Introduction

In unicellular organisms, cell volume is tightly regulated in coordination with nuclear division timing (1–3). Similarly, in multinucleated filamentous fungi, a correlation exists between cell volume and nuclear number (4, 5). However, the number of nuclei varies across species (6), the factors governing nuclear number and cell volume remain largely unknown. Filamentous fungi are industrially significant microorganisms, and the strains selected through breeding provide ideal research models for studying these evolutionary processes.

Filamentous fungi secrete a variety of hydrolytic enzymes to support their growth, and humans have long harnessed this enzyme secretion ability for the production of a wide range of fermented foods and beverages (7). Aspergillus oryzae has been used in Japan for over 1,000 years to produce traditional fermented foods like sake, soy sauce, and miso (8, 9). A distinctive feature of the fermentation techniques is solid-state cultivation using substrates like rice, soybeans, and wheat bran. The inoculum of A. oryzae used in fermentation, known as "koji," has been commercially produced for around 700 years (10). Koji is created by cultivating koji-fungus on grains, and koji starter manufacturers have selected and bred strains to improve the flavor and color of various fermented products. The A. oryzae strains currently in use are mainly preserved and managed by koji manufacturers and research institutions. A notable characteristic of A. oryzae is its high ability to secrete starch-degrading enzymes and proteases, which has led to extensive research into its genome and gene expression regulation (11–17).

Although A. oryzae shares 99.5% genomic similarity with Aspergillus flavus, which produces the carcinogenic aflatoxin (18), there have been no reports of A. oryzae producing aflatoxins, and its safety has been confirmed at the molecular level (19, 20). Phylogenomic analyses suggest that A. oryzae and A. flavus diverged approximately 50,000 to 189,000 years ago (21, 22). Over the following millennia, human use of A. oryzae is thought to have driven significant genomic recombination, resulting in its evolution into a "cell factory" specifically optimized for the breakdown of sugars and proteins (23).

Additionally, modern biotechnology utilizes filamentous fungi as cell factories to produce organic acids, enzymes, and pharmaceuticals (24–26). A. oryzae also shows high production capacity as a host for both homologous and heterologous protein production in modern biotechnology (27, 28). Recently, its application has expanded to include the production of pharmaceutical proteins and secondary metabolites (29, 30). Furthermore, filamentous fungi involved in fermentation play a crucial role in creating a more sustainable food system, such as by upcycling agricultural by-products into food through fungal fermentation and using A. oryzae mycelium for alternative meats (31, 32).

While basic and applied research on A. oryzae continues, fundamental questions remain unanswered, such as why A. oryzae exhibits high enzyme production capacity and why it excels as a host for heterologous expression. This study shows that the key lies in the increase in cell volume and nuclear number, and analyzes the molecular mechanisms.

Result

Increase in nuclear number and cell volume in Aspergillus oryzae

In the model strain A. oryzae RIB40, hyphae on day 1 of cultivation contain 10-20 nuclei per cell, but by day 2 of cultivation, apical cells with more than 200 nuclei appear (33). We found that the proportion of hyphae with a higher number of nuclei (represented by class III, over 200 nuclei in the cell, see methods) increased as the culture period increased from 1 to 3 days (Fig. 1A, B) (33). This phenotype was not observed in the model fungus Aspergillus nidulans or the closely related species A. flavus (Fig. 1A, B) (33). As far as we know, such phenotypes have not been observed in other strains of Aspergillus nidulans, Aspergillus niger, Aspergillus fumigatus, or other strains commonly used in research. Among other Aspergillus species used in fermentation (9, 10), the proportion of hyphae with a higher number of nuclei increased in Aspergillus sojae, which is used for miso and soy sauce production. This phenotype was not observed in Aspergillus luchuensis nor in its albino mutant Aspergillus luchuensis mut. kawachii, which are used for distilled spirits such as awamori and shochu. Among several strains of A. oryzae used in industrial applications, the RIB915 strain used for soy sauce fermentation does not show an increase in the number of nuclei, while the RIB128 and RIB430 strains used in sake brewing exhibit a significant increase in the number of nuclei, suggesting that different phenotypes were selected through distinct breeding processes.

Figure 1.

