Introduction

Metarhizium, one of the most important genera of entomopathogenic fungi, can kill insects directly by penetration of the cuticle and has emerged as an excellent model organism for exploring many questions at the molecular and biochemical levels (Clarkson and Charnley 1996; Gao et al. 2011). In Australia, Africa and China, M. acridum has been exploited in commercial biological insecticides for grasshoppers and locusts (Langewald and Kooyman 2007; Milner et al. 2002; Peng et al. 2008). As with other entomopathogenic fungi, Metarhizium has been considered as a promising alternative or supplement to chemical pesticides because of its environmental safety, high insect specificity and low likelihood of the development of insect resistance (Charnley and Collins 2007). However, some disadvantages have retarded widespread application, including short-term storage, sensitivity to environmental conditions and the resources required for mass production of conidia (Daoust and Roberts 1983; Ekesi et al. 2003). Therefore, an understanding of the mechanism of resistance to adversity and conidia production in M. acridum is necessary for commercial development and improvement of the biocontrol fungus.

The fungal cell wall, which comprises a polysaccharide–protein complex, can protect the cell from harsh environments and plays a prominent role in resisting host response (Latgé 2007). β-1,3-Glucan and chitin are the major polysaccharides in the basic framework of the fungal cell wall (Lesage and Bussey 2006). Thus, β-1,3-glucan synthesis is crucial for cell wall formation and fungal development (Bowman and Free 2006). β-1,3-Glucan chains are synthesized by a plasma membrane-localized enzyme, β-1,3-glucan synthase (GS; EC 2.4.1.34), which can use UDP-glucose as a substrate and transfer glucose to linear β-1,3-glucan chains (Douglas et al. 1994). The GS complex can be divided into two parts: a catalytic subunit (FKS) and a regulatory subunit (RHO1). FKS encodes a large protein, which is greater than 200 kDa and has 10–16 potential transmembrane helices (Beauvais et al. 2001). In yeast, there are usually multiple FKS genes, which are essential for normal spore wall formation and proper assembly (Ishihara et al. 2007; Martín et al. 2000). In filamentous fungi, such as Aspergillus nidulans, Cordyceps militaris and A. fumigatus, only one FKS homolog gene has been identified (Kelly et al. 1996; Beauvais et al. 2001; Ujita et al. 2006). A colony of A. fumigatus on potato dextrose agar (PDA) containing 1 μg/ml caspofungin showed no filamentous growth and the mycelium was deformed around the edges (Kahn et al. 2006). In Fusarium solani, depleted expression of FsFKS1 reduced spore viability and caused lysis of spores and hyphae (Ha et al. 2006). The heterokaryon ∆FKS1::hygB of Coccidioides posadasii, which has a null FKS1 allele, has an abnormal appearance, with swelling of hyphal elements (Kellner et al. 2005). Nevertheless, the characteristics of FKS in entomopathogenic fungi have not been investigated and its exact function is still far from clear.

In the present study, we cloned a single gene from M. acridum with similarity to other filamentous ascomycete FKS genes. To investigate its function, the expression level of MaFKS was down-regulated by RNA interference (RNAi). Phenotypic examinations of FKS-RNAi transformants demonstrated that MaFKS is involved in cell wall integrity, hyperosmotic pressure tolerance and conidiation in M. acridum.

Materials and methods

Strain, media and growth conditions

Metarhizium acridum strain CQMa102 was as previously described (Peng et al. 2009). For the stress tolerance test, 2 μl of conidial suspension (5 × 105 conidia/ml) was inoculated onto PDA plates containing KCl (1 M), sorbitol (1 M) or mannitol (1 M) and incubated for 9 days or onto PDA plates containing sodium dodecyl sulfate (SDS; 0.01% w/v), calcofluor white (CFW; 50 μg/ml) or Congo red (CR; 500 μg/ml) and incubated at 28°C for 4 days. Agrobacterium tumefaciens AGL1 was used for fungal transformation. Escherichia coli XL-blue was used for routine bacterial transformations and maintenance of plasmids.

