A base-editing and auxotrophic selection platform for linking epigenetic regulators to silent biosynthetic gene clusters in Aspergillus nidulans Authors: Meng Pang, Donghui Guo, Yanling Li, Yuze Chen, Rumeng Du, Jiayang Liu, Jiahui Hu, Hang Li, Zhuofan Li, Guojun Pan, Yuan Tian, Fanglong Zhao Published: 2026 ABSTRACT AbstractFilamentous fungi harbor numerous silent biosynthetic gene clusters (BGCs), representing an underexplored reservoir of natural products. Chemical inhibition of epigenetic regulators is an effective strategy for activating silent BGCs, but it often lacks specificity and cannot readily link activated clusters to their cognate regulators. Here, we developed a coupled cytosine base editing and auxotrophic selection platform in Aspergillus nidulans to systematically identify epigenetic regulators of silent gene clusters. An sgRNA library targeting 24 core epigenetic regulators was introduced, together with a cytosine base editor, into 21 pyrG-based PKS reporter strains, enabling growth-based selection and direct sgRNA recovery by Sanger sequencing. This screen identified five regulator-cluster activation pairs, including known interactions and two previously uncharacterized mechanisms. Specifically, KdmB represses the AN7084 gene cluster by erasing promoter H3K4me3, whereas Kmt5 silences the cichorine core pks gene through H4K20me3 deposition. This scalable platform enables rapid matching of silent BGCs with their epigenetic regulators and advances chromatin-guided fungal genome mining.Keywords: Aspergillus nidulans, Base editing screen, Epigenetic regulation, Activation of silent gene clusters, Polyketide synthase Introduction Filamentous fungi are a rich source of structurally complex natural products [1]. Many of these metabolites, including the antibiotic penicillin, the cholesterol-lowering drug lovastatin, and the antifungal echinocandins, have become cornerstones of modern therapeutic chemistry [2]. With the advent of whole-genome sequencing, genomic analyses have revealed that filamentous fungal genomes typically harbor 30-80 biosynthetic gene clusters (BGCs) [1]. However, under standard laboratory culture conditions, most of these BGCs remain transcriptionally silent or are expressed only at low levels [3,4]. This gap between genomic potential and experimentally accessible chemical diversity continues to limit the discovery of new bioactive molecules. Epigenetic regulation is a key mechanism controlling fungal secondary metabolism [5]. The transcriptional activity of secondary metabolite gene clusters is tightly governed by chromatin state [6]. Interfering with epigenetic processes such as histone acetylation, methylation, phosphorylation, and ubiquitination can relieve transcriptional repression, increase chromatin accessibility, and activate otherwise silent BGCs [[4], [5], [6]]. Since the first reports in 2009 demonstrating that fungal secondary metabolism is influenced by epigenetic regulation [7], chemical perturbation of histone modifications using histone deacetylase inhibitors, histone methyltransferase inhibitors, and related compounds has become a common strategy for silent cluster activation. This approach has been applied across diverse fungal species and has enabled the discovery of more than 540 new natural products [8,9]. However, chemical intervention has inherent limitations, particularly target promiscuity and pleiotropic effects. Small-molecule inhibitors often affect multiple families of epigenetic enzymes simultaneously, making it difficult to identify the specific regulatory factor responsible for activation of a given BGC. In contrast, conventional targeted gene deletion can establish causal links between regulators and gene clusters, but this approach relies on homologous recombination and requires the construction and analysis of individual mutants one at a time [[10], [11], [12]]. Given that filamentous fungi encode dozens of epigenetic modifiers [5,6], current genome-mining strategies lack an efficient method to rapidly and precisely match activated BGCs with their cognate epigenetic regulators. CRISPR-Cas9 cytosine base editors (CBEs) provide a powerful genetic tool to address this challenge [13]. CBEs enable precise C•G-to-T•A conversions at target loci without inducing DNA double-strand breaks. Through rational sgRNA design, this single-nucleotide editing capability can convert selected codons within target genes into premature stop codons, thereby enabling efficient gene inactivation without introducing extensive genome