Recombinant AnAFP Reveals Genetic Contributors to Antifungal Protein Tolerance and Fungal Development in Aspergillus flavus

Aspergillus flavus (A. flavus) is a saprophytic filamentous fungus with broad ecological adaptability, commonly colonizing plant debris, animal silage, and deteriorated grain-derived substrates [1], with a particular prevalence in maize, cottonseed, and tree nuts [2]. Beyond its environmental and agricultural relevance, A. flavus is also recognized as an opportunistic human pathogen and has been recovered from tracheal aspirates of patients with COVID-19 [3]. A major public health concern associated with this species is its capacity to biosynthesize aflatoxins, including aflatoxin B₁ (AFB₁), B₂, G₁, G₂, and related derivatives, among which AFB₁ is regarded as one of the most potent naturally occurring hepatocarcinogens [4]. Conventional strategies to control A. flavus contamination have relied largely on fungicides. However, the extensive use of azole-class antifungal agents has contributed to the emergence of isolates with clinically relevant resistance phenotypes. For example, a clinical surveillance study conducted between 2020 and 2023 in central China reported that 18.5% of clinical A. flavus isolates exhibited resistance to itraconazole [5]. These observations underscore the urgent need to develop alternative, environmentally compatible strategies to suppress A. flavus proliferation and reduce aflatoxin contamination. In this context, antifungal proteins and peptides have attracted increasing attention as promising “green” antifungal agents because of their biodegradability, suitability for fermentative production, and generally low cytotoxicity toward mammalian cells [6].

Antimicrobial proteins and peptides comprise a diverse group of bioactive molecules produced across a wide phylogenetic spectrum, ranging from bacteria to humans. Among them, antifungal proteins, hereafter referred to as AFPs, represent a distinct class of antifungal biomolecules. In filamentous ascomycetes, AFPs are predominantly produced by species belonging to Aspergillus, Penicillium, Neosartorya, and related genera [7]. Representative examples include the antifungal protein AFP from Aspergillus giganteus [8], the Penicillium chrysogenum antifungal protein PAF [8], and the anti-yeast protein NFAP2 from Neosartorya fischeri [9]. These proteins generally exhibit potent antifungal activity against agriculturally important phytopathogens, including Fusarium graminearum and Botrytis cinerea [10,11]. AFPs have been isolated from native fungal cultures; however, this approach is frequently constrained by low production yields, labor-intensive purification procedures, and limited scalability [12]. Recent work from our group has demonstrated that codon optimization and fusion with solubility-enhancing partners, such as thioredoxin (Trx), can substantially improve the recombinant production of AFPs in Escherichia coli (E. coli) [13]. Nevertheless, whether this strategy can be broadly applied to other AFPs or structurally related antifungal proteins remains to be further validated through experimental investigation.

Small cysteine-rich antifungal proteins have been recognized for several decades; nevertheless, their precise mechanisms of action remain incompletely understood. Current evidence indicates that these proteins exert antifungal activity through multiple cellular processes. One major mechanism involves disruption of fungal cell wall integrity, partly through inhibition of chitin synthase activity in susceptible fungi, including Aspergillus niger (A. niger), Aspergillus oryzae, and Fusarium oxysporum [14,15,16,17]. In addition, some AFPs can compromise plasma membrane integrity by inducing pore formation and potassium efflux, whereas PAF has been reported to promote membrane hyperpolarization followed by calcium influx [18]. Beyond cell wall disruption and membrane perturbation, AFPs can also interfere with key intracellular signaling networks. For example, the Penicillium chrysogenum antifungal protein PAF has been shown to activate heterotrimeric G-protein signaling and modulate downstream PKC/MPK and cAMP/PKA pathways in sensitive fungal cells [19]. Recently, our group identified Npt1 as a previously uncharacterized protein that regulates AFP activity, further suggesting that the antifungal effects of AFPs may involve additional molecular determinants [13]. In this study, we focus on AnAFP, a cysteine-rich antifungal protein originally isolated from A. niger [20]. AnAFP is of particular interest because its antifungal spectrum, feasibility for recombinant production, and mechanistic basis remain insufficiently characterized. Moreover, whether our previously established expression strategy can be effectively applied to AnAFP requires further investigation.