Gene cloning and phylogenetic analysis

From the whole genomic sequence of M. acridum (Gao et al. 2011), we found a gene with high similarity to the glucan synthase gene of C. militaris (BAE94194.1). The full cDNA sequence was acquired using RACE PCR. First-strand cDNA was synthesized with reverse transcriptase and the oligo dT-3sites adaptor primer (Takara, China). The 3′-end fragment of the MaFKS cDNA sequence was amplified using a 3′-Full RACE kit (Takara, China) according to the manufacturer’s instructions with the three sites adaptor primer and FKS-3S (5′-GCT TGC TGT ATG GGT CCC GTA TT-3′). 5′-RACE PCR was performed according to the manufacturer’s instructions (5′-Full RACE Core Set, Takara, China). The primers were RT-Prime (5′-AAG CAG TGG TAT CAA CGC AGA GT-3′ and FKS-5A (5′-CGA CCG TAG TAG TCC ATT TCA GC-3′). The full-length MaFKS cDNA sequence (5,820 bp) was amplified with the gene-specific forward primers MaFKS-F2 (5′-TTT GGG CAA CAA TGT CGG-3′) and MaFKS-R2 (5′-GTC AGT TCA CCT AGG ATG TC-3′). Each PCR product was subcloned into the pMD19-T vector and transformed into E. coli XL-blue for determination by GenScript (Nanjing, China). MEGA v4.1 (http://www.megasoftware.net) was used to construct phylogenetic dendrograms. Analysis of the transmembrane domains was performed using DAS (http://www.expasy.ch).

Construction of an RNAi vector

To study the function of MaFKS, an RNAi vector was constructed. Plasmid pPK2-pB was generated by inserting a bar (herbicide resistance gene) cassette into EcoRI and HindIII sites of the binary vector pPK2 (McCluskey 2003) in which an hph (hygromycin B resistance gene) cassette was deleted. A partial sequence of MaFKS (4,660–4,924 bp) was amplified using FKSF (5′-CGG AAT TCC ACC AAT TCG CAT GGA AC-3′) and FKSR primers (5′-GCG ATA TCC GAG GTA CGG TCT CAA GGT C-3′) from the cDNA of MaFKS. The resultant PCR product was digested with EcoRI/EcoRV and inserted into the dual promoter vector pDPB. Then the MaFKS cassette with dual promoters was excised from pDPB by HindIII and XbaI digestion and inserted into pPK2-pB to form pPK2-pB-MaFKS-RNAi.

Transformation of M. acridum and transformant screening

Agrobacterium-mediated transformation was carried out as described by dos Reis et al. (2004). To confirm the expression level of MaFKS, wild-type (WT) cells and FKS-RNAi transformants were grown in 1/4 SDAY (1% dextrose, 0.25% peptone, 0.5% yeast extract and 2% agar) for 3 days and the mycelia were collected and washed with sterile water. Then the mycelium (2 g fresh wt.) was incubated in 20 ml of PDA liquid medium containing 500 μg/ml CR at 28°C for 6 h. Total RNA was isolated using an SV Total RNA Isolation System (Promega, USA). Synthesis of cDNA and real-time RT-PCR were performed using the method described by Peng et al. (2009). Primers FKSRF (5′-TGG TGA AGC ATC GGC AAA AC-3′) and FKSRR (5′-GTA CGG TCT CAA GGT CCA G-3′) were used to detect MaFKS expression levels. Gdpdh encoding glyceraldehyde-3-phosphate dehydrogenase was selected as the reference gene and primers 5′-AGA TGG AGG AGT TGG TGT TG-3′ and 5′-GAC TGC CCG CAT TGA GAA G-3′ were used.

Microscopy

Aliquots of 100 μl of conidial suspension (2 × 107 conidia/ml) for various strains were inoculated into 20 ml of 1/4 SDAY liquid medium and 1/4 SDAY containing 500 μg/ml CR, 50 μg/ml CFW or 0.01% SDS and incubated at 28°C on a rotary shaker (100 rpm) for 48 h to observe hyphal bodies with a microscope. Mycelium of WT cells and FKS-RNAi transformants was harvested from 1/4 SDAY liquid medium and stained with aniline blue for detection of β-1,3-glucans as previously described (Nicholas et al. 1994).

