GPI-Anchored Protein Homolog IcFBR1 Functions Directly in Morphological Development of Isaria cicadae

Isaria cicadae is a famous edible and medicinal fungus in China and Asia. The molecular basis of morphogenesis and synnemal formation needs to be understood in more detail because this is the main source of biomass production in I. cicadae. In the present study, a fruiting body formation-related gene with a glycosylphosphatidylinositol (GPI) anchoring protein (GPI-Ap) gene homolog IcFBR1 was identified by screening random insertion mutants. Targeted deletion of IcFBR1 resulted in abnormal formation of synnemata, impairing aerial hyphae growth and sporulation. The IcFBR1 mutants were defective in the utilization of carbon sources with reduced polysaccharide contents and the regulation of amylase and protease activities. Transcriptome analysis of ΔIcfbr1 showed that IcFBR1 deletion influenced 49 gene ontology terms, including 23 biological processes, 9 molecular functions, and 14 cellular components. IcFBR1 is therefore necessary for regulating synnemal development, secondary metabolism, and nutrient utilization in this important edible and medicinal fungus. This is the first report illustrating that the function of IcFBR1 is associated with the synnemata in I. cicadae.


Introduction
Isaria cicadae, also called cicada flower in China, is an insect-derived fungus parasitic on cicada nymphs that has been used as food and herbal medicine for centuries [1]. Wild I. cicadae acts on a variety of diseases and is used to treat children with convulsions, children's night terrors, palpitations, malaria, and so on, in the practice of traditional Chinese medicine [2]. In recent years, scientists have investigated various biological activities of I. cicadae and identified useful compounds, such as amino acids, cordycepin, polysaccharides, and mannitol, which are similar to compounds in Cordyceps sinensis [3]. Some substances are helpful to human health by regulating the immune system, activating macrophages [4,5], promoting phagocytosis of macrophages [6], improving the immune response of the liver, kidney, spleen, and thymus [7], having antiaging properties [8], protecting renal function; alleviating glomerulosclerosis and improving chronic renal failure [9,10], alleviating renal anemia, decreasing blood glucose [11], having antitumor properties [12], and improving vision [13].
I. cicadae can form a special type of asexual fruiting body referred to as coremia or synnema [4]. Mature synnemata are composed of large aggregations of mostly parallel hyphae to form a yellow-to-brown pigmented bundle of hyphae growing erect from the

ATMT Transformation and TAIL-PCR
The wild-type (WT) strain 2-2 of I. cicadae was used for random insertion mutant construction by plasmid pBHt2. The fungal transformation process followed the method of Chen et al. [22]. DNA flanking transfer DNA (T-DNA) was determined using the TAIL-PCR method according to the procedure by Mullins et al. [23].

Vector Construction and Verification of Mutant
The upstream and downstream 1.0-1.2 kb fragments of the IcFBR1 gene were amplified with primers Up-F/Up-R and Dn-F/Dn-R from the wild-type strain genome. Using the plasmid pBHt2 as a template, the hygromycin B resistance gene HPH of about 1.4 kb was obtained using primers HPH-F and HPH-R. These three DNA fragments were fused to the HindIII and XhoI-linearized pCAMBIA1300 using a Fusion Enzyme Kit (Vazyme, Nanjing, China) ( Figure 1A). The knockout cassette was transformed into the wild-type strain via Agrobacterium tumefaciens-mediated transformation (ATMT) according to the reported method [24,25]. Briefly, mutants were screened on a selective medium containing 350 µg/mL hygromycin B. Then, mutants were verified by PCR with primers ck-F/ck-R to determine whether HPH recombined into the IcFBR1 deletion site ( Figure 1B) to create a target gene knockout mutant. Finally, the insertion copies of HPH in the mutant were identified by Southern blotting following methods used by Chen et al. [22]. Briefly, 5 µg of genomic DNA of each sample was digested overnight at 37 • C with XbaI and HindIII, respectively. Digested DNA fragments were separated by Electrophoresis in 1% agarose gel and blotted onto nylon membrane Hybond-N + (Amersham Pharmacia Biotech, Amersham, UK). Using the knockout vector as a template, a 1065 bp DNA fragment containing a 633 bp partial genomic DNA and a 432 bp partial HPH gene was amplified with primers S-F/S-R and used as a probe template. The probe was labeled using the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche, Mannheim, Germany) described in the manufacturer's manual. A single copy of knockout mutants was also confirmed by the Q-PCR method according to the assay described by Lu et al. [24]. Briefly, when tubulin is the reference gene, a single copy of the target gene was determined as ∆∆CT = 0.9~1.3, where ∆∆CT = (CT HPH -CT tubulin-m ) -(CT gene -CT tubulin-w ), CT HPH is the CT value of HPH in the mutant, CT tubulin-m is the CT value of tubulin in the mutant, CT gene is the CT value of target gene in the wild-type strain, and CT tubulin-w is the CT value of tubulin in the wild-type strain.

