The strains, plasmids, and primers used in this study are listed in Table 1. S. spinosa CCTCC M206084 was initially activated on corn starch medium (CSM) solid agar at 30 °C for 48 h. Escherichia coli strains were routinely cultured in Luria–Bertani medium, with antibiotic supplementation (when required) based on plasmid selection markers. Tryptic soy broth (TSB) and brain heart infusion (BHI) solid media were specifically used to comparatively analyze the sporulation capacities of WT and recombinant strains. Bacterial conjugation between S. spinosa and E. coli was performed using R6 medium, as previously described (Liu et al., 2021). For spinosad production analysis, seed cultures were inoculated into synthetic fermentation medium (SFM), with amino acid supplementation (1 mg/mL final concentration) where specified. All liquid cultures of the strains were incubated in 300 mL flasks containing 50 mL of medium, and all cultivation experiments were conducted in triplicate under aseptic conditions. The composition of the medium is listed in the Supplementary material (Supplementary: Composition of Culture Media). The relevant information pertaining to the plasmids has been deposited in Benchling.

(pET28A,¹

pOJ260,²

pSET-dCas9)³.

This study focused on PCC, encoded by the pcc gene cluster comprising pccA (A0A2N3XQX1), pccB1 (A0A2N3XWG9), and pccB2 (A0A0N3Y507). The strain engineering workflow comprised two parallel strategies: (1) overexpression strain construction and (2) CRISPRi strain development. A representative protocol for manipulating the pccB1 gene is described below. (1) Overexpression strain construction: The pccB1 gene (UniProt ID: A0A2N3XWG9) was polymerase chain reaction (PCR)–amplified from S. spinosa genomic DNA using primers pccB1-F/pccB1-R (Table 2). Simultaneously, the constitutive P_ermE promoter was amplified from the plasmid pOJ260-P_ermE using primers P_ermE-F/P_ermE-R. These fragments were fused via overlap extension PCR (Supplementary Figure 1) to generate the P_ermE-pccB1 transcriptional unit. The fusion product was subsequently digested with EcoR I/Hind III (Thermo Scientific) and ligated into a similarly restricted pOJ260 backbone (Supplementary Figure 2), yielding the recombinant plasmid pOJ260-P_ermE-pccB1. 2. CRISPRi strain construction: For targeted gene suppression, a 200-bp sgRNA-pccB1 cassette was amplified from the pSET-dCas9 template using primers pccB1-sgRNA-F/sgRNA-R (Table 2). This fragment was cloned into Spe I/EcoR I sites of pSET-dCas9 using DNA ligase, generating the CRISPRi plasmid pSET-dCas9-pccB1 (Supplementary Figure 3). Both recombinant plasmids were introduced into WT S. spinosa via E. coli-mediated conjugation. Engineered strains were obtained via antibiotic selection (apramycin 50 μg/mL for pOJ260 derivatives; thiostrepton 25 μg/mL for pSET derivatives) (Cao et al., 2024).

To evaluate the growth dynamics and morphological alterations, WT and engineered S. spinosa strains were precultured in CSM for 48 h at 30 °C, followed by submerged fermentation in SFM under identical conditions. Hyphal morphology was monitored daily using an electron microscope. For standardized protocols regarding biomass measurement and extracellular metabolite profiling, refer to supplementary: Analysis of Physiological and Biochemical Differences Among Strains (He et al., 2021).

Fermentation broths (10-days old, 30 °C) from triplicate biological replicates were extracted with methanol (1:3 v/v), after lyophilizing the methanol, ethyl acetate was employed for the back-extraction of spinosad (1:1 v/v). After solvent evaporation, the residues were reconstituted in methanol for HPLC and LC–MS/MS analysis. The spinosad pure product (50 mg/L), supplied by Merck, was used as the stock solution, and a gradient dilution was performed using a methanol solution. An Agilent 1290 HPLC system equipped with a C18 reverse-phase chromatographic column (such as ZORBAX Eclipse Plus C18, 4.6 × 150 mm, 5 μm) was used, and the mobile phase comprised methanol:acetonitrile:ammonium acetate in a volumetric ratio of 42:42:16. The column temperature was 30 °C, the flow rate was 1.0 mL/min, the injection volume was 10 μL, and the detection wavelength was 246 nm. Linear regression was performed with the spinosad peak area against the concentration (mg/L) to obtain the standard curve equation. MS conditions were as follows: electrospray ionization source (ESI±), ion source temperature 500 °C, spray voltage ± 4500 V, curtain gas 35 psi, and collision gas (CAD) 8 psi. The target compounds were detected using the multiple reaction monitoring mode.