Measurement of β-1,3-glucan

The β-1,3-glucan content of the hyphal wall was measured using the silkworm larvae plasma (SLP) reagent set (Wako Pure Chemical Industries Ltd, Osaka), as previously described (Tsuchiya et al. 1996). The hyphal wall was isolated from hyphae cultured in liquid 1/4 SDAY medium for 60 h, as described by Dallies et al. (1998). The β-1,3-glucan content was normalized to the fresh weight of mycelium.

Conidial yield of transformants

Conidial yield was determined using the method described by Fang et al. (2004). Conidia were spread onto PDA and incubated on PDA plates containing CR (500 μg/ml), CFW (50 μg/ml), SDS (0.01%), KCl (1 M), sorbitol (1 M) or mannitol (1 M) for 12 days.

Statistical analysis

All experiments were repeated at least three times, and data for RNAi efficiency, conidial production and β-1,3-glucan content were analyzed by one-way analysis of variance (ANOVA). The Tukey’s post hoc test was used to compare the means. SAS 2004 for Windows was used for all statistical analyses.

Results

Cloning MaFKS from M. acridum

Using information on the whole M. acridum genome, we cloned one homolog to the FKS1 gene of Saccharomyces cerevisiae and other fungi, named MaFKS (HQ441252). Alignment with the cDNA sequence demonstrated that MaFKS contains a 5,820 bp open reading frame (ORF) and two introns located at the N terminus (177–238 bp) and the C terminus (5,539–5,600 bp). The complete ORF of MaFKS encodes a predicted protein of 1938 amino acids (aa) with a molecular mass of 221.7 kDa and a pI of 7.94. The predicted MaFKS protein has the highest degree of similarity to the putative Fksp of C. militaris (88% identity, BAE94194.1). In comparison with the putative FKS protein from A. fumigatus, the putative β-1,3-glucan synthase domain (aa 852–1,683) contains the putative UDP-glucose binding consensus sequence K/RXTG at the C terminus (MaFKS, RITG: aa 1573-1576) (Inoue et al. 1996; Beauvais et al. 2001). Analysis of the transmembrane domains showed that the deduced MaFKS protein has 16 transmembrane helices like FKS from A. nidulans (Kelly et al. 1996) (Fig. 1a). We also found two domains (domains 1 and 2) in M. acridum FKS that are homologous (44 and 29% identity, respectively) to cellulose synthase of Aerobacter xylinum (SP19449) as described by Kelly et al. (1996) (Fig. 1b). A phylogenetic tree was constructed using the neighbor-joining method and shows the relationship of FKS-like proteins from S. cerevisiae and other fungi. The putative MaFKS protein is most closely related to the putative FKS protein from two other entomopathogenic fungi, Beauveria bassiana and C. militaris (Fig. 1c).

Fig. 1
figure 1

Analysis of putative FKS from M. acridum. a Plot of M. acridum FKS transmembrane regions. The solid line indicates a strict cutoff and the dashed line a loose cutoff. The numbers above each peak correspond to the predicted transmembrane domains. b Domains 1 and 2 are conserved regions between putative MaFKS and BcsAp (cellulose synthase of A. xylinum). Asterisks denote similar amino acids. c Neighbor-joining tree inferred from FKS protein sequence alignments. Numbers above and below the nodes represent the results of bootstrap analyses (100 replicates) using the neighbor-joining method. The tree shows that the M. acridum cloned gene is closely related to homologous sequences obtained from entomopathogenic fungi. Fungal species: Af, A. fumigatus; An, A. nidulans; Ba, B. bassiana; Cm, C. militaris; Cn, C. neoformans; Cp, C. posadasii; Ed, Exophiala dermatitidis; Ma, M. acridum; Nc, N. crassa; Pb, Paracoccidioides brasiliensis; Pm, Penicillium marneffei; Sc, S. cerevisiae; Sp, S. pombe

