Identification of Genes Involved in the Synthesis of the Fungal Cell Wall Component Nigeran and Regulation of Its Polymerization in Aspergillus luchuensis

ABSTRACT

Certain Aspergillus and Penicillium spp. produce the fungal cell wall component nigeran, an unbranched d-glucan with alternating α-1,3- and α-1,4-glucoside linkages, under nitrogen starvation. The mechanism underlying nigeran biosynthesis and the physiological role of nigeran in fungal survival are not clear. We used RNA sequencing (RNA-seq) to identify genes involved in nigeran synthesis in the filamentous fungus Aspergillus luchuensis when grown under nitrogen-free conditions. agsB, which encodes a putative α-1,3-glucan synthase, and two adjacent genes (agtC and gnsA) were upregulated under conditions of nitrogen starvation. Disruption of agsB in A. luchuensis (ΔagsB) resulted in the complete loss of nigeran synthesis. Furthermore, the overexpression of agsB in an Aspergillus oryzae strain that cannot produce nigeran resulted in nigeran synthesis. These results indicated that agsB encodes a nigeran synthase. Therefore, we have renamed the A. luchuensis agsB gene the nigeran synthase gene (nisA). Nigeran synthesis in an agtC mutant (ΔagtC) increased to 121%; conversely, those in the ΔgnsA and ΔagtC ΔgnsA strains decreased to 64% and 63%, respectively, compared to that in the wild-type strain. Our results revealed that AgtC and GnsA play an important role in regulating not only the quantity of nigeran but also its polymerization. Collectively, our results demonstrated that nisA (agsB) is essential for nigeran synthesis in A. luchuensis, whereas agtC and gnsA contribute to the regulation of nigeran synthesis and its polymerization. This research provides insights into fungal cell wall biosynthesis, specifically the molecular evolution of fungal α-glucan synthase genes and the potential utilization of nigeran as a novel biopolymer.

IMPORTANCE The fungal cell wall is composed mainly of polysaccharides. Under nitrogen-free conditions, some Aspergillus and Penicillium spp. produce significant levels of nigeran, a fungal cell wall polysaccharide composed of alternating α-1,3/1,4-glucosidic linkages. The mechanisms regulating the biosynthesis and function of nigeran are unknown. Here, we performed RNA sequencing of Aspergillus luchuensis cultured under nitrogen-free or low-nitrogen conditions. A putative α-1,3-glucan synthase gene, whose transcriptional level was upregulated under nitrogen-free conditions, was demonstrated to encode nigeran synthase. Furthermore, two genes encoding an α-glucanotransferase and a hypothetical protein were shown to be involved in controlling the nigeran content and molecular weight. This study reveals genes involved in the synthesis of nigeran, a potential biopolymer, and provides a deeper understanding of fungal cell wall biosynthesis.

INTRODUCTION

Nigeran, an unbranched d-glucan consisting of alternating α-1,3- and α-1,4-glucoside linkages, was first identified in the fungal cell wall of Penicillium expansum and Aspergillus niger (1, 2). Additional species of Penicillium and Aspergillus have since been shown to synthesize nigeran, whereas the model filamentous fungus Aspergillus nidulans and the pathogenic filamentous fungus A. flavus do not synthesize nigeran (3). Gold et al. reported that the accumulation of nigeran in Aspergillus hyphae is significantly activated under nitrogen-starved conditions (4). Thus, nigeran is not thought to be a key component of fungal cell walls, at least in the presence of nitrogen sources. Unlike other major cell wall polysaccharides, such as β-1,3-glucan, chitin, and α-1,3-glucan, neither the physiological function nor the biosynthetic mechanism of nigeran is known. Nigeran and α-1,3-glucan are classified into the α-glucan group of fungal cell wall polysaccharides (5). Fungal α-1,3-glucans have been widely studied to determine their role in cell morphology, cell wall integrity, and virulence. Although α-1,3-glucan is not essential for the survival of most filamentous fungi, including Aspergillus spp., it plays a key role in the aggregation of conidia and the formation of hyphal pellets in shaking cultures (6, 7). The genome of Aspergillus fumigatus contains three α-1,3-glucan synthase (AGS) genes. A strain in which all three AGS genes were disrupted did not produce α-1,3-glucan and exhibited wild-type growth; however, it was less pathogenic (8). At least five AGS genes have been annotated in Aspergillus section Nigri, such as A. luchuensis, A. niger, and A. aculeatus (9–11). Damveld et al. reported that in A. niger, the expression of agsA and agsE was upregulated by cell wall stress-inducing compounds such as calcofluor white (12). We previously revealed that AgsE (ortholog of A. niger AgsE) functions as a major synthesizer of α-1,3-glucan in A. luchuensis (13). In contrast, the gene encoding nigeran synthase has not been identified, and the physiological functions of nigeran remain unclear.

Aspergillus luchuensis, a member of Aspergillus section Nigri and used to brew traditional Japanese distilled liquor, produces a sufficient amount of nigeran. In this study, we performed RNA sequencing (RNA-seq) analysis of A. luchuensis grown under nitrogen-free or low-nitrogen conditions to identify the gene encoding nigeran synthase and characterize the mechanism of nigeran synthesis. One of the genes annotated as the putative α-1,3-glucan synthase (agsB) in A. luchuensis encodes nigeran synthase. In the model fungus A. nidulans, agsB encodes the major α-1,3-glucan synthase (7). To avoid misunderstanding in subsequent studies on α-1,3-glucan and nigeran, we propose that the A. luchuensis agsB gene should be named the nigeran synthase gene (nisA). These findings expand our understanding of fungal cell walls, including the molecular evolution of fungal α-glucan synthase genes, the development of antifungal agents, and the utilization of nigeran as a novel biopolymer.

