Integration of Transcriptomic and Metabolomic Profiles Provides Insights into the Influence of Nitrogen on Secondary Metabolism in Fusarium sacchari

Nitrogen availability might play an essential role in plant diseases by enhancing fungal cell growth and influencing the expression of genes required for successful pathogenesis. Nitrogen availability could modulate secondary metabolic pathways as evidenced by the significant differential expression of several core genes involved in mycotoxin biosynthesis and genes encoding polyketide synthase/nonribosomal peptide synthetases, cytochrome P450 and carbohydrate-active enzymes in Fusarium sacchari, grown on different nitrogen sources. A combined analysis was carried out on the transcript and metabolite profiles of regulatory metabolic processes and the virulence of Fusarium sacchari grown on various nitrogen sources. The nitrogen regulation of the gibberellin gene cluster included the metabolic flux and multiple steps of gibberellin synthesis. UHPLC-MS/MS-based metabolome analysis revealed the coordination of these related transcripts and the accumulation of gibberellin metabolites. This integrated analysis allowed us to uncover additional information for a more comprehensive understanding of biological events relevant to fungal secondary metabolic regulation in response to nitrogen availability.


Introduction
The critical factors affecting fungal growth are nitrogen sources that affect the growth and synthesis of secondary metabolites, including mycotoxins. Nitrogen commonly limits plant growth in terrestrial ecosystems and affects competitive interactions within plantmicrobe communities, including pathogen and symbiotic microorganisms [1]. Although nitrogen fertilizers improve the nutrient status of plant hosts, they can also promote disease by affecting physiological and biochemical balances in plants [2,3]. However, the overuse of nitrogen fertilizers and the inadequate timing of fertilizer application can promote plant disease outbreaks [4].

Transcriptome Analysis
To identify the critical genes in response to nitrogen availability and the production of secondary metabolites in F. sacchari, we performed the transcriptomic analyses of FsCNO-1, including 'urea', 'NaNO 3 ', and '(NH 4 ) 2 SO 4 ' group. We obtained 67,379,846 clean reads, with an average GC content of 52.73%. The raw reads were submitted to the GenBank Sequence Read Archive (SRR18439752, SRR18439753, and SRR18439754). Approximately 94.93% of the clean reads mapped to the F. sacchari CNO-1 genome and more than 93.65% could be uniquely mapped.

Gibberellins Production
FsCNO-1 strains displayed variation in pigmentation three days after inoculation with different nitrogen treatments ( Figure 1). We investigated the ability of the FsCNO-1 to produce gibberellins (GA7, GA3, and GA4 were the main gibberellins). GA7 and GA4 accumulated up to 10,000 and 6000 µg/L after culturing for three days in (NH 4 ) 2 SO 4 and NaNO 3 culture conditions. Meanwhile, cultures grown with urea had GA7 and GA4 levels at 600 µg/L. GA3 produced about 3-fold more than GA4 and GA7 when cultured in the urea, whereas GA7 was approximately 2.0-fold more than what GA3 produced when cultured in (NH 4 ) 2 SO 4 ( Figure 1).

Genes Associated with GA Production in FsCNO-1
The GA gene cluster contains a total of seven functional genes, including DES (FSAC_09644), P450-4 (FSAC_12979), P450-1 (FSAC_10060), P450-2 (FSAC_08504), GGS2 (FSAC_05737), CPS/KS (FSAC_08126), and P450-3 (FSAC_03987) which were present in FsCNO-1. In addition, genes encoding the ankyrin (ank), sugar membrane transporter (smt), alcohol (alc-DH), and aldehyde dehydrogenase (ald-DH) were located upstream of the GA cluster ( Figure 2A). Transcriptome data revealed that the expression of genes within and flanking the GA cluster had high transcriptional plasticity in response to nitrogen availability ( Figure 2B). We performed qRT-PCR to assess the expression of GA cluster genes and found that the calculated fold changes were consistent with those determined by RNA-seq ( Figure 2C).
Noticeably, growth in media having urea as the nitrogen source completely repressed the expression of the GA cluster core genes, which had an overall effect of decreasing the pool of GA cluster transcripts, limiting GA metabolic flux. In particular, genes encoding cytochrome P450 monooxygenase (P450-3), geranylgeranyl diphosphate synthase (GGS2), and/or ent-copalyl diphosphate/ent-kaurene synthase (CPS/KS), cytochrome P450 monooxygenase (P450-2), GA14-synthase (P450-1), and GA4 desaturase (DES) showed significantly higher transcript levels on either (NH4)2SO4 or NaNO3, as compared to urea. In addition, several intermediates and final biosynthetic steps are catalyzed by P450-2 and DES ( Figure 2D) [14]. Although the P450-3 gene expression was accompanied by relatively high levels of GA3 production in (NH4)2SO4 media, only low production levels were seen for growth on NaNO3. Our results show that the total GA production largely corresponded with the gene expression levels. Six out of seven GA genes in FsCNO-1 were regulated by nitrogen, except for ent-kaurene oxidase (P450-4), which showed only faint expression and remained low (fold change < 1.0) regardless of the nitrogen source. However, the expression levels of seven GA-flanking genes were highly variable ( Figure 2B). Genes flanking the GA cluster are often regulated by a transcriptional regulator in the cluster itself that could co-regulate the transcription of cluster genes [15]. However, our data could not define how the expression of GA-flanking genes is affected by the nitrogen source or whether these genes coordinate the expression of GA cluster genes during GA synthesis in response to specific nitrogen signals.

