Genome-Wide Characterization of the RNA Exosome Complex in Relation to Growth, Development, and Pathogenicity of Fusarium graminearum

ABSTRACT The RNA exosome complex is a conserved, multisubunit RNase complex that contributes to the processing and degradation of RNAs in mammalian cells. However, the roles of the RNA exosome in phytopathogenic fungi and how it relates to fungal development and pathogenicity remain unclear. Herein, we identified 12 components of the RNA exosome in the wheat fungal pathogen Fusarium graminearum. Live-cell imaging showed that all the components of the RNA exosome complex are localized in the nucleus. FgEXOSC1 and FgEXOSCA were successfully knocked out; they are both involved in the vegetative growth, sexual reproduction, and pathogenicity of F. graminearum. Moreover, deletion of FgEXOSC1 resulted in abnormal toxisomes, decreased deoxynivalenol (DON) production, and downregulation of the expression levels of DON biosynthesis genes. The RNA-binding domain and N-terminal region of FgExosc1 are required for its normal localization and functions. Transcriptome sequencing (RNA-seq) showed that the disruption of FgEXOSC1 resulted in differential expression of 3,439 genes. Genes involved in processing of noncoding RNA (ncRNA), rRNA and ncRNA metabolism, ribosome biogenesis, and ribonucleoprotein complex biogenesis were significantly upregulated. Furthermore, subcellular localization, green fluorescent protein (GFP) pulldown, and coimmunoprecipitation (co-IP) assays demonstrated that FgExosc1 associates with the other components of the RNA exosome to form the RNA exosome complex in F. graminearum. Deletion of FgEXOSC1 and FgEXOSCA reduced the relative expression of some of the other subunits of the RNA exosome. Deletion of FgEXOSC1 affected the localization of FgExosc4, FgExosc6, and FgExosc7. In summary, our study reveals that the RNA exosome is involved in vegetative growth, sexual reproduction, DON production, and pathogenicity of F. graminearum. IMPORTANCE The RNA exosome complex is the most versatile RNA degradation machinery in eukaryotes. However, little is known about how this complex regulates the development and pathogenicity of plant-pathogenic fungi. In this study, we systematically identified 12 components of the RNA exosome complex in Fusarium head blight fungus Fusarium graminearum and first unveiled their subcellular localizations and established their biological functions in relation to the fungal development and pathogenesis. All the RNA exosome components are localized in the nucleus. FgExosc1 and FgExoscA are both required for the vegetative growth, sexual reproduction, DON production and pathogenicity in F. graminearum. FgExosc1 is involved in ncRNA processing, rRNA and ncRNA metabolism process, ribosome biogenesis and ribonucleoprotein complex biogenesis. FgExosc1 associates with the other components of RNA exosome complex and form the exosome complex in F. graminearum. Our study provides new insights into the role of the RNA exosome in regulating RNA metabolism, which is associated with fungal development and pathogenicity.

The RNA exosome plays an important role in the degradation and processing of various RNAs and controls the quality of mRNA in response to cell differentiation and environmental changes (11). The specific recognition of the target transcript by RNA exosome is carried out by different cofactors that direct the RNA exosome to specific RNAs for processing or degradation (12,13). Recently, researchers have found that mutation in genes that encode the various subunits of RNA exosome and their cofactors are associated with human diseases. Mutations in EXOSC1, EXOSC2, EXOSC3, EXOSC5, EXOSC8, and EXOSC9 lead to a variety of neurodegenerative diseases (14)(15)(16)(17)(18)(19). Knockdown of the RNA exosome complex cofactors RBM7, SKIV2L, and TTC37 also leads to tissue-specific diseases with complex phenotypes (20,21). In addition, it has been determined that the expression of the component of RNA exosome Dis3 is stage specific, and deletion of DIS3 in Drosophila delays growth and induces the formation of melanoma, leading to death at the second-instar larval stage (22). Also in Drosophila, absence of Rrp6 and Rrp40 resulted in accumulation of small nucleolar RNA (snoRNA), while silencing of dMTR3, dRRP6, dRRP41, and dRRP42 by RNA interference (RNAi) resulted in death of the adult insects (23). The loss of Exosc3 in b cells led to increased expression of transcription start site (TSS)-related antisense transcription RNA (xTSS-RNA) and accumulation of intracellular noncoding RNA (ncRNA), resulting in a high degree of somatic mutation (24). In Arabidopsis thaliana, mutations in Csl4 have no obvious phenotypic changes (25), and homozygous insertion mutations in transfer DNA (T-DNA) of AtRrp44, AtRrp45, and AtRrp46 are lethal (10,26). Heterozygous mutations of AtRrp4, AtRrp41, and AtRrp42 lead to the death of female gametophytes during development (25,27).
