The AGC Kinase SsAgc1 Regulates Sporisorium scitamineum Mating/Filamentation and Pathogenicity

The AGC signaling pathway represents a conserved distinct signaling pathway in regulation of fungal differentiation and virulence, while it has not been identified or characterized in the sugarcane smut fungus Sporisorium scitamineum. In this study, we identified a PAS domain-containing AGC kinase, SsAgc1, in S. scitamineum. Functional analysis revealed that SsAgc1 plays a regulatory role on the fungal dimorphic switch.

unknown in S. scitamineum, although it has been extensively investigated in the model organism Ustilago maydis, the causal fungus of corn smut (3)(4)(5)(6). The conserved matingtype loci, a locus and b locus, are present in S. scitamineum. Functional study of the a locus genes in S. scitamineum has been limited to MFA2 (7), and the role of b locus in fungal mating has been validated through deletion analysis (1).
Quorum sensing (QS) in bacteria plays a role in coordinating cell density-dependent gene expression, including virulence genes in bacterial pathogens (8,9). Studying fungal QS has recently started. In Candida albicans or Saccharomyces cerevisiae, two aromatic alcohols tyrosol and tryptophol have been reported as filamentationpromoting QS molecules (10)(11)(12), and both are degradation products of aromatic amino acids in the Ehrlich pathway (13). Feedback regulation of S. cerevisiae morphology by tryptophol is mediated by the cyclic AMP (cAMP)/protein kinase A (PKA) signaling pathway, integrating sensing of nitrogen availability (11). However, there have not been any reports of any signaling pathway regulating the production of aromatic alcohols essential for C. albicans filamentation. Another fungal QS molecule, farnesol, identified in C. albicans as a negative regulator of hyphal formation, was shown to be regulated by a two-component system (14). The Cek1 mitogen-associated protein kinase (MAPK) involved in filamentous growth in C. albicans has been shown to be regulated by the QS molecule farnesol through suppression of Sho1 (a transmembrane adaptor protein)-dependent Cek1 phosphorylation (15). Our preliminary bioinformatic analysis identified several S. scitamineum orthologs of C. albicans Cek1 MAPK signaling components, but the functional study is lacking.
MAPK and cAMP/PKA signaling pathways, which are highly conserved in eukaryotic organisms, perceive and respond to a variety of external or internal signals for the regulation of cell growth, development, stress response, and pathogenicity (16)(17)(18)(19). They have been reported to control a range of physiological processes, including fungal development and infection, in various pathogenic fungi such as Magnaporthe oryzae (20), Fusarium graminearum (21), U. maydis (22), C. albicans (23), Aspergillus flavus (24), and Cryptococcus neoformans (16). The AGC kinase family includes cAMP-dependent protein kinase 1 (PKA), cGMP-dependent protein kinase (PKG), and protein kinase C (PKC), which share sequence similarities in their catalytic domains and are widely conserved among eukaryotic genomes (25). Although AGC kinases are important regulators of cell growth, metabolism, and cell death, much remains to be learned about their molecular functions and targets in plant or fungal organisms (26).
We recently identified a serine/threonine kinase, annotated as a homolog of Schizosaccharomyces pombe Cek1, from our transcriptome analysis between the wild-type S. scitamineum and its b locus deletion mutant under mating conditions (27). This serine/threonine kinase belongs to the AGC kinase family and is thus named SsAgc1 (for S. scitamineum Agc1). We found that the SsAGC1 gene is differentially expressed between mating and nonmating S. scitamineum cultures (27), indicative of a potential function in fungal mating/filamentation. In this study, we generated and characterized the ssagc1Δ mutant in mating/filamentation and disease symptom development. We found that SsAgc1 was required for S. scitamineum mating/filamentation, and transcriptional profiling demonstrated that SsAgc1 regulates expression of genes controlling fungal mating/filamentation and encoding enzymes for tryptophan metabolism. Exogenous addition of tryptophan or tryptophol could restore the defective mating/filamentation in the ssagc1Δ mutant. SsAgc1 is also required for S. scitamineum pathogenicity, possibly due to its function in dikaryotic hyphal formation and/or oxidative stress tolerance. We infer that SsAgc1 may participate in a signaling pathway to promote fungal mating/filamentation.

