The high osmotic response and cell wall integrity pathways cooperate to regulate morphology, microsclerotia development, and virulence in Metarhizium rileyi

Microsclerotia (MS) formation was successfully induced in Metarhizium rileyi under changing liquid culture conditions. Mitogen-activated protein kinases (MAPKs) play important roles in fungal development and in coordinating many stress responses. To investigate how M. rileyi transduces growth stress and regulates MS differentiation, we characterized the roles of two MAPKs, Hog1- and Slt2-type orthologues, in M. rileyi. Compared with the wild-type strain, the deletion mutants of Mrhog1 (ΔMrhog1) and Mrslt2 (ΔMrslt2) delayed germination and vegetative growth, displayed sensitivities to various stress, and produced morphologically abnormal clones. The ΔMrhog1 and ΔMrslt2 mutants significantly reduced conidial (42–99%) and MS (96–99%) yields. A transcriptional analysis showed that the two MAPKs regulate MS development in a cooperative manner. Insect bioassays revealed that ΔMrhog1 and ΔMrslt2 had decreased virulence levels in topical (36–56%) and injection (78–93%) bioassays. Our results confirmed the roles of MrHog1 and MrSlt2 in sensing growth-related stress and in regulating MS differentiation.

Scientific RepoRts | 6:38765 | DOI: 10.1038/srep38765 Adaptation to changing environments is crucial for fungi viability 14 . Fungi contain complex signalling pathway networks to handle with these stresses 15 . Mitogen-activated protein kinase (MAPK) signalling pathways are conserved and play important roles in sensing environmental stimuli, in transmitting extracellular signals to the nucleus to modulate gene expression, in regulating morphology, in responding to abiotic and biotic stresses, and in virulence/pathogenicity [16][17][18] . Five MAPK signal pathways in fungi are characterized and known to regulate different processes, such as mating, filamentous growth, high osmolarity responses, maintenance of cellular integrity, and ascospore formation. The HOG pathway is used to stimulate transcriptional responses to the osmolarity 14 , and CWI controls the maintenance of cell shape and integrity 17 . Although fungi have mechanisms that guarantee the specificity of MAPK signalling pathways and that insulate other pathways, the pathways respond to many stress in a coordinated manner 17,19 . In some fungi, the HOG and CWI pathways interact in responses the cell wall [20][21][22] and oxidative stresses 23,24 .
MAPK pathways are comprised of three protein kinases: MAP kinase kinase kinase (MAPKKK), MAP kinase kinase (MAPKK), and MAP kinase (MAPK), which are conserved among the MAPK pathways. Traditionally, Hog1p, the core MAPK of the HOG pathway, is activated in sense and responds to fluctuations in environmental osmotic pressure 25,26 . Slt2p, the central MAPK of the CWI pathway, is activated in response to cell wall stress to maintain cell wall integrity [27][28][29] . However, the two MAPKs are not restricted to individual stimuli, but their responses can be elicited by various environmental stresses [30][31][32] . Furthermore, there is some evidence of collaboration between the two MAPKs to counteract various antifungal drugs and environmental stresses [20][21][22][23][24]33 . However, there is no evidence of such a role in regulating MS differentiation.
To investigated the effects of the Mrhog1 and Mrslt2 genes on MS formation and verify whether the two MAPKs pathway regulate MS development in a cooperative way, the Mrhog1 and Mrslt2 genes were cloned, based on a comparative transcriptome library 10 , in the present study. We used a homologous recombination strategy to construct disruption mutants and investigated their involvement in hyphal polarized growth, conidiation, virulence, and MS formation. These mutants appeared to have defective conidial and MS cell differentiation. Insect bioassays revealed a significant decrease in the virulence of the mutants. Thus, Mrhog1 and Mrslt2 play important roles in fungal development and virulence. Furthermore, the expression patterns of the genes involved in CWI and HOG signalling pathways were analysed during MS development. These results provided new information about the cooperation of Mrhog1 and Mrslt2 in the regulation of M. rileyi MS differentiation.  35 , and Mhog1p (98.0% identity) in Magnaporthe oryzae 36 . Furthermore, the deduced amino acid sequence of Mrslt2p presented similarities with those of Slt2p (98.6% identity) in M. robertsii 34 and Slt2p (94.5% identity) in B. bassiana 35 . A phylogenetic tree analysis demonstrated that the two genes are conserved and have the closest genetic relationships with those of Metarhizium spp. (Fig. S1).