Total RNA was isolated from Day 2/4/6 SFM cultures using TRIzol^® (Invitrogen), followed by DNase I treatment (Takara). PrimeScript RT Master Mix (TaKaRa) was used for cDNA synthesis. qRT-PCR was performed on a QuantStudio 5 System (Applied Biosystems) with SYBR^® Green chemistry using 16S rRNA (F: 5′-AGAGTTTGATCCTGGCTCAG-3′; R: 5′-GGTTACCTTGTTACGACTT-3′) as the endogenous control (Liu et al., 2020). Detailed procedures are described in the supplementary (Supplementary: RNA Isolation and Quantitative Real-Time PCR Analysis).

The WT and recombinant strains were fermented and cultured, and 5 mL of the SFM fermentation broth on the 2nd, 4th, and 6th days was used for testing. The sample supernatant was used to determine the concentrations of propionyl-CoA and methylmalonyl-CoA according to the instructions provided in the enzyme-linked immunosorbent assay kit. Sample processing and detection protocol: 5 mL of SFM fermentation broth collected at different time points (2 d, 4 d, and 6 d) was immediately placed on ice, washed three times with ultrapure water. The bacterial cells were ground in liquid nitrogen and then dissolved in phosphate buffered saline (PBS). 50 mM N-ethylmaleimide was added to inhibit degradation, and centrifuged at 12,000 × g for 15 min at 4 °C to obtain the supernatant. It was precipitated using prechilled 10% perchloric acid, incubated on ice for 10 min, and then centrifuged under the same conditions. The acidic extract was neutralized with 2 M KOH on an ice bath and centrifuged again to remove the precipitates. The resulting samples were aliquoted and stored at −80 °C for later use.

A standard curve was plotted using a series of concentrations (0.1–100 nM). Samples and standards were added to antibody-precoated plates and incubated at 37 °C for 1 h. After washing, the biotinylated secondary antibody horseradish peroxidase–streptavidin and the 3,3′,5,5′-tetramethylbenzidine substrate were sequentially added. The reaction was developed in the dark for 15 min and then terminated with 2 M H2SO4. The absorbance at 450 nm was measured, and the concentrations of propionyl-CoA and methylmalonyl-CoA were calculated using a four-parameter fitting curve.

Strains were sonicated on ice in lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% NP-40, 1 mM PMSF) and then centrifuged at 12,000 × g for 20 min at 4 °C to collect the supernatant. After determining the protein concentration using the bicinchoninic acid assay, 30 μg of total protein was mixed with 4 × Laemmli buffer and denatured at 95 °C for 5 min (Hassan et al., 2025). The samples were separated using 12% SDS-PAGE (constant voltage: 80 V for the stacking gel and 120 V for the resolving gel). Then, they were wet-transferred to a polyvinylidene fluoride membrane at 100 V and 4 °C for 90 min. The membrane was blocked with 5% skim milk in TBST (Tris- buffered saline + 0.1% Tween 20) at room temperature for 1 h, followed by sequential incubations: primary antibody: mouse anti-PccB1 polyclonal antibody (1:2000 in blocking buffer, overnight at 4 °C); secondary antibody: DyLight 488-conjugated goat antimouse IgG (1:5000 in TBST, 1 h at room temperature). After washing with TBST (three 10-min washes), signals were detected using a fluorescence imaging system. The fluorescence intensity of the target band (53 kDa) was quantified using the ImageQuant TL software (Tang et al., 2021).

