Transcriptional analysis of the conidiation pattern shift of the entomopathogenic fungus Metarhizium acridum in response to different nutrients

Background Most fungi, including entomopathogenic fungi, have two different conidiation patterns, normal and microcycle conidiation, under different culture conditions, eg, in media containing different nutrients. However, the mechanisms underlying the conidiation pattern shift are poorly understood. Results In this study, Metarhizium acridum undergoing microcycle conidiation on sucrose yeast extract agar (SYA) medium shifted to normal conidiation when the medium was supplemented with sucrose, nitrate, or phosphate. By linking changes in nutrients with the conidiation pattern shift and transcriptional changes, we obtained conidiation pattern shift libraries by Solexa/Illumina deep-sequencing technology. A comparative analysis demonstrated that the expression of 137 genes was up-regulated during the shift to normal conidiation, while the expression of 436 genes was up-regulated at the microcycle conidiation stage. A comparison of subtractive libraries revealed that 83, 216, and 168 genes were related to sucrose-induced, nitrate-induced, and phosphate-induced conidiation pattern shifts, respectively. The expression of 217 genes whose expression was specific to microcycle conidiation was further analyzed by the gene expression profiling via multigene concatemers method using mRNA isolated from M. acridum grown on SYA and the four normal conidiation media. The expression of 142 genes was confirmed to be up-regulated on standard SYA medium. Of these 142 genes, 101 encode hypothetical proteins or proteins of unknown function, and only 41 genes encode proteins with putative functions. Of these 41 genes, 18 are related to cell growth, 10 are related to cell proliferation, three are related to the cell cycle, three are related to cell differentiation, two are related to cell wall synthesis, two are related to cell division, and seven have other functions. These results indicate that the conidiation pattern shift in M. acridum mainly results from changes in cell growth and proliferation. Conclusions The results indicate that M. acridum shifts conidiation pattern from microcycle conidiation to normal conidiation when there is increased sucrose, nitrate, or phosphate in the medium during microcycle conidiation. The regulation of conidiation patterning is a complex process involving the cell cycle and metabolism of M. acridum. This study provides essential information about the molecular mechanism of the induction of the conidiation pattern shift by single nutrients. Electronic supplementary material The online version of this article (doi:10.1186/s12864-016-2971-0) contains supplementary material, which is available to authorized users.


Background
Conidia (spores) are the beginning and end of the differentiation process in the lifecycle of fungi [1,2], and they play important roles in reproduction and survival [2]. Most filamentous fungi have two conidiation patterns: normal and microcycle conidiation [3]. Normal conidiation is the most common reproductive mode of filamentous fungi [4], and microcycle conidiation is a survival mechanism under stress conditions, whereby the normal lifecycle is bypassed [3,[5][6][7][8]. To date, microcycle conidiation has been described in more than 100 fungal species [3,5,7], and it has been divided into four basic categories based on the morphological characteristics of conidia [9].
The conidiation patterns can be shifted from normal to microcycle conidiation under various conditions, such as high and/or low temperature [6,10,11], high and/or low pH [12,13], high salt concentration [14], and the presence of certain nutrients [7,10,15,16]. Nutrients are the most common factors that affect fungal conidiation patterns. In Colletotrichum gloeosporioides, microcycle conidiation occurs in substrate-limited liquid cultures [17]. In Beauveria bassiana, microcycle conidiation is observed in the absence of a carbon source in the basal medium [15]. In Aspergillus flavus, exogenous putrescine inhibits microcycle conidiation and induces mycelial development [18]. In Neurospora intermedia, microcycle conidiation occurs under low sugar and nitrogen conditions [19]. The culture conditions for normal and microcycle conidiation are different, and even a subtle change in culture conditions can cause a substantially different conidiation pattern. The conidiation pattern shift in response to nutrients may be regulated by some sensors and pathways [20]. However, the molecular mechanisms of the conidiation pattern shift in response to nutrients have not been elucidated.
The conidia of entomopathogenic fungi are formulated as myco-insecticides [21][22][23]. Metarhizium acridum is a model system for entomopathogenic fungi, and it is widely used for locust control in Africa, Asia, and Australia [24][25][26]. The entomopathogenic fungus M. acridum displays two conidiation patterns: normal conidiation on 1/4 strength Sabouraud's dextrose agar medium (1/4 SDAY), but microcycle conidiation on sucrose yeast extract (SYA) medium [27]. In the present study, the effects of single nutrients on the conidiation pattern of M. acridum were investigated by the addition of 7.5 % sucrose (sucrose-rich (SR) medium), 0.75 % nitrate (nitrate-rich (NR) medium) or 0.25 % phosphate (phosphate-rich (PR) medium) to the microcycle conidiation medium (SYA). The results showed that normal conidiation occurred on 1/4 SDAY and the three nutrient-rich media, and conidiophores and normal conidiation occurred 21 h post-inoculation (hpi), while microcycle conidia were produced on SYA medium during this period. The transcripts of M. acridum derived from SYA medium, the three nutrient-rich media, and 1/4 SDAY medium were compared. The genes involved in the conidiation pattern shift and the genes involved in the regulation of the conidiation pattern shift in the three nutrient-rich media were identified. Then, the mechanisms of the conidiation pattern shift of M. acridum in response to different nutrients were explored.

