|Home | About | Journals | Submit | Contact Us | Français|
The accumulation of viral RNA depends on many host cellular factors. The hexagonal peroxisome (Hex1) protein is a fungal protein that is highly expressed when the DK21 strain of Fusarium graminearum virus 1 (FgV1) infects its host, and Hex1 affects the accumulation of FgV1 RNA. The Hex1 protein is the major constituent of the Woronin body (WB), which is a peroxisome-derived electron-dense core organelle that seals the septal pore in response to hyphal wounding. To clarify the role of Hex1 and the WB in the relationship between FgV1 and Fusarium graminearum, we generated targeted gene deletion and overexpression mutants. Although neither HEX1 gene deletion nor overexpression substantially affected vegetative growth, both changes reduced the production of asexual spores and reduced virulence on wheat spikelets in the absence of FgV1 infection. However, the vegetative growth of deletion and overexpression mutants was increased and decreased, respectively, upon FgV1 infection compared to that of an FgV1-infected wild-type isolate. Viral RNA accumulation was significantly decreased in deletion mutants but was significantly increased in overexpression mutants compared to the viral RNA accumulation in the virus-infected wild-type control. Overall, these data indicate that the HEX1 gene plays a direct role in the asexual reproduction and virulence of F. graminearum and facilitates viral RNA accumulation in the FgV1-infected host fungus.
The interactions between viral elements and host factors are important for maintaining the infection cycles of RNA viruses in host cells. Viruses utilize numerous host factors that play essential roles in virus infection. Therefore, understanding the role(s) of host factors can provide insight into the molecular mechanism(s) of host and virus interactions. Relative to host factors affecting RNA viruses of animals and plants, host factors affecting fungal viruses (mycoviruses) are poorly understood.
Several host and viral components required for virus life in cells have been identified and characterized in the model organism Saccharomyces cerevisiae. The knowledge obtained by studying S. cerevisiae as a host for several double-stranded RNA (dsRNA) and single-stranded RNA (ssRNA) viruses has greatly extended our understanding of mycovirus-host interaction (1). With respect to filamentous fungi, the host factors required for mycovirus replication and symptom induction have been well described for the interaction between the prototypic hypovirus Cryphonectria hypovirus 1 strain EP713 (CHV1) and its host, the chestnut blight fungus (Cryphonectria parasitica). One of these host factors, NAM-1, modulates symptom induction in the fungus in response to CHV1 infection (2). The hypovirus-responsive host transcription factor gene pro1 is required for female fertility of C. parasitica, development of its asexual spores, and the maintenance of CHV1 infection (3). The host gene Cpbir1 (bir1 of C. parasitica), which encodes the IAP (inhibitor of apoptosis protein) CpBir1 and which is required for fungal conidiation, virulence, and antiapoptosis, is considerably downregulated as a consequence of hypovirus infection and affects hypovirus transmission in C. parasitica (4). Although much has been learned about the functional roles of host factors in the interaction between CHV1 and C. parasitica, how host components affect replication and movement in other mycovirus-fungus systems is largely unknown.
The current paper concerns the interaction of a mycovirus with the plant-pathogenic fungus Fusarium graminearum Schwabe [teleomorph: Gibberella zeae (Schwein.) Petch]. F. graminearum is the causal agent of cereal head blight on crops such as maize, barley, and wheat (5–7). A number of mycoviruses have been reported to infect F. graminearum (8–17), and some of these result in fungal hypovirulence (8, 13, 15, 16). We previously reported that the dsRNA mycovirus Fusarium graminearum virus 1 strain DK21 (currently named FgV1), reduces the virulence of F. graminearum and also delays mycelial growth, increases pigmentation, and reduces the production of mycotoxin (8). To identify host factor(s) involved in FgV1-F. graminearum interactions, we screened putative genes or gene products based on transcriptional and proteomic analysis (18, 19). Research on this mycovirus-fungus interaction will be facilitated by a well-developed DNA-mediated transformation system for F. graminearum that enables efficient genetic alterations, such as targeted gene deletion and overexpression of endogenous genes by promoter switching (5). Research on this mycovirus-fungus interaction will also be facilitated by the genomewide functional analysis of transcription factors (20).