In yeast, all the RNA exosome complex genes were indispensable except RRP6 (Exosc10) (28)(29)(30)(31)(32). In Ustilago maydis, posttranscriptional processing is the key factor for the regulation of pathogenicity (33)(34)(35). In Magnaporthe oryzae, the fungus-specific RNAbinding protein Rbp35 regulates the length of 39 untranslated regions (UTRs) of transcripts having developmental and virulence-associated functions (36). MoCwf15, an essential splicing factor of the Prp19-associated component, regulates fungal growth and infection-related developments by modulating the intron splicing efficiency of a subset of genes in the rice blast fungus (37). In Botrytis cinerea, Nop53, a late-acting factor for 60S ribosomal subunit maturation, is crucial for the pathogen's development and virulence (38). As an important part of the RNA quality control system, the RNA exosome complex is mainly involved in normal RNA degradation, abnormal transcription clearance, nuclear rRNA, and snoRNA and ncRNA processing in eukaryotes. In yeast and Arabidopsis, the RNA exosome is necessary for the maintenance of normal growth and development of the organism. In addition, mutations in subunits and cofactors of RNA exosome complex lead to a variety of human diseases. However, the functions of RNA exosome complex in plant-pathogenic fungi are still largely unknown.
Fusarium head blight (FHB) which is caused by Fusarium graminearum can result in serious economic losses, and mycotoxin contamination of wheat grains seriously threatens the health of humans and animals (39). F. graminearum can penetrate plant cells through the stomata, surface openings, or infection structures such as appressoria and infection cushions (40)(41)(42). The phytopathogen injects virulence factors into the host that interfere with the host immune defense factors, and this helps the fungus to spread in and colonize the host and eventually cause disease (43)(44)(45). Deoxynivalenol (DON) is one of the virulence factors of F. graminearum during infection (46). FgTri1 and FgTri4 are two cytochrome P450 oxygenases involved in early and late steps in trichothecene biosynthesis under trichothecene-inducing conditions in F. graminearum, and they are localized to spherical organelles, called toxisomes, that were presumed to be the site of trichothecene biosynthesis (39). PRP4 encodes the only kinase among the spliceosome components; deletion of FgPRP4 not only affects the intron splicing efficiency in over 60% of F. graminearum genes but also causes severe growth defects (47). Similarly, deletion of the RNA-binding protein FgRbp1 leads to reduced splicing efficiency in 47% of the F. graminearum intron-containing gene transcripts that are involved in various cellular processes, including vegetative growth, development, and virulence (48). Deletion mutants of FgSrp1, an SR (serine/arginine)-rich protein, rarely produce conidia, have reduced ascospore ejection and DON production, and cause only limited disease symptoms on wheat heads and corn silks (49). In addition, genetic mutations in specific selective splicing and A-to-I RNA editing during sexual reproduction in F. graminearum can cause defects in ascospore release (50)(51)(52)(53)(54). However, how RNA processing and degradation in pathogenic fungi regulate the mechanism of their unique infectivity and pathogenicity is still unclear.
In this study, we identified 12 candidate components of the RNA exosome complex in F. graminearum, and named them FgExosc1, FgExosc2, FgExosc3, FgExosc4, FgExosc5, FgExosc6, FgExosc7, FgExosc8, FgExosc9, FgExosc10, FgExosc11, FgExoscA, respectively. Among them, ExoscA is specifically distributed in some fungi. Furthermore, all the components of RNA exosome are localized in the nucleus in F. graminearum. FgExosc1 and FgExoscA are important for F. graminearum vegetative growth, development, and pathogenicity. The RBD and N-terminal region of FgExosc1 are both required for its normal localization and function. Transcriptome sequencing (RNA-seq) revealed that FgExosc1 is involved in ncRNA processing, rRNA and ncRNA metabolism, ribosome biogenesis and ribonucleoprotein complex biogenesis. Furthermore, GFP pulldown and coimmunoprecipitation (co-IP) experiments showed that FgExosc1 associates with the other components of the RNA exosome complex in F. graminearum.
To further determine the architecture of RNA exosome components in F. graminearum, we analyzed the conserved domains and motifs of the 12 RNA exosome components. As shown in Fig. S1 in the supplemental material, FgExosc1, FgExosc2, and FgExosc3 have the typical conserved RNA binding domain (RBD) S1; FgExosc4, FgExosc5, FgExosc6, FgExosc7, FgExosc8 and FgExosc9 have the typical RNase_PH domain; FgExosc10 contained the PMC2NT domain at the N terminus and HRDC (helicase and RNase D C-terminal) domain at the C terminus. FgExosc11 contains a PilT N terminus (PIN) and a Vac domain. FgExoscA contains only one SAS10-UTP domain.