The distribution of nuclei in the hyphae of A. oryzae RIB40 was visualized through 3D reconstruction (Fig. 1C, Movie 1). In the class I hyphae, the nuclei were spaced relatively evenly, whereas in the class III hyphae with increased nuclei, the nuclei were densely packed and arranged in a disordered manner. Furthermore, the nuclear size in the class III hyphae was significantly smaller than that in the class I hyphae (Fig. 1D, E).

Strains without an increase in nuclear number had hyphae with diameters ranging from 2 to 7 μm, while strains with an increase in nuclear number had some hyphae of similar thickness but often exhibited thicker hyphae exceeding 10 μm (Fig. 1F). The number of nuclei and cell volume in hyphae at 100 μm from the tips were compared among A. oryzae strains (RIB40, RIB128, RIB915) and A. flavus (Fig. 1G). A strong positive correlation was observed between number of nuclei and cell volume. This is consistent with previously reports that the number of nuclei per hyphal volume remains constant (4, 5).

Thick hyphae with increased nuclei emerge by branching

Time-lapse imaging using the A. oryzae RIB40 expressing H2B-GFP indicated that thick hyphae with increased nuclei emerged by branching from hyphae without increased nuclei (Fig. 2A, Movie 2). In the emerged thick hyphae, rapid nuclear division occurred at the sites of branching (Fig. 2B, Movie 3). Immediately after branching, the thick hyphae showed a significantly faster rate of nuclear proliferation, increasing 4–5 times within 8 hours compared to the thin hyphae immediately after branching (Fig. 2C). Branches from hyphae with increased nuclei produced both hyphae with and without increased nuclei (Fig. 2D).

Figure 2.

Time-lapse imaging was conducted over a wider area, and image processing was applied to color-code the nuclei: green for those in hyphae with increased nuclei and white for those in hyphae without increased nuclei (Fig. 2E, Movie 4, Fig.S1A, see methods). Up to 24 hours after inoculation, most hyphae were thin with no increase in nuclei number, but thick hyphae with increased nuclei began to appear as time progressed (Fig. 2E, F). Since thick hyphae elongated faster than thin hyphae (Fig. 2G), thick hyphae outgrew and overtook thin hyphae, eventually dominating the colony perimeter (Fig. 2E, F).

Transmission electron microscopy confirmed basic nuclear structures, such as the nuclear membrane and nucleolus, in both thick and thin hyphae (Fig. S1B). As hyphal diameter increased, cell walls became more uneven and thicker (Fig. 2H, I). This finding corresponds with Raman spectroscopy results showing pronounced peaks corresponding to cell wall polysaccharides in thick hyphae (Fig. S1C)(34). Additionally, peaks associated with active mitochondria were prominent in thick hyphae (Fig. S1C)(35), which was further confirmed by fluorescence staining (Fig. S1D). The surface structure and mechanical properties of the cell wall were analyzed using atomic force microscopy. Thick hyphae exhibited greater surface roughness and higher elasticity compared to thin hyphae (Fig. 2J, K, Fig. S1E, F).

Correlation between nuclear number and enzyme secretion

We compared the amounts of secreted proteins in A. oryzae RIB40, RIB128, RIB915, A. nidulans, and A. flavus (Fig. 3A). The strains that show an increase in nuclear number had significantly higher levels of protein secretion than the strains that do not. In this condition, α-amylases are the dominant proteins secreted by A. oryzae (15). The amylase enzyme activity was compared between in A. oryzae RIB40, which shows an increase in nuclear number, and A. oryzae RIB915, which does not show an increase. In RIB40, starting from day 2, when hyphae with increased nuclear numbers appeared, the enzyme activity per fungal biomass significantly increased more than RIB915 (Fig. 3B).

Figure 3.

To compare the amylase activity between hyphae with a higher and a lower number of nuclei, amylase activity was monitored under conditions where a single hypha grew in a microfluidic channel with a fluorescent substrate (Fig. 3C, S2A, Movie 5)(36). After 12 hours, the fluorescence intensity in the channel with thick hyphae was about three times higher than that in the channel with thin hyphae, indicating that thicker hyphae secreted more amylase than thinner hyphae (Fig. 3D).