We recently reported that at least 22 proteins of F. graminearum are differentially expressed in response to FgV1 infection (19). One of the highly expressed proteins is the peroxisome-derived hexagonal protein (Hex1) (locus tag FGSG_08737). The Hex1 protein self-assembles to form the Woronin body (WB). The WB is a peroxisome-derived dense-core microbody that is specific to filamentous ascomycetes (21–28, 46). WBs maintain cellular integrity by sealing the septal pore after cellular damage in Neurospora crassa (23–25). WBs are also involved in pathogenesis and survival under nitrogen starvation conditions in Magnaporthe oryzae; deletion of the HEX1 gene causes appressorial defects and further reduces M. oryzae pathogenicity on barley and rice leaves (28).
Here, we report that Hex1 functions in the maintenance of cellular integrity, the production of asexual spores, and the pathogenicity of F. graminearum. We also provide evidence that the HEX1 gene acts as a host factor that enhances the accumulation of FgV1 RNA in F. graminearum. This is the first report concerning the effect of a host factor on the accumulation of a dsRNA mycovirus in an infected filamentous fungus.
All strains used in this study (Table 1) were stored in 25% (vol/vol) glycerol at −80°C and were reactivated on potato dextrose agar (PDA; Difco). For nucleic acid manipulation, all strains of F. graminearum were grown in 50 ml of a liquid complete medium (CM) (29) at 25°C with shaking (150 rpm) for 5 days, while strains of C. parasitica were grown in 50 ml of Endothia parasitica complete medium (29) at 26°C with shaking (120 rpm) for 5 days. Mycelia were harvested by filtration through miracloth (Calbiochem) and ground to a fine powder with liquid nitrogen in a mortar and pestle.
Nucleotide sequences from the NCBI database were assembled using the SeqMan program in DNASTAR. Sequence similarity searches of HEX1 and HEX1 homologs were conducted with the NCBI BLAST program. The alignment of Hex1 and Hex1 ortholog amino acid sequences was performed by using the MegAlign program in DNASTAR, using a default setting and GeneDoc programs (http://www.nrbsc.org/gfx/genedoc/). Phylogenetic analysis of amino acid sequences was inferred using the neighbor-joining method as previously described (11, 12). Evolutionary analysis was conducted in MEGA5 (http://www.megasoftware.net).
For extraction of genomic DNA, we used a previously described procedure (29). To construct a PCR fragment for deletion, overexpression, and complementation, a slightly modified double-joint (DJ) PCR strategy was applied for fusion of PCR products (20). The PCR construct for overexpression of the target gene was generated by the same procedure that was used for the deletion mutants. The gen sequence and elongation factor 1α (EF1α) promoter were amplified from the pSKGEN plasmid (5). For the complementation of deletion mutants, the hygromycin resistance gene cassette (hph) was amplified from the pBCATPH plasmid. General PCR was performed following the manufacturer's instructions (TaKaRa). The PCR primers used in this study (available upon request) were produced at an oligonucleotide synthesis facility (Bioneer).
Protoplasts of the WT-VF strain (hereinafter, VF indicates virus free and VI indicates virus infected) were prepared by treating fresh mycelia grown in YPG liquid culture medium (0.3% yeast extract, 1% peptone, 2% glucose) for 3 h at 30°C with 1 M NH4Cl containing 10 mg/ml of Driselase (InterSpex Products) as previously described (29). Amplified PCR products were directly added, with 1 ml of polyethylene glycol (PEG) solution (60% PEG 3350, 10 mM Tris-HCl, pH 7.5, 10 mM CaCl2), to protoplast suspensions as described above. Transformants with resistance to Geneticin or hygromycin B were selected on regeneration medium containing 150 μg/ml of Geneticin (Duchefa) or hygromycin B (Calbiochem). Selected transformants were infected by FgV1 using hyphal-fusion-mediated virus transmission, and viral infection was confirmed by RT-PCR (described below). For the Southern blot hybridization of all mutants, the extracted genomic DNAs were digested with PstI for 16 h at 37°C. A 10-μg quantity of digested DNA was separated on an 0.8% agarose gel for 8 h. The gels were capillary blotted onto positively charged nylon membranes (GE Healthcare) in 0.4 N NaOH for 12 h. The 32P-labeled DNA probes were generated following standard techniques as previously described (29). The hybridization reaction was performed at 65°C for 16 h. After hybridization, unhybridized DNA probe was removed by washing with low-stringency (2× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate] and 0.1% SDS) and high-stringency (0.1× SSC and 0.1% SDS) buffers. The blotting image was visualized using a Fuji BAS-2500 phosphorimager and corresponding imaging software (Fuji).
