oil medium. Among them, the FPKM value of fadB3 was the highest, and that of fadB7 was the lowest and less than the lower limit of gene expression level (FPKM value <10) defined in this study. The transcriptional levels of five of the nine fadBs were significantly upregulated in the soybean oil medium (Figures 1D and S3). Surprisingly, fadB3 was found to locate inside the biosynthetic gene cluster of salinomycin and was positively regulated by the pathway‐specific regulator SlnR (Figure S2B). According to our previous study, the transcriptional level of fadB3 was decreased over 90% in the slnR deletion mutant and was significantly increased in the slnR overexpression mutant compared to that in the WT strain (Jiang et al., 2012).

FIGURE 1.

FIGURE 1

Open in a new tab

Multiple analysis of the genes predicted to involve in the β‐oxidation pathway of S. albus ZD11. (A) Quantities of predicted β‐oxidation genes in S. albus ZD11, S. albidoflavus J1074, S. coelicolor A3(2) and S. lividans TK24. (B) Quantities of genes involved in each step of β‐oxidation pathway. (C) The sum of FPKM values of genes in each step of β‐oxidation pathway when the S. albus ZD11 WT strain was cultured in the medium using soybean oil or glucose as sole carbon source. (D) The fold changes in expression levels of nine 3‐hydroxyacyl‐CoA dehydrogenase genes in the oil medium relative to those in the glucose medium. RNA fold change is the normalized log2 ratio of FPKM values (FPKMoil/FPKMglucose). (E) Analysis of the domain composition of nine 3‐hydroxyacyl‐CoA dehydrogenases. (F) Phylogenetic tree of the 3HCDH_N‐3HCDH units from nine 3‐hydroxyacyl‐CoA dehydrogenases. Phylogenetic tree was reconstructed using the neighbour‐joining algorithm implemented in MEGA11. Scale bar represents 0.1 estimated sequence divergence. D1 and D2, respectively, represent the N‐terminal and C‐terminal 3HCDH_N‐3HCDH units located in the corresponding 3‐hydroxyacyl‐CoA dehydrogenases.

According to the analysis of conserved domains base on Pfam database (Mistry et al., 2021), the nine predicted 3‐hydroxyacyl‐CoA dehydrogenases were divided into four classes (Figure 1E). In the first class, FadB1 and FadB2 consist of an N‐terminal crotonase‐like family/enoyl‐CoA hydratase domain (pfam00378), a central 3‐hydroxyacyl‐CoA dehydrogenase‐NAD binding domain (3HCDH_N, pfam02737) and C‐terminal 3‐hydroxyacyl‐CoA dehydrogenase domain (3HCDH, pfam00725). These two proteins showed high identities to the known fatty acid oxidative multifunctional enzyme SCO6732 (84.49%) and SCO6026 (88.59%) from S. coelicolor (Menendez‐Bravo et al., 2017), implying FadB1 and FadB2 are the multifunctional proteins with both ECH and 3HCDH activities. In the second class, FadB3 and FadB9 have two 3HCDH_N domains and two 3HCDH domains. The domain composition of FadB5 was similar to the second class, which had a 3HCDH_RFF (pfam18321, a 3HCDH_N missing the catalytic residues and an NAD [H] binding cleft) instead of 3HCDH_N at the N‐terminal part. The last class includes FadB4, FadB6, FadB7 and FadB8, which consist of only one 3HCDH_N domain and one 3HCDH domain. According to the previous reports (Cox et al., 2019; Menendez‐Bravo et al., 2017; Taylor et al., 2010), the members in the last class of 3‐hydroxyacyl‐CoA dehydrogenase perhaps function by assembling into a dimer. In addition, phylogenetic analysis showed that all 3HCDH_N‐3HCDH units from the four classes of 3‐hydroxyacyl‐CoA dehydrogenase could be classified into two clades (Figure 1F). Clade I contains eight 3HCDH_N‐3HCDH units located in FadB3, FadB4, FadB6, FadB7, FadB8 and FadB9, respectively. Clades II contains three 3HCDH_N‐3HCDH units belonging to FadB1, FadB2 and FadB5. It suggests that the function of members in each group might be more similar.

FadB3 presents a core status among the 3‐hydroxyacyl‐CoA dehydrogenases in S. albus ZD11

To investigate the role of predicted 3‐hydroxyacyl‐CoA dehydrogenases in S. albus ZD11, we tried to knock out each 3‐hydroxyacyl‐CoA dehydrogenase gene by PCR targeting (Zhu et al., 2017) and constructed a series of single‐gene deletion (SGD) mutants, except fadB8 because of its low transcriptional level under the existing fermentation conditions. Compared to the WT strain, the salinomycin yields of eight SGD mutants were all decreased. The ΔfadB1, ΔfadB2, ΔfadB3 and ΔfadB4 mutants showed more significant reduction in salinomycin production than the other SGD mutants. Among them, the ΔfadB3 mutant exhibited the least yield of salinomycin (Figure 2A). Moreover, the biomass of each SGD mutant was almost the same as that of the WT strain, indicating that knockout of each 3‐hydroxyacyl‐CoA dehydrogenase gene alone did not affect cell growth (Figure S4A).

FIGURE 2.

