Effects of light on secondary metabolism and fungal development of Fusarium graminearum

Authors

  • H. Kim,

    1. Department of Agricultural Biotechnology and Center for Fungal Pathogenesis, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • H. Son,

    1. Department of Agricultural Biotechnology and Center for Fungal Pathogenesis, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Y.-W. Lee

    Corresponding author
    1. Department of Agricultural Biotechnology and Center for Fungal Pathogenesis, Seoul National University, Seoul, Korea
    • Correspondence

      Yin-Won Lee, Department of Agricultural Biotechnology and Center for Fungal Pathogenesis, Seoul National University, Seoul 151-921, Korea. E-mail: lee2443@snu.ac.kr

    Search for more papers by this author

Abstract

Aims

The objectives of this study were (i) to characterize white-collar (WC) orthologues of the filamentous fungus Fusarium graminearum, (ii) to investigate light-responsive phenotypes by the deletion of Fgwc-1 and Fgwc-2 genes and (iii) to examine the roles of those genes in constant light and darkness in relation to secondary metabolite synthesis and development.

Methods and Results

Production of secondary metabolites and asexual/sexual development of deletion mutants, ΔFgwc-1 and ΔFgwc-2, were assessed in constant light and darkness compared to the wild-type strain. The results showed that deletion of Fgwc-1 and Fgwc-2 impaired early onset of carotenogenesis, photoreactivation and the maturity of perithecia during sexual development. Conidiation of the ΔFgwc-1 and ΔFgwc-2 mutants was derepressed in constant light, but not in darkness. Moreover, the individual mutants produced more aurofusarin and trichothecenes than the wild-type strain in both constant light and darkness.

Conclusions

Both Fgwc-1 and Fgwc-2 are required for light-dependent processes in F. graminearum, whereas light-independent processes such as aurofusarin and trichothecene biosynthesis are derepressed by deletion of these genes. Thus, Fgwc-1 and Fgwc-2 play roles as positive and negative regulators, depending on the requirement of light for biological activity.

Significance and Impact of the Study

These results will extend the knowledge of the photobiology of Fusarium graminearum and will increase current understanding of light regulatory mechanisms mediated by white collar in secondary metabolism and fungal development.

Introduction

Light is an environmental signal that regulates numerous biological processes in various organisms. Fungi have also evolved intricate molecular mechanisms to detect and respond to light (Idnurm et al. 2010; Rodriguez-Romero et al. 2010). Various aspects of fungal responses to light have been studied extensively, including circadian rhythm, morphogenesis, reproduction, secondary metabolism and phototropism (Loros and Dunlap 2006; Corrochano 2007; Avalos and Estrada 2010; Bayram et al. 2010; Kim et al. 2011a; Fuller et al. 2013). Most known fungal responses to light are mediated by blue light, although other wavelengths can have an effect, indicating that photoreceptor systems complicatedly occur in the fungal photoresponses (Linden et al. 1997; Kim et al. 2011a).

Many photoreceptors have been identified and characterized in fungi, including white collar, vivid, phytochrome, opsin, rhodopsin and cryptochrome (Corrochano 2007; Avalos and Estrada 2010; Corrochano and Avalos 2010). The white-collar photoreceptor proteins, encoded by wc-1 and wc-2 genes, physically interact with each other to form the heterodimeric white-collar complex (WCC), a key element in light signal transduction pathway that was initially identified and characterized in Neurospora crassa (Ballario et al. 1996; Linden and Macino 1997). The WC-1 of N. crassa contains multiple domains: a poly-glutamine (poly-Q) region, a light, oxygen, voltage (LOV) domain, a per-ARNT-sim (PAS)-Fold domain, a PAS domain and a zinc-finger DNA binding domain (Froehlich et al. 2002; He et al. 2002; Cheng et al. 2003). BLAST analysis with the amino acid sequences of WC-1 homologues from ascomycetes, zygomycetes and basidiomycetes shows that although the central regions containing both the LOV and PAS-fold domains are highly conserved in most homologues, most of the divergent regions are found near the N- and C-termini (Kim et al. 2011a). The poly-Q terminal regions in the divergent N-termini have been suggested to be involved in transcriptional activation, resulting in a novel mechanism of activation or repression in response to light (Kim et al. 2011a). Furthermore, the presence of a zinc-finger DNA binding domain and the diversity of poly-Q regions suggest that transcriptional regulatory mechanism mediated by white collar may differ among fungal species (Ambra et al. 2004; Kim et al. 2011a).