The fully sequenced A. niger genome and available transcriptomic and proteomic datasets for the reference strain CBS 513.88 [21] provide useful resources for studying AnAFP activity and its potential molecular mechanisms. Genetic screening of the plant defensin HsAFP1 in Saccharomyces cerevisiae identified 13 resistant mutants and 71 hypersensitive mutants, many of which carried mutations in genes associated with mitochondrial function [22]. Notably, ado1, a gene essential for S-adenosylmethionine (SAM) biosynthesis, was identified as a key determinant of HsAFP1 resistance [23]. In addition, def1, which encodes a nuclear RNA polymerase II degradation factor, and adk1, which encodes adenylate kinase involved in adenine nucleotide homeostasis, ATP metabolism, and DNA replication, have been shown to modulate cellular sensitivity to HsAFP1 [22,24]. These findings suggest that susceptibility to antifungal proteins is closely associated with mitochondrial function, nucleotide metabolism, and broader cellular energy-homeostasis pathways in fungi. However, the roles of the corresponding ado1, def1, and adk1 homologs in the pathogenic and aflatoxigenic fungus A. flavus, particularly in relation to AnAFP sensitivity, remain to be elucidated. Investigating these homologs may provide new mechanistic insights into how AnAFP affects pathogenic fungal physiology and may help identify conserved determinants of susceptibility to cysteine-rich antifungal proteins.

In this study, we expressed and purified the recombinant antifungal proteins AnAFP, PgAFP, and AfAFP in E. coli and confirmed their inhibitory activities against A. flavus. To elucidate the molecular determinants underlying their antifungal effects, we generated A. flavus deletion mutants of ado1, def1, and adk1 using homologous recombination and examined their sensitivities to each AFP. These findings provide a foundation for further studies on the antifungal activity and potential mechanisms of AFPs in A. flavus.

The E. coli strains DH5α and BL21 were used for plasmid DNA preparation and recombinant protein expression, respectively. In this study, the wild-type A. flavus (WT) strain (Δku70) was maintained in our laboratory. The CA14PTs strain (Δku70; ΔpyrG), used to construct other mutant strains, was kindly provided by Dr. Perng-Kuang Chang [25]. The growth of the A. flavus strains was evaluated on YGT media (comprising 6 g/L yeast extract, 20 g/L glucose, 1 mL/L trace elements, and 1.5% agar) at 37 °C or 28 °C. Sclerotia production of the strains was assessed on CM media (complete medium containing 6 g/L yeast extract, 6 g/L peptone, 10 g/L sucrose, and 1.5% agar), and aflatoxin biosynthesis was examined in a YES liquid culture (yeast extract/sucrose medium; 2% yeast extract, 150 g/L sucrose, and 1 g/L MgSO₄·7H₂O) at 28 °C. The trace element solution (100 mL) consisted of 2.2 g of ZnSO₄·7H₂O, 1.1 g of H₃BO₃, 0.5 g of MnCl₂·4H₂O, 0.5 of g 0.5 g of FeSO₄·7H₂O, 0.17 g of CoCl₂·5H₂O, 0.16 g of CuSO₄·5H₂O, 0.005 g of (NH₄)₆Mo₇O₂₄·5H₂O, and 4.45 g of Na₂EDTA.