FIGURE 2

Open in a new tab

Deletion of the fadBs decreased the salinomycin production of S. albus ZD11. (A) The relative salinomycin yields of the SGD mutants. (B) The relative salinomycin yields of the MGD mutants. Error bars indicate the SD for samples tested in triplicate. (C) Expression level analysis of the remaining 3‐hydroxyacyl‐CoA dehydrogenase genes in each SGD mutant as compared to the WT strain. hrdB transcription was monitored and used as the internal control. Each fold change was calculated by RT‐qPCR using the 2−△△CT method. The data are presented as the average of three parallel samples.

To figure out how these 3‐hydroxyacyl‐CoA dehydrogenases work together in S. albus ZD11, the expression levels of remaining fadBs in each SGD mutant were measured. As shown in Figure 2C, the expression level of each remaining fadB gene in the ΔfadB3 mutant was more or less upregulated. At the same time, the expression level of fadB3 was upregulated in almost all other SGD mutants except the ΔfadB1 mutants. This indicates that fadB3 could be upregulated to recover the loss of other fadB genes and vice versa, suggesting its central role among the fadBs in S. albus ZD11.

Then several multi‐gene deletion (MGD) mutants were constructed in order to further analyse the effect of FadB1, FadB2, FadB3 or FadB4 on salinomycin production. As shown in Figure 2B, the salinomycin yield of the ΔfadB2ΔfadB1 mutant was less than that of the ΔfadB1 mutant or the ΔfadB2 mutant. Compared to the ΔfadB2ΔfadB1 mutant, the ΔfadB2ΔfadB1ΔfadB3 mutant showed a remarkable drawdown in salinomycin yield, while the ΔfadB2ΔfadB1ΔfadB4 mutant exhibited a relative less reduction. These results further determined the critical role of FadB3. In addition, these MGD mutants presented no obvious reduction in biomass compared to the WT strain, implying that the remaining fadBs in the mutants are still able to keep the β‐oxidation pathways working properly for primary metabolism in S. albus ZD11 (Figure S4B).

The 3‐hydroxyacyl‐CoA dehydrogenase orthologous located inside the PKS BGCs show no inevitable connection with the biosynthesis of EM‐CoA

Since fadB3 locates within the salinomycin BGC, previous studies suggested that FadB3 might participate in the biosynthesis of EM‐CoA precursor (Jiang et al., 2012; Lu et al., 2016), but there was no experimental evidence to confirm this. To investigate the association between the β‐oxidation genes within PKS BGCs and the biosynthesis of corresponding polyketides, we collected 272 PKS BGCs from the MIBiG database (Kautsar et al., 2020) (Appendix S1). It was found that more than half of these BGCs do not contain the predicted β‐oxidation genes and 10 of such BGCs are considered to use EM‐CoA as the biosynthetic precursor. On the contrary, there are 126 PKS BGCs containing the predicted β‐oxidation genes, and 30 of them harbour at least one 3‐hydroxyacyl‐CoA dehydrogenase gene. 11 of the 30 BGCs are supposed to use EM‐CoA as the biosynthetic precursor, implying that the 3‐hydroxyacyl‐CoA dehydrogenase genes located within the BGCs, at least, might be not just related to the biosynthesis of EM‐CoA (Figure 3A). Except the 3‐hydroxyacyl‐CoA dehydrogenase genes, the genes predicted to encode acyl‐CoA synthetase, acyl‐CoA dehydrogenase or enoyl‐CoA hydratase are also found in these PKS BGCs (Figure 3B, Appendix S1). And the acyl‐CoA synthetase genes account for the highest proportion. Phylogenetic analysis showed that the 3HCDH_N‐3HCDH units from the thirty‐one 3‐hydroxyacyl‐CoA dehydrogenase orthologous could be classified into three clades (Figure 3C). Six 3‐hydroxyacyl‐CoA dehydrogenases (including FadB3 in S. albus ZD11) containing two 3HCDH_N‐3HCDH units all belong to clade III, indicating their closer evolutionary relationships. These results demonstrate that the 3‐hydroxyacyl‐CoA dehydrogenase orthologous found in the PKS BGCs are not necessarily associated with EM‐CoA biosynthesis.

FIGURE 3.

FIGURE 3

Open in a new tab

Relationship between the PKS BGCs and the β‐oxidation genes within these BGCs. (A) Distribution of the PKS BGCs collected from the MIBiG database. (B) Distribution of the PKS BGCs containing β‐oxidation genes. (C) Phylogenetic tree of 3HCDH_N‐3HCDH units from the 3‐hydroxyacyl‐CoA dehydrogenases located in the 30 PKS BGCs. Phylogenetic tree was reconstructed using the neighbour‐joining algorithm implemented in MEGA 11. Scale bar represents 0.1 estimated sequence divergence. D1 unit and D2 unit represent the N‐terminal and C‐terminal 3HCDH_N‐3HCDH units located in the corresponding 3‐hydroxyacyl‐CoA dehydrogenases. Black dots represent the 3‐hydroxyacyl‐CoA dehydrogenase from the PKS BGCs using EM‐CoA as a biosynthetic precursor.

