Carot-4-en-9,10-Diol, a Conidiation-Inducing Sesquiterpene Diol Produced by Trichoderma virens PS1-7 upon Exposure to Chemical Stress from Highly Active Iron Chelators

Abstract

To screen biocontrol agents against Burkholderia plantarii, the causative agent of rice seedling blight, we employed catechol, an analog of the virulence factor tropolone, to obtain chemical stress-resistant microorganisms. The fungal isolate PS1-7, identified as a strain of Trichoderma virens, showed the highest resistance to catechol (20 mM) and exhibited efficacy as a biocontrol agent for rice seedling blight. During investigation of metabolic traits of T. virens PS1-7 exposed to catechol, we found a secondary metabolite that was released extracellularly and uniquely accumulated in the culture. The compound induced by chemical stress due to catechol was subsequently isolated and identified as a sesquiterpene diol, carot-4-en-9,10-diol, based on spectroscopic analyses. T. virens PS1-7 produced carot-4-en-9,10-diol as a metabolic response to tropolone at concentrations from 0.05 to 0.2 mM, and the response was enhanced in a dose-dependent manner, similar to its response to catechol at concentrations from 0.1 to 1 mM. Some iron chelators, such as pyrogallol, gallic acid, salicylic acid, and citric acid, at 0.5 mM also showed activation of T. virens PS1-7 production of carot-4-en-9,10-diol. This sesquiterpene diol, formed in response to chemical stress, promoted conidiation of T. virens PS1-7, suggesting that it is involved in an autoregulatory signaling system. In a bioassay of the metabolic and morphological responses of T. virens PS1-7, conidiation in hyphae grown on potato dextrose agar (PDA) plates was either promoted or induced by carot-4-en-9,10-diol. Carot-4-en-9,10-diol can thus be regarded as an autoregulatory signal in T. virens, and our findings demonstrate that intrinsic intracellular signaling regulates conidiation of T. virens.

INTRODUCTION

In 1985, it was first reported that an Omigawa isolate of Pseudomonas sp. caused rice seedling blight in nursery boxes in Chiba Prefecture, Japan (1). This isolate, initially named Pseudomonas plantarii, was reclassified into the genus Burkholderia in 1994 (2, 3), and tropolone produced by Burkholderia plantarii was characterized as the virulence factor responsible for rice seedling blight (2).

Tropolone-type compounds possess a unique seven-member aromatic ring system with 1-keto-2-hydroxy and other substitutions; several natural products, such as stipitatic acid, colchicine, and hinokitiol, contain this unique moiety (4, 5). As a nonbenzenoid aromatic compound, purified tropolone possesses unique properties as a phenol and a highly active iron chelator (6, 7). Before discovery of its association with symptoms of B. plantarii-caused rice seedling blight, tropolone was defined as a potent antibiotic toward various bacteria and fungi (8, 9). Afterwards, it was found that tropolone exhibited cytotoxicity on plants—e.g., inhibition of ethylene production of excised peach seeds (10), as well as inhibition of some enzymes, such as grape polyphenol oxidase (11), mushroom tyrosinase (12, 13), and metalloproteases of Tyrophagus putrescentiae and Dermatophagoides farinae (14). In addition, this compound and its derivatives reportedly inhibit growth of human and murine cell lines (5) and of methicillin-resistant Staphylococcus aureus (15). These reports indicate that the virulence of tropolone in rice seedlings can be attributed to its potent cationic metal-chelating effect (15, 16).