Genes Associated with GA Production in FsCNO-1
The GA gene cluster contains a total of seven functional genes, including DES (FSAC_09644), P450-4 (FSAC_12979), P450-1 (FSAC_10060), P450-2 (FSAC_08504), GGS2 (FSAC_05737), CPS/KS (FSAC_08126), and P450-3 (FSAC_03987) which were present in FsCNO-1. In addition, genes encoding the ankyrin (ank), sugar membrane transporter (smt), alcohol (alc-DH), and aldehyde dehydrogenase (ald-DH) were located upstream of the GA cluster ( Figure 2A). Transcriptome data revealed that the expression of genes within and flanking the GA cluster had high transcriptional plasticity in response to nitrogen availability ( Figure 2B). We performed qRT-PCR to assess the expression of GA cluster genes and found that the calculated fold changes were consistent with those determined by RNA-seq ( Figure 2C).   Noticeably, growth in media having urea as the nitrogen source completely repressed the expression of the GA cluster core genes, which had an overall effect of decreasing the pool of GA cluster transcripts, limiting GA metabolic flux. In particular, genes encoding cytochrome P450 monooxygenase (P450-3), geranylgeranyl diphosphate synthase (GGS2), and/or ent-copalyl diphosphate/ent-kaurene synthase (CPS/KS), cytochrome P450 monooxygenase (P450-2), GA14-synthase (P450-1), and GA4 desaturase (DES) showed significantly higher transcript levels on either (NH 4 ) 2 SO 4 or NaNO 3 , as compared to urea. In addition, several intermediates and final biosynthetic steps are catalyzed by P450-2 and DES ( Figure 2D) [14]. Although the P450-3 gene expression was accompanied by relatively high levels of GA3 production in (NH 4 ) 2 SO 4 media, only low production levels were seen for growth on NaNO 3 . Our results show that the total GA production largely corresponded with the gene expression levels. Six out of seven GA genes in FsCNO-1 were regulated by nitrogen, except for ent-kaurene oxidase (P450-4), which showed only faint expression and remained low (fold change < 1.0) regardless of the nitrogen source. However, the expression levels of seven GA-flanking genes were highly variable ( Figure 2B). Genes flanking the GA cluster are often regulated by a transcriptional regulator in the cluster itself that could co-regulate the transcription of cluster genes [15]. However, our data could not define how the expression of GA-flanking genes is affected by the nitrogen source or whether these genes coordinate the expression of GA cluster genes during GA synthesis in response to specific nitrogen signals.