Bioinformatics analysis of
Construction of Mrhog1 and Mrslt2 knockout mutants. As there is no public M. rileyi genome, fusion primer and nested integrated PCR 37 was used to obtain the up-and down-stream genomic DNA flanking sequences (data not shown). Using the genomic DNA of the wild-type (WT) strain CQNr01 as the template, the up-(~1.1 Kb and ~1.2 Kb) and down-stream (~1.5 Kb and ~1.3 Kb) flanking sequences of Mrhog1 and Mrslt2, respectively, were obtained by PCR using primers HLF1/HLR1, HRF1/HRR1, SLF1/SLR1, and SRF1/SRR1 (Table S1), respectively. The PCR fragments were digested by restriction endonucleases and ligated into the EcoRI/XhoI and XbaI/HingIII sites of pPZP-Hph-Knock 37 . The resultant plasmids were designated as pPZP-Hph-Mrhog1 and pPZP-Hph-Mrslt2.
The Mrhog1 and Mrslt2 genes were disrupted using the hygromycin phosphotransferase gene (hph) and Agrobacterium-mediated transformation (Fig. S2A). There were approximately 120 and 30 hygromycin B (Hyg B)resistant colonies harbouring the Hog1-and Slt2-replacement constructs, respectively. PCR screening confirmed that the replacement constructs were integrated by homologous recombination at the targeted loci, and three Mrhog1 disrupted mutant strains (ΔH13, ΔH74, and ΔH96) and two Mrslt2 disrupted mutant strains (ΔS6 and ΔS11) were obtained. Genomic Southern hybridization was performed to further confirm four of the mutants (ΔH13, ΔH74, ΔS6, and ΔS11). As shown in Fig. S2B, all of the putative homologous knockout mutants had lost the Mrhog1 or Mrslt2 coding sequences. The four disrupted mutants (ΔH13, ΔH74, ΔS6, and ΔS11) were used in following experiments.
Conidial yield and vegetative growth of ΔMrhog1 and ΔMrslt2. The ability of conidia to germinate was tested on Sabouraud maltose agar fortified with 1% (w/v) yeast extract (SMAY) and is shown in Fig. S3. The germination of mutants was somewhat retarded compared with WT. After 14 h, their germination rates (41.3-42.2% for ΔH13 and ΔH74; 60.2-61.3% for ΔS6 and ΔS11) were significantly lower than that of WT (83.3 ± 2.5%). At 24 h, the germination rates were also different between the mutants and WT. However, at 36 h, the differences in the germination rates were no longer observed (data not shown).
After incubation on SMAY plates, the colony morphologies were observed. There were morphological abnormalities in the mutants (Fig. 1A). At 3 days, the dimorphic switch was delayed in the mutants. After 6 days, the colony surfaces of the mutants were considerably more convoluted compared with the normal smooth colony surfaces of the WT. Furthermore, the mycelial growth of the mutants was restricted (Fig. 1B). After 12 days, the Scientific RepoRts | 6:38765 | DOI: 10.1038/srep38765 induced convolution continues, especially in ΔMrslt2 strains. Additionally, the mutants had significantly reduced conidial yields compared with that of WT (Fig. 1C).
To examine whether the mutant strains were defective in response to abiotic stress, the tested strains were analysed on SMAY plates containing salt, osmotic, cell, and oxidative stress agents. Under the various stress conditions, the conidial yields of the mutant strains were severely affected, exhibiting 41.9-99.1% decreases, which indicated the high sensitivity of the mutant (Fig. 2). Furthermore, the restricted mycelial growth and convoluted colony surfaces were present on these stress-inducing media (data not shown).
Fungal growth of ΔMrhog1 and ΔMrslt2 in liquid culture. The tested strains were cultured in liquid AM according to previous methods 13 . The microscopic observations of conidial germination, hyphal growth, and MS formation are shown in Fig. 3. In 24 h of cultivation, the mutants had slower germination rates (Fig. 3A). After 48 h of vegetative growth, the WT strain started to form MS, however, germinated conidia were observed in the mutants. After 144 h, the MS produced from the WT strain, had matured and was accompanied by secondary mycelial growth, whereas there were few radial growths in the mutant cultures and abnormal structures were observed. The hyphae of ΔMrslt2 mutants were inflated, particularly at the hyphal tips (Fig. S4). Additionally, compared with the WT, the fermentation broth of ΔMrhog1 exhibited a lower viscosity and a higher transparency, while the ΔMrslt2 mutants exhibited lower degree of pigmentation (Fig. 3B). The biomass of ΔMrhog1 was reduced approximately 70% in the AM culture and 42% in the minimal medium (MM) culture (AM without basal salt sources), whereas the ΔMrslt2 was reduced 43% in the AM culture and 22% in the MM culture (Table 1). Furthermore, the MS yields of ΔMrhog1 and ΔMrslt2 were reduced by approximately 99% and 96%, respectively.