The public availability of several fungal genome databases has led to the identification in many fungi of homologues of N. crassa wc-1 and wc-2, as well as to investigations of the light regulatory mechanism mediated by those genes (Casas-Flores et al. 2004; Idnurm and Heitman 2005; Lee et al. 2006; Kihara et al. 2007; Estrada and Avalos 2008; Kim et al. 2011a; Fuller et al. 2013). Orthologues of N. crassa wc-1 have been identified and characterized in Fusarium fujikuroi (wcoA) and Fusarium oxysporum (wc1) (Estrada and Avalos 2008; Ruiz-Roldan et al. 2008). Disruptions of these genes were found to result in pleiotropic phenotypes including partial reduction of light-dependent carotenoid biosynthesis (Estrada and Avalos 2008; Ruiz-Roldan et al. 2008). In particular, strains in which wcoA was disrupted produced less gibberellins and more bikaverins than the wild-type strain under nitrogen-limiting conditions when grown in either constant light or darkness, suggesting that WcoA plays a light-independent role in secondary metabolic pathways of F. fujikuroi (Estrada and Avalos 2008). To date, however, light-responsive phenotypes mediated by white-collar in F. graminearum have not yet been determined. Thus, assessing the phenotypes resulting from deletion of white-collar genes in F. graminearum is relevant to provide a better understanding how light regulates the physiology and development of this fungus.

Light is known to be essential for the induction of sexual development of the homothallic fungus F. graminearum, a pathogen that causes losses of economically important crops such as wheat, barley and corn and produces mycotoxins harmful to humans and animals (Leslie and Summerell 2006; Cavinder et al. 2012). The requirement of light for the sexual development gives rise to the question of whether light may also influence other biological processes in F. graminearum. In our previous study investigating the cellular and developmental responses regulated by transcription factors, we systematically deleted all genes encoding putative transcription factors, including the wc-1 and wc-2 homologues, designated Fgwc-1 and Fgwc-2 (Son et al. 2011). However, we did not assess the differences between light and dark conditions on the morphological and physiological changes in 17 phenotype categories resulting from the deletion of Fgwc-1 and Fgwc-2 (Son et al. 2011). Therefore, we have investigated the light-responsive phenotypes of Fgwc-1 and Fgwc-2 deletion mutants of F. graminearum and examined the role of these two genes in secondary metabolism and fungal development under conditions of constant light and darkness.

Materials and methods

Fungal strains, media and light treatment

The wild-type strain Z-3639 of F. graminearum (Bowden and Leslie 1999) and the deletion mutants (Son et al. 2011) derived from this strain were stored in 20% glycerol solution at −80°C. The media used in this study were formulated and used as described (Leslie and Summerell 2006). To determine fungal responses to specific wavelengths of light, customized fungal growth chambers were constructed from blue (400–530 nm), green (450–600 nm), orange (540–700 nm) and red (600–700 nm) cellophane filters. For all experiments, light was provided by conventional 22 W fluorescent bulbs (KumHo Electric, Inc., Seoul, Korea).

Phenotype assessments

A mutant library of 657 F. graminearum genes encoding putative transcription factors, including Fgwc-1 and Fgwc-2, was generated at our previous work, in which all genes were replaced by single copy of geneticin-resistance gene GEN by homologous recombination (Son et al. 2011). The two individual mutants of Fgwc-1 and Fgwc-2 were obtained, respectively, and showed identical phenotypes each other. For further assessing phenotypes, one of two mutants was used and designated ΔFgwc-1 and ΔFgwc-2, respectively. Mutants of each gene were used for phenotyping with three replications. Conidial suspensions generated in carboxymethyl cellulose (CMC) liquid medium (Leslie and Summerell 2006) were used for all assessments of phenotype. To measure conidiation in CMC liquid medium, the same sized agar blocks of wild-type, ΔFgwc-1 and ΔFgwc-2 cultures grown on complete agar medium (CM) were used to inoculate 5 ml of CMC liquid medium, and the cultures were incubated for 5 days at 25°C on a rotary shaker. For investigation of pigments and conidiation from CM cultures, one microlitre of a spore suspension (106 conidia ml−1) of each strain was point-inoculated onto CM, and the cultures were incubated for 7 days in constant white light or darkness. Changes in pigmentation on the culture surface were observed visually. To measure conidiation, an agar block (4 mm in diameter) from each culture was resuspended in 1 ml of distilled water in a microcentrifuge tube by vortexing for several seconds, and the number of conidia was counted using a hemocytometer (Superior, Marienfeld, Germany).

