The FgNot3 Subunit of the Ccr4-Not Complex Regulates Vegetative Growth, Sporulation, and Virulence in Fusarium graminearum

The Ccr4-Not complex is evolutionarily conserved and important for multiple cellular functions in eukaryotic cells. In this study, the biological roles of the FgNot3 subunit of this complex were investigated in the plant pathogenic fungus Fusarium graminearum. Deletion of FgNOT3 resulted in retarded vegetative growth, retarded spore germination, swollen hyphae, and hyper-branching. The ΔFgnot3 mutants also showed impaired sexual and asexual sporulation, decreased virulence, and reduced expression of genes related to conidiogenesis. Fgnot3 deletion mutants were sensitive to thermal stress, whereas NOT3 orthologs in other model eukaryotes are known to be required for cell wall integrity. We found that FgNot3 functions as a negative regulator of the production of secondary metabolites, including trichothecenes and zearalenone. Further functional characterization of other components of the Not module of the Ccr4-Not complex demonstrated that the module is conserved. Each subunit primarily functions within the context of a complex and might have distinct roles outside of the complex in F. graminearum. This is the first study to functionally characterize the Not module in filamentous fungi and provides novel insights into signal transduction pathways in fungal development.


Introduction
Fusarium graminearum is an ascomycetous fungus that causes Fusarium head blight in cereal crops worldwide, including wheat, barley, and rice, as well as ear and stalk rot in maize [1,2]. Fungal infection of F. graminearum leads to yield and quality losses as well as contamination of grains by the production of mycotoxins (trichothecenes and zearalenone) that threaten human and animal health [3]. F. graminearum produces both sexual (ascospores) and asexual (conidia) spores [4]. Ascospores are produced and discharged from the perithecia during

Nucleic acid manipulation, primers, and PCR conditions
The gDNA was extracted as previously described [4]. Restriction endonuclease digestion, agarose gel electrophoresis, gel blotting, and DNA blot hybridization were performed in accordance with standard techniques [36]. The polymerase chain reaction (PCR) primers (S1 Table) used in this study were synthesized by an oligonucleotide synthesis facility (Bionics, Seoul, Korea).

Genetic manipulations and fungal transformations
For complementation of the ΔFgnot3 deletion mutants, the wild-type FgNOT3 allele from F. graminearum strain Z-3639 was amplified using the Not3-5F com/Not3-3N com primer pair. The hygromycin resistance cassette (HYG) was amplified from the pBCATPH vector using the pBCATPH/comp 5 0 For/pBCATPH/comp 3 0 Rev primer pair [37]. The resulting amplicons were fused by double-joint (DJ) PCR as previously described [38]. The final PCR constructs were obtained by nested PCR and transformed into the ΔFgnot3 deletion mutants as described previously [39].
To generate FgNOT2 deletion mutants, the 5 0 -and 3 0 -flanking regions of the FgNOT2 gene and a geneticin resistance cassette (GEN) were amplified from Z-3639 and pII99, respectively, and were fused by DJ PCR. The subsequent procedures for the third round of PCR and transformation were the same as for complementation using the FgNOT3 gene of F. graminearum. The FgNOT4 deletion mutants were produced using the same strategy. The same strategy used for the generation of FgNot3c strains was also applied for the complementation of the ΔFgnot2 and ΔFgnot4.

Conidial production and morphology
After each strain was incubated in 50 ml of CM for 72 h at 25°C on a rotary shaker (150 rpm), mycelia of each strain were harvested and washed twice with distilled water. To induce conidiation, harvested mycelia were spread on YMA and incubated for 48 h at 25°C under near-UV light (wavelength: 365 nm, HKiv Import & Export Co., Ltd., Xiamen, China). Conidia were collected using distilled water, filtered through cheesecloth, washed, and resuspended in distilled water. After inoculating a 1 ml conidial suspension (1 × 10 6 conidia/ml) of each strain in 50 ml of CMC and incubating for 5 days at 25°C on a rotary shaker (150 rpm), the number of conidia produced was counted to measure conidial production with a hemocytometer (Superior, Marienfeld, Germany). For observation of conidial morphology, the conidia produced by each strain on YMA were harvested, and differential interference contrast (DIC) images were obtained using a DE/Axio Imager A1 microscope (Carl Zeiss, Oberkochen, Germany).