Transcriptional level analysis. The expression levels of Mrhog1 and Mrslt2 during MS development were analysed. When compared with the transcriptional levels at the germinating spore stage, the transcriptional levels of Mrhog1 and Mrslt2 peaked when MS initiation occurred (72 h) ( Fig. 4A and B), indicating that the expression of Mrhog1 and Mrslt2 may be involved in regulating MS formation.
Mrhog1 and Mrslt2 were up-regulated when the AM culture was amended with acetic acid, HCl, H 2 O 2 , and iron cations. In particular, the expression levels of the two genes were sharply up-regulated, 10 min after the addition of H 2 O 2 ( Fig. 4C and D). To further analyse the expression levels of genes under changing culture conditions, we used in the MM culture. The results were the same as in the AM investigation ( Fig. 4E and F). This indicated that Mrhog1 and Mrslt2 were up-regulated under changing culture conditions.
To investigate whether the cooperation of Mrhog1 and Mrslt2 was required to regulate MS differentiation, the genes of up-and down-stream in the two signalling pathway were analysed. The genes of the CWI signalling pathway were significantly up-regulated, whereas the expression of Mrmsn 2,4 , downstream of Mrhog1, was significantly reduced in the ΔMrhog1 mutants (Fig. 4G). In the ΔMrslt2 mutants, the genes of the HOG signalling pathway were significantly up-regulated, whereas the expression of gene Mrswi 4,6 , downstream of Mrslt2, was significantly reduced (Fig. 4H). The reduced pigmentation in ΔMrhog1 and ΔMrslt2 mutants suggested that melanin biosynthesis might be compromised. We thus analysed the transcript levels of the MrpksP gene, which is involved in melanin biosynthesis. Consistent with reduced pigmentation, the expression of MrpksP was significantly reduced in the ΔMrhog1 and ΔMrslt2 mutants (Fig. 4G and H). To evaluate whether the abnormal hyphal phenotypes of the mutants were due to changes in cell wall structure, we examined the transcription levels of three chitin synthase genes (class I, II, and IV chitin synthases, Mrchs1, Mrchs2, and Mrchs4, respectively). The results suggested that the expression levels of these genes were significantly increased in the ΔMrslt2 mutants, however, no expression changes were observed in ΔMrhog1 mutants ( Fig. 4G and H).   and ΔMrslt2 mutants increased to 9.5 ± 0.2 and 8.3 ± 0.3 days, respectively, in the topical bioassay, 8.1 ± 0.3 and 7.5 ± 0.2 days, respectively, in the injection bioassays.

Discussion
In this study, using the transformation protocol recently developed in our laboratory 38 , we examined the functions of the Hog1-and Slt2-type MAPKs in M. rileyi in response to stress, and their roles in morphology and pathogenicity. Consistent with findings in other fungi, two MAPKs are important in developmental processes and in pathogenesis 26,28,39,40 . Moreover, our current work indicated that the two MAPKs regulate MS development in a cooperative manner. The M. rileyi exhibits a dimorphic life style and the yeast-to-hyphae transition is a multi-step process 41,42 . To sense the signal, fungi may use several signalling pathways to coordinate the appropriate cellular actions 14 . The Hog1 and Slt2 kinases are various signal-activated proteins. Intriguingly, the inactivation of Mrhog1 or Mrslt2 appears to interfere with the dimorphic transition (Fig. 1). In addition, the disruption of Mrhog1 or Mrslt2 leads to restricted vegetative growth and abnormal morphological clones on SMAY plates. This result was unlike the observations in ΔBbhog1 26 and ΔBbslt2 27,28 . Moreover, similar phenotypic changes have been observed in other fungi [26][27][28]40 , in which the single-gene disruptions increased sensitivity to cell wall damage, high osmolarity, and oxidative stress, as well as showing convoluted colonies. As in other fungi, these results indicated that the CWI and HOG pathways in M. rileyi could functionally overlap in responding to cell wall damage, osmoregulation, and oxidative stress 17,19,24 .
