MaOpy2, a Transmembrane Protein, Is Involved in Stress Tolerances and Pathogenicity and Negatively Regulates Conidial Yield by Shifting the Conidiation Pattern in Metarhizium acridum

Opy2 is an important membrane-anchored protein upstream of the HOG-MAPK signaling pathway and plays important roles in both the HOG-MAPK and Fus3/Kss1 MAPK. In this study, the roles of MaOpy2 in Metarhizium acridum were systematically elucidated. The results showed that the MaOpy2 disruption significantly reduced fungal tolerances to UV, heat shock and cell-wall-disrupting agents. Bioassays showed that the decreased fungal pathogenicity by topical inoculation mainly resulted from the impaired penetration ability. However, the growth ability of ∆MaOpy2 was enhanced in insect hemolymph. Importantly, MaOpy2 deletion could significantly increase the conidial yield of M. acridum by shifting the conidiation pattern from normal conidiation to microcycle conidiation on the 1/4SDAY medium. Sixty-two differentially expressed genes (DEGs) during the conidiation pattern shift, including 37 up-regulated genes and 25 down-regulated genes in ∆MaOpy2, were identified by RNA-seq. Further analysis revealed that some DEGs were related to conidiation and hyphal development. This study will provide not only the theoretical basis for elucidating the regulation mechanism for improving the conidial yield and quality in M. acridum but also theoretical guidance for the molecular improvement of entomopathogenic fungi.


Introduction
Entomopathogenic fungi can penetrate directly into the hemocoel of the host to utilize its nutrition [1], which offers great potential for insect pest control in the insect population and a low possibility of inducing the resistance of insects [2][3][4]. However, some disadvantages have limited their wide application, such as the long time required to kill insect pests, their sensitivity to diverse environmental conditions and the high cost for production [5][6][7]. As the main infective unit of entomopathogenic fungi and the active components of mycopesticides, conidia can tolerate various types of environmental stress and lead to an epizootic development in the insect populations [8]. Therefore, the conidial quality and yield of entomopathogenic fungi directly determine the efficiency of mycopesticides and the production cost. Elucidating the regulatory mechanism for improving the conidial yield and quality (stress tolerance, virulence, etc.) is expected to provide a theoretical basis which will promote the development of mycopesticides.
Two conidiation patterns, the normal conidiation pattern and the microcycle conidaition pattern, are found in most filamentous fungi [9,10]. Normal conidiation must go through a period of hyphae elongation and then form conidiophores at the tip of the long hyphae; however, microcycle conidiation can bypass the long hyphae growth and produce conidia directly [10]. These two conidiation patterns could be interconvertible under some specific circumstances [11][12][13][14]. In the locust-specific pathogenic fungus, Metarhizium acridum, microcycle conidiation exhibited a greater potential in large-scale applications than normal conidiation [15]. However, the mechanism of this shift remains unclear. Therefore, elucidating the underlying mechanism(s) of the conidiation pattern shift in order to improve the conidial productivity and quality is of great importance in entomopathogenic fungi.
As a cell transmembrane protein in fungi, Opy2 plays distinct roles in different species. In Saccharomyces cerevisiae, Opy2 participates in the SHO1 branch of the high osmolarity glycerol (HOG) pathway and is identified as an osmosensor interacting with Ste50 [16]. In Candida albicans, CaOpy2 plays a crucial role in cell wall stress tolerance and is essential for the phosphorylation of Cek1 [17]. In Metarhizium robertsii, an alternative transcription start site is achieved, and the regulation of the MrOpy2 transcription level affects the fungal lifestyle and contributes to virulence [18]. In Magnaporthe oryzae, MoOpy2 plays essential roles in pathogenicity, turgor pressure, appressorium formation, conidiation and hyphal growth [19]. However, in the entomopathogenic fungus M. acridum, the biological functions of MaOpy2 have not been systematically elucidated.
Here, the functions of Opy2 were identified in M. acridum. It was found that MaOpy2 plays important roles in fungal growth, stress tolerance and fungal pathogenicity in M. acridum. Importantly, we present that MaOpy2 disruption resulted in an increased conidial yield through the regulation of the conidiation pattern shift. RNA-seq showed that MaOpy2 governs the conidiation pattern shift by regulating some gene expressions related to hyphae growth and conidiation in M. acridum.