damage [14]. In synthetic biology, CRISPR-Cas systems combined with sgRNA library screening have been widely used to interrogate cellular responses to genetic perturbations [15,16]. A key advantage of this approach is that sequencing the sgRNA barcode from cells displaying a selected phenotype allows the causal genetic perturbation to be traced directly. This paradigm provides an attractive framework for systematically using loss-of-function perturbations of epigenetic regulators to dissect the activation mechanisms of silent fungal BGCs. However, owing to the distinct genetic and experimental challenges of filamentous fungi, high-throughput screening approaches remain rarely implemented in these organisms [17]. In recent years, base-editing technologies have been successfully introduced into filamentous fungi, with targeted editing efficiencies of 60–90% achieved in Aspergillus species [18]. Nevertheless, the field still lacks a standardized and broadly applicable platform that directly links base editor-mediated high-throughput inactivation of regulatory factors to the activation phenotypes of specific silent BGCs. Here, we develop an efficient genetic screening strategy to rapidly elucidate the epigenetic regulatory logic of fungal secondary metabolism and validate it in A. nidulans. We first constructed an sgRNA library specifically targeting 24 core epigenetic regulators in A. nidulans and introduced this library, together with a CBE expression plasmid, into 21 engineered reporter strains. In each reporter strain, the auxotrophic marker pyrG was placed downstream of the promoter of a silent polyketide synthase (pks) gene, thereby coupling gene activation to growth. Under auxotrophic selection, the sgRNA cassette is directly linked to the activation phenotype of the corresponding cell, allowing the regulator-target BGC relationship to be established by Sanger sequencing alone. Using this system, we identified five epigenetic regulator-BGC relationships and uncovered two previously unknown regulatory mechanisms. Specifically, Kmt5-mediated H4K20me3 deposition silences the cichorine gene cluster, whereas KdmB maintains repression of the AN7084 cluster by erasing the activating H3K4me3 mark. This work establishes a scalable screening framework for targeted activation of silent core pks gene and rapid mechanistic interrogation of their upstream regulatory networks. Results Design of a coupled base editing and auxotrophic selection platform In this study, we employed a cytosine base editor library to inactivate epigenetic regulators, thereby perturbing the expression of silent gene clusters. We then directly coupled cluster activation to a growth phenotype, enabling efficient isolation of positive clones and precise mapping between regulatory factors and their target silent gene clusters (Fig. 1). To validate the feasibility of this strategy, we selected the previously validated regulatory pair consisting of the histone methyltransferase CclA and mdpG [7], the core gene of the silent emodin biosynthetic gene cluster (mdp BGC) in A. nidulans, as a proof of concept (Fig. 2A). Specifically, we placed the pyrG reporter gene under the control of the mdpG (pks) promoter and introduced targeted loss-of-function mutations in cclA using CBE in this engineered strain. As expected, disruption of cclA relieved transcriptional repression at the mdpG promoter, thereby activating pyrG expression and restoring growth on medium lacking uridine and uracil (Fig. 2B). These results provided initial validation of the logic and practical feasibility of the screening strategy. Schematic of the coupled base editing and auxotrophic selection platform for identifying epigenetic regulators of silent BGCs in filamentous fungi. A cytosine base editor (CBE) introduces premature stop codons in 24 core epigenetic regulators guided by specific sgRNAs. The sgRNAs are expressed as a tRNA-sgRNA array. The promoters of 21 silent polyketide synthase (pks) genes drive the expression of the pyrG auxotrophic marker. After introducing the regulator inactivation library into the reporter strains, positive clones are selected on medium lacking uracil and uridine, linking cluster activation to a growth phenotype. The causative epigenetic regulators are identified by Sanger sequencing of recovered sgRNAs, and transcriptional activation and associated histone modification changes are verified by RT-qPCR and ChIP-qPCR, respectively. Construction of pyrG-based reporter strains and validation of the screening system in