The A. flavus strain referred to as WT in this study was the Δku70 strain maintained in our laboratory. The CA14PTs strain, a Δku70; ΔpyrG derivative, was used as the recipient strain for targeted gene deletion because the ΔpyrG background enables selection with the A. fumigatus pyrG marker. The Δado1, Δdef1, and Δadk1 mutants were generated in the CA14PTs/Δku70 background by homologous recombination and were compared with the Δku70 WT strain in subsequent phenotypic analyses. A homologous recombination strategy was used to generate a deficient strain. In brief, the upstream and downstream flanking regions of the target gene and the A. fumigatus pyrG nutritional marker were amplified using specific primers. The fragments were then purified using a gel extraction kit and subsequently fused using overlap extension PCR to construct a homologous recombination cassette. The final fusion PCR product was further purified and transformed into CA14PTs strain protoplasts. The protoplasts were prepared by digesting the cell wall of freshly collected mycelium using an enzymatic solution containing 20 mM NaH₂PO₄ (pH 5.8), 20 mM CaCl₂, 1.2 M NaCl, Snailase (7.5 g/mL), lysozyme (7.5 g/mL), lywallzyme (7.5 g/mL), and 3 mg/mL driselase. The mixture was incubated at 29 °C with shaking at 180 rpm for 1–1.5 h. The efficiency of the enzymatic lysis of the mycelium’s cell wall was assessed under a microscope after 1 h. The truncated mutation strains of ado1, def1, and adk1 were all generated in this manner. Putative positive transformants were validated using diagnostic PCR [26].

In this study, we chose the mature form of AnAFP to express in the E. coli system. The gene encoding AnAFP was produced by using full gene synthesis with codon optimization and subsequently cloned into a pET-32a expression vector. The resulting constructs included a 6×His-tag and a Trx-tag at the N-terminus, followed by a PreScission protease recognition site before the target protein. After verification, the plasmids were individually transformed into BL21. The transformed cells were propagated in 10 mL of LB medium supplemented with 100 µg/mL ampicillin, and incubated at 37 °C with 180 rpm agitation for 12 h. Subsequently, cultures expressing either AnAFP were scaled up to 4 L of fresh LB medium with 100 µg/mL ampicillin and maintained under identical conditions until the optical density at 600 nm reached 0.5. Cells were then collected, resuspended in pre-cooled lysis buffer (comprising 50 mM Tris-HCl, 500 mM NaCl, and 20 mM imidazole, adjusted to pH 7.4), and disrupted by sonication on ice. The lysates were clarified by centrifugation at 12,000× g for 20 min at 4 °C for three times, and the supernatant was then applied to a Ni-NTA affinity column. A stepwise elution was performed using buffers with increasing concentrations of imidazole (from 20 to 300 mM) to selectively bind and subsequently elute the His-tagged fusion protein. The fusion tag was cleaved by incubating the protein with PreScission protease 3C at 4 °C for 16 h, followed by post-dialysis of the eluted fractions against a buffer composed of 50 mM Tris-HCl and 10 mM NaCl. The proteolytic reaction mixture was further subjected to ion-exchange chromatography using a Cytiva S cation exchange chromatography column. A linear gradient of NaCl was used for elution in an ӒKTA pure™ chromatography system, which yielded highly purified AnAFP. Size-exclusion chromatography was performed for the final purification using a Superdex 75 10/300 GL column, which was equilibrated and eluted with a buffer of 50 mM Tris-HCl and 150 mM NaCl. The recombinant protein AnAFP was resolved by electrophoresis [26].

Each 1.5-mL Eppendorf tube was filled with 1 mL of YGT liquid medium. Gradient concentrations of PgAFP, AfAFP, and AnAFP proteins were added to the medium. The MIC was defined as the lowest protein concentration required to completely inhibit macroscopic fungal growth. To achieve a final spore concentration of 1 × 10⁴ spores/mL, 1 µL of a spore suspension containing 1 × 10⁷ spores/mL was introduced into each tube. The fungal strains ado1, def1, adk1, or the WT strain were incubated at 37 °C with continuous shaking at 180 rpm for 48 h [27].

RT–qPCR was performed as previously described, with minor modifications. [28,29]. To remove possible residual genomic DNA (gDNA), the total RNA (5 µg) was first treated with DNase I (Thermo Fisher Scientific, Waltham, MA, USA), gDNA free RNA (1 µg) was reverse-transcribed into cDNA by using the HRbioTMⅢ Aid First-strand cDNA Synthesis kit (Heruibio, Fuzhou, China). Finally, the quantitative reverse transcription-polymerase chain reaction (RT-qPCR) was performed with the SYBR Green Premix kit (Takara, Dalian, China), and the instrument M×3000p thermocycler (Agilent Technologies, Santa Clara, CA, USA). The 2^−ΔΔCt method was applied to evaluate the expression levels of corresponding target genes. The primers of the RT-qPCR assay are listed in Table S1. The β-actin gene was selected as the reference gene.