RESULTS
Identification of an AGC kinase in S. scitamineum. A serine/threonine kinaseencoding gene was previously identified as a significantly (P Յ 0.05) differentially expressed gene in a nonmating and nonfilamentous mutant compared to the wild-type strain (27); therefore, it may serve a function in fungal mating/filamentation. The reads obtained by RNA sequencing with the wild-type sample (27) covers the gene as annotated by NCBI (https://www.ncbi.nlm.nih.gov/nuccore/1003330579; locus_tag "SPSC_00276") and skipped the region annotated as intron. Therefore, we followed the NCBI annotation of this gene for prediction of its encoding product. The predicted polypeptide (NCBI:protein accession no. CDR87150.1) consists of 4,170 amino acid residues, with a molecular weight (MW) of 435.72 kDa and pI of 6.15. As it belongs to the AGC (cAMP-dependent, cGMP-dependent, and protein kinase C) kinase family, we named it SsAgc1. SsAgc1 protein contains a PAS domain (amino acids [aa] 2486 to 2546), a protein kinase domain (aa 3162 to 3561) harboring an ATP-binding domain (aa 3168 to 3304), the kinase active site (aa 3307 to 3502) with two activation loop (A-loop) motifs (aa 3303 to 3313 and 3457 to 3465), an AGC-kinase-Cter domain (aa 3562 to 3655), and a response regulatory (RR) domain of the CheY-like superfamily (aa 3893 to 4010) (Fig. 1A).
Serine/threonine protein kinase homologs were identified using a BLASTP search against the nonredundant protein sequence (nr) database with SsAgc1 protein sequence as bait. The top 10 hits (sequence coverage of Ͼ80% and identity of Ͼ60%) were all from smut fungi. We also performed BLink search using SsAgc1 protein sequence and selected the top hits from ascomycetes, including S. pombe, S. cerevisiae, Individual domains of SsAgc1 are represented by black or gray boxes, with the domain names given above the boxes. The PAS domain is the Per-Arnt-Sim domain, usually found in a protein acting as a signal sensor. The response regulatory (RR) domain contains a predicted phosphorylation site at 3943D residue. The domains of SsAgc1 were identified with interPro-Scan (http://www.ebi.ac.uk/interpro/). (B) Phylogenetic analysis of the deduced amino acid sequences of SsAgc1. The phylogenetic tree is constructed using the neighbor-joining phylogeny of COBALT, the constraint-based multiple alignment tool (https://www .ncbi.nlm.nih.gov/tools/cobalt/re_cobalt.cgi) (50). The position of SsAgc1 in the phylogenetic tree is indicated by the pink highlighting. The NCBI:protein accession numbers of all the aligned kinases are given in parentheses (species names are given in brackets as in the figure): SsAgc1 (CDR87150. and C. albicans, from plants, including Arabidopsis thaliana and Glycine max, and from the nematode Caenorhabditis elegans, to generate the phylogenetic tree together with the top 10 most similar AGC kinases from smut fungi. The reconstructed phylogenetic tree is shown in Fig. 1B. SsAgc1 was clustered with the serine/threonine protein kinase from Sporisorium reilianum SRZ2, and most closely related to the kinases of three other smut fungi, Ustilago maydis, Ustilago hordei, and Ustilago bromivora. The serine/threonine protein kinases from two plant species were clustered in the other clade and related to ascomycetes. The nematode kinase clustered separately in an outgroup (Fig. 1B). No signature motif of the AGC groups (28) was identified in the SsAgc1 sequence, and the related serine/threonine kinases in the phylogenetic tree were classified as members of the other group AGC kinase family, including S. pombe Cek1 (SpCek1), S. cerevisiae Rim15 (ScRim15), and C. albicans Rim15 (CaRim15) (28), suggesting that the SsAgc1 protein is a member of the "other type" AGC kinases.
The amino acid sequence conservation between SsAgc1 and SpCek1 or ScRim15 is low and restricted in the protein kinase domain. Thus, the function of SsAgc1 may not be conserved with these two ascomycetous orthologs. The function of this AGC family protein has not been characterized in smut fungi. We generated the ssagc1Δ mutants in S. scitamineum and characterized them in sexual mating, dikaryotic hyphal formation, and pathogenicity.