Radial growth was measured on PDA, CM, and minimal medium (MM; 0.05% KCl, 0.2% NaNO3, 3% sucrose, 1% KH2PO4, 0.05% MgSO4 · 7H2O, 0.02% trace element, 2% agar) at 5 days postinoculation (d.p.i.) as previously described (20).
To test the effects of HEX1 deletion and overexpression on the virulence of F. graminearum, wheat head florets at early-mid-anthesis were inoculated with conidial suspensions as described previously (29) on wheat cv. Jokyoung. Approximately 6-week-old wheat plants with flowering heads were used. For production of conidial inoculum, 10 mycelial plugs were incubated in CMC liquid medium (1.5% carboxymethyl cellulose, 0.1% yeast extract, 0.05% MgSO4 · 7H2O, 0.1% NH4NO3, 0.1% KH2PO4) at 25°C with shaking (150 rpm) for 5 to 7 days. Conidia were collected by filtering through sterile miracloth. A 10-μl volume of the spore suspension (105 conidia/ml) in 0.01% (vol/vol) Tween 20 was injected into one floret of each flowering wheat head. Wheat plants inoculated with 0.01% (vol/vol) Tween 20 alone served as controls. For each treatment, 10 replicate wheat heads were inoculated. Because the overexpression mutant produced no conidia when infected with FgV1 (see Results), the wheat heads were inoculated with 10 μl of the CMC culture medium after that mutant had grown in the medium for 7 days. Inoculated plants were placed in a growth chamber (25°C, 80% relative humidity, 14-/10-h light/dark cycle). The percentage of wheat heads with head blight symptoms was determined at 14 d.p.i. The test was repeated three times. Statistical analysis was performed using PASW statistics software (SPSS, Inc.).
Samples for transmission electron microscopy (TEM) were prepared as described previously (10). All strains were grown on PDA plates at 25°C. Thin sections of mycelial agar plugs were used for preparing the samples. The ultrastructure of mycelia was observed with a transmission electron microscope (JEM 1010; JEOL).
For the measurement of the intercalary lengths of hyphae (i.e., cell lengths of hyphae), sections of mycelia on the agar plates were microscopically examined as previously described (30). Differential interference contrast (DIC) images of conidia and hyphae were captured on a DE/Axio imager A1 microscope (Carl Zeiss) with a charge-coupled device (CCD) camera. The lengths of hyphal cells were measured using the AxioVision release 4.8 software program (Carl Zeiss).
For selecting mycelia at the same growth stage, we attempted to determine those parts of the colony that had undergone cytokinesis (the creation of septa) for a similar number of times. We did this by first determining the average intercalary length (average hyphal cell size) of the strain. We then identified the location of three growth stages by multiplying the intercalary length by 0, 1,000, and 2,000. In other words, the initial inoculation point on the agar was established as “cytokinesis starting point 0” and the locations where cytokinesis had occurred 1,000 and 2,000 times were determined by multiplying the average intercalary length of the strain by 1,000 or 2,000. For example, when the intercalary length of the WT-VI strain was 20.9 μm, then cytokinesis points 1,000 and 2,000 were determined to be located 2.09 cm and 4.18 cm, respectively, from cytokinesis starting point 0.