RESULTS

Screening for nigeran synthesis genes by RNA-seq transcriptome analysis.

To identify the genes encoding nigeran biosynthetic enzymes, we sequenced the transcriptome of A. luchuensis, which was cultured in low-nitrogen and nitrogen-free media. A total of 366 genes were differentially expressed, among which 198 genes exhibited higher transcript levels in cells cultured in nitrogen-free medium than in cells cultured in low-nitrogen medium. Differentially expressed genes involved in the synthesis or degradation of cell wall polysaccharides or nitrogen metabolism are summarized in Table S1 in the supplemental material. Although the genes encoding cell wall polysaccharide synthases were not included in the top 50 differentially expressed genes, a gene annotated as a putative α-1,3-glucan synthase (RIB2604_01100220, referred to as agsB [nisA]) was found among the other 148 upregulated genes (log fold change [logFC] of 2.90) (Table S1). A. luchuensis agsB (nisA) is an ortholog of the A. niger α-1,3-glucan synthase gene agsB and exhibits 89% nucleotide sequence identity and 92% amino acid sequence identity with it. A. luchuensis agsB (nisA) comprises an extracellular amylase domain, an intracellular glycosyltransferase domain, and a C-terminal transmembrane domain. Furthermore, the expression levels of two genes adjacent to agsB (nisA) were elevated under nitrogen-free conditions. One of these genes, agtC, annotated as a fungal α-glucanotransferase (RIB2604_01100210; orthologous gene of A. niger α-glucanotransferase C), comprised an N-terminal signal peptide, an α-amylase-like domain, and a C-terminal glycosylphosphatidylinositol (GPI)-anchoring site (logFC of 3.38) (Table S1). A. luchuensis AgtC exhibits 82% amino acid sequence identity with A. niger AgtC. The second adjacent gene, annotated as hypothetical protein RIB2604_01100200, comprised an N-terminal signal peptide, a region of unknown function, a Ser/Thr-rich region, and a C-terminal GPI-anchoring site, which we designated gnsA (GPI-anchored nigeran synthase-related enzyme) (logFC of 4.83) (Table S1). A. luchuensis agsB (nisA) and agtC-gnsA are located on opposite chromosomal strands, suggesting that these genes share upstream regulatory elements. The loci of the corresponding genes were conserved in the A. niger genome. For α-1,3-glucan synthesis in Aspergillus spp., most major and some minor AGS genes form a cluster with genes encoding α-glucanotransferase or α-amylase-like protein, which exhibit transglycosylation activity (14). Therefore, at least agsB (nisA) and agtC, and perhaps gnsA, appear to be related to nigeran synthesis in A. luchuensis.

To further examine the expression of the three candidate nigeran synthesis genes, we performed reverse transcription-quantitative PCR (RT-qPCR) to measure changes in the transcript levels over time in A. luchuensis cells cultured in nitrogen-free medium (Fig. 1). Before the induction of nitrogen starvation (0 h), the expression levels of all genes were barely detectable compared to that of the housekeeping gene β-actin. After 24 h of induction, the gene expression levels of agsB (nisA), agtC, and gnsA significantly increased by 0.55-, 0.68-, and 0.84-fold (P < 0.01), respectively, relative to that of β-actin. The increased mRNA levels were sustained until 72 h. The transcript levels of each gene were suppressed under low-nitrogen conditions compared to that of β-actin (<0.05 relative mRNA level). These results indicated that the expression of agsB (nisA), agtC, and gnsA in A. luchuensis was strictly repressed in the presence of nitrogen and significantly activated after eliminating nitrogen from the medium.

FIG 1.

Transcript levels of three nigeran synthesis-related candidate genes in the A. luchuensis wild-type strain under nitrogen-free conditions. A. luchuensis mycelium cultured in low-nitrogen medium was transferred to nitrogen-free medium, and total RNA was extracted for RT-qPCR analysis 0, 24, 48, and 72 h after induction. The mRNA levels of nisA (agsB) (A), agtC (B), and gnsA (C) are shown compared to the levels of β-actin mRNA as a control. Values represent the means from three independent quantification tests with triplicate samples at each time ± standard deviations (SD). Dunnett’s test was performed to compare the means of the nitrogen-free groups with the means from the 0-h time point (*, P < 0.01).

Construction of agsB (nisA), agtC, gnsA, and agtC-gnsA mutants and their effect on nigeran synthesis.

agsB (nisA), agtC, and gnsA were predicted to form a gene cluster and to be involved in the synthesis of nigeran in A. luchuensis (Fig. 2A). Hence, we constructed mutants of these genes, ΔagsB (ΔnisA), ΔagtC, and ΔgnsA, as well as the ΔagtC ΔgnsA double mutant, by homologous recombination (Fig. 2B and Fig. S1A). The A. luchuensis ΔligD (wild-type) strain was transformed with the disruption cassette using the agrobacterium-mediated transformation method to replace each gene with an hph marker cassette. A. luchuensis transformants were screened for hygromycin resistance and selected for colony PCR testing (Fig. S1B). The generation of the ΔagsB (ΔnisA), ΔagtC, ΔgnsA, and ΔagtC ΔgnsA mutant strains was further confirmed by Southern blotting (Fig. 2B), which confirmed gene disruption by homologous recombination at each locus. The hyphal morphology, conidiation, and colony diameter of the disruptants on potato dextrose (PD) and Czapek-Dox (CD) plates were similar to those of the wild-type strain (Fig. 2C). No significant differences in sensitivity to cell wall stress-inducing compounds such as micafungin, calcofluor white, and Congo red were observed in the mutants compared to the wild-type strain (Fig. 2C).