Trichothecene analysis

Conidial suspensions (105) were inoculated into 10 ml of minimal liquid medium supplemented with 5 mmol l−1 agmatine (MMA, Gardiner et al. 2009) and incubated for 7 days in constant white light or darkness. Trichothecenes were extracted from 150 μl of culture filtrates by mixing with 250 μl of ethyl acetate/methanol solution (4 : 1, v/v) (He et al. 2007). Extracts (10 μl) transferred to a new vial were dried and derivatized in 50 μl of Sylon BTZ (BSA + TMCS + TMS1, 3 : 2 :3, Supelco, Bellefonte, PA, USA) by heating at 60°C for 5 min. Each solution was extracted sequentially with 200 μl of n-hexane and 200 μl of distilled water (Park et al. 2012). The trichothecenes in the upper layer were analysed using a Shimadzu QP-5000 gas chromatograph–mass spectrometer (GC-MS; Shimadzu, Kyoto, Japan) as previously described (Seo et al. 1996).

Photoreactivation

Conidial suspensions (106 conidia ml−1) of wild-type and mutant strains were serially diluted and point-inoculated onto CM. The plates were exposed to UV light (4·2 W in m2) for 3, 5 or 7 min and allowed to recover in constant white light or darkness for 3 days.

QRT-PCR analysis

Extraction of total RNA and generation of cDNA were performed as previously described (Kim et al. 2011a). Primers used in this study were listed in Table S1. Expression of Phr1 (Broad Institute ID: FGSG_00797), Gip1 (FGSG_02328), Pks12 (FGSG_02324), Tri6 (FGSG_03536), CarB (FGSG_03065) and CarRA (FGSG_03066) was measured by quantitative PCR. The PCRs were performed in a 7500 real-time PCR systems (Applied Biosystems, Foster City, CA), and the reaction conditions were followed by previously described (Kim et al. 2011a). Expression of each gene was normalized to the expression of house-keeping gene Ubh (FGSG_01231) and calculated as fold differences based on 2−ΔΔCt method.

Sexual reproduction

Each strain was grown on carrot agar medium at 25°C for 5 days. The mycelia in each culture were removed with 1 ml of 2·5% Tween 60 solution to induce sexual development, and the plates were incubated in constant white light at 25°C. Seven to 10 days after sexual induction, perithecia were dissected on glass slides in a drop of 20% glycerol, and asci were flattened under a coverslip. Asci rosettes and ascospores were observed under a DE/Axio Image A1 microscope (Carl Zeiss, Oberkochen, Germany).

Results

Identification of WC-1 and WC-2 orthologs in F. graminearum

FgWC-1 and FgWC-2 in the F. graminearum genome showed high sequence similarity to WC-1 (69%) and WC-2 (71%) of N. crassa. Fgwc-1 (FGSG_07941, chromosome 4) of F. graminearum contains a 3161 nt open reading frame (ORF) encoding a 1035 aa protein annotated as a white collar-1, and Fgwc-2 (FGSG_00710, chromosome 1) contains a 1633 nt ORF encoding a 483 aa protein annotated as a zinc-finger protein white collar-2 (MIPS F. graminearum Genome Database, http://mips.gsf.de/genre/proj/FGDB/). Similar to WC-1 of N. crassa, FgWC-1 contains a poly-Q region at the N terminus, a LOV domain, a PAS-Fold domain, a PAS domain and a zinc-finger DNA binding domain. FgWC-2 also contains a PAS domain and a zinc-finger DNA binding domain. The presence of zinc-finger DNA binding domains in both FgWC-1 and FgWC-2 suggests that they likely function as transcription factors.