As far as we know, a practical manner of controlling B. plantarii-caused rice seedling blight using sterilization of rice seeds with chemical pesticides is sometimes ineffective and often environmentally unfriendly (17). B. plantarii infects rice either preemergence or postemergence as a seed-borne pathogen, and chemical bactericides are rarely effective for controlling this disease (18). Furthermore, almost all rice seedlings in Japan are grown for machine transplanting in nursery boxes under well-controlled environments with relatively high temperature and humidity (2), which aggravates emergence of this disease and makes its control difficult. It is thus desirable to identify biocontrol agents that will be effective in preventing this disease. Tropolone is potently toxic toward a wide spectrum of bacteria and fungi (9, 10). Thus, microorganisms suitable for biocontrol of tropolone-responsive rice seedling blight caused by B. plantarii should be capable of survival and growth in the presence of tropolone.

To obtain candidate microorganisms for practical biocontrol of rice seedling blight, we first screened microorganisms from the rhizosphere of paddy rice. Under selection for microorganisms resistant to chemical stress from catechol, which has iron-chelating properties, a fungus resistant to high levels of catechol was selected and its efficacy in biocontrol was tested. During further investigation of its metabolic traits, we found that its production of a sesquiterpene was uniquely enhanced upon exposure to either catechol or tropolone at appropriate concentrations. In this article, we describe isolation and identification of the catechol-resistant fungus and of the sesquiterpene produced by this fungus and further focus on its role as an autoinducer of morphodifferentiation induction in this fungus.

MATERIALS AND METHODS

Sampling sites and preparation of soil samples.

Soil samples (5 to 10 g each) were collected from the rice rhizosphere in triplicate at six paddy field sites (total of 18 samples) in Hokkaido, Japan, after the harvest period in late October 2010. A small portion of each soil (10 mg, two replicates for each soil sample) was suspended in a sterilized 18-cm test tube containing 10 ml of water sterilized on a Milli-Q Advantage-A10 system (Millipore, MA) and then vortexed for 1 min until the soil was evenly suspended. The soil suspension was left to stand for 15 min, after which 50 μl of the supernatant was used as an inoculant.

Incubation and isolation of culturable rhizosphere microorganisms.

Potato dextrose agar (PDA: 1× potato dextrose broth [PDB], pH 6.2, solidified with 1.5% powdered agar) was used as a culture medium. The inoculant was spread evenly onto PDA plates with a glass spreader and incubated for 14 days at 25°C in the dark. All distinguishable bacterial colonies were isolated and were purified by streaking on fresh PDA plates. Fungi were isolated by cutting a hyphal plug from the margin of the mycelium and subculturing it on fresh PDA plates. These isolated microbes were used for screening chemical stress-resistant biocontrol agents.

Screening of microorganisms resistant to iron chelators.

Potent resistance to tropolone, which is cytotoxic due to its iron-chelating properties, was set as the screening criterion for identification of candidate biocontrol agents for rice seedling blight. To screen for iron chelator-resistant microorganisms, we employed a simple analog of tropolone, catechol, which has relatively low cytotoxicity for bacteria and fungi (19–21). Catechol (Tokyo Chemical Industry, Tokyo, Japan) was added as a supplement to PDA at 0.2, 1, 5, 10, 15, and 20 mM. The initial screening was done on PDA plates containing 0.2 mM catechol, from which isolates that obviously grew after a 14-day incubation were selected and further screened on PDA plates containing a higher concentration of catechol (1, 5, 10, 15, and 20 mM). Catechol-resistant strains were preserved in 10% glycerol solution at −80°C. PS1-7, the fungal isolate most resistant to catechol (20 mM), was selected for identification and bioassay.

Identification of fungus PS1-7.

PS1-7 was identified by two methods: by the morphology of its mycelia and conidia and through the sequences of its internal transcribed spacer (ITS) and 5.8S rRNA gene. Characteristics of hyphae, phialide arrangement, and conidia were observed under an Olympus IX70 light microscope (Olympus, Tokyo, Japan) at magnifications of 60 to 300×.

