The role of VosA/VelB-activated developmental gene vadA in Aspergillus nidulans

The filamentous fungus Aspergillus nidulans primarily reproduces by forming asexual spores called conidia, the integrity of which is governed by the NF-κB type velvet regulators VosA and VelB. The VosA-VelB hetero-complex regulates the expression of spore-specific structural and regulatory genes during conidiogenesis. Here, we characterize one of the VosA/VelB-activated developmental genes, called vadA, the expression of which in conidia requires activity of both VosA and VelB. VadA (AN5709) is predicted to be a 532-amino acid length fungal-specific protein with a highly conserved domain of unknown function (DUF) at the N-terminus. This DUF was found to be conserved in many Ascomycota and some Glomeromycota species, suggesting a potential evolutionarily conserved function of this domain in fungi. Deletion studies of vadA indicate that VadA is required for proper downregulation of brlA, fksA, and rodA, and for proper expression of tpsA and orlA during sporogenesis. Moreover, vadA null mutant conidia exhibit decreased trehalose content, but increased β(1,3)-glucan levels, lower viability, and reduced tolerance to oxidative stress. We further demonstrate that the vadA null mutant shows increased production of the mycotoxin sterigmatocystin. In summary, VadA is a dual-function novel regulator that controls development and secondary metabolism, and participates in bridging differentiation and viability of newly formed conidia in A. nidulans.


Introduction
Species of the genus Aspergillus are widespread in nature and have both beneficial and detrimental effects on humankind [1][2][3]. This genus includes plant and human pathogenic fungi, such as Aspergillus fumigatus and Aspergillus flavus, and other species that are of tremendous importance to the industrial production of enzymes, organic acids, and foods [1,4]. Aspergillus nidulans is a model ascomycetous fungus for studying fungal development and secondary metabolism [5,6]. All Aspergillus species produce asexual spores called conidia, which are the main reproductive propagule and the infectious particles [7]. Conidia are formed on a a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 tested strains were inoculated on MMG or MM with 100 mM threonine as the sole carbon source (MMT). All strains were grown on solid or liquid media with appropriate supplements at 37˚C. To check the number of conidia and cleistothecia, wild-type (WT), mutants, and complemented strains were point inoculated and cultured on solid MM or SM for four or seven days at 37˚C. Escherichia coli DH5α cells were grown in Luria-Bertani medium with ampicillin (100 μg/mL) for plasmid amplification.
For Northern blot analysis, samples were collected as previously described [28]. Briefly, for conidia, the conidia of WT and mutant strains were spread onto solid media and incubated at 37˚C. After two days of culture, the conidia were filtered, collected, and stored at −80˚C. For hyphal samples, conidia of WT and mutant strains were inoculated into 200 mL liquid MM in 1 L flasks and incubated at 37˚C. Samples of the submerged cultures were collected at designated time points and stored at -80˚C. For developmental induction, the conidia of WT and mutant strains were inoculated into liquid MM and incubated for 18 h. Mycelia were filtered, washed, and spread in a monolayer on solid MM, and the plates were air exposed for asexual developmental induction, or sealed of air and blocked from light for sexual developmental induction.

Construction of the vadA mutants
The oligonucleotides used in this study are listed in Table 2. The vadA deletion (ΔvadA) mutant strain was generated by double-joint PCR (DJ-PCR) as previously described [42]. The flanking regions of the vadA gene were amplified using the primer pairs OHS859/OHS861 and OHS860/OHS862 from A. nidulans FGSC4 genomic DNA as a template. The A. fumigatus pyrG marker was amplified from A. fumigatus AF293 genomic DNA with the primer pair OJH84/OJH85. The vadA deletion cassette was amplified with primer pair OHS863/OHS864 and was introduced into a RJMP1.59 [40] protoplasts generated by the Vinoflow FCE lysing enzyme (Novozymes) [43]. To complement ΔvadA, the WT vadA gene region, including its predicted promoter, was amplified with the primer pair OHS888/OHS889, digested with EcoRI and NotI, and cloned into pHS13 [23]. The resulting plasmid pHSN85 was then introduced into the recipient ΔvadA strain THS33.1 to give rise to THS34.1. To generate the alcA (p)::vadA fusion construct, the vadA ORF derived from genomic DNA was amplified using the primer pair OHS887/OHS889. The PCR product was then double digested with EcoRI and NotI and cloned into pHS82 [23]. The resulting plasmid pHS82 was then introduced into RJMP1.59. The vadA-overexpressing strains among the transformants were screened by Northern blot analysis using a vadA ORF probe followed by PCR confirmation.