Germination assay
To evaluate germination rates, conidial suspensions (1 × 10 6 conidia/ml) of each strain were inoculated into 20 ml of CM and MM and incubated at 25°C on a rotary shaker (150 rpm). The germinated conidia per 100 total conidia were counted at 0, 4,6,8,10,12,24,36,48, and 60 h after inoculation. Conidial germination was defined as the point at which the length of the germ tube is the same as the width of the conidium. The experiments were performed twice with three replicates for each time point.

Outcrosses and virulence test
For self-fertilization, mycelia grown on carrot agar for 5 days were mock-fertilized with a 2.5% Tween 60 solution to induce sexual reproduction as previously described [4]. For outcrosses, mycelia of the female strain grown on carrot agar plates were fertilized with 1 ml of male strain conidia (1 × 10 6 conidia/ml). The heterothallic mat1g (Δmat1-1::GEN hH1::hH1-GFP-HYG) strain [35] was used as a tester strain for outcrosses. After sexual induction, the fertilized cultures were incubated for 7 days under near-UV light (HKiv Import & Export Co., Ltd.) at 25°C. A virulence test of the fungal strains was performed using the wheat cultivar Eunpamil as previously described [40]. Briefly, 10 μl of conidial suspensions (1 × 10 6 conidia/ml) obtained from each strain was point-inoculated into a spikelet of the wheat head at early anthesis. Infected plants were incubated in a humidity chamber for 3 days and subsequently transferred to a green house. After 21 days, the number of spikelets showing disease symptoms was counted.

Quantification of mycotoxins and fungal ergosterol
For trichothecene analysis, the 3-week-old rice cultures were ground and extracted with an ethyl acetate/methanol mixture (4:1, v/v) as previously described [34]. The extracts were purified using MycoSep 1 225 Trich Multifunctional columns (Romer Labs, Inc., Union, MO, USA) and then concentrated to dryness. A portion of each extract was derivatized with Sylon BZT (BSA + TMCS + TMSI, 3:2:3, Supelco, Bellefonte, PA, USA) and analyzed using a Shimadzu QP-5000 gas chromatograph-mass spectrometer (GC-MS; Shimadzu, Kyoto, Japan). ZEA was extracted from rice cultures using the same strategy and analyzed using a HPLC system with a RF-10A XL fluorescence detector (Shimadzu) [34]. To quantify fungal ergosterol, ground rice cultures (1 g) were extracted in 4 ml of chloroform/methanol (2 :1, v/v) as previously described [41]. Ergosterol was analyzed using a HPLC system with a 4.6 U ODS column (250×4.6 mm, Phenomenex, Madrid Avenue Torrance, CA, USA) and an UV detector (Shimadzu) set to measure absorbance at 282 nm. Quantities were determined by comparing peak areas of samples to those of a standard curve generated from HPLC-grade ergosterol (Sigma-Aldrich, St. Louis, Missouri, USA). The experiments were repeated three times.

Quantitative real-time (qRT)-PCR
The total RNA of the wild-type and ΔFgnot3 strains was extracted from mycelia at 18 h after inoculation in CMC using an Easy-Spin Total RNA Extraction kit (Intron Biotech, Seongnam, Korea) [42]. First-strand cDNA was synthesized with the SuperScript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA) using oligo(dT) 20 according to the manufacturer's recommendations. qRT-PCR was performed using iQ SYBR Green Master Mix (Bio-Rad, Hercules, CA, USA) and a 7500 real-time PCR system (Applied Biosystems, Foster City, CA, USA). The endogenous housekeeping gene, ubiquitin C-terminal hydrolase (UBH; FGSG_01231), was used for normalization [22]. The PCR assays were repeated three times with two biological replicates. The threshold cycle (ΔC T ) value of gene expression was subtracted from the ΔC T value of each sample to obtain a ΔΔC T value. The transcript level relative to the calibrator was expressed as 2 ˗ ΔΔCT [43].

