Enhanced spinosad production in Saccharopolyspora spinosa by employing mannose as an extracellular carbon reservoir and optimizing acetyl-CoA supply pathway

Abstract

Spinosad, a potent broad-spectrum insecticidal polyketide produced by Saccharopolyspora spinosa, faces limitations in industrial-scale production due to inherent inefficiencies in its native biosynthetic pathways. To overcome this constraint, we devised and implemented a systematic, stepwise combinatorial strategy integrating classical strain improvement, fermentation optimization, and rational genetic engineering to augment spinosad titres. The approach commenced with iterative UV mutagenesis, yielding a stable high-producing mutant, U7, which demonstrated a significant elevation in spinosad production compared to the original strain. Subsequent fermentation optimization via single-factor shake-flask experiments identified mannose as a superior extracellular carbon source over glucose, markedly enhancing mutant U7 spinosad titres. To further augment metabolic efficiency, we employed rational genetic engineering, demonstrating that deletion of manB (beta-mannosidase) and leuA (2-isopropylmalate synthase) coupled with overexpression of pdhC (pyruvate dehydrogenase subunit) synergistically boosted spinosad biosynthesis. By integrating these modifications into mutant U7, we achieved a final spinosad titer of 537.6 mg/L—a 6.1-fold increase over the wild-type strain. This study presents a combinatorial metabolic engineering approach that not only significantly improves spinosad production but also provides a generalizable framework for optimizing polyketide biosynthesis in Streptomyces and related actinomycetes. Critically, the identification of mannose as the preferred carbon source not only directly enhanced precursor supply but also proved pivotal in redirecting metabolic flux through the glycolytic pathway, thereby generating elevated levels of acetyl-CoA. The acetyl-CoA pathway engineering strategy developed here can be readily adapted to enhance the production of other high-value polyketides.

Graphical abstract

Mannose-based metabolic flux reprogramming enabled enhanced acetyl-CoA supply and significantly improved spinosad production in Saccharopolyspora spinosa.

1. Introduction

Spinosad, a potent insecticidal secondary metabolite produced by Saccharopolyspora spinosa (S. spinosa), exhibits broad-spectrum pesticidal activity [1]. However, the yield of spinosad produced by wild-type S. spinosa strains obtained from the natural environment is often low. In recent years, researchers have adopted various strategies to enhance the ability of wild-type S. spinosa to produce polybactericides. Among them, random mutation, metabolic engineering of the precursor supply pathway and medium optimization are the most commonly used and effective strategies [2]. For example, Zhu et al. [3] enhanced spinosad production to 858.3 ± 27.7 mg/L by applying a combination mutagenesis of ARTP and NTG, corresponding to a 5.12-fold increase compared with the contemporaneous parental strain. Wang et al. [4] and Zhu et al. [5] reprogrammed primary metabolic pathways in S. spinosa using CRISPRi-based strategies, thereby achieving a better balance between cellular growth and secondary metabolism and increasing spinosad production to 2.9-fold and 1.9-fold, respectively. Focusing on intracellular triacylglycerol degradation, Cao et al. [6,7] co-overexpressed two key genes involved in triacylglycerol metabolism and three critical genes associated with fatty acid degradation in S. spinosa. When combined with culture medium optimization, these strategies resulted in 5.5-fold and 23.6-fold (1293.43 mg/L) improvements in spinosad production, respectively.

To date, numerous studies have integrated strategies such as mutagenesis-based strain improvement, overexpression of genes encoding rate-limiting enzymes, deletion of genes in competing or bypass pathways, and optimization of fermentation media to engineer the entire metabolic network of target natural products, an approach commonly referred to as combinatorial engineering [[8], [9], [10]]. The synergistic application of multiple metabolic strategies enables more precise optimization of complex biosynthetic pathways and has been demonstrated to exert a significant impact on the production of target natural products [11]. As a polyketide compound, spinosad is biosynthesized through multiple metabolic pathways and enzymatic reactions, making it particularly amenable to improvement through combinatorial metabolic engineering. For instance, Wang et al. [12] systematically combined rhamnose precursor overexpression, biosynthetic gene cluster amplification, enhancement of short-chain acyl-CoA supply, and genome reduction using a genome-scale model-guided framework, constructed a high-producing strain, NHF132-BAC-SP43-NCM, which achieved a spinosad titer of 1816.8 mg/L, representing a 553.3 % increase over the parental strain. Xia et al. [13] significantly improved spinosad production through a combination of genome shuffling, ultraviolet mutagenesis, and metabolic engineering of fatty acid degradation pathways, achieving a titer of 1120 ± 108 mg/L, approximately 12-fold higher than that of the original strain. Li et al. [14] further demonstrated that combining ultraviolet mutagenesis with the introduction of an exogenous NCM pathway (pyruvate→3-oxopropionate ester→malonyl-CoA) effectively optimized the intracellular pool of short-chain acyl-CoAs, resulting in a 11.2-fold increase in spinosad production and a final titer of 4637 ± 18 mg/L. Although recent advances have enabled certain high-performance engineered strains to achieve spinosad titers exceeding 1 g/L, and in some cases approaching 5 g/L, further improvements are still required to enable economically competitive industrial production [2]. Therefore, the identification of novel genetic targets that promote spinosad biosynthesis, as well as the development of more effective combinatorial engineering strategies, remains of significant importance.