Impairment of ΔFgwc-1 and ΔFgwc-2 in photoreactivation

Photoreactivation is a light-dependent response in which UV-induced DNA damage is repaired by light-dependent enzymatic reactions (Thoma 1999). To investigate whether photoreactivation requires Fgwc-1 and Fgwc-2, we evaluated the spore germination and subsequent mycelial growth of spores exposed to UV light and allowed to recover in light or darkness. Following exposure of spores to UV light for 5 or 7 min, the wild-type strain grew robustly over 3 days in constant light while did not grow in constant darkness, supporting a role for photoreactivation in the F. graminearum wild-type strain (Fig. 1a). However, when exposed to UV light for 7 min, neither the ΔFgwc-1 nor ΔFgwc-2 mutant grew in constant white light or darkness, indicating defects in the photoreactivation mechanism of ΔFgwc-1 and ΔFgwc-2 mutants (Fig. 1a). To clarify the mechanism through which Fgwc-1 and Fgwc-2 regulate UV damage repair, we measured the expression of Phr1 in the wild type and mutants during photoreactivation. The gene Phr1 is predicted to encode a deoxyribodipyrimidine photolyase in F. graminearum, which is homologue of photolyase gene Phr1 in F. oxysporum (Ruiz-Roldan et al. 2008). After 1 h of photoreactivation in constant light, expression of Phr1 was highly induced (approximately 7 fold) in the wild-type strain, whereas no induction of Phr1 was observed in the ΔFgwc-1 and ΔFgwc-2 strains (Fig. 1b). These results were consistent with findings in F. oxysporum, Cercospora zeae-maydis and Bipolaris oryzae, in which the disruption of wc-1 homologues resulted in defects in photoreactivation, resulted from the nonexpression of genes encoding photolyases (Kihara et al. 2007; Ruiz-Roldan et al. 2008; Kim et al. 2011a). Taken together, these results suggest that Fgwc-1 and Fgwc-2 are involved in photoreactivation by regulating photolyase genes.

Figure 1.

Photoreactivation of Fusarium graminearum strains. (a) Serial dilutions of all strains were point-inoculated onto complete agar medium (CM) (106, 105, 104, and 103 conidia ml−1). After UV irradiation for 3, 5 and 7 min, the plates were incubated for 3 days in constant white light (top) or darkness (bottom). (b) Regulation of Phr1 in photoreactivation by Fgwc-1 and Fgwc-2. After exposure to UV light for 7 min, cultures were incubated under constant light or darkness for 1 h. The experiments were performed with three biological replicates. (■) WT; (image_n/jam12381-gra-0001.png) Fgwc-1 and (□) Fgwc-2.

Aspects of pigment productions in ΔFgwc-1 and ΔFgwc-2

To assess the roles of FgWC-1 and FgWC-2 in the synthesis of pigmented secondary metabolites, we evaluated the production of pigments by the wild-type and mutant strains in constant white light or darkness. The biosynthesis of carotenoids, which accumulate as yellow or orange pigments, is induced by light in various fungi (Nelson et al. 1989; Estrada and Avalos 2008; Jin et al. 2010). We found that F. graminearum wild-type strain grown on CM for 3 days in constant white light visually exhibited the production of an orange pigmented material which is likely a carotenoids, whereas both ΔFgwc-1 and ΔFgwc-2 mutants did not show production of similar colour pigments (Fig. 2a). Furthermore, gene expression results that transcript levels of carotenoid biosynthetic genes CarB and CarRA were approximately twofold decreased in both mutants compared with the wild type (Fig. 3) support our observations. In darkness, as expected, none of these strains produced an orange pigment over 3 days (data not shown). Thus, this observation suggests that Fgwc-1 and Fgwc-2 are involved in earlier onset of carotenoid biosynthesis in the presence of light.

Figure 2.

Pigmentation of Fusarium graminearum strains in constant white light and darkness. WT, F. graminearum wild-type strain Z-3639; ΔFgwc-1, Fgwc-1 deletion mutant; ΔFgwc-2, Fgwc-2 deletion mutant. (a) Cultures in the first row were incubated for 3 days, and cultures in the second and third rows were incubated for 7 days in constant white light. Photos were taken of the bottoms (first and second rows) and tops (third row) of the plates. (b) Cultures were grown in darkness for 4 days (top row) and 7 days (bottom row). Photos were taken of the bottoms of all plates.

Figure 3.

qRT-PCR analyses of genes involved in aurofusarin and carotenoid biosynthesis. Total RNA was obtained from cultures grown in constant light condition for 1 day. Expression of each gene was normalized to Ubh and calculated by 2ΔΔCt. Data represent fold differences in expression. The experiments were performed with three biological replicates. (■) WT; (image_n/jam12381-gra-0001.png) Fgwc-1 and (□) Fgwc-2.