SsAgc1 is essential for S. scitamineum mating/dikaryotic hyphal formation. The ssagc1Δ mutants were generated in the wild-type (WT) S. scitamineum MAT-1 background (see Fig. S1A in the supplemental material), and two independent strains were confirmed by Southern blotting (Fig. S1B). We assessed S. scitamineum mating/filamentation by mixing and spotting sporidia of opposite mating type on solid PDA medium. The in vitro culture of mixed WT MAT-1 and MAT-2 sporidia gave rise to dikaryotic hyphae, and thus, the colonies had a fluffy appearance ( Fig. 2A). Mixed cultured WT MAT-2 with the ssagc1Δ mutants, however, displayed obviously reduced filamentous and radial growth ( Fig. 2A), although the ssagc1Δ sporidial colony was comparable to the WT sporidial colony. Genetic complementation of the SsAGC1 gene ( Fig. S1C) was performed using a truncated fragment encoding the region containing only annotated domains (aa 2082 to 4170) of the SsAgc1 protein, as detailed in Materials and Methods. These genetically complemented strains could exhibit fully restored mating/filamentation, although the sporidial colony of the complemented strains looked different (drier) from that of the WT or ssagc1Δ mutants ( Fig. 2A). This result confirmed that SsAGC1 is required for S. scitamineum mating/filamentation. Interestingly, ssagc1Δ mating/filamentation could be restored when cultured under nitrogen-starved conditions (MM-N in Fig. S1D), indicating that nutrient-restricted conditions may promote filamentation even without SsAgc1 function. We further tested the effect of cAMP on ssagc1Δ filamentation. The addition of 10 mM cAMP could effectively promote filamentous growth in the mixed culture of MAT-2 and ssagc1Δ (Fig. S1D). Next, we determined the intracellular concentration of cAMP in the wild type and ssagc1Δ mutant, with the deletion mutant of SsUAC1 (29), an ortholog to the U. maydis adenylyl cyclase-encoding gene UAC1 (30), as a positive control. The results showed that out of 10 7 sporidia, the cAMP concentration of the wild-type strain was 27.87 pmol/ml, while that in the ssuac1Δ mutant was 0.74 pmol/ml, confirming that cAMP production was indeed blocked due to the loss of adenylyl cyclase function. However, the cAMP concentration of the ssagc1Δ mutant was 25.39 pmol/ml, comparable to that in the wild type. These results indicate that SsAgc1 may function downstream of cAMP or in a parallel pathway in regulating S. scitamineum mating/ filamentation.
We examined the cell morphology of WT and ssagc1Δ mutant by microscopy, after staining nuclei with fluorescent dye Hoechst. Most of the WT sporidia appeared like typical rod-shaped cells, containing one nucleus per cell, as expected (Fig. 2B). In contrast, the ssagc1Δ sporidia were frequently seen as elongated, "chain-like" cells that resembled pseudohyphae with one nucleus per septum (Fig. 2B). Such altered cell morphology was likely due to failed cytokinesis. The percentage of such normal/ abnormal (pseudohyphal) cell morphology was quantified in both WT and ssagc1Δ sporidia (n Ն 300; Fig. 2C), and the results indicated that SsAgc1 kinase was involved in regulation of sporidial cell morphology and/or proper mitosis. We further compared the morphology of dikaryotic hyphae after sexual mating between the WT and the ssagc1Δ mutant. The WT dikaryotic hyphae were long and smooth and contained two nuclei per section, as shown in Fig. S2A, which was rarely seen in the ssagc1Δ mutant. In most cases, ssagc1Δ sporidia mixed with the WT sporidia remained as individual yeast-like cells (Fig. S2B), confirming that sexual mating was blocked due to the loss of SsAgc1 function. We occasionally saw cells in a chain in the WT ϫ ssagc1Δ culture (Fig. 2D), but the hyphal morphology was distinct from that of WT dikaryotic hypha formed after sexual mating (Fig. S2A). Costaining with Calcofluor white (for cell wall and septum) and Hoechst (nucleus) confirmed that such structures were pseudohyphae, as they contained only one nucleus per septum (Fig. 2D). Overall, we confirmed that the ssagc1Δ mutant was defective in sexual mating and dikaryotic hypha formation, which could be restored by the addition of cAMP.