FadB3 prefers to catalyse the conversion of 3‐hydroxyacyl‐CoA to the 3‐ketoyl acid to enhance the β‐oxidation pathway in S. albus ZD11 

3‐Hydroxyacyl‐CoA dehydrogenase is the penultimate enzyme in β‐oxidation cycle and reversibly catalyses the conversion of 3‐hydroxyacyl‐CoA to 3‐ketoacyl‐CoA. To evaluate its catalytic activity, fadB3 was cloned from S. albus ZD11 and heterologously expressed in E. coli BL21 (DE3) to get the purified protein (Figure 4A). According to the previous reports, the activity of 3‐hydroxyacyl‐CoA dehydrogenase was usually measured by NAD+/NADH‐dependent dehydrogenase reaction using acetoacetyl‐CoA or (3S)‐hydroxybutyryl‐CoA as the short‐chain acyl‐CoA substrate (Elena Volodina1, 2014; Taylor et al., 2010). As shown in Figure 4B, under the condition of pH 7.6 which is within the range of physiological pH in Streptomyces cell, FadB3 could catalyse the conversion of (3S)‐hydroxybutyryl‐CoA to acetoacetyl‐CoA much better than the reverse reaction. After 90 minutes of catalysis, more than two‐thirds of (3S)‐hydroxybutyryl‐CoA was transformed to acetoacetyl‐CoA, but only few acetoacetyl‐CoA was converted to (3S)‐hydroxybutyryl‐CoA. Since one of the pathways involved in EM‐CoA biosynthesis is started with converting acetoacetyl‐CoA to (3S)‐hydroxybutyryl‐CoA (Figure S1), our result gave the evidence that FadB3 prefers to participate in the β‐oxidation pathway rather than the synthesis of EM‐CoA.

FIGURE 4.

FIGURE 4

Open in a new tab

Evaluation of the catalytic activity of FadB3. (A) SDS‐PAGE gel image of the purified FadB3 (64.9 kDa). Lane 1: purified FadB3 with N‐terminal and C‐terminal His6 tag. Lane M: protein marker. Adobe Photoshop (PS) CC 2018 was used for image editing. (B) FadB3 catalysed the NAD+/NADH‐dependent reaction between (3S)‐hydroxybutyryl‐CoA (HB‐CoA) and acetoacetyl‐CoA (AA‐CoA). The contents of acetoacetyl‐CoA and (3S)‐hydroxybutyryl‐CoA were determined by HPLC analysis (high performance liquid chromatography).

To validate the catalytic results above and figure out the possibility of long‐chain acyl‐CoAs catalysed by FadB3, we thus turned to computational approaches to gain mechanistic insight into the enzymatic properties of FadB3. As the docking results of short‐chain acyl‐CoAs showed, AA‐CoA and HB‐CoA were docking with FadB3‐D1 (N‐terminal 3HCDH_N‐3HCDH unit of FadB3) and FadB3‐D2 (C‐terminal 3HCDH_N‐3HCDH unit of FadB3), respectively. Two representative binding conformations were selected due to their stability for further analysis (Figure S5A,B). RMSF (Root Mean Square Fluctuation) and RMSD (Root Mean Square Deviation) showed that FadB3 could combine with short‐chain acyl‐CoAs appropriately and stably (Figures S6 and 5A,B). Analysis of the conformation structures revealed that the short‐chain acyl‐CoAs were probably positioned near the residues Thr18, Thr229, His130 and Lys264 in FadB3‐D1 (Figures 6A,B and S7A,C). And Asn480, His430, Asn433 and Pro515 could stabilize the short‐chains probably in FadB3‐D2 (Figures 6C,D and S7B,D). As shown in Figure S8, sequence alignment showed that His130 in FadB3‐D1 and His430 in FadB3‐D2 were conserved compared to His143 in FadB2 from M. tuberculosis and His140 in PaaH1 from Ralstonia eutropha, which were reported as the conserved catalytic residues (Cox et al., 2019; Kim et al., 2014). Therefore, there is a great possibility that short‐chain acyl‐CoAs could be catalysed in FadB3‐D1 and FadB3‐D2 according to the results of MD simulations, which can be mutually verified with above experimental results (Figure 4B). Then the substrates were extended to long‐chain acyl‐CoAs based on the validity of MD simulations. Because the main fatty acids of soybean oil is oleic acid, (3S)‐hydroxy‐9Z‐octadecenoyl‐CoA and (9Z)‐3‐oxooctadecenoyl‐CoA were chosen as the long‐chain substrates of FadB3. Representative binding conformations between long‐chain acyl‐CoAs and FadB3 were also selected due to their stability for further mechanistic study. Our results showed that the interaction energy between long‐chain acyl‐CoAs and FadB3 was mainly van der Waals and with no obvious difference comparing with short‐chain acyl‐CoAs (Figure S9). However, FadB3‐D2 could combine with long‐chain acyl‐CoAs more stable than FadB3‐D1 according to the less fluctuation of RMSD in FadB3‐D2 (Figure 5C,D), which might be due to the flexible long carbon chains. Furthermore, the possible catalytic residue Asn480 was observed in the conformation of long‐chain acyl‐CoAs with FadB3‐D2 (Figures 6E,F and S7E,F). However, any possible catalytic residues could not be found at the binding site when long‐chain acyl‐CoAs were combined with the FadB3‐D1. In conclusion, these results indicate that FadB3‐D2 can not only bind short‐chain acyl‐CoAs but also long‐chains acyl‐CoAs. Based on the above analysis, FadB3 can bind short‐chain and long‐chain acyl‐CoAs with different preferences, suggesting its ability to act as a key 3‐hydroxyacyl‐CoA dehydrogenase to catalyse oxidation of both short‐chain and long‐chain acyl‐CoAs in the β‐oxidation process, but the oxidation of short‐chain acyl‐CoAs may occur more easily.