Yeast strains and complementation assay
The S. cerevisiae strains, BY4741 (wild-type) and YPR072w (ΔScnot5), were obtained from EUROSCARF (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/) and maintained on yeast extract peptone dextrose (YPD) medium. The FgNOT3 ORF was amplified from first-strand cDNA of the F. graminearum wild-type strain, Z-3639, with the Not3-cloning-F/Not3-cloning-R primer pair (S1 Table) by PCR, digested by BstXI and XbaI restriction enzymes, and subsequently cloned into the BstXI and XbaI sites of pYES2 (Invitrogen). The pYES2-FgNOT3 construct was introduced into the ΔScnot5 strain using the lithium acetate method [44] after verification of the construct by sequencing (Macrogen Inc., Seoul, Korea). In addition, the pYES2 empty vector was simultaneously introduced into yeast wild-type BY4741 and ΔScnot5. Synthetic complete medium lacking uracil (SC-Ura) and supplemented with ampicillin (0.2 mg/ml) was used for the selection and isolation of transformants [44]. For the complementation assay, yeast cells were cultured for 3 days at 30°C on a rotary shaker (200 rpm) in SC-Ura supplemented with ampicillin, and the cells were then harvested and diluted in distilled water. Aliquots of 10 μl were point-inoculated on SC-Ura supplemented with ampicillin followed by incubation for 4 days at 30°C.

Molecular characterization of the FgNOT3 gene
The Ccr4-Not complex of S. cerevisiae consists of two major modules, the catalytic module (ScCaf1 and ScCcr4) and the Not module (ScNot1, ScNot2, ScNot3, ScNot4, and ScNot5), and an additional two subunits, namely ScCaf40 and ScCaf130 [17,18]. The occurrence of the subunit genes in the Ccr4-Not complex in representative species based on the STRING database [45] showed that most of these genes are highly conserved in eukaryotes ( Fig 1A). However, Caf130 homologs are specifically conserved in members of the Saccharomycetaceae, such as S. cerevisiae and Candida glabrata.
Except for fungal species of the Saccharomycetaceae, only single gene-encoding proteins similar to ScNot3 or ScNot5 have been identified in other eukaryotic genomes ( Fig 1A). Although Not3/5 of oomycetes, Phytophthora infestans and Pythium ultimum showed higher sequence identity with ScNot5 than ScNot3, the rest of the single proteins were homologs for ScNot3. Previous reports have also shown that Not5 is not conserved in animals and is specific for the Saccharomycetaceae [46]. Phylogenetic analyses of Not3 homologs showed that Not3 homologs in filamentous fungi were clustered into a separate group relative to yeasts and animals ( Fig 1B).
BLASTp searches for both ScNot3 and ScNot5 in the F. graminearum genome (http://www. broadinstitute.org) identified the FGSG_13746 locus encoding 582 amino acids (30% and 27% overall identity to ScNot3 and ScNot5, respectively). The protein harbored two significant domains (IPR007207 and IPR007282) similar to those of both ScNot3 and ScNot5 (Fig 1C). Further analysis of the conserved protein sequences of partial N-terminal regions showed that FGSG_13746 shares 55% and 50% identity with those of ScNot3 and ScNot5, respectively. The human C-terminal region of Not3 also shows 24% overall identity to both ScNot3 and ScNot5 although their N-terminal regions share 41% and 39% identity with those of ScNot3 and ScNot5, respectively [47]. Based on these combined results, we designated the protein encoded by FGSG_13746 as FgNot3.