A. nidulans. (A) Schematic of reporter strain construction. The pyrG gene and bar resistance marker were precisely integrated into the open reading frame of target pks genes in A. nidulans via homologous recombination, placing pyrG expression under the control of the native pks promoter. (B) Proof-of-concept validation of the system. Inactivation of the known epigenetic regulator cclA in the AN0150 pks promoter-driven pyrG reporter strain (AN0150 pks(P)-pyrG cclA-) restored growth on medium lacking uracil and uridine. (C) Evaluation of basal growth of the 21 pks(P)-pyrG reporter strains on selective media. The x-axis indicates colony diameter. Most strains exhibited growth inhibition on uracil/uridine-deficient medium, which could be further enhanced by the pyrimidine synthesis inhibitor pyrazofurin. Strain with high basal expression (AN6791) was excluded from subsequent screening. Among the 21 constructed reporter strains, growth on selective medium varied considerably (Fig. 2C), reflecting the differential basal expression levels of each pks gene under standard conditions. Of these, 14 strains failed to grow completely on selective medium (Fig. 2C), 6 strains exhibited extremely slow growth, and 1 strain (AN6791(P)-pyrG) grew robustly. To suppress the growth of the fast-growing strains, we supplemented the medium with the PyrG inhibitor pyrazofurin [19]. This treatment inhibited the slow-growing strains, whereas the strain AN6791(P)-pyrG maintained robust growth (Fig. 2C). This strain was therefore excluded from subsequent activation screening. Together, these experiments established a clean and reliable background for downstream forward genetic screening. Coupled epigenetic perturbation and auxotrophic screening identify regulator-BGC pairs Next, we constructed an epigenetic regulator inactivation library (Fig. 3A). This library targeted 24 core epigenetic factors (Table S2), encompassing histone methyltransferases (e.g., kmt5, setB, and ash1), demethylases (kdmA and kdmB), acetyltransferases (gcn5, esaA, and rtt109), deacetylases (hdaA, hosA, and hstA), and a reader domain protein (sntB) [5,6]. For each target gene, two specific sgRNAs targeting the coding region were designed to introduce premature stop codons via C-to-T single-base editing (Table S2), thereby enabling efficient loss-of-function. All sgRNA expression cassettes were cloned into an AMA1-based vector harboring a pyroA selection marker, while the CBE was expressed from a separate AMA1-based plasmid carrying a riboB marker [18]. High-throughput screening using the epigenetic regulator inactivation library and identification of positive clones. (A) Schematic of the screening workflow. The nCas9-CBE expression plasmid harboring the AMA1 replicon and the sgRNA plasmid library targeting 24 core epigenetic regulators were co-transformed into A. nidulans protoplasts, followed by phenotypic selection on medium lacking uracil and uridine. (B) Validation of the five regulator-cluster activation pairs identified from the screen. The left panels show positive colonies growing on selective medium for each reporter strain compared to non-activated controls. The right panels show the direct identification of recovered sgRNAs from positive clones via Sanger sequencing, with dashed boxes indicating the target sequences and the corresponding identified regulator genes (cclA, hosA, kdmB, and kmt5). The CBE expression plasmid and the sgRNA library were co-transformed into each of the 20 validated reporter strains (Fig. 3A). Transformants were initially selected on medium lacking pyridoxine and riboflavin to ensure the maintenance of both plasmids. Subsequently, they were replica-plated onto either uracil/uridine-deficient medium or medium supplemented with the PyrG inhibitor pyrazofurin. Colonies exhibiting robust growth were isolated, and their resident sgRNA cassettes were PCR-amplified and sequenced to identify the inactivated regulators (Fig. 3B). Screening of the 20 reporter strains yielded five regulator-pks activation pairs, underscoring the efficiency of our high-throughput perturbation strategy in capturing regulatory interactions in a single screen (Fig. 3B). Among these, three combinations had been previously reported, namely cclA-AN0150 [7], cclA-AN7909 [7], and hosA-AN7909 [20]. Sequencing of the corresponding target genes revealed the expected C-to-T substitutions at the edited positions, generating premature stop codons and thereby inactivating the target genes (Fig. 3B). More importantly, we identified two previously