The MS, used as an alternative propagule for biocontrol production, were formed under changing culture conditions 13 . The gens of CWI and HOG pathways were found up-regulated in the comparative transcriptome analysis 10 . As Mrhog1 and Mrslt2 are the central of the HOG and CWI pathways, our first RT-qPCR analysis demonstrated that both genes play important roles in regulating MS formation. Iron cations are the critical nutritional factor that induced M. rileyi MS differentiation 2 and, as in Saccharomyces cerevisiae 43 , exogenous iron cations could produce oxidative stress (data not shown). Similar to the results in other studies, the two genes are activated to respond to low pH levels and oxidative stress 33,44 . The CWI pathway is involved in the maintenance of cell wall integrity 18 and is the main tolerance mechanism to acidic pH 33 , whereas chitin is a main component of the cell wall 45 . The vegetative growth and MS production defects observed in ΔMrslt2 mutants suggested a role in morphogenesis. Our observations of enhanced chitin synthesis and compromised melanisation, resulting in breached cell wall integrity, are underlying causes of the morphological defects. Additionally, as in Verticillium dahliae 40 , the down-regulation of melanin biosynthetic genes led to lower melanin production levels in ΔMrhog1 mutants. To survive under changing culture conditions, as described previously, single-gene disruptions interact to regulate cellular multi-stress responses. Furthermore, two transcriptional factors MrMsn 2,4 , downstream of Mrhog1, and Mrswi4,6, downstream of Mrslt2, were found up-regulated in the comparative transcriptome analysis 10 . Therefore, we chose the MrMsn 2,4 , Mrswi4,6, and the upstream component of CWI and HOG pathways and analysed the transcription level in the Mrhog1 and Mrslt2 mutants. This hypothesis was confirmed by the RT-qPCR analysis (Fig. 4). Further experiments are needed to elucidate the cross-talk between the two MAPKs in regulating MS development.
Hog1p and Slt2p are known to contribute to the regulation of pathogenesis in entomopathogenic and phytopathogenic fungi [26][27][28]40,46 . In M. rileyi, our data indicated that ΔMrhog1 and ΔMrslt2 were significantly less pathogenic than WT in both types of insect bioassays. One explanation involves morphogenic defects in the mutant. The inactivation of Bbhog1 in B. bassiana 26 and Moslt in M. oryzae 46 result in defects in appressorium formation. In contrast, MgHog1 mutants of Mycosphaerella graminicola 47 impair the initiation of infectious germ tubes. Furthermore, in some plant and human fungal pathogens harbouring these mutations are not as virulent due to the defects in invasive growth and conidiation 48 . Our results showed that the mutants were defective in vegetative growth in the hemocoel (data not shown). Another explanation is that fungal pathogens counter the oxidative stress as part of their armoury in the host 49 , and that Hog1p and Slt2p are coordinated to resist oxidative stress 17,19 . As in B. bassiana, the mutants would be hypersensitive to the oxidative stress [26][27][28] .
In conclusion, two MAPKs genes were characterized in the entomopathogenic fungus M. rileyi. The two genes contributed to stress responses and to regulating MS development by acting in a cooperative manner. This study provides new insight into M. rileyi MS development. Further studies are underway to identify the down-stream effectors and mechanism responsible for the Mrhog1 and Mrslt2 interaction.

Materials and Methods
Microbial strains and growth conditions. The fungal strains were cultured according to previously described methods 12 . Escherichia coli DH5α (Invitrogen, Carlsbad, CA, USA) was used for plasmid propagation and cultured in Luria-Bertani medium containing various concentrations of antibiotics based on plasmid resistance. Agrobacterium tumefaciens AGL-1 was used in fungal transformations and grown as described by Shao et al. 38 .
Gene cloning and phylogenetic analysis. Based on the partial sequences in the transcriptome library 10 , we designed gene-specific primers for the amplification of cDNA and genomic DNA sequences. Total RNA was extracted using TRIzol ® reagent (Invitrogen) and first-stand cDNA was synthesized using SuperScript II Reverse Transcriptase (Invitrogen), following the manufacturer's instructions. The amino acid sequences were aligned with DNAMAN software (http://www.lynnon.com) and phylogenetic trees were generated using MEGA 6.0 software (http://www.megasoftware.net).
Generation of Mrhog1 and Mrslt2 gene knockouts. The pPZP-Hph-Mrhog1 and pPZP-Hph-Mrslt2 plasmids were proliferated in E. coli DH5α and transformed into the WT strain using Agrobacterium-mediated transformation 38 . Putative gene disruption mutants were screened on SMAY supplemented with 450 μ g/ml Hyg B. To determine the desired recombination event had occurred in the transformants, their genomic DNA was extracted, PCR screening was performed with primers homologous to the hph and genomic sequence outside the flank regions (HF/hph-R/HR/hph-F and SF/hph-R/SR/hph-F, respectively) (Table S1) and the amplicons were sequenced. Putative mutants were confirmed by additional Southern blotting. The target fragments were probed with 732-bp and 440-bp probe amplified with primers HosF/HosR and SlsF/SlsR. respectively (Table S1).