FIG 2.

Construction and phenotypes of the ΔnisA (ΔagsB), ΔagtC, ΔgnsA, and ΔagtC ΔgnsA strains. (A) Strategy for homologous recombination of the A. luchuensis ΔligD strain for nisA (agsB) gene disruption using the hph gene as a selectable marker. (Strategies for agtC, gnsA, and agtC-gnsA are shown in Fig. S1A to C in the supplemental material.) The gray bar indicates the hybridization position of the probe to confirm gene replacement by Southern blotting. HindIII restriction sites are indicated by the letter H. The black arrows indicate the primer set for confirmation of ΔnisA. (B) Southern blot analysis of the genomic DNA from the transformants. (Left) Each lane contained 20 μg of genomic DNA of the ΔnisA mutant and the wild-type strain (ΔligD) cut with HindIII and the ΔagtC and ΔligD strains cut with EcoRI. (Right) Each lane contained 20 μg of BamHI-digested genomic DNA derived from the ΔligD, ΔgnsA, or ΔagtC ΔgnsA strain. EcoRI and BamHI restriction sites in the genomic DNA are indicated in Fig. S1A to C. (C) Phenotypes and sensitivity of wild-type (ΔligD), ΔnisA, ΔagtC, ΔgnsA, and ΔagtC ΔgnsA strains on CD medium, PD agar (PDA), CD medium plus micafungin (Mica) (10 ng/ml), CD medium plus calcofluor white (CFW) (10 μg/ml), and CD medium plus Congo red (CR) (400 μg/ml). The cells (1 × 10⁴) were cultured on PDA plates for 3 days and on CD plates for 4 days at 30°C.

To examine nigeran production by the ΔagsB (ΔnisA), ΔagtC, ΔgnsA, and ΔagtC ΔgnsA strains, the mutants were initially cultured in low-nitrogen medium for 72 h and then transferred to nitrogen-free medium and harvested after 72 h. The dry weight of each mycelium was comparable to that of the wild-type strain. The capacity to synthesize nigeran was almost completely lost in the ΔagsB (ΔnisA) strain, whereas the nigeran contents in both the ΔgnsA and ΔagtC ΔgnsA strains were reduced to approximately 63% of that of the wild-type strain (Fig. 3A). In contrast, the amount of nigeran produced by the ΔagtC strain was 121% of that of the wild-type strain. Analysis of the carbohydrate composition of nigeran produced by each strain revealed the presence of only glucose. The nigeran samples were hydrolyzed to nigerose (α-1,3-d-glucose disaccharide) using mycodextranase from Bacillus sp. strain NHB-1 (Fig. S2), which is a specific nigeran-degrading enzyme (15). The ¹H nuclear magnetic resonance (NMR) spectra of the nigeran samples from the mutant were consistent with that of the wild type (Fig. S3). In addition, the abundance ratio of α-1,4-linked glucose to α-1,3-linked glucose was 1:1.0 to 1:1.1. The ratio was calculated using the integrated value of the anomeric proton, which indicated that agtC and gnsA were not involved in regulating the alternating α-1,3- and α-1,4-glucoside linkage pattern in nigeran. Taken together, these results suggested that AgsB (NisA) is the nigeran synthase and that AgtC and GnsA are involved in nigeran synthesis in A. luchuensis.

FIG 3.

Polysaccharide contents of cell wall fractions of A. luchuensis wild-type (WT) and mutant strains (ΔnisA [agsB], ΔagtC, ΔgnsA, and ΔagtC ΔgnsA). A. luchuensis was cultured in low-nitrogen medium, followed by 3 days of culture in nitrogen-free medium. Cell wall polysaccharides were fractionated into nigeran (A), alkali-soluble (AS) (B), and alkali-insoluble (AI) (C) fractions. Values represent the means from three independent quantifications ± SD. Dunnett’s test was performed to compare the means of mutants with that of the wild-type strain. Statistical significance is indicated (*, P < 0.05; **, P < 0.01; n.d., not determined).

Analysis of the nonnigeran cell wall polysaccharides revealed that the yield of the alkali-soluble fraction of the ΔagtC strain was significantly reduced compared to that of the wild type (P < 0.01) (Fig. 3B), although the nigeran level was significantly increased (P < 0.05) (Fig. 3A). In contrast, the yield of the alkali-insoluble fraction of the ΔagsB (ΔnisA) and ΔgnsA strains was significantly increased (P < 0.01 and P < 0.05, respectively) (Fig. 3C), whereas the nigeran levels were significantly reduced (P < 0.01) (Fig. 3A). To compare the abundance ratios of β-1,3-glucan and chitin in the alkali-insoluble fraction, the fraction was acid hydrolyzed, followed by sugar composition analysis. The glucose/glucosamine ratio in the alkali-insoluble fraction of each strain did not change significantly (Fig. S4). Nigeran is likely one of the compensatory components of the fungal cell wall under nitrogen-free conditions. This is reflected by the fact that the alkali-insoluble fraction increased in the ΔagsB (ΔnisA) strain concomitant with a reduced nigeran content, and the alkali-soluble fraction was decreased in the ΔagtC strain concomitant with an increased nigeran content.

Nigeran production by the A. luchuensis agsB (nisA)-overexpressing strain in A. oryzae.