SsAgc1 regulates tolerance to oxidative stress. Next we tested tolerance of S. scitamineum sporidia toward various stressful conditions, including oxidative stress (H 2 O 2 ), osmotic stress (NaCl), and cell wall stress (SDS). We noticed that a hypersensitivity response to H 2 O 2 resulted from SsAGC1 deletion, specifically in cultures grown on PDA or MM-N but not in the cultures grown on YePSA medium (Fig. 3A). No obvious difference was observed in terms of osmotic stress or cell wall stress resistance (Fig. 3A). We tested the effects of different carbon sources by comparing YePSA (sucrose) and MM-N (glucose) media and did not find any change in the hypersensitivity toward oxidative stress in the ssagc1Δ mutant (Fig. 3B), indicating that the tolerance to oxidative stress was regulated by SsAgc1 in a glucose-independent manner. We further assessed the effect of cAMP on oxidative stress tolerance. The results showed that the addition of 10 mM cAMP could effectively restore resistance to 1 mM H 2 O 2 in the ssagc1Δ sporidia, on either PDA or MM-N medium (Fig. 3C), indicating an overlap between the SsAgc1 and cAMP signaling pathways in regulating tolerance to oxidative stress.
Overall, we found that SsAgc1 was required for tolerance to oxidative stress, but it was not directly related to glucose availability as reported for other PAS domaincontaining AGC kinases (31,32).
SsAgc1 is essential for S. scitamineum pathogenicity. To determine whether SsAGC1 is involved in pathogenicity, we inoculated the susceptible sugarcane seedlings with the mixed S. scitamineum sporidial cells of opposite mating types in the combinations of WT ϫ WT or WT ϫ ssagc1Δ. The sugarcane seedlings inoculated with WT ϫ WT sporidial mixture displayed the characteristic symptoms of "smut whip" emerging from the shoots at 90 to 120 days postinoculation (Fig. 4A). In contrast, seedlings infected by WT ϫ ssagc1Δ combination remained healthy through the observed period of 180 days and never developed whip symptom ( Fig. 4A and B). Therefore, we conclude that SsAGC1 is required for S. scitamineum pathogenicity, which is likely due to its function in regulating mating/filamentation and/or oxidative stress tolerance. SsAgc1 controls expression of the genes governing fungal mating/filamentation and tryptophan metabolism. To obtain a better understanding of SsAgc1 regulation of S. scitamineum mating/filamentation, we performed a transcriptome analysis to identify potential target genes of SsAgc1 signaling, under mating conditions, comparing the WT ϫ WT combination and the WT ϫ ssagc1Δ combination. We identified 1,019 differentially expressed genes (DEGs; |log2| Ն 1 and P Յ 0.05), of which 344 overlapped in the two biological replicates (Fig. S3A and Data Set S1). The pathways enriched with these overlapping DEGs suggest that SsAgc1 may promote transcription of genes involved in tryptophan metabolism (Data Set S1, Fig. S3B, and Fig. S4), as well as MAPK signaling pathway involved in fungal mating (Fig. S5). The a locus gene PRA2, fungal mating regulator encoding gene PRF1, and the genes encoding enzymes involved in tryptophan metabolism were among the DEGs (Data Set S1). We then performed quantitative real-time PCR (qRT-PCR) to verify the differential expression of selected genes in the ssagc1Δ mutant compared to the WT, mixed with compatible WT sporidia of the opposite mating type to induce mating/filamentation. These selected genes included a and b locus genes and transcriptional factor PRF1 involved in fungal mating/filamentation, and ARO8, ARO9, DC, TYNA and amiE genes for tryptophol or indoleacetate (IAA) biosynthesis from tryptophan. Our results showed that the a locus genes MFA1, PRA1, and PRA2 were indeed significantly (P Յ 0.05) downregulated, while the b locus gene bE or bW was not changed significantly (Table 1). TYNA1 was upregulated (P Յ 0.01), while amiE downregulated (P Յ 0.01) in the ssagc1Δ mutant during S. scitamineum mating/filamentation (Table 1). DEGs were consistently identified by transcriptome analysis and qRT-PCR as listed in Table 1 and schematically illustrated as Fig. S6, suggesting that SsAgc1 may regulate transcription of genes controlling fungal mating/filamentation and tryptophan metabolism in S. scitamineum.