To verify the expression of pccB1 in recombinant and wild-type strains, the heterologous expression vector pET28a-pccB1 was constructed. Primers pccB1-F and pccB1-R were designed to amplify the pccB1 gene from the genome of S. spinosa. After double digestion with EcoR I and Hind III, the pccB1 gene was cloned into the plasmid pET28a to generate the recombinant vector pET28a-pccB1, which was then transformed into E. coli BL21. The recombinant strain was cultured in LB medium containing 50 μg/mL kanamycin to induce the expression of the PCCB1 protein. The expressed protein was injected into mice as an antigen, and Western blot analysis was subsequently performed to detect the expression of PCCB1 in both the recombinant and wild-type strains.

The 144-h SFM fermentation broth was used for proteomic testing. MS proteomics data were deposited in the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD062497. The detailed methods of proteomic treatment are listed in the Supplementary section (Supplementary: Proteomic Sample Processing, Proteomics data processing and statistical analysis).

Spinosyns were extracted from the fermentation broth and incorporated into the artificial diet at a final concentration of 1 mg/L. The diet formulation contained (per liter) 60 g of flour, 120 g of soybean powder, 20 g of yeast, 13 g of agar, 2 g of ethyl p-hydroxybenzoate, and 1 g of sorbic acid, along with vitamin supplements (20 × vitamin C and 10 × vitamin B2 tablets). Control groups received equivalent volumes of sterile water instead of spinosyn extract. Both diets were dispensed into 24-well plates containing fourth-instar Helicoverpa armigera larvae (n = 24 per group). Mortality rates were recorded daily over a 4-days exposure period (Xia et al., 2024).

Error bars represent the standard error of the mean (based on n = 3 independent biological replicates). Student’s t-test was used to compare differences between groups (assuming normal distribution of data and homogeneity of variance). Statistical significance thresholds were set as p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***).

Propionyl-CoA carboxylase catalyzes the ATP-dependent conversion of propionyl-CoA to methylmalonyl-CoA, a critical extender unit that synergizes with malonyl-CoA to assemble the polyketide backbone of spinosad. Because of the inherently low intracellular concentrations of these precursors in S. spinosa, a dual genetic strategy was employed to enhance PCC activity: (1) CRISPRi for targeted gene repression and (2) pOJ260 plasmid-mediated overexpression. Six engineered strains with modified pccA, pccB1, or pccB2 expression were constructed and cultured for 14 days under standardized fermentation conditions.

Quantitative HPLC analysis revealed significant variations in spinosad production across strains (Figures 1A, B, Supplementary Figure 4). The CRISPRi-mediated repression of pccB1 (strain S. spinosa-pSET-dCas9-pccB1) paradoxically led to a 2.6-fold increase in spinosad yield compared with the WT. In contrast, the overexpression of pccA and pccB2 resulted in a 1.8- and 1.5-fold increase in production, respectively. The structure of spinosad was determined using high-resolution LC–MS (Supplementary Figure 5), with characteristic molecular ions that matched the reference spectra.

The observed production patterns were aligned with PCC’s proposed role in precursor supply: overexpression strains exhibited elevated methylmalonyl-CoA levels, consistent with enhanced PKS initiation and elongation efficiency. However, the unexpected hyperproduction in the pccB1-repressed strain suggests complex regulatory dynamics; follow-up analyses identified a compensatory regulation mechanism.

This counterintuitive finding highlights that pccB1 is a potential central point of metabolic flux at which partial repression triggers cellular adaptation mechanisms that ultimately optimize spinosad biosynthesis. Further investigation of post-translational regulation and pathway crosstalk is critical for understanding this phenomenon.

Quantitative RT-qPCR analysis confirmed the effective transcriptional repression of pccB1 in CRISPRi strains (Figures 1C–E, Supplementary Figure 6), showing a reduction compared with the WT. To validate protein-level effects, a heterologous pccB1 expression system was constructed (Supplementary Figure 7) and immunoblotting was performed with anti-PccB1 polyclonal antibodies generated from mouse cardiac serum. SDS-PAGE of recombinant protein extracts revealed distinct bands at the predicted 53 kDa molecular weight (Figure 1F). Subsequent western blot using DyLight 488–conjugated goat antimouse IgG secondary antibody demonstrated significantly higher fluorescence intensity in overexpression strains compared with the WT. In contrast, CRISPRi strains exhibited signal reduction (Figure 1G), confirming successful translational regulation.