To investigate whether A. luchuensis AgsB (NisA) alone can synthesize nigeran, we constructed an agsB (nisA)-overexpressing strain (OEnisA) using the A. oryzae niaD300 strain, which does not synthesize nigeran, as the host and the plasmid pNEnisA (Fig. S5A). The OEnisA candidates were screened for nitrate assimilation and confirmed by colony PCR (Fig. S5B). A. oryzae OEnisA was inoculated into PD liquid medium including 2% maltose. After 96 h of incubation, the mycelium was harvested and lyophilized. The dry weight of the OEnisA mycelium was similar to that of the parent strain (A. oryzae niaD300), but only the OEnisA strain produced nigeran (Fig. S5C), which comprised solely glucose and was hydrolyzed to nigerose by mycodextranase (Fig. S2). The ¹³C NMR spectra of nigeran from the A. luchuensis wild-type and A. oryzae OEnisA strains are shown in Fig. 4. The spectra were almost identical to each other and agreed well with the ¹³C NMR spectrum of nigeran prepared from A. niger (16). A. luchuensis agsB encodes the nigeran synthase NisA. As there are no agtC or gnsA homologs in the A. oryzae genome, our results provide evidence that NisA alone is sufficient to synthesize nigeran.

FIG 4.

NMR spectra of nigeran. Shown are ¹³C NMR spectra of the nigeran fractions from the A. luchuensis wild-type (A) and A. oryzae OEnisA (B) strains in DMSO-d₆ at 50°C. Chemical shifts are expressed as δ ppm.

Determination of the molecular weight of nigeran.

We predicted that the deletion of AgtC would affect the molecular weight of nigeran because AgtC possesses an N-terminal signal peptide sequence, an α-glucanotransferase domain, and a C-terminal GPI-anchoring sequence that localizes the protein to the plasma membrane of A. luchuensis. Therefore, we analyzed the molecular weights of nigeran prepared from the A. luchuensis wild type and the mutants (ΔagtC, ΔgnsA, and ΔagtC ΔgnsA) by size exclusion chromatography. The average molecular weight (M_w), number-average molecular weight (M_n), and broadness of the molecular weight distribution of the polymer (M_w/M_n) were calculated and are summarized in Table S2. After being transferred to nitrogen-free medium for 24 h, the size exclusion chromatograms of nigeran from all samples were similar (Fig. 5). The peak top molecular weight (M_p) of nigeran (peak 1 [P1]) was detected at approximately 9.0 min and was approximately 650,000, with a degree of polymerization estimated at 4,012. As the culture time increased from 24 to 48 and 72 h in the wild-type strain, a new peak (peak 2 [P2]) appeared with an increased peak area (Fig. 5A). Its M_p (around 10.5 min) and the degree of polymerization were 45,000 and 277, respectively. The M_w/M_n value of nigeran from the wild-type strain gradually increased from 6.2 to 10.7 with increasing induction time (Table S2). ΔgnsA nigeran also exhibited an increased peak area of lower-molecular-weight nigeran (P2) with an increasing induction time, similar to that of the wild-type strain (Fig. 5C). In contrast, the peak top higher-molecular-weight nigeran from the ΔgnsA strain shifted to approximately 8.4 min (peak 1′ [P1′]) after 48 and 72 h of incubation, and the calculated M_p and degree of polymerization were over 1,620,000 and 10,000, respectively. Consequently, the M_w/M_n value of the ΔgnsA strain nigeran was higher than those of the wild-type strain and other gene disruptants (Table S2). The molecular weight distribution of nigeran from the ΔagtC and ΔagtC ΔgnsA strains did not change significantly during the nigeran induction period (Fig. 5B and D). P2 was observed only in the wild-type strain and the ΔgnsA strain in which AgtC is functional; therefore, it was speculated that AgtC contributed to the decrease in the molecular weight of nigeran. In addition, we believe that GnsA affected AgtC and subsequently inhibited the polymerization of nigeran and promoted the depolymerization of nigeran.

FIG 5.

Temporal change in the molecular weight of nigeran derived from A. luchuensis wild-type and mutant strains as determined by size exclusion chromatography. Chromatograms of nigeran from the wild-type (A), ΔagtC (B), ΔgnsA (C), and ΔagtC ΔgnsA (D) strains after 24, 48, and 72 h of culture in nitrogen-free medium are shown. P1 (or P1′) and P2 indicate higher- and lower-molecular-weight peaks in the wild-type and ΔgnsA strains, respectively. Dextran standards (peak top molecular weight [M_p] values of 401,300, 276,500, 196,300, 123,600, 66,700, and 21,400) are indicated by black dot symbols and were used for calibration curve creation.

DISCUSSION

In this study, we identified genes responsible for nigeran synthase in A. luchuensis using RNA-seq analysis of genes differentially expressed under nitrogen-free relative to those under low-nitrogen culture conditions. Nitrogen starvation is known to stimulate the induction of nigeran production in A. luchuensis. After identifying the three candidate key genes, we analyzed the effects of their disruption in A. luchuensis and the overexpression of nisA (agsB) in A. oryzae. Our results clearly showed that nisA is a nigeran synthase. Comparing the phenotypes of the wild type and the mutants did not provide clear insights into the importance of nigeran production. However, the identification of the nigeran synthesis-related genes could help clarify the importance of nigeran production during nitrogen starvation in filamentous fungi.