The colony diameter was evaluated using YG solid media plates inoculated with a spore suspension at a concentration of 10⁷ spores/mL. The plates were then incubated at 37 °C for 4 days in the dark. Post-incubation, photographs were obtained to document the colony morphology, and the colony diameters were measured. Conidia production was compared using YGT solid media plates similarly inoculated with 10⁷ spores/mL and incubated at 37 °C for 4 days in the dark. Following incubation, the plates were washed with sterile water to collect the spores. Conidial count was determined using a hemocytometer. Sclerotia production was analyzed using fresh spore suspensions of both mutant and WT strains at 10⁷ spores/mL inoculated onto CM solid medium plate, and incubated in the dark at 37 °C for 7 days. The aerial mycelium and conidia were removed by washing with 75% ethanol, and the plates were photographed. The sclerotia morphology was captured using a stereomicroscope, and the sclerotia were quantified from the magnified images based on a quarter section of the colony [30].

To extract aflatoxin from the cultured liquid medium, fresh spore suspensions of various A. flavus strains, including WT strain and gene-edited strains, at a concentration of 10⁷ spores/mL were inoculated into 10 mL of YES liquid medium. The cultures were incubated in the dark at 29 °C for 5 days. Subsequently, 7 mL of the post-cultivation medium was collected, and an equivalent volume of dichloromethane was added to each sample. After thorough mixing, the resulting mixture was shaken at 180 rpm for 30 min and then immediately centrifuged. A 5 mL aliquot of the dichloromethane layer (the lower layer) was extracted for subsequent analysis. For the analysis, 10 µL of each dissolved sample was applied to silica gel plates. A developing solvent with a dichloromethane to acetone ratio of 9:1 was used. Plates were imaged under UV illumination using identical exposure settings within each experiment. AFB₁ band intensity was quantified by densitometric analysis using GeneTools software (version 4.3.9). The integrated density of each AFB₁ band was measured, and the local background signal was subtracted from the same lane. The background-corrected intensity of each sample was normalized to the corresponding WT control on the same TLC plate, which was set as 1.0. Data were calculated from three independent biological replicates and are presented as mean ± SD [31].

A pathogenicity analysis of deletion strains on peanut and maize seeds was conducted following the method. To prevent germination, embryos were removed from both types of seeds. Uniformly sized seeds were selected and sterilized by washing with 0.5‰ sodium hypochlorite for 3 min, followed by rinsing with sterile water to remove any sodium hypochlorite residue. Subsequently, the seeds were soaked in 75% ethanol for 15 s, washed several times with sterile water, and soaked in sterile water for 5 min to completely remove the ethanol. The sterilized seeds were then placed on sterilized filter paper to remove excess water. For testing the pathogenicity of the deletion strains, sterilized peanut or maize seeds were immersed in a spore suspension with 10³ spores/mL for 30 min. Filter paper was placed on Petri dishes, 1 mL of sterile water was added, and the seeds were placed on the filter paper. The seeds were incubated at 29 °C in the dark for 7 days. For the spore count analysis, 20 mL of sterile water was added to the cultured seeds, and the spore suspension was diluted before counting using a hemocytometer. For metabolite extraction, dichloromethane was added to the spore suspension at a 1:1 ratio. The resulting mixture was incubated at 37 °C with continuous shaking at 180 rpm for 10 min, Thin-layer chromatography analysis of aflatoxins was conducted following the method described above [32].

The samples were chosen through a random selection process to ensure an unbiased and representative dataset. Data analysis was conducted using GraphPad Prism 10.0 and OriginPro 2023, and statistical significance was assessed by means of one-way ANOVA. Differences were considered statistically significant as * p < 0.05, ** p < 0.01, and *** p < 0.001. All analyzes were performed on a minimum of three biological replicates unless otherwise specified.