Further, we analyzed the phylogenetic relationship of RNA exosome components in F. graminearum in comparison with those in Saccharomyces cerevisiae, Magnaporthe oryzae, Aspergillus niger, Arabidopsis thaliana, and Homo sapiens, where we constructed a phylogenetic tree of the RNA exosome components in these species (Table S2). As shown in Fig. 1, the RNA exosome can be divided into three clades. Exosc1, -2, and -3 make up clade 1, Exosc4, -5, -6, and 10 make up clade 2, and Exosc7, -8, -9, -11, and -A make up clade 3, suggesting the divergence of the RNA exosome components in evolution. The evolutionary relationships of the components of RNA exosome have adaptive changes through evolution of species, which may lead to the different functions of those components in different species.
RNA exosome subunits are localized to nucleolus in F. graminearum. To investigate the subcellular localization of RNA exosome in F. graminearum, we tagged each of the protein components with a green fluorescent protein (GFP), under the control of their native promoters. As shown in Fig. 2, the GFP signal of exosome complex displays distinct punctate signals along the cytoplasm, which we presumed to be nuclei. To check whether the GFP signals are really localized in the nuclei, we cotransformed the pGFP-FgExosc1 construct with a construct expressing the nuclear marker protein Histone1-mCherry into the protoplasts of the wild-type strain PH-1. We found that GFP-FgExosc1 colocalizes with the Histone1-mCherry signals in the nuclei of hyphae and conidia ( Fig. 2A). To further confirm the precise localization of the RNA exosome within the nucleus, we cotransformed GFP-FgExosc1 with each of two constructs expressing the nucleolus marker FgNucleolin-mCherry and the nucleoplasm marker FgNcbp2-mCherry (a nuclear cap-binding protein) into the protoplasts of the wild-type strain PH-1 and observed their colocalizations by confocal microscopy. The results showed that GFP-FgExosc1 mainly colocalizes with FgNucleolin-mCherry ( Fig. 2B and C), indicating that FgExosc1 is mainly localized to the nucleolus. Furthermore, we found that the other components of the RNA exosome are all colocalized with mCherry-FgExosc1 in the fungal mycelia and conidia (Fig. 2D). Taken together, these results indicate that all the components of the RNA exosome are mainly localized in the nucleolus of F. graminearum.
FgExosc1 and FgExoscA are important for vegetative growth, conidiation, and sexual development of F. graminearum. To investigate the biological functions of RNA exosome in F. graminearum, FgEXOSC1 and FgEXOSCA were successfully knocked out by homologous recombination (Fig. S2). However, we could not obtain mutants of the other 10 genes after several attempts. Phenotypic analysis of the DFgexosc1 mutant shows that its growth rate was severely decreased compared to that of wild-type strain PH-1 ( Fig. 3A and B). On the other hand, deletion of FgEXOSCA has only a slight effect on the vegetative growth of F. graminearum ( Fig. 3C and D). Moreover, as shown in Fig. 3E and F, the DFgexosc1 mutant displayed totally flattened mycelia on complete medium (CM) agar plates and in glass tubes containing CM agar. In addition, the DFgexosc1 mutant displayed aberrant mycelia with swelling and vacuolization in liquid CM compared to wild-type strain PH-1 (Fig. 3E). These results suggest that FgExosc1 is critical for vegetative growth and aerial hyphal development in F. graminearum. We performed gene complementation by reintroducing FgEXOSC1 and FgEXOSCA genes (along with their respective native promoters) into the protoplasts of the DFgexosc1 and DFgexoscA mutants, respectively. We succeeded in generating the complemented DFgexosc1-C and DFgexoscA-C strains, respectively. Consistently, DFgexosc1-C and DFgexoscA-C strains were both found to have the phenotypic defects observed in all the mutants restored to the levels observed in the wild type.