Beyond the productions of carotenoid by F. graminearum, this fungus also produces two other pigments such as aurofusarin and rubrofusarin, which the colour of aurofusarin is dependent on the ambient pH, ranging from golden yellow to red/purple (Ashley et al. 1937; Kim et al. 2005; Frandsen et al. 2006). In constant light, all strains showed substantial accumulation of reddish pigments, compared with cultures grown in darkness, indicating that light is conducive for aurofusarin biosynthesis on CM (Fig. 2). Furthermore, the ΔFgwc-1 and ΔFgwc-2 strains produced more reddish pigments than those of the wild type in either constant white light or darkness (Fig. 2). In particular, we observed that mutants grown in darkness accumulated reddish pigments into the growth medium after 4 days of incubation, whereas the wild-type strain did not (Fig. 2b), suggesting that FgWC-1 and FgWC-2 play a role in the repression of aurofusarin biosynthesis light-independently. To investigate how FgWC-1 and FgWC-2 regulate aurofusarin biosynthetic genes, we measured the expression of Gip1 and Pks12, aurofusarin biosynthetic genes, from cultures grown in light condition. The expression of Gip1 and Pks12 from ΔFgwc-1 and ΔFgwc-2 strains was highly induced compared to the wild type (Fig. 3), indicating that FgWC-1 and FgWC-2 regulate aurofusarin biosynthetic genes. These results support our hypothesis that FgWC-1 and FgWC-2 are involved in the repression of aurofusarin biosynthesis.

To determine whether specific wavelengths of light are responsible for the biosynthesis of aurofusarin, we incubated all strains in customized fungal growth chambers with cellophane filters that specifically transmitted blue, green, orange or red light. However, we observed no differences in reddish pigment production in response to specific wavelengths, with cultures grown in constant white light showing similar appearance to cultures grown in constant blue, green, orange or red light (data not shown).

Trichothecene productions mediated by light

To investigate whether light affects trichothecene biosynthesis, we measured the production of deoxynivalenol (DON) and 15-acetyl-deoxynivalenol (15-ADON) by wild-type, ΔFgwc-1 and ΔFgwc-2 strains grown in either constant light or darkness. All three strains produced lower quantities of these trichothecenes when grown in constant light than in darkness; in constant light, DON was not detected, and 15-ADON production was approximately 20- to 30-fold lower than in darkness (Table 1). Moreover, the results were supported by gene expression data that transcript level of Tri6, a global transcriptional regulator for trichothecene biosynthetic genes, from cultures grown in constant light was induced, approximately ranging from threefold to sevenfold, compared with the cultures grown in darkness (Table 1). Interestingly, the ΔFgwc-1 and ΔFgwc-2 mutants produced approximately two- to threefold more trichothecenes than did the wild-type strain, except that DON was not detected from cultures grown in constant light (Table 1). Taken together, these results indicate that constant light represses trichothencene biosynthesis and that the deletion of Fgwc-1 and Fgwc-2 derepresses the production of trichothecenes in both constant light and darkness.

Table 1. Effects of light on trichothecene production and Tri6 gene expression of Fusarium graminearum strainsa
 LightDark
DON15-ADON Tri6 b DON15-ADON Tri6
  1. a

    Production of DON (deoxynivalenol) and 15-ADON (15-acetyl-DON) was measured in μg ml−1 of medium, ± standard deviation. WT, Fusarium graminearum wild-type strain Z-3639; ΔFgwc-1, Fgwc-1 deletion mutant; ΔFgwc-2, Fgwc-2 deletion mutant; n.d., not detected.

  2. b

    Expression of Tri6 by qPCR was normalized to Ubh gene and was calculated relative to expression in wild type grown on light condition. Data represent fold differences in expression, and the relative expression of each gene was calculated as 2ΔΔCt. The range of expression, in parentheses, for each gene = (2ΔΔCt−s – 2ΔΔCt+s), where s = the standard deviation of the ΔΔCt value.