The biosynthesis of polyketides in actinomycetes is fundamentally dependent on the condensation of short-chain acyl-CoA precursors, with acetyl-CoA, malonyl-CoA, and methylmalonyl-CoA serving as the primary building blocks for polyketide backbone assembly [15,16]. Acetyl-CoA, occupying a central position in central metabolism, acts as the critical entry point for polyketide biosynthesis [17]. Glucose catabolism, primarily via glycolysis and the tricarboxylic acid (TCA) cycle, constitutes a major source of acetyl-CoA. Crucially, acetyl-CoA is the direct precursor for malonyl-CoA synthesis, catalyzed by acetyl-CoA carboxylase (ACC), and can be indirectly converted to methylmalonyl-CoA through intermediates such as succinyl-CoA [18,19]. Consequently, the availability and metabolic flux of acetyl-CoA are rate-limiting determinants for polyketide titers. For instance, in Streptomyces rimosus M527, malonyl-CoA serves as a key precursor for rimocidin biosynthesis. Overexpression of accsr, which encodes acetyl-CoA carboxylase (ACC), increased ACC enzymatic activity and intracellular malonyl-CoA levels by 1.0-fold and 1.5-fold, respectively, resulting in a 34 % enhancement in rimocidin production [20]. Methylmalonyl-CoA biosynthesis represents a rate-limiting step in FK506 production. Overexpression of pccB1, encoding propionyl-CoA carboxylase, in combination with supplementation of isoleucine and valine, increased the FK506 titer to 929.6 mg/L, corresponding to a 56.6 % improvement over the parental strain [21]. Similarly, Tian et al. [22] employed CRISPRi in Streptomyces rapamycinicus to downregulate competing pathways consuming acetyl-CoA, thereby diverting flux towards rapamycin biosynthesis and yielding a 6.6-fold titer improvement. Notably, high-yielding mutants obtained from mutagenesis such as NT24 [3] and YX2 [6], consistently exhibit elevated acetyl-CoA pools, underscoring the pivotal role of precursor supply in spinosad biosynthesis. Collectively, these studies demonstrate that optimizing the supply of acyl-CoA precursors represents a critical strategy for enhancing the biosynthesis of polyketide natural products, including spinosad.

Our study reveals a previously unrecognized link between mannose metabolism and acetyl-CoA homeostasis in spinosad biosynthesis. Initial isolation of a UV-mutagenized strain (U7) with enhanced spinosad production uncovered its preferential utilization of mannose over glucose—a finding with profound metabolic implications. Subsequent engineering targeted three key nodes: (1) disruption of manB to block mannose-1-phosphate conversion, (2) knockout of leuA to reduce the competitive utilization of acetyl-CoA, and (3) overexpression of pdhC to strengthen pyruvate dehydrogenase activity. The resulting strain U7-ΔmanBΔleuA:pdhC achieved a spinosad titer of 537.6 mg/L, representing a 6.1-fold improvement over parental strains. Comparative proteomics demonstrated that this combinatorial strategy uniquely rewires central carbon metabolism by simultaneously upregulating glycolysis, fatty acid β-oxidation, and TCA cycle activity—all converging to amplify acetyl-CoA availability. Importantly, our work establishes mannose as a superior carbon source for spinosad production, as mannose metabolism more efficiently channels carbon flux through glycolysis and pyruvate dehydrogenase, resulting in enhanced acetyl-CoA availability for polyketide biosynthesis. These findings provide both a practical blueprint for spinosad overproduction and a conceptual framework for leveraging alternative carbon routes to enhance polyketide biosynthesis in Streptomyces.

2. Materials and methods

2.1. Bacterial strains, plasmids, media and growth conditions

The bacterial strains and plasmids used in this study are listed in Table S1. The media and growth conditions used in this study are described in the Supplementary Materials and Methods.

2.2. The UV mutagenesis and screening of S. spinosa

To determine the proper conditions for UV mutagenesis, 3 mL of diluted S. spinosa spore suspension (10⁶/mL) was transferred into a sterilized 60-mm disposable plate and subsequently exposed to UV irradiation in a dark room from UV lamp with a wavelength of 254 nm and 30 W, and then cultivated at 30 °C in the dark for 7 days to recover viable colonies. To determine the UV mutagenesis lethality and positive mutation rates, various treatment times (0, 25, 50, 75, 100, 125, 150 s) were measured. 3 mL of the bacterial suspension treated with UV irradiation for 50 s was activated for 2 h and then diluted to 50–100 cfu/mL. The diluted culture was subjected to primary screening by measuring the OD₆₀₀ using a single-cell microliter-droplet culture omics system (MISS cell culture omics, Luoyang Huaqing Tianmu Biotechnology Co., Ltd.). A total of 227 droplets with OD values higher than 3.4 were collected. The top 60 droplets with the highest OD values were selected in descending order for scale-up cultivation and fermentation, followed by HPLC analysis, to screen for mutant strains with spinosad production higher than that of the parental strain. The mutant strain that produced the highest spinosad was subjected to 5 generations of subculture in a 2 L flask (the transduction was performed once every 5 days, during the stable phase), and its spinosad titer was determined by HPLC to ensure the genetic stability.