Strains and Vector Construction
For gene deletion construction, the wild type (WT) strain CQMa102 was cultured on 1/4 SDAY (1/4 Sabouraud dextrose agar plus yeast, 10 g glucose, 2.5 g yeast extract, 2.5 g peptone and 18 g agar per liter) for 15 days. Escherichia coli DH5α (Bioground Biotech, Beijing, China) and Agrobacterium tumefaciens AGL1 (Weidi Biotechnology, Shanghai, China) were adopted for vector construction and fungal transformation, respectively. For MaOpy2 deletion vector construction, the fragments of the up-and downstream of MaOpy2 were amplified with the primer pairs Opy2-LF/Opy2-LR and Opy2-RF/Opy2-RR, which were clonedinto pK2-SM-F and pK2-SM-R to form pK2-SM-MaOpy2-F and pK2-SM-MaOpy2-R. The 5 -flanking region and MaOpy2 coding region (3397 bp) and downstream sequence (1398 bp) were cloned into the plasmid pK2-Nat to yield the pK2-MaOpy2-Nat vector, which was used for MaOpy2 complementation (which was resistant to 75 µg/mL nourseothricin sulfate (Harveybio, Beijing, China)). Fungal transformation and transformant validation were conducted as reported previously [20]. The primers used in this work are listed in Table S1.

Phenotypic and Pathogenicity Analyses
Conidial germination and conidial yield assays were conducted as described previously [21]. Briefly, the conidial suspensions (1 × 10 7 conidia/mL) of fungal strains were evenly spread on a 1/4 SDAY medium, and every 2 h, conidial germination was recorded. For the conidial yield, 2 µL suspensions (1 × 10 7 conidia/mL) were spotted onto 24-well plates, and every 3 days, conidial yields were calculated. Additionally, for fungal sensitivity to heat shock, conidial suspensions (1 × 10 7 conidia/mL) treated with a water bath at 45 • C for 3 h, 6 h, 9 h and 12 h were spread on a 1/4 SDAY medium and then cultured at 28 • C for 20 h to access the conidial germination rates. Similarly, for the UV-B tolerance assay, the conidial suspensions (1 × 10 7 conidia/mL) were spread on 1/4 SDAY and exposed to UV-B radiance with a dose of 1350 mW/m 2 for 1.5 h, 3.0 h, 4.5 h and 6.0 h and then cultured at 28 • C for 20 h to access the conidial germination. For fungal tolerances to different chemicals, 2 µL of the suspension (1 × 10 7 conidia/mL) was spotted onto a 1/4 SDAY medium or 1/4 SDAY medium supplied with SDS, calcofluor white (CFW), Congo red (CR), H 2 O 2 , NaCl and sorbitol. After a 5-day cultivation, the fungal colonies were photographed, and the diameters of the colonies were recorded to calculate the relative growth inhibition rates. The fungal pathogenicity assays were determined by two methods. For topical inoculation, the suspensions of WT, ∆MaOpy2 and CP prepared in paraffin (5 µL, 1 × 10 7 conidia/mL) were inoculated to the fifth-instar locust. For intrahemocoel injection, the suspensions of the fungal strains prepared in ddH 2 O (1 × 10 6 conidia/mL) were injected into locust hemocoel. The negative controls for these two methods were treated with pure paraffin oil and ddH 2 O, respectively. Every half day, the survival of the locusts was recorded.

Growth of Fungi in Locust Hemolymph In Vitro and Appressorium Induction on Locust Wings
To determine the appressorium formation and conidial germination, locust wings were used for analyses, as described previously [22]. Briefly, 100 µL of the conidial suspensions (1 × 10 7 conidia/mL) was inoculated into the autoclaved locust wings and then cultured at 28 • C for different times. The conidial germination and appressorium formation were recorded under a microscope. To stain the neutral lipids, Nile red (Sigma-Aldrich, Gillingham, UK) was used [23]; then, it was photographed with fluorescence microscopy. For the fungal growth in hemolymph in vitro, the conidial suspensions (1 × 10 6 conidia/mL, 10 µL) were inoculated into 500 µL of locust hemolymp without blood cells, which was incubated on a rotary shaker at 28 • C and 250 rpm for 4 d or 6 d. To measure the gDNA concentrations for the fungal growth in the locust blood, quantitative real-time PCR was used [24]. The detection of the conidial cell surface components was conducted as previously described [25]. The chitin of the fungal cell wall was stained with wheat germ agglutinin (WGA) (Invitrogen, Carlsbad, CA, USA). Cell wall α-1,3-glucan and β-1,3-glucan were detected using the IgMg MOPC-104E (Sigma, St, Louis, MO, USA) and β-1,3-glucan-specific antibodies (Biosupplies, Parkville, Australia) as the first antibody overnight at 37 • C, respectively, and the Alexa Fluor 488 goat anti-mouse IgM antibody (Invitrogen, Carlsbad, CA, USA) and Alexa Fluor 594 goat anti-mouse IgG antibody (Invitrogen, Carlsbad, CA, USA) were used as the secondary antibody at 37 • C for 4 h in the dark, respectively. The fluorescence of the stained conidia was detected and photographed by fluorescence microscopy (Nikon Y-TV55, Tokyo, Japan).