Results
Conidiation pattern shift of M. acridum in response to different nutrients To investigate the effects of single nutrients on the conidiation pattern shift, M. acridum was grown on the microcycle conidiation medium (SYA), normal conidiation medium (1/4 SDAY), and SYA medium supplemented with sucrose, nitrate, or phosphate. On 1/4 SDAY, SR, NR, and PR media, conidiophores appeared without conidia at 21 hpi. Normal conidiation took place after 24 hpi, while microcycle conidia were produced on SYA medium during this period. The morphology of the normal and microcycle conidia differed significantly, with microcycle conidia having a more uniform size than normal conidia (Fig. 1). These results indicate that all the nutrients, including sucrose, nitrate, and phosphate, can influence the conidiation pattern shift and cause morphological changes in the conidia of M. acridum.

Characterization of a digital gene expression (DGE) database
To elucidate the molecular mechanisms of the conidiation pattern shift that was regulated by single nutrients, mRNA derived from M. acridum cultured on SYA, SR, NR, PR, and 1/4 SDAY media was used to construct five digital gene expression (DGE) libraries. Approximately 6 million sequence tags, 2 million of which were distinct, were obtained for all five DGE libraries. For each library, more than 60 % of the tags were mapped to the transcription reference database of M. acridum [28]. Major characteristics of the libraries are shown in Table 1. The copy number of a tag reflected the mRNA expression level in clean tags, and the distribution of clean tag expression could be used to evaluate the normality of all the data. The distribution of total tags and distinct tags suggests that a small number of mRNAs were highly abundant, but the majority of mRNAs were expressed at low levels, thus meeting the heterogeneity law of gene expression (Additional file 1). The results indicated that our sequencing data are credible and suitable for further analysis.
Using a gene ontology (GO) analysis of the genes mapped in the reference database of the M. acridum genome, we constructed a particular GO hierarchy of (i) biological process, (ii) cellular component, and (iii) molecular function for each library (http://wego.genomics.org.cn/ cgi-bin/wego/index.pl) (Fig. 2). For molecular function, the most significant enrichment was observed among various binding genes (GO: 0005488) and catalytic activity genes (GO: 0003824). For biological process, the most significant enrichment was observed among cellular process (GO: 0009987) and metabolic process (0008152). For cellular component, about 71 % of the differentially expressed genes (DEGs) were found to be involved in "cell" structure; these included genes related to the plasma membrane and external encapsulating structures, such as the cell wall and cell envelope (Additional file 2).