As noted in the introduction, the nomenclature of AFPs has expanded with the continued discovery of new family members [33,34]. To further investigate AFP-mediated inhibition of fungal growth, we analyzed an AFP from A. niger, hereafter referred to as AnAFP. Protein sequence information and structural annotation of AnAFP were obtained from the UniProt database. AnAFP was identified from A. niger through sequence alignment (Figure 1A). ProtScale analysis indicated that hydrophilic regions, characterized by negative GRAVY values, were more prominent than hydrophobic regions (Figure 1B). SignalP 6.0 predicted a Sec/SPI signal peptide at residues 1–18, with a cleavage site between residues 18 and 19 (Figure 1C). Accordingly, the signal peptide was excluded from the recombinant expression construct. Consistent with this prediction, UniProt analysis showed low confidence scores across residues 1–36 (Figure 1D), suggesting that this region may be intrinsically disordered [35]. Because intrinsically disordered regions often mediate weak, context-dependent interactions and may affect recombinant expression, this N-terminal segment was removed for expression in E. coli. Homology modeling using SWISS-MODEL predicted that AnAFP adopts a tertiary structure similar to that of PeAFPB (Figure 1E), and this structural similarity was further supported by alignment using RCSB PDB tools (Figure 1F). Together, these bioinformatic analyses suggest that AnAFP shares structural features with known AFPs and may possess antifungal activity.

The coding sequence of AnAFP was obtained by whole-gene synthesis and subsequently inserted into the pET-32a expression vector to generate the pET-32a-AnAFP construct (Figure 2A). Recombinant clones were verified by PCR using T7 promoter and T7 terminator primers, and the correct insert was further confirmed by Sanger sequencing. After confirmation, small-scale expression was performed to confirm the production of Trx-6×His-AnAFP (Figure 2B). The recombinant fusion protein was then purified by Ni-NTA affinity chromatography (Figure 2C). The Trx-6His fusion tag was then removed as shown in Figure 2C, and the target protein was further purified by cation-exchange chromatography using a Cytiva cation-exchange column. In a low-salt buffer containing 50 mM Tris-HCl and 10 mM NaCl at pH 7.4, the basic properties of AnAFP allowed selective binding to the cation-exchange column, whereas most contaminating proteins were removed (Figure 2D). SDS-PAGE analysis showed that the resulting protein preparation had relatively high purity (Figure 2E). The purity and homogeneity of AnAFP were further assessed by size-exclusion chromatography (Figure 2F), confirming a purified protein preparation (Figure 2G). PgAFP and AfAFP were purified according to a previously described method [13]. In vitro antifungal assays using concentration gradients showed that all three purified antifungal proteins, AnAFP, PgAFP, and AfAFP, inhibited the macroscopic growth of A. flavus under the tested purified-protein assay conditions. Based on the macroscopic growth endpoint after 48 h of incubation, the MIC values for AnAFP, PgAFP, and AfAFP were determined to be 16, 8, and 4 µM, respectively. Representative cultures at these concentrations are shown in Figure 2H.

Previous studies have shown that the fungal genes ado1, def1, and adk1 are involved in resistance to the Heuchera sanguinea-derived antifungal protein HsAFP1 [22]. To determine whether these genes also contribute to the response of A. flavus to AFPs, the Δado1, Δdef1, and Δadk1 deletion strains were constructed by homologous recombination. The WT strain and three deletion mutants were then treated with AnAFP, PgAFP, and AfAFP, respectively. Compared with the WT strain, the Δado1, Δdef1, and Δadk1 mutants showed increased sensitivity to all three AFPs (Figure 3A). To further examine the response of A. flavus to AnAFP, the transcript levels of three regulatory genes, laeA, veA, and wetA, were analyzed. laeA and veA are key regulators involved in fungal secondary metabolism and development [36,37], whereas wetA plays an important role in conidial maturation and spore integrity [38]. Following AnAFP treatment, laeA expression exhibited distinct response patterns between the WT and mutant strains. Compared with the corresponding untreated controls, laeA expression was markedly downregulated in the WT and Δado1 strains but was induced in the Δadk1 mutant (Figure 3B). The veA expression was significantly reduced in the WT strain and Δadk1 strains (Figure 3C). A wetA expression was strongly downregulated in the WT strain and Δdef1 strains, whereas it was upregulated in the Δadk1 mutant (Figure 3D). Because AnAFP treatment and deletion of ado1, def1, and adk1 affected fungal development and secondary metabolism-related phenotypes, we selected laeA, veA, and wetA as representative regulatory genes associated with these processes. These results suggest that ado1, def1, and adk1 influence AFP sensitivity in A. flavus and that their deletion alters AnAFP-induced transcriptional responses of genes associated with fungal development and secondary metabolism.