The conidia produced by F. graminearum are known to be the main inocula infecting flowering wheat heads (55). To characterize the role of the RNA exosome in conidium production by F. graminearum, wild-type strain PH-1, the DFgexosc1 and DFgexoscA mutants, and the complemented DFgexosc1-C and DFgexoscA-C strains were inoculated into liquid carboxymethyl cellulose (CMC) medium. As shown in Fig. 4A and B, the conidiation of the DFgexosc1 mutant is clearly reduced in comparison to that produced by wild-type strain PH-1, while the DFgexoscA mutant had no significant difference. In fact, at 6 days postinoculation in CMC medium, the DFgexosc1 mutant failed to produce any conidia. However, very few conidia were recorded when the incubation time of the FIG 1 Phylogenetic analysis of subunits of RNA exosome complex. The phylogenetic tree was constructed with MEGAX software using the maximumlikelihood algorithm. The unrooted phylogenetic tree was plotted using the circular mode in the Interactive Tree Of Life (iTOL) v.6. The accession numbers of all the exosome components are provided in Table S2. mutant culture was extended to 9 days. The number of septa per conidium in the DFgexosc1 and DFgexoscA mutants was similar to that observed in the wild-type strain conidia (Fig. 4C). In addition, we found that the DFgexosc1 mutant lost its ability to form perithecia (Fig. 4D). Although the DFgexoscA mutant produced perithecia normally, the number of ascospores formed was significantly reduced compared to that of wild-type strain PH-1 ( Fig. 4D and E). These results indicate that FgExosc1 is required for conidiation and sexual development of F. graminearum, while FgExoscA is involved only in sexual development.
FgExosc1 and FgExoscA are important for pathogenicity and DON production. To characterize the role of FgExosc1 and FgExoscA in the pathogenicity of F. graminearum, wild-type strain PH-1 and the DFgexosc1, DFgexoscA, DFgexosc1-C, and DFgexoscA-C strains were inoculated on flowering wheat heads. As shown in Fig. 5A, the blight symptoms caused by DFgexosc1 and DFgexoscA mutants spread to the nearby spikelets at much lower rates than those caused by the wild type and the complemented strains under the same conditions. The average disease indices (number of diseased spikelets per head) of the DFgexosc1 and DFgexoscA mutants are 1.42 and 4.74, respectively, while the wild-type strain PH-1, DFgexosc1-C, and DFgexoscA-C strains have indices of 13.44, 13.00, and 13.33, respectively (Fig. 5B). This suggests that FgExosc1 and FgExoscA are both important for the pathogenicity of F. graminearum.
Deoxynivalenol (DON) is one of the virulence factors of F. graminearum during infection (46). To verify whether the reduction in virulence of DFgexosc1 and DFgexoscA mutants is linked to reduction in DON production, we examined whether the FgEXOSC1 and FgEXOSCA deletion mutants are deformed in DON production. To achieve this, we inoculated wild-type strain PH-1, the DFgexosc1 and DFgexoscA mutants, and the complemented DFgexosc1-C and DFgexoscA-C strains into liquid trichothecene biosynthesis induction (TBI) medium. After 7 days of static culture at 28°C, DON production was determined. As shown in Fig. 5C, the amount of DON produced by DFgexosc1 mutant was significantly low (0.89 mg/g) compared to the amount produced by the wild type (34.74 mg/g). In addition, the amount of DON produced by the DFgexoscA strain was nearly half of that produced by the wild type (Fig. 5D). To further support these results, we assayed the expression levels of DON biosynthesis genes (FgTRI1, FgTRI4, FgTRI5, FgTRI6, FgTRI10, and FgTRI12) in the wild-type strain PH-1 and the DFgexosc1 mutant cultured in TBI medium. We found that the expression levels of these genes in the DFgexosc1 mutant were significantly downregulated compared to those in the wild type (Fig. 5E). Consistently, the fluorescence intensities of FgTri1-GFP-and FgTri4-GFP-labeled toxisomes in the DFgexosc1 mutant were significantly weaker than those observed in the wild type under same conditions, while the fluorescence of FgTri1-GFP-and FgTri4-GFP-labeled toxisomes were not significantly difference from those in wild-type strain PH-1 (Fig. 5F). In addition, the relative expression level of FgTRI10 in the DFgexoscA mutant was significantly decreased compared to that in wild-type strain PH-1 (Fig. S3A), while the fluorescence intensities of FgTri1-GFPand FgTri4-GFP-labeled toxisomes were not significantly different from those in wild-type strain PH-1 ( Fig. S3B and C). Together, our data indicated that FgExosc1 and FgExoscA are both required for normal DON production and pathogenicity of F. graminearum.