DEGs of M. acridum during normal and microcycle conidiation
To elucidate the molecular mechanism of the conidiation pattern shift, genes whose expression was up-regulated genes during normal and microcycle conidiation were screened by constructing four subtractive libraries based on the five DGE libraries. A comparison of the four subtractive libraries revealed that the expression of 137 genes was up-regulated in the four normal conidiation media (Fig. 3a), and the expression of 436 genes was upregulated in the microcycle conidiation medium (SYA) (Fig. 3b). Among the 137 genes whose expression was up-regulated during the normal conidiation stage, there were three transcription factors, seven absorption-and transportation-related genes, including one amino acid transporter, two major facilitator superfamily (MFS) transporters, two ATP-binding cassette (ABC) transporters, one sulfate transporter, and one oligopeptide transporter, six stress-related genes, including two cytochrome P450 genes, and four nutrient and energy metabolismrelated genes, such as one glycolysis-related gene, one tricarboxylic acid (TCA)-cycle-related genes, and two phosphorylation-related genes (Additional file 3). A GO analysis showed that the genes are mainly involved in catalytic activity, transporter activity, and binding function processes (Additional file 2). These genes are mainly involved in amino acid metabolism, cell growth and death, energy metabolism, lipid metabolism, metabolism of terpenoids and polyketides, xenobiotics biodegradation and metabolism, biosynthesis of secondary metabolites, and carbohydrate metabolism pathways (Additional file 4).
The results indicate that normal conidiation in M. acridum is a complex process that involves multiple genes and biological processes.
These results indicate that genes related to both normal and microcycle conidiation are mainly involved in amino acid metabolism, cell growth and death, energy metabolism, lipid metabolism, metabolism of terpenoids and polyketides, translation pathways, and other pathways (Additional file 4). These pathways participate in cell proliferation, cell development, cell cycle, and cytoskeletal rearrangement processes. The common pathways in the conidiation pattern shift indicate that normal and microcycle conidiation have similar developmental processes and can be regulated through some common pathways, eg, the FluG pathway. However, compared with normal conidiation, microcycle conidiation involved two more pathways, more genes in 10 pathways (Additional file 4), and higher gene expression in common pathways, such as the FluG pathway. These genes and pathways might be related to the shift between normal and microcycle conidiation.