WTn.d.19 ± 31·0 (0·9–1·1)14 ± 3414 ± 713·5 (3·1–3·8)
ΔFgwc-1n.d.39 ± 91·2 (1·1–1·3)65 ± 131176 ± 2087·6 (6·8–8·5)
ΔFgwc-2n.d.37 ± 81·2 (1·1–1·3)24 ± 1852 ± 433·1 (2·6–3·5)

Derepression of conidiation in ΔFgwc-1 and ΔFgwc-2

Although the growth rates of mutants and wild-type strain on CM did not differ significantly, either in constant light or darkness (data not shown), the wild type exhibited more fluffy mycelia on the media in constant light (Fig. 2a). To determine whether the deletion of Fgwc-1 and Fgwc-2 affects asexual development, we compared the abilities of the wild-type and mutant strains to produce conidia in CMC liquid medium (Leslie and Summerell 2006). All strains produced comparable amounts of conidia, with no significant differences between the wild type and mutants (Fig. 4a). For further investigation, we also tested the abilities of these three strains to produce conidia in CM, a culture medium that does not favour conidiation of F. graminearum (Leslie and Summerell 2006). As expected, the wild-type strain produced no conidia after 7 days incubation in darkness with few produced in constant light (Fig. 4b). In contrast, ΔFgwc-1 and ΔFgwc-2 mutants grown in constant light produced large numbers of conidia, but not in darkness (Fig. 4b), suggesting that the asexual development of Fgraminearum is mediated by FgWC-1 and FgWC-2.

Figure 4.

Effects of Fgwc-1 and Fgwc-2 on conidiation of Fusarium graminearum. Conidia were quantified in cultures grown in (a) carboxymethyl cellulose (CMC) liquid media and (b) complete agar medium (CM). n.d., not detected. (image_n/jam12381-gra-0004.png) WT; (image_n/jam12381-gra-0002.png) Fgwc-1 and (image_n/jam12381-gra-0003.png) Fgwc-2.

Deletion of Fgwc-1 and Fgwc-2 affects the maturity of perithecia during sexual development

To determine the self-fertility of all strains used in this study, cultures grown on carrot agar plates were induced to undergo sexual development in constant white light. Self-fertility of the wild-type and mutant strains was determined by production of perithecia and asci containing ascospores. All strains began to produce perithecia 3 days after sexual induction. At 7 days after sexual induction, the wild type produced abundant mature perithecia, whereas the perithecia produced by the ΔFgwc-1 and ΔFgwc-2 mutants contained immature asci that did not develop mature ascospores (Fig. 5). However, the numbers of mature ascospores were higher in the ΔFgwc-1 and ΔFgwc-2 mutants at 10 days than at 7 days (data not shown), suggesting that the deletion of Fgwc-1 and Fgwc-2 delays the maturity of perithecia in F. graminearum.

Figure 5.

Sexual development of ΔFgwc-1 and ΔFgwc-2. The top and middle rows show perithecium production on carrot agar medium. The white arrows in the bottom row indicate immature asci. All photos were taken 7 days after sexual induction. Size bar = 50 μm.

Discussion

The functions of genes related to toxin biosynthesis, sexual reproduction, pigmentation and pathogenicity have been extensively studied in F. graminearum, and their regulatory mechanisms have been investigated in the presence of various environmental factors. Regarding light as an environmental signal, the role of white collar in F. graminearum photobiology has not been determined. The results presented in this study showed that light has effects on secondary metabolic process in F. graminearum, including the biosynthesis of aurofusarin and trichothecenes. Recently, Fanelli et al. (2012) presented that white pulsing light inhibits fumonisin biosynthesis in F. verticillioides with the reduction of FUM gene expression. In this study, we also found that light has inhibition effects on the production of trichothecenes by F. graminearum. The ΔFgwc-1 and ΔFgwc-2 mutants produced more trichothecenes compared with the wild-type strain in constant light, leading to a hypothesis that the light inhibition of trichothecene biosynthesis is likely due to the role of Fgwc-1 and Fgwc-2. However, our finding, that the mutants grown in darkness produced more trichothecenes compared with the wild-type strain, suggests that FgWC-1 and FgWC-2 play negative roles in trichothecene biosynthesis regardless of light. Furthermore, we observed that both the ΔFgwc-1 and ΔFgwc-2 mutants produced more reddish pigments compared with the wild-type strains, although light enhanced the biosynthesis of aurofusarin. Taken together, our results suggest that FgWC-1 and FgWC-2 act as negative regulators, either directly or indirectly, of light-independent secondary metabolism such as the biosynthesis of aurofusarin and trichothecenes.