FIGURE 5.

FIGURE 5

Open in a new tab

RMSD of acyl‐CoAs located in different domains of FadB3. (A) acetoacetyl‐CoA (AA‐CoA). (B) (3S)‐hydroxybutyryl‐CoA (HB‐CoA). (C) (3S)‐hydroxy‐9Z‐octadecenoyl‐CoA (HCO‐CoA). (D) (9Z)‐3‐oxooctadecenoyl‐CoA (OCO‐CoA).

FIGURE 6.

FIGURE 6

Open in a new tab

Binding configurations of substrates and FadB3. (A) FadB3‐D1 combines with AA‐CoA; (B) FadB3‐D1 combines with HB‐CoA; (C) FadB3‐D2 combines with AA‐CoA; (D) FadB3‐D2 combines with HB‐CoA; (E) FadB3‐D2 combines with HCO‐CoA; (F) FadB3‐D2 combines with OCO‐CoA.

Heterologous biosynthesis of daunorubicin (DNR) further confirmed the key role of FadB3 played in the β‐oxidation pathway of S. albus ZD11

Since FadB3 could also catalyse the conversion from acetoacetyl‐CoA to (3S)‐hydroxybutyryl‐CoA with lower efficiency in vitro, whether the significant reduction in salinomycin yield of ∆fadB3 mutant is more influenced by β‐oxidation or EM‐CoA biosynthesis is still uncertain. To explain that, a daunorubicin (DNR) BGC was cloned from Streptomyces coeruleorubidus and heterogeneously expressed in S. albus ZD11. DNR is synthesized by a type II PKS. This PKS assembles 10 precursors consisting of one P‐CoA starter unit and nine M‐CoA extender units into the anthracyclinone skeleton of DNR (Blumauerova et al., 1977; Malla et al., 2010) (Figure 7A). Most of the genes in the DNR BGC are regulated by a pathway‐specific regulator DnrI (Tang et al., 1996). The S2 mutant (ΔsalΔfadB3) based on a S1 (Δsal) mutant that cannot produce salinomycin anymore (Figure S2C) (Dong et al., 2021) was constructed in this study. The vector (pDNR4) carrying an engineered DNR BGC was integrated into the chromosome of S2 or S1 mutants, respectively. As shown in Figure 7B,C, the S4 (ΔsalΔfadB3::pDNR4) mutant produced 47% of DNR less than the S3 (Δsal::pDNR4) mutant when using soybean oil as sole carbon source. We suggest that the β‐oxidation pathway lacking fadB3 decreases the generation of acetyl‐CoA, which in turn reduces M‐CoA production in S4 mutant and leads to a reduction in DNR production compared to the S3 mutant. Since biosynthesis of DNR does not use EM‐CoA as a precursor, this result further confirmed the key role of FadB3 played in the β‐oxidation pathway of S. albus ZD11.

FIGURE 7.

FIGURE 7

Open in a new tab

Heterologous biosynthesis of daunorubicin in S. albus ZD11. (A) Engineering the promoter of dnrI (a pathway‐specific regulator gene) and blocking the shunt pathways in DNR BGC. (B) HPLC analysis of the yield of daunorubicin in the S1 (Δsal), S2 (ΔsalΔfadB3), S3 (Δsal::pDNR4) or S4 (ΔsalΔfadB3::pDNR4) mutant. 1# and 2# peaks represent 13‐dihydrodaunorubicin and DNR, respectively. (C) The production of daunorubicin in the S3 and S4 mutants. Error bars indicate the SD for samples tested in triplicate.

DISCUSSION

In Streptomyces, understanding the coupling mechanisms of primary and secondary metabolism could enable the development of strategies to improve the production of target secondary metabolites. Genomic and transcriptomic analysis indicates that S. albus ZD11 has a remarkable capability to generate abundant acyl‐CoA precursors for salinomycin biosynthesis with the help of its enhanced β‐oxidation pathway. In this work, 3‐hydroxyacyl‐CoA dehydrogenase that is the penultimate enzyme in the β‐oxidation cycle was carefully studied. We found that a 3‐hydroxyacyl‐CoA dehydrogenase gene (fadB3) carried by the salinomycin BGC played a very important role in adjusting the rate of β‐oxidation in S. albus ZD11. Deletion of fadB3 significantly reduced the production of salinomycin. The evidences from in vitro enzymatic activity detection, protein–substrate docking and heterologous expression of DNR BGC all proved that FadB3 mainly involved in the β‐oxidation pathway rather than EM‐CoA biosynthesis.

Since fadB3 is positively regulated by the pathway‐specific regulator SlnR, it should start to function only after the biosynthesis of salinomycin is initiated. As mentioned earlier, the amount of 3‐hydroxyacyl‐CoA dehydrogenase genes is minimal among the five steps of β‐oxidation pathway in S. albus ZD11, suggesting 3‐hydroxyacyl‐CoA dehydrogenase might be the rate‐limiting step. Thus, we proposed a mechanism of how salinomycin biosynthesis regulates the rate of fatty acid oxidation in S. albus ZD11. As shown in Figure 8, when salinomycin synthesis was initiated, fadB3 was then highly expressed, which relieved the rate limitation caused by the lack of 3‐hydroxyl‐CoA dehydrogenases. This would result in more generation of acetyl‐CoAs, most of which were subsequently converted to various acyl‐CoA precursors for salinomycin biosynthesis (Figure S1). On the contrary, when the strain did not synthesize salinomycin, fadB3 would not be activated to express, and the remaining 3‐hydroxyacyl‐CoA dehydrogenases would enable the β‐oxidation pathway to operate at a relatively low rate to economically utilize nutrients in the environment.