2.3. Single-factor medium optimization of mutant U7

Based on the original production medium, 1 g/L KNO₃, 0.5 g/L K₂HPO₄·3H₂O, 0.5 g/L MgSO₄·7H₂O, 0.01 g/L FeSO₄, 4 g/L tryptone and 4 g/L yeast extract were set as the constant components, and the effects of various carbon sources on the production of spinosad by mutant U7 were then evaluated. Four common six carbon monosaccharides, including the mannose, fructose, glucose and galactose, were tested at the concentration of 20 g/L, respectively. After the optimal carbon sources has been assigned, its optimal dosages have been further determined to finally formulate the optimized production medium, which has been applied in the subsequent fermentation studies.

2.4. Construction of the single gene modified engineering strain

Taking the construction of leuA deletion and overexpression engineering strain as an example (Fig. S1). The leuA was deleted using CRISPR/Cas9-mediated gene editing, and specific measures are as follows. The upstream and downstream homologous arms of leuA were amplified with primers leuA-U-F/R and leuA-D-F/R, respectively. The sgRNA of leuA was designed by CRISPick (https://portals. broadinstitute.org/gppx/crispick/public) and fused with upstream and downstream homologous arms of the leuA. Then, the fusion fragment was digested with Spe I and Hind III and cloned into the corresponding restriction sites of pKCcas9dO by T4 DNA ligase, yielding the deletion plasmid pKCcas9dleuA. Finally, this plasmid was transformed into S. spinosa by protoplast transformation and screened by 50 μg/mL apramycin [23]. The leuA locus of several Apra^R colonies was analyzed by PCR using primers leuA-1-F/leuA-1-R and by sequencing, and the transformant that was successfully verifed was named mutant ΔleuA. For leuA overexpression, plasmid pOJ260 was selected as the editing tool. The whole ORF of leuA and strong promoter ermEp∗ were amplified with primers leuA-F/R and leuA-ermEp-F/R, respectively. The fusion fragment of ermEp∗ and leuA was digested with Xba I and Hind III and cloned into the corresponding restriction sites of pOJ260, yielding the overexpressed plasmid pOJ260-ermEp∗-leuA. Then, this plasmid was transformed into S. spinosa by protoplast transformation and screened by 50 μg/mL apramycin. The chromosome structure of several Apra^R colonies was analyzed by PCR using primers apr-F/R, and the transformant that was successfully verified was namedleuA. The primers used in this study are listed in Table S1.

2.5. Construction of the acyl-CoAs supply pathway optimized engineering strain

To improve spinosad biosynthesis by optimizing acetyl-CoA supply pathway in mutant U7, we plan simultaneous manB and leuA deletion and overexpression of pdhC. The pKCcas9dO is a temperature sensitive plasmid, it cannot be replicated in S. spinosa at over 37 °C. Based on this function, we can complete the deletion of manB and leuA by using knockout plasmids pKCcas9dmanB and pKCcas9dleuA through two round operation, and obtained the mutant U7-ΔmanBΔleuA without apramycin resistance gene. Subsequently, the overexpressed plasmid pOJ260-ermEp∗-pdhC was transformed into U7-ΔmanBΔleuA by protoplast transformation and screened by 50 μg/mL apramycin, and the transformant that was successfully verified was named U7-ΔmanBΔleuA:pdhC.

2.6. Cultivation profile analysis of the wild-type strain and its derivative strains

Collect samples on day 10 to determine the spinosad production by using an HPLC 1290 system (Agilent, USA). The samples were cultured with methanol at a 1:1 volumetric ratio, incubated for 12 h, and then, the methanol was removed by a freeze concentrator. An equal volume of ethyl acetate was added for shaking treatment for approximately 5 h. After centrifugation and standing for 30 min, the ethyl acetate layer was removed by a freeze concentrator. Finally, the sample was dissolved in 100 μL methanol. A 20 μL aliquot of each supernatant was loaded onto a C18 column (AQ12S05-1546WT, YMC, Kyoto, Japan) and eluted with elution bufer (bufer A: 10 % (v/v) acetonitrile, bufer B: 90 % (v/v) methanol) at 1.0 mL/min, and the program ran for 30 min. The detection wavelength was set to 250 nm during the analysis. Spinosad, which was previously purified and identified in our laboratory, was used as an internal standard. An optical density of 600 nm (OD₆₀₀) was used to determine the cell concentration. Cells were collected every 24 h during fermentation for growth curve measurements. Supernatants were collected every 24 h during fermentation to determine the reducing sugar concentration by using a assay kit until the reducing sugar was fully consumed. Collect samples on day 2 and day 4 to measure the concentration of acetyl-CoA using the acetyl-CoA assay kit (Shanghai FANKEL Industrial Co., Ltd.). All of the experiments were performed in triplicate.

