A Novel Nitrogen and Carbon Metabolism Regulatory Cascade Is Implicated in Entomopathogenicity of the Fungus Metarhizium robertsii.

The entomopathogenic fungus Metarhizium robertsii can switch among parasitic, saprophytic, and symbiotic lifestyles in response to changing nutritional conditions, which is attributed to its extremely versatile metabolism. Here, we found that the Fus3-mitogen-activated protein kinase (MAPK) and the transcription factor regulator of nutrient selection 1 (RNS1) constitute a novel fungal cascade that regulates the degradation of insect cuticular lipids, proteins, and chitin to obtain nutrients for hyphal growth and enter the insect hemocoel for subsequent colonization. On the insect cuticle, Fus3-MAPK physically contacts and phosphorylates RNS1, which facilitates the entry of RNS1 into nuclei. The phosphorylated RNS1 binds to the DNA motif BM2 (ACCAGAC) in its own promoter to self-induce expression, which then activates the expression of genes for degrading cuticular proteins, chitin, and lipids. We further found that the Fus3-MAPK/RNS1 cascade also activates genes for utilizing complex and less-favored nitrogen and carbon sources (casein, colloid chitin, and hydrocarbons) that were not derived from insects, which is repressed by favored organic carbon and nitrogen nutrients, including glucose and glutamine. In conclusion, we discovered a novel regulatory cascade that controls fungal nitrogen and carbon metabolism and is implicated in the entomopathogenicity of M. robertsii. IMPORTANCE Penetration of the cuticle, the first physical barrier to pathogenic fungi, is the prerequisite for fungal infection of insects. In the entomopathogenic fungus Metarhizium robertsii, we found that the Fus3-mitogen-activated protein kinase (MAPK) and the transcription factor regulator of nutrient selection 1 (RNS1) constitute a novel cascade that controls cuticle penetration by regulating degradation of cuticular lipids, proteins, and chitin to obtain nutrients for hyphal growth and entry into the insect hemocoel. In addition, during saprophytic growth, the Fus3-MAPK/RNS1 cascade also activates utilization of complex and less-favored carbon and nitrogen sources that are not derived from insects. The homologs of Fus3-MAPK and RNS1 are widely found in ascomycete filamentous fungi, including saprophytes and pathogens with diverse hosts, suggesting that the regulation of utilization of nitrogen and carbon sources by the Fus3-MAPK/RNS1 cascade could be widespread. Our work provides significant insights into regulation of carbon and nitrogen metabolism in fungi and fungal pathogenesis in insects.

INTRODUCTION

Fungi have evolved sophisticated regulatory mechanisms to enable them to respond rapidly to fluctuating environmental nutrient availability. Utilizing environmental carbon and nitrogen sources is a major physiological process to obtain the needed building blocks to sustain life and grow (1). Two regulatory mechanisms have been found to ensure that fungi preferentially utilize favored carbon (e.g., glucose) and nitrogen (e.g., ammonium [NH4+] and glutamine) sources and that fungi express genes for less-favored carbon or nitrogen source usage only in the absence of the favored nutrients (2). One pathway is the nitrogen metabolite repression (NMR). The protein NmrA deactivates the GATA transcription factor AREA by physical interaction in the presence of favored nitrogen, whereas in the absence of favored nitrogen sources, NAD+ binds to NmrA, which in turn dissociates from AREA, thereby allowing it to activate expression of the genes to utilize less-favored nitrogen sources (3). The other pathway is the carbon catabolite repression (CCR); via the repressor CreA, CCR ensures glucose is preferentially utilized by preventing the expression of genes for the metabolism of less-favored carbon sources (4–6). CCR and NMR converge on genes required for metabolizing several compounds that can be used as both carbon and nitrogen sources such as proline and arginine (7, 8). For example, in the plant-pathogenic fungus Magnaporthe oryzae, the trehalose-6-phosphate synthase Tps1 regulates NMR and CCR in response to glucose-6-phosphate (9). Although regulation of fungal nitrogen and carbon metabolism has been extensively investigated, the regulatory mechanisms remain to be fully understood, especially in multicellular fungal pathogens (9, 10).

The entomopathogenic fungus Metarhizium robertsii can switch among parasitic, saprophytic, and symbiotic lifestyles in response to changing environmental conditions because of its extremely versatile metabolism (11). Insect infection initiates when single-celled conidia adhere to the insect cuticle and produce multicellular hyphae, which tip cells differentiate into infection structures called appressoria. The fungus then penetrates the cuticle via mechanical pressure and cuticle-degrading enzymes. In the host hemocoel, the fungus proliferates as yeast-like hyphal bodies and kills the insects by a combination of fungal growth and toxins. Therefore, the insect cuticle is the first physical barrier to pathogenic fungi, and penetrating the cuticle is a prerequisite for fungal infection (12).

The insect cuticle is a nutrient-sparse environment, with lipids in the outermost layer and chitin and protein in the inner layer. Major signaling components, including mitogen-activated protein kinase (MAPK) cascades, Mr-OPY2, and cyclic adenosine 3′,5′-monophosphate protein kinase A (cAMP-PKA), and several transcription factors such as AFTF1 and MrSkn7 have been found to regulate cuticle penetration (13–15). These proteins control penetration of the cuticle by regulating genes for appressorial formation or for cuticle degradation. Proteases, chitinases, lipases, and P450s are responsible for degrading cuticular lipids, protein, and chitin, and the resulting products are used as nutrients for fungal growth (11, 16–18). However, how the upstream signaling pathways regulate the cuticle-degrading enzymes remains to be explored. The Crr1 gene in CCR was described in M. robertsii, but its roles in utilization of cuticular nutrients are still unknown (19).

In our previous study, we reported that the Fus3-MAPK cascade positively regulates many cuticle-degrading genes (14). In this study, we found that regulation of cuticle degradation by Fus3-MAPK (named Fus3 in this paper for brevity) was mediated by the transcription factor RNS1. On the insect cuticle, Fus3 phosphorylates RNS1 to induce its own expression, which in turn activates the expression of cuticle-degrading genes. During saprophytic growth, the Fus3/RNS1 cascade also regulates utilization of complex less-favored carbon and nitrogen sources that are not derived from insects. Therefore, we discovered a novel cascade that regulates fungal carbon and nitrogen metabolism and is implicated in degradation of the insect cuticle to obtain nutrients for hyphal growth and entry into the hemocoel by the fungus M. robertsii.

RESULTS

RNS1 is regulated by Fus3 during cuticle penetration.

Our previous transcription sequencing (RNA-Seq) analysis showed that a transcription factor (XP_007823070) was positively regulated by Fus3 during cuticle penetration (14). XP_007823070 is designated RNS1, as it is a regulator of nutrient selection (see below). RNS1 is a Myb transcription factor containing a SANT domain (pfam00249). Homologs of RNS1 were identified in many other fungi, including the saprophytes Neurospora crassa (XP_960002) and Aspergillus ochraceoroseus (KKK12761), the insect pathogens Beauveria bassiana (XP_008593929) and Cordyceps militaris (XP_006673394), the plant pathogens M. oryzae (XP_003711780) and Fusarium graminearum (XP_011327324), and the mammal pathogens Histoplasma capsulatum (EEH11129) and Blastomyces dermatitidis (EQL37237).

Quantitative reverse transcription-PCR (qRT-PCR) showed that Rns1 was ∼11-fold more highly expressed during cuticle penetration (grown on the hindwings of Locusta migratoria locust adults) than during saprophytic growth (grown in the nutrient-rich medium Sabroud dextrose broth plus 1% yeast extract [SDY]) and hemocoel colonization (grown in the silkworm hemolymph), but no significant difference was found between saprophytic growth and hemocoel colonization (Fig. 1A). During cuticle penetration, Rns1 was ∼5-fold more highly expressed in the wild-type (WT) strain than in the deletion mutants of the three kinase genes (Fus3-MAPK, Ste7-MAPKK, and Ste11-MAPKKK) in the Fus3-MAPK cascade (Fig. 1B), but no difference was found between the three mutants. During saprophytic growth and hemocoel colonization, no difference in Rns1 expression was found between the WT strain and the ΔFus3 mutant (see Fig. S1A in the supplemental material).

FIG 1

Expression pattern of Rns1 and its involvement in cuticle penetration. (A) qRT-PCR analysis of Rns1 expression during saprophytic growth (SDY), cuticle penetration (cuticle), and hemocoel colonization (hemolymph). All qRT-PCR analyses in this study were repeated three times, and the values in each figure represent the fold changes of expression of a gene in treatments compared with expression in their respective controls, which are set to 1. (B) qRT-PCR analysis of Rns1 expression during cuticle penetration in the WT strain and the three deletion mutants in the Fus3-MAPK cascade: ΔFus3 (MAPK), ΔSte7 (MAPKK), and ΔSte11 (MAPKKK). (C) LT50 values when the insects were inoculated by topical application of conidia on the cuticle. Data are expressed as the means ± standard errors (SEs). Values with different lowercase letters are significantly different (n = 3, P < 0.05, Tukey’s test in one-way analysis of variance [ANOVA]). (D) Appressorial formation on a hydrophobic plastic surface. Within each time point, data with different lowercase letters are significantly different (n = 3, P < 0.05, Tukey’s test in one-way ANOVA). The experiment was repeated three times. Data are expressed as the means ± SEs. WT, wild-type strain; ΔRns1, deletion mutant of Rns1; C-ΔRns1, the complemented strain of the ΔRns1 mutant.