FIGURE 8.

FIGURE 8

Open in a new tab

Schematic diagram of the proposed regulatory mechanism of the salinomycin biosynthesis on the β‐oxidation pathway in S. albus ZD11.

Here, we report a very interesting mechanism by which a secondary metabolic pathway reversely regulates the primary metabolic processes. The PKS BGCs could regulate the β‐oxidation pathway by carrying the β‐oxidation genes, enabling Streptomyces strains to efficiently synthesize polyketides. By analysing more PKS BGCs, we found such mechanism might be widespread. Nearly 50% of the investigated PKS BGCs contain predicted β‐oxidation genes, including the genes encoding acyl‐CoA synthetase, acyl‐CoA dehydrogenase, 3‐hydroxyacyl‐CoA dehydrogenase or enoyl‐CoA hydratase, suggesting that they may play the similar roles like FadB3.

Although we used computational approaches to confirm that FadB3 could bind short‐chain and long‐chain acyl‐CoAs with different preferences, it is still unclear how FadB3 catalyses the acyl‐CoAs with different chain length. Besides, the protein–substrate docking predicted several key catalytic sites in FadB3, and it has not been experimentally verified, which is worth to study in our future work. It is also worth noticing that the existing methods for protein function prediction cannot accurately distinguish 3‐hydroxyacyl‐CoA dehydrogenase and 3‐hydroxybutyryl‐CoA dehydrogenase. Thus, functions of the other predicted 3‐hydroxyacyl‐CoA dehydrogenase in S. albus ZD11 still remain uncertain and need further study in the future.

AUTHOR CONTRIBUTIONS

J. X. Wei and W. J. Guan designed the study. J. X. Wei wrote the manuscript. J. X. Wei, J. X. Dong and X. Y. Wang performed the wet‐lab experiments. B. B. Chen and Y. C. Liu designed and performed the protein‐ligand docking experiments. J. X. Wei and B. B. Chen analysed the data. W. J. Guan and Y. C. Liu revised the manuscript. All authors read and approved the final manuscript.

CONFLICT OF INTEREST

The authors declare no competing interests.

Supporting information

Table S1

Click here for additional data file. (6.4MB, docx)

Appendix S1

Click here for additional data file. (288.4KB, pdf)

ACKNOWLEDGMENTS

This work was financially supported by the National Key R&D Program of China (2019YFA0905400, 2018YFA0903200).

Wei, J. , Chen, B. , Dong, J. , Wang, X. , Li, Y. & Liu, Y. et al. (2022) Salinomycin biosynthesis reversely regulates the β‐oxidation pathway in Streptomyces albus by carrying a 3‐hydroxyacyl‐CoA dehydrogenase gene in its biosynthetic gene cluster. Microbial Biotechnology, 15, 2890–2904. Available from: 10.1111/1751-7915.14145

Jiaxiu Wei and Binbin Chen contributed equally to this work.

Contributor Information

Yingchun Liu, Email: liuyingch@zju.edu.cn.

Wenjun Guan, Email: guanwj@zju.edu.cn.

DATA AVAILABILITY STATEMENT

Genomic sequences of S. albus ZD11 (CP033071.1), S. albidoflavus J1074 (CP004370.1), S. coelicolor A3(2) (AL645882.2) and S. lividans TK24 (NZ_CP009124.1) were obtained from GenBank in GenBank format. The sequence reads obtained by RNA‐seq were obtained from in the SRA database under accession numbers SAMN12251778, SAMN12251779, SAMN12251780 and SAMN12251781, which were described in our previous work (Li et al., 2020).