2.7. Extraction, preparation and LC–MS/MS analysis of whole proteins

The cells of the mutants U7-ΔmanBΔleuA::pdhC and U7 were harvested after 6 days of culture, washed four times by resuspending the cell pellet in 20 mL fresh PBS (10 mM, pH 7.8, prechilled at 4 °C), and quickly frozen in liquid nitrogen (three independent repeats). Protein extraction and 2D LC-MS/MS analysis were performed as described previously [24,25]. The utilized protein database was the protein sequence set of all Saccharopolyspora strains. The differentially abundant proteins (DAPs) were defined in the iTraq experiment by the following criteria: P value < 0.01, and fold change >1.5 or < 0.67. The metabolic pathway analysis of DAPs was conducted using the KEGG database (http://www.genome.jp/kegg/) [26].

2.8. RNA isolation, cDNA synthesis, and quantitative reverse transcription PCR (qRT–PCR)

The total RNAs of the wild-type and engineering strains were separately isolated by using TRIzol Reagent (Sangon Biotech, Shanghai). cDNA synthesis was performed by using a HighCapacity cDNA Archive Kit (Fermentas). Real-time qPCR amplification was performed using Power SYBRR Green PCR Master Mix (Applied Biosystems) as previously described. The primer sequences used in qRT-PCR were designed with Primer Premier 5.0 and are listed in Table S2. The transcript generated from the 16S rRNA gene was used for normalization. The results are expressed as the means from three replicate experiments.

3. Results

3.1. Enhancement of spinosad production through UV mutagenesis

To address the limited spinosad biosynthesis capacity in wild-type S. spinosa, we employed UV mutagenesis to enhance its production. Based on the UV lethality curve analysis (Fig. 1A), a 50-s exposure was selected as the optimal mutagenesis duration for generating high-yield mutants. Spinosad titers were quantified by HPLC using standard curves generated from authentic spinosad (Merck) (Fig. S2). The obtained mutant strain was subjected to the MISS cell culture omics, and a total of 1488 individual droplets were generated and cultivated for 4.5d independently. Based on optical density measurements, 227 droplets with OD₆₀₀ values higher than 3.4 were selected as growth-positive candidates (Fig. S3). From these enriched droplets, the top 60 candidates (ranked by OD₆₀₀ from high to low) were subsequently recovered, expanded in shake flasks, and quantitatively analyzed for spinosad production by HPLC. Among these, mutant U7 demonstrated the most significant improvement, which was 2.6-fold higher than that of wild-type strain (Fig. 1B). Additionally, U7 was confirmed to exhibit stable spinosad production after five subcultures (Fig. S4).

Fig. 1.

A physiological and biochemical comparison analysis of U7 and S. spinosa. (A) The lethal rate of different UV treatment times on strain S. spinosa. (B) Comparison of spinosad production in mutant strains produced by UV treatment. (C) Comparison of bacterial cell densities between U7 and S. spinosa. (D) The content of acetyl-CoA in U7 and S. spinosa at 48 and 96 h. Error bars are calculated from three independent determinations in each sample. Univariate variance, ^nsP > 0.05, ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.001.

Further characterization revealed that mutant U7 possessed significantly enhanced growth, maximum cell density, and acetyl-CoA synthesis capacity relative to the wild-type strain (Fig. 1C and D), suggesting these physiological improvements underlie its elevated spinosad biosynthesis capability.

3.2. Mannose replacing glucose as extracellular carbon reservoir could promote spinosad biosynthesis in mutant U7

Recognizing that the carbon source critically governs precursor availability and metabolic flux for polyketide biosynthesis, we systematically evaluated alternative carbon substrates (glucose, mannose, fructose, galactose) in the fermentation medium to enhance spinosad production in the high-yielding mutant U7. Comparative analysis revealed that mannose was the optimal sugar source for spinosad biosynthesis in mutant U7 (Fig. 2A). Notably, mutant U7 exhibited superior acetyl-CoA and malonyl-CoA production when cultured with mannose as the sole sugar source (Fig. 2B and C), suggesting efficient carbon flux redirection toward spinosad biosynthesis. In addition, when cultured with mannose, the growth states of mutant U7 at day 4 were comparable to those obtained with glucose (Fig. S5), but its transcriptional profiling of key mannose metabolism genes manA, pfkA, pyk and pdhA significantly upregulated in this period (Fig. 2D). This genetic reprogramming likely enhances the strain's ability to channel mannose-derived carbon into spinosad synthesis. Further analysis identified two competing metabolic branches in mannose utilization: (1) the GDP-mannose pathway, governed by manB, and (2) the leucine biosynthesis pathway, regulated by leuA (Fig. 2E). Both pathways exhibited elevated expression, potentially diverting acetyl-CoA—a critical spinosad precursor—toward alternative metabolic sinks. These findings suggest that fine-tuning acetyl-CoA allocation is essential to maximize spinosad yields. Based on qRT-PCR data, pdhC, manB, and leuA were prioritized as targets for further metabolic engineering to optimize carbon flux toward spinosad production.