Rns1 is not regulated by Fus3 during hemocoel colonization and saprophytic growth. (A) qRT-PCR analysis of Rns1 expression during saprophytic growth in the SDY medium (SDY) and during hemocoel colonization (hemolymph). (B) LT50 values when the insects were inoculated by injection of conidia into the hemocoel. Data are expressed as the means ± SEs. Values with same lowercase letters are not significantly different (n = 3, P > 0.05, Tukey’s test in one-way ANOVA). (C) Collapse rates of appressoria in the polyethylene glycol 8000 (PEG 8000) solution (80% [wt/vol]). The assays were repeated three times with three replicates per repeat. Data are expressed as the means ± SEs. Download FIG S1, TIF file, 0.7 MB.

To investigate the biological functions of Rns1, we constructed the deletion mutant of Rns1 (ΔRns1) and its complemented strain (C-ΔRns1) (see Fig. S2A, B, and C). The genes and fungal strains used in this study are listed in the Table 1. Bioassays were conducted on Galleria mellonella larvae either by topically applying conidia to the insect cuticle or by direct injection of conidia into the hemocoel (thus bypassing the cuticle). Following topical application, the time taken to kill 50% of insects (LT50) of the ΔRns1 mutant (20.1 ± 1.1 days) was nearly 2-fold higher (P < 0.05) than the WT strain (10.2 ± 0.3 days), but no significant difference (P > 0.05) was found between the WT strain and the C-ΔRns1 strain (11.5 ± 1.4 days) (Fig. 1C). Via injection, no significant difference (P > 0.05) in LT50 was found among the WT, ΔRns1, and C-ΔRns1 strains (Fig. S1B). On a normally inductive milieu (the hydrophobic surfaces of plastic petri dishes) in the presence of low levels of nitrogenous nutrients, compared with that in the WT strain, the appressorial formation of the ΔRns1 mutant was delayed, and again, no difference was found between the WT and C-ΔRns1 strains (Fig. 1D). No significant difference in appressorial turgor pressure was found between the WT, ΔRns1, and C-ΔRns1 strains (Fig. S1C).

TABLE 1

Plasmids, fusion proteins, and fungal strains used in this study

Plasmid, protein, or strain	Description	Reference
Plasmids
pPK2-Bar-Ptef-HA	Expression of a protein tagged with HA	25
pPK2-bar-Ptef-YFPN	Expression of a protein tagged with the N terminus of YFP	This study
pPK2-sur-Ptef-YFPC	Expression of a protein tagged with the C terminus of YFP	This study
pPK2-Sur-Ptef-FLAG	Expression of a protein tagged with FLAG	25
pPK2-Sur-Ptef-GFP-N	Expression of a protein tagged with GFP	25
Fusion proteins
Fus3::HA	Fus3 tagged with HA	25
RNS1-DBD::YFPN	RNS1-DBD tagged with the N terminus of YFP	This study
Fus3::YFPC	Fus3 tagged with the C terminus of YFP	This study
RNS1::FLAG	RNS1 tagged with FLAG	This study
RNS1T215A::FLAG	Thr-215 replaced by alanine in RNS1::FLAG	This study
RNS1S226A::FLAG	Ser-226 replaced by alanine in RNS1::FLAG	This study
RNS1T215A /S226A::FLAG	Thr-215 and Ser-226 replaced by alanines in RNS1::FLAG	This study
RNS1-DBD	RNS1 portion (Glu-61 to Pro-267)	This study
RNS1-DBD::GFP	RNS1-DBD tagged with GFP	This study
RNS1T215A-DBD::GFP	Thr-215 replaced by alanine in RNS1-DBD::GFP	This study
RNS1S226A-DBD::GFP	Ser-226 replaced by alanine in RNS1-DBD::GFP	This study
Promoters
PRns1	Promoter region (1,724 bp upstream of the ORF) of the gene Rns1	This study
PRns1ΔBM2	Mutated PRns1 with all 7 nt in the motif BM2 changed to As	This study
Genomic clones
gRns1	Genomic clone of the Rns1 gene	This study
gRns1T215A	gRns1 mutated to replace Thr-215 with alanine in RNS1 protein	This study
gRns1S226A	gRns1 mutated to replace Ser-226 with alanine in RNS1 protein	This study
gRns1ΔBM2	gRns1 with 7 nt of BM2 changed to As in the promoter PRns1	This study
Fungal strains
WT	Wild-type strain of M. robertsii ARSEF 2575	This study
ΔRns1	Deletion mutant of the Rns1 gene	This study
C-ΔRns1	Complemented strain of the ΔRns1 mutant	This study
WT-FLAG	Expressing a FLAG tag in the WT strain	This study
WT-RNS1-FLAG	Expressing RNS1::FLAG in the WT strain	This study
WT-RNS1T215A-FLAG	Expressing RNS1T215A::FLAG in the WT strain	This study
WT-RNS1S226A-FLAG	Expressing RNS1S226A::FLAG in the WT strain	This study
WT-RNS1T215A/S226A-FLAG	Expressing RNS1T215A/S226A::FLAG in the WT strain	This study
ΔFus3-RNS1-FLAG	Expressing RNS1::FLAG in the ΔFus3 mutant	This study
ΔFus3-RNS1T215A-FLAG	Expressing RNS1T215A::FLAG in the ΔFus3 mutant	This study
ΔFus3-RNS1S226A-FLAG	Expressing RNS1S226A::FLAG in thet ΔFus3 mutan	This study
ΔFus3-RNS1T215A/S226A-FLAG	Expressing RNS1T215A/S226A::FLAG in the WT strain	This study
ΔRns1-RNS1-FLAG	Expressing RNS1::FLAG in the ΔRns1 mutant	This study
ΔRns1-FLAG	Expressing a FLAG tag in the ΔRns1 mutant	This study
RNS1-FLAG/Fus3-HA	Expressing RNS1::FLAG and Fus3::HA in the WT strain	This study
RNS1-DBD-YFPN/Fus3-YFPC	Expressing RNS1-DBD::YFPN and Fus3::YFPC in the WT strain	This study
RNS1-DBD-YFPN/YFPC	Expressing RNS1-DBD::YFPN and YFPC in the WT strain	This study
WT-RNS1-DBD-GFP	Expressing RNS1-DBD::GFP in the WT strain	This study
WT-RNS1T215A-DBD-GFP	Expressing RNS1T215A-DBD::GFP in the WT strain	This study
WT-RNS1S226A-DBD-GFP	Expressing RNS1S226A-DBD::GFP in the WT strain	This study
ΔFus3-RNS1-DBD-GFP	Expressing RNS1-DBD::GFP in the ΔFus3 mutant	This study
ΔFus3-RNS1T215A-DBD-GFP	Expressing RNS1T215A-DBD::GFP in the ΔFus3 mutant	This study
ΔFus3-RNS1S226A-DBD-GFP	Expressing RNS1S226A-DBD::GFP in the ΔFus3 mutant	This study
ΔRns1-gRns1T215A	ΔRns1 mutant transformed with the genomic clone gRns1T215A	This study
ΔRns1-gRns1S226A	ΔRns1 mutant transformed with the genomic clone gRns1S226A	This study
ΔRns1-gRns1ΔBM2	ΔRns1 mutant transformed with the genomic clone gRns1ΔBM2	This study
WT-PRns1-gfp	gfp gene driven by the promoter PRns1 in the WT strain	This study
WT-PRns1ΔBM2-gfp	gfp gene driven by the mutant promoter PRns1ΔBM2 in the WT strain	This study
ΔRns1-PRns1-gfp	gfp gene driven by the promoter PRns1 in theΔRns1 mutant	This study
ΔRns1-PRns1ΔBM2-gfp	gfp gene driven by the mutant promoter PRns1ΔBM2 in the ΔRns1 mutant	This study
ΔFus3-PRns1-gfp	gfp gene driven by the promoter PRns1 in the ΔFus3 mutant	This study
ΔFus3-PRns1ΔBM2-gfp	gfp gene driven by the mutant promoter PRns1ΔBM2 in the ΔFus3 mutant	This study

Confirmation of gene deletion and knockdown. (A) Schematic diagram of deleting a gene based on homologous recombination. Map of a deletion plasmid (bottom) and its relative position in the fungal genome (top). (B) Confirmation of the deletion of Rns1 and complementation of the mutant with PCR. D1 and D2, two independent deletion mutants; WT, wild-type strain; M, DNA ladder. (Top) PCR performed with primers CF1/CF2 (the relative positions of all primers are shown in panel A); PCR products were obtained only in the WT strain. (Middle) PCR with Bar-up/CF2; PCR products were obtained only from the gene deletion mutants. (Bottom) Confirmation of the complementation of the ΔRns1 deletion mutants using PCR. C, complemented strain; D, ΔRns1 mutant. (C) Confirmation of insertion of the mutants of the genomic clone gRns1 into the ΔRns1 mutant. 1, ΔRns1-gRns1 mutant; 2, ΔRns1-gRns1 mutant; 3, ΔRns1-gRns1Δ mutant; D, ΔRns1 mutant; WT, wild-type strain. Confirmation of the deletion of AreA (D), Crr1 (E), and Snf1 (F) with PCR. (Top) PCR with primers CF1/CF2; (bottom) PCR with Bar-up/CF2. (G) qRT-PCR analysis of Tor, Tps1, and G6PD in the WT strain and their knockdown mutants. #1, #2, and #3, three independent knockdown isolates. Download FIG S2, TIF file, 2.9 MB.

Fus3 phosphorylates RNS1 during cuticle penetration.