To obtain fresh mycelia for extraction of genomic DNA, PS1-7 was shake-cultured in 50 ml of PDB medium for 2 days at 110 rpm at 25°C in the dark. Mycelia collected from 1.5 ml of the culture by a brief centrifugation were washed with Milli-Q water several times and then transferred to 2-ml Eppendorf tubes filled with zirconia beads and then frozen in liquid nitrogen. The mycelia in the tube were disrupted in a Multi-beads shocker (Yasui Kikai Co., Osaka, Japan) at 3,000 rpm for 2 min. DNA was extracted from the disrupted mycelia using an Isoplant II DNA kit (Nippon Gene, Toyama, Japan). Using the resulting DNA as the template, the ITS region was amplified by PCR using a pair of universal primers (forward ITS 1, 5′-TCCGTAGGTGAACCTGCGG-3′; and reverse ITS 4, 5′-TCCTCCGCTTATTGATATGC-3′) as reported before (22). The PCR amplicon (600 bp) was sequenced using a BigDye Terminator v 3.1 cycle sequencing kit (Applied Biosystems, Tokyo, Japan) with primer ITS 4 according to the protocol recommended for an ABI Prism 310 genetic analyzer (Applied Biosystems, CA).

Biocontrol assay.

Burkholderia plantarii, originally isolated from a rice seedling infested with blight, was a kind gift from Yuichi Takikawa (Faculty of Agriculture, Shizuoka University) and Kumiai Chemical Industry Co., Ltd. (Tokyo, Japan). We confirmed its production of tropolone and the sequence of its 16S rRNA gene. The culture was preserved in 10% glycerol solution at −80°C and routinely grown on PDA plates. To test the biocontrol efficacy of Trichoderma virens PS1-7 on rice seedlings infested with B. plantarii, healthy rice seeds (Oryza sativa cv. Koshihikari) were placed in 1% NaCl solution and sterilized with 70% ethanol for 2 min and then surface sterilized with 2% NaClO for 30 min and thoroughly washed with sterile distilled water. The surface-sterilized seeds were inoculated with B. plantarii by soaking them in a petri dish containing 10 ml of bacterial cell suspension (10⁶ CFU ml⁻¹). The seeds were simultaneously inoculated with 100 μl of a conidial suspension of T. virens PS1-7 (10⁶ conidia ml⁻¹). Surface-sterilized rice seeds incubated with B. plantarii only (control), PS1-7 only (blank 1), or neither B. plantarii nor PS1-7 (blank 2) were also prepared.

All seeds were incubated at 25°C for 2 days until germination. We selected seeds at an early stage of germination and transplanted them into a 6-cm-high glass dish (40 seeds per dish) containing 25 ml of Hoagland's no. 2 solution solidified with 0.3% gellan gum (Wako, Osaka, Japan). After a 5-day incubation in a plant growth incubator (25°C, 12-h photoperiod), we measured the lengths of the stem and root as parameters of growth performance of the seedlings in each dish to assess the efficacy of the biocontrol treatment.

Semiquantitation and precise quantification of a secondary metabolite produced by T. virens PS1-7 upon exposure to catechol.

To monitor the metabolic traits of T. virens PS1-7 with exposure to 0.5 mM catechol, we inoculated 50 μl of PS1-7 conidial suspension (10⁶ conidia ml⁻¹) into 5 ml of PDB containing 0.5 mM catechol in an 18-cm test tube. For a series of experiments, we prepared 18 or more of the culture tubes, and the tubes inoculated with T. virens PS1-7 were shake-cultured at 110 rpm at 25°C in the dark. Three culture tubes were harvested at 24, 36, and 48 h or later in some experiments. The entire culture medium harvested was transferred into a 15-ml Falcon tube and centrifuged at 10,000 × g for 5 min. The resulting supernatant (4 ml) was pipetted to an 18-cm test tube, and the pH was adjusted to 3.5 to 4.0 with HCl, at which point 1.5 ml of ethyl acetate (EtAOc) was added and the mixture was vigorously vortexed for 1 min. For semiquantitation, a portion of the organic layer (5 μl) was applied using a volumetric glass capillary tube to a Kieselgel 60 GF₂₅₄ silica gel thin-layer chromatography (TLC) plate (0.25 mm; Merck, Darmstadt, Germany) and developed in EtOAc-hexane (3:2 [vol/vol]). The TLC plate was sprayed with vanillin-H₂SO₄ reagent for detection.