We found that addition of yeast extract increased in both hyphal width and nuclear number in the RIB915 strain (Fig. 3E, Fig. S2B), whereas adding yeast extract to A. nidulans did not result in an increase in hyphal width or nuclear number. In A. oryzae RIB40, the addition of yeast extract did not make the already thick hyphae any thicker but increased the proportion of thicker hyphae. Addition of yeast extract increased protein secretion approximately 1.8 times in RIB40 and 8.5 times in RIB915 (Fig. 3F).

The addition of nucleic acids, peptone, or casamino acids to the minimal medium increased the number of nuclei in A. oryzae RIB915, while vitamin B did not cause a significant change in nuclear number (Fig. S2C). Although the addition of individual amino acids did not show the same effect as yeast extract, the addition of asparagine, proline, glutamine, and branched-chain amino acids increased the proportion of hyphae with increased nuclei by 40-50% (Fig. 3G, Fig. S2D). The amino acids that induced nuclear number increase also led to an increase in protein secretion, indicating a positive correlation between the ratio of class III hyphae and the increase in secreted proteins (Fig. 3H, Fig. S2E). Since the amino acids inducing an increase in nuclear number may activate the TOR pathway (37), A. oryzae RIB40 was grown with low concentration of rapamycin, a TOR pathway inhibitor, which minimally affects colony size. Under the conditions, RIB40 did not produce hyphae with an increased number of nuclei (Fig. 3I-K, Fig. S2BF).

Transcriptome analyses in hyphae with increased nuclei

To investigate gene expression changes in hyphae with increased nuclei, transcriptome analysis was performed under six conditions: In minimal medium; A. oryzae RIB40, which shows increased nuclei, A. oryzae RIB915 and A. nidulans, which do not show increased nuclei. In minimal medium with yeast extract, A. oryzae RIB40 and RIB915, which shows increased nuclei, and A. nidulans, which do not show increased nuclei (Fig. 4A, Fig. S2B). In RIB915, supplementation with yeast extract led to a more than four-fold increase in the expression of 660 genes, among which 449 genes did not show such increase in RIB40 or A. nidulans (Fig. 4B, Table S1). GO analysis of these genes revealed that processes related to cell wall synthesis and divalent metal ion transport were significantly enriched (Fig. 4C).

Figure 4.

To capture gene expression changes associated with increased nuclei with higher resolution, thick and thin hyphae from RIB40 were dissected using laser microdissection, collected separately, and subjected to transcriptome analysis (Fig. 4D, Movie 6). Clear differences in gene expression pattern were observed between thick and thin hyphae (Fig. 4E). 558 genes that were more than four-fold upregulated in thick hyphae but not in thin hyphae (Table S2). GO analysis of the upregulated genes indicated an enrichment of processes related to biogenesis of ribosome and cellular component (Fig. 4F). Among them, 21 matched those from Table S1 (Fig. 4G, Table S3). Of these, 13 had GO annotations, which were related to conidiation, cell wall synthesis, Ca²⁺ transport and rRNA processing (Fig. 3H). KEGG map visualization indicates that overall ribosomal gene expression increased in the thick hyphae than the thin hyphae (Fig. S3A).

Among the gene list, we focused on msy1 and msy2, which are involved in Ca²⁺ transport and maintaining cell volume homeostasis in Schizosaccharomyces pombe (38, 39). Disruption of the orthologue genes, msyA and msyB, in A. oryzae RIB40 did not have a significant impact on colony growth, hyphal morphology and number of nuclei in minimal medium (Fig. S3B). Under hypoosmotic shock conditions, the ΔmsyA often indicated cell lysis near the hyphal tips (Fig. 4I, J). Under hypoosmotic conditions, water influx into the cells causes hyphal cell expansion, but the volume and turgor pressure are regulated to prevent cell lysis. The msyA disruption, however, impairs the regulation of the hyphal cell expansion, leading to hyphal lysis (Fig. S3C-E).

SNP analysis among industrial strains with distinct phenotypes

The fact that currently used A. oryzae strains are managed and preserved by koji manufacturers and research institutions is a notable advantage of A. oryzae research. Whole-genome sequencing and comparative genome analysis of 82 A. oryzae strains with different applications have shown that these strains cluster into several clades (24). We investigated whether nuclear numbers increase in 20 strains selected from six different clades (Fig. 5A). In the minimal media supplemented with yeast extract, all strains exhibited an increase in nuclear number (Fig. S4A). In the minimal media without yeast extract, three strains from Clade C showed an increase in nuclear number, whereas strains from Clades A, B, and E did not (Fig. 5A, B, Fig. S4A). Clades F and G included both strains with increased nuclear numbers and strains without. SNP analysis of ORFs in Clades F and G revealed 108 mutations in Clade F between TK-32 and TK-38, and 23 mutations in Clade G between TK-41 and TK-47 (Table S4 and S5).

