Homeobox proteins are essential for fungal differentiation and secondary metabolism in Aspergillus nidulans

The homeobox domain-containing transcription factors play an important role in the growth, development, and secondary metabolism in fungi and other eukaryotes. In this study, we characterized the roles of the genes coding for homeobox-type proteins in the model organism Aspergillus nidulans. To examine their roles in A. nidulans, the deletion mutant strains for each gene coding for homeobox-type protein were generated, and their phenotypes were examined. Phenotypic analyses revealed that two homeobox proteins, HbxA and HbxB, were required for conidia production. Deletion of hbxA caused abnormal conidiophore production, decreased the number of conidia in both light and dark conditions, and decreased the size of cleistothecia structures. Overexpressing hbxA enhanced the production of asexual spores and formation of conidiophore under the liquid submerged conditions. The hbxB deletion mutant strains exhibited decreased asexual spore production but increased cleistothecia production. The absence of hbxB decreased the trehalose content in asexual spores and increased their sensitivity against thermal and oxidative stresses. The ΔhbxA strains produced more sterigmatocystin, which was decreased in the ΔhbxB strain. Overall, our results show that HbxA and HbxB play crucial roles in the differentiation and secondary metabolism of the fungus A. nidulans.


Results
Genes coding for homeobox-type proteins in A. nidulans. To identify genes coding for homeobox-type proteins in A. nidulans genome, the protein sequence of the homeobox domain (IPR001356) was queried into the ASPGD database (www.aspgd.org). Consequently, eight genes were identified in the genome of A. nidulans FGSC4. To name these genes, their predicted protein sequences were aligned with the sequences of A. fumigatus and A. flavus homeobox proteins using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/), and the aligned sequences were input to MEGA software. Based on A. flavus homeobox proteins, the names of HbxA-HbxF were chosen (Fig. 1A). Domain analysis found that these proteins contained a 60-amino-acid-long helix-turn-helix DNA-binding domain (Fig. S1). Three proteins, HbxC, HbxE and HbxF, have a C 2 H 2 zinc finger domain at their C termini. In addition, HbxE contains a GAL4-like Zn(2) C 6 fungal type DNA binding domain (Fig. 1B).
The roles of homeobox proteins in fungal development. To predict the functions of hbxA-hbxF, mRNA levels of these genes were examined during the life cycle of A. nidulans. As shown in Fig. 2A, hbxA-hbxF were expressed during asexual development. Especially, hbxB mRNA levels were high in conidia. To further study the roles of genes coding for homeobox-type proteins, deletion mutants for each hbx gene were generated, and their developmental phenotypes were examined. WT or mutant strains were point-inoculated onto MMG or SM and cultured in the light or dark (Fig. 2B). In both conditions, the colony phenotypes of the ΔhbxA and ΔhbxB mutant strains were different from those of the WT strains. Importantly, the ΔhbxA, ΔhbxB, and ΔhbxD strains produced significantly less amount of conidia than the WT strains under both light and dark conditions (Fig. 2C). Taken together, hbxA and hbxB seem to play important roles in fungal development, and thus the functions of hbxA and hbxB were further studied. the roles of hbxA in development. To further assess the roles of hbxA, hbxA-complemented strains (C'hbxA) were generated. WT, ΔhbxA, and C'hbxA strains were point-inoculated onto MMG and their conidiophore structures were examined. Under the light condition, the ΔhbxA strain exhibited abnormal conidiophore morphology (Fig. 3A). The number of conidia in ΔhbxA strain was significantly less than in WT and C'hbxA strains (Fig. 3B). To test whether the deletion of hbxA affected the expression pattern of brlA, a key gene for asexual development, the mRNA levels of brlA in WT, ΔhbxA, and C'hbxA strains grown under the conditions that preferentially induce asexual development were examined. As shown in Fig. 3C, the deletion of hbxA reduced brlA expression by 12 h and 24 h of the induction of asexual development.