Nucleic acid isolation and manipulation
Total RNA isolation and Northern blot analyses were performed as previously described [42,44]. The DNA probes for Northern blot analysis were amplified using the appropriate oligonucleotide pairs (Table 2) from the coding regions of individual genes using FGSC4 genomic DNA as a template. 32 P-labeled probes were prepared using the Random Primer DNA Labeling Kit (Clontech) with [α-32 P]-dCTP. Genomic DNA extractions were carried out as previously described [43].

Spore viability test
To check spore viability, fresh conidia from two-day-old cultured WT and mutant strains were spread on solid MM and incubated at 37˚C. Conidia from five-and ten-day-old cultures were then collected. About 100 conidia were inoculated onto solid MM and incubated at 37˚C for 48 h in triplicate. Survival rates were calculated as the ratio of the number of viable colonies to the number of spores inoculated.  Spore trehalose assay The spore trehalose assay was performed as previously described [22]. Briefly, conidia from two-day-old cultured WT and mutant strains were collected, washed with ddH 2 O, resuspended in 200 μL of ddH 2 O, and incubated at 95˚C for 20 min. The supernatant was collected by centrifugation, mixed with an equal volume of 0.2 M sodium citrate (pH 5.5), and incubated at 37˚C for 8 h with or without (as a control) 3 mU of trehalase (Sigma, St Louis, MO, USA). The amount of glucose generated from the trehalose was assayed with a Glucose Assay Kit (Sigma, St Louis, MO, USA) in triplicate.

Oxidative stress tolerance test
The hydrogen peroxide sensitivity of conidia was tested by spotting 10 μL of serially diluted conidia (10 to 10 5 ) on solid MM with 0 and 2.5 mM of H 2 O 2 and incubating at 37˚C for 48 h.

β(1,3)-glucan analysis
The β(1,3)-glucan assay was performed as previously described [36]. Briefly, two-day old conidia of WT and mutants were collected in ddH 2 O. Conidia suspensions (10 2 to 10 5 ) were resuspended in 25 μL of ddH 2 O and were mixed with Glucatell 1 reagents (Associates of Cape Cod, East Falmouth, MA, USA) following the manufacturer's instructions. After incubation, diazo reagents were added to stop the reaction. The optical density at 540 nm was determined. This test was performed in triplicate.

Sterigmatocystin (ST) extraction and HPLC conditions
Briefly, 10 6 conidia of each strain were inoculated into 2 mL liquid complete medium (CM) and cultured at 30˚C for 7 days. Secondary metabolites were extracted by adding 2 mL of CHCl 3 , the organic (CHCl 3 ) phase separated by centrifugation and transferred to new glass vials, evaporated in the fume hood, and resuspended in 1 mL HPLC-grade acetonitrile:methanol (50:50, v/v). The samples were filtered through a 0.45-μm pore filter. High-performance liquid chromatography with diode-array detection (HPLC-DAD) analysis was performed with a Series 1100 binary pump with an auto sampler and Nova-Pak C-18 column (Agilent Technologies, Waldbronn, Germany). Ten microliters of the samples were injected in to the column. The mobile phase consisted of acetonitrile:water (60:40, v/v). The flow rate was 0.8 mL/min. The ST stock solution (Sigma, St Louis, MO, USA) was dissolved in acetonitrile:methanol (50:50, v/v). A linear calibration curve (R 2 = 0.998) was constructed with a ST dilution series of 10, 1, 0.5, 0.1, and 0.005 μg/mL. ST was detected at a wavelength of 246 nm. The retention time for ST was approximately 5.6 min.

Microscopy
The colony pictures were taken using Sony digital (DSC-F828) camera. Micrographs were taken using a Zeiss M2Bio microscope equipped with AxioCam and AxioVision digital imaging software.

Statistical analysis
Statistical differences between WT and mutant strains were evaluated by Student's unpaired ttest. Mean ± SD are shown. P values < 0.05 were considered to be significant.

Expression and phylogenetic analyses of vadA
Our previous data showed that the accumulation of vadA mRNA in conidia is dependent on both VosA and VelB [35]. To further verify the vadA expression during asexual development, we examined the levels of vadA mRNA in WT, ΔvosA, and ΔvelB strains under asexual developmental conditions. While accumulation of vadA mRNA was detectable at 24 h after developmental induction and was high in the WT conidia, vadA mRNA was decreased in both the ΔvosA and ΔvelB strains compared to WT (Fig 1A). We also examined levels of the vadA transcript throughout the life cycle and found that the levels of the vadA transcript were high during the late phase of asexual development (Fig 1B). These results suggest that vadA is specifically expressed during conidiogenesis, and its expression is largely dependent on both VelB and VosA. Multiple sequence alignment of Aspergillus spp. suggests that the predicted VadA protein contains a highly conserved domain of unknown function (DUF) at the N-terminus (Fig 1C). VadA orthologues are found in many Ascomycota (Neurospora, Sordaria, and Marssonina) and some Glomeromycota (Rhizophagus), implying that this DUF might be ancient, and has uniquely evolved in fungi, but not plants and animals ( Fig 1D).