To investigate how Fus3 regulates Rns1 expression, we first assayed whether Fus3 physically contacted RNS1. Yeast two-hybrid assays showed that Fus3 physically interacted with RNS1 (Fig. 2A). Using an RNS1-FLAG/Fus3-HA strain that expressed RNS1::FLAG (a protein with the FLAG fused to RNS1) and Fus3::HA (a protein with Fus3 tagged with hemagglutinin), a coimmunoprecipitation (Co-IP) assay confirmed that RNS1 physically interacted with Fus3 in vivo (Fig. 2B). Unless otherwise indicated, all fusion proteins are driven by the constitutive promoter Ptef from Aureobasidium pullulans in this study. For a biomolecular fluorescence complementation (BiFC) assay, construction of a strain expressing a protein with RNS1 fused with the N terminus of YFP (YFPN) failed, but we successfully constructed a protein RNS1-DBD::YFPN with YFPN fused to a portion of RNS1 (named RNS1-DBD, Glu-61 to Pro-267) containing the nuclear localization signal (NLS), DNA binding domain (DBD), and Fus3 phosphorylation sites (see below). A clear YFP signal was detected in the cells of the RNS1-DBD-YFP strain that expressed the fusion proteins RNS1-DBD::YFPN and Fus3::YFPC (Fig. 2C). YFPC is the C terminus of YFP. YFP signal was not observed in the negative-control RNS1-DBD-YFP strain expressing RNS1-DBD::YFPN and YFPC (Fig. 2C).

FIG 2

Phosphorylation of RNS1 by Fus3 is important for its entry into the nucleus. (A) Yeast two-hybrid analysis showing Fus3 interacts with RNS1. (Left) Colonies grown in SD-His-Ade-Leu-Trp plus X-α-Gal plus AbA. Fus3/RNS1, cells expressing Fus3 and RNS1. (Right) Colonies grown in SD-His-Trp-Ade plus X-α-Gal. BD-Fus3, Y2HGold cells expressing Fus3; NC, negative control; PC, positive control. (B) Co-IP confirmation of the interaction of Fus3 with RNS1. Immunoprecipitation was conducted with anti-FLAG antibody. Proteins were detected by immunoblot analysis with anti-HA or anti-FLAG antibodies. (C) BiFC analysis showing Fus3 interacted with RNS1. (Left) Strain expressing RNS1-DBD-YFPN and Fus3-YFPC. (Right) Strain expressing RNS1-DBD-YFPN and YFPC. Bars, 20 μm. (D) Phos-tag analysis of RNS1 phosphorylation by Fus3. 1, ΔFus3-RNS1-FLAG strain; 2, WT-RNS1-FLAG strain; 3, WT-RNS1 strain; 4, ΔFus3-RNS1 strain; 5, WT-RNS1 strain; 6, ΔFus3-RNS1 strain; SDY, SDY medium; cuticle, cuticle medium. (E) Phosphorylation of the sites Thr-215 and Ser-226 by Fus3 is important for RNS1 entry into the nucleus. The strain names are shown on the left. DAPI, 4′,6-diamidino-2-phenylindole; c, conidium; n, nucleus. Bars, 5 μm. Detailed description of the strains is shown in Table 1. In this study, all shown images are representative of at least three independent experiments.

To investigate whether Fus3 phosphorylates RNS1, the WT-RNS1-FLAG and ΔFus3-RNS1-FLAG strains were constructed by expressing RNS1::FLAG in the WT strain and the ΔFus3 mutant, respectively (see Fig. S3A). RNS1 phosphorylation was assayed with the Phos-tag method. Since it was not possible to obtain enough biomass to get sufficient protein for this assay by growing M. robertsii on the locust hindwings, we cultivated the fungus in the cuticle medium with the insect cuticle as a sole carbon and nitrogen sources to approximate cuticle penetration. In the cuticle medium, compared with that of the ΔFus3-RNS1-FLAG strain, the band of the RNS1::FLAG protein shifted in the WT-RNS1-FLAG strain (Fig. 2D). When the WT-RNS1-FLAG strain was transferred from the SDY medium to the cuticle medium, this band shift was also seen, but no obvious difference in the positions of the RNS1::FLAG protein was seen between the WT-RNS1-FLAG strain in the SDY medium and the ΔFus3-RNS1-FLAG strain in the cuticle medium or the SDY medium (Fig. 2D).

Confirmation of expression of fusion proteins. (A) Expression of the fusion proteins RNS1::FLAG, RNS1T215A::FLAG or RNS1S226A::FLAG in the WT strain or in the ΔFus3 mutant. 1, WT-RNS1-FLAG; mutant; 2, ΔFus3-RNS1-FLAG mutant; 3, WT-RNS1 mutant; 4, ΔFus3-RNS1 mutant; 5, WT-RNS1 mutant; 6, ΔFus3-RNS1 mutant; 7, ΔRns1-RNS1-FLAG mutant; 8, WT-FLAG strain. Note that the band corresponding to the FLAG tag was not seen due to its small size. M, protein ladder (Thermo Fisher Scientific). (B) qRT-PCR analysis of the RNS1::FLAG-encoding gene in the WT-RNS1-FLAG, ΔFus3-RNS1-FLAG, WT-RNS1, and ΔFus3-RNS1 strains. Two randomly selected isolates of the WT-RNS1 and ΔFus3-RNS1 strains were analyzed. (C) Expression of RNS1::FLAG and RNS1T215A/S226A::FLAG. 1, WT-RNS1-FLAG mutant; 2, ΔFus3-RNS1-FLAG mutant; 3 and 4, two isolates of the WT-RNS1 strain; 5 and 6, two isolates of the ΔFus3-RNS1 strain. (D) Impact of the proteasome inhibitor MG132 (0.2 mM) and the autophagy inhibitor 3-methyladenine (3-MA; 0.2 mM) treatment on the expression of RNS1T215A/S226A::FLAG in the WT-RNS1 strain. 1, not treated with inhibitors; 2, treatment with MG132 for 4 h; 3, treatment with MG132 for 8 h; 4, treatment with 3-MA for 4 h; 5, treatment with 3-MA for 8 h. Note that the protein RNS1T215A/S226A::FLAG was not detected. (E) Expression of the proteins RNS1-DBD::GFP, RNS1T215A-DBD::GFP, and RNS1S226A-DBD::GFP in the WT strain or in the ΔFus3 mutant. 1, WT-RNS1-DBD-GFP mutant; 2, ΔFus3-RNS1-DBD-GFP mutant; 3, WT-RNS1 mutant; 4, ΔFus3-RNS1; mutant 5, WT-RNS1 mutant; 6, ΔFus3-RNS1 mutant. Detailed description of the strains is shown in Table 1. In this study, all shown images are representative of at least three independent experiments. Download FIG S3, TIF file, 0.8 MB.

According to the characteristics of the serine/threonine amino acids targeted by MAPK kinases (20), the threonine at position 215 (Thr-215) and the serine at position 226 (Ser-226) in RNS1 were predicted to be Fus3 phosphorylation sites. To confirm these predictions, we constructed two RNS1 mutants (RNS1T215A::FLAG or RNS1S226A::FLAG) with Thr-215 or Ser-226 in the protein RNS1::FLAG, respectively, replaced with alanine. They were expressed in the WT strain to produce the strains WT-RNS1 and WT-RNS1 and in the ΔFus3 mutant to generate ΔFus3-RNS1 and ΔFus3-RNS1 strains (Fig. S3A). In the cuticle medium, comparison of WT-RNS1-FLAG, WT-RNS1, and WT-RNS1 strains showed that the mutations of Thr-215 and Ser-226 significantly affected RNS1 phosphorylation (Fig. 2D). Three bands were seen in WT-RNS1 and WT-RNS1 strains, and two of them had the same migration speed, indicating that the differential bands between the two strains corresponded to the proteins with Thr-215 or Ser-226 phosphorylated. But the difference in migration of these two differential bands was not seen in ΔFus3-RNS1 and ΔFus3-RNS1 strains.

We also constructed a WT-RNS1 strain expressing the protein RNS1T215A/S226A::FLAG with both Thr-215 and Ser-226 replaced by alanines. RNS1 was constitutively transcribed, but its protein was not detectable, which was not attributed to the proteasome pathway and autophagy (Fig. S3B to D). Therefore, phosphorylation of Thr-215 and Ser-226 could not be simultaneously assayed.

Fus3-mediated phosphorylation facilitates RNS1 entry into nuclei.

To investigate the impact of Fus3-mediated phosphorylation on RNS1 entry into nuclei, we constructed three green fluorescent protein (GFP)-tagged proteins: RNS1-DBD::GFP, RNS1T215A-DBD::GFP, and RNS1S226A-DBD::GFP, which were then expressed in the WT strain to produce WT-RNS1-DBD-GFP, WT-RNS1, and WT-RNS1 strains, respectively, and in the ΔFus3 mutant to form ΔFus3-RNS1-DBD-GFP, ΔFus3-RNS1, and ΔFus3-RNS1 strains, respectively (Fig. S3E). On the insect cuticle, the GFP signal was concentrated in the nuclei of the WT-RNS1-DBD-GFP strain, but it was almost evenly distributed in the cells of all five other strains (Fig. 2E).

Identification of the DNA motifs recognized by RNS1.

Chromatin immunoprecipitation sequencing (ChIP-Seq) analysis was conducted to identify the DNA motif recognized by RNS1. To this end, the ΔRns1-RNS1-FLAG strain was constructed with RNS1::FLAG expressed in the ΔRns1 mutant (Fig. S3A). A total of 3, 616 peaks were identified (see Fig. S4A), and six consensus motifs were present with a high confidence score (Fig. S4B).