Dynamic changes in spinosad precursors were observed across engineered strains (Figures 2A–F). The pccB1-repressed strain accumulated higher concentrations of propionyl-CoA (Day 6) and methylmalonyl-CoA (Day 4) compared with the WT, despite reduced pccB1 expression. Concurrently, glucose consumption rates increased in repression strains (Figures 2G–I), indicating enhanced carbon flux via these pathways. These paradoxical findings prompted the investigation of PCC subunit cross-regulation. Transcript analysis demonstrated reciprocal compensation: pccA and pccB2 expression increased in pccB1-repressed strains, while pccB1 repression in pccB2-repressed strains reduced pccB1 transcripts (Figures 3A–C). This regulatory crosstalk likely maintains functional PCC holoenzyme assembly, as evidenced by the sustained production of methylmalonyl-CoA.

The transcriptional activation of PKS genes (spnA–J) in pccB1-repressed strains exceeded WT levels (Figures 3D–F). This coordinated upregulation of both precursor supply (methylmalonyl-CoA) and biosynthetic machinery (spn cluster) creates a feed-forward loop that drives spinosad hyperproduction. The metabolic rewiring appears to involve two compensatory mechanisms: subunit redundancy in PCC complex assembly, which allows pccB2 to compensate for pccB1 deficiency, and carbon flux redistribution via activation of the tricarboxylic acid (TCA) cycle.

These findings redefine the regulatory role of PCC in secondary metabolism, demonstrating that targeted repression of specific subunits can trigger global metabolic adaptations that paradoxically enhance antibiotic biosynthesis.

Comparative proteomic analysis of S. spinosa-pSET-dCas9-pccB1 vs. the WT via the Kyoto Encyclopedia of Genes and Genomes pathway enrichment and eggNOG functional annotation revealed extensive metabolic remodeling (Supplementary Figure 8). The engineered strain exhibited a significant upregulation (>1.5-fold change, p < 0.05) of proteins linked to spinosad biosynthesis and energy metabolism, indicating a redirection of metabolic resources toward secondary metabolite production.

Focusing on spinosad precursor pathways (Figure 4), an 8.7-fold increase was observed in PCC α-subunit (PccA) expression, driving the accelerated conversion of propionyl-CoA to methylmalonyl-CoA. This catalytic surge facilitated dual metabolic routing: channeling of methylmalonyl-CoA into the TCA cycle via succinyl-CoA (1.8-fold flux increase) and its enhanced incorporation into polyketide backbones via the coordinated activation of macrolide synthases (12-, 14-, and 16-acetyl transferases upregulated 3.1–4.5 fold) (Supplementary Figure 9).

Global carbon metabolism analysis (Figure 5) demonstrated a multilevel phenomenon. In acetyl-CoA amplification, pyruvate kinase (2.97-fold upregulation) and L-serine dehydratase (5.2-fold upregulation) enhanced pyruvate synthesis. In oxidative decarboxylation, the upregulation of the pyruvate dehydrogenase complex (7.4-fold) increased the acetyl-CoA yield. In TCA cycle activation, key enzymes, including aconitate hydratase (2.3-fold upregulation), isocitrate dehydrogenase (2.6-fold upregulation), and 2-oxoglutarate ferredoxin oxidoreductase (3.9-fold upregulation), collectively increased ATP generation (1.7-fold) and the production of metabolic intermediates. The fatty acid degradation pathway showed remarkable activation: long-chain acyl-CoA synthetase (5.2-fold upregulation), acyl-CoA oxidase (5.3-fold upregulation), and enoyl-CoA hydratase (3.9-fold upregulation). This enzymatic triad enhanced β-oxidation efficiency, converting stored lipids into acetyl-CoA at higher rates than in the WT.