unreported pairs. Specifically, in the activated clones of the AN7084(P)-pyrG reporter strain, we recovered an sgRNA targeting kdmB [21], which encodes a JmjC-domain-containing H3K4me3 demethylase (Fig. 3B). Furthermore, in the activated clones of the AN6448(P)-pyrG reporter strain, we identified an sgRNA targeting kmt5 [22], which encodes the sole H4K20 methyltransferase in A. nidulans (Fig. 3B). To assess the specificity of these regulatory effects and exclude false-positive signals caused by global transcriptional perturbation, we randomly selected five additional non-target pks genes and quantified their expression in the kdmB- and kmt5-inactivated strains (Fig. S2). Most of these non-target genes showed only modest changes in expression or no statistically significant differences, suggesting that the KdmB-AN7084 and Kmt5-AN6448 regulatory relationships do not result from a general activation of pks genes, but instead exhibit a degree of target-cluster preference. Because neither of these regulator-pks associations had been previously documented, we selected them for further investigation to characterize their activation capacity and the specific histone modification changes underlying cluster derepression. KdmB silences the AN7084 gene via H3K4me3 demethylation To rule out potential off-target effects of the CBE, we generated a clean kdmB deletion mutant (kdmB KO) via homologous recombination in the AN7084-pyrG reporter strain. Compared to the parental control, the kdmB KO strain exhibited markedly enhanced growth on selective media (Fig. 4B and C), confirming that specific regulator inactivation, rather than off-target editing, drives reporter activation. KdmB represses the AN7084 gene cluster through H3K4me3 demethylation. (A) Schematic representation of the genomic organization of the unknown AN7084 BGC. (B) Phenotypic validation. The kdmB knockout (kdmB KO) reporter strain exhibited markedly enhanced growth on uracil/uridine-deficient plates compared to the control AN7084(P)-pyrG strain, indicating activation of the AN7084 promoter. Colonies expanded over time (Days 3-5) with reddish-brown pigmentation. (C) Growth curve analysis. The cell dry weight of the kdmB KO strain was significantly higher than that of the control in liquid selective medium, further confirming the pyrG-mediated growth restoration. (D) Transcriptional analysis by RT-qPCR. Deletion of kdmB resulted in significant upregulation (approximately 12-fold) of the core gene AN7084 and its adjacent cluster genes (AN7083 and AN7085). Data are presented as mean ± SD from three independent experiments. (E) Chromatin immunoprecipitation-qPCR (ChIP-qPCR) analysis. (F) IgG background control. ChIP-qPCR using a non-specific IgG antibody showed minimal background enrichment at the AN7083-AN7085 loci in both WT and kdmB KO strains. Although AN7084 is predicted to encode a PKS, its cognate metabolic product remains uncharacterized. RT-qPCR analysis revealed that kdmB deletion upregulated AN7084 expression by ∼12-fold, accompanied by the concomitant upregulation of adjacent cluster genes (AN7083 and AN7085) (Fig. 4D). Furthermore, ChIP-qPCR using an anti-H3K4me3 antibody demonstrated that the kdmB KO strain accumulated significantly higher H3K4me3 levels at the AN7084 and AN7083 promoters relative to the wild type (Fig. 4E). Collectively, these data establish KdmB as a negative regulator of the AN7084 gene, maintaining its transcriptional silencing by actively removing H3K4me3, a canonical mark of active transcription-from the cluster promoters. To determine whether the derepression of the AN7084 gene induced by KdmB inactivation translates into the production of novel metabolites, we deleted kdmB in the WT strain and further generated a kdmB/AN7084 double mutant (Fig. S3). However, comparative metabolic profiling of their fermentation extracts revealed no discernible differences (Fig. S4). This result suggests that while kdmB deletion successfully initiates cluster transcription, it is insufficient to trigger the complete biosynthesis of the target product. Kmt5 silences the AN6448 gene via H4K20me3 deposition The cichorine cluster directs the biosynthesis of the phytotoxic isocoumarin cichorine (Fig. 5A), a pathway previously elucidated through systematic gene deletions [23]. Deletion of kmt5 in the AN6448(P)-pyrG reporter strain increased its growth rate on both solid plates and in liquid culture, suggesting that the gene is activated (Fig. 5B and C). RT-qPCR analysis revealed that kmt5 deletion upregulated