ATMT Transformation and TAIL-PCR
The wild-type (WT) strain 2-2 of I. cicadae was used for random insertion mutant construction by plasmid pBHt2. The fungal transformation process followed the method of Chen et al. [22]. DNA flanking transfer DNA (T-DNA) was determined using the TAIL-PCR method according to the procedure by Mullins et al. [23].

Vector Construction and Verification of Mutant
The upstream and downstream 1.0-1.2 kb fragments of the IcFBR1 gene were amplified with primers Up-F/Up-R and Dn-F/Dn-R from the wild-type strain genome. Using the plasmid pBHt2 as a template, the hygromycin B resistance gene HPH of about 1.4 kb was obtained using primers HPH-F and HPH-R. These three DNA fragments were fused to the HindIII and XhoI-linearized pCAMBIA1300 using a Fusion Enzyme Kit (Vazyme, Nanjing, China) ( Figure 1A). The knockout cassette was transformed into the wild-type strain via Agrobacterium tumefaciens-mediated transformation (ATMT) according to the reported method [24,25]. Briefly, mutants were screened on a selective medium containing 350 µg/mL hygromycin B. Then, mutants were verified by PCR with primers ck-F/ck-R to determine whether HPH recombined into the IcFBR1 deletion site ( Figure 1B) to create a target gene knockout mutant. Finally, the insertion copies of HPH in the mutant were identified by Southern blotting following methods used by Chen et al. [22]. Briefly, 5 µg of genomic DNA of each sample was digested overnight at 37 °C with XbaI and HindШ, respectively. Digested DNA fragments were separated by Electrophoresis in 1% agarose gel and blotted onto nylon membrane Hybond-N + (Amersham Pharmacia Biotech, America). Using the knockout vector as a template, a 1065 bp DNA fragment containing a 633 bp partial genomic DNA and a 432 bp partial HPH gene was amplified with primers S-F/S-R and used as a probe template. The probe was labeled using the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche, Mannheim, Germany) described in the manufacturer's manual. A single copy of knockout mutants was also confirmed by the Q-PCR method according to the assay described by Lu et al. [24]. Briefly, when tubulin is the reference gene, a single copy of the target gene was determined as ∆∆CT = 0.9~1.3, where ∆∆CT = (CTHPH -CTtubulin-m) -(CTgene -CTtubulin-w), CTHPH is the CT value of HPH in the mutant, CTtubulin-m is the CT value of tubulin in the mutant, CTgene is the CT value of target gene in the wild-type strain, and CTtubulin-w is the CT value of tubulin in the wild-type strain. For the complementation experiments, a 2.5 kb native IcFBR1 gene (containing a promoter and coding sequence (CDS), but no termination codon) was cloned from the I. cicadae genome by primers FBR1-CF/FBR1-CR and inserted into the PstI-linearized vector For the complementation experiments, a 2.5 kb native IcFBR1 gene (containing a promoter and coding sequence (CDS), but no termination codon) was cloned from the I. cicadae genome by primers FBR1-CF/FBR1-CR and inserted into the PstI-linearized vector pEC 1 . pEC 1 was derived from pCAMBIA1300 containing a neomycin gene (G418) and green fluorescent protein gene (GFP). The complementary vector was transformed into the mutant ∆Icfbr1 via ATMT. The complementary mutants of ∆Icfbr1-c were selected on a selection medium containing 1100 µg/mL neomycin and confirmed by Southern blotting. All primers used in this paper are listed in Table S1 (Supporting Material).