Identification of DNA motifs recognized by RNS1. (A) ChIP-Seq analysis of genome-wide RNS1 binding motifs. Shown is the distribution of all RNS1 peak locations within the genome. (B) Top six motifs enriched by HOMER software. (C) SDS-PAGE analysis of the expression and purification of the recombinant protein RNS1-DBD in E. coli. M, protein ladder; 1, crude extract from the cells with the plasmid pET-28a-SUMO (control); 2, crude extract from the cells expressing RNS1-DBD; 3, supernatant of the crude extract shown in lane 2; 4, proteins purified from the supernatant shown in lane 3 with the HisPur Ni-NTA resin; 5, proteins after the protease ULP1 treatment with the proteins from lane 4 to remove the SUMO tag; 6, homogenous RNS1-DBD purified from the proteins of lane 5 with the HisPur Ni-NTA resin. (D) Western blot analysis confirming expression of the protein SUMO::RNS1-DBD using the anti-His tag antibody. 1 and 2 are proteins from lane 3 and 4, respectively, described for panel C. EMSAs of the in vitro binding of RNS1-DBD to motif 3 (E), motif 4 (F), motif 5 (G), and motif 6 (H). Note that the shift of the labeled DNA was not seen with these four DNA motifs. The names of the genes whose promoters contain a motif are shown at the bottom. (I) EMSA of the importance of the nucleotides in the BM1 motif in its binding to RNS1-DBD. (Top) Sequences of the DNA probes containing the BM1 motif (10 nucleotides shown in bold) within the promoter of the gene MAA_10686. The name of the DNA probe is shown in the left, and its sequence is in the right. BM1, the WT DNA probe; M1, the nucleotide at position 1 in the BM1 motif is mutated into a nucleotide shown in bold lowercase letter. The naming system is also used for all other mutated DNA probes. (Bottom) EMSA of the binding of the DNA probe to RNS1-DBD. The binding activity was demonstrated by the labeled DNA band shift prior to the addition of the specific competitor (the unlabeled WT DNA probe) in a 200-fold excess. The importance of each nucleotide in the BM1 motif in the binding of the DNA probe to RNS1-DBD is shown by the impact of the addition of an unlabeled mutated probe as a competitor in a 200-fold excess on the labeled DNA band shift. The names of the unlabeled mutated DNA probes are shown above their respective lanes. (J) Importance of the nucleotides in the BM2 motif in its binding to RNS1-DBD. The legends for the top and bottom panels are the same as described for panel I for the BM1 motif. The DNA probe containing the BM2 motif is within the promoter of the gene MAA_05782. Download FIG S4, TIF file, 2.9 MB.

For electrophoretic mobility shift assays (EMSAs), we failed to express the whole protein of RNS1 in Escherichia coli. However, RNS1-DBD was successfully expressed and purified (Fig. S4C and D) for assaying the binding of RNS1-DBD to six biotin-labeled DNA probes, each containing one of the six consensus motifs (Fig. S4E to H). The recombinant RNS1-DBD protein bound to two DNA probes, which contained the 10-nucleotide (nt) motif BM1 [G(A/T)T(C)CA(G)AC(T/G)T(A)GG(C/A)T(C)] and the 7-nt motif BM2 (ACCAGAC) (Fig. 3A and D). For these two DNA probes, their respective specific competitor (unlabeled DNA probe) abolished the DNA band shift (Fig. 3B and E). Chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) analysis showed that the copy number of the enriched DNA fragment containing the BM1 motif in the promoter of the gene MAA_10686 from the ΔRns1-RNS1-FLAG strain was 9-fold higher than in the ΔRns1-FLAG strain that expressed the FLAG tag (Fig. 3C). For the motif BM2 in the promoter of the gene MAA_05782, the copy number of the enriched DNA fragment from the ΔRns1-RNS1-FLAG strain was 4-fold higher than in the ΔRns1-FLAG strain (Fig. 3F).

FIG 3

Identification of the DNA motifs bound by RNS1. (A) Consensus DNA motif BM1. (B) EMSA confirmed the in vitro binding of the biotin-labeled BM1 to the recombinant protein RNS1-DBD. The binding activity was demonstrated by the shift of the labeled DNA band prior to the addition of the specific competitor (the unlabeled DNA probe) in a 200-fold excess. The tested motif BM1 is in the promoter of the gene MAA_10686. (C) ChIP-qPCR analysis confirms that RNS1 in vivo binds to the motif BM1 in the MAA_10686 promoter. Detailed description of the ΔRns1-RNS1-FLAG and ΔRns1-FLAG strains is shown in Table 1. (D) Consensus DNA motif BM2. (E) EMSA confirmed the in vitro binding of the biotin-labeled BM2 to the protein RNS1-DBD. The tested motif BM2 is in the promoter of the gene MAA_05782. (F) ChIP-qPCR analysis shows that RNS1 in vivo binds to the motif BM2 in the MAA_05782 promoter.

To assay the importance of the 10 nucleotides of the BM1 motif in its binding to RNS1-DBD, unlabeled BM1 mutants were constructed by replacing a consensus nucleotide with each of other three, which were subsequently used as competitors for the binding of RNS1-DBD to the biotin-labeled WT DNA probe. None of the mutations in the 10 consensus nucleotides completely abolished the shift of the band of the biotin-labeled WT DNA probe, but the mutations of the nucleotides at positions 8 and 9 had the least impact on the band shift (Fig. S4I). Similar experiments were also conducted for the motif BM2, and mutations of the nucleotides at positions 5 and 7 had no impact on the band shift of the biotin-labeled WT DNA probe, but all other mutated DNA probes completely abolished the band shift (Fig. S4J).

Phosphorylated RNS1 upregulates its own expression.

We then investigated how Rns1 was upregulated during cuticle penetration. A BM1 motif (GTCGACTGGC) and a BM2 motif (CCCAGAC) were found in the Rns1 promoter. During the cuticle penetration, ChIP-qPCR analysis showed that a higher copy number of the BM2 motif was enriched from the WT-RNS1-FLAG strain than from the WT-FLAG strain that expressed the FLAG (Fig. 4A). EMSA also showed that the recombinant RNS1-DBD protein bound to the BM2 motif (Fig. 4B). ChIP-qPCR analysis showed that RNS1 did not bind to the BM1 motif, though EMSA showed that RNS1-DBD in vitro weakly bound to this motif (see Fig. S5A and B).

FIG 4

Phosphorylation of RNS1 by Fus3 facilitates binding to its own promoter to self-induce expression during cuticle penetration. (A) ChIP-qPCR analysis of in vivo binding of RNS1 to the BM2 motif in its own promoter during cuticle penetration. Detailed description of the strains in this figure is shown in Table 1. (B) EMSA confirmed in vitro binding of the recombinant protein RNS1-DBD to the biotin-labeled BM2 motif in the Rns1 promoter. The specific competitor (unlabeled DNA motifs) was added in a 200-fold excess. (C) qRT-PCR analysis of Rns1 expression in the WT strain, the ΔFus3 mutant, and the strains obtained by transforming the ΔRns1 mutant with the wild-type genomic clone gRns1 or mutants of gRns1. qRT-PCR analysis (D) and Western blot analysis (E) of the expression of GFP in the strains with gfp driven by the wild-type Rns1 promoter PRns1 or its mutant PRns1Δ (PRns1 with the motif BM2 mutated) in cuticle medium (cuticle) or SDY medium (SDY).

RNS1 did not bind to the BM1 but BM2 motif during cuticle penetration. (A) EMSA confirmed in vitro binding of the recombinant protein RNS1-DBD to the biotin-labeled BM1 motif in the Rns1 promoter. The specific competitor (unlabeled DNA motifs) was added in a 200-fold excess. (B) ChIP-qPCR analysis of the binding of the fusion proteins RNS1::FLAG, RNS1T215A::FLAG, and RNS1S226A::FLAG to the BM1 motif in the Rns1 promoter in different strains during cuticle penetration. Note that RNS1 did not bind to the putative BM1 motif in its own promoter. (C) qRT-PCR analysis of the transcription of the genes encoding the fusion proteins RNS1::FLAG, RNS1T215A::FLAG, and RNS1S226A::FLAG in different strains. (D) ChIP-qPCR analysis of the binding of the protein RNS1::FLAG to the BM2 motif in the Pr1a (MAA_05675) and Pr1b (MAA_08168) promoters during cuticle penetration. Note that RNS1 bound to the BM2 motif. (E) ChIP-qPCR analysis of the binding of the fusion proteins RNS1::FLAG to the BM1 motif in the MAA_10686 promoter during cuticle penetration. Note that RNS1 did not bind to the BM1 motif. Detailed description of the strains is shown in Table 1. Download FIG S5, TIF file, 2.9 MB.

To assay the impact of the Fus3-mediated phosphorylation on the binding of RNS1 to the BM2 motif in the Rns1 promoter, we constructed three strains: WT-RNS1 expressing the protein RNS1T215A::FLAG, WT-RNS1 expressing RNS1S226A::FLAG, and ΔFus3-RNS1-FLAG with RNS1::FLAG expressed in the ΔFus3 mutant (Fig. S3A). During cuticle penetration, no difference in the transcription level of the fusion genes was found among the WT-RNS1-FLAG, WT-RNS1, WT-RNS1, and ΔFus3-RNS1-FLAG strains (Fig. S5C). ChIP-qPCR analysis showed that a higher copy number of the BM2 motif was enriched in the WT-RNS1-FLAG strain than in the WT-RNS1, WT-RNS1, and ΔFus3-RNS1-FLAG strains, but no significant difference was found between the latter three strains (Fig. 4A).