The coordinated upregulation of complementary acetyl-CoA production routes, spanning glycolysis, amino acid catabolism, and lipid degradation, establishes the metabolic flexibility of the strain. This system-level adaptation not only sustains energy demands via enhanced spinosad biosynthesis but also maintains redox balance by reinforcing the TCA cycle. The pccB1 repression appears to trigger a global metabolic state that favors secondary metabolism, as revealed by proteomic data, which show that many of the upregulated proteins are directly associated with precursor supply or energy transduction.

Analysis of growth kinetics demonstrated distinct physiological adaptations in S. spinosa-pSET-dCas9-pccB1 compared with the WT. The engineered strain exhibited a shorter stationary phase and an earlier entry into the decline phase (Figures 6A–C), which correlated with accelerated glucose consumption rates. This metabolic prioritization was accompanied by a reduction in the final biomass, indicating a reallocation of resources toward secondary metabolism at the expense of vegetative growth.

Scanning electron microscopic examination of 48-h mycelia revealed that the hyphal morphology and branching patterns were conserved between strains (Figure 6D). Nonetheless, sporulation assays on CSM, TSB, and BHI media showed significantly reduced spore formation in the CRISPRi strain (Figure 6E). In addition, proteomic profiling identified complete suppression of the spore germination protein YaaH, coupled with the downregulation of the developmental regulators WhiG and SsgA. This phenotypic shift suggests the adoption of a metabolically oriented survival strategy, favoring the rapid utilization of precursors over developmental competency.

The observed growth–sporulation trade-off aligns with the overflow metabolism hypothesis, which posits that enhanced carbon flux through acetyl-CoA nodes redirects cellular resources from differentiation programs to secondary metabolite synthesis. The concurrent loss of sporulation efficiency and activation of spinosad biosynthetic genes (spn cluster) signifies hierarchical metabolic prioritization under genetic perturbation.

Building upon proteomic findings that revealed a significant upregulation of amino acid metabolic enzymes (log2FC > 1.5, p < 0.01) in the pccB1-repressed strain, a metabolic engineering strategy was devised to enhance precursor availability for spinosad biosynthesis. Pathway analysis identified three key metabolic nodes capable of generating acetyl-CoA and propionyl-CoA precursors: cysteine/methionine metabolism, branched-chain amino acid (valine/leucine/isoleucine) degradation, and propanoate metabolism (Figure 7). Targeted supplementation of 1 mg/mL valine, leucine, isoleucine, methionine, alanine, or propionate during mid-log phase fermentation resulted in differential enhancement of spinosad production (Figure 8A). The pccB1-repressed strain exhibited superior precursor utilization efficiency, with alanine supplementation yielding maximal titers of 590 ± 31 mg/L, a 6.2-fold increase over WT baseline production (95 ± 13 mg/L; Figure 8B, Supplementary Figure 10).

The exceptional performance of alanine stems from its dual metabolic fates: direct conversion to pyruvate via alanine dehydrogenase, subsequent oxidative decarboxylation to acetyl-CoA via the pyruvate dehydrogenase complex, and anaplerotic entry into propionyl-CoA synthesis via methylmalonyl-CoA mutase. This metabolic flexibility synergizes with the enhanced amino acid catabolic capacity of the strain, as evidenced by the upregulation of branched-chain amino acid transaminases and methionine γ-lyases. Upon supplementation of alanine to the pccB1-repressed strain, spinosad yield was significantly enhanced. This improvement is likely attributed to the redirection of exogenous alanine toward optimizing metabolic flux operation, thereby refining precursor distribution within the strain.

To confirm the functional enhancement of spinosad production, standardized bioassays were performed using third-instar H. armigera larvae (Figures 9A–C). The pccB1-repressed strain exhibited significantly increased insecticidal efficacy (Figure 9D). Symptomatology analysis revealed rapid-onset neurotoxic effects consistent with spinosad’s mode of action, including neuromuscular hyperexcitation, progressive paralysis, and synaptic dysfunction. These observations agree with spinosad’s known disruption of insect neuronal signaling, confirming that the enhanced titers in the pccB1-repressed strain are directly correlated with functional bioactivity. The dose-dependent mortality curves and symptom progression validated the quantitative and qualitative improvements in spinosad production.