For fungal protein extraction, the powdered mycelia were suspended with lysis buffer (21). This lysate was filtered twice through miracloth, and the filtrate was centrifuged at 100 × g for 5 min to remove unlysed cells. The lysate was then centrifuged at 10,000 × g for 5 min, and the pellet was suspended in lysis buffer. A 200-μg quantity of total protein was resolved on a Mini-Protean TGX precast gel (Bio-Rad) and transferred to a Hybond-P membrane (GE Healthcare) in transfer buffer (25 mM Tris, 380 mM glycine, 20% methanol) at 100 V for 40 min for protein blot analysis. The membrane was blocked for 16 h in Tris-buffered saline–Tween 20 (TBST) plus 2.5% skim milk (10 mM Tris, pH 7.4, 100 mM NaCl, 0.05% [vol/vol] Tween 20). The membrane was then probed with anti-Hex1 antibody (1:200) in TBST plus 1% skim milk for 1 h at room temperature. The membrane was washed five times with TBST over the course of 1 h and probed with horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody at 1:2,000 dilution for 1 h. The membrane was then washed five times in TBST over the course of 1 h and evaluated for chemiluminescence using the Amersham ECL Western blotting detection reagents and analysis system (GE Healthcare) according to the manufacturer's protocol.
For total RNA preparation, the powdered mycelia were suspended in Isol-RNA lysis reagent (5 PRIME). Nucleic acid was extracted by following the manufacturer's protocol, with slight modifications. The extracted total RNAs were purified twice with acid phenol/chloroform (1:1), precipitated with isopropanol, suspended in diethyl pyrocarbonate (DEPC)-treated water, and further treated with Turbo DNA-free (Ambion) to remove genomic DNA. The cDNAs were synthesized with Moloney murine leukemia virus (MMLV) reverse transcriptase (Promega) and oligo(dT) primer to quantify HEX1 mRNA expression and viral RNA accumulation. Quantitative real-time reverse transcription (RT)-PCR (qRT-PCR) was performed on a CFX96 real-time PCR system (Bio-Rad) using SsoFast EvaGreen supermix (Bio-Rad) according to the manufacturer's instructions. After initial denaturation at 95°C for 10 min, 40 cycles consisting of 5 s at 95°C and 5 s at 58°C were run. Two endogenous reference genes, the cyclophilin 1 gene (CYP1, locus FGSG_07439) and the elongation factor 1α gene (EF1α, locus FGSG_08811), were used in each experiment. The endogenous reference gene for C. parasitica, the glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH), was used according to a previous report (31).
For the following experiment, all virus-infected strains were incubated in square culture dishes (245 by 245 by 28 mm) (SPL Life Sciences) containing PDA. Mycelia were harvested from 1-cm2 areas of the cultures at cytokinesis points 0, 1,000, and 2,000 and were ground to a fine powder with liquid nitrogen in a mortar and pestle. Total RNAs and cDNAs for semiquantitative RT-PCR (sqRT-PCR) were prepared using the same procedures described above. The sqRT-PCR was performed on a C1000 thermal cycler (Bio-Rad). After initial denaturation at 95°C for 5 min, 35 or 40 cycles consisting of 30 s at 95°C, 30 s at 58°C, and 30 s at 72°C were run. Amplified DNA bands were analyzed by agarose gel electrophoresis.
The F. graminearum genomic DNA and mRNA of HEX1 were sequenced using oligonucleotide primers specific for HEX1 in PCR and RT-PCR. The 561-bp coding region contains a single 81-bp intron, which could encode a polypeptide with 186 amino acids and a molecular mass of about 20 kDa. The protein-coding region was predicted according to previous reports on Hex1 (21–23, 26).
According to phylogenetic analysis using the putative amino acid sequence of F. graminearum Hex1 and homologs from other filamentous fungi, sequence identity with F. graminearum Hex1 was highest among Fusarium species, including F. oxysporum and F. verticillioides (Fig. 1A). Putative amino acid sequences of Hex1 homologs were more closely related among plant-pathogenic fungi than among animal-pathogenic fungi except in the case of Chaetomium globosum (see Table S1 in the supplemental material).
Computer-aided comparison of the deduced amino acid sequence of F. graminearum Hex1 showed high levels of sequence identity with Hex1 homologs from F. oxysporum (84%), F. verticillioides (82%), N. crassa (72%), and M. oryzae (67%) (Fig. 1B). All predicted amino acid sequences of Hex1 and its homologs contained 14 motifs, including 3 helix-turn-helix and 11 beta sheet motifs and peroxisome targeting sequence 1 (S/R/L). Given the high level of sequence identity, we designated locus FGSG_08737 the Hexagonal protein 1 gene (HEX1) in F. graminearum.