To investigated the impact of the binding of RNS1 to its own promoter on its expression, we constructed three mutated genomic clones of gRns1 (used for complementing the ΔRns1 mutant as described above): gRns1Δ by changing all consensus nucleotides in the motif BM2 to adenine, gRns1 by substituting alanine for Thr-215, and gRns1 by replacing Ser-226 with alanine. gRns1Δ, gRns1, and gRns1 were then transformed into the ΔRns1 mutant to produce ΔRns1-gRns1Δ, ΔRns1-gRns1, and ΔRns1-gRns1 strains, respectively (Fig. S2C). During cuticle penetration, qRT-PCR showed that Rns1 was more highly expressed in the WT strain than in the ΔRns1-gRns1Δ, ΔRns1-gRns1, ΔRns1-gRns1, and ΔFus3 strains, and no significant difference was found between the latter four strains. The C-ΔRns1 strain had the same transcription level of Rns1 as the WT strain (Fig. 4C).

Another assay was conducted to confirm that Rns1 self-induces expression. To this end, we constructed two gfp gene cassettes: PRns1-gfp with gfp driven by Rns1 promoter PRns1, and PRns1Δ with gfp driven by the promoter PRns1Δ with the motif BM2 mutated in PRns1. These two cassettes were transformed into the WT strain to produce WT-PRns1-gfp and WT-Pns1Δ strains, into the ΔRns1 mutant to generate ΔRns1-PRns1-gfp and ΔRns1-PRns1Δ strains, and into the ΔFus3 mutant to form ΔFus3-PRns1-gfp and ΔFus3-PRns1Δ strains. For all strains, two independent isolates showed the same results, and so only one of them is presented here. For the WT-PRns1-gfp strain, compared with that during saprophytic growth, the expression levels of GFP transcript and protein were upregulated during cuticle penetration (Fig. 4D and E), showing that the Rns1 promoter PRns1 retains its activity in its nonnative chromosomal positions. During cuticle penetration, GFP transcript and protein were more highly expressed in the WT-PRns1-gfp strain than in the WT-PRns1Δ, ΔRns1-PRns1-gfp, ΔRns1-PRns1Δ, ΔFus3-PRns1-gfp, and ΔFus3-PRns1Δ strains, whereas no differences were found between the latter five strains (Fig. 4D and E).

RNS1 induces cuticle-degrading genes.

RNA-Seq analysis was used to profile genes regulated by RNS1 during cuticle penetration. Compared with expression in the WT strain, 285 genes were downregulated and 233 genes upregulated in the ΔRns1 mutant. ChIP-qPCR analysis showed that during cuticle penetration, RNS1 bound to the BM2 motif in the promoters of two protease genes (Pr1a [MAA_05675] and Pr1b [MAA_08168]) that were regulated by RNS1 (see below) but not to the BM1 motif in the promoter of MAA_10686 that was not regulated by RNS1 during cuticle penetration (Fig. S5D and E). Therefore, RNS1 only recognized the BM2 motif during cuticle penetration, and we thus searched for this motif in the promoters of the differentially expressed genes profiled by the RNA-Seq analysis. Among the 285 downregulated genes in the ΔRns1 mutant, 262 had the BM2 motif in their promoters, and this motif was also found in the promoters of 226 upregulated genes. Particularly, among the 262 downregulated BM1 motif-containing genes, 28 encoded cuticle-degrading enzymes, including 15 for lipid utilization, 11 proteases, and 2 chitinases.

Cuticle-degrading enzymes are functionally redundant, and their expression has synergistic impacts (12); we thus investigated the extent to which RNS1 controlled overall activity of cuticle-degrading enzymes. Compared with that in the WT strain, the ΔRns1 mutant secreted significantly less Pr1 subtilisin protease activity (P < 0.05), though total extracellular protease activity was not altered in the mutant (see Fig. S6B and C). Likewise, the WT strain produced significantly more chitinases and lipases than the ΔRns1 mutant (Fig. S6D and E). We further quantified the cuticle degradation products following 12-h fungal growth in the cuticle medium. Although the WT strain had the same biomass as the ΔRns1 mutant, the mutant released significantly (P < 0.05) fewer amino acids from the cuticle (Fig. S6A, F, and G).

RNS1 regulates cuticle-degrading enzymes during cuticle penetration. For quantification of free amino acids, peptides, and activities of cuticle degrading enzymes, mycelium grown in SDY medium for 36 h was transferred to cuticle medium using the locust cuticle as the sole carbon and nitrogen source. (A) Wet weight of mycelium after 12 h of growth in cuticle medium. Total extracellular protease (B), Pr1 protease (C), chitinase (D), and lipase (E) activity in the culture supernatant. Concentrations of peptides (F) and free amino acids (G) in culture supernatants. The control was cuticle medium that was not inoculated with the mycelium. All assays were repeated three times. Values with different lowercase letters are significantly different (n = 3, P < 0.05, Tukey’s test in one-way ANOVA). WT, WT strain; ΔRns1, Rns1 deletion mutant; C-ΔRns1, complemented ΔRns1 strain. Download FIG S6, TIF file, 1.7 MB.

The Fus3/RNS1 cascade regulates utilization of less-favored carbon and nitrogen sources.

As described above, RNS1 regulates the degradation of cuticular protein, chitin, and lipids, i.e., less-favored carbon and nitrogen sources. We thus postulated that it also regulated utilization of non-insect-derived less-favored carbon and nitrogen sources to support saprophytic growth. To test this postulation, we first compared mycelial dry weight of the WT with the that of the ΔRns1 mutant when they were grown for 48 h in the liquid medium containing less-favored carbon and nitrogen sources that were not derived from insects. As in other fungi, glucose is also a favored carbon source for Metarhizium fungi (21). It has not been documented which amino acids Metarhizium fungi prefer, and we assume that glutamine and ammonium are preferred by this group of fungi, because these two nutrients are reported to be favored nitrogen sources for many fungi (10). In this study, protein, lipids, and chitin are designated complex less-favored carbon or nitrogen sources, while some sugars such as raffinose are defined as less-favored carbon sources and amino acids such as arginine and proline are less-favored carbon and nitrogen sources. In the nutrient-rich medium SDY, the mycelial dry weight of the ΔRns1 mutant was the same as the WT strain (see Fig. S7A). In the medium with the non-insect-derived protein casein (named casein medium) or colloid chitin (chitin medium) as the sole carbon and nitrogen sources, the dry weight of the WT mycelium was significantly higher than that of the ΔRns1 mutant, but no significant difference was found between the two strains when proline, arginine, glutamine, or N-acetylglucosamine (the chitin monomer) was used as the sole carbon and nitrogen source (Fig. S7B and D). In the medium using the non-insect-derived lipid pentadecane (lipid medium) as the sole carbon source (ammonium was the nitrogen source), the mycelial dry weight of the ΔRns1 mutant was significantly lower than that of the WT strain (Fig. S7C), and no difference was found between these two strains when glucose, trehalose, or raffinose was used as the sole carbon source (Fig. S7C). In all media, no significant difference in mycelial dry weight was found between the WT strain and the complemented C-ΔRns1 strain (Fig. S7). On the nutrient-rich medium potato dextrose agar (PDA), the conidial yield of the ΔRns1 mutant was significantly lower than that of the WT strain and the complemented C-ΔRns1 strain. On the lipid, chitin, or casein medium, the WT strain still produced more conidia than the ΔRns1 mutant (Fig. S7E).

Mycelial dry weight and conidial yield in different media. (A) Nutrient-rich medium SDY. Conidia (108) were grown in 100 ml of the medium for 36 h. (B) Media using free amino acids or casein as the sole carbon and nitrogen source. (C) Media using sugars or the lipid pentadecane as the carbon source and ammonium as the nitrogen source. (D) Media using colloid chitin or N-acetylglucosamine (chitin monomer) as the sole carbon and nitrogen source. In panels B to D, the mycelium was first grown in SDY medium, which (0.2 g [wet weight]) was then inoculated into the specified media. After 48 h of growth at 26°C, the mycelial dry weight was measured. Values with different lowercase letters are significantly different (P < 0.05, Kruskal-Wallis test). (E) Conidial yield on PDA, lipid, casein, or chitin medium. WT, wild-type strain; ΔRns1, Rns1 deletion mutant; C-ΔRns1, complemented ΔRns1 strain. Download FIG S7, TIF file, 2.8 MB.

We then compared Rns1 expression between the media with different carbon and nitrogen sources. The expression level of Rns1 in the SDY medium was lower than in the lipid, casein, and chitin media, but no difference was found between the SDY medium and the media without carbon or nitrogen sources, i.e., under carbon or nitrogen starvation (see Fig. S8A). Rns1 was more highly expressed in the lipid medium than in the medium using glucose, trehalose, maltose, sucrose, fructose, or raffinose as the sole carbon source, and no difference was found between these six sugar-containing media (Fig. 5A). Glucose suppressed Rns1 expression in the lipid medium (Fig. 5A). Rns1 was 5-fold more highly expressed in the casein medium than in the media with a free amino acid (glutamine, proline, or arginine) as the sole carbon and nitrogen source, and no difference was found between these amino acid-containing media. A free amino acid significantly reduced Rns1 expression in the casein medium, but the inorganic favored nitrogen source ammonium had no impact on Rns1 expression in the casein medium (Fig. 5B). In the chitin medium, the expression level of Rns1 was 5-fold higher than that with N-acetylglucosamine as the sole carbon and nitrogen source, and N-acetylglucosamine inhibited Rns1 expression in the chitin medium (Fig. 5C). Glucose, glutamine, arginine, proline, and N-acetylglucosamine all repressed Rns1 expression in the cuticle medium (Fig. 5D), and again, ammonium had no impact on Rns1 expression in the cuticle medium (Fig. 5D).