Tryptophan metabolites restored ssagc1⌬ mating/filamentation. Our transcriptional profiling results suggest that SsAgc1 plays a key role in regulation of tryptophan metabolism, which produces a known yeast quorum-sensing molecule tryptophol (13), and phytohormone indoleacetate (IAA). Thus, we tested the effects of tryptophan, tryptophol, and IAA on S. scitamineum mating/filamentation. Our results showed that tryptophan could partially restore ssagc1Δ filamentation only at a very low concentration (3.3 M; Fig. 5A), and tryptophol (25 M) could fully restore ssagc1Δ filamentation (Fig. 5A). Microscopic observation was performed to closely examine the hyphal morphology in the WT and ssagc1Δ mutant. We observed abundant nonmating sporidia in the ssagc1Δ mutant without any treatment (Fig. 5B, black arrowhead), and hyphae were formed occasionally. In contrast, long and smooth hyphae were formed in the WT cultured under the same conditions (Fig. 5B). Treatment with tryptophan or tryptophol promoted hyphal formation in the ssagc1Δ mutant, which were morphologically identical to WT hyphae (Fig. 5B). However, IAA could not restore ssagc1Δ mating/filamentation (Fig. S7A). In summary, we found that tryptophan and tryptophol could differentially promote S. scitamineum mating/filamentation at various concentrations and were able to at least partially restore the defects of ssagc1Δ in mating/ filamentation.
Tryptophol production during S. scitamineum sporidial budding growth. As tryptophol could promote hyphal formation in the WT strain and fully restore the defects of ssagc1Δ in mating/filamentation, we suspected that it may also act as a quorum-sensing molecule in S. scitamineum. To prove that, we examined the relationship of tryptophol with S. scitamineum cell density. Interestingly, the addition of tryptophol (25 M) displayed a growth-promoting activity on WT sporidia when the initial sporidial density was adjusted to 10 5 cells/ml, but such growth-promoting property was not observed when the initial inoculation of WT sporidia was increased to 10 7 cells/ml (Fig. 5C). However, when the initial sporidial density was as low as 10 3 cells/ml, growth (budding) was suppressed by tryptophol (Fig. 5C). We further measured tryptophol production in S. scitamineum sporidial growth (budding) by high-resolution electrospray ionization mass spectrometry (HR-ESI-MS). Our results showed that in the first 8-h growth, WT sporidia increased steadily, with concurrent increase in tryptophol production ( Fig. 6 and Fig. S7B and C), confirming that tryptophol production is dependent on cell density in S. scitamineum. Sporidial growth of two ssagc1Δ mutants was slightly faster than that of the WT. However, tryptophol production in these two mutants did not show significant differences compared to that in the WT sporidia (Fig. 6); therefore, loss of SsAGC1 did not cause changes in tryptophol production.

DISCUSSION
In multicellular eukaryotes, including fungi, AGC kinases have been reported to link various signaling events to regulate cell growth, development, and aging. The evolutionarily conserved TORC1-Sch9 in S. cerevisiae was shown to link nutrient sensing and aging (33), and it likely acts to provide functional compensation to cAMP/PKA-deficient yeast mutants (34). The cAMP/PKA pathway downstream kinase Rim15 cross-talks with the TORC1-Sch9 pathway in S. cerevisiae and Yarrowia lipolytica to regulate yeast dimorphic transition (33,35). Homologs of the AGC kinase Sch9 have been identified in several pathogenic fungi, including Aspergillus nidulans (36), Aspergillus fumigatus (37), and F. graminearum (38), where it contributes to fungal virulence independently or via partial overlap with the cAMP/PKA pathway. In smut fungi, however, the AGC signaling pathway has not been identified or characterized.