Microscopic Observation of the Conidiation Pattern and Different Expression Genes (DEGs) by RNA-seq
Fungal fresh conidia cultivated on 1/4 SDAY for 15 days were used for the conidial suspensions. An aliquot of 100 µL of the conidial suspension (1 × 10 7 conidia/mL) of each fungal strain was spread evenly onto 1/4 SDAY plates and incubated at 28 • C. About 1 cm 2 media was cut to detect the fungal growth by microscopic observation every 2 h under the digital light microscope (MOTIC, Xiamen, China). The RNA-sequencing was accomplished by the Beijing Genomics Institution (Wuhan, China). The DEGs were defined with a false discovery rate ≤ 0.005 and a fold change ≥ 2. The RNA-seq data were validated by quantitative reverse-transcription PCR (qRT-PCR).

Statistical Analysis
Tukey's honestly significant difference (HSD) and a one-way ANOVA were applied to access the statistical differences and the phenotypic estimate, respectively.

Characteristics of MaOpy2
Bioinformatics analysis showed that the full length of The sequences of the TM domain were conserved in filamentous fungi, indicating that Opy2 is a conserved membrane-anchor protein ( Figure 1A). Mega v7.0 was used for the phylogenetic tree construction with the neighbor-joining method, showing that, based on the sequence homology, MaOpy2 was well conserved in filamentous fungi ( Figure 1B).

Disruption of MaOpy2 Decreased Fungal Tolerances to UV-B Irradiation, Heat Shock, and Cell-Wall-Perturbing Agents
To reveal the roles of MaOpy2, the strategies of homologous recombination were used to construct the MaOpy2 deletion strain (ΔMaOpy2) and the complementation strain (CP) ( Figure S1A,B). Southern blotting was used to confirm the successful transformants ( Figure  S2C). Conidial germination assays were conducted after the treatment with UV-B irradiation and heat shock. The UV-B tolerance in ΔMaOpy2 was reduced significantly (Figure 2A). The half-inhibition times of germination (IT50s) by UV-B irradiation for the WT and CP strains were 4.42 ± 0.88 h and 4.21 ± 0.48 h, respectively, while the IT50 of ΔMaOpy2 was 3.42 ± 0.29 h ( Figure 2B; p < 0.01). Additionally, the tolerance of ΔMaOpy2 to heat shock was also significantly impaired ( Figure 2C; p < 0.01). The IT50 of ΔMaOpy2 was 8.65 ± 0.76 h-significantly decreased compared to that of the WT (12.20 ± 0.89 h) and CP strains (11.10 ± 0.97 h,

Disruption of MaOpy2 Decreased Fungal Tolerances to UV-B Irradiation, Heat Shock, and Cell-Wall-Perturbing Agents
To reveal the roles of MaOpy2, the strategies of homologous recombination were used to construct the MaOpy2 deletion strain (∆MaOpy2) and the complementation strain (CP) ( Figure S1A,B). Southern blotting was used to confirm the successful transformants ( Figure S2C). Conidial germination assays were conducted after the treatment with UV-B irradiation and heat shock. The UV-B tolerance in ∆MaOpy2 was reduced significantly (Figure 2A). The half-inhibition times of germination (IT 50 s) by UV-B irradiation for the WT and CP strains were 4.42 ± 0.88 h and 4.21 ± 0.48 h, respectively, while the IT 50 of ∆MaOpy2 was 3.42 ± 0.29 h ( Figure 2B; p < 0.01). Additionally, the tolerance of ∆MaOpy2 to heat shock was also significantly impaired ( Figure 2C; p < 0.01). The IT 50 of ∆MaOpy2 was 8.65 ± 0.76 h-significantly decreased compared to that of the WT (12.20 ± 0.89 h) and CP strains (11.10 ± 0.97 h, Figure 2D; p < 0.01). The spot assays revealed that the disruption of MaOpy2 rendered the fungi more sensitive to the cell-wall-disturbing agents SDS and CR; however, the deletion of MaOpy2 did not change the fungal tolerances to oxidative (H 2 O 2 ) and hyperosmotic stresses (NaCl and Sorbitol) ( Figure 3). Figure 2D; p < 0.01). The spot assays revealed that the disruption of MaOpy2 rendered the fungi more sensitive to the cell-wall-disturbing agents SDS and CR; however, the deletion of MaOpy2 did not change the fungal tolerances to oxidative (H2O2) and hyperosmotic stresses (NaCl and Sorbitol) ( Figure 3).    Figure 2D; p < 0.01). The spot assays revealed that the disruption of MaOpy2 rendered the fungi more sensitive to the cell-wall-disturbing agents SDS and CR; however, the deletion of MaOpy2 did not change the fungal tolerances to oxidative (H2O2) and hyperosmotic stresses (NaCl and Sorbitol) ( Figure 3).