Genes Related to Secondary Metabolite Production in FsCNO-1
Mycotoxin synthesis. To enhance our understanding of secondary metabolism events associated with the nitrogen regulation of biosynthetic gene clusters, we analyzed FSR, BIK, FUB, FUS, and FUM biosynthetic gene clusters related to mycotoxin synthesis. FSR mediates the biosynthesis of fusarubin, BIK for bikaverin, FUB for fusaric acid, FUS for fusarin C, and FUM for fumonisin. We identified complete FSR, BIK, FUS, and FUB biosynthetic gene clusters in FsCNO-1 genomes, lacking a complete FUM gene cluster ( Figure 3A).   Based on the clustering analysis, bikaverin, and fusarubin-associated genes were expressed to significantly higher levels in (NH 4 ) 2 SO 4 media relative to urea or NaNO 3 media. Specifically, the genes were related to bikaverin synthesis, including FSAC_04364 (polyketide synthase), FSAC_09746 (hypothetical protein), FSAC_00970 (O-methyltransferase), FSAC_03389 (transcription factor enhancer), FSAC_09877 (transcription factor), and FSAC_11009 (efflux pump), and showed a markedly induced expression in (NH 4 ) 2 SO 4 media. Moreover, the pattern of the fusarubin synthesis is similar. The fusarubin biosynthetic gene cluster mainly contains six functional genes (FSAC_04059, FSAC_04742, FSAC_03718, FSAC_09095, FSAC_12585, and FSAC_09765), and most genes also showed an up-regulated expression in (NH 4 ) 2 SO 4 media ( Figure 3B). Significantly, in contrast to the utterly repressed expression pattern of GA genes with urea as the nitrogen source, the expression of genes associated with fusarin C (FSAC_14605, FSAC_10601, FSAC_07424, FSAC_09464, FSAC_09939, FSAC_10936, FSAC_12527, FSAC_08908, and FSAC_10763) and fusaric acid (FSAC_02491, FSAC_08022, FSAC_08313, and FSAC_02703) production was distinctly activated on urea ( Figure 3B). Furthermore, fusaric acid-related gene expression was obviously higher than other mycotoxins ( Table 1).  (Table S2). A prediction tool identified a complete set of conserved domains comprising each PKS or NRPS gene ( Figure 4). Based on Clustal W v2.0 and MEGA v7.0 software, a phylogenetic tree of the PKSs and NRPSs protein sequences was performed using the neighbor-joining (NJ) method. In addition, in the expression of PKSs and NRPS genes, significant variation was detected for the FsCNO-1 transcriptome. Only two putative PKS genes (FSAC_10998 and FSAC_10909) were silent and could not express, and the other twenty-eight genes showed distinctly different expression patterns for growth on different nitrogen sources (Figure 4). , x FOR PEER REVIEW 6 of 14 PKS/NRPS. Sixteen putative PKS genes, fourteen NRPS genes, and four PKS-NRPS hybrid genes were identified in FsCNO-1 (Table S2). A prediction tool identified a complete set of conserved domains comprising each PKS or NRPS gene ( Figure 4). Based on Clustal W v2.0 and MEGA v7.0 software, a phylogenetic tree of the PKSs and NRPSs protein sequences was performed using the neighbor-joining (NJ) method. In addition, in the expression of PKSs and NRPS genes, significant variation was detected for the FsCNO-1 transcriptome. Only two putative PKS genes (FSAC_10998 and FSAC_10909) were silent and could not express, and the other twenty-eight genes showed distinctly different expression patterns for growth on different nitrogen sources (Figure 4). Cytochrome P450. Cytochrome P450 is a member of the CYP superfamily of hemecontaining monooxygenases involved in the production and/or modification of diverse compounds [16]. We performed a batch BLAST search against the InterPro database and identified 156 putative cytochrome P450-coding genes in FsCNO-1 (Table S3). Among these genes, 111 genes had (log2FC > 1.0, p < 0.05) different expressions according to our RNA-Seq analysis, which suggested that nitrogen availability could influence the expression of most cytochrome P450 genes. Among these DEGs, the transcription levels of most DEGs were significantly down-regulated in NaNO3 media, whereas many DEGs were upregulated in (NH4)2SO4 and urea media ( Figure S3).
Cytochrome P450 has a multifunctional role in fungi-like secondary, primary, xenobiotic, and drug metabolism. RNA-seq analysis was performed to examine changes in the expression of several representative P450 genes with known functions. As expected, a series of P450-encoding genes were assigned to the secondary metabolite class, involved in the biosynthesis of trichothecene and sterigmatocystin. Interestingly, cytochrome P450-Pdm (FSAC_01874, FSAC_10736, FSAC_07317, and FSAC_13209) plays a defensive role against factors secreted by the host and increases fungal pathogenicity, indicating a significant expression induction when (NH4)2SO4 or urea was used as the nitrogen source Cytochrome P450. Cytochrome P450 is a member of the CYP superfamily of hemecontaining monooxygenases involved in the production and/or modification of diverse compounds [16]. We performed a batch BLAST search against the InterPro database and identified 156 putative cytochrome P450-coding genes in FsCNO-1 (Table S3). Among these genes, 111 genes had (log2FC > 1.0, p < 0.05) different expressions according to our RNA-Seq analysis, which suggested that nitrogen availability could influence the expression of most cytochrome P450 genes. Among these DEGs, the transcription levels of most DEGs were significantly down-regulated in NaNO 3 media, whereas many DEGs were up-regulated in (NH 4 ) 2 SO 4 and urea media ( Figure S3).
Cytochrome P450 has a multifunctional role in fungi-like secondary, primary, xenobiotic, and drug metabolism. RNA-seq analysis was performed to examine changes in the expression of several representative P450 genes with known functions. As expected, a series of P450-encoding genes were assigned to the secondary metabolite class, involved in the biosynthesis of trichothecene and sterigmatocystin. Interestingly, cytochrome P450-Pdm (FSAC_01874, FSAC_10736, FSAC_07317, and FSAC_13209) plays a defensive role against factors secreted by the host and increases fungal pathogenicity, indicating a significant expression induction when (NH 4 ) 2 SO 4 or urea was used as the nitrogen source (Table 2). CAZyme families. CAZyme families in fungi orchestrate the bioactivity and metabolism of small molecules that comprise various glycosylated compounds (e.g., glycolipids, glycoproteins, lignin precursors, and secondary metabolites), including glycoside hydrolases (GHs), glycosyl transferases (GTs), polysaccharide lyases (PLs), carbohydrate esterases (CEs), auxiliary activities (AAs), and carbohydrate-binding modules (CBMs) [17]. A total of 786 CAZY genes were detected in FsCNO-1, including 315 GHs (315), 174 CEs (174), 120 AAs, 109 GTs, 48 CBMs, and 20 PLs.
Moreover, 93 CAZy genes were differentially expressed among the nitrogen treatments, including 39 GHs, 11 GTs, 20 AAs, 6 CBMs, 16 CEs, and 1 PLs ( Figure 5A). Next, we performed hierarchical clustering and heat map analysis, which found significant differences in FsCNO-1 grown on urea ( Figure 5B). A large proportion of alpha-glucosidase, chitinase, and galactinol synthase genes could be induced with a high expression of growth on urea when compared with NaNO 3 or (NH 4 ) 2 SO 4 . Overall, the type of nitrogen source significantly affected the expression of the prominent CAZy families in our studies and was mainly involved in regulating the biological processes related to SMs in Fusarium, such as mycotoxin, PKS/NRPS, and cytochrome P450.