Many fungal species, including N. crassa, Trichoderma atroviride, and Bipolaris oryzae, display photoinducible conidiation, with conidiation reduced by deletions of the wc-1 homologues of these fungi (Lauter et al. 1997; Casas-Flores et al. 2004; Kihara et al. 2007). Although the effects of various wavelengths of light on conidiation have been investigated in Fusarium spp. (Chaudhary and Prasad 1975; Das and Busse 1990; Fanelli et al. 2012), the general features of conidiation regulated by light cannot be interpreted because of variations in strains, experimental approaches and culture conditions (Avalos and Estrada 2010). For example, conidiation was stimulated by light in the F. fujikuroi wild-type strain IMI58289 (Avalos et al. 1985), but not in a different wild-type strain (Prado et al. 2004). Furthermore, light regulatory mechanisms have been associated with mechanisms regulated by nutrients such as nitrogen and carbon sources (Avalos and Estrada 2010). The F. graminearum wild-type strain used in this study did not show extensive conidiation in response to light (data not shown). Conidiation of F. graminearum can be induced on CMC liquid and YMA media, but not on CM and PDA media; thus, F. graminearum wild-type strain did not produce significant amount of conidia on CM or PDA regardless of the presence of light. Unlike in N. crassa, both the ΔFgwc-1 and ΔFgwc-2 mutants of F. graminearum derepressed conidiation, producing large numbers of conidia on both CM and PDA (data not shown) and indicating that F. graminearum possesses a mechanism for repression of asexual development mediated by white-collar proteins. Similar derepression of conidiation was observed in Cercospora zeae-maydis, a foliar pathogen of maize (Kim et al. 2011a). The disruption of CRP1, the wc-1 homologue in C. zeae-maydis, resulted in the production of large numbers of mature conidia in constant light, a condition typically unfavourable for conidiation, whereas wild type did not produce conidia in constant light, suggesting that disruption of CRP1 resulted in a global derepression of conidiation (Kim et al. 2011a). Interestingly, our results also showed that derepression of conidiation in the ΔFgwc-1 and ΔFgwc-2 mutants exclusively occurs in constant light, but not in darkness, suggesting that other photoreceptors or light-responsive regulatory proteins may be involved in conidia formation. This hypothesis is supported by recent results, showing that the deletion from F. fujikuroi of the cryD cryptochrome gene, which encodes the blue/UV-A light photoreceptor, induces the formation of macroconidia in the presence of light, although F. fujikuroi wild-type strain usually produces microconidia (Castrillo et al. 2013). Moreover, deletion of either FgVeA or FgVelB, genes encoding members of the velvet complex that coordinates light signals for fungal development and secondary metabolism, was found to lead the derepression of conidiation in F. graminearum (Jiang et al. 2011, 2012). Taken together, our results demonstrate that derepression of conidiation in F. graminearum is mediated by FgWC-1 and FgWC-2, possibly by working together with other photoreceptors or light-responsive regulatory proteins.

The wc-1 homologues of various fungi have been found to be required for light-inducible biological processes such as photoreactivation and carotenoid biosynthesis (Corrochano and Avalos 2010). In F. graminearum, either white light or near-UV light is needed for the induction of sexual development under laboratory conditions (Leslie and Summerell 2006). Thus, it is not surprising that Fgwc-1 and Fgwc-2 are essential for photoreactivation, early onset of carotenoid biosynthesis and maturity of perithecia during sexual development. Mutations of the wc-1 orthologues of F. oxysporum and F. fujikuroi resulted in partial reductions in amounts of carotenoids, with a lower transcriptional induction of carB, a structural gene for the carotenoid pathway (Estrada and Avalos 2008; Ruiz-Roldan et al. 2008). In contrast, we observed that the ΔFgwc-1 and ΔFgwc-2 mutants and F. graminearum wild-type strain yielded visibly distinguishable phenotypes in carotenoid production. These discrepancies may warrant further investigations of the exact mechanism by which FgWC-1 and FgWC-2 mediate carotenoid biosynthesis in F. graminearum.

Sexual development is another light-dependent process in the homothallic fungus Fgraminearum. Somewhat surprisingly, despite the requirement of light for sexual development in F. graminearum, relatively little is known about the regulatory mechanisms how photoreceptors are involved in sexual development. Furthermore, considering the importance of Fgwc-1 and Fgwc-2 for light-dependent biological processes, we expected that deletions of Fgwc-1 and Fgwc-2 would have effects on sexual developments (e.g. abolishment of perithecia formation or ascospores). Contrary to these expectations, we only found the delay of perithecial maturity in the ΔFgwc-1 and ΔFgwc-2 mutants. These little effects on sexual development led us to postulate that other light-responsive regulatory proteins, including other photoreceptors, are involved in sexual development.