Targeted gene deletion (Δhex1) and overexpression (HEX1 OE) mutants were generated to examine the role(s) of the HEX1 gene in the growth and pathogenicity of F. graminearum, as well the effect of the gene on the accumulation of FgV1. HEX1 was successfully replaced with gen by homologous recombination and complemented with an hph construct (Fig. 2A and andC).C). We also generated Hex1 overexpression mutants in which the HEX1 gene is under the control of the EF1α promoter (PEF1α) from F. verticillioides (Fig. 2B). All of the genetically manipulated strains were confirmed by Southern hybridization (Fig. 2D).
Quantitative RT-PCR was performed to confirm the transcript levels of deletion and overexpression mutants. The expression level of HEX1 mRNA in the HEX1 OE strain was upregulated by about three times under the control of the EF1α promoter. All of these HEX1 gene mutants were without virus infection and are denoted as virus free (VF). Virus-infected strains obtained by hyphal fusion-mediated transmission using a virus-infected wild-type strain of F. graminearum are denoted as VI. The genetic characteristics of all fungal strains used in this study are listed in Table 1.
On PDA at 5 d.p.i., the colony morphologies were very similar among WT-VF, Δhex1-VF, HEX1 OE-VF, and Δhex1::HEX1-VF strains under standard laboratory conditions (Fig. 3A, top row). All strains infected with FgV1 showed virus-associated phenotypic changes, including reduced growth of aerial mycelia, increased pigmentation, and irregular colony morphology. However, there were distinguishable phenotypic characteristics among VI strains: the colony diameter was greatest for Δhex1-VI, smallest for HEX1 OE-VI, and intermediate for WT-VI and Δhex1::HEX1-VI (Fig. 3A, bottom row).
We also measured the radial growth on PDA, CM, and MM to obtain quantitative data concerning vegetative growth. Among virus-free strains, radial growth on PDA was not significantly affected by gene deletion or overexpression (Fig. 3B). Among virus-infected strains, radial growth was fastest for Δhex1-VI, slowest for HEX1 OE-VI, and intermediate for WT-VI and Δhex1::HEX1-VI strains. Consistent with colony morphology, radial growth on PDA was slowest for the HEX1 OE-VI strain (Fig. 3B). The results on CM and MM were similar (data not shown).
Among virus-free strains, conidial production was significantly lower for the Δhex1-VF and HEX1 OE-VF strains than for the WT-VF and Δhex1::HEX1-VF strains (Fig. 3C). Relative to the conidial production by the WT-VF strain, conidial production was reduced for all four virus-infected strains. Among the virus-infected strains, the conidial production was about 3-fold lower for the Δhex1-VI strain than for the WT-VI strain, and the HEX1 OE-VI strain did not produce any conidia in CMC liquid medium (Fig. 3C).
TEM indicated that WBs were localized near septa but were absent in deletion mutants (Fig. 4A). More WBs were detected in the WT-VI than in the WT-VF strain. WBs formed in HEX1 OE strains and Δhex1::HEX1 strains with and without virus infection. In general, more WBs were observed in FgV1-infected than in VF HEX1 OE and Δhex1::HEX1 strains.
Western blot analysis was carried out to verify genetic alteration and to support microscopic observations at the protein level using anti-HEX1 antibody. One major band with a molecular mass of approximately 20 kDa was detected in all strains except the deletion mutants (Δhex1-VF and Δhex1-VI) and with slightly increased accumulation in WT-VI and Δhex1::HEX1-VI strains (Fig. 4B).
According to a previous report, deletion of the HEX1 gene reduces the ability of N. crassa to maintain cellular integrity (21). To determine whether the HEX1 gene has a similar role in F. graminearum, we used light microscopy to observe the growth rates of mycelia after the mycelia were wounded by amputation. The mycelial growth after amputation was delayed when the HEX1 gene was deleted. Six hours after amputation, the quantity of new growth was much less in Δhex1-VF than in WT-VF, HEX1 OE-VF, or Δhex1::HEX1-VF strains (Fig. 5). We also assessed cytoplasmic bleeding in all virus-free strains according to previous reports (23, 28). Cytoplasmic bleeding was observed only in the Δhex1-VF strain at 3 h postinoculation in a cutting plane and the area around it (Fig. 6).