To date, the whole genomes of several Aspergillus species have been sequenced, revealing a different number of AGS genes depending on the species, as follows (6, 17): A. nidulans (agsA and agsB), A. fumigatus (AGS1, AGS2, and AGS3), A. oryzae (agsA, agsB, and agsC), and A. niger (agsA, agsB, agsC, agsD, and agsE). The five AGS genes in A. luchuensis (agsA, nisA [agsB], agsC, agsD, and agsE) correspond to those of A. niger, respectively. The nigeran synthase NisA (AgsB) showed 48 to 54% sequence identity with the other four AGS proteins in A. luchuensis. The identities between the sequence of A. luchuensis NisA and those of the other AGS proteins are summarized in Table S3 in the supplemental material. The phylogenetic tree of Aspergilli AGS protein sequences, including those of A. luchuensis, revealed a clade consisting of A. luchuensis NisA (AgsB) and A. niger AgsB, which was clearly separated from other functionally known AGS protein groups (Fig. 6). The genes for nigeran synthase and α-1,3-glucan synthase likely evolved from a common ancestor gene based on their high sequence similarities and the results of this study. In addition, at least for Aspergillus spp., orthologs of A. luchuensis NisA were found only in the genomes of Aspergillus section Nigri members by a BLAST search, such as A. piperis, A. niger, A. tubingensis, A. aculeatus, A. japonicus, and A. neoniger (80 to 93% sequence identity). This observation is consistent with those of previous reports of some Aspergillus spp. producing nigeran, such as A. niger, A. aculeatus, and A. awamori (renamed A. luchuensis) (2, 18, 19).

FIG 6.

Phylogenetic tree of α-1,3-glucan synthases in Aspergillus spp. The alignment and phylogenetic analysis were performed with MEGA X. The tree was inferred using the maximum likelihood method. Gene standard names and accession numbers are shown. α-1,3-Glucan synthase in S. pombe Ags1p was chosen as the outgroup. Putative AGS proteins in A. luchuensis are shown in bold. The percentages of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) are shown next to the branches.

Most AGS genes in Aspergillus spp. form a cluster with amylase-like protein or α-glucanotransferase genes (14). A. nidulans agsB, the major AGS gene for α-1,3-glucan synthesis, forms a gene cluster with amyD and amyG (14, 20). AmyD, which comprises an N-terminal signal peptide, an amylase-like domain, and a GPI-anchoring sequence, represses α-1,3-glucan accumulation (14). The A. nidulans ΔamyD strain exhibited no phenotypic change compared to the wild-type strain grown on solid or in liquid medium, but the α-1,3-glucan content was increased by 1.5-fold compared to that of the wild type (14). The A. luchuensis AgtC protein identified in this study showed the same domain composition as A. nidulans AmyD. The nigeran content of the A. luchuensis ΔagtC strain also increased by 1.2-fold compared to that of the wild type (Fig. 3A). The protein products of three of the genes are common between A. luchuensis and A. niger (AgtA, AgtB, and AgtC) and lack the highly conserved His residue found in the conserved region I of fungal α-amylase, which is important for hydrolase activity (21). The recombinant AgtA (ortholog of AmyD) and AgtB derived from A. niger exhibit high transglycosylation activity and low hydrolase activity against starch and malto-oligosaccharides, whereas the enzymatic property of AgtC is unknown (21). In this study, A. luchuensis AgtC influenced the molecular weight of nigeran and functioned as a polymerization and depolymerization enzyme (Fig. 5). Bobbitt and Nordin reported that A. awamori (renamed A. luchuensis) secretes nigeran into the growth medium (19). This observation may result from the depolymerization activity of AgtC. Our results and these findings suggested that Agt proteins control α-1,3-glucan and nigeran accumulation and degradation in fungal hyphae through repeating transglycosylation and hydrolysis reactions. We plan to verify the enzymatic property of A. luchuensis AgtC in our future studies.

amyG in A. nidulans and its orthologs AMY1 in Histoplasma capsulatum and AMY1 in Paracoccidioides brasiliensis, which encode intracellular amylases, were suggested to be crucial for α-1,3-glucan synthesis because Amy1p activity provides malto-oligosaccharides for α-1,3-glucan synthesis (22, 23). In this study, GnsA did not exhibit any amino acid sequence similarity with AmyG or Amy1p, and unlike AmyG and Amy1p, GnsA was predicted to localize to the plasma membrane or galactomannan linked to cell wall β-1,3-glucan (24) via its C-terminal GPI-anchoring site. Therefore, GnsA may not provide substrates for nigeran synthesis in A. luchuensis. It is difficult to identify the function of GnsA because the sequence of the unknown-function region shows low similarities with fungal GPI-anchored proteins of known function. Although further studies are needed, we predict that GnsA supports nigeran synthesis by NisA and nigeran degradation by AgtC in the extracellular matrix.

In conclusion, we demonstrate that NisA is essential for nigeran synthesis in A. luchuensis and that AgtC and GnsA are also involved in nigeran synthesis. AgtC and GnsA control the nigeran content and molecular weight of nigeran. These data will be useful to determine the physiological importance of nigeran production under nitrogen starvation in filamentous fungi. In addition, characterization of these nigeran synthesis-related enzymes and establishment of mass production systems for nigeran will assist in a variety of nigeran-based applications, such as its use as a novel biopolymer (25).

MATERIALS AND METHODS

Strains and media.

Standard Escherichia coli manipulations were performed as described previously (26). E. coli strain DH5α (Nippon Gene Co. Ltd., Tokyo, Japan) was used for plasmid propagation. A. luchuensis NBRC4314 (RIB2604; National Research Institute of Brewing Stock Culture) was previously used for the genome sequencing project (9) and was used here as a nigeran-producing strain and a source of genomic DNA. A. oryzae niaD300 was used as a host for overexpressing A. luchuensis nisA (agsB). These strains were grown in potato dextrose (PD) medium (BD, Tokyo, Japan) for the preparation of conidial suspensions. Low-nitrogen medium (3% sucrose, 0.1% NH₄NO₃, 0.03% MgSO₄·7H₂O, 0.126% KH₂PO₄, 0.001% yeast extract, and 0.001% peptone, adjusted to pH 5.5 with KOH) and nitrogen-free medium (3% sucrose, 0.03% MgSO₄·7H₂O, and 0.126% KH₂PO₄, adjusted to pH 5.5 with KOH) were used for nigeran production in A. luchuensis. CD minimal medium (27) with 0.8 M NaCl was used as the selection medium for niaD gene transfer derivatives from A. oryzae niaD300.