Effects of FgNOT3 deletion on vegetative growth, conidiogenesis, and germination
The FgNOT3 deletion mutants were obtained from a mutant library of F. graminearum TF deletions [9]. For genetic complementation, the construct containing the FgNOT3 open reading frame (ORF) fused with HYG was introduced into the protoplast of the ΔFgnot3 strain (S1 Fig). Southern blot analysis showed that the construct successfully replaced GEN in the genome of the complementation strain, resulting in FgNot3c strains. The ΔFgnot3 strains showed markedly reduced radial growth (~50%) and aerial mycelia on both complete medium (CM) and minimal medium (MM) (Fig 2A).
To further determine the features affecting the defective growth of the ΔFgnot3 mutants, we performed microscopic observation. Deletion of FgNOT3 resulted in a hyper-branching phenotype compared to the wild-type strain ( Fig 2B). Furthermore, the hyphae of ΔFgnot3 mutants tended to be abnormally swollen, and the swollen hyphae resulted in distorted branching ( Fig 2C). These results demonstrated that FgNOT3 is required for normal growth and mycelial morphology in F. graminearum.
Deletion of FgNOT3 also resulted in severe defects in asexual sporulation. The conidial production of the ΔFgnot3 strain in CMC medium was significantly reduced compared to the wild-type and complemented strains ( Fig 3A). Moreover, conidia of ΔFgnot3 strains were abnormally shaped (Fig 3B). The conidia of the ΔFgnot3 strains were shorter and wider than the wild-type ( Table 2 and Fig 3B). Deletion of FgNOT3 also resulted in a reduced septum number.
To determine how deletion of FgNOT3 affects conidiogenesis in F. graminearum, we generated ΔFgnot3-g strains (ΔFgnot3::GEN hH1-GFP-HYG) by an outcross between the mat1g [35] and ΔFgnot3 strains. Dozens of ascospores were isolated, and their genotypes were confirmed by fluorescence microscopy and PCR. The hH1-GFP strain carrying the wild-type allele of FgNOT3 initially produced phialides from the hyphae, and mature phialide cells continuously produced conidia ( Fig 3C). Additionally, conidia were often directly produced from the hyphae. In contrast, deletion of the FgNOT3 mostly abolished phialide production, and most conidia were directly produced from the hyphae (Fig 3C). All of these defects were restored to wild-type levels in the FgNot3c complemented strains.
To test the hypothesis that FgNot3 plays a role in regulating the expression of genes related to conidiogenesis, we compared the transcript levels of representative conidiation-related genes in the wild-type and ΔFgnot3 deletion mutant strains [6][7][8][9]. Transcript levels of four genes, namely STUA, HTF1, REN1, and FLBC, were significantly decreased in the ΔFgnot3 mutants compared to wild-type ( Fig 3D). Interestingly, transcript levels of ABAA and WETA, the transcription factors specifically involved in conidiogenesis in F. graminearum [10,11], were not altered when FgNOT3 was deleted (data not shown).
The conidia germination rates of ΔFgnot3 mutants were greatly reduced in both CM and MM compared to wild-type ( Fig 3E). Approximately 90% of conidia in the wild-type strain germinated 6 h after inoculation in CM, whereas only approximately 10% of conidia germinated in the ΔFgnot3 mutants ( Fig 3E). Furthermore, only approximately 16% of ΔFgnot3 mutant conidia germinated in MM 24 h after inoculation, whereas most wild-type conidia were germinated after 24 h. Although the germinated hyphae of all strains showed an identical morphology  Fig and Fig 2C). All defects of the ΔFgnot3 mutants were restored in the FgNot3c complemented strains.

FgNOT3 is important for sexual development and virulence
The fertility of the F. graminearum strains was determined on carrot agar. In self-fertility, the wild-type strains began to produce detectable perithecial initials 3 days after sexual induction, and mature perithecia were produced after an additional 3 or 4 days of incubation (Fig 4A). In contrast to wild-type, the ΔFgnot3 mutants only produced a few perithecium initials that did not mature.
Subsequently, we determined the female and male fertilities of the ΔFgnot3 mutants based on outcross analyses. When the ΔFgnot3 mutant was spermatized with wild-type or Δmat1 strains, no mature perithecium was produced, similar to the self-cross of the ΔFgnot3 mutants ( Fig 4A). However, when the ΔFgnot3 mutant was used as a male in the outcross of Δmat1 (female) × ΔFgnot3 (male), normal perithecia were produced, and the progeny with or without the hH1-Gfp signal were segregated 1:1 in accordance with Mendelian genetics, suggesting that FgNOT3 is not necessary for male fertility (Fig 4B). The sexual defects of the ΔFgnot3 mutants were recovered in the FgNot3c strains.
To evaluate the pathogenicity of the ΔFgnot3 mutants in flowering wheat heads, conidial suspensions of each strain were point-inoculated on spikelets. The results showed that wildtype and FgNot3c strains caused typical head blight symptoms 21 days after inoculation, whereas the ΔFgnot3 strains were unable to spread from the inoculated spikelet to adjacent spikelets on the heads (Table 2 and Fig 4C).