In the pathogenicity test, deletion of HEX1 in the virus-free strains significantly reduced head blight symptoms (Fig. 7A and andB).B). Surprisingly, overexpression of HEX1 also reduced head blight symptoms. Head blight symptoms were less severe with virus-infected than with virus-free strains. The HEX1 OE-VI strain, which produced no conidia (Fig. 3C), did not induce any symptoms when wheat heads were inoculated with its culture filtrate.
qRT-PCR was conducted to measure the levels of expression of HEX1 and its homologs in the parental strain and in the other plant-pathogenic fungus used (C. parasitica). The references and sources for VF and VI strains of these isolates were described previously (Table 1) (29). The qRT-PCR analysis revealed that the HEX1 expression level was more than 6-fold greater in the WT-VI than in the WT-VF strain at 5 d.p.i. (Fig. 8A). Relative to the expression in the WT-VF strain, the expression level was increased by approximately 3.5-fold in HEX1 OE-VF and by more than 5-fold in the HEX1 OE-VI strain (Fig. 8A). The HEX1 transcript was not detected in either of the HEX1 deletion mutants. In C. parasitica, the expression of the HEX1 homolog was much greater with virus infection than without and was much greater with FgV1 infection than with CHV1 infection (Fig. 8B).
qRT-PCR was used to measure the accumulation of FgV1 RNA in the four F. graminearum strains that were infected with the virus. FgV1 RNA accumulation was substantially lower in the HEX1 deletion mutant than in the WT but was similar in the complemented mutant and the WT (Fig. 9A). FgV1 RNA accumulation was much greater in the overexpression mutant than in the other strains (Fig. 9A).
sqRT-PCR was also conducted to verify FgV1 accumulation based on the number of times that cytokinesis had occurred (i.e., based on growth stage) in growing mycelia of all infected strains (Materials and Methods and Table 2). The intensities of amplified DNA bands obtained using the FgV1-specific primer set were verified in the same manner as in the qRT-PCR analysis. Viral RNA accumulation was decreased in the Δhex1-VI strain, recovered in the Δhex1::HEX1-VI strain, and increased in the HEX1 OE-VI strain. All amplified DNA fragments showed different intensities based on the number of times cytokinesis had occurred. As the number of times cytokinesis occurred increased, the intensities diminished in all four kinds of VI strains (Fig. 9B).
In this study, we characterized the F. graminearum HEX1 gene, which encodes a protein that forms Woronin bodies (WBs) in filamentous fungi. The Δhex1 strain of F. graminearum did not form these cellular organelles, and although its colony morphology was normal, the Δhex1 strain exhibited reduced radial growth after injury, indicating that WBs play an important role in maintaining cellular integrity. Our observations of the Δhex1 strain also confirmed that WBs are important for fungal virulence. Although these functions of WBs have been reported for other fungi (21, 23, 24), it is noteworthy that WBs also have the same functions in F. graminearum, which is an important plant pathogen.
According to the Fusarium comparative database of the Broad Institute (http://www.broadinstitute.org/annotation/genome/fusarium_group), the coding region of the FGSG_08737 gene (designated HEX1 in this study) consists of 1,859 bp, and the predicted protein has 514 amino acids and a molecular mass of about 74 kDa. The results of our sequence analysis and Western blot analysis indicated a coding region of 561 bp and a 20-kDa protein. Even though our Western blot analysis used antibody that can cover sequences upstream from the 561-bp coding region, we could not detect any specific protein bands corresponding to the 74-kDa protein band in repeated experiments (data not shown). Previous Western blot analyses using the same antibody from N. crassa and M. oryzae also demonstrated that the molecular mass of the protein encoded by the HEX1 gene is about 20 kDa (21, 28). Moreover, analyses based on nucleic acid and amino acid sequences and on the function of the Hex1 protein in previous studies and in the present study strongly suggest that Hex1 shares both sequence and functional homologies among filamentous fungi. We therefore conclude that the molecular mass of the protein encoded by the HEX1 gene is 20 kDa in F. graminearum, as it is in other filamentous fungi.