RNA preparation for RNA-seq.

Total RNA extracted from the A. luchuensis NBRC4314 strain grown under low-nitrogen or nitrogen-deficient conditions was used for RNA-seq analysis as follows. All experiments under each growth condition were performed in triplicate as independent biological experiments. Total spores of A. luchuensis (10⁸ cells) were incubated in 200 ml low-nitrogen medium and grown for 72 h at 30°C on a rotary shaker at 100 rpm. The mycelium was harvested by suction filtration, washed with a sterilized saline solution, transferred into 200 ml of low-nitrogen medium or nitrogen-free medium, and cultured at 30°C for 72 h at 100 rpm. Total RNA was isolated from the mycelium using an SV total RNA isolation system (Promega, Madison, WI, USA), and the RNA concentration was quantified using a Qubit fluorometer with the Qubit RNA HS assay kit (Thermo Fisher Scientific, Waltham, MA, USA). Total RNA (150 ng) was used for cDNA synthesis using a SureSelect strand-specific RNA library prep kit (Agilent Technologies, Santa Clara, CA, USA). RNA-seq analysis was performed using paired-end sequencing on an Illumina MiSeq platform (Illumina, San Diego, CA, USA).

RNA-seq data analysis.

All libraries were processed, and expression analysis was performed as follows. First, the quality of the obtained sequence reads was checked using FastQC (28), and the adaptor sequences were removed from all libraries using cutadapt 1.4.2 (29) using default parameters. The trimming of low-quality reads was performed using SolexaQA v2.5 (30) with a Phred score cutoff of 28 in DynamicTrim.pl and a minimum trimmed read length of 23 in LengthSort.pl. These reads were mapped to the A. luchuensis RIB2604 strain reference genome using HISAT2 (31) with default parameters. Reads were counted using featureCounts v1.5.2 (32). Differential gene expression levels were compared using a generalized linear model approach implemented using the edgeR 3.18.1 Bioconductor package (33). Normalization factors for each library were calculated using the trimmed mean of M-values method (34). A false discovery rate of <0.05 was used as the cutoff for differential expression.

RT-qPCR.

Spores of A. luchuensis (10⁷ cells) were incubated in 100 ml low-nitrogen medium and cultured for 72 h at 30°C at 100 rpm. The mycelium was washed, transferred into 100 ml of nitrogen-free medium, and cultured at 30°C at 100 rpm. Total RNA was extracted from the A. luchuensis mycelium cultured in nitrogen-free medium for 0, 24, 48, and 72 h, and 100 ng RNA was used for cDNA synthesis, using the procedures described above. Quantitative PCR was performed using the StepOne Plus real-time PCR system (Thermo Fisher Scientific) in a total volume of 10 μl containing cDNA at the appropriate dilution and FastSYBR green mix (Thermo Fisher Scientific). RT-qPCR conditions were set as follows: 30 s at 95°C for initial denaturation, followed by 40 cycles of 95°C for 3 s and 60°C for 30 s. Primers designed for nisA (agsB) (RT_nisAFw and RT_nisARv), agtC (RT_agtCFw and RT_agtCRv), gnsA (RT_gnsAFw and RT_gnsARv), and the β-actin gene (RT_actinFw and RT_actinRv) are listed in Table 1. Expression levels were normalized to that of the β-actin gene and calculated using the standard curve method. Three biological replicates were analyzed with three technical replicates.

TABLE 1.

Primers used in this study

Primer name	Sequence (5–3′)a
RT_nisAFw	TCTTCTTTGTCGTGGTGTGGG
RT_nisARv	GATGGGGATGAACCAGGCAT
RT_agtCFw	CGCGTGTACTATCCCACAGA
RT_agtCRv	TTGGTTTGGGCTGACGGATG
RT_gnsAFw	ACGTCCGCGTCTTCTACCT
RT_gnsARv	CCAATCAGAGCCACAAGCCC
RT_actinFw	GGTATGGGTCAGAAGGACTC
RT_actinRv	CTCCATGTCATCCCAGTTCG
5nisAFw	CGTTTTTAATGAATTGTTCAGGGCACAAAAAGATGATTACCGACC
5nisARv	GAGCTCAATGGCCCGGACGGTGTAGGTATCCCGACCAGCGAGTC
3nisAFw	TGGACCCCGAAGGCGCTAACTCTTTTGGACCAGTCTCTTGCTTG
3nisARv	AGTTTAAACTGAATTGAGTACCCTTCATATACTGTCTGCTCTTCG
5agtCFw	CGTTTTTAATGAATTGTAACACGCCACGCCACCTTGCTCGG
5agtCRv	GAGCTCAATGGCCCGGATTATGACCTGCGTAACCTTCTGTTGAGC
3agtCFw	TGGACCCCGAAGGCGGGCCTGTCTTTCCCTGTTTCGCATTACCC
3agtCRv	AGTTTAAACTGAATTGGGTTCCCTTTGACGGGATCGTATAACG
5gnsAFw	CGTTTTTAATGAATTGGCCTGTCTTTCCCTGTTTCGCATTACCC
5gnsARv	GAGCTCAATGGCCCGGGGTTCCCTTTGACGGGATCGTATAACG
3gnsAFw	TGGACCCCGAAGGCGCGGGATCATGCACACGCTTGCTTTTAGTC
3gnsARv	AGTTTAAACTGAATTGACTTCCGGCGGTGTGAAGATTGCCATG
OEnisAFw	GCTTGCGGCCGCCACAAAATGTCCACCAAATGTGGCGTGTATGTCAGC
OEnisARv	CATATGACTAGTCACCTATGGTTTTGAAAGCTGCTCTTTGCGG
OEnisA_confFw	CAGTTGCTGGAGCCATTCTCGGACGG
OEnisA_confRv	GATTCTGCCTTCCTGCTCAACGCGTG
OEnisA_puriFw	GTAGTAGCTCCGCTCTAACAGCCGTG
OEnisA_puriRv	GGTTTCGGTCGTCAAAGGTCTCATTCGTG
DnisA_Fw	ACCATCCAGCTCTGCGCAATCCTTACAG
hph_Rv	GCAGTTCGGTTTCAGGCAGGTCTTGCAAC
DagtC_Fw	GTGGAGAAATCCCATTCAGTGCCTGAGG
DgnsA_Fw	CCGTGCCGATTACACTACAATTATTGAGATTACC
probeA-Fw	GGGAGGACAAGAGTGGAGGAAAGCTATGG
probeA-Rv	GTCTATCCCACTCACATTTCCCTCCTTGGC
probeB-Fw	CTTTGGCTTAAGGGCGCACTAGAAAATGAG
probeB-Rv	CTCGTAATCACTCACCGGTAGTCGCAAC