Fig. 2.

Mannose can increase the spinosad production by promoting the supply capacity of U7 acetyl-CoA. (A) Comparative analysis of the effects of glucose, mannose, fructose, and galactose on spinosad synthesis in U7. (B) A comparison of acetyl-CoA levels in U7 when mannose and glucose were used as the sole sugar sources respectively. (C) A comparison of malonyl-CoA levels in U7 when mannose and glucose were used as the sole sugar sources respectively. (D) Analysis of the transcriptional levels of several important functional genes in the acetyl-CoA synthesis pathway. (E) Analysis of the metabolic network diagram of U7 when using mannose as the carbon source. Error bars are calculated from three independent determinations in each sample. Univariate variance, ^nsP > 0.05, ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.001.

3.3. Impact of pdhC and leuA genes expression level alterations on strain growth and spinosad biosynthesis

While previous studies have established that manB deletion enhances spinosad biosynthesis [27], the specific regulatory roles of pdhC (pyruvate dehydrogenase subunit) and leuA (2-isopropylmalate synthase) in this pathway remained unclear. To elucidate their functions, we constructed S. spinosa mutants harboring targeted deletions or overexpression of pdhC and leuA, respectively. These strains were then characterized to systematically assess the impact of these genetic perturbations on cellular metabolism and spinosad production.

Cell density measurements revealed distinct growth patterns among the engineered strains compared to wild-type S. spinosa. Both the pdhC-overexpressing strain (::pdhC) and the leuA knockout strain (ΔleuA) consistently exhibited higher strain density than the wild-type strain throughout cultivation. In contrast, the pdhC knockout strain (ΔpdhC) and leuA-overexpressing strain (::leuA) showed significantly reduced strain density (Fig. 3A and D). Furthermore, detections of acetyl-CoA and malonyl-CoA revealed that overexpression of pdhC and deletion of leuA significantly enhanced the supply of these two short chain acyl-CoAs, redirecting more of them towards spinosad production (Fig. 3B, C, E, F). Notably, these growth phenotypes correlated directly with spinosad production. As anticipated, ::pdhC and ΔleuA demonstrated remarkable spinosad yields, reaching 2.11-fold and 1.57-fold of wild-type strain levels, respectively. Conversely, ΔpdhC and ::leuA produced substantially lower spinosad titers (Fig. 3G). Furthermore, the insecticidal activity against H. armigera of ::pdhC and ΔleuA increased significantly (Fig. 3H and I), with the median lethal time (LT₅₀) reduced by 1.15 days and 0.57 days, respectively (Table 1). To elucidate how changes in the expression levels of pdhC and leuA affect spinosad production, we compared the ability to supply acetyl-CoA among the different strains. These findings demonstrate that the overexpression of pdhC and the deletion of leuA enhance both strain growth and spinosad biosynthesis by increasing the supply of acetyl-CoA, highlighting their crucial roles in redirecting metabolic flux toward spinosad production.

Fig. 3.

The functional study of pdhC and leuA in regulating the supply of spinosad biosynthesis precursors in S. spinosa. (A) Comparison of bacterial cell densities between S. spinosa and engineered strains with pdhC deletion and overexpression. (B) The content of acetyl-CoA in S. spinosa and engineered strains with pdhC deletion and overexpression at 48 and 96 h. (C) The content of malonyl-CoA in S. spinosa and engineered strains with pdhC deletion and overexpression at 48 and 96 h. (D) Comparison of bacterial cell densities between S. spinosa and engineered strains with leuA deletion and overexpression. (E) The content of acetyl-CoA in S. spinosa and engineered strains with leuA deletion and overexpression at 48 and 96 h. (F) The content of malonyl-CoA in S. spinosa and engineered strains with leuA deletion and overexpression at 48 and 96 h. (G) Comparison of spinosad production in engineered strains ::pdhC, ΔleuA, ΔpdhC, ::leuA and wild-type strain S. spinosa. (H) Survival curve of H. armigera under the action of the fermentation broth of pdhC genetically engineered strains. (I) Survival curve of H. armigera under the action of the fermentation broth of leuA genetically engineered strains. Error bars are calculated from three independent determinations in each sample. Univariate variance, ^nsP > 0.05, ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.001.

Table 1.

Determination of insecticidal activity of pdhC and leuA genetically engineered strain fermentation broth against H. armigera.

Strains	LT50 value(d)	95 % confidence interval
S. spinosa	4.838	4.550–5.180
ΔpdhC	5.643	5.240–6.200
::pdhC	3.693	3.455–3.936
ΔleuA	4.273	4.0164.554
::leuA	5.381	5.033–5.836

3.4. Combinatorial optimization of pdhC, manB and leuA expression improves spinosad biosynthesis in mutant U7

Informed by our prior functional analysis demonstrating the regulatory roles of manB, leuA, and pdhC, we designed a targeted gene circuit strategy to further enhance spinosad production in mutant U7. This strategy involves the concurrent disruption of competing metabolic pathways (manB, leuA deletion) coupled with the upregulation of pdhC expression to augment precursor flux. To implement this multiplexed genetic modification efficiently in S. spinosa, we employed the temperature-sensitive CRISPR-Cas9 vector pKCcas9dO, which enables precise combinatorial genome editing in Streptomyces. Following two rounds of CRISPR-Cas9-mediated editing, we successfully generated the strain U7-ΔmanBΔleuA, which lacks the apr resistance gene and carries deletions in both manB and leuA. Subsequently, pdhC was placed under the control of the strong constitutive promoter ermEp∗ and integrated into U7-ΔmanBΔleuA, resulting in the final three-gene modified strain, designated U7-ΔmanBΔleuA::pdhC (Fig. 4A). The expression level of pdhC was verified by qRT-PCR (Fig. S6).