Effects of IcFBR1 on Nutrition Utilization and Phenotypic Characterization
For the fungal growth assay, the WT, ∆Icfbr1, and the complimentary strain ∆Icfbr1-c were cultured on 7 cm diameter plates containing 17.5 mL PDA medium at 25 • C for 10 days. To investigate germination of conidia, 20 µL of the spore suspensions (1 × 10 5 conidia/mL in liquid PDB medium) was separately smeared on PDA medium and water agar, respectively, and incubated at 25 • C for 24 h.
The utilization of carbon sources affected by IcFBR1 was as follows: single 1% carbon sources (glucose, sucrose, fructose, lactose, maltose, soluble starch, sodium carboxymethyl cellulose, ethanol, or glycerol) were separately configured in minimal media (MM) [NaNO 3 6 g/L, KCl 0.52 g/L, MgSO 4 ·7H 2 O 0.52 g/L, KH 2 PO 4 0.25 g/L, Agar 15 g/L]. A piece of the colony of each strain tested was incubated on 7 cm plates containing 17.5 mL configured medium at 25 • C with a 12 h/12 h light/dark cycle for 10 days. Stress tests were performed as follows: A drop of the spore suspensions (30 µL) was dripped onto separate PDA plates with 0.01% sodium dodecyl sulfate (SDS), 200 µg/mL Calcofluor white (CFW), or 200 µg/mL Congo red, and incubated at 25 • C in the dark for 7 days. All experiments were performed at least twice, with three replicates of each treatment.

Observation of Fluorescence Fusion Proteins in I. cicadae and Quantitative Real-Time PCR
GFP expression of the complementary strain ∆Icfbr1-c was observed using a fluorescence microscope (Filter: FITC and TRITC) (Nikon Eclipse 80i, Nikon, Japan) after culturing on a PDA plate or rice substrate according to conditions described in Section 2.8. In order to ensure the reliability of the results, RT-PCR was also performed. Total RNAs were extracted from spores and mycelia with Trizol following the manufacturer's procedure (TaKaRa Bio, Beijing, China). Reverse transcription of cDNA was conducted using the PrimeScript TM RT reagent kit with gDNA Eraser (TaKaRa, Japan). Expression of IcFBR1 was performed on a Real-Time PCR Detection System Mastercycler (Eppendorf, Germany) with an SYBR Premix Ex Taq kit (Tli RNaseH Plus, TaKaRa Bio, Beijing, China) according to the manufacturer's instructions. Tubulin was selected as the reference gene, and IcFBR1 gene expression was calculated using 2 −∆CT , where ∆CT = CT gene -CT Tubulin . The primers used in this experiment are listed in Table S1 (Supporting Material).

Lipid Droplet Stain with BODIPY
BODIPY was dissolved to 1 µg/mL with 2% DMSO for lipid droplet staining. The conidia were stained by BODIPY solution for 5 min, then observed by fluorescence microscopy.

Assay of Determination of Polysaccharide and Enzyme Activity
A total of 10 5 spores of each strain tested were transferred to a 500 mL flask containing 200 mL potato dextrose broth, incubated at 25 • C while shaking at 150 rpm for 7 days. The mycelia of each strain were harvested by filtering with 3 layers of gauze and washed with distilled water 3 times. Subsequently, the mycelia were dried by using the freeze dryer. The total polysaccharide content of mycelia was determined using the phenol sulfuric acid method, according to Zhang et al. [26]. The activities of enzymes were investigated by inoculating a 0.5 cm diameter of mycelium on an enzyme-induced medium and incubating at 25 • C in the dark for 7 d to observe the enzymatic hydrolysis cycle. Amylase was assayed on a medium with 1% starch, 1.5% agar, and 0.1% potassium iodide solution. Protease was assayed on a medium containing 1% milk and 1.5% agar. Chitinase was detected on medium with 1.5 × 10 −4 g/L Bromocresol purple, 0.03% MgSO 4 •7H 2 O, 0.3% (NH 4 ) 2 SO 4 , 0.2% KH 2 PO 4 , 0.1% citric acid, 0.45% colloidal chitin, and 200 µL/L Tween-80.

Effects of IcFBR1 on the Growth and Formation of Synnema
The wild-type strain 2-2, knockout mutant ∆Icfbr1, and the complementary strain ∆Icfbr1-c were cultured on PDA medium in a growth chamber at 25 • C with a 12 h light/12 h dark cycle for phenotypic observation. Synnema was induced in rice or flour medium, according to previous research [21,27]. Briefly, fungal strains were inoculated on rice or flour medium in darkness at 25 • C for one week, then transferred to 25 • C with a day/night cycle of 12 h light/12 h darkness for two weeks to form synnemata.