Our present study identified an AGC kinase-encoding gene in S. scitamineum, named SsAGC1. The predicted polypeptide encoded by SsAGC1 possesses the conserved catalytic domains while lacking an N-terminal lipid binding domain and a C-terminal regulatory domain containing a consensus hydrophobic sequence, as reported in ascomycetous AGC kinases. Instead, it contains a PAS domain in front of the catalytic domains, and a response regulatory (RR) domain of the CheY-like superfamily at the C terminus. Our genetic complementation was performed using a truncated polypeptide containing all the conserved domains (2082 to 4170 aa), which could fully restore the mating/filamentation defect in the ssagc1Δ mutant, indicating that this region is sufficient for SsAgc1 function in S. scitamineum mating/filamentation. The long FIG 6 Detection of tryptophol in the WT and the ssagc1Δ mutant during sporidial growth. Sporidia from the wild-type strain or ssagc1Δ mutants (two independent strains) were cultured in the liquid YePS medium at 28°C, with shaking at 200 rpm, until an OD600 of 2.0 was reached. The sporidia were then diluted in 1,000 ml fresh YePS liquid medium, adjusting to a cell density of 10 5 cells per ml, and cultured for another 8 h under the same conditions. The number of cells for each hour was acquired by measuring the OD absorption at 600 nm. The growth curve for each culture was prepared by plotting the logarithmic values of OD600 versus incubation time. For every hour, 10 ml of sporidial culture was extracted with 1.5ϫ volume of ethyl acetate, and dried by rotary evaporation (EYELA, OSB-2100). The crude extracts were then dissolved in methanol and filtered with 0.22-m PVDF membrane (Nylon6) for HR-ESI-MS analysis to determine the concentration of tryptophol. Means Ϯ standard deviations (error bars) are shown for the tryptophol concentrations. N-terminal peptide (1 to 2081 aa) of SsAgc1 contains no identified domains, but the gene locus encoding this region was shown to be transcribed, in our previous highthroughput RNA sequencing (RNA-seq) data (27).
The presence of a PAS domain in the middle of SsAgc1 indicates that it may also be a PAS kinase, which is broadly evolutionarily conserved among eukaryotes, but not in C. elegans (32,39). S. cerevisiae contains several PAS kinases, including Rim15, and they function in nutrient sensing, glucose homeostasis, and/or oxidative stress tolerance (31,32,40). However, the molecular mechanisms of PAS kinase regulation and function are largely unknown. There have not been reports on any function of PAS kinases in fungal filamentation/dimorphic switch. Our study showed that the filamentation defect of the ssagc1Δ mutant could be restored by culturing the fungus under nitrogen starvation conditions; therefore, it may represent the first PAS kinase essential for proper filamentation in the pathogenic fungus S. scitamineum in a nutrient-dependent manner.
The ssagc1Δ mutants could not mate/filament when mixed with the WT sporidia of the opposite mating type, in which a wild-type copy of SsAGC1 gene was present. We inferred that SsAgc1 may predispose sporidia to mating, and the loss of SsAgc1 may make sporidia unable to respond to pheromone-induced sexual mating, even though the sporidia of the other mating type are functional.
Our results showed that deletion of SsAGC1 led to upregulation of the TYNA-1 gene and downregulation of the amiE gene. Production of the potential fungal quorumsensing molecule tryptophol and phytohormone IAA both rely on the tryptophandependent biosynthesis pathway and share many intermediate metabolites (41)(42)(43). However, we noticed that tryptophol and IAA had different effects on S. scitamineum differentiation and morphogenesis. The addition of tryptophan or tryptophol could effectively promote S. scitamineum mating/filamentation and at least partially restore ssagc1Δ defects, while the addition of the phytohormone IAA could not. We also tested the effect of another aromatic alcohol, tyrosol, and it was not able to restore ssagc1Δ mating/filamentation (data not shown). Tryptophol might act as a potential quorumsensing compound, as it shortened the lag phase of WT haploid sporidia grown in liquid culture, dependent on initial cell density. WT sporidia displayed a trend of cell density-dependent production of endogenous tryptophol. We speculate that S. scitamineum may manipulate IAA biosynthesis pathway to fungal QS molecule production as a part of its colonization strategy especially during its in planta growth stage. Detour of IAA biosynthesis or transport has recently been reported in S. reilianum, the closest smut relative of S. scitamineum (44,45).
Overall, our study suggests a connection between AGC kinase and tryptophol production and/or signaling, but at present, we are not clear whether, or how, SsAgc1 regulates the responses to quorum sensing. There are also several interesting questions remaining unanswered in our work, such as whether SsAgc1 directly interacts with the signal molecule(s), given the presence of a PAS domain. The C-terminal RR domain implies nuclear translocation and direct transcriptional regulation of downstream gene expression by SsAgc1, pending further verification. Overall, we believe our work represents the first report on the AGC signaling pathway in S. scitamineum growth and differentiation. Furthermore, SsAgc1 signaling may integrate cell density, nutrient availability, and fungal mating/filamentation.