In accordance with a previous proteomic study (19), we used an incubation period of 5 days for all fungal strains except in the production of conidia for the virulence assay. We first measured the conidial production of strains at 5 d.p.i. (Fig. 3C) and then extended the incubation for 2 days to collect more conidia from both deletion and overexpression mutants. Surprisingly, the rates of sporulation for the virus-infected strains were reduced by about half relative to the rates obtained in our previous study (8). This difference between the two studies might result from differences in the fungal host strain, the amount of initial inoculum, or the incubation time in CMC medium. The difference might also result from an unknown alteration of the fungal host or mycovirus during successive serial culturing after isolation.
The virulence test with wheat plants showed that deletion of the HEX1 gene reduced F. graminearum virulence by about 50%. This indicates that WBs are important to F. graminearum pathogenesis, which is in agreement with a previous study with M. oryzae (28). While deletion of the HEX1 gene did not affect conidial production by M. oryzae, it reduced conidial production by F. graminearum (Fig. 3C). In the case of M. oryzae, deletion of HEX1 resulted in severe morphological defects in appressoria and a delay in host penetration (28), and such abnormalities evidently explained the diminished virulence of the HEX1 gene deletion mutant of M. oryzae. Overall, the reduced virulence of HEX1 deletion mutants in both F. graminearum and M. oryzae seems to be caused by reduced viability of conidia.
Like the deletion mutant, the HEX1 OE mutant exhibited reduced conidial production and virulence while showing normal vegetative growth (Fig. 3C and andB).B). The results obtained with the HEX1 OE mutant might be explained by the constitutive expression of HEX1 by PEF1α. In other words, abnormal overexpression of HEX1 might explain the defects in asexual reproduction and conidial viability during pathogenesis. In this regard, the WT-VI strain, which had the highest level of expression of Hex1 among all strains, also showed reduced conidial production and virulence. Also, more FgV1 viral RNA accumulated in the HEX1 OE-VI mutant than in any of the other virus-infected strains, even though the HEX1 expression in the other virus-infected strains (except for the virus-infected deletion strain) was equivalent to that in the HEX1 OE-VI mutant (Fig. 8A and and9A).9A). This result might also be explained by the constitutive expression of HEX1 by PEF1α. In other words, continuous production of HEX1 seemed to facilitate the constant accumulation of FgV1 viral RNA. Taken together, these results suggest that a moderate level of HEX1 gene expression might be required for normal asexual reproduction and pathogenesis by F. graminearum and for FgV1 viral RNA accumulation in F. graminearum.
Semiquantitative RT-PCR was conducted to investigate the cause of differences in colony morphology among VI strains. To sample mycelia at similar growth stages, we harvested growing mycelia on PDA according to the number of times the cells had undergone cytokinesis rather than according to a set location of the growing hyphae. In agreement with the qRT-PCR results, the accumulation of FgV1 viral RNA in a solid culture medium depended on the genetic alteration of HEX1 (Fig. 9A and andB).B). We conclude that the relative levels of accumulation of viral RNA among the VI strains were inversely related to their colony growth. Estimating the number of times cells had undergone cytokinesis by measuring the intercalary lengths (cell size) of all strains provided additional evidence that the deletion of HEX1 and FgV1 infection decreased cell size (Table 2). However, the moderate decrease of cell size in the Δhex1-VF strain compared to the cell sizes in the other VF strains may not affect vegetative growth. In contrast, FgV1 infection severely decreased cell size in all VI strains. This result suggested that the severe decrease of cell size induced by FgV1 infection resulted in reduced growth of the fungal host.
Many mycoviruses have been identified and characterized from various filamentous fungi (32, 33), but only a few confer hypovirulence to their fungal host; these include Botrytis cinerea debilitation-related virus (BcDRV) (34), Sclerotinia sclerotiorum debilitation-associated RNA virus (SsDRV) (35), Sclerotinia sclerotiorum hypovirulence-associated DNA virus 1 (SsHADV-1) (36), and Rosellinia necatrix megabirnavirus 1 (RnMBV1) (37). Although these are valuable resources for studying hypovirulence caused by mycoviruses, CHV1 in C. parasitica remains a well-studied model system for understanding mycovirus-fungal host interactions (38–40).