Deletion of MaOpy2 Impaired the Insect Cuticle Penetration of M. acridum
For topical inoculation, the disruption of MaOpy2 decreased fungal pathogenicity compared to WT and CP. The locusts all died at day 7 when inoculated with WT and CP; however, when infected with ∆MaOpy2, the locusts died at day 9 ( Figure 4A). In addition, the ∆MaOpy2 strain exhibited significantly longer LT 50 (6.19 ± 0.20 d), while the LT 50 s of the WT and CP strains were 4.70 ± 0.11 d and 4.48 ± 0.26 d, respectively (p < 0.01; Figure 4B). The fungal infection assays showed that, after the inoculation of the locusts for 4 days and 6 days, there were less hyphae bodies in the locusts incubated with ∆MaOpy2 than those in the other strains ( Figure 4C). After 6 d post-inoculation, the fungal gDNA concentrations were decreased in ∆MaOpy2 (0.69 ± 0.03 ng/µL) when compared to those in WT (1.55 ± 0.11 ng/µL) and CP (1.34 ± 0.1 ng/µL, p < 0.05; Figure 4D). To further explore whether MaOpy2 deletion affects fungal penetration, a spot assay was performed on the locusts' hind wings, and the results indicated that the fungal colony of ∆MaOpy2 was smaller than that of the other strains ( Figure 4E). Furthermore, conidial germination on the locust wings and appressorium formation accesses were both conducted. When cultured for 6 h, the conidia germination rate of ∆MaOpy2 was 33.3%-significantly lower than that of the WT (69.0%) and CP (57.6%) strains (p < 0.01; Figure 5A). At 20 h, the conidial germination rates of all the strains reached the same level ( Figure 5A). The GT 50 of ∆MaOpy2 (7.25 ± 0.46 h) was significantly decreased compared with that of WT (4.77 ± 0.54) and CP (5.26 ± 0.54 h, p < 0.01; Figure 5B). In addition, compared with that of the WT (34.3%) and CP (27.0%) strains, the appressorium formation of ∆MaOpy2 (3.3%) was significantly reduced when cultured for 12 h ( Figure 5C). In addition, Nile red was used to measure the pressure in appressorium. However, it was revealed that ∆MaOpy2 presented a weaker fluorescence intensity compared to that presented by WT and CP ( Figure 5D), suggesting that the content of neutral lipids in ∆MaOpy2 was reduced. To investigate the turgor pressure in the fungal appressoria, different concentrations of PEG-8000 were used, and the results revealed that the appressoria of ∆MaOpy2 collapsed more difficultly than those of WT and CP ( Figure 5E For topical inoculation, the disruption of MaOpy2 decreased fungal pathogenicity compared to WT and CP. The locusts all died at day 7 when inoculated with WT and CP; however, when infected with ΔMaOpy2, the locusts died at day 9 ( Figure 4A). In addition, the ΔMaOpy2 strain exhibited significantly longer LT50 (6.19 ± 0.20 d), while the LT50s of the WT and CP strains were 4.70 ± 0.11 d and 4.48 ± 0.26 d, respectively (p < 0.01; Figure 4B). The fungal infection assays showed that, after the inoculation of the locusts for 4 days and 6 days, there were less hyphae bodies in the locusts incubated with ΔMaOpy2 than those in the other strains (Figure 4C). After 6 d post-inoculation, the fungal gDNA concentrations were decreased in ΔMaOpy2 (0.69 ± 0.03 ng/μL) when compared to those in WT (1.55 ± 0.11 ng/μL) and CP (1.34 ± 0.1 ng/μL, p < 0.05; Figure 4D). To further explore whether MaOpy2 deletion affects fungal penetration, a spot assay was performed on the locusts' hind wings, and the results indicated that the fungal colony of ΔMaOpy2 was smaller than that of the other strains ( Figure 4E). Furthermore, conidial germination on the locust wings and appressorium formation accesses were both conducted. When cultured for 6 h, the conidia germination rate of ΔMaOpy2 was 33.3%-significantly lower than that of the WT (69.0%) and CP (57.6%) strains (p < 0.01; Figure 5A). At 20 h, the conidial germination rates of all the strains reached the same level ( Figure 5A). The GT50 of ΔMaOpy2 (7.25 ± 0.46 h) was significantly decreased compared with that of WT (4.77 ± 0.54) and CP (5.26 ± 0.54 h, p < 0.01; Figure 5B). In addition, compared with that of the WT (34.3%) and CP (27.0%) strains, the appressorium formation of ΔMaOpy2 (3.3%) was significantly reduced when cultured for 12 h ( Figure 5C). In addition, Nile red was used to measure the pressure in appressorium. However, it was revealed that ΔMaOpy2 presented a weaker fluorescence intensity compared to that presented by WT and CP ( Figure 5D), suggesting that the content of neutral lipids in ΔMaOpy2 was reduced. To investigate the turgor pressure in the fungal appressoria, different concentrations of PEG-8000 were used, and the results revealed that the appressoria of ΔMaOpy2 collapsed more difficultly than those of WT and CP ( Figure 5E,F), suggesting an increased turgor pressure in ΔMaOpy2. The results above indicate that the MaOpy2 disruption impaired the pre-penetration and penetration of insect cuticle in M. acridum.