REFERENCES

  1. Abraham, M.J. , Murtola, T. , Schulz, R. , Páll, S. , Smith, J.C. , Hess, B. et al. (2015) GROMACS: high performance molecular simulations through multi‐level parallelism from laptops to supercomputers. SoftwareX, 1‐2, 19–25. [Google Scholar]
  2. Alvarez, H.M. & Steinbuchel, A. (2002) Triacylglycerols in prokaryotic microorganisms. Applied Microbiology and Biotechnology, 60, 367–376. [DOI] [PubMed] [Google Scholar]
  3. Bahram, A. & El‐Shehawy, S.A. (2015) Study of the convergence of the increments of gaussian process. Applied Mathematics, 06, 933–939. [Google Scholar]
  4. Banchio, C. & Gramajo, H.C. (1997) Medium‐ and long‐chain fatty acid uptake and utilization by Streptomyces coelicolor A3(2): first characterization of a gram‐positive bacterial system. Microbiol‐Sgm, 143, 2439–2447. [DOI] [PubMed] [Google Scholar]
  5. Berdy, J. (2005) Bioactive microbial metabolites. Journal of Antibiotics (Tokyo), 58, 1–26. [DOI] [PubMed] [Google Scholar]
  6. Hess, B. , Bekker, H. , Berendsen, H.J.C. & Fraaije, J.G.E.M. (1997) LINCS: a Linear Constraint Solver for Molecular Simulations. Journal of Computational Chemistry, 18, 1463–1472. [Google Scholar]
  7. Black, P.N. , Dirusso, C.C. , Metzger, A.K. & Heimert, T.L. (1992) Cloning, sequencing, and expression of the fadD gene of Escherichia coli encoding acyl coenzyme A synthetase. The Journal of Biological Chemistry, 267, 25513–25520. [PubMed] [Google Scholar]
  8. Blumauerova, M. , Mateju, J. , Stajner, K. & Vanek, Z. (1977) Studies on the production of daunomycinone‐derived glycosides and related metabolites in Streptomyces coeruleorubidus and Streptomyces peucetius . Folia Microbiologia (Praha), 22, 275–285. [DOI] [PubMed] [Google Scholar]
  9. Chan, Y.A. , Podevels, A.M. , Kevany, B.M. & Thomas, M.G. (2009) Biosynthesis of polyketide synthase extender units. Natural Product Reports, 26, 90–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cox, J.A.G. , Taylor, R.C. , Brown, A.K. , Attoe, S. , Besra, G.S. & Futterer, K. (2019) Crystal structure of Mycobacterium tuberculosis FadB2 implicated in mycobacterial β‐oxidation. Acta Crystallographica Section D: Structural Biology, 75, 101–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. DeLano, W.L. (2002) PyMOL: An Open‐Source Molecular Graphics Tool. CCP4 Newsl Protein Crystallography, 40, 82–92. [Google Scholar]
  12. DiRusso, C.C. (1990) Primary sequence of the Escherichia coli fadBA operon, encoding the fatty acid‐oxidizing multienzyme complex, indicates a high degree of homology to eucaryotic enzymes. Journal of Bacteriology, 172, 6459–6468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dong, J. , Wei, J. , Li, H. , Zhao, S. & Guan, W. (2021) An efficient markerless deletion system suitable for the industrial strains of Streptomyces . Journal of Microbiology and Biotechnology, 31, 1722–1731. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Efthimiou, G. , Thumser, A.E. & Avignone‐Rossa, C.A. (2008) A novel finding that Streptomyces clavuligerus can produce the antibiotic clavulanic acid using olive oil as a sole carbon source. Journal of Applied Microbiology, 105, 2058–2064. [DOI] [PubMed] [Google Scholar]
  15. Elena Volodina1, A.S. (2014) (S)‐3‐hydroxyacyl‐CoA dehydrogenase/enoyl‐CoA hydratase (FadB’) from fatty acid degradation operon of Ralstonia eutropha H16. AMB Express, 4, 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fuchs, D. , Daniel, V. , Sadeghi, M. , Opelz, G. & Naujokat, C. (2010) Salinomycin overcomes ABC transporter‐mediated multidrug and apoptosis resistance in human leukemia stem cell‐like KG‐1a cells. Biochemical and Biophysical Research Communications, 394, 1098–1104. [DOI] [PubMed] [Google Scholar]
  17. Fujita, Y. , Matsuoka, H. & Hirooka, K. (2007) Regulation of fatty acid metabolism in bacteria. Molecular Microbiology, 66, 829–839. [DOI] [PubMed] [Google Scholar]
  18. Ghisla, S. & Thorpe, C. (2004) Acyl‐CoA dehydrogenases. A mechanistic overview. European Journal of Biochemistry, 271, 494–508. [DOI] [PubMed] [Google Scholar]
  19. Gumila, C. , Ancelin, M.L. , Delort, A.M. , Jeminet, G. & Vial, H.J. (1997) Characterization of the potent in vitro and in vivo antimalarial activities of ionophore compounds. Antimicrobial Agents and Chemotherapy, 41, 523–529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gust, B. , Challis, G.L. , Fowler, K. , Kieser, T. & Chater, K.F. (2003) PCR‐targeted Streptomyces gene replacement replacement identifies a protein domain needed for biosynthesis of the sesquiterpene soil odor geosmin. Proceedings of the National Academy of Sciences of the United States of America, 100, 1541–1546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hobbs, D.H. , Kim, H.J. , Chater, K.F. & Hills, M.J. (1997) Mutants of lipid synthesis in Streptomyces coelicolor . Biochemical Society Transactions, 25, S674. [DOI] [PubMed] [Google Scholar]