FIG 5

Fus3 regulates RNS1 during utilization of the non-insect-derived less-favored nutrients chitin, casein, and lipid (pentadecane is used as a representative of lipids). qRT-PCR analysis was conducted to analyze gene expression. (A) Rns1 expression in the media with favored (glucose) and less-favored (lipid, fructose, maltose, trehalose, raffinose, and sucrose) carbon sources. The nitrogen source was (NH4)2SO4. (B) Rns1 expression in the media with favored [glutamine, (NH4)2SO4] and less-favored (casein, proline, or arginine) carbon and nitrogen sources. (C) Rns1 expression in the media containing chitin or N-acetylglucosamine (GlcNAc; the chitin monomer) as the sole carbon and nitrogen source. (D) Rns1 expression in te cuticle medium with or without glucose, glutamine, proline, arginine, (NH4)2SO4, or N-acetylglucosamine. (E) Rns1 expression in the WT strain and the ΔFus3 mutant when they were grown in lipid, chitin, or casein medium. (F) Phos-tag analysis of RNS1 phosphorylation in lipid, chitin, or casein medium. 1, WT-RNS1-FLAG strain; 2, ΔFus3-RNS1-FLAG strain.

RNS1 upregulates its own expression in media with non-insect-derived complex less-favored carbon and nitrogen sources (casein, chitin or lipids [pentadecane is used as a representative]). (A) qRT-PCR analysis of Rns1 expression in SDY, lipid, casein, and chitin media and in the media without carbon or nitrogen source, i.e., carbon and nitrogen starvation. Expression of the GFP transcript (qRT-PCR analysis) (B) and GFP protein (Western blotting) (C) by the WT-PRns1-gfp strain in the media with different carbon and nitrogen sources. When glucose, raffinose, or lipid was used as the sole carbon source, the nitrogen source was ammonium. In other media, the sole carbon and nitrogen source was shown as their name: glutamine, glutamine was the sole carbon and nitrogen source; GlcNAc, N-acetylglucosamine. Expression of the GFP transcript (top) and protein (bottom) in six different strains when they were grown in lipid medium (D), in casein medium (E), and in chitin medium (F). Detailed description of the strains is shown in Table 1. Download FIG S8, TIF file, 2.9 MB.

Consistent with results regarding Rns1 transcription (Fig. 5), in the WT-PRns1-gfp strain, the expression levels of GFP transcript and protein in the casein, chitin, or lipid medium were higher than in the media containing glucose, raffinose, glutamine, arginine, proline, or N-acetylglucosamine (Fig. S8B and C).

Compared with that in the WT strain, the expression level of Rns1 was lower in the ΔFus3 mutant in the casein, chitin, or lipid medium (Fig. 5E). Also, in these three media, Phos-tag assays showed that Fus3 phosphorylated the RNS1 protein (Fig. 5F).

Another assay also showed that the Fus3 regulated Rns1 expression for utilizing protein, chitin, and lipids. When grown in the casein, chitin, or lipid medium, the expression levels of the GFP transcript and protein in the WT-PRns1-gfp strain were significantly higher than in the WT-PRns1Δ, ΔRns1-PRns1-gfp, ΔRns1-PRns1Δ, ΔFus3-PRns1-gfp, and ΔFus3-Pns1Δ strains (Fig. S8D to F), but no differences were found between the latter five strains (Fig. S8D to F).

RNS1 regulates genes for utilizing protein, chitin, and lipids.

We further investigated whether RNS1 regulated the genes for utilizing casein, chitin, and lipids. In the M. robertsii genome, there are genes encoding 34 lipases, 29 chitinases, and 122 proteases (11). The BM2 motif was identified in the promoters of 29 chitinase genes, 34 lipase genes, and 107 protease genes. Using two proteases (Pr1a and Pr1b), two lipases (MAA_08921 and MAA_01415), and two chitinases (MAA_01212 and MAA_10080) as their representatives, we assayed whether RNS1 regulated these BM2 motif-containing genes. Compared with that during growth in the SDY medium, the expression of these representative protease, chitinase, and lipase genes was upregulated in the casein, chitin, and lipid media, respectively (Fig. 6A, B, and C). Compared with expression in the WT strain, these representative genes were all downregulated in the ΔRns1 mutant (Fig. 6D). Ammonium had no impact on Pr1b expression in the casein medium but suppressed Pr1a expression. An amino acid (glutamine, arginine, or proline) repressed the expression of Pr1a and Pr1b in the casein medium (Fig. 6A). In the chitin medium, ammonium had no impact on the expression of two chitinase genes, but N-acetylglucosamine inhibited their expression (Fig. 6B). Glucose inhibited the expression of the two lipase genes in the lipid medium (Fig. 6C).

FIG 6

RNS1 regulates the expression of proteases, chitinases, and lipases for utilization of the protein casein, chitin, and lipids that are not derived from insects. qRT-PCR analysis was conducted to analyze gene expression. (A) Expression of two protease genes (Pr1a and Pr1b) in SDY (SDY) and casein medium with or without free amino acids and ammonium. (B) Expression of two chitinase genes (MAA_01212 and MAA_10080) in SDY and the chitin medium with or without N-acetylglucosamine and ammonium sulfate. (C) Expression of two lipase genes (MAA_01415 and MAA_08921) in SDY medium and the lipid medium with or without glucose. (D) Expression in the ΔRns1 deletion mutant of the protease genes in casein medium, chitinase genes in chitin medium, and lipase genes in lipid medium relative to that in the WT strain.

Relationship between RNS1 and other regulators of carbon and nitrogen metabolism.

We further investigated the relationship between RNS1 and other previously known regulators of carbon and nitrogen metabolism. We identified the regulator AREA (MAA_03820) in NMR, and CRR1 (MAA_06444) in CCR, and constructed their deletion mutants, ΔAreA and ΔCrr1 (Fig. S2D and E). In the cuticle medium with or without glucose, no differences in Rns1 expression were found between the WT strain and the ΔCrr1 mutant (see Fig. S9A). Since RNS1 upregulated the lipase genes MAA_08921 and MAA_01415 in the lipid medium (Fig. 6C) and glucose suppressed their expression (Fig. S9C), these two lipase genes were used as indicators to assay whether RNS1 regulated CRR1-mediated glucose repression. qRT-PCR analysis showed that CRR1 mediated glucose repression of MAA_08921, but RNS1 had no impact on this repression (Fig. S9C). Both CRR1 and RNS1 had no impact on the glucose repression of MAA_01415 (Fig. S9C). In the lipid medium, the mycelial dry weight of the WT strain was significantly higher than that of the ΔCrr1 mutant, which was in turn higher than that in the ΔRns1 mutant (Fig. S9D).

The Fus3/RNS1 cascade is not directly related to other known regulators of carbon and nitrogen metabolism. (A) Rns1 expression in the WT strain and the ΔCrr1 mutant in cuticle medium with or without glucose. Unless otherwise indicated, gene expression presented in this figure was analyzed by qRT-PCR analysis. (B) Rns1 expression in the ΔAreA and ΔSnf1 mutants and the knockdown mutants of the gene Tor (Tor), Tps1 (Tps1), and G6PD (G6PD) in cuticle medium. (C) Expression of two lipase genes in the WT strain and the ΔRns1 and ΔCrr1 deletion mutants in lipid medium with or without glucose. (D) Mycelial dry weight of the WT strain and the ΔRns1, ΔCrr1 and ΔAreA mutants in lipid, casein, or chitin medium. Values with different lowercase letters are significantly different (P < 0.05, Kruskal-Wallis test). (E) NAD+ production when the WT and ΔRns1 strains were grown in SDY medium or insect cuticle medium. (F) RNA-Seq analysis of the expression of 10 Nmr genes in the WT and ΔRns1 strains in cuticle medium. The expression level of a gene in the #1 repeat of the WT strain is set to 1; the values represent the log2-transformed fold changes of differential gene expression in other treatments. #1, #2, and #3, three repeats. Expression of two protease genes in casein medium (G) and two chitinase genes in chitin medium (H) in the WT strain and the ΔRns1 and ΔAreA mutants. (I) Expression of five proline utilization genes in the WT strain and the ΔRns1 and ΔAreA mutants in the medium with proline as the sole carbon and nitrogen source. PrnA, MAA_08437 (GenBank accession number); PrnB, MAA_01021; PrnC, MAA_07805; PrnD, MAA_07804; PrnX, MAA_10225. (J) Expression in the WT strain and the ΔRns1 mutant of previously reported regulators (AreA, Crr1, Snf1, Tps1, Tor, and G6PD) in cuticle medium. Download FIG S9, TIF file, 2.9 MB.