For precise quantification of this chemical stress-responsive metabolite, a DB-1 capillary column (30 m by 0.25 mm; J&W Scientific, Folsom, CA) was installed on a GC-2025 gas chromatograph (Shimadzu, Kyoto, Japan) equipped with a flame ionization detector and using ultrapure helium (99.999%) as the carrier gas. The oven temperature was initially held at 120°C for 5 min and raised to 270°C at a rate of 10°C min⁻¹. The temperatures of the injector and detector were set to 250°C and 280°C, respectively. The chemical stress-responsive metabolite was diluted in methanol containing 100 μM m-tert-butylphenol as the internal standard into a series of solutions of 1, 10, 50, 100, and 1,000 μM, from each of which 2 μl was injected into the gas chromatograph. Peaks of the chemical stress-responsive metabolite and m-tert-butylphenol were detected at t_R (times of retention) of 15.6 and 7.8 min, respectively. The peak intensity of the chemical stress-responsive metabolite relative to that of the internal standard showed a linear relationship of y = 50.0x + 26.7 (r² = 0.989), where y is the concentration of the metabolite (in μM) in culture fluid and x is the peak intensity ratio (ratio of the metabolite to internal standard).

Isolation and identification of the chemical stress-responsive metabolite.

For isolation of the chemical stress-responsive metabolite, large-scale culturing of T. virens PS1-7 was done in 1,500 ml of PDB containing 0.5 mM catechol. The PS1-7-inoculated medium was shake-cultured at 110 rpm at 25°C in the dark. After a 3-day incubation, the culture medium was centrifuged at 10,000 × g for 10 min and then filtered through no. 101 filter paper (Advantec, Tokyo, Japan). The culture filtrate was adjusted to pH 3.5and then extracted exhaustively with EtOAc (500 ml × 3). The organic layer was combined and dried over anhydrous Na₂SO₄ and then concentrated under low pressure. The concentrates thus obtained (547 mg) were resuspended in hexane-EtOAc (95:5 [vol/vol]) and then separated by chromatography on a GF₆₀ silica gel column (50 g, 35-to-70 mesh; Merck) with stepwise elution at 5 to 100% EtOAc in n-hexane. Fractions containing the chemical stress-responsive metabolite (those eluted with 25% and 30% EtOAc in n-hexane) were combined (113 mg), purified by preparative TLC, and recrystallized in chloroform to afford 47.3 mg of colorless needles.

Field desorption mass spectroscopy (FD-MS) and electron ionization MS (EI-MS) were conducted with a JEOL JMS-T100GCV and a JMS-SX-102, respectively. One-dimensional nuclear magnetic resonance (1D-NMR) and 2D-NMR were conducted using a JEOL JNM-EX270 and a Bruker AM 500, respectively. The spectroscopic data of the isolated compound are as follows: FD-MS and FD-HR-MS, [M]⁺ at m/z 238.1936 (C₁₅H₂₆O₂, calculated 238.1932) (see Fig. S2 in the supplemental material). EI-MS at m/z (relative intensity [%]), 238 (13%, [M]⁺), 220 (11%, [M-H₂O]⁺), 202 (12%, [M-2H₂O]⁺), 195 (98%, [M-Me₂CH]⁺), 177 (100%), 159 (69%), 123 (40%), 107 (42%), 93 (42%), and 43 (88%). The compound was acetylated with Ac₂O-pyridine at 60°C and yielded a monoacetylated compound as a colorless syrup (24 mg, 68% yield). ¹H NMR (nondecoupling [NON], H-H correlation spectroscopy [COSY], and nuclear Overhauser effect spectroscopy [NOESY]) and ¹³C-NMR (bilevel complete decoupling [BCM], distortionless enhancement by polarization transfer [DEPT], heteronuclear multiple-quantum correlation [HMQC], and heteronuclear multiple-bond correlation [HMBC]) spectra of the monoacetylated compound were taken in CDCl₃ to confirm its chemical structure, including relative configuration.