VadA balances development
To study functions of vadA, we generated the vadA deletion (ΔvadA) mutant, complemented strains (C'), and examined their phenotypes. The ΔvadA mutant produced light green conidia distinct from those of WT and C' strains (Fig 2A). The ΔvadA mutant produced a similar number of conidia as WT and C' strains ( Fig 2B). We then checked whether the absence of vadA altered the patterns of brlA, vosA, and velB mRNA accumulation during vegetative growth and asexual development. As shown Fig 2C, brlA mRNA accumulation in the ΔvadA mutant was detectable in vegetative growth, early asexual development (~6 h), and conidia, but was not observed in WT cells. The levels of vosA and velB mRNAs in the ΔvadA mutant were decreased compared to that of WT. These results indicate that VadA is required for full conidial pigmentation and proper expression of asexual developmental genes.
To test whether VadA is also associated with sexual development, WT, ΔvadA, and the complemented strains were inoculated on SM, and the numbers of sexual fruiting bodies were counted. As shown Fig 2D and 2E, the ΔvadA mutant produced significantly higher numbers of sexual fruiting bodies compared to the WT and complemented strains, suggesting that VadA is also required for proper sexual development in A. nidulans.

VadA governs the conidial integrity
As described above, vadA is a conidia-specific gene (Fig 1) and is required for proper expression of brlA, vosA, and velB in conidia (Fig 2), implying that VadA has a potential role in conidiogenesis. To test this idea, the conidial viability, trehalose content, oxidative stress tolerance, β-glucan level, and expression of genes associated with conidiogenesis were compared between WT, ΔvadA, and C' strains (Fig 3). When we checked the viability of conidia of colonies grown for five and ten days, the ΔvadA mutant conidia had a slight loss of viability at ten days ( Fig  3A). We measured trehalose in two-day conidia of WT, ΔvadA, and C' strains, and showed that the trehalose content in the ΔvadA mutant conidia was significantly decreased compared to that in the WT and complemented strains ( Fig 3B). As trehalose acts as a protectant against several stresses, we examined the conidial tolerance of the strains to oxidative stress. As shown in Fig 3C, the ΔvadA mutant conidia were more sensitive to oxidative stress than WT and C' conidia. We also checked β(1,3)-glucan levels and found that its levels in the ΔvadA conidia were higher than that of WT and C' conidia ( Fig 3D).
To correlate phenotypic changes caused by the deletion of vadA with molecular events, we examined mRNA levels of brlA, wetA, tpsA, orlA, rodA, and fksA in conidia. The ΔvadA mutant conidia showed decreased mRNA levels of tpsA and orlA, which are associated with trehalose biosynthesis, and increased transcript levels of brlA, rodA, and fksA, suggesting that VadA is required for properly controlling expression of certain spore-metabolic and developmental regulatory genes in conidia (Fig 3E). Taken together, VadA is expressed during conidiogenesis and globally affects metabolism, spore wall structure, and feed-back control thereby exerting the integrity of conidia.

The absence of vadA leads to elevated ST production
We then tested whether the absence of vadA would affect the biosynthesis of secondary metabolites in A. nidulans. TLC image showed that all ΔvadA mutant strains produced increased amounts of ST compared to WT (S1 Fig). To further examine this result, we extracted ST in WT, ΔvadA, and C' strains and analyzed these sample using HPLC. As shown Fig 4, the ΔvadA mutant produced increased amounts of ST compared to WT and C' strains, suggesting that VadA may repress ST production in A. nidulans (Fig 4). Overall, these results imply that VadA functions controlling development, spore primary metabolism, and hyphal secondary metabolism.
Overexpression of vadA leads to enhanced conidiation As mentioned above, the absence of vadA caused enhanced production of cleistothecia and altered brlA expression. To further test the sufficiency of VadA influencing fungal development, we constructed the vadA overexpression (OE) mutant and examined the developmental phenotypes upon induction. Under non-inducing conditions, the OEvadA mutant produced a comparable number of asexual spores compared to WT. However, when induced, overexpression of vadA resulted in significantly enhanced production of conidiospores (Fig 5). Collectively, these results support the idea that vadA is necessary for balancing asexual and sexual development, and VadA may act as an activator of asexual development.