FgNOT3 is required for normal growth under high-temperature conditions
To characterize the roles of FgNOT3 in environmental stress responses, we examined the sensitivity of the ΔFgnot3 mutants to various stresses, including carbon and nitrogen starvation, osmotic and oxidative stresses, cell wall-damaging agents, fungicide exposure, and thermal stresses. There were no specific stresses or agents that affected the growth of ΔFgnot3 mutants as previously described (data not shown). However, we identified the role of FgNOT3 in adaptation to thermal stress in F. graminearum. The ΔFgnot3 mutants exhibited increased sensitivity to high temperature and could not grow at 31°C (Fig 5). FgNot3 functions together with other Not subunits of the Ccr4-Not complex The effects of FgNOT3 deletion on the transcript levels of nine putative Ccr4-Not complex subunit genes were analyzed during conidiation (Fig 6). Five genes (homologs for CCR4, CAF1, CAF40, CAF130, and DDH1) showed similar transcript levels between the wild-type and ΔFgnot3 strains. However, the transcript levels of FgNOT1 and FgNOT4 in the ΔFgnot3 mutants were significantly decreased compared with the wild type (Fig 6). The FgNOT2 transcript levels of the ΔFgnot3 mutants were greatly increased, showing more than 6-fold higher expression than wild type (Fig 6). Because FgNOT1 is an essential gene [9] and ΔFgnot3 showed pleiotropic defects, we sought to determine how other Not subunits affect the developmental stages of F. graminearum. To characterize their biological functions, we generated deletion and complementation mutants of FgNOT2 (FGSG_12685) and FgNOT4 (FGSG_09233) via homologous recombination (S4 Fig). We found that the radial growth, conidial production, conidial morphology, sexual development, and virulence of both ΔFgnot2 and ΔFgnot4 mutants were severely impaired compared to wild type, similar to the observed phenotypes of the ΔFgnot3 strains (Figs 7 and 8 and Table 3). In particular, deletion of FgNOT2 and FgNOT4 resulted in markedly reduced radial growth compared to the wild-type and complemented strains (Fig 7). Furthermore, ΔFgnot2 and ΔFgnot4 mutants also exhibited increased sensitivity to high temperature (Fig 5).
Conidial production of both the ΔFgnot2 and ΔFgnot4 mutants was similarly reduced compared to wild type ( Table 3). Deletion of FgNOT2 resulted in more severe defects in conidial Roles of FgNot3 in Fusarium graminearum morphologies than FgNOT4 deletion (Table 3 and Fig 8A). Whereas the ΔFgnot3 mutants showed an almost complete lack of phialide formation, the ΔFgnot2 and ΔFgnot4 mutants mainly produced normal phialides and conidia (Fig 8B). Approximately 30% of the phialides of both ΔFgnot2 and ΔFgnot4 mutants were produced as cluster forms with abnormal shapes. × 100, where C is colony diameter of the control (at 25°C), and N is that of treatments (at 29°C and 31°C) as previously described [48]. The values were generated based on three biological replicates. The significance of differences between the wild-type and each strain was calculated using Student's t-test. Significant differences (P < 0.05) are indicated with an asterisk for the 29°C condition or a double asterisk for the 31°C condition. WT, F. graminearum wildtype strain Z-3639; ΔFgnot2, FgNOT2 deletion mutant; ΔFgnot3, FgNOT3 deletion mutant; ΔFgnot4, FgNOT4 deletion mutant; FgNot2c, ΔFgnot2-derived strain complemented with FgNOT2; FgNot3c, ΔFgnot3-derived strain complemented with FgNOT3; FgNot4c, ΔFgnot4-derived strain complemented with FgNOT4. doi:10.1371/journal.pone.0147481.g005