MATERIALS AND METHODS
Growth conditions and fungal strains used in this study. Teliospores of sugarcane smut collected from the fields in Guangdong province of China (21°12= 36== N; 101°10= 12== E) by Yan et al. (1) were maintained in Lian-Hui Zhang's lab, and the MAT-1 or MAT-2 haploid sporidia isolated from these teliospores were used in this study. The culture medium used in this study include YePSA medium (1% yeast extract, 2% peptone, 2% sugar, 1.5% agar), YePS liquid medium (1% yeast extract, 2% peptone, 2% sugar [pH 7.0]), YePS soft medium (1% yeast extract, 2% peptone, 2% sugar, 0.65% agar), YePSS medium (1% yeast extract, 2% peptone, 2% sugar, 18 volumes of haploid sporidia of MAT-1 or MAT-2 were mixed, plated on the solid medium, and kept in dark at 28°C incubator for 2 to 3 days before photographing. For the growth assay, wild-type (MAT-1) and ssagc1Δ mutant S. scitamineum sporidia were first cultured in 5 ml of YePS liquid medium at 28°C with shaking at 200 rpm for 24 h. Such cultured sporidia were then diluted to 200 ml fresh YePS liquid medium, adjusting to a cell density of 10 5 cells per ml, and cultured for another 8 h under the same conditions. Measurement of the optical density at 600 nm (OD600) with a spectrophotometer (NanoDrop 2000C; Thermo Fisher Scientific) was performed hourly to monitor the yeast-like (budding) growth of wild-type or mutant strains, with or without the addition of chemical reagents as described above in Results.
Characterization of the SsAgc1 sequence. The deduced protein sequence of SsAgc1 was analyzed using the Compute pI/M w tool (http://web.expasy.org/compute_pi/) to determine the theoretical isoelectric point (pI) and molecular weight (MW), and interPro-Scan (http://www.ebi.ac.uk/interpro/) to identify domains. A phylogenetic tree was constructed using a neighbor-joining method following alignment with other protein kinases, using COBALT (constraint-based multiple alignment tool; https:// www.ncbi.nlm.nih.gov/tools/cobalt/cobalt.cgi?CMDϭWeb).
Nucleic acid-related manipulation. Fungal genomic DNA was extracted using a HP Fungal DNA kit (catalog no. D3195-01; Omega). PCR amplification was performed using Phusion high-fidelity DNA polymerase (lot 00528748; Thermo Fisher Scientific). DNA fragment elution was performed using Gel Extraction kit (catalog no. D2500-02; Omega) and/or Cycle Pure kit (catalog no. D6492-02; Omega). In the Southern blot assay, NEB restriction enzyme ClaI (R0197V) was used for digestion of genomic DNA. DIG-High Prime DNA Labeling and Detection Starter kit I (catalog no. 11745832910; Roche) was used for labeling of PCR-amplified fragments as probes. Amersham Hybond TM-Nϩ (catalog no. RFN303B; GE Healthcare) membrane was used for blotting. DIG-High Prime DNA Labeling and Detection Starter kit I (catalog no. 11745832910; Roche) was used for labeling of PCR-amplified fragments as probes. NBT/BCIP stock solution (catalog no. 11681451001; Roche) was used for probed band detection. For total RNA extraction, Qiagen RNeasy plant minikit (catalog no. 74104) was used. For removing contaminating DNA from RNA preparations, Ambion TURBO DNA-free kit (catalog no. AM1907; Invitrogen) was used. TransScript First-Strand cDNA Synthesis Super Mix (catalog no. AT301-02; Transgen) was used for cDNA synthesis. For quantitative real-time PCR (qRT-PCR), we used PowerUp SYBR green master mix (catalog no. A25742: Applied Biosystems) with the primers as listed in Table S1 in the supplemental material, and the reaction was run on QuantStudio 6 Flex Real-Time PCR system (Thermo Fisher Scientific).