Transcriptome Analysis
Mycelia of strains tested were cultured on rice medium at 25 • C for 7 days, harvested, and ground into a fine powder in liquid nitrogen using a mortar and pestle for RNA isolation. Total RNA was extracted using Trizol, according to the manufacturer's instructions (TaKaRa, Japan). Then, the NovaSeq 6000 system was used for transcriptome sequencing. RNA-Seq was performed using three biological replicates of each sample. Fast QC was used to detect the quality of transcriptome data [28]. Trimmomatic was used to compare transcriptome data after removing low-mass bases [29]. HISAT was used to compare transcriptome data to the reference genome of WT CCAD02. Htseq-count was used to count the number of genes [30]. DEseq2 (R2) was used to screen differentially expressed genes (DEGs) with a preset threshold of p < 0.05 and log 2 FC (fold change) > 1 [31]. Blast2 was used to annotate the Gene Ontology (GO) of DEGs [32]. Enrichment analysis of GO annotation results was carried out using WEGO online tool [33]. Differences in KEGG pathway analysis were performed with the KEGG database (https://www.genome.jp/kaas-bin/kaas_main, accessed on 13 July 2022) [34].

Statistical Analysis
The significance of treatments was determined by DPS 9.5 software, and means were separated by a least significant difference (LSD) test (p ≤ 0.05).

The IcFBR1 Gene Screening in I. cicadae
A random insertion mutation library was constructed with plasmid pHt2 by ATMT in I. cicadae. Six mutants were obtained with reduced aerial hyphae growth compared with the phenotype of WT strain 2-2. Southern blot confirmed that single-copy insertion induced the change in hyphae growth in these mutants. The gene location of the genomic insertion in mutant ChGD was screened using the TAIL-PCR method, and a 260 bp DNA fragment was obtained after sequencing. BLAST research in the Genbank database showed that the gene fragment had 90% homology with Isaria fumosorosea ARSEF 2679 hypothetical protein (ISF_08023, GenBank accession No.: XM_018851626). After searching the genome of I. cicadae (GenBank accession no. MWMN00000000), we obtained its homologous gene. Since it is a fruiting body-related gene, we named it IcFBR1. The sequence length of the IcFBR1 gene is 916 bp, putatively encoding 235 amino acids (Supporting Information Supplementary File S1). Amino acid BLAST using NCBI showed more than 50% identity with a putative GPI-AP in the fungal genomes KAF5707377 of Fusarium mundagurra, XP_008598424 of Beauveria bassiana, TQV90811 of Cordyceps javanica, OAQ78299 of Purpureocillium lilacinum, and OAQ58852 of Pochonia chlamydosporia. Conserved domains search in the GenBank dataset showed that it belongs to the lytic polysaccharide monooxygenase (LPMO) auxiliary-like protein (Supporting Material Figure S1A). Members of this family are related to LPMOs but have no enzymatic activity in the oxidative cleavage of polysaccharides [19]. Predicting structure analysis indicated that this protein has two transmembrane structures (Supporting Material Figure S1B). GPI prediction analysis showed a GPI-anchor signal sequence at C-terminal, and there is a signal peptide at the N-terminal composed of 17 amino acids, MKLNLIISLALAGLAAA (http://gpi.unibe.ch/, accessed on 13 July 2022).

Validation of Gene Knockout Mutants
The mutant ∆Icfbr1 was obtained by homologous recombination. The primers ck-F/ck-R were used to detect whether HPH replaced IcFBR1. The bands of ∆Icfbr1 were different from that of WT, but the same as that of the control (pK-hph) ( Figure 1B), demonstrating that the IcFBR1 gene was successfully replaced by the HPH gene. Moreover, we used Southern blotting to detect the gene copy number and showed that the mutant ∆Icfbr1 and wild-type strain 2-2 both had one band with different sizes (Supporting Material Figure  S2). We select NO. 2 and NO. 5 mutants as single-copy mutants and use mutant NO. 2 for further research. Single-copy knockout mutants were also confirmed by Q-PCR in that the ∆∆CT value is 1.1 and 1.2, respectively. According to Lu et al. [24], this means that the gene is a single copy. The results above indicated that the strains gained in this experiment were consistent with the expected results but also indicated that the HPH gene exactly substituted for the IcFBR1 gene.