Fig. 4.

Metabolic engineering of mannose utilization in strain U7 and its impact on spinosad production. (A) Schematic representation of the optimized mannose metabolic pathway in U7. (B) Biomass comparison of U7-ΔmanBΔleuA::pdhC and U7. (C) Acetyl-CoA content comparison of U7-ΔmanBΔleuA::pdhC and U7. (D) Malonyl-CoA content comparison of U7-ΔmanBΔleuA::pdhC and U7. (E) spinosad production comparison of U7-ΔmanBΔleuA::pdhC and U7. Error bars are calculated from three independent determinations in each sample. Univariate variance, ^nsP > 0.05, ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.001.

To assess the impact of combinatorial gene optimization on strain U7, we analyzed biomass accumulation and phenotypic differences between the parental strain U7 and the engineered mutant U7-ΔmanBΔleuA::pdhC. The modified strain exhibited a 25.8 % increase in maximum biomass (7.94 g/L vs. 6.31 g/L in U7) and demonstrated accelerated growth rate during the logarithmic phase (Fig. 4B). Furthermore, U7-ΔmanBΔleuA::pdhC showed enhanced acetyl-CoA and malonyl-CoA accumulations during logarithmic growth (Fig. 4C and D), which, as anticipated, supported increased spinosad biosynthesis in stationary phase. The titer cumulation detection showed that, the engineered strain exhibited consistently higher spinosad accumulation throughout the fermentation process except for the fourth day of fermentation, and final spinosad titers reached 537.6 mg/L in the engineered strain, representing a 1.88-fold improvement over U7 and a 6.1-fold increase compared to the wild-type strain (Fig. 4E). These results demonstrate that utilizing mannose as an extracellular carbon source, coupled with coordinated regulation of pdhC, manB, and leuA expression, significantly enhances spinosad production efficiency.

3.5. The mechanism underlying the promoting effect of pdhC, manB and leuA combinatorial optimization on spinosad production

To mechanistically dissect how combinatorial engineering of pdhC, manB, and leuA enhances spinosad biosynthesis, we performed comparative proteomics on mutant U7 and the engineered strain U7-ΔmanBΔleuA::pdhC using mid-production phase (144 h) protein samples. Among 3367 reproducibly quantified proteins, 1995 proteins were differentially expressed (fold change >1.5 or < 0.67, p < 0.01) in the engineered strain, with 1071 upregulated proteins and 924 downregulated proteins (Fig. 5A, DataSet S1). Functional annotation via KEGG pathway analysis revealed that these differentially expressed proteins were primarily associated with core metabolic processes, including amino acid biosynthesis, carbon metabolism, glycolysis, and the TCA cycle (Fig. 5B). This systemic modulation of central carbon metabolism underscores the pleiotropic effects of pdhC overexpression coupled with manB and leuA deletion, demonstrating their coordinated role in redirecting metabolic flux toward spinosad biosynthesis.

Fig. 5.

Mechanistic analysis of enhanced spinosad biosynthesis in engineered strain U7-ΔmanBΔleuA::pdhC. (A) Quantitative comparison of differentially expressed proteins between U7-ΔmanBΔleuA::pdhC and parental strain U7. The volcano plot illustrates the distribution of up- and down-regulated proteins. (B) KEGG pathway enrichment analysis of up-regulated proteins in U7-ΔmanBΔleuA::pdhC. (C) Schematic representation of protein expression profiles in key metabolic pathways, including glycolysis, pentose phosphate pathway, fatty acid degradation, and TCA cycle. Up-regulated and down-regulated proteins are highlighted in red and blue, respectively.