Discussion
Pokkah boeng disease has spread rapidly in China during all seasons in recent years. Guangxi, the most extensive sugarcane planting and production area in China, was first reported to be seriously affected due to PBD [18]. Here, we performed the RNA-seq and UHPLC-MS/MS analysis of FsCNO-1, which are reported to be the significant representative pathogens of PBD [7]. Nitrogen was associated with F. sacchari cell growth speed and plant disease severity. The expression of genes required for successful pathogenesis was affected, such as those involved in nitrogen assimilation, nitrogen sensing and regulation, and potential virulence-and pathogenicity-associated functions. Many genes associated with secondary metabolism were also affected by the nitrogen source [13]. Several regulators were found to be associated with specific biological processes related to nitrogen utilization [19][20][21]. The genes associated with secondary metabolite synthesis in FsCNO-1 were investigated in silico. Nitrogen availability could modulate secondary metabolic pathways as evidenced by the significant differential expression of cluster genes involved in mycotoxin biosynthesis and genes encoding polyketide synthase/nonribosomal peptide synthetases, cytochrome P450, and carbohydrate-active enzymes in F. sacchari grown on different nitrogen sources. The nitrogen regulation of the gibberellin gene cluster included metabolic flux and multiple steps of gibberellin synthesis. RNA-seq and UHPLC-MS/MS analysis revealed the coordination of these related transcripts and the accumulation of gibberellin metabolites. For sugarcane, the pathogenic FsCNO-1 of PBD resulted in the significant inhibition of leaf NR, GS, and GOGAT activities, decreased NO 3− -N and NH 4+ -N assimilation rates, and continued to promote RNA degradation, thereby reducing ribosome and protein synthesis. Glutamine plays an important protective role by regulating the biosynthesis of proline and polyamines. The metabolic pathways of cyanamide, glutamic acid, proline, tyrosine, and arachidonic acid actively respond to the stress of pathogens [22]. The results provide a scientific basis for the formulation of technical regulations for the rational fertilization of sugarcane ( Figure 6).