Genes involved in the conidiation pattern shift related to different nutrients
Because the conidiation pattern of M. acridum could be regulated by single nutrients, including sucrose, nitrate, and phosphate, we constructed three subtractive libraries to screen for genes that are specifically expressed in the conidiation pattern shift in response to these nutrients. The three subtractive libraries were compared individually with the genes expressed in 1/4 SDAY medium and the other two nutrient-rich media libraries, which showed that 83, 216, and 168 genes were specifically expressed on SR, NR, and PR media, respectively, indicating that they are involved in the regulation of the conidiation pattern shift by these nutrients (Fig. 4). Among the 83 sucrose-regulated genes, the expression of 75 genes was up-regulated, and the expression of eight genes was down-regulated (Additional file 6). Among them, there were four transcription factors, five stress-related genes, including three cytochrome P450 genes, one lipoxygenase, and one phytanoyl-CoA dioxygenase, six metabolism-related genes, including three glycolysis-related genes, two binding proteins, and one protein tyrosine phosphatase, and two genes involved in the cell cycle process. A GO analysis found that most of the genes played roles in catalytic activity, oxidoreductase activity, hydrolase activity, and metal ion binding (Additional file 6). A pathway analysis found that these DEGs are mainly involved in amino acid metabolism, biosynthesis of secondary metabolites, lipid metabolism, xenobiotics biodegradation and metabolism, metabolism of cofactors and vitamins, and transcription pathways (Additional file 7). The results indicate that sucrose could facilitate cell growth and inhibit sporulation by changing metabolic pathways.
With respect to the nitrate-rich medium, 216 specifically expressed genes were screened in the target libraries. Of these, the expression of 203 genes was up-regulated, and the expression of 13 genes was down-regulated (Additional file 8). Among these, there were seven transcription factors, including four zinc finger transcription factors and one GATA-binding transcription factor, three stress-related genes, including two cytochrome P450 genes and one dioxygenase, 41 metabolism-related genes, including three binding proteins and one protein tyrosine phosphatase, and five genes involved in the cell cycle process. A pathway analysis found that these genes are mainly involved in amino acid metabolism, carbohydrate metabolism, cell growth and death, lipid metabolism, xenobiotics biodegradation and metabolism, nucleotide metabolism, transport and catabolism pathways, and microbial metabolism in diverse environments pathways (Additional file 7). A GO analysis found that most of the genes had roles in catalytic activity, protein kinase activity, transporter activity, transferase activity, hydrolase activity, and adenyl ribonucleotide binding. The results showed that at a high nitrate level, the expression of catalytic-, hydrolase-, transferase-, and ribonucleotide-bindingrelated genes was up-regulated. A reasonable explanation is that nitrogen affects cell morphology by controlling amino acid metabolism.
For the phosphate-rich medium, 168 specifically expressed genes were filtered in the target libraries. Of these, the expression of 151 genes was upregulated, and the expression of 17 genes was downregulated (Additional file 9). Among these, there were five transcription factors, including three zinc finger transcription factors, seven stress-related genes, including two cytochrome P450 genes and three dioxygenases, 24 metabolism-related genes, including two TCA-cyclerelated genes, one phosphorylation-related gene, five binding proteins, and three phosphatases, and two cell cycle genes. A pathway analysis found that the genes are mainly involved in amino acid metabolism, carbohydrate metabolism, cell growth and death, lipid metabolism, xenobiotics biodegradation and metabolism, transport and catabolism, and microbial metabolism in diverse environment pathways (Additional file 7). A GO analysis found that most of the genes are involved in catalytic activity, nucleic acid binding, protein kinase activity, transferase activity, adenyl ribonucleotide binding, ATPase activity, and peptidase activity. The results showed that phosphate might affect cell division, proliferation, and differentiation by controlling the formation of ATP and affecting the cytoskeleton.
Carbon, nitrogen, and phosphorus are basic components of the cell. They are constituents of nucleic acids, sugar-phosphate backbones, and phospholipid bilayers, and they are required for cell division and membrane formation [44]. However, the conidiation pattern shifts in response to these nutrients were controlled by different genes, some of which played roles in the same pathway. MAP kinase [45], the origin recognition complex subunit [46], and a serine/threonine protein kinase [47] are located in the cell growth and death pathway, and they played roles in the conidiation pattern shift in response to sucrose, nitrate, and phosphate, respectively. The conidiation pattern shift in response to nutrients involved different pathways. Exo-beta-D-glucosaminidase is involved in the sucrose-induced conidiation pattern shift, and it has an effect on hyphal growth at low sugar concentrations [48]. Acyltransferase is involved in the nitrate-induced conidiation pattern shift, and it participates in the composition of the membrane at high nitrogen concentrations [49]. L-asparaginase is involved in the phosphate-induced conidiation pattern shift, and it has an effect on growth under certain phosphate concentrations [50,51]. These results indicate that carbon, nitrogen, and phosphorus influence the fungal conidiation pattern by perturbing the cell cycle, nutrient metabolism, and related pathways of M. acridum.

DEGs involved in conidiation were confirmed by the gene expression profiling via multigene concatemers (MgC-GEP) method
To confirm the reliability of the DEGs related to conidiation patterns and their shifts, the expression profiles of the 217 genes that were specific to microcycle conidiation were further analyzed by the MgC-GEP method [52] using the mRNA of M. acridum grown on SYA and the four normal conidiation media. One hundred and eighty genes were found using primer pairs targeting the 217 genes, and the expression of 142 genes was confirmed to be up-regulated on SYA medium in comparison with the four normal conidiation media (Additional files 10 and 11). Of these 142 genes, 101 genes encode hypothetical proteins or proteins of unclear function, and only 41 genes encode proteins with putative functions. Of these 41 genes, 18 are related to cell growth, 10 are related to cell proliferation, three are related to the cell cycle, three are related to cell differentiation, two are related to cell wall synthesis, two are related to cell division, and seven have other functions (Table 2). These results indicate that the conidiation pattern shift in M. acridum mainly results from changes in cell growth and proliferation.