In the cuticle medium, no difference in Rns1 expression was found between the WT and the ΔAreA mutant (Fig. S9B). The protease genes (Pr1a and Pr1b) and the chitinases (MAA_01212 and MAA_10080) were used as indicators to test if RNS1 regulated AREA to control genes for utilizing alternative carbon and nitrogen sources. In the casein medium, no difference in Pr1b expression was found between the WT strain and the ΔAreA mutant, but both strains had higher expression levels than the ΔRns1 mutant (Fig. S9G). Pr1a was more highly expressed in the WT strain than in the ΔAreA and ΔRns1 mutants, and no difference was found between these two mutants (Fig. S9G). In the casein medium, the mycelial dry weight of the ΔRns1 mutant was the same as that of the ΔAreA mutant (Fig. S9D). In the chitin medium, the two chitinase genes were regulated by Rns1 (Fig. S9H) but not by AreA; the mycelial dry weight of the ΔAreA mutant was the same as that of the ΔRns1 mutant but was lower than that of the WT strain (Fig. S9D). AREA positively regulated four proline utilization genes, but no difference in expression of these four genes was found between the WT strain and the ΔRns1 mutant (Fig. S9I). In the cuticle medium, the WT strain had the same NAD+ level as the ΔRns1 mutant (Fig. S9E), and RNA-Seq analysis showed that no difference in expression of the 10 Nmr genes was found between the WT and ΔRns1 strains (Fig. S9F).

In addition to AREA and CRR1, the Tps1 pathway (9), Snf1 (22), and a TOR kinase (23) have been reported to regulate fungal carbon and nitrogen metabolism. The genes encoding Snf1 (MAA_04401), the TOR kinase (MAA_02388), and two genes in the Tps1 pathway (Tsp1 [MAA_04676], and G6PD [MAA_00144]) were also identified in M. robertsii. In the cuticle medium, the expression levels of Tsp1, G6PD, Snf1, and Tor in the WT strain were not different from the those in the ΔRns1 mutant (Fig. S9J). The deletion mutant of Snf1 and the knockdown mutants of Tor, Tps1, and G6PD were constructed (Fig. S2F and G), and in the cuticle medium, no significant difference in Rns1 expression was found between the WT strain and the mutants (Fig. S9B).

DISCUSSION

The entomopathogenic fungus M. robertsii secretes numerous enzymes to degrade cuticular lipids, protein, and chitin and thereby obtains nutrients for hyphal growth and enters the hemocoel for further colonization. The Fus3-MAPK cascade regulates cuticle-degrading genes and is indispensable for cuticle penetration (14). In this study, we found that Fus3 activated the expression of the cuticle-degrading genes via the transcription factor RNS1. On the insect cuticle, Fus3 phosphorylates RNS1. The phosphorylated RNS1 migrates into the nucleus, binds to its own promoter, and self-upregulates its transcription, which then induces the expression of the cuticle-degrading genes. The deletion mutant of the Fus3 gene was nonpathogenic (14), whereas deleting Rns1 only resulted in partial loss of virulence. Therefore, RNS1 regulates a subset of pathogenicity genes controlled by the Fus3-MAPK cascade.

On an inductive milieu, single-celled conidia of M. robertsii need external nutrients for germination and following hyphal growth. The infection structure appressorium then forms on the tip of a multicellular hypha (15). Deleting Rns1 suppressed the expression of cuticle-degrading genes and reduced the ability to exploit the cuticular lipids, protein, and chitin as nutrients and thereby slowed hyphal growth and delayed appressorial formation. In addition to the cuticle-degrading genes, RNA-Seq analysis showed that Rns1 also regulated many other genes, but none of them were previously characterized genes related to appressorium formation. Rns1 is also not involved in appressorial turgor pressure. Therefore, it remains to be resolved whether Rns1 regulates appressorial formation itself. When M. robertsii reaches the hemocoel where free amino acids and sugars such as trehalose and glucose are available, these nutrients suppress Rns1 expression, explaining why no significant difference in virulence was observed between the WT strain and the ΔRns1 mutant when inoculation was conducted by direct injection of conidia into the hemocoel. This is consistent with our previous findings that the Fus3-MAPK cascade is not involved in hemocoel colonization (14).

In addition to insect cuticular lipids, protein, and chitin, Rns1 also regulates exploitation of non-insect-derived less-favored carbon and nitrogen sources as nutrients for saprophytic growth. RNS1 differs in two aspects from the transcription factor AREA. First, RNS1 is only involved in utilization of the complex less-favored carbon and nitrogen sources (proteins and chitin), whereas in addition to the complex carbon and nitrogen nutrients, AREA also regulates utilization of less-complex alternative carbon and nitrogen sources such as free amino acids. Second, the favored nitrogen sources affect AREA and RNS1 activity differently. Both ammonium and glutamine inhibit AREA activity (3). Glutamine also suppresses RNS1 activity, but ammonium’s impacts on RNS1 vary depending on its target genes. Ammonium had no impact on RNS1-mediated upregulation of Pr1b and two chitinase genes, whereas it suppressed the induction of Pr1a by RNS1. Nevertheless, both RNS1 and AREA activated Pr1a expression, and their mutants showed reduced mycelial growth in the casein medium, suggesting RNS1 and AREA have overlap in controlling the utilization of alternative nitrogen sources. However, RNS1 and AREA did not regulate each other’s transcription, and RNS1 also did not control two other components (NAD+ and nmr genes) in NMR. Therefore, the relationship between RNS1 and NMR remains to be resolved. Genetic and transcriptional analysis also failed to identify the relationship between RNS1 and other regulators of fungal carbon and nitrogen metabolism, including CRR1, Tps1, Snf1, and TOR kinase.

In conclusion, the Fus3-MAPK and the transcription factor RNS1 constitute a regulatory cascade that activates the utilization of the complex less-favored nitrogen and carbon sources by M. robertsii. This cascade enables this fungus to degrade cuticular protein, lipid, and chitin to obtain nutrients for hyphal growth and enter the hemocoel for further infection. The Fus3-MAPK cascade exists across the fungal kingdom, and RNS1 homologs are also widely found in ascomycete filamentous fungi, including saprophytes and pathogens with diverse hosts, suggesting that regulation of the utilization of alternative nitrogen and carbon sources by the Fus3/RNS1 cascade could be widespread.

MATERIALS AND METHODS

Gene deletion and gene knockdown.

M. robertsii ARSEF 2575 was obtained from the Agricultural Research Service Collection of Entomopathogenic Fungi (US Department of Agriculture). E. coli DH5α was used for plasmid construction. Agrobacterium tumefaciens AGL1 was used for fungal transformation as previously described (24).

Gene deletion and complementation of the deletion mutants were performed as previously described (25). Gene knockdown using the antisense RNA method was performed as described previously (12). Briefly, an ∼200-bp DNA fragment corresponding to part of the coding sequence of a target gene was inserted downstream of the constitutive promoter Ptef in the plasmid pPK2-TEF (15) to form pPK2-bar-GFP-RNAi, which was then transformed into M. robertsii mediated by A. tumefaciens. Gene knockdown was confirmed by qRT-PCR analysis. All primers used in this study are presented in Table S1 in the supplemental material.

Primers used in this study. Download Table S1, DOCX file, 0.03 MB.

Pathogenicity assays.

Pathogenicity assays were conducted using last-instar G. mellonella larvae (RuiQing Bait Co., Shanghai, China). Inoculations were performed by topical application of conidia on the insect cuticle or by injection of conidia into the insect hemocoel (26). Mortality was recorded daily, and the LT50 values were determined using the SPSS statistical package (SSPS, Chicago, IL, USA). All bioassays were repeated three times with 40 insects per repeat.

Appressorial formation on the hydrophobic surface of a petri dish (Corning, NY, USA) was assayed as previously described (15). Turgor pressure of appressorium was measured as previously described (27).

Assays of mycelial growth and conidial yield.

To assay the mycelial growth in SDY medium, 108 conidia were inoculated into 100 ml of SDY and cultured at 26°C for 36 h with 160-rpm shaking. The mycelium was then collected by filtration and dried by lyophilization, and the mycelial dry weight was then measured. To measure mycelial growth in other media, the mycelium collected from the SDY medium (0.3 g [wet weight]) was inoculated into a medium. After 48 h of growth at 26°C, the mycelial dry weight was measured.

Yeast two-hybrid assays.

The interaction between RNS1 and Fus3 was assayed using yeast two-hybrid assays (Clontech, Japan). The coding sequences of RNS1 and Fus3 were cloned by PCR and inserted into the plasmids pGADT7 and pGBKT7 to produce the plasmids pGADT7-RNS1 and pGBKT7-Fus3, respectively. The plasmid pGBKT7-Fus3 was transformed into Y2HGold cells, and pGADT7-RNS1 was transformed into Y187 cells. The resulting strain from mating was grown on medium (SD-His-Ade-Leu-Trp) supplemented with X-α-Gal (5-bromo-4-chloro-3-indolyl-α-d-galactopyranoside) and aureobasidin A (AbA; TaKaRa Bio). The autoactivation of Fus3 was tested by inoculating the strain containing the plasmid pGBKT7-Fus3 on the medium (SD-His-Ade-Trp) with X-α-Gal. Yeast two-hybrid and autoactivation assays were repeated three times.

Co-IP analysis.

The master plasmid pPK2-sur-Ptef-FLAG or pPK2-bar-Ptef-HA was used for producing a protein fused with FLAG or HA, respectively (26). To assay the in vivo interaction between RNS1 and Fus3, an RNS1-FLAG/Fus3-HA strain expressing the fusion protein RNS1::FLAG and Fus3::HA was constructed. The coding sequence of RNS1 was inserted into pPK2-sur-Ptef-FLAG, and the resulting plasmid was then transformed into the WT strain to produce the WT-RNS1-FLAG strain. The coding sequencing of Fus3 was inserted into the plasmid pPK2-bar-Ptef-HA to produce pPK2-sur-Ptef-Fus3-HA, which was transformed into the WT-RNS1-FLAG strain to produce the RNS1-FLAG/Fus3-HA strain.