Carot-4-en-9,10-diol production in T. virens PS1-7 exposed to catechol, tropolone, and other antifungal iron chelators.

In a time course experiment, shake cultures of T. virens PS1-7 conidia in PDB containing 0.5 mM catechol were sampled at 0, 12, 24, 48, 60, 72, 84, 96, 108, 120, 132, and 144 h after inoculation (2 ml each) and analyzed quantitatively by capillary gas chromatography, as described above. T. virens PS1-7 was inoculated into PDB containing different concentrations of catechol (0.1, 0.5, 1, 2, or 5 mM) and cultured under the same incubation conditions to investigate the effect on carot-4-en-9,10-diol production. Samples of the culture medium (2 ml) were taken at 24, 48, 72, 96, 120, and 144 h for quantitative analyses.

Production of carot-4-en-9,10-diol by T. virens PS1-7 was examined in the same manner. The metabolic effects of other iron chelators, including pyrogallol, gallic acid, citric acid, and salicylic acid, were also examined in T. virens PS1-7. These chelators were tested at 0.5 mM; in addition, EDTA was tested, but at 0.2 mM, because it was toxic at 0.5 mM, as shown by prevention of hyphal growth. Cinnamic acid, which has neither antifungal nor iron-chelating properties, was used at 0.5 mM as a negative control.

Cleanup of culture medium for analysis of carot-4-en-9,10-diol by gas chromatography.

To analyze the carot-4-en-9,10-diol in the culture fluid, the culture medium (2 ml) taken at each sampling was centrifuged at 10,000 × g for 5 min and then the resulting supernatant (1.5 ml) underwent solid-phase extraction (SPE) using a Sep-Pak C₁₈ (3-ml Vac cartridge) containing 200 mg resin (Waters, MA). Methanol (3 ml) and water (3 ml) were added successively in a vacuum chamber to a cartridge preconditioned with acetone (3 ml). The supernatant was then gently loaded and passed through the column under low pressure. The column was then washed with water (1 ml) to remove the void culture fluid in the column and then eluted with methanol (2 ml). The methanolic eluates thus obtained were concentrated and redissolved in 150 μl methanol containing 100 μM m-tert-butylphenol as an internal standard; 2 μl of this mixture was injected into a gas chromatograph for quantification of carot-4-en-9,10-diol.

Tropolone dynamics in cultures of T. virens PS1-7.

We analyzed tropolone dynamics in medium inoculated with PS1-7 to investigate the capacity of PS1-7 to degrade tropolone. Authentic tropolone (Wako) was dissolved in sterilized Milli-Q water to a 10 mM stock solution, from which 100 μl was added to 5 ml of PDB in an 18-cm test tube. PDB containing 0.2 mM tropolone was inoculated with 50 μl of a PS1-7 conidial suspension (10⁶ conidia ml⁻¹); water was used in place of tropolone as the control. The culture was shaken at 110 rpm at 25°C in the dark and sampled at 0, 12, 24, 48, 72, 96, and 120 h. To analyze tropolone dynamics in the culture fluid, the culture (2 ml) was centrifuged at 10,000 × g for 5 min and the resulting supernatant (1.5 ml) was subjected to solid-phase extraction, as mentioned above. The methanolic elutes from the cartridge were concentrated and redissolved in a volumetric 150-μl methanol, from which 10 μl was injected into a high-performance liquid chromatograph (HPLC) for quantitative analysis of remaining tropolone.