IcFBR1 Is Involved in the Development of I. cicadae
The complementary strain ∆Icfbr1-c was gained by transforming the complementary vector with fused IcFBR1 and GFP gene into knockout mutant ∆Icfbr1.
To investigate the possible effect of IcFBR1 on colony growth, sporulation, and conidia germination, the mutants and wild-type strain were cultured on PDA at 25 • C for 10 days. After measuring, we found that the colony diameters of ∆Icfbr1 were larger than the WT and the complementary ∆Icfbr1-c (p < 0.05) (Figure 2A), but the hyphae were very sparse and short, and the colony thickness of ∆Icfbr1 were thinner than the wild-type strain about 3 times (Supporting Material Figure S3). Compared with the wild type, sporulation in ∆Icfbr1 was significantly diminished by almost twofold ( Figure 2B). When spores germinated in a water agar medium, the germination rate of ∆Icfbr1 was slower than the wild-type strain at 8 h (p < 0.01) ( Figure 2C), with the same tendency on PDA (Supporting Material Figure S4). The results showed that the IcFBR1 gene is crucial for the colony growth and sporulation development of I. cicadae.

IcFBR1 Is Involved in the Utilization of Carbon Sources by I. cicadae
To measure whether IcFBR1 affects the utilization of carbon sources, fresh colonies of the wild-type strain, ∆Icfbr1, and ∆Icfbr1-c were inoculated on MM media added with different carbon sources and incubated at 25 • C, with 12 h light/12 h dark cycles for 7 days. The morphological characteristics and colony diameters are shown in Figure 3A. After measuring the colony diameters, we found that on the flour medium, the growth of ∆Icfbr1 was faster than the wild type and complementary ∆Icfbr1-c (p < 0.05). Measuring the colony diameters on media with sucrose, maltose, and fructose, we found that the growth of ∆Icfbr1 was slower than the wild-type strain and complementary strain ∆Icfbr1-c (p < 0.05). On the lactose medium, the colony diameters of the three strains had no differences, but the aerial hyphae of ∆Icfbr1 were significantly reduced compared with the WT and complementary ∆Icfbr1-c. In other media, there was no obvious difference ( Figure 3B). The results suggested that the IcFBR1 gene may influence the utilization of partial carbon sources in I. cicadae.

IcFBR1 Has No Effect on the Integrity of the Cell Wall of I. cicadae
When spores were inoculated on different cell wall stress media for 7 days, there was no significant difference in the growth of the WT, ∆Icfbr1, or complementary ∆Icfbr1-c (Supporting Material Figure S5), indicating that IcFBR1 does not affect cell wall integrity.

Expression of GFP-Tagged IcFBR1 in Different Stages of I. cicadae Life Cycle
Observation of the GFP fluorescence of the spores and hyphae of the complementary strain ∆Icfbr1-c under the microscope indicated that IcFBR1 was expressed in cells of hyphae, phialides, and conidia ( Figure 4B-E) when incubated on PDA. GFP was highly expressed in a type of globular structure in mature conidia but was very faint in young conidia ( Figure 4B). To confirm if these globular structures are lipid droplets, conidia of the wild-type strain were stained with BODIPY and observed by fluorescence microscopy. The results showed that the globular structures in mature conidia are lipid droplets ( Figure 4A). These results illustrated that the IcFBR1 protein was located on lipid droplets in mature conidia. GFP expression in phialides was very strong ( Figure 4C). In hyphae, GFP was highly expressed in the condensed protoplast structure of mature cells (curved hyphae cells) and cells of conidiophore but was very faint in young cells in hyphae ( Figure 4D-F). To investigate the GFP expression on synnemata, a hand section was made and observed under a fluorescence microscope ( Figure 4F-L). GFP was highly expressed in young synnemal cells compared with the mature synnemal cells ( Figure 4H,I), but GFP can be observed at the surface of mature synnemal cell wall, especially at the sites of branch and septa ( Figure 4J,K). Other cells of highly expressed GFP are mature cells growing from the outside surface of synnema ( Figure 4L). These results indicate that IcFBR1 proteins are located in mature cells and production structures.  The RNA of spores, synnemata, and hyphae of the wild-type strain were extracted, and RT-PCR verification was conducted after reverse transcription. Compared with the internal reference gene tubulin, the expression of the IcFBR1 gene in the spore and synnemata was much higher than that in the hyphae ( Figure 5). There are 17.7, 13.2, and 8.2 times more IcFBR1 gene expressions in synnemata, conidiophore, and conidium than in mycelium, respectively.