Functional characterization of the RBD and the N-and C-terminal regions of FgExosc1. As mentioned above, FgExosc1 possesses an RNA-binding domain (RBD; amino acids [aa] 83 to 173) in F. graminearum (Fig. S1). To analyze the roles of this domain, we generated the GFP fusion constructs GFP-FgEXOSC1 DRBD (DRBD), GFP-FgEXOSC1 DN (DN), and GFP-FgEXOSC1 DC (DC) with deletions of the RBD (aa 83 to 173), N-terminal domain (aa 1 to 82), and C-terminal domain (aa 174 to 208), respectively (Fig. 6A). Next, these constructs were transformed into the protoplasts of the DFgexosc1 mutant to obtain the GFP-Fgexosc1 DRBD (DRBD), GFP-Fgexosc1 DN (DN), and GFP-Fgexosc1 DC (DC) mutants, respectively. Analyses of the phenotypes of these mutants revealed that the full-length RBD is required for the full functions of FgExosc1 in vegetative growth, conidiation, and pathogenicity, while the N-terminal region is only partially required (Fig. 6B to E and Fig. S4). However, the C-terminal region is dispensable for the functions of FgExosc1 in F. graminearum.
After deleting the N-terminal region of FgExosc1, we noticed that the GFP signal of GFP-FgExosc1 DN appeared diffuse in the cytoplasm and did not accumulate in the nucleolus (Fig. 6F), suggesting an important role of the N terminus in the nucleolus  localization of FgExosc1. Interestingly, the GFP signal of GFP-Fgexosc1 DRBD displays a ring-or crescent-shaped appearance around the nucleolus but does not concentrate in the nucleolus (Fig. 6F), suggesting that deletion of the RBD of FgExosc1 disrupts its nucleolus localization. In addition, the localization of GFP-FgExosc1 DC (DC) was not significantly affected when its C-terminal region was deleted (Fig. 6F). From these results, we conclude that the N-terminal region and RBD of FgExosc1 are both required for the correct localization and functions of the protein, whereas the C-terminal region is dispensable not only for its function but also for its normal subcellular localization in F. graminearum.
FgExosc1 is involved in processing and metabolism of various RNAs. The eukaryotic RNA exosome complex can process and degrade a variety of RNAs (13). We carried out RNA-seq for transcriptomic analyses of wild-type strain PH-1 and the DFgexosc1  mutants under nutrient-rich conditions. The results revealed that 1,788 genes were downregulated while 1,651 genes were upregulated in the DFgexosc1 mutant, compared to wildtype strain PH-1 (Fig. 7A). Based on GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) analyses, it was found that these differentially expressed genes are mainly related to small-molecule metabolism, ribosome biogenesis, rRNA metabolism and processing, and ncRNA processing (Fig. 7B). Genes that participate in rRNA metabolism and processing, ncRNA processing, and ribosome biogenesis are significantly upregulated (Fig. 7C). These results suggest that FgExosc1 plays an important role in RNA (ncRNA and rRNA) processing and metabolism in F. graminearum.
FgExosc1 associates with the other components of the RNA exosome. The results presented above showed that FgExosc1 colocalizes with the other components of the RNA exosome in F. graminearum. To further establish the relationship between FgExosc1 and the other components of the RNA exosome, co-IP experiments were performed to search for possible interactions among the proteins. Consistently, the results showed that FgExosc1 interacts with FgExosc2, FgExosc3, FgExosc4, FgExosc5, FgExosc6, FgExosc7, FgExosc8, FgExosc9, FgExosc10, FgExosc11, and FgExoscA (Fig. 8). Moreover, the GFP-FgExosc1 strain was immunoprecipitated using a GFP-Trap kit to confirm the above interactions. The proteins bound to the GFP-Trap beads were eluted and identified by liquid chromatographytandem mass spectrometry (LC-MS/MS). Again, FgExosc2, FgExosc3, FgExosc4, FgExosc5,  FgExosc6, FgExosc7, FgExosc8, FgExosc9, FgExosc10, FgExosc11, and FgExoscA were identified in the pulldown assays of GFP-FgExosc1 (Table 1). For in vivo confirmation of these results, a bimolecular fluorescence complementation (BiFC) assay was performed to visualize the interactions in living cells. As shown in Fig. S5  To investigate whether disruption of any one of the RNA exosome components could affect the homeostasis of other components, we first checked the expression levels of other subunits of the RNA exosome in DFgexosc1 and DFgexoscA mutants. As shown in Fig. 10A, deletion of FgEXOSC1 caused downregulation of FgEXOSC5, FgEXOSC8, and FgEXOSC9, while the expression levels of FgEXOSC2, FgEXOSC3, FgEXOSC4, FgEXOSC6, FgEXOSC7, FgEXOSC10, and FgEXOSC11 were not significantly affected. In addition, the expression levels of FgEXOSC2, FgEXOSC3, and FgEXOSC4 in the DFgexoscA mutant were significantly downregulated compared to that in PH-1. These results suggest that loss of a single component of the RNA exosome may affect the homeostasis of other components in F. graminearum.