AN6448 transcription by ∼3.5-fold, whereas the expression of the neighboring genes (AN6447 and AN6449) remained largely unchanged (Fig. 5D). This pattern suggests that Kmt5 exerts its regulatory effects in a locus-specific manner, primarily targeting the core pks promoter rather than broadly repressing the entire cluster. ChIP-qPCR further confirmed that wild-type cells harbor higher levels of H4K20me3 at the AN6448 promoter, a modification that was drastically reduced in the kmt5 KO mutant (Fig. 5E). Together, these results identify Kmt5 as a locus-specific silencer that maintains AN6448 in a transcriptionally repressed state via H4K20me3 deposition. Kmt5 silences the cichorine gene cluster via H4K20me3 deposition. (A) Schematic representation of the genomic organization of the cichorine biosynthetic gene cluster (AN6447-AN6449). (B) Phenotypic validation. The kmt5 knockout (kmt5 KO) reporter strain restored growth on uracil/uridine-deficient plates compared to the control AN6448(P)-pyrG strain, indicating activation of the AN6448 promoter. Colonies of the knockout strain expanded over time (Days 3-5) with dark pigmentation. (C) Growth curve analysis. The cell dry weight of the kmt5 KO strain was higher than that of the control in liquid selective medium, further supporting pyrG-mediated growth advantage. (D) Transcriptional analysis by RT-qPCR. (E) Chromatin immunoprecipitation-qPCR (ChIP-qPCR) analysis. High levels of H4K20me3 enrichment were detected at the AN6448 promoter in the wild-type (WT); however, this repressive mark was reduced in the kmt5 deletion mutant. Data are presented as mean ± SD from three independent experiments. (F) IgG background control. ChIP-qPCR using a non-specific IgG antibody showed minimal background enrichment at the AN6447-AN6449 loci in both WT and kmt5 KO strains. To determine whether the targeted derepression of AN6448 translates into metabolite production, we generated a kmt5 single mutant and a kmt5 and AN6448 double mutant in the WT strain (Fig. S5). However, comparative metabolic profiling of their fermentation extracts revealed no discernible differences (Fig. S6). This indicates that kmt5 deletion alone is insufficient to trigger the complete biosynthesis of the target product. Consistent with our transcriptional data, the lack of metabolic output in this case is likely due to the highly localized nature of the derepression-while the core pks gene (AN6448) was successfully activated, the accessory genes required for the full biosynthetic pathway remained silent. Discussion Filamentous fungi harbor numerous cryptic biosynthetic gene clusters (BGCs) that represent a rich reservoir of natural products, yet the majority remain silent under laboratory conditions [1,3,4]. Epigenetic regulation, particularly through histone modifications, has emerged as a key mechanism governing BGC expression [5], and small-molecule inhibitors of histone-modifying enzymes have been widely used to activate these clusters [8]. However, such pharmacological interventions act globally and lack a rapid means of assigning specific epigenetic regulators to individual BGCs, constraining mechanistic understanding. To overcome this limitation, we developed a platform integrating CRISPR-Cas9 cytosine base editing (CBE) of epigenetic regulators with an auxotrophic reporter-based selection system in Aspergillus nidulans. This strategy converts the transcriptional activation of core pks gene into a selectable growth phenotype, enabling direct, efficient pairing of perturbed regulators with activated clusters. The implementation of this method relies primarily on base-editing library perturbation and auxotrophic selection to achieve rapid activation screening. First, we constructed an sgRNA library targeting 24 core epigenetic regulators, using CBE to introduce premature stop codons for efficient gene inactivation. The modular tRNA-sgRNA expression design and the AMA1-based episomal vector ensured stable sgRNA expression and facilitated downstream sequencing-based identification of the causative regulator. Second, by integrating the pyrG auxotrophic marker downstream of 21 silent pks core gene promoters, we transformed cluster activation into a simple growth readout on uracil/uridine-deficient medium. With the rapid development of genome-editing technologies in filamentous fungi, including CRISPR-Cas9-mediated knock-in Ref. [24], prime-editing-mediated insertion of large DNA fragments [25], and Cre recombinase-mediated genomic integration [26], the rapid construction of