Conidial yield, germination, and vegetative growth. To characterize the role of Mrhog1 and Mrslt2
in the yeast-to-hyphae transition, vegetative growth, and conidial development, conidia of indicated strains were harvested and suspended according to previous methods 13 . The conidial suspensions were inoculated onto SMAY plates at 25 °C to record conidial germination rate. Beginning 14 h after inoculation, germination was assessed every 2 h. Three random fields were observed by microscope. In each field, the number of germinated conidia out of 100 conidia was recorded. The average numbers from these visual fields were recorded as the germination frequency. Then, 3 μ l of conidial suspensions were dripped onto SMAY plates and cultured under continuous light at 25 °C for 12 days. The colony morphology was examined and images were collected using a digital camera (60-mm Macro lens, Canon Inc., Japan) and microscope.
To analyse abiotic stress tolerance, 3 μ l of conidial suspensions were dripped onto SMAY plates supplemented: (I) with NaCl (0.5 mol/l) and KCl (0.5 mol/l) for the salt stress assay; (II) with sorbitol (0.5 mol/l) and glycerinum (1 mol/l) for the osmosensitivity assay; (III) with Congo red (100 μ g/ml), Calcofluor (20 μ g/ml), and SDS (0.06%) for the cell stress assay; and (IV) with menadione (0.3 mmol/l) and H 2 O 2 (3 mmol/l) for the oxidative stress assay. All the plates were incubated for 12 days. The conidial yields of each strain were assessed according to previous methods 12 .
Fungal growth in liquid culture. The mycelia and MS morphologies were observed using digital camera and microscope. After a 6-day incubation, the biomass was quantified in AM and MM cultures. Furthermore, the MS yield was examined in the AM culture. The biomass and MS yields were determined according to previous methods 12 . Meanwhile, a wet-mount of 72-h and 144-h cultures in 0.1% Calcofluor (Fluorescent Brightener 28, Sigma) were viewed and photographed using a fluorescent attachment (Nikon Ni, Nikon Inc., Japan).
Transcriptional analysis. Three total RNA samples were collected. One collection occurred following the stages of MS development, as described previously 13 . To assess the changing culture conditions effect on Mrhog1 and Mrslt2 expression levels, the incubated AM or MM medium was supplemented with exogenous acetic acid, HCl, H 2 O 2 (3 mmol/l), or iron cation, independently 13 . Then, the mycelia were harvested for another total RNA extraction. To investigate the cooperation of Mrhog1 and Mrslt2 to regulate MS differentiation, the WT and mutant strains separately were incubated in liquid AM cultures. After 3 days of incubation, the mycelia were collected and total RNA was extracted.
All of the samples were collected by vacuum filtration and washed twice with sterile water. After digesting with DNase I (TaKaRa), the first-strand cDNA fragment was synthesized. RT-qPCR was performed using a Mini Opticon Real-time PCR System (Bio-Rad) with SYBR Green detection. As internal standard, β -tubulin (Mrtub) and translation elongation factor (Mrtef) genes were used, and the specific primer pairs are listed in Table S1. The relative quantification of the target gene's expression level was evaluated using the 2 −ΔΔCt method 50 .
Scientific RepoRts | 6:38765 | DOI: 10.1038/srep38765 Virulence assays. Virulence was assayed against third-instar Spodoptera litura larvae by topically immersion into the conidial suspension (5 μ l of a 1 × 10 6 conidia/ml solution in cottonseed oil) or by injection into a conidial suspension (5 μ l of a 1 × 10 6 conidia/ml solution in sterile water with 0.01% Tween 80). Three replicates comprised of 30 larvae each were tested. For the control, 5 μ l of pure cotton seed oil alone or sterile water with 0.01% Tween 80 was applied to or injected into the larvae, respectively. The treated larvae were reared as described previously 12 . The larval mortality rate was recorded every day, and the LT 50 values were estimated by probit analysis.
Statistical analysis. All of the experiments were repeated three times. The data obtained were analysed using SPSS 17.0 software. The mean LT 50 was estimated using SAS version 9.1 software 51 . The graphs were constructed with GraphPad Prism 5 software. The results were recorded as mean ± standard error (SE).