Full size image

RNAi effectively down-regulates MaFKS expression

To study the function of MaFKS, an RNAi strategy was applied. Phosphinothricin-resistant transformants of M. acridum were generated by transformation with the vector pPK2-pB-MaFKS-RNAi (Fig. 2a). Transcripts of MaFKS were detected by real-time RT–PCR. In comparison with the WT, MaFKS transcription was inhibited in RNAi transformants, with down-regulation ranging from 40 to 64.2% (RNAi-73, -55, -80, -71 and -69) (F 5,17 = 21.353; P = 0.0001) (Fig. 2b).

Fig. 2
figure 2

Effect of RNAi on MaFKS expression in M. acridum. a Maps of pPK2-pB-MaFKS-RNAi, the silencing vector for MaFKS. PgpdA is the promoter of gpd from A. nidulans, bar is the herbicide resistance gene, PtrpC is the promoter of trpC from A. nidulans, and MaFKS (4,660–4,924 bp) is the partial sequence of the β-1,3-glucan synthase gene in M. acridum. b Quantitative RT-PCR. Analysis of MaFKS expression on PDA containing CR for transformants RNAi-55, -69, -71, -73 and -80 compared to the WT (F 5,17 = 21.353; P = 0.0001). Results were first standardized against actin, with WT expression arbitrarily set to 100

Full size image

MaFKS is involved in cell wall integrity

To determine the cell wall integrity of FKS-RNAi transformants, the sensitivity of two transformants (RNAi-69 and RNAi-71) and the WT to agents that disturb the cell wall or cell membrane was investigated. A 2 μl aliquot of conidial suspension, containing approximately 10,000 conidia, was spotted with a micropipette onto PDA plates containing CR, CFW or SDS. In comparison with WT, FKS-RNAi transformants had much less aerial mycelium on PDA plates with or without CR (500 μg/ml), CFW (50 μg/ml) or SDS (0.01%) after 4 days of incubation (Fig. 3a). A viability test for conidia demonstrated that there was no significant difference in germination between the conidia on PDA with and without CR, CFW or SDS (data not shown). The two transformants showed a similar phenotype, and thus RNAi-69 was further examined.

Fig. 3
figure 3

Growth characteristics of FKS-RNAi transformants. a Fungal growth on solid media. Colonies of WT and FKS-RNAi transformants were cultured for 4 days on PDA or PDA plates containing 500 μg/ml CR, 50 μg/ml CFW or 0.01% SDS. b Mycelium development in liquid media. The WT and FKS-RNAi transformants were inoculated into 1/4 SDAY liquid medium and 1/4 SDAY supplemented with 500 μg/ml CR and incubated at 28°C on a rotary shaker (100 rpm) for 48 h. Arrows indicate abnormally shaped hyphal compartments

Full size image

To further clarify the effect of agents that disturb the cell wall or cell membrane, we observed mycelial growth in 1/4 SDAY liquid medium supplemented with CR, CFW or SDS. Most of the mycelium elements in the transformants were indistinguishable from WT hyphae in liquid medium. However, closer inspection revealed some regions of the mycelium with frequent aberrant conformations, such as blowing out of cell wall elements at hyphal tips and internal compartments in the presence of CR (Fig. 3b). However, this phenotype was not observed in 1/4 SDAY liquid medium and 1/4 SDAY supplemented with CFW or SDS (data not shown). All the results demonstrate that depleted MaFKS expression affected mycelium growth and increased sensitivity to the agents used, which indicates that MaFKS plays an important role in cell wall integrity in M. acridum.