Figure 5.

Among these, AO090038000626; rseA, predicted glycosyl transferase, was the only common gene. This gene has been identified as a mutation in a A. sojae mutant with high extracellular enzyme production (40). It is also shown that deletion of rseA in A. nidulans leads to cell wall defects and promotes enzyme secretion (40). Comparison of mutations in rseA among strains in Clade C, F and G indicates a few amino acid substitutions within the ORF, but no common mutations corresponded to the phenotype (Fig. 5C, Fig. S4B). Structural prediction suggests that the mutations are located inside the enzyme domain (Fig. S4C).

The disruption of the rseA in A. oryzae RIB40 resulted in significant growth delay, increased branching, and the absence of hyphae with increased nuclei (Fig. 5D, Fig. S4D). The rseA gene DNA fragments from the strains with increased nuclei (RIB40, TK-47) or from the strain without increased nuclei (TK-41) were amplified by PCR and introduced ectopically into RIB915 (Fig. 5E). The strains expressing rseA from RIB40 or TK-47 exhibited an increase in hyphal width and a corresponding increase in the proportion of hyphae with more nuclei, whereas the strain expressing rseA from TK41 showed no increase in hyphal width or nuclear number (Fig. 5E, F, Fig. S4E).

Industrial breeding strains from other genera

To investigate whether the phenomenon of increased nuclear numbers also applies to other industrial fungi, we compared the bred strain of Trichoderma reesei (QM9414) for cellulase production and its control strain (QM6a)(41), as well as the bred strain of Penicillium chrysogenum (IFO4688) for penicillin production and wild type strain (42). In minimal medium, no significant differences in hyphal width or nuclear number were observed between the bred and control strains, whereas with the addition of yeast extract, a notable increase in nuclear number and hyphal width was observed only in the bred strains (Fig. 5G, H, Fig. S5A-C). The addition of yeast extract significantly increased cellulase and protease activity only in the bred strain of T. reesei and P. chrysogenum, respectively (Fig. 5I, J). These results suggest that even in modern biotechnology, where high enzyme or compound producing strains are bred, strains with increased nuclear numbers were selected, like the traditional breeding practices of koji-fungi.

Discussion

In A. oryzae some strains, hyphae grow about three times thicker and contain roughly ten times more nuclei than the initial hyphae (Fig. 1). These thicker hyphae emerge from existing branches, where nuclear division is promoted, resulting in the formation of hyphae with an increased number of nuclei (Fig. 2). A clear correlation was observed between hyphal cell volume and nuclear number (Fig. 1G). This consistency aligns with findings in other organisms, from unicellular yeast to multicellular plants and animals, where the nuclear-to-cytoplasmic ratio is tightly regulated, allowing cell cycle progression only upon reaching a critical cell size (1–3). Both the G1/S and G2/M transitions are controled by size checkpoints, with specific mechanisms differing between species (44, 45).

In A. nidulans, it is known that the duration of the G1 and G2 phases is regulated by temperature (46). While the mechanisms governing the G1/S phase transition remain unclear (47), the G2/M phase transition has been shown to involve the following: During the interphase of A. nidulans, NimX (CDK1) binds to NimE (cyclin B), and its activity is regulated through phosphorylation by AnkA (Wee1 kinase) and NimT (Cdc25 phosphatase)(48–50). The activated NimX-NimE complex subsequently dephosphorylates by NimT and activates NimA (Never-in-Mitosis A) protein kinase (51, 52). NimA is essential for the transition from G2 to mitosis, and its expression is tightly regulated during the cell cycle (52). The mRNA level of nimA and nimT begin to increase in G2, reach a plateau during late G2 and M phases, and are sharply degraded at the end of mitosis upon re-entry into interphase (52, 53), meaning high expression levels of nimA and nimT result in a longer G2 phase. In thick hyphae, nimA and nimT mRNA was hardly detected, suggesting that thick hyphae sense cell volume in some way and subsequently shorten the G2 phase (Fig. S6A).