Discussion
Pokkah boeng disease has spread rapidly in China during all seasons in recent years. Guangxi, the most extensive sugarcane planting and production area in China, was first reported to be seriously affected due to PBD [18]. Here, we performed the RNA-seq and UHPLC-MS/MS analysis of FsCNO-1, which are reported to be the significant representative pathogens of PBD [7]. Nitrogen was associated with F. sacchari cell growth speed and plant disease severity. The expression of genes required for successful pathogenesis was affected, such as those involved in nitrogen assimilation, nitrogen sensing and regulation, and potential virulence-and pathogenicity-associated functions. Many genes associated with secondary metabolism were also affected by the nitrogen source [13]. Several regulators were found to be associated with specific biological processes related to nitrogen utilization [19][20][21]. The genes associated with secondary metabolite synthesis in FsCNO-1 were investigated in silico. Nitrogen availability could modulate secondary metabolic pathways as evidenced by the significant differential expression of cluster genes involved in mycotoxin biosynthesis and genes encoding polyketide synthase/nonribosomal peptide synthetases, cytochrome P450, and carbohydrate-active enzymes in F. sacchari grown on different nitrogen sources. The nitrogen regulation of the gibberellin gene cluster included metabolic flux and multiple steps of gibberellin synthesis. RNA-seq and UHPLC-MS/MS analysis revealed the coordination of these related transcripts and the accumulation of gibberellin metabolites. For sugarcane, the pathogenic FsCNO-1 of PBD resulted in the significant inhibition of leaf NR, GS, and GOGAT activities, decreased NO 3− -N and NH 4+ -N assimilation rates, and continued to promote RNA degradation, thereby reducing ribosome and protein synthesis. Glutamine plays an important protective role by regulating the biosynthesis of proline and polyamines. The metabolic pathways of cyanamide, glutamic acid, proline, tyrosine, and arachidonic acid actively respond to the stress of pathogens [22]. The results provide a scientific basis for the formulation of technical regulations for the rational fertilization of sugarcane ( Figure 6). FsCNO-1 cultured in different nitrogen sources showed differences in the production of pigments associated with secondary metabolites. However, the detailed mechanisms still need to be clarified. Notably, many genes central to producing fusarubin, bikaverin, fusaric acid, and fusarin C had an altered expression according to the different nitrogen sources, which strongly implied the existence of specific fungal metabolites that might play a role in mycotoxin production. Previous studies revealed a correlation between nitrogen availability and production of other SMs (e.g., gibberellins, mycotoxin, and CAZYmes, etc.) by Fusarium (Figure 6), such as that arising from carotenoids, the PKSderived mycelial reddish pigments bikaverin, and fusarubins [9], and the brown or black melanin pigments [23]. In this study, the results showed that the appearance of the reddish pigment observed in cultured ammonium as the sole nitrogen source, was coincident with the ammonium-induced abundant expression of BIK in FsCNO-1.
The expression profile of CAZyme family genes suggested that the expression of AA and CE genes induced by growth on ammonium could preferentially contribute to fungal pathogenicity and fungal biomass production, as consistent with our previous study [13]. A diverse arsenal of CAZymes might thus be needed to access the plant xylem and to breach plant defense structures (e.g., xyloses and pectin gels) [17,24]. Genes encoding cytochrome P450 appear in gene clusters with neighboring PKS or mycotoxin genes and encode proteins that can modify or customize the polyketide backbone to form functional secondary metabolites [25].
Fusarium can generate a vast array of bioactive secondary metabolites, which are linked to the biosynthetic genes responsible for their production [26], such as PKS and NRPS, although many of these genes have unknown products [27,28]. The PKS and NRPS genes provide a basis for further studies to identify genes and pathways in F. sacchari. From an evolutionary perspective, the origin of fungal gibberellin biosynthesis is independent of that in plants [29]. The earlier speculation was that fungi acquired GA biosyn- FsCNO-1 cultured in different nitrogen sources showed differences in the production of pigments associated with secondary metabolites. However, the detailed mechanisms still need to be clarified. Notably, many genes central to producing fusarubin, bikaverin, fusaric acid, and fusarin C had an altered expression according to the different nitrogen sources, which strongly implied the existence of specific fungal metabolites that might play a role in mycotoxin production. Previous studies revealed a correlation between nitrogen availability and production of other SMs (e.g., gibberellins, mycotoxin, and CAZYmes, etc.) by Fusarium (Figure 6), such as that arising from carotenoids, the PKS-derived mycelial reddish pigments bikaverin, and fusarubins [9], and the brown or black melanin pigments [23]. In this study, the results showed that the appearance of the reddish pigment observed in cultured ammonium as the sole nitrogen source, was coincident with the ammonium-induced abundant expression of BIK in FsCNO-1.