Next, we further studied whether the deletion of FgEXOSC1 affects the subcellular localizations of the other components of the RNA exosome complex by checking the localizations of the other subunits in the DFgexosc1 mutant. FgExosc2-GFP, FgExosc3-GFP, FgExosc4-GFP, FgExosc5-GFP, FgExosc6-GFP, FgExosc7-GFP, GFP-FgExosc8, FgExosc9-GFP, GFP-FgExosc10, FgExosc11-GFP, and FgExoscA-GFP vectors were transferred into the protoplasts of DFgexosc1 mutant. As shown in Fig. 10B, the localizations of FgExosc2, FgExosc3, FgExosc5, FgExosc8, FgExosc9, FgExosc10, FgExosc11, and FgExoscA were not significantly altered in the absence of FgEXOSC1 compared to the wild type. However, the localizations of FgExosc4, FgExosc6, and FgExosc7 were significantly different from those in the wild type; the signals of FgExosc4-GFP, FgExosc6-GFP, and FgExosc7GFP accumulated in the rings around the nucleolus in the DFgexosc1 mutant. Taken together, these results suggest that FgExosc1 is not necessary for the correct localizations of FgExosc2, FgExosc3, FgExosc5, FgExosc8, FgExosc9, FgExosc10, FgExosc11, and FgExoscA in F. graminearum but is required for the localization of FgExosc4, FgExosc6, and FgExosc7.

DISCUSSION
The RNA exosome complex is a conserved, multisubunit RNase complex that contributes to the 39-to-59 processing and degradation of RNAs in mammalian cells (1). However, the biological functions of the RNA exosome complex have not been reported in plant-pathogenic fungi, and how this complex regulates the fungal development and pathogenicity remain unclear. We identified 12 components of RNA exosome complex in the wheat fungal pathogen F. graminearum. We first found that all 12 subunits are all localized in the nucleus and that they form a complex through interactions. FgEXOSC1 and FgEXOSCA were successfully knocked out, while we could not obtain mutants of the other 10 genes after several attempts. FgExosc1 and FgExoscA were both found to be required for the vegetative growth, conidiation, DON production, and pathogenicity of F. graminearum. Furthermore, FgEXOSC1 is involved in ncRNA processing, rRNA metabolism and processing, and ribosome and ribonucleoprotein complex biogenesis.
In eukaryotes, the RNA exosome complex consists of about 11 components (one may have multiple subunits) (1,2,5,56). So far, the components of the RNA exosome complex in fungi have not been well investigated. In this study, 12 candidate components of RNA exosome complex were identified; of these, FgExoscA is a specific component that is widely distributed in fungi, but its homologs are absent in plant and mammalian cells. These 12 candidate components have similar localizations. Furthermore, subcellular localization, GFP pulldown, and co-IP experiments all showed that FgExosc1 associates with the other RNA exosome components. Taking these observations together, we speculated that these 12 candidate components form the RNA exosome complex in F. graminearum.
In yeast, humans, and fruit flies, RNA exosome complex is a key factor in growth and development (23,57,58). Deletion of AtCSL4 (EXOSC1) does not cause significant phenotypic changes, but homozygous mutations of the remaining components are deadly, while heterozygous mutations lead to different phenotypic defects (10,26,27,59,60). Deletion of ScRRP6 (EXOSC10) in yeast leads to slow growth of the mutants, while the other components (ScCsl4, ScRrp4, ScRrp40, ScRrp41, ScRrp42, ScRrp43, ScRrp44, ScRrp45, ScRrp46, and ScMtr3) are all essential for the survival of the organism (30,32,61). Unlike in yeast, FgExosc10, the homolog protein of ScRrp6, is essential in F. graminearum. In contrast, FgExosc1, the homolog protein of yeast ScCsl4, is dispensable in F. graminearum. Collectively, these results indicate that different RNA exosome components have different functions in the growth and development of different organisms. Through real-time PCR, we found that deletions of FgEXOSC1 and FgEXOSCA resulted in decreased expression of some of the other components of RNA exosome. Furthermore, disruption of FgEXOSC1 affected the localizations of FgExosc4, FgExosc6, and FgExosc7. These results suggest that FgEXOSC1 and FgEXOSCA deletions may affect the stability of the RNA exosome complex.