Branching emerges as a result of cell wall loosening at the branching site. SNP analysis between industrial strains with different phenotypes identified RseA, a predicted glycosyltransferase (Fig. 5). Deletion of rseA caused defects in cell wall synthesis in A. nidulans, leading to increased enzyme secretion (40, 54). Although the targets of RseA are unknown, RseA could be involved in the cell wall integrity by modifying glycosyl chains on enzymes responsible for cell wall synthesis and degradation. Mutations in rseA may impair the regulation of cell wall loosening at the branching site, resulting in the formation of thicker hyphae. In fact, in the thick hyphae of RIB40, localized imbalances in cell wall components have been demonstrated by TEM and AFM.

Thicker hyphae can only form if high turgor pressure is maintained within the hyphae, in addition to changes in cell wall remodeling during branching. msyA and msyB, whose orthologues are calcium channels involved in the regulation of cell volume, were identified among the genes that show a significant increase in expression in thicker hyphae (Fig. 4)(38, 39). Analysis of the knockout strains demonstrated the function of cell volume regulation adapted to increased turgor pressure in hypoosmotic conditions. The function of these calcium channels is likely crucial for maintaining the balance of turgor pressure and cell volume.

The Target of Rapamycin (TOR) pathway is a central regulator of cellular growth, responding to a variety of nutrients, such as amino acids, and carbon sources, as well as environmental factors like oxygen levels, osmotic pressure, pH, and temperature stress (55, 56). Upon sensing these inputs, the TOR pathway influences translation activity, metabolic flow, and cell size (56, 57). In both yeast and mammalian models, amino acids like branched-chain amino acids (BCAAs), glutamine, and alanine are crucial for nutrient sensing, forming the basis for TOR signaling activation (57–59). These amino acids are like those contributing to nuclear increase in the RIB915 strain (Fig. 3G). Furthermore, TOR inhibition by rapamycin in the RIB40 strain suppressed the formation of hyphae with increased nuclei (Fig. 3I-K). Consistent with the activation of translation by the TOR pathway, the expression of ribosome-related genes was elevated in thicker hyphae (Fig. S3A). The TOR pathway directly regulates mitochondrial activity (60). Thick hyphae exhibited heightened mitochondrial density and activity (Fig. S1C, D), suggesting that the TOR pathway support the high ATP demand associated with rapid biosynthesis.

Cell volume and nuclear number mutually regulate each other, meaning that thicker hyphae always possess more nuclei. For the formation of thicker hyphae with numerous nuclei, an increase in cell volume and a corresponding increase in nuclei must occur simultaneously (Fig. S6A). In our model, mutations in rseA loosen the cell wall during branching, and the regulation of turgor pressure and cell volume by MsyA represents the necessary condition to increase cell volume and maintain thick hyphae. At the same time, enhanced translational capacity involving TOR activation and mechanisms that accelerate the cell cycle constitute another essential condition enabling an increase in nuclear number.

A. nidulans and A. flavus did not exhibit an increase in nuclear number. A. sojae displayed increased nuclear numbers, whereas A. kawachii and A. luchuensis did not (Fig. 1). The 20 strains of A. oryzae selected from different clades showed an increase in nuclear number in the minimal medium supplemented with yeast extract. In the medium without yeast extract, there were strains that showed an increase in nuclear number and others that did not. These differences might be due to variations in their applications or the selection of strains with diverse characteristics, such as not only enzyme activity but also their impact on the flavor, aroma, and color of the final products. Additionally, there might be dominant mutations that activate TOR pathway even in the minimal medium. Multiple gene mutations were identified as contributing to the phenotype of increased hyphal volume and nuclear number, suggesting that the phenotype was not the result of a single mutation but rather the gradual accumulation of mutations through breeding. The koji-fungus strains used to produce high-quality sake and soy sauce have been selectively bred over many years, long before the advent of modern biotechnology, to enhance their saccharolytic and proteolytic activities. Modern biotechnological bred strains such as T. reesei QM9414 and P. chrysogenum IFO4688 also exhibit a phenotype of thicker hyphae with increased nuclear numbers under nutrient-rich conditions (Fig. 5). This consistency suggests that by breeding naturally multinucleate filamentous fungi under nutrient-rich conditions, it is possible to obtain strains with increased nuclear numbers and enhanced enzyme or compound production.