Discussion
Most fungi have two different conidiation patterns, and the mechanisms of normal conidiation have been well studied in most industrial and model species [1,2]. In Aspergillus nidulans, snaD encodes a spindle pole body protein, which is a positive factor in the conidiation process, by promoting the nucleation of tip cells and cell division [53]. In Neurospora crassa, GNAT negatively regulates conidiation through a cAMP-dependent pathway on solid media, and deletion of GNAT leads to decreased aerial hyphal height and premature conidiation [54]. In M. acridum, FKS is involved in cell wall integrity and conidiation [55]. However, the mechanisms of the conidiation pattern shift are not clear.
Previous research found that M. acridum produces microcycle conidia on insect cadavers, while producing normal conidia on PDAY medium [16], suggesting that microcycle conidia maybe produced under nutritive stress. In the present study, microcycle conidiation of M. acridum was shifted to normal conidiation after adding single nutrients to SYA medium. This phenotypic change is similar to the conidiation pattern shift in B. bassiana [14] and A. flavus [17]. Our results revealed that nutrients can cause the conidiation pattern shift in M. acridum. However, the underlying mechanisms are far from clear.
To understand the mechanism of the conidiation pattern shift, we used a statistical comparison analysis to obtain 1040 DEGs that are involved in the conidiation pattern shift. These genes are involved in catalytic activity, cellular transport, cell cycle, developmental processes, signal transduction, stress responses, and metabolic processes. The results indicated that a very complex molecular network is involved in the conidiation pattern shift.
The conidiation pattern shift can be induced in some fungi by different nutrients [7,14,56]. Therefore, some genes that are regulated by nutrients might be involved in the conidiation pattern shift. The glucose level-sensing gene snf3 [57] and the nitrogen-starvation gene glnA [58] were detected in our conidiation pattern shift-related gene library. The results indicate that exogenous nutrients affected the conidiation pattern shift by controlling the expression of nutrient-related genes. Our research identified pathways and molecular mechanisms for the conidiation pattern shift that are regulated by a single nutrient.
pcvA [62], gasA [63], fadA [43], GNAT [54] and pkaA [42] were reported to be specific, negative regulators of conidiation, but their transcription levels were higher during microcycle conidiation than normal conidiation. It may be that: (1) these genes are related to nutrient starvation, and on nutrient-rich media, their function is inhibited by nutrients, which results in normal conidiation; and (2) the transcription levels of these genes may be related to spore production. In summary, there are differences in the morphological and genetic features between the two conidiation patterns. However, little information has been reported about the process of the conidiation pattern shift. The present study used Solexa/Illumina deep-sequencing technology to show that a large number of DEGs are associated with the conidiation pattern shift in M. acridum in the presence of single nutrients. Further studies could confirm these DEGs using transgenic technology to enhance the yields of biopesticides. This study also provides a template for future investigations of the conidiation pattern shift in fungi, and it offers hints about the molecular mechanism of conidiation regulation by different nutrients.

Conclusions
Our results indicate that the conidiation pattern shift is a complex biological process in which the cell cycle and metabolism of M. acridum are perturbed by sucrose, nitrate, and phosphate. This research provides essential information about the molecular mechanism of the conidiation pattern shift that is induced by single nutrients.

Fungal strain and culture conditions
The M. acridum strain CQMa102 used in this study is a locust-specific strain that was isolated by the Genetic Engineering Research Center of Chongqing University and used in a previous study [25]. Conidia were harvested after the fungus was cultured on 1/4 SDAY medium (SDAY, 1 % dextrose, 0.25 % mycological peptone, 2 % agar and 0.5 % yeast extract, w/v) at 28°C for 12-15 d.
SYA medium was used for microcycle conidiation under previously described conditions [64]. Sucrose-, nitrate-, and phosphate-rich media, which were used for normal conidiation, had the same composition as SYA medium, except the amount of sucrose, nitrate and phosphate increased to 75, 7.5, and 2.5 g/L, respectively.