There are different forms of RNA exosome in the nucleus and cytoplasm that perform RNA processing or degradation functions in yeast (4). However, our result showed that all the components of RNA exosome complex in F. graminearum are mainly localized in the nucleus, although FgExosc1, FgExosc2, FgExosc3, FgExosc4, FgExosc5, FgExosc6, FgExosc7, FgExosc8, and FgExosc9 also have weak fluorescence signals in the cytoplasm, while the other three subunits (FgExosc10, FgExosc11, and FgExoscA) could not be detected in the cytoplasm. In yeast, Dis3/Rrp44 (Exosc11) is localized in both the nucleus and cytoplasm, while Rrp6 (Exosc10) is present only in the nucleus; in humans, Rrp6 is similarly present in both the nucleus and cytoplasm (62), suggesting different localization pattern of the RNA exosome in F. graminearum.
The RNA exosome complex not only regulates the processing of rRNAs, tRNAs, snRNAs, and snoRNAs but also degrades mRNA and misprocessed RNA, playing an important role in RNA surveillance (63). The RNA exosome has been shown to be the primary regulatory element after transcription, which controls cell differentiation by regulating the transcriptome and proteome (64). In humans and yeast, the RNA exosome complex processes all nascent RNA in the nucleus, especially pre-rRNA, and mutations in any of the exosome complex subunits lead to accumulation of pre-rRNAs within the cell (28). Cytoplasmic RNA exosome complex participates in the quality control and degradation of mRNA, maintaining the stability of mRNA in vivo. Eukaryotic rRNAs are transcribed, folded, modified, and processed to form ribosomes through a complex series of assembly and maturation pathways under the mediation of ribosomal assembly factors (AFs) and snoRNAs, bound to approximately 80 ribosome proteins (65,66). In the present study, RNA-seq analysis of the transcriptome of the wild type and DFgexosc1 mutant under nutritional conditions revealed that the deletion of FgEXOSC1 caused the expression of 1,651 genes to be upregulated, while 1,788 genes were downregulated. These differential levels of gene expression are mainly linked to ncRNA processing, rRNA metabolism and processing, and ribosome and ribonucleoprotein complex biogenesis, indicating that the RNA exosome complex of F. graminearum plays an important role in ribosome biogenesis as well as processing of rRNA and ncRNA. In addition, we found that some secreted pathogenesis-related genes were downregulated (Fig. S6), including OSP24, OSP25, and OSP44 (67). Among those genes, OSP24 was identified as an important virulence factor of F. graminearum which modulates host immunity by mediating proteasomal degradation of TaSnRK1a (67), suggesting that the RNA exosome may be involved in regulating the functions of virulence factors at transcription level in F. graminearum.
In summary, our study identified 12 components of the RNA exosome complex in F. graminearum. All the RNA exosome components are localized in the nucleus. FgExosc1 and FgExoscA are both required for vegetative growth, sexual reproduction, DON production, and pathogenicity. FgExosc1 is involved in ncRNA processing, rRNA and ncRNA metabolism, ribosome biogenesis, and ribonucleoprotein complex biogenesis. FgExosc1 associates with the other components of RNA exosome complex to form the RNA exosome complex in F. graminearum. Our findings provide new insights into the roles of RNA exosome in regulating fungal development and pathogenicity.

MATERIALS AND METHODS
Fungal strains and culture conditions. Fusarium graminearum PH-1 was used as the wild type, from which all the mutants were generated (Table S3). All the strains were cultured on CM, starch yeast medium (SYM), and minimal medium (MM) at 28°C for 3 days (68). For cultivation in liquid medium, mycelia from 2-day-old colonies were transferred into liquid CM and incubated for 2 days with shaking (180 rpm) at 28°C.
Gene deletion, GFP/mCherry fusion vector constructions, and complementation. Protoplast preparation and fungal transformation of F. graminearum were performed by following standard protocols (69). FgEXOSC1 (FGSG_13120) and FgEXOSCA (FGSG_08866) genes were deleted using split-marker approach (70). The primers used to amplify the flanking sequences for each gene are listed in Table S4. Knockout candidates were verified by Southern blotting. For complementation, the resulting DFgexosc1 and DFgexoscA mutants were complemented with GFP-FgEXOSC1 and FgEXOSCA-GFP vectors, respectively. In brief, the GFP-FgEXOSC1 vector was generated by amplifying the open reading frame (ORF) and native promoter of FgEXOSC1 and GFP genes using the primer pair in Table S4. The amplicons were ligated and then cloned into a pKNT vector and the product was sequenced for verification. The same method was used to construct the mCherry-FgEXOSC1 vector. The FgEXOSCA-GFP vector was generated by amplifying the ORF and native promoter of FgEXOSCA using the primer pair in Table S4. The gene sequence was tagged with GFP at its C terminus and then cloned into a pKNTG2 vector and finally sequenced for verification. The same method was used to generate the GFP fusion constructs for other RNA exosome complex subunits.