Roles of FgNot3 in Fusarium graminearum
We found that deletion of FgNOT2 and FgNOT4 also resulted in a loss of self and female fertilities (Fig 8C and 8D) and a significant decrease in virulence on wheat heads (Table 3 and Fig  8E). All of these defects of the ΔFgnot2 and ΔFgnot4 mutants were restored to wide type levels in the corresponding complemented strains.   Whereas ΔFgnot4 mutants only produced significantly higher levels of ZEA than the wild-type strain, deletion of both FgNOT2 and FgNOT3 resulted in much higher production of both trichothecenes and ZEA in rice cultures (Fig 9). These observed defects in the ΔFgnot2, ΔFgnot3, and ΔFgnot4 mutants were restored in the corresponding complemented strains.

Discussion
In this study, FgNOT3 was found to be involved in numerous developmental stages in F. graminearum, including vegetative growth, asexual reproduction, sexual reproduction, secondary metabolite production, and virulence. Moreover, we provided genetic evidence that other Not subunits also have conserved roles in this fungus. Taken together, these results demonstrated Table 3. Vegetative growth, conidial production, conidial morphology, and virulence of ΔFgnot2 and ΔFgnot4 mutants.

Strain
Radial growth (mm) a Conidial production(10 6   Mycotoxin production of ΔFgnot2, ΔFgnot3, and ΔFgnot4 mutants. Total trichothecene production or zearalenone (ZEA) production was normalized to fungal ergosterol levels with the following equation as previously described [41]: [trichothecene or ZEA production (μg/g) / ergosterol contents (μg/g)] × 100. The values were generated based on three biological replicates. The significance of differences between the wild-type and each strain was calculated using Student's t-test. that the Not module of the Ccr4-Not complex plays critical roles in the regulation of multiple and complex cellular processes and differentiation in F. graminearum. FgNot3 contributes to hyphal morphogenesis and virulence in F. graminearum. Deletion of FgNOT3 led to a significant reduction in mycelial production and abnormal shaped hyphae in F. graminearum, phenotypes similar to other model eukaryotes. Yeast ScNOT5 is involved in diverse processes, including cell wall integrity and filamentation [14,24]. Deletion of CaNOT5 also results in defective morphogenesis and virulence in Candida albicans [26]. Although the ΔFgnot3 mutants produced a considerable amount of trichothecenes, which are virulence factors [49], virulence of the mutants was completely abolished, and fungal growth was restricted to infection sites. We believe that the ΔFgnot3 mutant hyphae could not differentiate into structures required for host infection due to the attenuated morphogenesis and cellular differentiation. It has been reported that F. graminearum hyphae develop mats and appressoria-like structures to penetrate the host cell wall [50,51].
FgNot3 also regulates hyphal differentiation required for both sexual and asexual reproduction in F. graminearum. Both ascospores and conidia are produced from highly differentiated hyphal structures such as perithecia and phialides, respectively. ΔFgnot3 mutants completely lost female fertility and failed to produce normal phialide cells, leading to markedly impaired spore production and abnormal conidium morphologies. The significant down-regulation of genes involved in conidiation, including STUA, HTF1, REN1, and FLBC, but not ABAA and WETA, supports the hypothesis that FgNOT3 has global roles in asexual sporulation independent of the AbaA-WetA pathway. To our knowledge, this is the first report to implicate the contribution of Not3 homologs in both sexual and asexual developmental stages.
NOT3/5 mutations in yeasts affect cell wall structure. Therefore, the null mutants showed altered sensitivities to cell wall stress-inducing agents. Although ΔCanot5 mutants of C. albicans are highly sensitive to Calcofluor white, they are more resistant to β-glucanase zymolyase than wild type [26]. ΔScnot3 mutants show increased sensitivity to caffeine in S. cerevisiae [52]. However, FgNOT3 deletion mutants did not show any altered sensitivity to cell wall stressinducing agents as well as other specific stresses as previously reported (data not shown) [9]. Intriguingly, we identified a novel function of FgNOT3 in adaptation to thermal stress. Moreover, FgNOT3 did not complement the S. cerevisiae ΔScnot5 mutant (S3 Fig). Differences in the function of Not3/5 in cell wall integrity between yeast and filamentous fungi may not only be derived from evolutionary divergence between the two groups but also might explain the diverse roles of Not3/5 in cellular processes among eukaryotes. CNot3 in mice is essential for multiple functions, including embryonic development and control of heart function, and the metabolism of adult mice supports this notion [28,53].