Generation of deletion or complementation constructs. Construction of two fragments for the replacement of SsAGC1 gene by the HPT (Hyg r ) gene follows the strategy described previously (46,47). The flanking DNA (1 kb 5= and 3=) of the SSAGC1 gene was PCR amplified using WT S. scitamineum genomic DNA as the template, and the HPT gene with plasmid pEX2 (1) as the template. The 5=-and 3=-flanking sequences, fused with two half partially overlapping HPT sequences, were generated by fusion PCR and verified by sequencing. For constructing genetic complementation of the SsAGC1 gene, the SsAGC1 open reading frame (ORF) sequence from ϩ7860 to ϩ14163 (stop codon) was PCR amplified with the forward primer containing a start codon ATG (Table S2) and inserted into our complementation plasmid pEX2-Zeo-dsRed. The resultant plasmid was named EX2-Zeo-AGC1. A constitutive promoter G3PD Prom fragment was PCR amplified and inserted in the pEX2-Zeo-AGC1 plasmid in front of the SsAGC1 ORF to form plasmid pEX2-Zeo-G3PD Prom-AGC1. This complementation plasmid was transformed into the ssagc1Δ-1 mutant, and the complemented strains were first screened by resistance to 100 g/ml zeocin (catalog no. R25001; Invitrogen) and then verified by PCR amplification of zeocin fragment and qRT-PCR analysis of SsAGC1 gene expression. All the primers used for construction or verification of deletion or complementation strains are listed in Table S2.
PEG-mediated protoplast transformation. Polyethylene glycol (PEG)-mediated protoplast transformation follows the established protocol (47) modified as follows. Enzyme digestion of the wild-type MAT-1 sporidia was performed using the lysing enzyme (catalog no. L1412; Sigma) dissolved in SCS solution (20 mM trisodium citrate and 1 M D-sorbitol [pH 5.8]) to reach the final concentration of 0.02 g/ml, incubated at 28°C for about 30 min. A 40% PEG solution was prepared by dissolving 4 g PEG 3350 (catalog no. 202444; Sigma-Aldrich) in 10 ml STC solution (10 mM Tris-HCl, 1 M D-sorbitol, and 100 mM CaCl 2 [pH 7.5]). The PCR-amplified fragments of 1 to 5 g, 1 l heparin solution (15 mg/ml) (Beijing Dingguo, DH157) and the protoplasts were mixed and incubated with 40% PEG solution on ice for 10 min. The protoplasts were regenerated on the three-layer regeneration medium: one layer of YePS soft medium on top of two lays of YePSS medium, with only the bottom YePSS layer containing 400 g/ml hygromycin B (Calbiochem, CAS: 53-84-9).
Staining protocols and microscopy. Calcofluor white (catalog no. 18909; Sigma-Aldrich) was used at 3 g/ml (in 100 mM Tris-HCl buffer [pH 9.0] containing Triton X-100 at 1:1,000) to visualize cell wall and septa of the sporidia of the respective strains. To stain nuclei, S. scitamineum sporidia were incubated in 1 g/ml Hoechst (catalog no. 33258; Invitrogen-Molecular Probes) solution for 15 min at room temperature, followed by a thorough wash with sterile water. Epifluorescence microscopy and imaging were performed using an Axio Observer Z1 microscope (Zeiss, Jena, Germany) equipped with an sCMOS camera (PCO Edge, Kelheim, Germany).
Determination of intracellular cAMP concentration. Sporidia from wild-type ssagc1Δ, and ssuac1Δ strains were cultured in the liquid YePS medium at 28°C, with shaking at 200 rpm, until an optical density at 600 nm of 2.0 was reached. A total of 10 7 sporidia of each strain were aliquoted and streaked on YePSA medium for a further 24-h culture in the dark at 28°C, before they were harvested and ground with liquid nitrogen. The ground samples were subject to quantitative determination of cyclic AMP using a cAMP Enzyme Immunoassay kit (catalog no. CA201; Sigma-Aldrich) following the manufacturer's instructions.
Pathogenicity assay. The susceptible sugarcane variety ROC22 was inoculated with mixed fungal sporidia from WT or mutant combination (of opposite mating type) by injecting seedlings with four to five leaves. The disease symptoms were photographed at 90 to 180 days postinoculation (dpi).
Statistical analysis. Statistical significance of the expression data was determined at P Ͻ 0.01 (**) and P Ͻ 0.05 (*) using Student's t test.
Data availability. All the data present in this article are available in the main body or supplemental material.