Assays for asexual reproduction, sexual reproduction, and ascospore discharge. To assay asexual reproduction, PH-1 and the various mutants were inoculated into liquid carboxymethyl cellulose (CMC) medium to induce conidiation (9). The number of conidia produced by each strain was determined 3 days after incubation at 28°C using a hemacytometer (Qiujing, Shanghai, China) under an Olympus BX53F microscope (Olympus, Tokyo, Japan). To induce sexual reproduction, the wild type and the mutants were inoculated on carrot agar medium and incubated at 28°C for 5 to 7 days, and 1 mL of 2.5% sterile Tween 60 was pressed gently into each plate (Macklin, Shanghai, China). All of the sexual reproduction-induced cultures were incubated at 22°C under black light (F20T8/BLB [Danqi, Shanghai, China]; wavelength, 365 nm). The perithecia formed were photographed and recorded after 10 days. As for the ascospore discharge, the 7-day-old perithecia were extracted from the plates using a 0.5-cm hole punch, placed on hydrophobic slides, put in a black box to moisturize, and incubated at 22°C under black light for 3 days. Each experiment was independently repeated three times.
Plant infection and DON production assays. Infection assays on flowering wheat heads and wheat coleoptiles were conducted as previously described (68,71). Mycelial agar blocks (5 mm in diameter) were inoculated into liquid trichothecene biosynthesis induction (TBI) medium for 7 days in the dark to assay DON production. The DON produced was quantified using a vomitoxin detection kit (enzyme-linked immunosorbent assay) and a Berthold multifunctional microplate reader.
qRT-PCR. For quantitative real-time PCR (qRT-PCR) of DON biosynthesis genes, PH-1 and the DFgexosc1 mutant were inoculated in liquid TBI for 3 days. Total RNA was isolated from mycelia using an Eastep total RNA extraction kit, and first-strand cDNA was synthesized using Moloney murine leukemia virus (M-MLV) reverse transcriptase as previously reported (72). For relative expression, the data were analyzed using the 2 2DDCT (cycle threshold) method (73). The F. graminearum housekeeping genes for actin (FGSG_07335) and b-tubulin (FGSG_06611) were used as the endogenous reference genes. A similar method was used to analyze the relative expression levels of FgEXOSC2, FgEXOSC3, FgEXOSC4, FgEXOSC5, FgEXOSC6, FgEXOSC7, FgEXOSC8, FgEXOSC9, FgEXOSC10, and FgEXOSC11 in DFgexosc1 and DFgexoscA mutants. All experiments and qRT-PCR assays were repeated three times.
RNA-seq analysis. Mycelia from PH-1 and the DFgexosc1 mutant were cultured in CM for 2 and 4 days, respectively. Mycelia from the resulting cultures were then harvested and stored at 280°C for future use. RNA was extracted and detected using the RNA Nano 6000 assay kit of the Bioanalyzer 2100 system (Agilent Technologies, California, USA). The RNA library was sequenced on an Illumina NovaSeq platform and 150-bp paired-end reads were generated. Reference genome and gene model annotation files were downloaded directly from genome website. The mapped reads of each sample were assembled by StringTie (v1.3.3b) (74) in a reference-based approach. Differential expression analysis of two conditions/groups was performed using DESeq2 R package (1.20.0). Gene Ontology (GO) enrichment analysis of the differentially expressed genes was conducted using cluster Profiler R package, and the same package was used to test the statistical enrichment of the differentially expressed genes in KEGG pathways. These transcriptome analyses were completed by Novogene Bioinformatics Institute (Beijing, China).
Affinity capture-mass spectrometry analysis, immunoblotting, and co-IP assays. Briefly, GFP-FgExosc1 and PH-1-GFP strains were cultured in liquid CM for 2 days with shaking (180 rpm) at 28°C. Mycelia were then harvested and stored at 280°C. An affinity purification assay was then carried out as previously reported (75). For immunoblot analysis, total proteins were isolated from vegetative hyphae as described above. The extracted proteins were separated on 10% SDS-PAGE gels and then transferred onto a polyvinylidene difluoride (PVDF) membrane with a Bio-Rad electroblotting apparatus. The proteins were finally detected by anti-GFP (M20004L) and anti-mCherry (NBP1-96752) as previously reported (75).
Data availability. The RNA-seq data were deposited in the NCBI BioProject database under accession number PRJNA894853.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 1.1 MB.