Effects of IcFBR1 on I. cicadae Key Metabolites
Polysaccharides, cordycepin, and adenosine are important active substances of I. cicadae. In the present research, we found there are no effects on the contents of cordycepin and adenosine, but the contents of polysaccharides in ∆Icfbr1 were reduced when compared with the wild-type strain and ∆Icfbr1-c. The polysaccharide content of mutant ∆Icfbr1 was 2.84%, less than 3.91% in the wild-type strain and 3.87% in ∆Icfbr1-c. It suggested that IcFBR1 had an effect on the contents of polysaccharides.
We compared the activities of amylase, protease, and chitinase of the wild type, ∆Icfbr1, and complementary ∆Icfbr1-c. Our results showed that the hydrolytic circle of ∆Icfbr1 in the starch medium was significantly larger (p < 0.01) than that of the wild type (1.84 ± 0.09 cm of ∆Icfbr1 compared with 1.33 ± 0.05 cm of the wild-type strain 2-2) and complementary ∆Icfbr1-c, but the hydrolytic circle of the ∆Icfbr1 in milk powder medium was significantly smaller than that of the WT and complementary ∆Icfbr1-c (0.97 ± 0.05 cm diameter of ∆Icfbr1 compared with 1.20 ± 0.00 cm of wild the wild-type 2-2 and 1.18 ± 0.02 cm of ∆Icfbr1-c) (p < 0.01). There was no significant difference in the diameter of the hydrolytic circle on the chitin medium (1.67 ± 0.05 cm diameter of ∆Icfbr1 and wild-type strain 2-2) (Supporting Material Figure S6). The results showed that IcFBR1 inhibits the activity of amylase in I. cicadae and has no effect on the synthesis of chitinase but promotes the

Effects of IcFBR1 on I. cicadae Key Metabolites
Polysaccharides, cordycepin, and adenosine are important active substances of I. cicadae. In the present research, we found there are no effects on the contents of cordycepin and adenosine, but the contents of polysaccharides in ∆Icfbr1 were reduced when compared with the wild-type strain and ∆Icfbr1-c. The polysaccharide content of mutant ∆Icfbr1 was 2.84%, less than 3.91% in the wild-type strain and 3.87% in ∆Icfbr1-c. It suggested that IcFBR1 had an effect on the contents of polysaccharides.
We compared the activities of amylase, protease, and chitinase of the wild type, ∆Icfbr1, and complementary ∆Icfbr1-c. Our results showed that the hydrolytic circle of ∆Icfbr1 in the starch medium was significantly larger (p < 0.01) than that of the wild type (1.84 ± 0.09 cm of ∆Icfbr1 compared with 1.33 ± 0.05 cm of the wild-type strain 2-2) and complementary ∆Icfbr1-c, but the hydrolytic circle of the ∆Icfbr1 in milk powder medium was significantly smaller than that of the WT and complementary ∆Icfbr1-c (0.97 ± 0.05 cm diameter of ∆Icfbr1 compared with 1.20 ± 0.00 cm of wild the wild-type 2-2 and 1.18 ± 0.02 cm of ∆Icfbr1-c) (p < 0.01). There was no significant difference in the diameter of the hydrolytic circle on the chitin medium (1.67 ± 0.05 cm diameter of ∆Icfbr1 and wild-type strain 2-2) (Supporting Material Figure S6). The results showed that IcFBR1 inhibits the activity of amylase in I. cicadae and has no effect on the synthesis of chitinase but promotes the synthesis of protease. The main components of the insect epidermis are protein and chitin, 70% of which is protein, meaning that protein is the main component of the insect epidermis.
Secreting a large amount of protease is very important in the process of entomopathogenic fungi infecting insects [35]. Our result infers that the protease function of IcFBR1 may be essential in pathogenicity to the host.

IcFBR1 Is Involved in the Formation of Synnemata
Synnemata are the main parts of this kind of fungus that people process and eat; therefore, whether synnemata can be produced is extremely important for cultivating edible fungi. To check if synnemal formation was affected by IcFBR1, the wild-type strain, ∆Icfbr1, and complementary ∆Icfbr1-c were incubated on flour and rice culture. In wild-type strain 2-2, some mycelium curled up and aggregated to form hyphal knots (Supporting Material Figure S7) after vegetative mycelium covers the surface of the substrates. A large number of mycelia then grow in parallel and adhere together to form synnemata from the hyphal knots. However, ∆Icfbr1 did not produce synnemata even though there was no hyphal knots formation on the surface of the substrate. The results indicated that the IcFBR1 gene had a strong effect on the formation of the synnemata ( Figure 6); IcFBR1 gene defection can block the synnemal formation at the primary synnemata differentiation stage. synthesis of protease. The main components of the insect epidermis are protein and chitin, 70% of which is protein, meaning that protein is the main component of the insect epidermis. Secreting a large amount of protease is very important in the process of entomopathogenic fungi infecting insects [35]. Our result infers that the protease function of IcFBR1 may be essential in pathogenicity to the host.