The expression profile of CAZyme family genes suggested that the expression of AA and CE genes induced by growth on ammonium could preferentially contribute to fungal pathogenicity and fungal biomass production, as consistent with our previous study [13]. A diverse arsenal of CAZymes might thus be needed to access the plant xylem and to breach plant defense structures (e.g., xyloses and pectin gels) [17,24]. Genes encoding cytochrome P450 appear in gene clusters with neighboring PKS or mycotoxin genes and encode proteins that can modify or customize the polyketide backbone to form functional secondary metabolites [25].
Fusarium can generate a vast array of bioactive secondary metabolites, which are linked to the biosynthetic genes responsible for their production [26], such as PKS and NRPS, although many of these genes have unknown products [27,28]. The PKS and NRPS genes provide a basis for further studies to identify genes and pathways in F. sacchari. From an evolutionary perspective, the origin of fungal gibberellin biosynthesis is independent of that in plants [29]. The earlier speculation was that fungi acquired GA biosynthetic genes from higher plants via horizontal gene transfer [30]. However, the function of gibberellins in fungi is less well understood, although they can influence the growth and development of their host plants [31]. During the infection of the host plants, gibberellins contribute to the ability of F. fujikuroi to grow invasively in the symplasts of the parenchyma cells of the rice epidermis to confer a selective advantage [9]. Thus, the symbiosis of gibberellins produced by endophytic fungi could be a promising strategy to overcome the adverse effects of abiotic stresses on plants [32]. In the current study, several intermediates and final biosynthetic steps are catalyzed by P450-2 and DES (i.e., synthesis of GA9, GA25, GA12, GA13, or GA7), such that the relative accumulation of GA4 and GA7 in NaNO 3 or (NH 4 ) 2 SO 4 media could be due to the increased expression of these two genes. However, the up-regulation of a given GA gene was not always associated with the high production levels of the corresponding GAs. For example, NaNO 3 or (NH 4 ) 2 SO 4 media induced the high-level expression of 13-hydroxylase (P450-3), which catalyzes the hydroxylation on C-13 of GA4 or GA7 to GA3 in the last steps of the GA biosynthetic pathway [14]. Additionally, the gibberellin gene cluster expression was repressed by growth in urea relative to growth on NaNO 3 or (NH 4 ) 2 SO 4 . This repression could be associated with a dramatic decrease in the metabolic flux of gibberellins, a metabolic bottleneck that occurs downstream of GA4 and GA7 formation and upstream of ent-kaurene formation. This metabolic bottleneck is due, at least in part, to the nitrogen-dependent regulation of GA gene cluster expression. Accumulating evidence indicates that the production of bioactive gibberellins in fungi is tightly regulated by developmental and environmental cues [33]. Nevertheless, the modulation of the expression of the entire gibberellin gene cluster, including the bilateral flanking genes, is complex and requires additional study. Global regulatory proteins and miRNAs are also required to regulate secondary metabolite production [28].
Nitrogen fertilizers are commonly applied in farming to increase crop yield, but they also modify the ability of several plants to resist pathogen infections. Nitrogen availability might play an essential role in plant diseases by enhancing fungal cell growth and influencing the expression of genes required for successful pathogenesis. In this study, these integrated analyses allow us to uncover additional information for a more comprehensive understanding of biological events relevant to fungal secondary metabolic regulation in response to nitrogen availability. These findings can thus provide essential information for Fusarium management in the field, although further experiments are necessary to confirm whether such a scenario exists for F. sacchari that causes PBD in sugarcane.