The number of nuclei in filamentous fungi varies greatly among species (61). For example, Neurospora crassa can exceed 100 nuclei, which correlates with its exceptionally rapid growth rate and high protein secretion capacity (62). An increase in the number of nuclei per cell is expected to enhance transcription and translation per cell, thereby improving enzyme production and secretion capacity (Fig. S6B). However, maintaining a high nuclear number requires a substantial amount of energy, and in conditions where external nutrient supply is insufficient, cells rapidly die. Thus, the tendency to maintain a high nuclear number could be a selective pressure unlikely to occur in nature. Koji-kin strains were bred specifically for high enzyme production in nutrient-rich environments, and we propose that some of them acquired the specialized trait of increased cell volume and nuclear number through selection. This study elucidates the evolutionary processes driven by breeding selection in A. oryzae and the underlying mechanisms contributing to high enzyme productivity. These findings not only offer new approaches for optimizing filamentous fungi for bioindustry applications but also lead to the elucidation of the molecular mechanisms regulating cell volume and nuclear number in multinucleated organisms.

Methods

Strains and Culture Conditions

The filamentous fungal strains used in this study are listed in Table S6. A. oryzae RIB40 expressing H2B-GFP, UtH2BG stain, is described previously (33). Minimal medium for Aspergillus species was shown in Table S7. When peptone, casamino acids, or yeast extract was added, the final concentration was 1%. Amino acids were added at a final concentration of 0.1%, B vitamins (biotin, pyridoxine hydrochloride, thiamin hydrochloride, riboflavin, PABA, and nicotinic acid) were added at 100 μM, and nucleotides (inosine and uridine) were added at 10 mM.

Genetic Manipulation

To delete target genes, genome editing plasmids and circular donor DNA plasmids were constructed as described previously (63). The sgRNA cassette was designed under the U6 promoter with the following target sequences: GTGACGATCTACGGTAACGC for the msyA gene, GATCAGTTGAGCGTCCCCTA for the msyB gene, and AGCTAGCGCAGCCGGTCTGC for the rseA gene. The genome editing plasmid was constructed by introducing the sgRNA expression cassette into the SmaI restriction site of the pRGE-gRT6 plasmid (63). For the donor plasmid, 1 kb fragments upstream and downstream of the target genes were amplified, ligated, and inserted into the linearized pUC19 vector (Takara Bio). Both plasmids were applied for transformation using the A. oryzae RIB40 strain. To amplify the RseA region, spanning 1 kb upstream to 1 kb downstream, the templates used were A. oryzae RIB40, TK-42, and TK-47 strains. The amplified fragments were inserted into the HindIII site of the pPTRI vector (TaKaRa) to construct the plasmid. Transformations were performed using the protoplast-PEG method according to the manufacturer’s protocol for pPTRI. Transformants were selected on the minimal medium plates containing 0.1 mg/l pyrithiamine. Integration of constructs was confirmed by PCR analysis.

Microscopy

Images were captured using an Axio Observer Z1 inverted epifluorescence microscope (Carl Zeiss) equipped with Plan-Apochromat objectives, an AxioCam 506 mono camera, and a Colibri.2 LED light source. The stage temperature was maintained at 30°C with thermoplate (Tokai Hit). Nuclei were stained with SYBR Green (Takara), mitochondrial ware stained with Rhodamine 123 (Invitrogen), cell membrane were stained with FM4-64 (Invitrogen). Images were collected using Zen (Carl Zeiss) and ImageJ software.

Image Analysis

The distribution pattern of nuclei was classified into the following three types; Class I: nuclei distribute at a constant interval without overlapping; Class II: nuclei align but sometimes overlap; Class III: nuclei scattered throughout hyphae but not aligned. Nuclei, hyphal size, and fluorescence intensity were quantified using Zen and ImageJ software. Z-stack 3D reconstruction and nuclear color coding were performed with Imaris (Oxford Instruments). Hyphal elongation distances were analyzed using the MTrackJ plugin in ImageJ. For Fig. 2E, nuclei were binarized, and those in non-increased hyphae were displayed in white based on size and circularity.