The roles of light in regulation of fungal development and secondary metabolism raise a question of whether light influences additional components, such as pathogenesis. The direct relationship between wc-1 homologues and pathogenicity has been studied in various fungi (Idnurm and Heitman 2005; Ruiz-Roldan et al. 2008; Kim et al. 2011a,b). For example, the pathogenicity of disrupted mutants of wc1 was studied in F. oxysporum, which is pathogenic in both tomatoes and mice (Ruiz-Roldan et al. 2008). Although showing similar progression of the disease index in tomato as the wild-type strain, the mutant showed a fivefold lower mortality rate as the wild type in mice, suggesting that wc1 is dispensable for pathogenicity in plants, but is required for full virulence in mice. Similarly, the BWC1 gene of the human pathogen Cryptococcus neoformans was found to contribute to the severity of disease in mice (Idnurm and Heitman 2005). Light was found to suppress disease development by Magnaporthe oryzae, the causal agent of rice blast disease (Kim et al. 2011b). Consequently, mgwc-1 mutants were more virulent than wild type in the presence of light supporting a role for wc-1 in light suppression of disease. The results suggested that either M. oryzae recognize the darkness to mobilize virulence effectors or that other photoreceptors are involved in the light-dependent disease suppression (Kim et al. 2011b). In contrast to M. oryzae, light is required for C. zeae-maydis to induce foliar necrosis on maize leaves (Kim et al. 2011a). Kim et al. (2011a) showed that disruption of a gene encoding a putative blue-light photoreceptor homologous to WC-1 displayed the loss of ability in multiple aspects of pathogenesis, including tropism of hyphae to stomata and the formation of appressoria; a resulting mutant did not cause disease on maize plants. Thus, the authors demonstrated that C. zeae-maydis uses light as a key environmental signal to coordinate pathogenesis with maize photoperiodic responses (Kim et al. 2011a). In our previous work, to examine pathogenicity by deletion of either Fgwc-1 or Fgwc-2 in F. graminearum, we inoculated conidial suspensions of mutants and wild type onto wheat head followed by incubation in a greenhouse for 14 days, with exposure to daily cycles of day and night (Son et al. 2011). Under these conditions, we observed no significant differences in disease development between the mutants and wild type. Although deletion of Fgwc-1 and Fgwc-2 did not affect to pathogenesis at the previous work, further detailed experiments, in which these inoculated plants were exposed to constant light or darkness, are needed to better understand how light influences the development of pathogenesis through white collar.

Using single knockout strains of Fgwc-1 and Fgwc-2, we found that their changed phenotypes exhibited similar patterns compared to the wild type, indicating that FgWC-1 and FgWC-2 are involved in the same regulatory mechanisms of secondary metabolism and fungal development. Interestingly, we found that deletion of Fgwc-1 has slightly greater effects on secondary metabolism and fungal development compared to the deletion of Fgwc-2, particularly in conidiation and trichothecene productions. To explain these findings, we cannot exclude a possibility that FgWC-1 and FgWC-2 may have different features in the localization, phosphorylation and transcriptional activation, with respect to previous studies of N. crassa (Corrochano and Avalos 2010). However, further assessments of the differences between ΔFgwc-1 and ΔFgwc-2 mutants may include the determination of their abilities to activate or repress target genes or their interactions with other light-regulated proteins.

In summary, we have shown here that the deletion of Fgwc-1 and Fgwc-2 in F. graminearum had effects on secondary metabolism and fungal development, under constant white light and/or darkness. Of light-dependent responses, carotenoid biosynthesis, photoreactivation and sexual development were impaired in ΔFgwc-1 and ΔFgwc-2 mutants grown in constant white light. In contrast, the production of reddish pigments, trichothecenes and conidia was derepressed in the ΔFgwc-1 and ΔFgwc-2 mutants compared with the wild-type strain. These findings suggest that the roles of FgWC-1 and FgWC-2 as negative and positive regulators are likely to depend on the requirements for light of biological processes in F. graminearum.

Acknowledgements

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by Ministry of Education (No. 2013R1A1A2006103) and by the Basic Science Research Program through the NRF funded by Ministry of Science, ICT & Future Planning (No. 2008-0061897).

Conflict of interest

No conflict of interest declared.

Ancillary