Fungal Strain and Culture Conditions
F. sacchari CNO-1 (FsCNO-1) was isolated in 2012 from sugarcane plants exhibiting PBD symptoms in Guangxi, China. The isolate was maintained on a slant medium containing potato dextrose agar (PDA) at 4 • C, until use. FsCNO-1 spores (1 mL, 1.0 × 10 6 conidia/mL) were inoculated into 200 mL Erlenmeyer flasks containing 60 mL Czapek medium (per liter: K 2 HPO 4 1 g, MgSO 4 ·7H 2 O 0.5 g, KCl 0.5 g, Fe 2 SO 4 0.01 g, Sucrose 30 g) supplemented with 3 g/L of NaNO 3 , (NH 4 ) 2 SO 4 or urea as the nitrogen source. The flasks were shaken on a rotary shaker at 200 rpm and 28 • C for three days. Mycelia from the different nitrogen cultures were harvested by centrifugation at 8000 rpm for 15 min at 4 • C for the immediate extraction of total RNA.

Transcriptomic Analysis
Total RNA was extracted from FsCNO-1 grown in the three nitrogen mediums (NaNO 3 , (NH 4 ) 2 SO 4 , and urea) using RNAprep Pure plant Kit (Tiangen Biotech, Beijing, China). The quality and quantity of RNA were assessed using Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Libraries were produced using an Illumina TruSeq RNA Sample Prep Kit (Illumina, San Diego, CA, USA). The library products were sequenced using the Illumina Hiseq 2500 platform in Majorbio company (Majorbio, Shanghai, China).

Quantitative RT-PCR
Total RNA was extracted and reverse-transcribed in a 20 µL reaction volume using the PrimeScript TM RT reagent Kit with gDNA Eraser (Takara, Dalian, China). qRT-PCR analysis was performed three times for each extracted RNA sample using SYBR ® Premix Ex Taq™ (Tli RNaseH Plus) (Takara, Dalian, China) according to the manufacturer's protocol. PCR amplifications were performed using a Roche Light Cycler 480 (Indianapolis, IN, USA). cDNA was synthesized from each sample, and RNA levels were normalized to β-actin levels of FsCNO-1. A relative quantitative method (2 −∆∆Ct ) was used to evaluate the quantitative variation [45]. Primer sequences are listed in Table S4.

Extraction and Analysis of Gibberellins
Gibberellic acids were extracted from 200 mL culture filtrate after culturing for three days using ion exchange resins as previously described [46]. The experiments were carried out using a static method wherein 5 g D301G and D201I exchange resin (NKU, Tianjin, China) were added to 200 mL culture filtrate at pH 2.5 and 3.5, respectively. The exchange resins were shaken for 10 h to achieve saturation and improve gibberellic acid extraction. The gibberellic acid was then eluted with 10 mL acetone and dried in a Vacufuge concentrator (Eppendorf, Hamburg, Germany). The residue was dissolved in HPLC-grade methanol for UHPLC-MS/MS analysis. A 1290 series UHPLC system coupled with a 6460 Triple Quadrupole (QqQ) mass spectrometer (Agilent Technologies, Waldbronn, Germany) was used for gibberellin analysis. An aliquot (1.0 µL) of each solution containing gibberellins was injected into the UHPLC equipped with a Hypersil GOLD C18 column (1.9 µm, 2.1 × 100 mm, Thermo Scientific, Waltham, MA, USA). For MS/MS detection, the ionization source parameters in positive and negative modes were the same as that described in the literature [47]. A linear gradient from 20% to 100% acetonitrile in 10 mM HCOOH over 30 min was used for elution. The flow rate was 0.3 mL/min. The extracted samples and standard mixture of gibberellins (GA3, GA4, and GA7, Sigma-Aldrich, St. Louis, MO, USA) were injected under the same conditions for UHPLC analysis. Each run was repeated three times, and the detector response was measured in the peak area.

Conclusions
In summary, integrated transcriptome and UHPLC-MS/MS metabolite analysis indicated that nitrogen affected multiple steps in the gibberellin synthesis and metabolic flux to impact the net and relative accumulation of gibberellic acid compounds. Additionally, the transcriptome results illuminated the potential effects of nitrogen availability on the expression of several genes involved in fungal toxin synthesis or that encode PKS/NRPS, cytochrome P450, and carbohydrate-active enzymes, which expanded our knowledge of the role that nitrogen sources play in both secondary metabolite biosynthesis.  Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study were included in this published article and its Supplementary Information files. Further inquiries can be directed to the corresponding author.

Conflicts of Interest:
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.