Using genome-wide expression profiling analysis, we recently found that FgV1 causes transcriptional reprogramming of F. graminearum (18). Although the transcriptional expression profiles of many host genes were significantly altered by FgV1 infection, it is difficult to determine whether these transcriptional changes also represent translational changes. At the start of our study, we hypothesized that the upregulation in HEX1 transcription that occurs with FgV1 infection might be a cellular response to virus infection. This hypothesis was based on previous reports indicating that Hex1 and WBs function in the sealing of the septal pore after hyphal wounding (21, 24, 41). However, both qRT-PCR and semiquantitative RT-PCR demonstrated that the accumulation of FgV1 viral RNA decreased in deletion mutants and increased in overexpression mutants relative to its accumulation in the WT-VI strain (Fig. 9A and andB).B). The greater viral RNA accumulation in overexpression mutants compared to that in deletion mutants strongly suggests that FgV1 accumulation depends on HEX1 gene expression. The combined results suggest that Hex1 not only affects the formation of WBs (and therefore the sealing of septal pores) but also acts as a host factor that increases the accumulation of FgV1 viral RNA.
Although the present study demonstrates that a single gene influences virus accumulation in a fungal host, the underlying molecular mechanism remains unknown. We speculate that HEX1 functions as a host factor affecting cell-to-cell movement of FgV1 or replication of viral RNA for two reasons. First, Hex1 proteins self-assemble to form WBs, as indicated earlier (21). This process is promoted by the Woronin body sorting complex (WSC). The WSC is required to recruit Hex1 assemblies to the matrix face of the peroxisome membrane, where they can bud off to produce the WB organelle (42). The WSC can also trigger cortical association (enveloping) of the WB, which allows partitioning of the nascent WB and segregation of the newly formed organelle into a subapical compartment. After WB biogenesis in the peroxisome membrane, the newly formed WB should be moved near a septal pore if it is to contribute to septal-pore sealing. In this relocation of the organelle, the Leashin (Lah) gene product is responsible for WB tethering (22). It is also well known that some plant viruses use peroxisomes for their viral replication (43). Especially in the case of Tomato bushy stunt virus (TBSV), viral RNA replication occurs on the surface of the cytosolic membrane of peroxisomes in plants and in yeast (44). Based on the previous reports concerning WB biogenesis (21, 22, 42) and RNA viral replication on peroxisomes and on the current results concerning the presence of gene homologs related to WB biogenesis in F. graminearum (WSC for locus FGSG_01049 and Lah for locus FGSG_04119), it is tempting to speculate that FgV1 or other potential mycoviruses might use this tethering process to facilitate the accumulation and intracellular movement of viral RNAs.
A second reason for suspecting that HEX1 functions as a host factor affecting cell-to-cell movement of FgV1 or replication of viral RNA is that a previous report demonstrated that the Hex1 protein shares sequence homology with eukaryotic translational initiation factor 5A (eIF-5A) (protein secondary structures are designated in Fig. 1B) (24). The latter study also used X-ray diffraction crystallography to reveal that the tertiary structure of Hex1 is similar to that of eIF-5A in N. crassa. As a result of substantial research, the cellular functions of eIF-5A are now well known. It is interesting that eIF-5A binds to mRNA for translation elongation (45). Hence, we speculate that Hex1 might bind to FgV1 viral RNA and thereby enhance viral RNA replication.
Future research should focus on understanding how individual Hex1 proteins and/or WBs affect the accumulation of FgV1 viral RNA. Research is also needed to understand the molecular mechanisms of cell-to-cell movement and virus RNA replication.
This research was supported in part by grants from the Center for Fungal Pathogenesis (grant no. 20100001822), funded by the Ministry of Education, Science, and Technology, and from the Next-Generation BioGreen 21 Program (grant no. PJ00819801), Rural Development Administration, Republic of Korea. M. Son, K.-M. Lee, J. Yu, and M. Kang were supported by graduate research fellowships from the MEST through the Brain Korea 21 Project.
We thank G. Jedd and M. Tjota from the Temasek Life Sciences Laboratory at the National University of Singapore for the generous gift of anti-HEX1 antibody.
Published ahead of print 17 July 2013
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JVI.01026-13.