Library preparation and sequencing
After incubation for 21 h at 28°C, the fungal cell mass was harvested by centrifugation and washed with sterile double-distilled H 2 O. Total RNA was extracted from the fungal cell mass using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Oligo(dT) magnetic beads were used to purify mRNA from 6 μg of total RNA, and an oligo(dT) primer was used to synthesize first-and second-strand cDNAs. Sequencing was performed on a High-Seq2000 platform (Illumina, San Diego, CA, USA). Sequence analysis was conducted as described previously [28].

Screening of DEGs
Based on the five DGE databases, a statistical comparison was performed to identify DEGs between the libraries using the rigorous algorithm with a false discovery rate (FDR) < 0.001 and an absolute value for the log2 ratio > 1 [65]. To screen for genes involved in the shift between microcycle and normal conidiation, we constructed four subtractive libraries: an SYA vs. SR media library, an SYA vs. NR media library, an SYA vs. PR media library, and an SYA vs. 1/4 SDAY media library.
To identify genes involved in the regulation of the conidiation pattern shift in response to single nutrients, three subtractive libraries based on the in silico analysis were constructed to screen for genes that are specifically expressed in the conidiation pattern shift in response to sucrose, nitrate, or phosphate. To screen for genes that are specifically expressed in the conidiation pattern shift in response to sucrose, the genes expressed in the 1/4 SDAY, NR, and PR media libraries were removed from the SYA vs. SR media subtractive library, and the remaining genes were the sucrose-regulated genes that are involved in the conidiation pattern shift. Nitrate-and phosphate-regulated genes involved in the conidiation pattern shift were also identified using this method.

GO functional and pathway enrichment analysis
GO is a gene functional classification system with three ontologies: molecular function, cellular component, and biological process. In gene expression profiling analysis, a GO enrichment analysis of functional significance uses a hypergeometric test to map all DEGs to terms in the GO database (http://www.geneontology.org/), looking for significantly enriched GO terms in the DEGs compared to the genome background.
Pathway enrichment analysis identifies significantly enriched metabolic pathways or signal transduction pathways in DEGs by comparison with the whole genome background. The formula is the same as that used in the GO analysis; a pathway with a Q value ≤ 0.05 was defined as having a significant enrichment of DEGs. DEGs that are involved in most major biochemical, metabolism, and signal transduction pathways could be identified through the pathway of significant enrichment.

Verification of the DEGs by expression profiling via multigene concatemers
To verify the DEGs, 217 genes were randomly chosen from the 436 genes whose expression was up-regulated during microcycle conidiation, and gapdh (GenBank accession number: EFY84384), a gene encoding glyceraldehyde 3-phosphate dehydrogenase, was used as an internal control. Primer design was performed according to a previous method [52]. All the primers used in this study are listed in Additional file 11. The expression levels of the selected genes were analyzed in all five media (1/4 SDAY, SYA, SR, NR, and PR) using the MgC-GEP method [52], with minor modifications. For each sample, reverse transcription reaction mixtures were followed by polymerase chain reaction (PCR) amplification with 1 μM of a specific forward primer and 1 μM of a specific reverse primer. The PCR conditions consisted of an initial denaturation at 95°C for 5 min, followed by five cycles of 94°C for 30 s, 68°C for 30 s, 72°C for 30 s, and ending in a single extension cycle of 72°C for 5 min. All of the 218 PCR products were mixed and concentrated to 200 μl, and then extracted using the QIAquick Gel Extraction kit (Qiagen, Valencia, CA, USA). The extraction products were further amplified by PCR with 1 μM of a universal forward primer and 1 μM of a universal reverse primer. The amplification conditions were as follows: an initial denaturation at 95°C for 5 min, followed by 28 cycles of 94°C for 30 s, 53°C for 30 s, and 72°C for 30 s, and ending in a single extension cycle of 72°C for 5 min. The PCR products were extracted with the QIAquick Gel extraction kit and used for quantitative gene expression analysis as described previously [52].