IcFBR1 Is Involved in the Formation of Synnemata
Synnemata are the main parts of this kind of fungus that people process and eat; therefore, whether synnemata can be produced is extremely important for cultivating edible fungi. To check if synnemal formation was affected by IcFBR1, the wild-type strain, ∆Icfbr1, and complementary ∆Icfbr1-c were incubated on flour and rice culture. In wildtype strain 2-2, some mycelium curled up and aggregated to form hyphal knots (Supporting Material Figure S7) after vegetative mycelium covers the surface of the substrates. A large number of mycelia then grow in parallel and adhere together to form synnemata from the hyphal knots. However, ∆Icfbr1 did not produce synnemata even though there was no hyphal knots formation on the surface of the substrate. The results indicated that the IcFBR1 gene had a strong effect on the formation of the synnemata ( Figure 6); IcFBR1 gene defection can block the synnemal formation at the primary synnemata differentiation stage.

Transcriptome Analysis
A total of 21,701,446 raw reads were obtained by sequencing the total RNA, and 21,700,017 clean reads were retained in mutant ∆Icfbr1. Moreover, a total of 23,137,568 raw reads were obtained by sequencing the total RNA in the wild-type strain, and 23,136,222 clean reads were retained after trimming the low-quality bases. The clean reads were mapped to the reference genome of I. cicadae 2-2 and showed that 89.73% of ∆Icfbr1 RNA-Seq reads were mapped to the genome of I. cicadae 2-2, and 90.73% of wild-type reads were mapped to the genome of I. cicadae 2-2. In total, 378 differentially expressed genes, including 131 upregulated genes and 247 downregulated genes, were identified in the IcFBR1 mutant compared with the I. cicadae wild type ( Figure 7A).

Discussion
Entomopathogenic fungi, such as Cordyceps sinensis, have been used for medicinal purposes for several centuries, particularly in China, Japan, and other Asian countries [36].
As an important member of the Cordyceps family, I. cicadae has attracted research attention and has been utilized in traditional medicine for a long time [10]. Functional analysis of numerous genes in C. militaris has been conducted in recent years, including CmWC-1 [37], CmVVD [38], CmSnf1 [39], and so on. In contrast, there are few reports on gene functions in I. cicadae. In this study, we found that IcFBR1 affects the growth and development of I. cicadae and, most critically, that it may function directly in the formation of synnemata. For I. cicadae, growing hyphae in the early period and producing synnemata in the late period are extremely critical steps. Results in the present study indicated that IcFBR1gene could influence differentiation at the early stage of synnemal formation. This is the first report in which the gene function is related to synnemata (coremium) growth.
In general, LPMOs are copper-dependent enzymes that degrade polysaccharides through an oxidative mechanism. Recently, a new family of proteins (X325) with the LPMO fold having a His brace structure overturned the accepted thinking within the field of polysaccharide degradation and illustrated that these proteins are not LPMOs [19]. Reports indicated that the proteins in this family might regulate fungal growth in P. anserine [19] and control hyphal cell-cell fusion and maintenance of mutualistic interaction in an Epichloë festucae [40] and ryegrass system, as well as copper acquisition in Cyptococcus neoformans [41]. In the present study, a conserved domain search of IcFBR1 in GenBank showed that this gene has an LPMO-auxiliary-like domain shared with X325 of Laetisaria arvalis. Loss-of-function of IcFBR1 resulted in defects in coremium formation and reduction in conidial formation, but no phenotype of "crippled growth", "impaired for cell-cell fusion", or "effect of pathogenicity" [19]. Our results provide new evidence that this new protein family can regulate synnemata development in I. cicadae. The mechanism of IcFBR1 in regulating development will be further studied in the future.
Many genes play important roles in the formation and development of fruiting bodies, and their absence will have adverse effects on fruiting body development, even resulting in no fruiting bodies. For example, exg1 is involved in fruiting body development in Lentinula edodes [42], and FruA is an essential transcription factor for fruiting body development in Myxococcus xanthus [43]. In Flammulina velutipes, fv-pda is specifically expressed through the entirety of fruiting body development, and the transcript is abundant despite mature fruiting bodies. This suggests that chitin deacetylase (CDA) functions directly in the process of fruiting in F. velutipes [44]. In the present study, our results indicate that IcFBR1 regulated the colony growth, development of sporulation and synnemal formation, and secondary metabolism of I. cicadae. By observing the fluorescence, IcFBR1 can be basically localized on the plasma membrane but can localize at the inner surface of the cell wall in the mature synnemata, which is consistent with the structural analysis that this kind of protein has a GPI domain. Post-translation regulation may result in a variety of locations in different morpho-genesis stages. Numerous biological processes were affected by the disruption of IcFBR1, as illustrated by the transcriptome analyzed in this research. Some genes involved in cellular components composing cell structures, including external encAPSulating structure, cell wall, cell periphery, and plasma membrane, are absent or significantly downregulated in the IcFBR1 mutant, and these components are very important for fungal growth and differentiation. Through analysis of the transcriptome, we found that IcFBR1 seriously influenced the steroid biosynthesis pathway. Some genes involved in the MAPK signal pathway and calcium-transporting pathway were sharply downregulated when IcFBR1 defected. The roles of the signal pathway in synnemata formation need to be investigated in the future. The ability to produce synnemata is extremely important in artificial breeding, and ∆Icfbr1 produced no synnemata under the same conditions. Therefore, we think IcFBR1 is very important to I. cicadae.
In general, simple multicellularity in fungi is typically involved in linear processes such as vegetative mycelia, in which all cells are in direct contact with the external environment to absorb nutrients and O 2 , while complex multicellularity is defined as three-dimensional differentiated structure formation with a spatially and temporally integrated developmental program. In PDA and other media plates, I. cicadae grow only in a simple multicellular manner in which the growing hyphae rarely adhere to each other, and no synnemata are produced. In artificial substrates, however, such as rice and certain culture conditions, a large number of conidia are produced after vegetative hyphae cover the surface of the substrates. Then some hyphae begin to curl up and aggregate to form the primary hyphal knot. Synnemata grow up erect to the substrate along with the extension of primary hyphal knots. A large number of conidiophores and conidia are then produced on the surface of the synnemal stalks. Considering that the genetic mechanisms of synnemata are not yet fully understood, I. cicadae may be an excellent model to study the development of complex multicellularity since simple and complex multicellularity coexist in this species under certain conditions. In this study, high expression of IcFBR1 in curved hyphae cells indicates that the IcFBR1 gene may be involved in the formation of hyphal knots, the preliminary structure of synnemata. Mutants with a defective IcFBR1 gene exhibited a simple multicellular stage (clonal formation in plates) that can undergo morphogenesis but failed in synnema formation (complex multicellular stage), indicating that the growth and development mechanisms of synnemata differ from vegetative mycelium and conidial sporulation on PDA plates, despite the fact that both are known as asexual development stages in I. cicadae.