To further investigate the roles of HbxA in sexual development, these strains were inoculated onto SM. WT and C'hbxA strains produced normal sexual fruiting bodies, whereas the ΔhbxA strains formed small and abnormal cleistothecia (Fig. 3D). The size of cleistothecia in the ΔhbxA strain was smaller than that of WT or C'hbxA strains (Fig. 3E). Taken together these results demonstrated that HbxA was essential for sexual development in A. nidulans. www.nature.com/scientificreports www.nature.com/scientificreports/ overexpression of hbxA leads to enhanced conidiation. Because the absence of hbxA affected fungal development, we then tested whether the overexpression of hbxA could influence fungal development. To test this hypothesis, hbxA overexpression (OEhbxA) mutant strains were constructed. WT and OEhbxA strains were point-inoculated onto non-inducing (MMG) or inducing media (MMT), and their asexual developmental phenotypes were examined (Fig. 4A). Under the inducing conditions, the overexpression of hbxA enhanced the production of conidiospores (Fig. 4B). To further confirm this observation, OEhbxA strains were inoculated into liquid MMT (inducing condition). Whereas WT strains could not develop conidiophores in either of the inducing or non-inducing conditions, the OEhbxA strain exhibited the formation of conidiophores in liquid submerged culture (Fig. 4C). These results demonstrated that HbxA could act as an activator of asexual development in A. nidulans.
Deletion of hbxB enhances sexual development. As mentioned above, the deletion of hbxB decreased the number of conidia (Fig. 2), suggesting that HbxB may act as a developmental regulator. To test this hypothesis, WT, ΔhbxB, and hbxB-complemented strains (C'hbxB) were inoculated onto MM plates and cultured in the light or dark (Fig. 5A). Under both dark and light conditions, the ΔhbxB strains produced fewer conidia than WT or C'hbxB strains (Fig. 5B). However, the number of cleistothecia in the ΔhbxB strains was increased compared with that of WT or C'hbxB strains (Fig. 5C). The cleistothecia in the ΔhbxB strains were bigger than that of WT or C'hbxB strains under the conditions that induced sexual development (Fig. 5D,E). To further elucidate the developmental role of HbxB, the mRNA levels of two key developmental activators, BrlA and AbaA, were examined during asexual developmental processes. As shown in Fig. 5F, brlA mRNA levels were decreased in ΔhbxB strains after inducing asexual development. The abaA mRNA levels in ΔhbxB strains were also decreased by 24 and 48 h of the asexual development induction. Overall, these results suggest that HbxB is required for a correct timing of development and the balance between both developmental programs. overexpression of hbxB causes enhanced conidiation. To further investigate the role of hbxB in development, hbxB overexpression (OEhbxB) mutant strains were generated and checked the production of asexual spores. Whereas control strains can produce lots of sexual fruiting bodies, OEhbxA strains exhibited less sexual fruiting bodies under the inducing condition (Fig. 6A). In addition, the overexpression of hbxB increased the production of asexual spores (Fig. 6B), but decreased the production of cleistothecia (Fig. 6C). Taken together, these results propose that HbxB can act as a balancer between sexual and asexual development.
the roles of hbxB in trehalose biosynthesis in conidia. As mentioned above, hbxB mRNA levels were high in conidia ( Fig. 2A), implying that HbxB may participate in conidial formation and maturation. To test this hypothesis, trehalose content and stress tolerance were examined. The conidia of the ΔhbxB mutant strains contained less trehalose than the conidia of the WT or C'hbxB strains (Fig. 7A). Because trehalose is a key protective factor against environmental stresses 47,48 , the conidial tolerance against thermo and oxidative stresses was tested. At high temperature or high concentration of H 2 O 2 , the viability of the ΔhbxB mutant conidia was decreased compared to WT or C'hbxB conidia (Fig. 7B,C). We then examined the mRNA levels of tpsA, wetA, and vosA, which are associated with trehalose biosynthesis. As shown in Fig. 7D, the ΔhbxB mutant conidia exhibited www.nature.com/scientificreports www.nature.com/scientificreports/ decreased mRNA levels of tpsA, wetA, and vosA than WT or C'hbxB conidia. Overall, these results demonstrated that HbxB was a key regulator of trehalose biosynthesis. the roles of the hbxA and hbxB genes in ST production. Because the hbxA homolog hbx1 is essential for aflatoxin production in A. flavus 44 , ST production in the absence of genes coding for homeobox-type proteins was examined. Interestingly, the deletion of hbxA or hbxB affected production of ST (Fig. S2). Therefore, the roles of hbxA and hbxB in ST production were examined. ST in WT, ΔhbxA, and C' hbxA were extracted and these samples were analyzed using TLC analysis. As shown in Fig. 8A,B, the ΔhbxA mutant strains produced more ST than WT or C' hbxA strains, suggesting that HbxA may negatively affect ST production. To verify this observation, ST production was examined in WT, ΔhbxB, and C'hbxB strains. Indeed, the ST production in the ΔhbxB mutant was lower than in WT and C'hbxB strains (Fig. 8C,D). In addition, the mRNA level of aflR, a transcriptional activator of the ST biosynthesis gene cluster, was also decreased (Fig. 8E). Collectively, these results demonstrated that HbxB was necessary for proper ST biosynthesis. www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/

Discussion
Homeobox transcription factors are evolutionarily conserved proteins that play vital roles in the developmental processes of multicellular organisms, including Ascomycota and Basidiomycota 37 . These genes have been found to reside at the mating type loci of the fungal genomes and associate with fruiting body formation in various fungal species 37 . In addition, genes coding for homeobox-type proteins are associated with various fungal processes, including conidiation, virulence, and secondary metabolisms. Here, we characterized the functions of genes coding for homeobox-type proteins in the model fungus A. nidulans.