Transmission Electron Microscopy (TEM)

The culture samples were embedded in 1% water agar and fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.0) at room temperature for 2 hours. Sample preparation and sectioning were performed at the Institute of Medicine, University of Tsukuba. Observations were conducted using a JEM-1400 transmission electron microscope (JEOL).

Imaging with Microfluidic Devices

Microfluidic devices were fabricated as described previously (33). Two types of devices were constructed: a 2D observation device to visualize hyphal expansion (Fig. S1A) and a single-hypha isolation device for enzymatic activity measurement (Fig. S2A). Conidia suspensions were loaded into a 10 ml plastic syringe (SS-10ESZ, Terumo) and connected to polyethylene tubing (inner diameter 0.38 mm, outer diameter 1.09 mm, BD Intramedic). Air was expelled from the syringe, and the suspension was injected into the device inlet. The number of conidia flowing through the microfluidic channel was adjusted under a microscope. The syringe was then replaced with another containing medium, and the medium was delivered at the minimum flow rate using a syringe pump (YSP-101, YMC). The device was incubated at 30°C on a thermal stage (TOKAI HIT) and imaged using an Axio Observer Z1 microscope (Carl Zeiss). α-amylase activity in the device was measured using the EnzChek Ultra amylase assay kit (Invitrogen).

Enzyme Activity Assays

Fungal conidia 10⁴ were inoculated onto 25 ml of medium in a petri dish and statically cultured at 30 °C for 4 days. Protein concentration and enzymatic activities in the culture supernatant were measured. Protein quantification was performed using the Bradford assay (Protein Assay Kit, BIO-RAD), with bovine gamma globulin as the standard. Carboxymethyl cellulase (CMCase) activity was determined by measuring reducing sugars using the 3,5-dinitrosalicylic acid method. CMCase activity was assayed at 50 °C for 15 min in a reaction mixture containing 50 mM sodium acetate buffer (pH 5.0) and 1% carboxymethyl cellulose. One unit of enzyme activity was defined as the amount of enzyme required to release 1 μmol of reducing sugar (as glucose equivalent) per minute. α-amylase activity was measured using an α-Amylase Assay Kit (Kikkoman Biochemifa) according to the manufacturer’s instructions. Acid carboxypeptidase (ACP) activity was determined using the Acid Carboxypeptidase Assay Kit (Kikkoman Biochemifa) following the provided protocol.

RNA-seq Analysis

Sterilized cellophane was placed on agar plates, and conidia were inoculated. After culturing at 30 °C for 3 days, samples were frozen, homogenized with a mortar and pestle, and total RNA was extracted using the RNeasy Mini Kit (QIAGEN). Library preparation, sequencing, and partial data analysis were performed by the Faculty of Medicine, University of Tsukuba. Sequencing was conducted using 36 bp paired end reads on the Illumina NextSeq 500, and reads were mapped to the A. oryzae RIB40 reference genome using the CLC Genomic Workbench (QIAGEN). For laser microdissection, sterilized membrane slides (Carl Zeiss) were immersed in MM liquid, and conidia were inoculated onto the membrane. After static culturing at 30 °C for 3 days, sections were collected using a PALM MicroBeam laser microdissection system (Carl Zeiss). RNA was extracted from 500-700 sections respectively using the RNeasy Mini Kit. Library preparation and sequencing were performed by Takara Bio Inc., using 150 bp paired end reads on the Illumina NovaSeq 6000. Reads were mapped to the A. oryzae RIB40 reference genome using the CLC Genomic Workbench. Genes showing a ≥4-fold change in expression were considered significant, and gene ontology (GO) enrichment analysis of biological processes was performed using ShinyGO 0.81 (64). Heatmaps were visualized using the Python Seaborn package (65). Gene annotations were obtained from FungiDB.

Low Osmotic Pressure Shock Analysis

Conidia were spot inoculated onto the minimal medium containing 1 M NaCl and cultured at 30 °C for 3 days. Colony tips were treated with distilled water for 3 min, and the number of tip ruptures was counted. Hyphal tip swelling was assessed by staining hyphae cultured in the minimal medium liquid for 3 days with FM4-64, mounting them on a coverslip with a small volume of medium, and adding an equal volume of distilled water.