^aUnderlining indicates the Kozak sequence.

Construction of deletions in nigeran synthesis-related genes in A. luchuensis.

Four plasmids, ΔnisA::hph/pRIE, ΔagtC::hph/pRIE, ΔgnsA::hph/pRIE, and ΔagtCgnsA::hph/pRIE, used for each disruptant of the nigeran synthesis genes were constructed as described below. The 5′ and 3′ fragments of the nisA (agsB), agtC, gnsA, and agtC-gnsA genes were prepared by performing PCR using primers designed for nisA (primers 5nisAFw and 5nisARv and primers 3nisAFw and 3nisARv), agtC (primers 5agtCFw and 5agtCRv and primers 3agtCFw and 3agtCRv), gnsA (primers 5gnsAFw and 5gnsARv and primers 3gnsAFw and 3gnsARv), and agtC-gnsA (primers 5agtCFw and 5agtCRv and primers 3gnsAFw and 3gnsARv) (Table 1) using A. luchuensis NBRC4314 genomic DNA as a template. The hygromycin B resistance gene (hph) cassette and the binary vector pRIE were prepared according to a previous report (13). Each of the four DNA fragments was ligated using an In-Fusion cloning kit (TaKaRa Bio Inc., Shiga, Japan), generating the four plasmids. In the ΔagtC ΔgnsA construction, agtC is flanked on the 3′ side by gnsA and the hph marker, and homologous regions upstream of agtC and downstream of gnsA were used for the double disruption.

Transformation of A. luchuensis ΔligD was performed using the agrobacterium-mediated transformation method (35). The transformants were screened for hygromycin (0.1 mg/ml) resistance and subcultured at least once on CD agar plates containing hygromycin. Homologous recombination in primary transformants was confirmed by colony PCR, as described previously (36), using primer set 1 (DnisAFw and hphRv) for the ΔnisA strain, set 2 (DagtCFw and hphRv) for the ΔagtC strain, set 3 (DgnsAFw and hphRv) for the ΔgnsA and ΔagtC ΔgnsA strains, and set 4 (DagtCFw and hphRv) for the ΔagtC ΔgnsA strain. A single correct homologous integration using each fragment for gene disruption was confirmed by Southern blotting. Each probe used for hybridization was prepared by performing PCR with a digoxigenin DNA labeling kit (Merck Millipore, Darmstadt, Germany). Primers probeA-Fw and probeA-Rv for the ΔnisA and ΔagtC strains, primers probeB-Fw and probeB-Rv for the ΔgnsA and ΔagtC ΔgnsA strains, and the A. luchuensis NBRC4314 genome as the template were used. Hybridization and detection were performed according to the instruction manual of the digoxigenin DNA labeling kit.

Cell wall fractionation and sugar composition analysis.

A. luchuensis wild-type and gene disruptant (ΔnisA [agsB], ΔagtC, ΔgnsA, and ΔagtC ΔgnsA) strains were cultured in low-nitrogen medium and then additionally cultured in nitrogen-free medium for 72 h. The resultant mycelium was collected by filtration, washed with distilled water, and frozen at −80°C for 2 h. Subsequently, the mycelium was resuspended in 50 mM Tris-HCl buffer (pH 8.0) containing 1 mM EDTA and disrupted by ultrasonication. The pellet containing the cell wall fraction was collected by centrifugation (14,000 × g for 10 min), washed twice with distilled water, and lyophilized. The dried cell wall (0.5 g) was suspended in 100 ml distilled water and autoclaved at 121°C for 30 min. After autoclaving, the nigeran fraction dissolved in hot water (>80°C) was collected by filtration. The filtrate was cooled at room temperature and then refrigerated at 4°C overnight. The resulting cold-water-insoluble precipitant (nigeran fraction) was collected by centrifugation and lyophilized. The residual cell wall fraction was dissolved in a 30-ml 1 M NaOH solution, and the extracted alkali-soluble polysaccharide solution was incubated overnight at 4°C with gentle shaking, followed by centrifugation. The supernatant (alkali-soluble fraction) was neutralized with acetic acid to pH 6.0 and then dialyzed against distilled water overnight. The precipitate (alkali-insoluble fraction) was dissolved in 30 ml distilled water, neutralized, and dialyzed similarly to the alkali-soluble fraction. The nigeran, alkali-soluble, and alkali-insoluble fractions were collected by centrifugation, lyophilized, and weighed.