Deletion of ado1, def1, and adk1 markedly affected the growth and development of A. flavus. Compared with the WT strain, the Δado1, Δdef1, and Δadk1 mutants all exhibited reduced colony diameters (Figure 4A), with the most pronounced reductions observed in the Δado1 and Δdef1 strains (Figure 4B). On YGT medium, colonies of the Δadk1 strain appeared completely white (Figure 4A), suggesting altered pigment production. Microscopic observation showed that all three deletion mutants formed smaller conidial heads and displayed a sparser distribution of conidiophores than the WT strain (Figure 4C). Consistently, spore quantification revealed that all three mutants produced markedly fewer spores than the WT strain (Figure 4D). Sclerotial development was also affected by the deletion of these genes. In the Δado1 strain, sclerotia were more concentrated near the center of the culture plate (Figure 4E), although the total number of sclerotia was comparable to that of the WT strain (Figure 4F). In contrast, the Δdef1 and Δadk1 strains produced substantially fewer sclerotia than the WT strain (Figure 4F), suggesting that def1 and adk1 contribute to sclerotial formation in A. flavus. Aflatoxin production was further examined by thin-layer chromatography (TLC) (Figure 4G). Quantification of TLC band intensity showed that the Δado1 and Δadk1 strains accumulated lower levels of aflatoxin than the WT strain (Figure 4H,J), whereas the Δdef1 strain showed increased aflatoxin production (Figure 4I).

To assess whether ado1, def1, and adk1 influence the expression of developmental and secondary metabolism-related regulators, we quantified the transcript levels of laeA, veA, and wetA by RT-qPCR. Under normal growth conditions without AnAFP treatment, laeA was upregulated in Δado1 relative to the WT strain, whereas veA was downregulated in Δdef1. In Δadk1, both laeA and wetA showed lower basal expression levels than those in the WT strain (Figure 4K).

Given the close association between fungal development, secondary metabolism, and stress adaptation, we further assessed the responses of the three deletion mutants to different stress conditions. Under SDS-induced plasma membrane stress, all three mutants exhibited higher growth inhibition rates than the WT strain, indicating increased sensitivity to membrane perturbation (Figure 5A). Under CFW-induced cell wall stress, the Δado1 mutant displayed an approximately threefold lower growth inhibition rate than the WT strain (Figure 5A). Under H₂O₂-induced oxidative stress, the Δado1 and Δdef1 mutants showed reduced growth inhibition relative to the WT strain, whereas the Δadk1 mutant showed a response comparable to that of the WT strain under the same condition (Figure 5B). Together, these results suggest that ado1, def1, and adk1 differentially affect colony growth, conidiation, sclerotial development, aflatoxin production, and stress in A. flavus.

To further assess the contribution of ado1, def1, and adk1 to infection-related pathogenicity in A. flavus, peanut and maize seed infection assays were performed. In peanut infection assays, the three deletion mutants (Δado1, Δdef1, and Δadk1) showed visibly reduced colonization compared with the WT strain, as indicated by weaker mycelial growth and reduced sporulation on the seed surface (Figure 6A). Quantification further confirmed that all three mutants produced markedly fewer spores on infected peanut seeds than the WT strain (Figure 6B). TLC analysis showed reduced AFB₁ accumulation in the Δado1 and Δadk1 mutants during peanut infection (Figure 6C,E,F), whereas the Δdef1 mutant produced AFB₁ at levels comparable to those of the WT strain (Figure 6C,D). A similar reduction in host colonization was observed in maize infection assays. Compared with the WT strain, the Δado1, Δdef1, and Δadk1 mutants displayed weaker fungal growth and sporulation on maize kernels (Figure 6G), accompanied by markedly reduced spore production (Figure 6H). TLC analysis further showed that AFB₁ accumulation was reduced in the Δado1 and Δadk1 mutants (Figure 6 I,K,L), whereas the Δdef1 mutant accumulated higher levels of AFB₁ than the WT strain during maize infection (Figure 6I,J). These results suggest that deletion of ado1, def1, or adk1 impairs seed colonization and sporulation by A. flavus, whereas their effects on AFB₁ accumulation vary depending on the deleted gene and host substrate.

Here, we characterized the interaction between AnAFP and A. flavus. We established a recombinant expression and purification strategy that yielded soluble AnAFP with inhibitory activity against A. flavus. We further investigated the contribution of ado1, def1, and adk1 to AFP susceptibility and fungal developmental processes in A. flavus. Our findings suggest that AnAFP has potential as an antifungal protein against A. flavus and that ado1, def1, and adk1 contribute to AFP tolerance, fungal development, stress responses, and aflatoxin production.

Cysteine-stabilized AFPs, particularly those produced by Eurotiales fungi, have attracted considerable interest as potential antifungal agents because of their small size and potent antifungal activity [39,40]. However, the heterologous production of cysteine-rich AFPs remains technically challenging. Unlike well-established yeast expression systems, filamentous fungi naturally form mycelial structures during cultivation, which makes large-scale and standardized protein production less convenient. In A. niger, AnAFP expression is closely associated with carbon starvation, nutrient recycling, and autophagy [41,42]. Previous studies have also reported a temporal disconnect between the accumulation of ANAFP transcripts and the extracellular detection of AnAFP [41,42]. In prokaryotic systems such as E. coli, inefficient disulfide-bond formation, protein misfolding, and inclusion body formation can further limit the yield of soluble and functional AFPs [12,43,44]. These factors highlight the difficulty of achieving reliable production of functional AFPs using native or conventional expression systems. To overcome these challenges, we established an E. coli-based expression strategy adapted to the predicted structural features of AnAFP. Similar expression strategies had previously been applied to PgAFP and AfAFP [13], prompting us to test whether this approach could also support the soluble production of AnAFP. Instead of expressing the full-length precursor, we focused on the predicted mature region of AnAFP, excluding the signal peptide, putative propeptide, and intrinsically disordered N-terminal segment. Although the removed N-terminal residues may contribute to secretion, processing, folding efficiency, or stability in the native fungal host, they are not predicted to constitute the mature cysteine-rich antifungal domain. The retained antifungal activity of recombinant AnAFP indicates that the predicted mature region is sufficient to inhibit A. flavus. This strategy enabled soluble expression and subsequent purification of recombinant AnAFP, providing a practical basis for evaluating its antifungal activity against A. flavus.

Mutant analysis showed that deletion of ado1, def1, or adk1 was associated with increased susceptibility of A. flavus to AnAFP, suggesting that these genes may contribute to AFP tolerance. adk1 encodes adenylate kinase, an enzyme involved in the interconversion of adenine nucleotides and the maintenance of cellular energy balance [45,46]. ado1 encodes adenosine kinase, which phosphorylates adenosine and participates in the recycling of adenosine generated through the methyl cycle [23,47]. Considering these previous findings together with our results, we speculate that the increased AnAFP sensitivity of the Δado1 and Δadk1 mutants may be related to a reduced ability to adjust nucleotide and energy metabolism under AFP-induced stress. Because adk1 is involved in adenine nucleotide homeostasis and ATP metabolism, deletion of adk1 may reduce the capacity of A. flavus to buffer AnAFP-induced stress. The induction of laeA after AnAFP treatment in the Δadk1 mutant may therefore represent stress-associated transcriptional rewiring rather than a simple restoration of the basal developmental programme. However, further biochemical and metabolic analyses will be required to test this possibility directly. In contrast, def1 may contribute to AnAFP tolerance through a different process, possibly related to transcriptional stress management. In contrast, def1 may contribute to AnAFP tolerance through a different process, possibly related to transcriptional stress management [24,48]. AFPs are known to impose cell-wall and membrane-associated stress in susceptible fungi, including inhibition of chitin synthesis, perturbation of membrane integrity, and activation of the cell-wall integrity pathway [14,18,49]. Therefore, rapid transcriptional adjustment may be important for fungal adaptation during AFP exposure. In this context, deletion of def1 could potentially impair the ability of A. flavus to coordinate stress-responsive transcriptional responses upon AnAFP treatment. Collectively, these findings suggest that the increased AnAFP sensitivity of Δado1, Δadk1, and Δdef1 may not result from disruption of a single resistance pathway. The current results should be interpreted as preliminary evidence that ado1, def1, and adk1 are associated with AFP sensitivity, fungal development, stress responses, aflatoxin production, and infection-related phenotypes in A. flavus. Future studies involving complemented strains, metabolic profiling, and transcriptome-wide analyses will be necessary to establish gene-specific causality and define the mechanisms by which these genes contribute to AFP tolerance and fungal physiology.

laeA, veA, and wetA are important regulatory genes associated with development, sporulation, stress responses, and secondary metabolism in A. flavus [38,50,51]. The velB-veA-laeA complex has been proposed to link developmental programmes with secondary metabolic output, and laeA is widely regarded as a global regulator of secondary metabolite gene clusters [52,53,54,55]. In A. flavus, veA and laeA have been reported to contribute to aflatoxin biosynthesis, sclerotial formation, and seed colonization [56]. By contrast, wetA is a multifunctional regulator that couples conidial maturation and survival with stress tolerance and chemical development in A. flavus [38]. In our study, AnAFP treatment induced different expression patterns of laeA, veA, and wetA in the WT strain and deletion mutants. These mutant transcriptional changes suggest that deletion of ado1, def1, or adk1 may alter the transcriptional response of A. flavus to AnAFP. Thus, the altered expression of laeA, veA, and wetA may represent an indirect consequence of broader changes in developmental and secondary metabolism-related regulatory networks upon AnAFP exposure. We acknowledge that the RT-qPCR analysis in this study was limited to selected developmental and secondary metabolism-related regulators. Although SDS, CFW, and H₂O₂ assays provided phenotypic evidence that deletion of ado1, def1, and adk1 alters stress responses, additional transcriptional analysis of cell wall-, membrane-, and oxidative stress-related genes will be needed to define the stress-response pathways involved in AFP tolerance more comprehensively.

The following supporting information can be found at: https://www.sciepublish.com/article/pii/1091, Table S1: The quantitative reverse transcription-polymerase chain reaction primers used in this study.

Conceptualization, Z.L. and Y.W.; Methodology, Y.W.; Software, Z.L., S.W. (Sen Wang) and X.H.; Validation, J.S., Z.L. and S.W. (Sen Wang); Formal Analysis, Y.W.; Investigation, S.W. (Shihua Wang); Resources, Y.W.; Data Curation, Z.L.; Writing—Original Draft Preparation, Z.L.; Writing—Review & Editing, Y.W. and S.W. (Shihua Wang); Visualization, Z.L., S.W. (Sen Wang), Y.W. and J.S.; Supervision, S.W. (Shihua Wang); Project Administration, S.W. (Shihua Wang); Funding Acquisition, S.W. (Shihua Wang).

Not applicable.

Not applicable.

The authors confirm that the data supporting the findings of this study are available within the article.

This work was supported by the National Natural Science Foundation of China (Grant No. 32370205), the Natural Science Foundation of Fujian Province (Grant No. 2026J001289) and the Science and Technology Innovation Project of FujianAgriculture and Forestry University (Grant No. KFB23082A).

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.