Extraction of total fungal proteins and Co-IP analysis were performed as described previously (26). Protein concentration was determined using the bicinchoninic acid (BCA) protein assay kit (Meilune, Dalian, China). The mouse anti-HA antibody was purchased from Abclonal (Hangzhou, China), and the rabbit anti-FLAG antibody was purchased from Sigma-Aldrich (MO, USA). Dynabeads protein G beads were purchased from Invitrogen (CA, USA). Detection of a target protein was conducted using Western blot analysis. All blots were imaged by a chemiluminescence detection system (Clarity Western ECL; Bio-Rad). All Co-IP assays were repeated three times.

BiFC assay.

The coding sequences of YFPN (Met-1 to Asp-174) and YFPC (Gly-175 to Lys-239) were cloned by PCR as described previously (28), and the resulting PCR products were inserted into the BamH I/EcoRV sites of pPK2-Ptef-bar and pPK2-Ptef-sur to produce the master plasmids pPK2-bar-Ptef-YFPN and pPK2-sur-Ptef-YFPC, respectively. The coding sequences of RNS1-DBD (Glu-61 to Pro-267) and Fus3 were inserted into the Bam I/Eco I sites of pPK2-bar-Ptef-YFPN and pPK2-sur-Ptef-YFPC, respectively, to produce pPK2-bar-Ptef-YFPN-RNS1-DBD and pPK2-sur-Ptef-YFPC-Fus3. The resulting plasmids were transformed into the WT strain to produce the RNS1-DBD-YFP strain. As a control, the plasmids pPK2-bar-Ptef-YFPN-RNS1-DBD and pPK2-sur-Ptef-YFPC were transformed into the WT strain to produce the RNS1-DBD-YFP strain. YFP fluorescence was analyzed using a Leica TCS SP2 laser confocal scanning microscope (Germany).

Phos-tag analysis.

The mutagenesis kit (NEB, UK) was used to construct the mutated RNS1 proteins RNS1T215A, RNS1S226A, and RNS1T215A/S226A (described in Results).

The mycelium cultured in SDY medium was used for analysis of protein phosphorylation during saprophytic growth. Since it was not possible to obtain enough biomass to get sufficient protein for Phos-tag analysis by growing M. robertsii on the locust hindwings, we prepared the mycelium by growing the fungus in cuticle medium (basal salt medium containing 1% locust cuticle as sole carbon and nitrogen sources). To do this, the mycelium from the SDY medium was collected by filtration, washed with ample sterile water three times, and then grown in the locust cuticle medium for 1 h with 160-rpm shaking.

The Phos-tag analysis was conducted as described previously (29). The mycelium (∼0.7 g) was ground into fine powder in liquid nitrogen and then suspended in 3 ml of lysis buffer (26). After being incubated at 4°C for 6 to 8 h, the mixture was centrifuged, and the supernatant was used for Phos-tag analysis. An SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) gel (8%) containing 25 μM Phos binding reagent acrylamide (APExBIO, USA) and 100 μM MnCl2 was used. After electrophoresis, proteins in the gel were blotted to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad, USA) at 100 V for 6 h, which was then subjected to Western blotting using the anti-FLAG antibody (Huabio, China).

ChIP-Seq and ChIP-qPCR analyses.

ChIP-Seq analysis was conducted by the company Igenebook (Wuhan, China). The anti-FLAG antibody was used for immunoprecipitation, and DNA fragments enriched by ChIP were sequenced on an Illumina HiSeq 2000. After removing sequencing adaptors and low-quality bases, the clean reads were mapped to the M. robertsii genome using the software BWA (version 0.7.15-r1140) (30). The peak caller MACS (version, 2.1.1.20160309; q value < 0.05) was used to localize the potential binding sites of RNS1 (31).

ChIP-qPCR analysis was conducted as previously described (26). ChIP-enriched DNA was used as a template for quantitative PCR analysis using Thunderbird SYBR qPCR mix without ROX (Toyobo, Japan). All ChIP-qPCR analyses were repeated three times.

Expression and preparation of RNS1-DBD in E. coli.

To express RNS1-DBD in E. coli, its coding sequence was inserted into the BamHI/EcoRI sites of the plasmid pET-SUMO (Invitrogen, USA), and the resulting plasmid was then transformed into E. coli BL-21. The expression of the recombinant protein was induced by isopropyl β-d-1-thiogalactopyranoside at 18°C for 16 h (Novagen, Madison, WI, USA). The fusion protein SUMO::RNS1-DBD was partially purified with HisPur Ni-nitrilotriacetic acid (NTA) resin (Thermo Fisher Scientific, MA, USA), and the SUMO tag was then removed with laboratory-prepared SUMO protease ULP1. The recombinant RNS1-DBD protein was further purified to homogeneity with HisPur Ni-NTA resin.

EMSA.

To prepare a biotin-labeled double-stranded DNA probe, one of the DNA strands was biotin-labeled and annealed to its complementary strand (unlabeled) according to the manufacturer’s instructions (TsingKe Biological Technology, Hangzhou, China). To prepare an unlabeled WT DNA probe, two regular DNA strands were commercially synthesized and annealed to form the probe. A mutated DNA probe was prepared exactly as the WT DNA probe with one nucleotide in the RNS1 binding motif replaced with each of the other three. EMSAs were conducted using the LightShift Chemiluminescent EMSA kit (Thermo Fischer Scientific, MA, USA). In the competition assays, an unlabeled DNA probe was added in a 200-fold excess.

RNA-Seq and qRT-PCR analyses.

Total RNA was extracted with TRIzol reagent (Life Technologies, USA). RNA-Seq analysis was conducted by Personal Gene Technology (Nanjing, China). Paired-end sequencing was performed on an Illumina HiSeq 2000 sequencing platform. Clean reads were mapped to the M. robertsii genome using software HISAT2. Reads that aligned uniquely to the reference sequence were used for gene expression quantification using the fragments per kilobases per million fragments (FPKM) method. Differential expression analysis was performed with DESeq software with an adjusted P value of 0.05 (Benjamini-Hochberg method).

For qRT-PCR analysis, cDNAs were synthesized with total RNAs using ReverTra Ace qPCR RT master mix (Toyobo, Osaka, Japan). qRT-PCR analysis was conducted using Thunderbird SYBR qPCR mix without ROX (Toyobo). The reference genes Gpd and tef were used as described previously (32). The relative expression level of a gene was calculated using the comparitive threshold cycle (2−ΔΔ) method (33). All qRT-PCR experiments were repeated three times.

Assay of subcellular localization of RNS1-DBD.

To assay the impact of phosphorylation of RNS1 by Fus3 on its cellular localization, GFP-tagged proteins RNS1-DBD::GFP, RNS1T215A-DBD::GFP, and RNS1S226A-DBD::GFP were constructed. To do this, the coding sequences of RNS1-DBD, RNS1T215A-DBD, and RNS1S226A-DBD were individually inserted into the BamHI/EcoRI sites of the plasmid pPK2-Ptef-GFP-N (26), and the resulting plasmids were transformed into the WT strain to produce WT-RNS1-DBD-GFP, WT-RNS1, and WT-RNS1 strains and into the ΔFus3 mutant to form ΔFus3-RNS1-DBD-GFP, ΔFus3-RNS1, and ΔFus3-RNS1 strains. The expression and integrity of the fusion proteins were analyzed with Western blot analysis using anti-GFP antibody (Huabio, Hangzhou, China). The subcellular localization of a fusion protein was determined by following the GFP signal. The nuclear was visualized by DAPI (4′,6-diamidino-2-phenylindole) staining.

Quantification of amino acids and peptides and assays of the activities of cuticle-degrading enzymes.

Quantification of amino acids and peptides released during enzymolysis of the insect cuticle was conducted as previously described (12). Briefly, the mycelium (1 g [wet weight]), collected from the SDY medium, was inoculated into 100 ml of the cuticle medium, followed by 12 h of incubation at 26°C with 160-rpm shaking. Free amino acids in the supernatant were quantified with an amino acid quantification kit (Solarbio, Shanghai), and peptides were quantified using the Bradford assay kit (Bio-Rad, USA). The activity of the total extracellular proteases was assayed using the Azocasein kit (Sigma, USA). One unit of proteolytic activity was defined as an increase of 0.01 absorbance at 440 nm after 1 h of incubation at 28°C. Pr1 subtilisin protease activity was assayed using the specific substrate Suc-(Ala)2-Pro-Phe-p-nitroanilide (NA) (Sigma, USA). One unit of activity was defined as the amount of enzyme that released 1 μmol of NA per minute at 28°C. The activity of chitinase was assayed using the chitinase assay test kit (Solarbio Life Sciences, Beijing, China). One unit of chitinase activity was defined as the amount of enzyme that released 1 μg of N-acetylglucosamine per minute at 37°C. Lipase activity assay was conducted as described previously (34). 4-Nitrophenyl palmitate was used as the substrate to measure the activity of lipases. One unit of enzyme activity was defined as the amount of enzyme to produce 1 μmol of p-nitrosophenol per min at 37°C.

Extraction of promoter sequences from M. robertsii genome.

The promoter sequence of a gene was extracted based on the reference genome of M. robertsii using an in-house Perl script. The promoter of a gene was determined as the 2-kb DNA fragment upstream of its open reading frame (ORF) start site or as the region, if shorter than 2 kb, between its ORF and the ORF of its adjacent gene.

Data availability.

RNA-Seq data from the ΔRns1 mutant (accession number PRJNA720174 [SRR14339752, SRR14339744, and SRR14657472]) and the WT strain (PRJNA637940 [SRR11946811, SRR11947115, SRR11947195]) were deposited in the GenBank database.