We identified eight genes coding for homeobox-type proteins in the A. nidulans genome. Among them, hbxA had the greatest effects on the development of A. nidulans. In the absence of hbxA, the conidiophore formation decreased along with decreased mRNA level of brlA, a key transcription factor for the initiation of conidiogenesis, during the early phase of conidiogenesis (Fig. 3), implying that HbxA is necessary for the proper asexual development. Similar results have been reported in other Aspergillus species 40,46 . In A. fumigatus, the ΔhbxA mutant strains exhibit an almost aconidial phenotype with fewer conidia. Moreover, the brlA mRNA level is also decreased in this mutant strain 46 . Deleting the hbxA homolog hbx1 shows similar results in A. flavus 44,45 . In other fungi, such as M. oryzae, Fusarium graminearum, and Ustilaginoidea virens, the putative hbxA orthologs (although similarity was low) are also involved in conidiogenesis. In M. oryzae, the deletion of Mohox2 results in no conidia production 40 . The Fghtf1 null mutant produces abnormal macroconidia in F. graminearum 49 . UvHox2 deletion mutants of U. virens also produce abnormal asexual structures 38 . Overall, these results suggest that HbxA has conserved roles in asexual development in fungal species.
Although HbxA plays a conserved role in asexual development, its role in secondary metabolism differs among fungi. Our results revealed the ΔhbxA mutant strains produced ST, a known aflatoxin precursor in A. nidulans (Fig. 8). However, aflatoxin and aflatrem production is abolished in the Δhbx1 mutant of A. flavus 44 . The ΔhbxA mutant of A. fumigatus produces less secondary metabolites, including fumigaclavines, fumiquinazolines, and chaetominine 46 . In both A. flavus and A. fumigatus, the hbxA homologs are involved in the regulation www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ of secondary metabolism. Although we measured ST production in these mutants, we did not check the whole chemical profiles in the mutants. Therefore, additional research should be conducted to elucidate the mechanism(s) underlying the HbxA and HbxB control of secondary metabolite production.
Unlike other genes coding for homeobox-type proteins, hbxB was highly transcribed in conidia (Fig. 2), suggesting that HbxB is involved in conidial maturation. As shown in Fig. 5, the ΔhbxB mutant conidia contained less trehalose and were more sensitive to thermal and oxidative stresses. In addition, the mRNA levels of the trehalose synthase gene (tpsA) and two spore-specific transcription factors (wetA and vosA) were decreased in the ΔhbxB mutant conidia. These observations suggest that HbxB can be a spore-specific gene. HbxB also affected fungal development. The deletion of hbxB increased the number of sexual fruiting bodies but decreased conidia number. Moreover, developmental phenotypes of OEhbxB strains exhibited the opposite phenotypes of ΔhbxB mutant strains, suggesting that HbxB acts as a balancer between the asexual and sexual developments in A. nidulans.
Although the functions of six (hbxC-hbxH) of the eight genes coding for homeobox-type proteins were not studied in detail in this study, HbxD also appears to affect fungal development. The absence of hbxD caused decreased conidial production (Fig. 2). Previous genetic analyses have demonstrated that the overexpression of hbxD suppresses fungal growth and asexual development, suggesting that HbxD is required for the proper development 50 . The orthologs of HbxD have been studied in other fungi, such as M. oryzae and C. albicans. The deletion of GRF10 suppresses hyphal growth and biofilm formation in C. albicans 43 . In M. oryzae, the absence of MoHOX7 abolishes the appressorium formation and pathogenicity 40 . These results suggest that HbxD plays diverse roles in filamentous fungi.