Our transcript analyses provide some clues for understanding the interactions among the subunits of Ccr4-Not complex. Transcript levels of all Not subunits, including FgNOT1, FgNOT2 and FgNOT4, were significantly altered, but those of other Ccr4-Not subunits were not, suggesting that they function within the context of the Not module. In addition, the transcript levels of FgNOT1 and FgNOT4 were decreased, whereas the FgNOT2 transcript level was highly increased in the ΔFgnot3 compared to wild type for unknown reasons. Because the feedback regulation often occurs in protein complexes [54], FgNot2 might be a direct interactor of FgNot3 but not of FgNot1 and FgNot4. The heterodimerization of the Not module in the ScNot1-ScNot2-ScNot5 or CNot1-CNot2-CNot3 forms a platform for macromolecular interactions [55,56]. While ScNot3/5 and ScNot2 seem to function together [16] and ScNot3 directly interacts with ScNot4 and ScNot5 in yeast [19,24], there is no interaction between CNot3 and CNot4 in humans [47]. The interaction between these proteins in F. graminearum needs to be confirmed, but these differences might be attributed to a different composition of the complex or the evolutionary divergence in yeast, filamentous fungi, and human.
We further functionally characterized the Not module of the Ccr4-Not complex in F. graminearum. FgNOT1 is an essential gene as reported in other eukaryotes [9]. ΔFgnot2 and ΔFgnot4 mutants had pleiotropic effects on phenotypes, including vegetative growth, sexual and asexual production, and virulence, similar to the impacts of ΔFgnot3, suggesting that the Not module composed of FgNot1-4 is also conserved in F. graminearum. In yeast, the association of all Ccr4-Not subunits is essential for cell viability although it remains unclear whether each subunit functions only within the context of a complex or has distinct roles outside of the complex [16]. For example, ScNot4 mainly functions in the regulation of proteasome integrity, whereas the ScNot2-3/5 module has more fundamental roles. Therefore, ΔScNot2 and ΔScNot5 showed more pronounced growth defects than ΔScNot4 [16]. Consistently, FgNOT4 deletion had a slight effect on phenotypes, but deletion of FgNOT2 resulted in mostly indistinguishable phenotypic defects compared to ΔFgnot3 mutants.
FgNOT2, FgNOT3, and FgNOT4 are negative regulators of ZEA and/or trichothecene production. Whereas ΔFgnot2 and ΔFgnot4 mutants produced significantly higher levels of ZEA and/or trichothecenes, ΔFgnot3 accumulated more than 40-fold higher levels of both ZEA and trichothecenes compared with wild type. The Not module of the Ccr4-Not complex appears to regulate upstream genes or transcriptional regulatory elements participating in the diverse regulation of multiple secondary metabolite biosynthetic clusters. Moreover, highly accumulated mycotoxins might affect the physiologies of the ΔFgnot2, ΔFgnot3, and ΔFgnot4 mutants. Although direct biological functions of mycotoxins have not been reported in F. graminearum, overproduction of secondary metabolites often leads to unexpected developmental defects. All of the 13 transcription factor mutants overproducing ZEA and/or trichothecenes show defective vegetative growth and/or reproduction [9], and FgFlbA deletion mutants accumulating both ZEA or trichothecenes at high levels also show pleiotropic defects in F. graminearum [57].
In summary, our study functionally characterized the Not3 subunit of the Ccr4-Not complex for the first time in filamentous fungi. FgNOT3 is involved in hyphal morphogenesis and cellular differentiation, which are related to sexual and asexual sporulation as well as virulence in F. graminearum. In addition, we found that the Not module of the Ccr4-Not complex of F. graminearum is conserved and involved in numerous characteristics, including vegetative growth, reproduction, virulence, and secondary metabolism.