engineered strains is becoming increasingly feasible. These advances could facilitate the systematic generation of reporter strains in which activation signals are coupled to growth, allowing comprehensive evaluation of how transcriptional or epigenetic regulators influence the expression of the core genes of individual BGCs. Applying this platform, we identified five regulator-pks activation pairs, demonstrating its efficiency and sensitivity. Among these, two novel associations were characterized in mechanistic detail. First, KdmB, a JmjC-domain H3K4me3 demethylase, was found to repress the AN7084 gene cluster (AN7083-AN7085). Deletion of kdmB caused ∼12-fold upregulation of the core pks gene AN7084, coordinated induction of neighboring genes, and enhanced enrichment of H3K4me3 at cluster promoters. Previous studies showed that KdmB globally regulates primary and secondary metabolism, with both repressive roles linked to H3K4me3 demethylation and activating roles that can be independent of its catalytic JmjC domain [21]. Our results identify the AN7084 gene as a clear example of demethylase-dependent repression. Second, Kmt5, the sole H4K20 methyltransferase in A. nidulans, was identified as a repressor of the AN6448 gene. Kmt5 inactivation increased AN6448 expression by ∼3.5-fold and nearly abolished H4K20me3 at its promoter. Unlike KdmB, whose loss activated multiple genes within the AN7084 gene cluster, Kmt5-mediated repression was largely confined to the core pks gene, suggesting a more localized mode of regulation. Our results also expand the known role of H4K20 methylation in fungal secondary metabolism. In Fusarium species, Kmt5 has been shown to mediate H4K20 methylation and influence secondary metabolite production [22], but direct chromatin-level evidence at individual clusters has been lacking. Here, we show that Kmt5-dependent H4K20me3 is enriched at the AN6448 promoter and contributes to gene silencing. Despite these findings, several limitations should be acknowledged. Although we observed clear transcriptional activation of pks gene, the entire gene cluster was not fully activated, and the corresponding metabolites were not detected. Previous studies of epigenetic regulation in secondary metabolism have shown that epigenetic-modifying enzymes often function combinatorically [27]; therefore, perturbation of a single epigenetic regulator may have limited effects, whereas combinatorial perturbation of multiple regulators could induce broader and more pronounced activation [18]. This possibility should be explored in future experiments. Nevertheless, these results do not diminish the value of our approach. Rather, they highlight a key advantage over conventional chemically based activation strategies, which are inherently product-driven and may overlook regulatory events that do not lead to detectable metabolite accumulation. By directly reporting promoter activity through a growth phenotype, our platform captures functional regulator-cluster relationships that might otherwise be missed. Finally, the rapidly expanding catalog of secondary-metabolite gene clusters revealed by large-scale genome sequencing, together with the broad implementation of CRISPR-Cas9-based genome-editing technologies in filamentous fungi, highlights the considerable potential of this platform for dissecting the regulatory networks underlying fungal secondary metabolism. With efficient protoplast-transformation methods and CBE systems now established in genera including Aspergillus, Penicillium, Myceliophthora, and Trichoderma [18,28,29], this framework offers a broadly applicable strategy for systematically elucidating the epigenetic regulation of secondary metabolism across diverse fungal species. Overall, this platform provides a scalable and generalizable approach for linking epigenetic perturbations to the activation of silent biosynthetic gene clusters and for decoding the regulatory networks that govern fungal secondary metabolism. Methods and materials Strains and culture conditions A. nidulans A1145 ΔSTΔEM [30] was used as the host strain in this study. In this strain, the core pks genes of the highly expressed sterigmatocystin (stcA, AN7825) and emericellamide (easA, AN2545) gene clusters were knocked out [30]. In this study, these two expressed genes were also excluded. The strain was cultured on glucose minimal medium (GMM) plates at 30°C for 3 days. Conidia were harvested with sterile water and inoculated into 500 mL flasks containing 100 mL of GMM, followed by overnight