Raman Spectroscopy

The conidia were inoculated into minimal medium liquid in a glass-bottom dish and cultured at 30 °C for 3 days. Raman spectra were acquired using an XploRA PLUS Raman microscope, as described previously (32). Hyphae attached to the dish bottom were irradiated with a 532 nm laser at 0.93 mW for 30 seconds. Spike artifacts caused by cosmic rays were manually removed. Ten spectra were averaged, and background signals attributed to the medium and glass were subtracted.

Protein Structure Analysis

RseA protein structure was predicted using AlphaFold2 and visualized with PyMOL (v1.20). Domains were annotated using InterPro.

Atomic Force Microscopy (AFM)

The conidia were inoculated into minimal medium liquid on a Poly-L-Lysine-coated (Sigma Aldrich) glass-bottom dish (ibidi) and cultured at 30 °C for 3 days. We used a JPK Nanowizard 4 (Bruker) equipped with an inverted fluorescence microscope (Eclipse Ti2, Nikon). The temperature was maintained at 30 °C using a dish heater integrated into the stage. BL-AC40TS-C2 cantilevers (Olympus, spring constant approximately 0.1 N/m) were used for AFM imaging under the following conditions: QI mode, 1.5 × 1.5 μm or 0.5 × 0.5 μm scan size, 128 × 128 pixels, Z-length 1 μm, setpoint 0.1 nN, and Z speed 166 μm/s. The Young’s modulus was calculated using JPKSPM Data Processing software (Bruker). After averaging and baseline subtraction, Young’s modulus was calculated using a Hertz model with a triangular pyramid half-cone angle of 35°.

SNP analysis

The genome data of A. oryzae required for SNP identification was obtained from the National Center for Biotechnology Information (NCBI). and mapped to A. oryzae RIB40 reference sequence using BWA-MEM2 (64). The SNPs were listed by FreeBayes (67), and detailed comparative analysis of genomes was performed on Integrative Genome Viewer (IGV) manually.

RNA seq

The RNA seq data have been deposited in DDBJ as BioProject PRJDB19992.

Acknowledgements

This work was funded by MEXT KAKENHI grant number 21H02095, 21K19062 and Scientific Research on Innovative Areas “Post-Koch Ecology” grant number 22H04878 to NT. Ohsumi Frontier Science Foundation, Noda Institute for Scientific Research Grant, and Japan Science and Technology Agency (JST) ERATO grant number JPMJER1502 to NT. AFM measurement was supported by the World Premier International Research Center Initiative (WPI).

Additional files

Figure supplements

Table S1. Annotated data of RNAseq in A. oryzae RIB40 and RIB915, and A. nidulans grown in the minimal medium with or without yeast extract.

Table S2. Annotated data of RNAseq in A. oryzae RIB40 thick or thin hyphae.

Table S3. Annotated data of RNAseq common in upregulated gene in A. oryzae RIB915 grown with yeast extract and in A. oryzae RIB40 thick hyphae.

Table S4. SNP analysis of ORFs in Clades F between TK-32 and TK-38.

Table S5. SNP analysis of ORFs in Clades G between TK-41 and TK-47.

Table S6. Strains used in this study.

Table S7. Composition of minimal medium.

Movie 1. 3D images of hyphae without increased nuclei (left) and with increased nuclei (right) in A. oryzae RIB40. Each nucleus is indicated with different colors by Imaris soft. Scale bar: 5 μm.

Movie 2. Emergence of thick hyphae with increased nuclei by branching in A. oryzae RIB40 expressing H2B-GFP. Scale bar: 20 μm. Elapsed time is indicated in min.

Movie 3. Successive nuclear division within the newly emerged thick branched hypha in A. oryzae RIB40 expressing H2B-GFP. Scale bar: 10 μm. Elapsed time is indicated in min.

Movie 4. Mycelial growth of hyphae with increased nuclei (green) and without increased nuclei (white). Scale bars: 200 μm. Elapsed time is indicated in h.

Movie 5. Amylase activity was monitored with a fluorescent substrate under conditions where a single thick or thin hypha grew in a microfluidic channel. Scale bar: 10 μm. Elapsed time is indicated in min.

Movie 6. Each hypha was dissected using laser microdissection and collected separately.