Conclusions
A gene encoding GPI-APS homologous protein (IcFBR1) was cloned, and its functions were investigated in the present study in I. cicadae. We found that IcFBR1 regulates the formation of synnemata, hypha growth, sporulation, biosynthesis of polysaccharides, the enzyme activity of amylase and protease, and utilization of carbon sources in this important edible and medicinal fungus. This is the first report illustrating the roles of IcFBR1 in synnemata formation, fungal growth, sporulation, and polysaccharides biosynthesis in filamentous fungi, and it provides a foundation for future research.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8111152/s1, Figure S1: analysis of IcFBR1 functional domain, Figure S2: Southern blot of T-DNA insertion mutants and Icfbr1 null mutant ∆Icfbr1, Figure S3: colony thickness of gene mutant and the wild-type strain, Figure S4: spore germination rate of I. cicadae on PDA, Figure S5: growth of strains of I. cicadae inoculated on different cell wall stress media incubated at 25 • C in the dark for 7 days, Figure S6: hydrolytic circle of strains of I. cicadae on the different medium for enzyme activity test, Figure S7: morphology of hyphal knot effected by IcFBR1 Gene, Figure S8: (A) gene ontology (GO) hierarchical graph analysis of the DEGs: biological process (BP), (B) gene ontology (GO) hierarchical graph analysis of the DEGs: molecular function (MF), (C) gene ontology (GO) hierarchical graph analysis of the DEGs. Cellular Component (CC), File S1: DNA sequences of IcFBR1 and amino acid sequence encoded by the target gene, Table S1: oligonucleotide primers used in this study, Table S2: the differentially expressed genes (DEGs) between IcFBR1 mutant and Isaria cicadae wild type, Table S3: GO enrichment analysis of the differentially expressed genes (DEGs) between IcFBR1 mutant and Isaria cicadae wild type, Table S4: KEGG enrichment analysis of the differentially expressed genes (DEGs) between IcFBR1 mutant and Isaria cicadae wild type, Table S5: protein sequence and it NR functional annotation.