incubation at 30°C and 220 rpm. GMM was composed as follows (g/L): 10 g/L D-(+)-glucose, 6.0 g/L NaNO3, 0.52 g/L KCl, 0.52 g/L MgSO4 7H2O, 1.52 g/L KH2PO4, 1.0 mL/L trace element solution (containing 22 g/L ZnSO4·7H2O, 11 g/L H3BO3, 5 g/L MnCl2·4H2O, 1.6 g/L FeSO4·7H2O, 1.6 g/L CoCl2·5H2O, 1.6 g/L CuSO4·5H2O, 1.1 g/L (NH4)6Mo7O24·4H2O, and 50 g/L Na4EDTA), and 20 g/L agar, with the pH adjusted to 6.5. Escherichia coli TStbl3 chemically competent cells (Tsingke Biotechnology Co., Ltd., Beijing) were used as the host for plasmid cloning. Construction of polyketide synthase gene expression cassettes driven by the native promoter for pyrG expression Auxotrophic screening strains were engineered for 21 pks genes (Table S1) in A. nidulans. The open reading frame of each target pks gene was replaced with the pyrG gene from A. oryzae via homologous recombination. The bar gene conferring glufosinate resistance was used for selection. The upstream and downstream homology arms of each pks gene were PCR-amplified from A. nidulans genomic DNA. Fusion PCR was then performed to link the upstream homology arm, pyrG, bar, and the downstream homology arm, generating the complete knockout cassette. Primers were listed in Table S3. Construction of an epigenetic regulator inactivation library A plasmid library for the inactivation of 24 epigenetic regulators was constructed as follows. Two sgRNAs were designed for each gene (Table S2) and individually cloned into the pYTP-sgRNAs vector using Gibson assembly. Each target sequence was fused to a gRNA scaffold to form an sgRNA, with the A. nidulans glycine tRNA used as the promoter to drive sgRNA expression. This vector harbors the AMA1 replicon and the pyroA selectable marker. The sgRNA expression cassette and the plasmid map are shown in Fig. S7. All sgRNA expression plasmids were mixed in equal concentrations to form the inactivation library. The DNA concentration of the mixed library was 1000 ng/μL. The nCas9-CBE expression plasmid used was identical to that in a previously reported case [18]. Construction of epigenetic regulator knockout cassettes For the regulatory genes kmt5 and kdmB, which were identified in the primary screen as activating pks expression, knockout cassettes were constructed. Upstream and downstream homology arms of each regulator were amplified from A. nidulans genomic DNA, and the riboB auxotrophic marker was amplified from pytR plasmid. The knockout cassettes were constructed via fusion PCR. Protoplast preparation and transformation of A. nidulans All genetic manipulations were performed using the polyethylene glycol (PEG)-mediated protoplast transformation method. For protoplast preparation, A. nidulans conidia were inoculated into liquid GMM and cultured overnight at 30°C with shaking at 220 rpm. Mycelia were collected by centrifugation and subjected to enzymatic digestion with Yatalase and Vinotaste Pro at 30°C and 80 rpm for 4–5 h to generate protoplasts. The resulting protoplasts were resuspended in STC buffer (1.2 M sorbitol, 10 mM CaCl2·2H2O, 10 mM Tris-HCl, pH 7.5). For constructing pks knockout strains, 5 μg of each DNA cassette was mixed with 100 μL of protoplast suspension and incubated on ice for 20 min. After adding 0.6 mL of PEG solution (60% PEG4000, 50 mM CaCl2·2H2O, 50 mM Tris-HCl, pH 7.5), the mixture was incubated at 30°C for 20 min. The mixture was then plated on GMM-sorbitol (1.2 M) plates containing 5% glufosinate (for bar resistance) for selection. For CRISPR library transformation, 5 μg of nCas9-CBE plasmid and 5 μg of the pYTP-sgRNAs library plasmid were co-transformed into protoplasts. Transformants were first screened on GMM plates lacking pyroA and riboB. Spores were collected and further plated on GMM plates lacking uracil and uridine for activation screening. Single colonies capable of growing on selective plates were ultimately picked, and genomic DNA was extracted for PCR amplification and sequencing verification of the sgRNA target sequences. Quantitative real-time PCR (qRT-PCR) analysis The wild-type strain (A1145 ΔSTΔEM) and mutant strains (kmt5 KO and kdmB KO) were inoculated in GMM and cultured at 37°C for 48 h. Mycelia were collected, snap-frozen in liquid nitrogen, and ground into a powder. Total RNA was extracted using the Quick-RNA Fungal/Bacterial Microprep Kit, and 1 μg of RNA was used as a template for reverse transcription. qPCR was performed on a QuantStudio real-time PCR system using the ChamQ SYBR qPCR Master Mix. The β-actin gene (acn, AN6542) from A. nidulans was used as the