In summary, we characterized the roles of genes coding for homeobox-type proteins in the model fungus A. nidulans. Our results indicate that HbxA and HbxB have multifunctional roles in governing asexual and sexual developmental cycles and ST production in A. nidulans. Additional studies will be needed to provide insight into the genetic regulatory networks and action mechanisms of these transcription factors. www.nature.com/scientificreports www.nature.com/scientificreports/

The distribution, phylogenetic analyses, and sequence logos of the Homeobox domains in
Aspergillus. Three representative Aspergillus species, A. flavus NRRL 3357, A. fumigatus Af293, and A. nidulans FGSC A4, were used in this study, and their genomic data were downloaded from AspGD (http://www. aspergillusgenome.org/) 51 . A phylogenetic tree of the homeobox orthologs in these three Aspergillus species was generated with MEGA 7 software (Maximum Likelihood method based on the JTT matrix-based model) (http:// www.megasoftware.net/). The bootstrap consensus tree inferred from 1000 replicates was assumed to represent the evolutionary history of the taxa analyzed.
Strains, media, and culture conditions. Fungal strains used in this study are listed in Table S1. For routine procedures, A. nidulans was grown using solid or liquid minimal media (MM) with 1% glucose (MMG) and appropriate supplements, such as uridine, uracil, or pyridoxine. Sexual medium (20 g/l glucose, 1,5 g/l glycine, 0.52 g/l MgSO 4 7H 2 O, 0.52 g/l KCl, 1.52 g/l KH 2 PO 4 , and 1 ml/l of 1000 x trace element solution; pH 6.5; simplified as SM) 52 was used to induce the sexual developmental cycle. For ST production, fungal strains were incubated in liquid complete medium (CM) at 30 °C for 7 days. To examine the effects of hbxA overexpression, solid MMG (non-inducing), MM with 100 mM threonine (MMT, inducing), or YLC (0.1% yeast extract, 1.5% lactose, 30 mM cyclopentanone) media were used 53 . Escherichia coli DH5α cells were grown in Luria-Bertani medium with ampicillin (100 μg/ml) for plasmid amplification.
For routine procedures, wild type (WT) or mutant strains were point-inoculated onto solid MMG plates, and the plates were incubated at 37 °C for 5-7 days in the light or dark as indicated. The photographs of the colonies were taken with a Pentax MX-1 digital camera. To analyze conidiophore structures, fresh conidia were spread onto solid MM plates and incubated at 37 °C for 2 days. The agar containing conidiophores of fungal strains was cut into small blocks and examined under a Zeiss Lab.A1 microscope equipped with AxioCam 105 C and AxioVision (Rel. 4.9) digital imaging software. To count of the number of conidia, conidia were collected from each plate, washed using ddH 2 O, passed through Miracloth (Calbiochem, San Diego, CA) to collect pure conidia, and counted using a hemocytometer. Experiments were performed in triplicate for each strain.
hbx deletion mutant strains. The oligonucleotides used to construct the deletion mutants are listed in Table S2. To generate the deletion mutant strains, gene disruption cassettes were generated by using the double-joint PCR (DJ-PCR) strategies, as previously described 54 . Briefly, the 5ʹ and 3ʹ regions of genes coding for homeobox-type proteins were amplified with primer pairs DF/TR and DR/TF, respectively, from A. nidulans FGCS4 genomic DNA. The auxotrophic selection marker pyrG (AfupyrG) was amplified with primers OHS089/ www.nature.com/scientificreports www.nature.com/scientificreports/ OHS090 by using A. fumigatus AF293 genomic DNA as the template. In the overlap PCR, genes coding for homeobox-type proteins disruption cassettes were amplified from the combined 5ʹ and 3ʹ regions of genes coding for homeobox-type proteins and the AfupyrG marker by using primer pair NF/NR. For transformation, RJMP 1.59 conidia (1 × 10 8 ) were inoculated in liquid YG (MM with 0.5% yeast extract) medium and cultured for 14 h at 30 °C. Afterward, the hyphae were harvested, washed, and incubated with the Vinoflow FCE lysing enzyme (Novozymes) to generate protoplasts 55 . The deletion cassettes were introduced into the protoplasts, and the transformed cells were cultured in the selection medium (MMG without uridine or uracil). The hbx genes deletion mutant strains were confirmed by PCR followed by restriction enzyme digestion. At least three colonies per deletion mutation were isolated and phenotypically characterized.