  22. Huerta‐Cepas, J. , Szklarczyk, D. , Heller, D. , Hernandez‐Plaza, A. , Forslund, S.K. , Cook, H. et al. (2019) eggNOG 5.0: a hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Research, 47, D309–D314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Humphrey, W. , Dalke, A. & Schulten, K. (1996) VMD: Visual molecular dynamics. Journal of Molecular Graphics, 14, 33–38. [DOI] [PubMed] [Google Scholar]
  24. Hutchinson, C.R. (1997) Biosynthetic studies of daunorubicin and tetracenomycin C. Chemical Reviews, 97, 2525–2536. [DOI] [PubMed] [Google Scholar]
  25. Hwang, K.S. , Kim, H.U. , Charusanti, P. , Palsson, B.O. & Lee, S.Y. (2014) Systems biology and biotechnology of Streptomyces species for the production of secondary metabolites. Biotechnology Advances, 32, 255–268. [DOI] [PubMed] [Google Scholar]
  26. Jiang, C. , Wang, H. , Kang, Q. , Liu, J. & Bai, L. (2012) Cloning and characterization of the polyether salinomycin biosynthesis gene cluster of Streptomyces albus XM211. Applied and Environmental Microbiology, 78, 994–1003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kallifidas, D. , Jiang, G. , Ding, Y. & Luesch, H. (2018) Rational engineering of Streptomyces albus J1074 for the overexpression of secondary metabolite gene clusters. Microbial Cell Factories, 17, 25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kanehisa, M. & Goto, S. (2000) KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research, 28, 27–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kautsar, S.A. , Blin, K. , Shaw, S. , Navarro‐Munoz, J.C. , Terlouw, B.R. , van der Hooft, J.J.J. et al. (2020) MIBiG 2.0: a repository for biosynthetic gene clusters of known function. Nucleic Acids Research, 48, D454–D458. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kim, D.E. , Chivian, D. & Baker, D. (2004) Protein structure prediction and analysis using the Robetta server. Nucleic Acids Research, 32, W526–W531. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim, J. , Chang, J.H. & Kim, K.J. (2014) Crystal structure and biochemical properties of the (S)‐3‐hydroxybutyryl‐CoA dehydrogenase PaaH1 from Ralstonia eutropha . Biochemical and Biophysical Research Communications, 448, 163–168. [DOI] [PubMed] [Google Scholar]
  32. Kuo, S.Z. , Blair, K.J. , Rahimy, E. , Kiang, A. , Abhold, E. , Fan, J.B. et al. (2012) Salinomycin induces cell death and differentiation in head and neck squamous cell carcinoma stem cells despite activation of epithelial‐mesenchymal transition and Akt. BMC Cancer, 12, 556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Larkin, M.A. , Blackshields, G. , Brown, N.P. , Chenna, R. , McGettigan, P.A. , McWilliam, H. et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics, 23, 2947–2948. [DOI] [PubMed] [Google Scholar]
  34. Letunic, I. & Bork, P. (2016) Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Research, 44, W242–W245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Li, C. , Hazzard, C. , Florova, G. & Reynolds, K.A. (2009) High titer production of tetracenomycins by heterologous expression of the pathway in a Streptomyces cinnamonensis industrial monensin producer strain. Metabolic Engineering, 11, 319–327. [DOI] [PubMed] [Google Scholar]
  36. Li, H. , Wei, J. , Dong, J. , Li, Y. , Li, Y. , Chen, Y. et al. (2020) Enhanced triacylglycerol metabolism contributes to efficient oil utilization and high‐level production of salinomycin in Streptomyces albus ZD11. Applied and Environmental Microbiology, 86, e00763‐20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Livak, K.J. & Schmittgen, T.D. (2001) Analysis of relative gene expression data using real‐time quantitative PCR and the 2(‐Delta Delta C[T]) Method. Methods, 25, 402–408. [DOI] [PubMed] [Google Scholar]
  38. Lu, C. , Zhang, X. , Jiang, M. & Bai, L. (2016) Enhanced salinomycin production by adjusting the supply of polyketide extender units in Streptomyces albus . Metabolic Engineering, 35, 129–137. [DOI] [PubMed] [Google Scholar]
  39. Malla, S. , Niraula, N.P. , Singh, B. , Liou, K. & Sohng, J.K. (2010) Limitations in doxorubicin production from Streptomyces peucetius . Microbiological Research, 165, 427–435. [DOI] [PubMed] [Google Scholar]
  40. Menendez‐Bravo, S. , Paganini, J. , Avignone‐Rossa, C. , Gramajo, H. & Arabolaza, A. (2017) Identification of FadAB complexes involved in fatty acid beta‐oxidation in Streptomyces coelicolor and construction of a triacylglycerol overproducing strain. Frontiers in Microbiology, 8, 1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mistry, J. , Chuguransky, S. , Williams, L. , Qureshi, M. , Salazar, G.A. , Sonnhammer, E.L.L. et al. (2021) Pfam: the protein families database in 2021. Nucleic Acids Research, 49, D412–D419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Olano, C. , Lombo, F. , Mendez, C. & Salas, J.A. (2008) Improving production of bioactive secondary metabolites in actinomycetes by metabolic engineering. Metabolic Engineering, 10, 281–292. [DOI] [PubMed] [Google Scholar]
  43. Peng, Q. , Gao, G. , Lu, J. , Long, Q. , Chen, X. , Zhang, F. et al. (2018) Engineered Streptomyces lividans strains for optimal identification and expression of cryptic biosynthetic gene clusters. Frontiers in Microbiology, 9, 3042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Schulz, H. (1991) Beta oxidation of fatty acids. Biochimica et Biophysica Acta, 1081, 109–120. [DOI] [PubMed] [Google Scholar]