Discussion
The velvet regulators are fungal NF-κB-type transcript factors that regulate both development and metabolism [25,35]. In particular, VosA and VelB (and their orthologues) form a heterocomplex that binds to the promoters of various developmental genes in A. nidulans and Histoplasma capsulatum in a sequence-specific manner [34,35,46]. Our previous studies demonstrated that the VosA/VelB complex regulates common targets grouped as VosA/VelB-activated developmental genes (VADs, e.g., tpsA) and VosA/VelB-inhibited developmental genes   .g., brlA and fksA), and that many VADs and VIDs are regulatory factors that subsequently control expression of downstream genes, leading to maturation of spores and completion of sporogenesis [35,36]. The vadA gene is one of VADs defined in A. nidulans. In the ΔvosA or ΔvelB mutant conidia, the levels of vadA transcript are radically decreased compare to WT conidia (Fig 1A). Our previous chromatin immunoprecipitation followed by microarray (ChIP-chip) analysis showed that the promoter region of vadA was enriched with VosA [35]. Based on the result of MEME (Multiple Em for Motif Elicitation) analysis, we proposed a predicted VosA-VelB binding site in the vadA promoter region (-245 CTACCCCAGGC -234). These results imply that the expression of vadA in conidia might be directly activated by VosA and VelB.   VadA is a hypothetical protein that contains a highly conserved DUF at the N-terminus (Fig 1C). The ePESTfind (http://emboss.bioinformatics.nl/cgi-bin/emboss/epestfind) program predicted that VadA has a putative PEST sequence for rapid degradation at the C-terminus (397-HTGFYPTSASSLSDPSSSAELLPTPR-422). In addition, VadA contains a nuclear localization signal (NLS)-pat4 (51-HKKR-54) and might be localized in the nucleus (56.5%), as predicted by PSORT II (http://psort.hgc.jp/form2.html). VadA is required for proper regulation of several developmental genes such as brlA, fkaA, and rodA in conidia. Taken together, we propose that VadA is a novel regulator involved in transcriptional control of spore-specific and metabolic genes during the lifecycle. Further studies defining the function and the molecular mechanisms for the role of VadA in sporogenesis are needed.
VadA is a conserved in many fungi including Aspergillus species, other Ascomycota (Neurospora, Sordaria, and Fusarium), and some Glomeromycota (Rhizophagus). However, orthologues of VadA were not found in Candida albicans or Saccharomyces cerevisiae. To further test the role of the VadA homologues in other Aspergillus species, we examined the expression of vadA mRNAs in two major pathogens: A. fumigatus and A. flavus. Our preliminary data showed that transcript levels of the vadA in A. fumigatus and A. flavus are high in conidia and during the late phase of conidiation (data not shown). Moreover, in A. fumigatus, the vadA deletion mutant, similar to the ΔvadA mutant in A. nidulans, produces light green conidia that differ from WT. Taken together, these results imply that VadA is a spore-specific regulator, which plays a crucial role in sporogenesis in Aspergillus spp.
Although the mRNA of vadA is primarily detectable in conidia, VadA is also required for balanced progression of asexual and sexual development (Figs 2 and 5). The ΔvadA mutant exhibited increased production of sexual fruiting bodies, and overexpression of vadA caused The role of vadA in A. nidulans elevated formation of conidiospores, suggesting that VadA may act as an activator of asexual development. We then examined the phenotypes of the vadA overexpression (OE) mutant strain in liquid submerged culture and found that the ΔvadA mutant cannot produce conidia or activate brlA expression, implying that VadA indirectly activates asexual development in A. nidulans.
Taken together, we present a working model for the VadA-mediated regulation of sporogenesis in A. nidulans (Fig 6). In phialides, vosA and velB are activated by AbaA [23], and the two proteins form a hetero-complex then localize in in the conidial nucleus. This VosA-VelB heterodimer binds directly to the VosA-Responsive Element (VRE) present in the promoter The role of vadA in A. nidulans region of vadA and activates expression of vadA. VadA is then involved in the downregulation of brlA, fksA, and rodA and the proper expression of tpsA and orlA in conidia, thereby exerting the integrity and vaiblity of conidia. The molecular mechanism of vadA-mediated sporogenesis, as well as the genetic position of VadA, will help to understand the regulatory networks governing sporogenesis in association with VosA/VelB.