A detailed metabolic network diagram was constructed to reveal the specific metabolic differences between the mutants U7-ΔmanBΔleuA::pdhC and U7 based on the comparative proteomic data. Key enzymes involved in mannose transport and acetyl-CoA biosynthesis—including phosphoenolpyruvate phosphotransferase (PtsI), fructose-bisphosphate aldolase (FbaA), glyceraldehyde-3-phosphate dehydrogenase (GapN), enolase (Eno), pyruvate kinase (Pyk), and pyruvate dehydrogenase subunits (PdhA, PdhB, PdhC)—were significantly upregulated in U7-ΔmanBΔleuA::pdhC (Fig. 5C). This systematic enhancement of glycolytic and pyruvate dehydrogenase activity demonstrates the targeted genetic modifications effectively potentiated acetyl-CoA generation from mannose. Conversely, enzymes in the pentose phosphate (PP) pathway—glucose-6-phosphate dehydrogenase (Zwf) and transaldolase (TalA)—were downregulated in U7-ΔmanBΔleuA::pdhC. We hypothesize this suppression reflects metabolic prioritization, where reduced intracellular mannose availability redirects flux from PP pathway toward glycolysis to maximize acetyl-CoA production. Notably, 19 of 28 differentially expressed proteins in fatty acid degradation were upregulated in the engineered strain, including acetyl-CoA C-acetyltransferase. Similarly, TCA cycle rate-limiting enzymes citrate synthase (GltA) and isocitrate dehydrogenase (Icd-2) exhibited elevated expression (Fig. 5C). These coordinated changes indicate U7-ΔmanBΔleuA::pdhC possesses enhanced catabolic capacity for both fatty acid β-oxidation and oxidative phosphorylation, providing not only increased acetyl-CoA pools but also critical biosynthetic precursors (e.g., α-ketoglutarate, succinyl-CoA). Collectively, these proteomic findings mechanistically explain the superior biomass accumulation and spinosad production in U7-ΔmanBΔleuA::pdhC, demonstrating how targeted pathway optimization can reprogram central metabolism to favor secondary metabolite biosynthesis.

4. Discussion

The biosynthesis of secondary metabolites critically depends on precursor supply from primary metabolic pathways [23,28,29]. As a polyketide-derived secondary metabolite produced by S. spinosa, spinosad production is fundamentally constrained by the availability of key precursors, particularly acetyl-CoA [13]. Spinosad is a secondary metabolite produced by Saccharopolyspora spinosa, so enhancing the supply of precursors is an efficient way to increase its production. Combinatorial metabolic engineering has emerged as a powerful strategy for enhancing natural product biosynthesis by simultaneously optimizing multiple metabolic nodes [30]. In the present study, we implemented an integrated strain improvement approach combining: (1) UV mutagenesis to generate genetic diversity, (2) systematic fermentation optimization to enhance precursor flux, and (3) targeted genetic circuit modification to redirect metabolic pathways. This multi-level engineering strategy was designed to synergistically overcome the inherent metabolic bottlenecks limiting spinosad production.

Mutagenesis breeding coupled with fermentation optimization represents a well-established and efficient strategy for rapidly improving microbial strains with enhanced secondary metabolite production, as extensively documented in Streptomyces species [[31], [32], [33]]. Notably, Li et al. demonstrated that this combinatorial approach achieved a 7.8-fold increase in rabelomycin titer compared to the parental strain [34], validating its effectiveness for polyketide biosynthesis. Therefore, this strategy was first adopted in this study to improve the spinosad production. Through iterative cycles of UV mutagenesis and high-throughput screening, we isolated mutant U7, which exhibited the highest spinosad yield among all generated variants. By conducting acetyl-CoA concentration analysis, we proposed that the enhancement of precursor supply capacity could be the main reason for the higher production of spinosad in mutant U7. Further fermentation optimization demonstrated that mannose was identified as a superior carbon source for spinosad production in the UV-mutagenized strain U7, and this advantage can be explained by its impact on central carbon metabolism and precursor supply. Compared with glucose, mannose utilization resulted in significantly elevated intracellular acetyl-CoA levels, which is the key rate-limiting precursor for spinosad biosynthesis. This metabolic phenotype was accompanied by coordinated transcriptional upregulation of genes involved in glycolysis and acetyl-CoA generation, including manA, pfkA, pyk, and pdhA. Mannose is assimilated through a distinct metabolic entry route that enables efficient conversion into fructose-6-phosphate and subsequent integration into the glycolytic pathway. In mutant U7, enhanced expression of manA likely strengthens mannose assimilation, while upregulation of pfkA and pyk promotes carbon flux toward pyruvate. Concurrently, increased expression of pdhA facilitates the conversion of pyruvate into acetyl-CoA, collectively resulting in an amplified precursor pool that supports enhanced spinosad biosynthesis.

Notably, although mannose metabolism also activates competing branches such as the GDP-mannose pathway (manB) and leucine biosynthesis (leuA), the overall carbon flux toward acetyl-CoA remains dominant, as evidenced by increased acetyl-CoA accumulation and spinosad production. Moreover, identification of these competing pathways provided a rational basis for subsequent targeted metabolic engineering, and their coordinated regulation further improved spinosad titers. Taken together, these results demonstrate that mannose enhances spinosad biosynthesis in strain U7 primarily by driving a more efficient and directed carbon flux toward acetyl-CoA.