Deletion of MaOpy2 Enhanced the Colonized Ability of M. acridum in Locust Hemolymph
For intrahemocoel injection, however, the deletion of MaOpy2 increased the fungal virulence significantly. The LT50 of ΔMaOpy2 was 3.7 ± 0.16 d-significantly decreased compared with that of the WT (5.32 ± 0.33 d) and CP (5.46 ± 0.34 d) strains (p < 0.01; Figure 6A,B). Consistently, the treatment with ΔMaOpy2 led to a higher fungal growth after 6 d injection (p < 0.05; Figure 6C,D). Moreover, the fungal growth in the locust hemolymph without hemocytes in vitro was strengthened, and after 3 d inoculation, the gDNA concentrations of WT, ΔMaOpy2 and CP were 12.66 ± 0.94 pg/μL, 14.7 ± 1.10 pg/μL, and 13.11 ± 0.43 pg/μL, respectively, (p < 0.05; Figure 7A), showing that the deletion of MaOpy2 promoted fungal growth in hemolymph in vitro. Moreover, fluorescent staining revealed that the conidia of ΔMaOpy2 showed dramatically decreased contents of chitin, α-1,3-glucan and β-1,3-glucan ( Figure  7B). These results indicate that the MaOpy2 deletion enhanced the colonized ability of M. acridum in locust hemolymph.

Deletion of MaOpy2 Increased Fungal Conidial Yield by Shifting the Conidiation Pattern
The fungal growth was examined by the conidial germination and conidial yield, and the results showed that the disruption of MaOpy2 accelerated conidial germination. Compared with those of WT (8.64 ± 0.88 h) and CP (8.44 ± 0.70 h), the GT 50 of ∆MaOpy2 was significantly decreased (6.88 ± 0.66 h, p < 0.01; Figure 8A,B). The deletion of MaOpy2 increased the conidial yield in 1/4 SDAY by one time (Figure 8C). To gain a deeper insight into the roles of MaOpy2 in the conidial yield, the conidiation process was microscopically observed. At 12 h, ∆MaOpy2 began to generate conidia at the apex of the hypha (black arrows in Figure 8D); however, the WT and CP strains both grew with long hyphae. At 16 h, ∆MaOpy2 began to exhibit the typical microcycle conidiation, whereas the WT and CP strains began to form conidiophores at the apex of the hyphae (white arrows in Figure 8D). At 36 h, ∆MaOpy2 generated a great number of conidia; however, the WT and CP strains still grew with long hyphae and exhibited normal conidiation ( Figure 8D). The MaOpy2 expression levels were determined by qRT-PCR during the conidiation pattern shift. The results showed that the expression level of MaOpy2 reached its peak at 12 h and was about threefold greater than that at 10 h ( Figure 8E). The results above indicate that MaOpy2 governed the conidiation pattern shift in M. acridum.