Depleted MaFKS expression increases the sensitivity to hyperosmotic pressure in M. acridum

Under hyperosmotic pressure, growth of both WT and FKS-RNAi transformants was partially inhibited, but MaFKS-RNAi made M. acridum more sensitive to hyperosmotic pressure (Fig. 4). Substrate hyphae of the transformants on PDA plates containing 1 M KCl were much fewer than for the WT. Moreover, the colony surface for transformants was smooth and tawny, while that for the WT was typically wrinkled and green. On PDA plates containing 1 M sorbitol or mannitol, both transformants and WT were wrinkled and green, but the transformants had sparse aerial hyphae compared with WT (Fig. 4). These results reveal that MaFKS influences the hyperosmotic pressure tolerance of M. acridum.

Fig. 4
figure 4

Sensitivity of FKS-RNAi transformants to hyperosmotic pressure. Colonies of WT and FKS-RNAi transformants were cultured for 9 days on PDA plates with 1 M KCl, 1 M sorbitol or 1 M mannitol

Full size image

MaFKS is involved in conidiation in M. acridum

Conidial yield was measured on PDA and PDA plates containing CR, CFW, SDS, KCl, sorbitol or mannitol after 12 days of incubation. There was no significant difference in conidial yield between WT and transformants on PDA plates (F 1,5 = 0.883; P = 0.401), but conidial yield/mm2 was significantly higher for WT than for FKS-RNAi transformants on PDA plates containing CR (F 1,5 = 24.874; P = 0.0076), CFW (F 1,5 = 113.032; P = 0.0004), SDS (F 1,5 = 248.914; P = 0.0001), KCl (F 1,5 = 122.129; P = 0.0004), sorbitol (F 1,5 = 197.936; P = 0.0001) or mannitol (F 1,5 = 33.314; P = 0.0045) (Fig. 5). The results show that the expression MaFKS level influences conidial yield in M. acridum under certain adverse conditions. In addition, we observed much fewer aerial hyphae for interference transformants than for WT (data not shown); thus, the reduced conidial yield may be caused by a direct effect on conidial yield and/or an indirect one resulting merely from reduced mycelia growth.

Fig. 5
figure 5

Conidial yield of WT and FKS-RNAi transformants. The conidia were spread onto PDA and PDA plates containing 500 μg/ml CR, 50 μg/ml CFW, 0.01% SDS, 1 M KCl, 1 M sorbitol or 1 M mannitol and incubated for 12 days. Error bars show the standard deviation for three trials (**P < 0.01)

Full size image

FKS-RNAi transformant are defective in β-1,3-glucan synthesis

The β-1,3-glucan content of hyphal wall in mycelium was quantified using the SLP reagent set and the results demonstrate that the β-1,3-glucan content in FKS-RNAi transformants was only 51.32% of the WT content (F 1,5 = 34.105; P = 0.0043) (Fig. 6a). To further investigate changes in β-1,3-glucan distribution, mycelia were stained with the β-1,3-glucan-specific fluorochrome aniline blue. Fluorescence microscopy revealed that WT and transformant hyphae fluoresced over the majority of the hyphal surfaces, but staining in septal regions was greatly decreased in transformant compared to WT hyphae (Fig. 6b). All these results indicate that depleted MaFKS transcription decreases the β-1,3-glucan content in hyphal walls in M. acridum, especially in septal regions.

Fig. 6
figure 6

β-1,3-Glucan content and staining of mycelium of WT and transformants. a β-1,3-Glucan content in hyphae of WT and RNAi-69 transformants was measured using the SLP reagent set (**P < 0.01). b WT and FKS-RNAi transformants were inoculated into 1/4 SDAY liquid medium, incubated at 28°C on a rotary shaker (100 rpm) for 48 h, and then stained with the β-1,3-glucan-specific fluorochrome aniline blue

Full size image

Discussion

β-1,3-Glucan synthesis is essential for life in fungi examined to date (Latgé 2007; Kellner et al. 2005). The β-1,3-glucan synthase catalytic subunit plays an important role in cell wall assembly and growth in fungi (Ishihara et al. 2007; Bernard and Latgé 2001). In this study, we identified a single gene, MaFKS, in M. acridum with significant similarity to known fungal FKS genes. As with other Fksp proteins, the predicted MaFKS protein has 16 transmembrane domains, the putative UDP-glucose binding consensus sequence R/KXTG at the C terminus (RITG) (Beauvais et al. 2001; Kelly et al. 1996; Thompson et al. 1999; Mazur et al. 1995; Mio et al. 1997) and two conserved domains (Kelly et al. 1996; Pereira et al. 2000). All the data presented support the hypothesis that the cloned MaFKS gene encodes β-1,3-glucan synthase.