  45. Shrestha, B. , Pokhrel, A.R. , Darsandhari, S. , Parajuli, P. , Sohng, J.K. & Pandey, R.P. (2019) Engineering Streptomyces peucetius for doxorubicin and daunorubicin biosynthesis. Environmental Chemistry for a Sustainable World, 26, 191–209. [Google Scholar]
  46. Staunton, J. & Weissman, K.J. (2001) Polyketide biosynthesis: a millennium review. Natural Product Reports, 18, 380–416. [DOI] [PubMed] [Google Scholar]
  47. Tamura, K. , Stecher, G. & Kumar, S. (2021) MEGA11: Molecular Evolutionary Genetics Analysis Version 11. Molecular Biology and Evolution, 38, 3022–3027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Tang, L. , Grimm, A. , Zhang, Y.X. & Hutchinson, C.R. (1996) Purification and characterization of the DNA‐binding protein DnrI, a transcriptional factor of daunorubicin biosynthesis in Streptomyces peucetius . Molecular Microbiology, 22, 801–813. [DOI] [PubMed] [Google Scholar]
  49. Taylor, R.C. , Brown, A.K. , Singh, A. , Bhatt, A. & Besra, G.S. (2010) Characterization of a beta‐hydroxybutyryl‐CoA dehydrogenase from Mycobacterium tuberculosis . Microbiology (Reading), 156, 1975–1982. [DOI] [PubMed] [Google Scholar]
  50. Trott, O. & Olson, A.J. (2010) AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Journal of Computational Chemistry, 31, 455–461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. van Keulen, G. , Siebring, J. & Dijkhuizen, L. (2011) Central carbon metabolic pathways in Streptomyces . In: Dyson, P. (Ed.) Streptomyces. Norfolk: Caister Academic Press, pp. 105–124. [Google Scholar]
  52. Vranken, A.W.S.d.S.W.F. (2012) ACPYPE ‐ AnteChamber PYthon Parser interfacE. BMC Research Notes, 5, 367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wang, W. , Li, X. , Wang, J. , Xiang, S. , Feng, X. & Yang, K. (2013) An engineered strong promoter for streptomycetes. Applied and Environmental Microbiology, 79, 4484–4492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wang, H. , Li, Z. , Jia, R. , Hou, Y. , Yin, J. , Bian, X. et al. (2016) RecET direct cloning and Redαβ recombineering of biosynthetic gene clusters, large operons or single genes for heterologous expression. Nature Protocols, 11, 1175–1190. [DOI] [PubMed] [Google Scholar]
  55. Wang, J. , Liu, H. , Huang, D. , Jin, L. , Wang, C. & Wen, J. (2017) Comparative proteomic and metabolomic analysis of Streptomyces tsukubaensis reveals the metabolic mechanism of FK506 overproduction by feeding soybean oil. Applied Microbiology and Biotechnology, 101, 2447–2465. [DOI] [PubMed] [Google Scholar]
  56. Wang, T.J. , Shan, Y.M. , Li, H. , Dou, W.W. , Jiang, X.H. , Mao, X.M. et al. (2017) Multiple transporters are involved in natamycin efflux in Streptomyces chattanoogensis L10. Molecular Microbiology, 103, 713–728. [DOI] [PubMed] [Google Scholar]
  57. Wang, W. , Li, S. , Li, Z. , Zhang, J. , Fan, K. , Tan, G. et al. (2020) Harnessing the intracellular triacylglycerols for titer improvement of polyketides in Streptomyces . Nature Biotechnology, 38, 76–83. [DOI] [PubMed] [Google Scholar]
  58. Yang, S.Y. , Li, J.M. , He, X.Y. , Cosloy, S.D. & Schulz, H. (1988) Evidence that the fadB gene of the fadAB operon of Escherichia coli encodes 3‐hydroxyacyl‐coenzyme‐A (CoA) epimerase, delta‐3‐cis‐delta 2‐trans‐enoyl‐CoA isomerase, and enoyl‐CoA hydratase in addition to 3‐hydroxyacyl‐CoA dehydrogenase. Journal of Bacteriology, 170, 2543–2548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Yang, S.Y. , Yang, X.Y. , Healy‐Louie, G. , Schulz, H. & Elzinga, M. (1990) Nucleotide sequence of the fadA gene. Primary structure of 3‐ketoacyl‐coenzyme A thiolase from Escherichia coli and the structural organization of the fadAB operon. Journal of Biological Chemistry, 265, 10424–10429. [PubMed] [Google Scholar]
  60. Zhang, X. , Lu, C. & Bai, L. (2017) Conversion of the high‐yield salinomycin producer Streptomyces albus BK3‐25 into a surrogate host for polyketide production. Science China Life Sciences, 60, 1000–1009. [DOI] [PubMed] [Google Scholar]
  61. Zhu, Z. , Li, H. , Yu, P. , Guo, Y. , Luo, S. , Chen, Z. et al. (2017) SlnR is a positive pathway‐specific regulator for salinomycin biosynthesis in Streptomyces albus . Applied Microbiology and Biotechnology, 101, 1547–1557. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1

Click here for additional data file. (6.4MB, docx)

Appendix S1

Click here for additional data file. (288.4KB, pdf)

Data Availability Statement

Genomic sequences of S. albus ZD11 (CP033071.1), S. albidoflavus J1074 (CP004370.1), S. coelicolor A3(2) (AL645882.2) and S. lividans TK24 (NZ_CP009124.1) were obtained from GenBank in GenBank format. The sequence reads obtained by RNA‐seq were obtained from in the SRA database under accession numbers SAMN12251778, SAMN12251779, SAMN12251780 and SAMN12251781, which were described in our previous work (Li et al., 2020).

Articles from Microbial Biotechnology are provided here courtesy of Wiley

RESOURCES