Deletion of MaOpy2 Increased Fungal Conidial Yield by Shifting the Conidiation Pattern
The fungal growth was examined by the conidial germination and conidial yield, and the results showed that the disruption of MaOpy2 accelerated conidial germination. Compared with those of WT (8.64 ± 0.88 h) and CP (8.44 ± 0.70 h), the GT50 of ΔMaOpy2 was significantly decreased (6.88 ± 0.66 h, p < 0.01; Figure 8A,B). The deletion of MaOpy2 increased the conidial yield in 1/4 SDAY by one time (Figure 8C). To gain a deeper insight into the roles of MaOpy2 in the conidial yield, the conidiation process was microscopically observed. At 12 h, ΔMaOpy2 began to generate conidia at the apex of the hypha (black arrows in Figure 8D); however, the WT and CP strains both grew with long hyphae. At 16 h, ΔMaOpy2 began to exhibit the typical microcycle conidiation, whereas the WT and CP strains began to form conidiophores at the apex of the hyphae (white arrows in Figure 8D). At 36 h, ΔMaOpy2 generated a great number of conidia; however, the WT and CP strains still grew with long hyphae and exhibited normal conidiation ( Figure 8D). The MaOpy2 expression levels were determined by qRT-PCR during the conidiation pattern shift. The results showed that the expression level of MaOpy2 reached its peak at 12 h and was about threefold greater than that at 10 h ( Figure 8E). The results above indicate that MaOpy2 governed the conidiation pattern shift in M. acridum.

Identification of the DEGs Regulated by MaOpy2 during a Conidiation Pattern Shift Using RNA-seq
To gain an insight into the role of MaOpy2 in shifting the conidiation pattern, RNAseq was adopted to identify the genes regulated by MaOpy2. Combined with the morphology that the strains exhibited ( Figure 8D) and the MaOpy2 expression ( Figure 8E), the total RNA from WT and ΔMaOpy2 on 1/4 SDAY at 12 h were isolated for sequencing. Among the 62 DEGs, 37 were upregulated and 25 were downregulated ( Figure 9A). Thirty-two DEGs were annotated as hypothetical proteins (Table S2). Sixteen DEGs, including eight upregulated and eight downregulated genes, were used for the qRT-PCR analysis to validate the reliability of the RNA-seq data. The results revealed that, compared to the RNA-seq data, all the selected genes displayed similar expression patterns ( Figure S2), indicating that the RNA-seq data were reliable. Through gene ontology (GO) annotation, the DEGs were classified into cat-

Identification of the DEGs Regulated by MaOpy2 during a Conidiation Pattern Shift Using RNA-seq
To gain an insight into the role of MaOpy2 in shifting the conidiation pattern, RNA-seq was adopted to identify the genes regulated by MaOpy2. Combined with the morphology that the strains exhibited ( Figure 8D) and the MaOpy2 expression ( Figure 8E), the total RNA from WT and ∆MaOpy2 on 1/4 SDAY at 12 h were isolated for sequencing. Among the 62 DEGs, 37 were upregulated and 25 were downregulated ( Figure 9A). Thirty-two DEGs were annotated as hypothetical proteins (Table S2). Sixteen DEGs, including eight upregulated and eight downregulated genes, were used for the qRT-PCR analysis to validate the reliability of the RNA-seq data. The results revealed that, compared to the RNA-seq data, all the selected genes displayed similar expression patterns ( Figure S2), indicating that the RNA-seq data were reliable. Through gene ontology (GO) annotation, the DEGs were classified into catalytic activity, binding, membrane, membrane part, metabolic processes and cellular processes ( Figure 9B). The genes involved in the conidiation pattern shift are listed in Table S3, such as a downregulated gene for cation-transporting ATPase 4 (MAC_09130) and two upregulated genes for the pantothenate transporter (MAC_06827) and putative endoglucanase (MAC_02571), which were related to hyphal growth and conidiation, demonstrating that MaOpy2 regulates the shift of the conidiation pattern. alytic activity, binding, membrane, membrane part, metabolic processes and cellular processes ( Figure 9B). The genes involved in the conidiation pattern shift are listed in Table S3, such as a downregulated gene for cation-transporting ATPase 4 (MAC_09130) and two upregulated genes for the pantothenate transporter (MAC_06827) and putative endoglucanase (MAC_02571), which were related to hyphal growth and conidiation, demonstrating that MaOpy2 regulates the shift of the conidiation pattern.