Genes encoding the β-1,3-glucan synthase catalytic subunit have been well studied in S. cerevisiae (Douglas et al. 1994) and several other fungi (Thompson et al. 1999; Beauvais et al. 2001; Kellner et al. 2005; Ujita et al. 2006). In this study, we constructed RNAi transformants of MaFKS. Microscopic examination of mycelia from transformants grown in the presence of CR revealed some hyphal swelling and blowing out of cell walls. This morphology is similar to that related to the effects of inhibitors of β-1,3-glucan synthase, which cause hyphal swelling in A. fumigatus and C. posadasii germlings (Kurtz et al. 1994; Kellner et al. 2005). In addition, previous research has demonstrated that mutations in genes involved in cell wall biosynthesis, such as Rho1, chsV and ChsVb in Fusarium oxysporum, leads to increased sensitivity to cell wall-disturbing agents (Madrid et al. 2003; Martínez-Rocha et al. 2008; Martín-Urdíroz et al. 2008). The phenotype is consistent with our transformants on solid media containing CR or CFW. This indicates that MaFKS plays an important role in cell wall organization and maintains cell wall integrity, and suggests that the cloned MaFKS gene encodes β-1,3-glucan synthase. In comparison with WT, β-1,3-glucan content in FKS-RNAi transformants was significantly decreased, particularly in septal regions, which confirms that MaFKS encodes β-1,3-glucan synthase.

Interestingly, unlike disruption of other genes related to cell wall biosynthesis, including rho1 (Martínez-Rocha et al. 2008) and gas1 (Caracuel et al. 2005) null mutants of F. oxysporum, the growth of FKS-RNAi M. acridum transformants could not be relieved under hyperosmotic pressure, which suggests that MaFKS plays a role not only in maintaining cell wall integrity, but also in hyperosmotic pressure tolerance. A possible explanation is that β-1,3-glucan is a helical structure that can assume various states by extension (Rees et al. 1982). In S. cerevisiae, the mechanical strength of the cell wall is mainly due to β-1,3-glucan (Klis et al. 2002). A decrease in MaFKS transcript levels can affect the amount of β-1,3-glucan in the hyphal wall. This makes the cell wall fragile, which influences the sensitivity to hyperosmotic pressure in M. acridum.

Increased cell wall synthesis is also a hallmark of conidiation (Peng et al. 2009). For instance, Ecm33 and GEL1 in A. fumigatus and csmB in A. nidulans are involved in fungal cell wall biosynthesis, but also play a role in filamentous fungal conidiation (Mouyna et al. 2005; Romano et al. 2006; Takeshita et al. 2006). Conidial yield analysis demonstrated that conidial production by MaFKS-RNAi transformants was reduced on PDA plates containing disturbing agents or KCl, sorbitol or mannitol. This suggests that MaFKS is important during conidiation under certain conditions. As with csmB mutants of A. nidulans (Takeshita et al. 2006), conidiation of these transformants might be repressed in response to certain signals derived from cell wall defects.

As an alternative to chemical pesticides, M. acridum has been intensively investigated as a commercial biopesticide for grasshoppers and locusts (Langewald and Kooyman 2007; Milner et al. 2002; Peng et al. 2008). Stress tolerance and conidiation are two important factors that could determine the success of entomopathogenic fungi for insect pest management. The dynamic cell wall of fungi is an intermediary between the fungus and the environment and can change, depending on the life cycle stage and environmental conditions (Latgé 2007). This study demonstrated that FKS is involved in cell wall integrity, hyperosmotic pressure tolerance and conidiation in M. acridum. Thus, our data should help in elucidating the mechanism of resistance to adversity and the conidiation process in entomopathogenic fungi.