However, our analysis revealed that two competing metabolic pathways in mutant U7 were simultaneously enhanced through upregulation of leuA and manB, while downregulation of pdhC expression potentially constrained acetyl-CoA biosynthesis efficiency. This observation aligns with previous findings demonstrating that leuA deletion enhances 2-ketoisovalerate production in E. coli by redirecting metabolic flux [35]. We hypothesized that LeuA competes for key precursors derived from mannose metabolism, thereby attenuating spinosad biosynthesis. This hypothesis was experimentally validated through genetic manipulation: leuA knockout significantly increased spinosad production (by 157 %), whereas its overexpression resulted in a marked decrease (by 59 %), confirming the competitive nature of this metabolic branch point. As for pdhC and manB, their function are related to the synthesis of acetyl-CoA [36,37]. Their dysregulated expression in mutant U7 likely created a bottleneck in acetyl-CoA supply for spinosad biosynthesis. Consistent with this hypothesis, our genetic engineering experiments demonstrated that pdhC overexpression enhanced spinosad production by 211 %, while the pdhC deletion had the opposite effect. Previous literature has reported that the manB deletion can promote the spinosad biosynthesis in S. spinosa. This effect likely stems from redirected carbon flux from mannose-6-phosphate isomerization toward glycolysis, thereby increasing acetyl-CoA generation [27]. These systematic genetic perturbations provide conclusive evidence that pdhC, manB, and leuA serve as key regulatory nodes controlling spinosad biosynthesis through their coordinated effects on precursor availability and metabolic flux distribution.

Insufficient supply of acetyl-CoA has become a key bottleneck of spinosad biosynthesis. Therefore, how to fine tune the complex metabolic flux of the acetyl-CoA supply pathway is a core challenge [38,39]. Previous studies have focused on modifying specific individual gene to solve this challenge [2]. However, compared with the previous local modifications, we systematically fine-tuned multiple metabolic flow nodes in the acetyl-CoA supply pathway to achieve higher acetyl-CoA concentration for spinosad biosynthesis, included that simultaneous manB and leuA deletion and overexpression of pdhC in mutant U7. As expected, the concentration of acetyl-CoA in the engineered strain obtained by the combination modification of three genes was significantly higher than that of the engineered strain obtained by single gene. Not only that, it also has the highest spinosad production. Hence, enhancing acetyl-CoA concentration during synthesis is an efficient strategy to boost spinosad production [6,7]. To better elucidate the mechanism by which optimization of acetyl-CoA supply pathway promotes spinosad biosynthesis, we performed comparative proteomic analysis. A total of 97 proteins were upregulated, while 133 proteins were downregulated in the mutant U7-ΔmanBΔleuA::pdhC, confirmed that the combinatorial optimization of manB, leuA and pdhC had a comprehensive effect on the strain metabolic network, especially central carbon metabolism. The upregulated proteins not only promote the conversion of extracellular mannose to phosphorylated mannose but also enhance glycolysis, fatty acid degradation and TCA cycle, provided sufficient precursors and energy for other biological processes. These results may explain why the mutant U7-ΔmanBΔleuA::pdhC exhibited high spinosad production.

5. Conclusion

Mannose is the carbon source that is most easily absorbed and efficiently converted to produce spinosad by S. spinosa compared with glucose. The mannose metabolism is a key acetyl-CoA synthetic pathway that significantly affects the production of spinosad. This pathway provides precursors for spinosad biosynthesis, and improves carbon conversion. Our results suggested that it is feasible to optimize the mannose metabolism by combining UV mutagenesis, fermentation optimization and gene wiring optimization to improve spinosad production, which combined with proteomic analysis can help us to clarify the high production mechanism of spinosad. Moreover, this study found three key functional genes that have an important impact on the biosynthesis of spinosad, which indicates that when optimizing the metabolism network of spinosad biosynthesis, it is available to modify these functional genes. This is of great significance in optimizing the synthetic route of other natural products in Streptomyces.

CRediT authorship contribution statement

Zirong Zhu: Writing – original draft, Validation, Methodology, Investigation, Data curation. Qin Zhao: Validation, Methodology, Investigation, Data curation. Yan Zhu: Methodology, Investigation, Funding acquisition. Ling Shuai: Methodology, Data curation. Jun Li: Methodology, Investigation, Data curation. Qing Liu: Validation, Methodology, Investigation. Xirong Liu: Methodology, Investigation, Data curation. Shanrui Wang: Methodology, Investigation. Duo Jin: Investigation. Zirui Dai: Investigation. Liqiu Xia: Writing – review & editing, Funding acquisition, Data curation, Conceptualization. Hongrong Liu: Data curation, Conceptualization. Jie Rang: Writing – review & editing, Writing – original draft, Project administration, Investigation, Funding acquisition, Data curation, Conceptualization.

Availability of data and materials

All data generated or analyzed during this study are included in this published article (and its supporting Information files).

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interest: Xirong Liu is currently employed by Hunan Norchem Pharmaceutical Co., Ltd.

Abbreviations

UV, ultraviolet irradiation; manB, beta-mannosidase; pdhC, pyruvate dehydrogenase subunit; leuA, 2-isopropylmalate synthase; SAM, S-adenosyl methionine; TCA, tricarboxylic acid cycle; HPLC, high-performance liquid chromatography; LC-MS, liquid chromatography-mass spectrometry; m/z, mass-to-charge ratios; qRT-PCR, quantitative reverse transcription polymerase chain reaction; LT50, median lethal time; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Appendix A. Supplementary data

The following are the Supplementary data to this article: