Peroxin Pex14/17 Is Required for Trap Formation, and Plays Pleiotropic Roles in Mycelial Development, Stress Response, and Secondary Metabolism in Arthrobotrys oligospora

ABSTRACT The peroxins encoded by PEX genes involved in peroxisome biogenesis play a crucial role in cellular metabolism and pathogenicity in fungi. Herein, we characterized a filamentous fungus-specific peroxin Pex14/17 in the Arthrobotrys oligospora, a representative species of nematode-trapping fungi. The deletion of AoPEX14/17 resulted in a remarkable reduction in mycelial growth, conidia yield, trap formation, and pathogenicity. Compared with the wild-type strain, the ΔAopex14/17 mutant exhibited more lipid droplet and reactive oxygen species accumulation accompanied with a significant decrease in fatty acid utilization and tolerance to oxidative stress. Transcriptomic analysis indicated that AoPEX14/17 was involved in the regulation of metabolism, genetic information processing, environmental information processing, and cellular processes. In subcellular morphology, the deletion of AoPEX14/17 resulted in a decrease in the number of cell nuclei, autophagosomes, and Woronin bodies. Metabolic profile analysis showed that AoPex14/17 affects the biosynthesis of secondary metabolites. Yeast two-hybrid assay revealed that AoPex14/17 interacted with AoPex14 but not with AoPex13. Taken together, our results suggest that Pex14/17 is the main factor for modulating growth, development, and pathogenicity in A. oligospora. IMPORTANCE Peroxisome biogenesis genes (PEX) play an important role in growth, development, and pathogenicity in pathogenic fungi. However, the roles of PEX genes remain largely unknown in nematode-trapping (NT) fungi. Here, we provide direct evidence that AoPex14/17 regulates mycelial growth, conidiation, trap formation, autophagy, endocytosis, catalase activity, stress response to oxidants, lipid metabolism, and reactive oxygen species production. Transcriptome analysis and metabolic profile suggested that AoPex14/17 is involved in multiple cellular processes and the regulation of secondary metabolism. Therefore, our study extends the functions of PEX genes, which helps to elucidate the mechanism of organelle development and trap formation in NT fungi and lays the foundation for the development of efficient nematode biocontrol agents.

influence of AoPex14/17 on global changes in gene transcription levels was analyzed by RNA sequencing. Meanwhile, the regulation of AoPex14/17 in the secondary metabolism was determined by the metabolome. Our analysis shows that AoPex14/17 is required for trap formation and regulates a variety of phenotypic features, including mycelial development, conidiation, fatty acid utilization, and stress response.

RESULTS
Identification of AoPEX14/17 in A. oligospora. The compute pI/MW tool of the ExPASy server (http://web.expasy.org) predicted that the AoPEX14/17 encodes a polypeptide of 456 amino acid residues, with a predicted molecular mass of 49.27 kDa and an isoelectric point of 4.78, and AoPex14/17 contained one conserved domain of Pex14_N (IPR006785). AoPex14/17 showed a high similarity (71.83 to 95.14%) to the orthologs of four other NT fungal species, including Arthrobotrys flagrans, Dactylellina haptotyla, Drechslerella stenobrocha, and Drechslerella brochopaga, and it has a moderate similarity to the orthologs of Aspergillus nidulans (44.44%) and other filamentous fungi (29.52 to 27.70%), such as Metarhizium robertsii, Fusarium graminearum, and Aspergillus fumigatus. As expected, Pex14/17 orthologs from NT fungi clustered into an independent branch (see Fig. S1A in the supplemental material).
Transcriptomic insights into the regulatory role of AoPEX14/17. The wild-type (WT) and DAopex14/17 strains were inoculated in PD broth for 5 and 7 days, and then we collected them and performed comparative RNA-seq analysis. A total of about 42.58 to 58.27 million reads were obtained per sample (see Table S1 in the supplemental material). The quality control data show that the base error rates were both less than 0.03%, while the Q20 was greater than 97.72%, the Q30 was greater than 93.62%, and the GC content was greater than 47.57% (Table S1). Principal-component analysis (PCA) suggested the gene expression patterns of WT and mutant strains were distributed in various quadrants at different time points with high similarity, indicating consistency between biological replicates (see Fig. S2A in the supplemental material). The accuracy of the transcriptome data was verified by reverse transcription-quantitative PCR (RT-qPCR). Twenty genes associated with the mitogen-activated protein kinase (MAPK) signaling pathway, glycolysis/gluconeogenesis, and ubiquitin-mediated proteolysis were selected for validation, and the results show that the RT-qPCR results were consistent with transcriptome clustering results ( Fig. S2B to D).
Compared with the WT strain, 2,819 and 2,431 differentially expressed genes (DEGs) were enriched in the DAopex14/17 strain at 5 and 7 days, respectively. The numbers of upregulated DEGs were 1,163 and 1,052, and those downregulated DEGs were 1,656 and 1,379, respectively (Fig. 1A); Venn analysis displayed that 1,576 DEGs were shared at 5 and 7 days in the DAopex14/17 strains (Fig. 1B). Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed that the DEGs involved in the metabolic pathways were divided into the following five categories: metabolism, genetic information processing, environmental information processing, cellular processes, and organismal systems. Among them, the highest proportion of pathways belonged to the metabolism category. The enrichment results were highly similar between day 5 and day 7 ( Fig. 1C and D), with regard to oxidative phosphorylation, glycolysis/gluconeogenesis, nitrogen metabolism, pentose phosphate pathway, fatty acid degradation, fatty acid biosynthesis, citrate cycle (tricarboxylic acid [TCA] cycle), ribosome, proteasome, protein processing in the endoplasmic reticulum, spliceosome, ubiquitin-mediated proteolysis, the MAPK signaling pathway, peroxisome, endocytosis, cell cycle, autophagy, phagosome, and the longevity-regulating pathway. The difference was that pyruvate metabolism in the metabolism category and hippo signaling pathwaymultiple species in the environmental information processing category were enriched at 5 days, while starch and sucrose metabolism, tyrosine metabolism and biosynthesis of unsaturated fatty acids, DNA replication in the genetic information processing category, and ABC transporters and mitophagy in the cellular processes category were enriched at 7 days ( Fig. 1C and D).
GO enrichment analysis of the DEGs was performed. On the 5th day, the most enriched GO term was biological process (BP), and the top three items of BP were cellular metabolic process (294 unigenes), biosynthetic process (175 unigenes), and oxidation-reduction process (99 unigenes); in the cellular component (CC) category, the top three items were intrinsic component of membrane (384 unigenes), non-membrane-bounded organelle (84 unigenes), and ribosome (72 unigenes); in the molecular function (MF) category, the top items were oxidoreductase activity (101 unigenes), transporter activity (61 unigenes), and RNA binding (41 unigenes) (see Fig. S3A and B in the supplemental material). On the 7th Pex14/17 is Required for Trap Formation and Virulence mSphere day, in the BP category, the most intrinsic items were metabolic process (303 unigenes), oxidation-reduction process (85 unigenes), and transmembrane transport (65 unigenes); the DEGs enriched in the CC category on day 7 were similar to the membrane intrinsic component (293 unigenes) on day 5; in the MF category, the top three items were catalytic activity (299 unigenes), oxidoreductase activity (87 unigenes), and transporter activity (51 unigenes) ( Fig. S3C and D). AoPEX14/17 regulates vegetative growth and hyphal morphogenesis. To characterize the effect of Aopex14/17 on the mycelium growth and differentiation in A. oligospora, we inoculated WT and DAopex14/17 mutant strains on tryptone-glucose (TG), potato dextrose agar (PDA), and tryptone yeast-extract glucose agar (TYGA) media, respectively, and compared their hyphal growth and colony morphology at 6 days postinoculation. Compared with that of the WT strain, the DAopex14/17 mutant showed a significant decreased radial growth on all three media ( Fig. 2A and B). With regard to the defects in radial growth, we visualized the mycelia septa and morphology using calcofluor white (CFW) staining. The results indicate that the mycelia septum in the DAopex14/17 mutant was not significantly different from that observed in the WT strain, and partial hyphal cells of the mutant became enlarged (Fig. 2C). Furthermore, the hyphal cells of the WT strain contained 8 to 10 nuclei, whereas the mutants contained 2 to 3 nuclei ( Fig. 2D and E).
AoPEX14/17 is required for conidiation, trap formation, and pathogenicity. The DAopex14/17 mutant produced approximately 2.0 Â 10 3 to 4.0 Â 10 3 conidia/mL, which is only 1% of that produced by the WT strain ( Fig. 3A and B). The RT-qPCR results demonstrate that the expression levels of several genes involved in conidiation from filamentous fungi, including wetA, abaA, velB, fluG, brlA, and hyp1, were significantly downregulated in the DAopex14/17 mutant (Fig. 3C). Moreover, we assessed the ability of WT and DAopex14/17 strains to produce traps by inducing with nematodes (Caenorhabditis elegans) to water agar (WA) plates. Our results demonstrate that the DAopex14/17 mutant was unable to produce mature traps (Fig. 3D). In the DAopex14/17 mutant, half rings or one ring appeared at 12 h, one and one-half ring appeared at 24 h, two rings appeared at 36 h, and three rings appeared at 48 h, but no mature traps were formed until 5 days ( Fig. 3D and E). Additionally, the nematode mortality of the WT strain was as high as 96% at 48 h, whereas the nematode mortality of the DAopex14/17 strain was less than 10% at each time point (Fig. 3F). Thus, deletion of the AoPEX14/17 gene significantly reduced the ability of the strain to infect nematodes.
AoPEX14/17 contributes to stress tolerance. To determine the function of Aopex14/ 17 in stress response, we treated strains to compounds that stimulate oxidative (menadione and H 2 O 2 ), cell wall (Congo red [CR] and sodium dodecyl sulfate [SDS]), and osmotic stress (NaCl and sorbitol). The DAopex14/17 strain was sensitive to high concentrations of menadione and H 2 O 2 , and the relative growth inhibition (RGI) values were significantly higher than those of the WT strain ( Fig. 4A and B). In the transcriptome analysis, eight genes involved in oxidative stress response were significantly decreased on day 5 and day 7, including cat (AOL_s00173g374 and AOL_s00188g243), cat-2 (AOL_s00006g411 and AOL_s00076g488), nox-1, nox-2, sod, sodB, and sod-2 (Table 1).
Moreover, the RGI value of the DAopex14/17 mutant was higher than that of the WT strain under stress of SDS (0.01-0.03%), and the DAopex14/17 mutant showed a decreased RGI value compared to the WT strain at 0.1 mg/L CR (see Fig. S4A and B in the supplemental material). Similarly, the DAopex14/17 mutant showed a higher RGI value than the WT strain at 0.3 M NaCl, whereas no obvious defects in stress responses to sorbitol were observed ( Fig. S4C and D). We also examined the expressions of genes involved in cell wall biosynthesis in the transcriptome; except for hex and glu, the expression levels of other genes were significantly decreased on day 5 and day 7, including chs-3, chs, trs, and gls (Table 1).
AoPEX14/17 regulates intracellular autophagy, ROS accumulation, and endocytosis. Monodansylcadaverine (MDC) staining was used to monitor the autophagy process, and autophagosomes in hyphae displayed blue fluorescence after MDC staining (30). Compared with that of WT, the DAopex14/17 strain showed an increased accumulation of autophagosomes (Fig. 5A). Accordingly, 10 and 17 genes related to autophagy were differentially expressed at days 5 and 7, respectively, and three genes encoding cell division control protein Cdc48 (AOL_s00004g527) and serine protease (AOL_s00170g103 and AOL_s00176g50) were upregulated in the DAopex14/17 strain at both day 5 and day 7 by transcriptome analysis (see Fig. S5A in the supplemental material).
After staining with dihydroethidium (DHE) and nitrotetrazolium blue chloride (NBT), the hyphae of the DAopex14/17 strain were more intensely stained by NBT than the WT strain, indicating greater reactive oxygen species (ROS) accumulation in the DAopex14/ 17 strain ( Fig. 5B and C). Moreover, the fluorescence intensity levels of DHE in the WT and DAopex14/17 strains were 57.5 and 69.5, respectively (Fig. 5D). Similarly, we determined catalase (CAT) activity in the WT and DAopex14/17 strains. The results show that the activity of CAT was reduced in the DAopex14/17 strain (Fig. 5E). This is consistent with our transcriptome analysis of the significantly reduced expression levels of CAT activity-related genes cat and cat-2 (Table 1).
To clarify the role of AoPEX14/17 in endocytosis, we stained the hyphal cells of WT and DAopex14/17 strains with FM4-64 and followed the uptake of the dye over time. In the WT strain, the dye first appeared on the plasma membrane and was taken up by the cells after 10 min. In contrast, in the DAopex14/17 strain, the dye uptake was accelerated and localized to the cytoplasm, which may be mature endosomes or vacuoles (Fig. 5F).
AoPEX14/17 is required for the formation of WBs and impairs biogenesis of other organelles. Transmission electron microscopy (TEM) revealed that WBs were observed near the septa in the hyphae of the WT, whereas they were not observed in DAopex14/17 hyphae (Fig. 5G), and the hyphal septum of the DAopex14/17 strain was broken up, consistent with the role of WBs that prevent the multiple organelles and intracellular substance from effluxing (Fig. 5G). We also found that the volume and number of vacuoles were significantly increased in the DAopex14/17 strain compared with those of the WT strain (Fig. 5H). However, the number of mitochondria and mitochondrial cristae decreased significantly, and the mitochondrial shape became longer, wider, and vacuolated in the DAopex14/17 hyphae (Fig. 5I).
Pex14/17 is Required for Trap Formation and Virulence mSphere a marked decrease in the expression level of AoPEX14/17, consistent with which Aopex14/ 17 is knocked out (Fig. S5B).
AoPEX14/17 deletion led to defects in lipid metabolism. Peroxisome plays a significant role in fatty acid b-oxidation (31). We investigated whether AoPEX14/17 affects fatty acid level by staining with the neutral lipid-specific fluorescent dye boron dipyrromethene (BODIPY). The number of lipid droplets (LDs) in the DAopex14/17 strain was reduced, but the volume was increased compared to that of the WT strain (Fig. 6A). To investigate the role of AoPEX14/17 in fatty acid utilization, the WT and DAopex14/17 mutant strains were inoculated on MM plates containing short-chain (sodium acetate), medium-chain (Tween 20), or long-chain (oleate) fatty acids as the sole carbon source, respectively (Fig. 6B). After 6 days, the growth rate of the DAopex14/17 strain was remarkably reduced on all three plates compared to the WT strain. The RGI values of the DAopex14/17 strain on plates supplemented with all three fatty acids were markedly increased (Fig. 6B). Additionally, we also observed larger LDs in the hyphae of the DAopex14/17 strain by TEM (Fig. 6C).
AoPEX14/17 involvement in the regulation of the biosynthesis of secondary metabolites. High-performance liquid chromatography (HPLC) results indicate that the metabolic abundance produced by the DAopex14/17 strain was significantly less than that of the WT strain, and the differential peaks were mainly concentrated at retention times (Rt) of 20 to 38 min (Fig. 7A). There were 2,232 upregulated and 3,314 downregulated   Table S2 in the supplemental material. KEGG analysis of the differential metabolic pathways between WT and DAopex14/17 strains showed that the upregulated pathways were involved in lipoxygenase, trichothecene biosynthesis, scopolin and esculin biosynthesis, aromatic compounds, etc. (Fig. S6C and Table S2), and the downregulated pathways mainly included aerobic toluene degradation, steroid hormone biosynthesis, rosmarinic acid degradation biosynthesis, and scopolin and esculin biosynthesis ( Fig. S6D and Table S2). Ultraperformance liquid chromatography with diode array detection mass spectrometry (UPLC-DAD/MS) spectral comparison showed that the ion peak at Rt 35 min was much higher in DAopex14/17 strains than in WT (Fig. 7C). MS analysis of this peak revealed a diagnosis of arthrobotrisins, a specific metabolite of A. oligospora, at m/ z 139, 393, and 429 (Fig. 7D).
AoPex14/17 can interact with AoPex14 but not with AoPex13. We attempted to probe the interaction between AoPex14/17 and two other docking complex members, AoPex13 and AoPex14, by yeast two-hybrid (Y2H) analysis. Firstly, we chose positive controls including plasmids pGBKT7-53 and pGADT7-T, while negative controls included pGBKT7-Lam and pGADT7-T. Furthermore, the growth of transformants was measured by serial dilution of yeast cells on SD/Trp/Leu, SD/Trp/Leu/His/Ade, and SD/Trp/Leu/His/ Ade/-bromo-4-chloro-3-indolyl-b-D-galactopyranoside (X-Gal) media, respectively. Ultimately, the results of Y2H experiments reflect that AoPex14/17 does not interact with AoPex13 but is able to interact with AoPex14 (see Fig. S7A and B in the supplemental material).

DISCUSSION
The maintenance of the peroxisome depends on the formation of the peroxisomal Pex14/17 is Required for Trap Formation and Virulence mSphere membrane and the import of both membrane and matrix proteins (18). The matrix proteins that are translocated to the peroxisomal membrane must interact with the docking machinery components, involving multiple peroxins, such as Pex1, Pex2, Pex5, Pex6, Pex7, Pex10, Pex12, Pex13, Pex14, and Pex26 (32,33). Pex14/17, a filamentous fungusspecific peroxin of the docking machinery of peroxisomes, is required for the import of PMPs via binding Pex5 and Pex14 and may act as ScPex14 and ScPex17 in the docking of the peroxisomal matrix by expressing PEX14-and PEX17-similar portions of the proteins (19,22). In this study, we first characterized the presence of homologs of PEX14/17 in the NT fungus A. oligospora and found that it is essential for the regulation of functions of peroxisomes and has a significant impact in a variety of phenotypic features through the disruption of AoPEX14/17. Peroxisome has been linked to the growth, development, and pathogenicity of fungi (29). In Penicillium chrysogenum, the loss of PEX14/17 leads to a defect in sporulation (19). In M. oryzae, the loss of PEX14/17 delayed conidial generation and appressorial formation, reduced appressorial turgor accumulation and penetration ability in host plants, and significant decreased fungal virulence; however, PEX14/17 deletion showed no significant difference with regard to colony morphology and aerial hyphal development and only a slight reduction in radial growth (21). In M. robertsii, the conidial germination, conidiation, and fungal virulence were all greatly reduced upon PEX14/17 depletion, but there were no appreciable changes between colony phenotypes and the colony diameter of the PEX14/17 mutant strains and control strains on various conditions in the vegetative growth assay (34). In this study, we provide direct evidence that the knockout of PEX14/17 represses conidial production and germination and impairs trap formation and pathogenicity in A. oligospora, which are consistent with previous reports. Comparatively, the loss of AoPEX14/17 significantly reduces hyphae growth on the three different media. The reduced nuclei involved in the regulation of meiotic recombination (35) and the enrichment of the cell cycle indicated by transcriptome analysis indicate that AoPex14/17 may play a positive role in vegetative growth in A. oligospora. In addition, the morphology of the hyphae was changed in the DAopex14/17 mutants, including inflated mycelial cells. These results indicate that AoPex14/17 is necessary for mycelial growth, conidial production and germination, and the pathogenicity of A. oligospora and other filamentous fungi.
In fungi, besides common functions including the b-oxidation of fatty acids and the detoxification of hydrogen peroxide, peroxisomes are crucial for the metabolism of various unique carbon and nitrogen sources, including oleic acid, methanol, primary amines, and uric acid, and play an important role in response to stress (36,37). Previous evidence has highlighted a decreasing ability for the control of fatty acid utilization, ROS degradation, stress tolerance, and cell wall integrity in fungi without Pex14/17 (21,34,38). Our results indicate that the disruption of AoPEX14/17 also resulted in a significant decrease in fatty acid utilization and tolerance to oxidative stress. In transcriptome analysis, there are nine genes involved in lipid metabolism that were downregulated in the DAopex14/17 strain versus WT strain, such as glycerophosphocholine phosphodiesterase, phospholipase D, alkaline ceramidase, and ethanolamine kinase; these genes are mainly involved in fatty acid catabolic processes, such as glycerophospholipid metabolic pathways and hydrolases and transferases of various lipids (39,40). Obviously, the hyphae of DAopex14/17 mutants exhibited bigger lipid droplets and a much greater accumulation of ROS associated with the lower enzyme activity of CAT. Moreover, Pex14/17 plays a crucial role in peroxisome biogenesis (21). These results reveal that the downregulation of peroxisomal functions caused by a lack of AoPEX14/17 was triggered by a defect of peroxisome biogenesis in A. oligospora. However, no obvious inhibition was observed in DAopex14/17 mutants under chemical stressors of CR, NaCl, and sorbitol, indicating that AoPex14/17 may play a dispensable role in cell wall integrity and hyperosmotic stresses in A. oligospora.
Intriguingly, we observed that the terms of autophagy, mitophagy, and ubiquitin-mediated proteolysis were significantly enriched by transcriptome analysis, and subsequently, the remarkable increase in the level of autophagosome was confirmed by fluorescent staining in the DAopex14/17 mutants. Autophagy plays a vital role in the maintenance of peroxisome mass, and autophagy further regulates peroxisomes and contributes to the production of ROS (41). Peroxisome abundance can be regulated by PEX genes. The upregulation of autophagy level may be caused by the shortage of integrated peroxisome in DAopex14/17 mutants. Whether the regulation of PEX14/17 in autophagy and endocytosis inhibits the peroxisome function or alters matrix protein transport remains unknown. In addition, DAopex14/17 mutants generally increase the endocytosis process but lack WBs, which are peroxisome-related organelles necessary for sealing the septal pore in the event of hyphal damage (42,43). Consistently, our transcriptome data suggest that Pex14/17 affects multiple metabolic pathways involved in carbohydrate metabolism, the MAPK signaling pathway, and mitochondrial function. TEM analysis revealed that the number of mitochondria and mitochondrial cristae decreased significantly in the DAopex14/17 mutants. Our findings support a model in which cellular energy metabolism relies heavily on the peroxisome, and the structure and function of peroxisomes and mitochondria are closely related in A. oligospora. The metabolome data also show that the abundance of secondary metabolites was significantly reduced upon PEX14/17 deletion in A. oligospora. Fungal secondary metabolites are synthesized by complex biosynthetic pathways catalyzed by enzymes located in different subcellular compartments, including the peroxisome (44). In Penicillium chrysogenum, Pex14/17 is involved in the penicillin biosynthesis process (19). Most importantly, the loss of AoPEX14/17 contributed to the accumulation of arthrobotrisins of A. oligospora. Arthrobotrisins, polyketide synthase-terpenoid synthase hybrid metabolites found in A. oligospora and other NT fungi, have been shown to inhibit hyphal growth and trap formation (35,45). Fungal hyphal growth and pathogenicity were inhibited in DAopex14/ 17 mutants associated with the upregulated arthrobotrisin content. The mechanism underlying the ability of Pex14/17 to regulate these processes in A. oligospora awaits further analysis.
AoPex14, AoPex13, and AoPex14/17 are the components of peroxisomal docking complex, and we demonstrated by Y2H that AoPex14/17 interacted with AoPex14 but not with AoPex13. A previous study found that MoPex14 interacted with MoPex13 and MoPex14/17, but MoPex13 did not bind to MoPex14/17 in M. oryzae (46), which is consistent with our results. The Y2H data support the conclusion that AoPex14 binds primarily to AoPex14/17 and is essential for the maintenance of AoPex14/17 on the peroxisomal membrane.
In our recent work, AoPex1 and AoPex6 are required for mycelial growth, trap formation, and conidiation of A. oligospora (29), and the DAopex1 and DAopex6 mutants showed a more significant phenotypic difference than the DAopex14/17 mutant, such as mycelial growth and trap formation. Pex1 and Pex6 are both cytosolic and membrane-associated AAA ATPases of the peroxisomal protein import machinery, which have been demonstrated to mediate the ATP-dependent dislocation of the PTS1 receptor from the peroxisomal membrane to the cytosol (47); while Pex14/17 is a docking complex component specific to filamentous fungi and is mainly involved in the docking of receptors and cargoes during peroxisome synthesis (21). Thus, Pex1 and Pex6 are functionally different from Pex14/17, which would result in their phenotypic differences in A. oligospora.
In conclusion, we characterized the function of Pex14/17 in A. oligospora involved in mycelial growth and development, conidial production, pathogenicity, lipid utilization, lipid metabolism, response to oxidative stress, ROS accumulation, autophagy, endocytosis, the maintenance of structure and function of organelles, and the regulation of secondary metabolites (Fig. 8). These findings provide a basis for exploring the features of peroxins that underlie the survival and environmental adaptations of NT fungi and contribute to developing an effective biocontrol strategy for pathogenic nematodes.

MATERIALS AND METHODS
Strains, media, and growth conditions. Arthrobotrys oligospora ATCC 24927 and its derived mutants were grown on potato dextrose agar (PDA) medium in a constant-temperature incubator (28°C). The plasmid pCSN44 containing the selective marker hygromycin resistance gene (hph) was stored in the Escherichia coli strain DH5a (TaKaRa, Shiga, Japan), and the knockout vector was created using the plasmid PRS426 (48). As previously mentioned, S. cerevisiae (FY834) was cultivated in yeast extract peptone dextrose (YPD) medium for the recombinational cloning technique (49). To investigate mycelial development and associated phenotypic characteristics, the following media were used: TYGA (10 g tryptone, 5 g yeast extract, 10 g glucose, and 5 g molasses per 1 L), TG (10 g tryptone and 10 g glucose per 1 L), and CMY (20 g maizena [corn starch] and 5 g yeast extract per 1 L); additionally, 20 g agar should be added per 1 L of the above media (50). In addition, protoplast regeneration was conducted with PDAS (200 g potato, 0.3 g yeast extract, 0.6 M sucrose, and 10 g molasses per 1 L) medium. The nematode Caenorhabditis elegans was cultured in WA (20 g agar per 1 L) medium. Besides this, the utilization of carbon sources was also tested in MM (20 g glucose, 2 g NaNO 3 , 0.01 g FeSO 4 Á7H 2 O, and 20 g agar per 1 L) medium.
Sequence and phylogenetic analyses of AoPex14/17. The amino acid sequences for Pex14/17 homologs were obtained from the genomic databases of the model fungi S. cerevisiae, N. crassa, and A. nidulans, and the orthologous AoPex14/17 (AOL_s00081g263) was identified in A. oligospora. The NCBI online website (http://www.ncbi.nlm.nih.gov/) was used to download all protein sequences. The sequence similarity of homologous Pex14/17 from diverse fungi was analyzed by DNAman software (Lynnon Biosoft, San Ramon, USA). ClustalW was used to perform multiple sequence alignments, Mega X was used to construct a neighbor-joining tree, and the analysis of bootstrap values was conducted using 1,000 replicates (51).
Targeted gene deletion and Southern blotting. Using the homologous recombination technique, the AoPEX14/17 gene was deleted in the wild-type (WT) strain (52). The upstream and downstream fragments of AoPEX14/17 were PCR amplified from A. oligospora using a modified yeast cloning procedure, and a hygromycin cassette (hph) was amplified using the vector pCSN44 as a template (53). The fragment, hph cassette, and gapped yeast shuttle vector pRS426 (digested with EcoRI and XhoI) were transformed into yeast strain FY834. Next, the recombinant cloned strains were screened by SC-Ura medium as described previously (53,54). The circular construct was created by homologous recombination, and the disruption vector (pRS426-AoPEX14/17-hph) was eventually recovered through transformation into E. coli DH5a. The gene disruption target fragment was amplified using paired primers (see Table S3 in the supplemental material) and transformed into A. oligospora protoplasts. As previously mentioned, hygromycinresistant transformants were selected on PDAS containing 200 mg/mL hygromycin (55). The candidates for the gene deletion mutant were screened from hygromycin-resistant transformants via genomic PCR and confirmed by Southern blotting.
Genomic DNA was digested with NcoI enzyme (New England Biolabs, Inc., Ipswich, MA, USA) for Southern blotting. The analysis was carried out in accordance with the directions included in the North2South chemiluminescent hybridization and detection kit (Pierce, Rockford, IL, USA).
RNA sequencing analysis. WT and mutant strains were incubated in PD (200 g potatoes and 20 g glucose per 1 L) broth medium at 28°C for 5 and 7 days, respectively. Three replicates were used for   Table S3 (23, 24). Finally, differentially expressed genes (DEGs) were screened out using P # 0.05 and Log 2 (fold change) $ 2 as thresholds. The GO and KEGG databases were used to functionally assess the selected DEGs (56).
Comparison of mycelial growth and conidia yield. To determine mycelial growth, WT and mutant strains were inoculated onto TG, PDA, and TYGA media at 28°C for 5 days, respectively, with growth rates and colony diameters measured at 24-h intervals (57). To analyze sporulation, conidia were collected from 14-day CMY agar plates with 5 mL sterile distilled water. Conidia produced by a given colony were counted and reported as the total number of conidia present per unit area of the colony using a hemocytometer (58).
Trap formation and bioassay. To observe trap formation, 50-mL conidia suspensions (4 Â 10 6 conidia mL 21 ) of WT and mutant strains were evenly distributed on WA plates at 28°C for 3 to 4 days. Then, to induce trap formation, approximately 300 to 400 nematodes (C. elegans) were added to each WA plate (59). The number of traps produced and nematodes captured by each strain were counted using a microscope (Olympus, Japan) at 12, 24, 36, and 48 h.
Analysis of stress tolerance. To test the sensitivity of each strain to various environmental stresses, mycelial plugs taken from the periphery of each strain were inoculated on TG medium amended with the following compounds: menadione (0.01, 0.02, and 0.03 mM) and H 2 O 2 (2.5, 5, and 10 mM) as oxidative stressors; Congo red (CR) (0.05, 0.07, and 0.1 mg/mL) and sodium dodecyl sulfate (SDS) (0.01, 0.02, and 0.03%) as cell wall stress agents; and NaCl (0.1, 0.2, and 0.3 M) and sorbitol (0.25, 0.5, and 0.75 M) as osmotic stressors. After incubation at 28°C for 6 days, the colony diameter in each plate was measured, and the relative growth inhibition (RGI) of each colony was evaluated as described previously (60). Each experiment was repeated three times.
Analysis of fatty acid utilization. For the lipid utilization assay, WT and mutant strains were inoculated on MM medium supplemented with sodium acetate (50 mM), oleic acid (0.12%), or Tween 20 (0.5%) as the sole carbon source. The size of the fungal colonies was measured and photographed postincubation at 28°C for 6 days, after which RGI values were determined (29).
Analyses of reactive oxygen species and endocytosis. By staining with nitrotetrazolium blue chloride (NBT) (Solarbio, Beijing, China), the ROS levels in the hyphae were determined. WT and mutant strains were grown on PDA at 28°C for 3 days, after which plates were stained with 20 mL of a 0.2% NBT solution and incubated in the dark at 28°C for 30 min. After draining the supernatant, ethanol was used to wash the plates. The plates were reincubated in the dark at 28°C for 30 min before imaging (61).
Cellular ROS accumulation in the fungal hyphae was determined by dihydroethidium (DHE) (Beyotime, Shanghai, China) staining. Briefly, the collected mycelia were mixed with 10 mg/mL DHE solution, incubated at 37°C for 30 min, and then washed with phosphate-buffered saline (PBS) twice (62). Subsequently, ROS levels were estimated from the intensity of DHE fluorescence with the same weight of mycelium via multimode microplate reader (Molecular Devices, Shanghai, China).
For TEM analysis, WT and mutant strains were harvested from colonies cultured on PD broth for 3 days. Then, the mycelia were collected and fixed with 2.5% glutaraldehyde for TEM observation (30).
Reverse transcription-quantitative PCR. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) from frozen fungal tissues. Then, the treated RNA samples were reversely transcribed into cDNA using the PrimeScript RT reagent kit (TaKaRa, Shiga, Japan). The above experiments were performed according to the manufacturer's instructions. RT-qPCR was accomplished by LightCycler 480 with SYBR green I master mix (Roche, Basel, Switzerland) and a specific primer set (Table S3). Besides this, b-tubulin (AOL_s00076g640) was used as an internal control to normalize the target gene's expression level (64). Finally, the relative expression level of duplicated samples was analyzed by the 2 2DDCT method, and each reaction was independently repeated three times with three biological replicates per sample.
Untargeted metabolomics analysis. WT and mutant strains were inoculated in 250 mL of PD broth at 28°C for 5 days. Then, the mycelium was dissolved in 250 mL ethyl acetate, and the fermentation broth was ultrasonically extracted and allowed to stand overnight. The ethyl acetate layer was dried with a rotary evaporator before being dissolved in 1 mL of chromatographic methanol. For liquid chromatography-mass spectrometry (LC-MS) analysis, the methanol solution was filtered through a 0.22-mm membrane filter. Analysis was performed using the Thermo Fisher Scientific Dionex Ultimate 3000 UHPLC system and a Thermo Fisher high-resolution Q precision focusing mass spectrometer (Thermo Fisher Scientific) (30,65,66). Subsequently, the Compound Discoverer 3.3 software package was used to perform metabolite deconvolutions and identifications (66,67).
Statistical analysis. GraphPad Prism (version 9.00) was used for statistical analysis, and Tukey's multiple comparison test was applied to assess the level of difference between different strains. Data are presented as means 6 standard deviation, and a P value of ,0.05 was used to determine significant differences.
Data availability. All data generated or analyzed during this study are included in the published paper and associated supplemental files. The sequencing data were deposited in the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) under accession number GSE216978.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.

ACKNOWLEDGMENTS
We are grateful to the Microbial Library of the Germplasm Bank of Wild Species from Southwest China for preserving and providing experimental strains and to Guo Ying-qi (Kunming Institute of Zoology, Chinese Academy of Sciences) for her help with taking and analyzing TEM images. We declare that no competing interests exist.