To analyze the polysaccharide composition of the alkali-insoluble fraction, 10 mg of the powdered fraction was dissolved in 1 ml of cooled 72% (wt/wt) H₂SO₄ in a glass tube, sonicated in ice water for 1 h, and diluted with distilled water to 2 N H₂SO₄. The sample was hydrolyzed at 121°C for 1 h using an autoclave. After cooling the sample at room temperature, barium carbonate was added to neutralize the supernatant, which was diluted with MilliQ water and filtered. The sugar composition was analyzed by high-performance anion-exchange chromatography with pulsed amperometric detection using a Dionex ICS-5000⁺ DC system (Thermo Fisher Scientific). Separation was carried out on a Thermo Fisher Scientific 4- by 250-mm CarboPac PA1 column coupled with a 4- by 50-mm CarboPac PA1 guard column at 30°C at a flow rate of 1 ml/min. The eluent gradient used for sugar separation started at 0 min and ended at 30 min, after which the sodium acetate concentration was raised from 0 to 150 mM sodium acetate and the NaOH concentration was maintained at 100 mM.

Construction of agsB (nisA)-overexpressing strain OEnisA in A. oryzae.

To construct the overexpression plasmid pNEnisA, nisA (agsB) was amplified by PCR using the primers OEnisAFw and OEnisARv (Table 1) with A. luchuensis NBRC4314 genomic DNA as a template. The amplified PCR product included a consensus Kozak sequence (37, 38) in front of the nisA start codon. The vector for gene overexpression, pNEN142 harboring a maltose-inducible promoter (39), was digested with PmaCI, and the two DNA fragments were ligated using an In-Fusion cloning kit, resulting in pNEnisA. The cloned nisA gene was sequenced. A. oryzae niaD300 was transformed with pNEnisA digested with SgrAI as previously described (40) (see Fig. S4A in the supplemental material). The transformants were screened for nitrate prototrophy and purified by subculturing the cells at least three times on CD agar plates. The transformants were subjected to colony PCR (35) using primer set 1 (OEnisA_confFw and OEnisA_confRv) and set 2 (OEnisA_puriFw and OEnisA_puriRv) to confirm integration, followed by purification (Table 1). When the digested pNEnisA was inserted into the targeted niaD locus, a 0.6-kb fragment was amplified using primer set 1. Following the purification of the transformants, a 4.4-kb fragment was amplified using primer set 2. In contrast, amplification of a 3.7-kb fragment with primer set 2 indicated that the transformant was still heterokaryotic.

To examine whether the A. oryzae OEnisA strain could produce nigeran, the strain was cultured in PD liquid medium including 2% maltose at 30°C for 96 h at 100 rpm. Nigeran was extracted from the resulting mycelium using hot water, as described above.

NMR analysis.

Nigeran prepared from A. luchuensis ΔligD (wild-type), the gene disruptants (ΔagtC, ΔgnsA, and ΔagtC ΔgnsA), and A. oryzae OEnisA was dissolved in 1 ml dimethyl sulfoxide-d₆ (DMSO-d₆) and 0.05% tetramethylsilane (Sigma-Aldrich, St. Louis, MO, USA), which served as an internal reference (δ_H,C, 0.0 ppm). ¹H and ¹³C NMR experiments were recorded using a Bruker Avance III 500 spectrometer (Bruker Biospin, Ettlingen, Germany) at 500 and 125 MHz at 50°C, respectively. The spectra were recorded and analyzed using the TopSpin 3.6 program (Bruker Biospin).

Determination of the molecular weight of nigeran prepared from A. luchuensis and its variants.

The A. luchuensis wild type and the gene disruptants were cultured in low-nitrogen medium for 3 days and then cultured in the nitrogen-free medium for 1 to 3 days. Nigeran was prepared from the mycelium by hot-water extraction, as described above. The molecular weight of the nigeran samples was analyzed by size exclusion chromatography using a Shimadzu LC-20AD pump and a Shimadzu RID-20A refractive index detector (Shimadzu, Kyoto, Japan). Separation was carried out on a Shodex sugar KS-805 column (8 by 300 mm) coupled to a Shodex KS-G guard column (6 by 50 mm) at 50°C and eluted with 0.1 M NaOH at a flow rate of 1 ml/min. Nigeran (5 mg) was dissolved in 0.1 ml 1 M NaOH and diluted to 10-fold with MilliQ water. Dextran standards (Sigma-Aldrich) were used for generating the calibration curve. The values of M_w, M_n, and M_w/M_n were calculated using CDS-Lite version 5.0 (LAsoft Ltd., Osaka, Japan) according to the manufacturer’s instructions.

Phylogenetic analysis.

Multiple-sequence alignments of the full-length α-1,3-glucan synthase protein sequences from Aspergillus spp. were performed with ClustalW. A maximum likelihood tree was constructed using MEGA X (41) using the JTT matrix-based model (42) with 1,000 bootstrap replicates. All positions with <80% site coverage were eliminated (partial deletion option). There were a total of 2,349 positions in the final data set. The sequence from Schizosaccharomyces pombe Ags1p was used as the outgroup to root the tree.

Data availability.

RNA-seq data were deposited in the DDBJ Sequence Read Archive database under accession number DRA011440. All remaining data are contained within the article.