Discussion
Opy2 is a conserved membrane-anchor protein upstream of the HOG-MAPK signaling pathway. In this study, the homolog recombination strategy was used to obtain the MaOpy2 deletion and complementation transformants, and the functions of MaOpy2 were systematically analyzed. The deletion of MaOpy2 significantly impaired the fungal tolerance to UV-B irradiation, heat shock and cell-wall-interfering compounds. Interestingly, the deletion of MaOpy2 decreased the fungal virulence through topical inoculation but enhanced the fungal virulence by injection, suggesting a unique role of MaOpy2 in fungal pathogenicity. More importantly, the disruption of MaOpy2 increased the conidial yield by shifting the conidiation pattern from normal conidiation to microcycle conidiation, and RNA-seq was used to explore the possible mechanism in the MaOpy2 regulation of the conidiation pattern shift.
For entomopathogenic fungi, stress tolerance is a vital factor to survive and infect hosts [2]. Our results gave the first insights into the functions of Opy2 in UV and heat shock tolerance in entomopathogenic fungi, which have never been reported before. It was revealed that the deletion of MaOpy2 significantly decreased the fungal tolerance to UV-B irradiation and heat shock. Based on the RNA-seq data, some DEGs related to the cell wall reorganization were found to be differentially expressed. MAC_01372, a gene for the Glycosyl hydrolase family 16 protein, which was involved in the process of β-1,3-glucan degradation [26], was upregulated in ΔMaOpy2. MAC_04906, a gene for the putative cell wall glycosyl hydrolase Dfg5, was downregulated. In S. cerevisiae, Dfg5 was an essential component of the cell wall and was vital for cell growth [26]. A gene for the serine/threonine-protein kinase GIN4 (MAC_00865) was found to be upregulated. In Fusarium graminearum, the disruption of a GIN4-like protein kinase gene increased the fungal tolerance to cell wall stress [27], suggesting a possibility that the Gin4 regulation of the fungal tolerance to cell wall stress by MaOpy2 functioned differentially in M. acridum. In addition, ΔMaOpy2 exhibited a higher susceptibility to the cell-wall-disturbing agent CR, which was consistent with that in C. albicans [17]. However, MaOpy2 was a dispensable fungal adaption to high osmolarity or oxidative stress, which was in accordance with that in C. albicans [17] but contrast with that in S. cerevisiae.
It was indicated that MaOpy2 played a totally distinct role in fungal pathogenicity by two different methods. On one hand, the fungal virulence was significantly attenuated by topical inoculation by impairing the appressorium formation. Our results are well in accordance with