hbxA-or hbxB-complemented strains. For the hbxA-or hbxB-complemented strains, the predicted promoters of hbxA and hbxB were amplified with the primer pairs OHS0657/OHS0658 and OHS0910/OHS0911, respectively, digested with NotI, and cloned into pHS13 56 . The resulting plasmids pYE4.1 and pSH1.1 were introduced into the recipient ΔhbxA (TYE14.1) and ΔhbxB (TSH1.1) strains to give rise to TYE27 and TSH7, respectively. The complemented strains were verified by PCR and quantitative reverse-transcription (qRT) PCR. hbxA and hbxB overexpression strains. To generate the alcA(p)::hbxA and alcA(p)::hbxB fusion construct, the hbxA and hbxB open reading frame derived from A. nidulans FGCS4 genomic DNA was amplified using the primer pair OHS0743/OHS0744 and OHS1130/OHS1131, respectively, digested with BamHI, and cloned into pHS3, which contains A. nidulans alcA promoter 56 . The resulting plasmid pYE5.1 and pSH3.1 were then introduced into TNJ36 57 to give rise to TYE19 and TSH13, respectively. Strains that overexpress hbxA and hbxB were selected from the transformants, screened by PCR and qRT-PCR after the induction of the promoter.
qRT-PCR analysis. For qRT-PCR analysis, samples were collected as previously described 58 . For vegetative samples, WT and mutant conidia were inoculated into liquid MMG and incubated at 37 °C for 12 or 16 h. The mycelia were collected, washed, squeeze-dried, and stored at −80 °C until RNA extraction. For conidium samples, WT and mutant conidia were inoculated onto solid MMG plates and incubated for 48 h. Then, conidia were collected from plates using Miracloth (Calbiochem, San Diego, CA) and stored at −80 °C until RNA extraction. To induce asexual development, WT and mutant conidia were inoculated in liquid MMG and incubated at 37 °C for 16 h. The mycelia were filtered, washed and spread onto solid MMG plate to exposure them to air. The plates were incubated at 37 °C with air-exposure to induce asexual development. Samples were collected at the designated time points following the induction of asexual development. All the samples were collected, squeeze-dried, and stored at −80 °C until RNA extraction. Conidial trehalose analysis. The conidial trehalose assay was performed as previously described 23 . WT or mutant strains were inoculated onto MMG and incubated at 37 °C for 2 days. After incubation, conidia (2 × 10 8 ) were collected, washed with ddH 2 O, resuspended in 200 mL of ddH 2 O, and incubated at 95 °C for 20 min. The supernatant was collected by centrifugation, mixed with 0.2 M sodium citrate (pH5.5), and incubated with or without trehalase (3 mU, Sigma), which hydrolyzes trehalose to glucose. The amount of glucose produced from trehalose was assayed with a glucose assay kit (Sigma). Samples untreated with trehalase served as negative controls.
Stress tolerance assay. The thermal tolerance test was carried out as described previously 26 . Approximately 10 3 conidia from plates that had been cultured for two-days were placed into ddH 2 O and incubated at 55 °C for 15 or 30 min. After incubation, the conidial samples were diluted, and approximately 100 conidia were spread onto solid MM plates. The plates were incubated at 37 °C for 48 h, and the colonies were counted. The survival rates were calculated as the ratio of the numbers of colonies on the heat-treated and untreated plates. All the experiments were carried out in triplicate.
The oxidative tolerance assay was conducted as described previously 26 . Approximately 10 3 conidia were incubated with varying concentrations (0, 0.05, or 0.1 M) H 2 O 2 for 30 min at 25 °C om temperature. After incubation, each conidial suspension was diluted, and the diluted solution was spread onto solid MM plates and cultured at 37 °C for 48 h. The numbers of colonies were determined and their ratios to the colony number in the untreated control were estimated.
ST extraction and thin-layer chromatography (TLC) analysis. To extract ST, approximately 10 5 conidia were inoculated into 5 mL liquid complete medium (CM) and cultured at 30 °C for 7 days in the dark. After cultivation, 5 mL CHCl 3 was added per sample and the samples were vigorously mixed for 1 min. The organic phases were separated by centrifugation and transferred to new glass vials. Each sample was evaporated, resuspended in 100 μl of CHCl 3 , and spotted onto a TLC silica plate (Kiesel gel 60, 0.25 mm; Merck). The TLC plate was placed into a chamber that contained the toluene:ethyl acetate:acetic acid (8:1:1 v/v) solution to resolve the samples. Afterward, the TLC plate was treated with 1% aluminum hydroxide hydrate (Sigma, St Louis, MO, USA). The images of the TLC plates were captured under UV light (366 nm). The intensities of the ST spots were quantitated using Image J software. Experiments were performed in triplicate for each strain.
Statistical analysis. Statistical differences between WT (or control) and mutant strains were evaluated using Student's unpaired t-test. Data are reported as mean ± standard deviation (SD). P values < 0.05 were considered significant.