internal reference, and relative gene expression was calculated using the 2–ΔΔCt method. Experiments were performed independently three times (biological replicates), and data are presented as mean ± standard deviation. For RT-qPCR analysis of non-target PKS clusters, five non-target silent PKS clusters [31] including AN1034 (asperfuranone BGC), AN3230 (aspernidine BGC), AN6000 (asperthecin BGC), AN7903 (felinone A BGC), and AN11191 ((+)-asperlin BGC), were selected to represent clusters located on different chromosomes and with distinct basal expression levels. The core pks gene of each cluster was quantified using the primers listed in Supplementary Table S5. Data are presented as mean ± SD from three biological replicates. Primers used for qPCR are listed in Supplementary Table S4. Chromatin immunoprecipitation-qPCR (ChIP-qPCR) analysis ChIP assays were performed according to established methods [18]. Briefly, mycelia cultured for 48 h were crosslinked with 1% formaldehyde for 10 min, and the reaction was quenched with 125 mM glycine. After grinding in liquid nitrogen, the mycelia were sonicated to shear the DNA into 200–500 bp fragments. Chromatin fragments (200 μg of protein) were incubated overnight at 4°C with 2 μg of specific antibodies, including anti-H3K4me3 (Abcam, ab8580), anti-H4K20me3 (Abcam, ab9053), and control antibodies targeting the C-terminus of histone H3 (Abcam, ab1791) and H4 (Abcam, ab10158). Subsequently, 30 μL of Dynabeads® Protein A magnetic beads were added to capture the antibody-chromatin complexes. Eluted DNA was purified using a PCR purification kit. Quantitative PCR analysis of target regions was performed using Luna® Universal qPCR Master Mix (primers listed in Table S5). Data are expressed as the ratio of DNA immunoprecipitated by the specific antibody (e.g., H3K4me3) relative to that immunoprecipitated by the control antibody (total H3), representing fold enrichment. Experiments were performed independently three times, and results are shown as mean ± standard deviation. Cell culture and metabolic profiling For agar plate cultures, A. nidulans mutants were grown on GMM at 30°C for 5 days. Subsequently, three agar plugs (7 mm in diameter) were excised using a cork borer and transferred to a 15 mL centrifuge tube. For metabolite extraction, 3 mL of methanol was first added, followed by 3 mL of ethyl acetate, with sonication for 1 h at each step. The combined extracts were pooled in a glass tube and evaporated to dryness. The extracted metabolites were redissolved in 0.3 mL of DMSO:MeOH (1:4). Finally, 10 μL of the sample was injected into an Ultraviolet High Performance Liquid Chromatography (UV HPLC, 1260 Infinity II, Agilent, USA) system equipped with a reverse-phase column (Shim-pack GIST C18-AQ column, 5 μm, 4.6 × 250 mm, Shimadzu). The binary gradient elution system consisted of (A) methanol and (B) water (containing 0.1% formic acid, v/v), and separation was achieved using the following gradient: 0–5 min, 5% A; 5–20 min, 5% to 95% A; 20–25 min, 95% A; 25–30 min, 5% A. The flow rate was 1.0 mL/min, and the column oven was operated at 30°C. Detection wavelengths were 210 nm and 254 nm. CRediT authorship contribution statement Meng Pang: Investigation, Data curation. Donghui Guo: Data curation. Yanling Li: Writing – original draft. Yuze Chen: Investigation. Rumeng Du: Investigation. Jiayang Liu: Investigation. Jiahui Hu: Investigation. Hang Li: Investigation. Zhuofan Li: Investigation. Guojun Pan: Writing – review & editing, Methodology. Yuan Tian: Writing – review & editing. Fanglong Zhao: Writing – review & editing, Funding acquisition. Declaration of generative AI in scientific writing Generative AI and AI-assisted technologies were only used in the writing process to improve the readability and language of the manuscript. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work was supported by the National Natural Science Foundation of China (Nos. 32300067), the Natural Science Foundation of Shandong Province (Nos. 2024HWYQ-057 and Nos. ZR2020MH368). Footnotes Peer review under the responsibility of Editorial Board of Synthetic and Systems Biotechnology. Supplementary data to this article can be found online at https://doi.org/10.1016/j.synbio.2026.10.001. Contributor Information Guojun Pan, Email: gjpan@sdfmu.edu.cn. Yuan Tian, Email: ytian@sdfmu.edu.cn. Fanglong Zhao, Email: zhaofanglong@sdfmu.edu.cn. Supplementary data The following is/are the supplementary data to this article: Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1