Discussion
Opy2 is a conserved membrane-anchor protein upstream of the HOG-MAPK signaling pathway. In this study, the homolog recombination strategy was used to obtain the MaOpy2 deletion and complementation transformants, and the functions of MaOpy2 were systematically analyzed. The deletion of MaOpy2 significantly impaired the fungal tolerance to UV-B irradiation, heat shock and cell-wall-interfering compounds. Interestingly, the deletion of MaOpy2 decreased the fungal virulence through topical inoculation but enhanced the fungal virulence by injection, suggesting a unique role of MaOpy2 in fungal pathogenicity. More importantly, the disruption of MaOpy2 increased the conidial yield by shifting the conidiation pattern from normal conidiation to microcycle conidiation, and RNA-seq was used to explore the possible mechanism in the MaOpy2 regulation of the conidiation pattern shift.
For entomopathogenic fungi, stress tolerance is a vital factor to survive and infect hosts [2]. Our results gave the first insights into the functions of Opy2 in UV and heat shock tolerance in entomopathogenic fungi, which have never been reported before. It was revealed that the deletion of MaOpy2 significantly decreased the fungal tolerance to UV-B irradiation and heat shock. Based on the RNA-seq data, some DEGs related to the cell wall reorganization were found to be differentially expressed. MAC_01372, a gene for the Glycosyl hydrolase family 16 protein, which was involved in the process of β-1,3-glucan degradation [26], was upregulated in ∆MaOpy2. MAC_04906, a gene for the putative cell wall glycosyl hydrolase Dfg5, was downregulated. In S. cerevisiae, Dfg5 was an essential component of the cell wall and was vital for cell growth [26]. A gene for the serine/threonine-protein kinase GIN4 (MAC_00865) was found to be upregulated. In Fusarium graminearum, the disruption of a GIN4-like protein kinase gene increased the fungal tolerance to cell wall stress [27], suggesting a possibility that the Gin4 regulation of the fungal tolerance to cell wall stress by MaOpy2 functioned differentially in M. acridum. In addition, ∆MaOpy2 exhibited a higher susceptibility to the cell-wall-disturbing agent CR, which was consistent with that in C. albicans [17]. However, MaOpy2 was a dispensable fungal adaption to high osmolarity or oxidative stress, which was in accordance with that in C. albicans [17] but contrast with that in S. cerevisiae.
It was indicated that MaOpy2 played a totally distinct role in fungal pathogenicity by two different methods. On one hand, the fungal virulence was significantly attenuated by topical inoculation by impairing the appressorium formation. Our results are well in accordance with the results in another fungus. In M. robertsii, both a low MrOpy2 protein level and the deletion of MrOpy2 could impair appressorium formation and thereby reduce fungal pathogenicity to hosts [18]. However, it was found that the deletion of MaOpy2 increased the appressorium pressure, in contrast with the result in M. oryzae, where MoOpy2 deletion reduced the appressorium pressure [19]. A previous study has shown that there are similar cases in which increased turgor pressure led to decreased virulence [21,28]. Intracellular glycerol is essential for the appressorium to generate huge pressure for the successful penetration of the host cuticle [29,30]. The deletion of MaOpy2 impaired fungal virulence, leading to a lower lipid content and increased turgor pressure, suggesting that the decreased virulence may result from the weakened pre-penetration and penetration process. In C. albicans, a CaOpy2 mutant also displayed a significantly reduced virulence in the Galleria mellonella model [17].
On the other hand, the deletion of MaOpy2 significantly enhanced the fungal virulence through injection, which had never been reported. Our results proved that the increased virulence was mainly due to the enhanced fungal colonization inside insects. The host innate immune system could recognize the specific components of the fungal cell wall [31]. Compared to WT and CP, the distribution of α-1,3-glucan, chitin and β-1,3-glucan on the conidial surface of ∆MaOpy2 was obviously decreased, which contributed to the increased fungal virulence by intrahemocoel injection.
More importantly, MaOpy2 deletion could increase the fungal conidial yield by shifting the conidiation pattern from normal conidiation to microcycle conidiation. Three membrane proteins, including Opy2, Sho1 and Msb2, comprise the SHO1-branch of the HOG-MAPK signaling pathway. A previous study has shown that the deletion of MaSho1 could also shift the conidiation pattern to microcycle conidiation in M. acridum [32]. However, MaHog1 made no contribution to shifting the conidiation pattern [32], suggesting that MaOpy2 may function together with MaSho1 during the conidiation pattern shift, which is independent of the conserved HOG-MAPK signaling pathway.
To reveal the mechanism of MaOpy2 regulating the conidiation pattern shift, RNAseq was performed to screen some possible genes that played important roles during the conidiation pattern shift. Among the DEGs, some genes related to the hyphae growth, cell cycle and conidiation were found to be differentially expressed. A gene for the pantothenate transporter (MAC_06827) was upregulated in ∆MaOpy2. A previous study revealed that a pantothenate transporter gene disruption mutant could not form spores [33]. In some fungi, the serine/threonine-protein kinase Gin4 was important for the septin organization and cell cycle [34][35][36]. The inactivation of Gin4 led to a prolonged mitotic delay [36]. In the RNA-seq data, the gene for the serine/threonine-protein kinase Gin4 (MAC_00865) was upregulated, which may accelerate the process of cell division, suggesting the possibility that MaOpy2 may regulate the expression of Gin4 to shift the conidiation pattern, thus promoting the conidiation-but this needs to be further explored.  Figure S2. Validation of the RNA-seq data by qRT-PCR. Table S1: Primers used in this study. Table S2: Differentially expressed genes. Table S3: DEGs involved in stress tolerance and conidiation pattern shift [26,27,[33][34][35][36].
Funding: This research was funded by the National Natural Science Foundation of China (32172479), the Venture & Innovation Support Program for Chongqing Overseas Returnees (cx2019035) and the Chongqing Talent Program of China (cstc2021ycjh-bgzxm0313).