For the HPLC system, an L-column2 ODS column (250 mm by 4.6 mm; inside diameter [i.d.], 5 μm) was installed on a Waters 600 HPLC system (Waters, MA) equipped with a photodiode array detector (wavelength, 270 nm) and using 5% water–CH₃CN containing 1 mM EDTA·2Na as the mobile phase. Standard solutions of tropolone (0.01, 0.1, 0.2, 2, and 20 mM) were prepared as a serial dilution series of the stock solution; 10 μl of each concentration was separated by HPLC, and then a standard curve was obtained, which fit the equation y = 0.0004x + 0.0289 (R² = 0.998), where y is the concentration of tropolone (in mM) and x is the absolute peak intensity of tropolone.

Autoregulatory function of carot-4-en-9,10-diol.

The autoregulatory function (23) of carot-4-en-9,10-diol on T. virens PS1-7 mycelium was tested by two experiments. For macroscopic observation, a 5-mm-diameter PSF-7 mycelial plug placed on a PDA plate was allowed to develop hyphae on a fresh PDA plate (at full strength, 1/4×, or 1/10× PDB) containing 10 μM carot-4-en-9,10-diol; controls were without carot-4-en-9,10-diol. After a 4-day incubation, formation of green concentric circles of conidiophores was recognized.

For microscopic observation, 10 ml of PDA (containing 1/10× PDB) was impregnated with 100 μl of a conidial suspension (10⁶ conidia ml⁻¹). After a 24-h incubation, an 8-mm-diameter, thick paper disc loaded with 50 μl of 10 μM carot-4-en-9,10-diol solution (in acetone) was placed on the resulting plates containing uniformly distributed PS1-7 hyphae. Acetone (50 μl) was used as the control. After a 4-day incubation, conidiation was observed along the PS1-7 hyphae around the paper disc under a light microscope.

Nucleotide sequence accession number.

The DNA sequence of isolate PS1-7 has been deposited in the DNA Data Bank of Japan (DDBJ) under accession no. AB744653.

RESULTS

Catechol-resistant microorganisms isolated from the paddy rice rhizosphere.

We obtained a total of 186 culturable and morphologically distinguishable microbial isolates from rhizosphere soil samples collected from six paddy fields (Table 1). Upon exposure to 0.2 mM catechol and later stepwise increases in concentration up to 20 mM, the number of surviving microbial isolates decreased (Fig. 1). Upon exposure to catechol beyond 10 mM, the number of resistant microbial isolates decreased to 12 bacterial isolates and 4 fungal isolates. Considering that the majority of the microbial isolates were capable of surviving exposure to less than 5 mM catechol, 10 mM catechol was likely above the threshold level of significant toxicity. As a result, 16 microbial isolates (Fig. 1) that survived at 10 mM were stored for further investigation. The most catechol-resistant fungal isolate, PS1-7, the only one that survived exposure to 20 mM catechol, was investigated for its biocontrol efficacy against rice seedling blight and for relevant metabolic traits upon exposure to chemical stress mediated by iron chelation.

Table 1.

Sampling sites and catechol-tolerant microbial isolates

Paddy field (latitude, longitude)	Isolate(s)a
	Fungal	Bacterial
Shimamatsu-Kitahiroshima Kyoei (43°00.147′N, 141°34.434′E)	PS1-7	PS1-1, PS2-3, PS2-8
Eniwa Kitajima (42°58.314′N, 141°36.647′E)		PK1-3, PK2-1
Chitose Komasato (42°56.627′N, 141°38.400′E)		PC1-3, PC1-25, PC2-15
Atsuma Kamiatsuma (42°38.692′N, 141°50.674′E)		PAK1-2, PAK2-13
Atsuma Shinmachi (42°43.126′N, 141°52.704′E)	PAS1-1, PAS2-11
Hayakita Hokusin (42°46.558′N, 141°49.961′E)	PH1-6	PH1-12, PH2-11

^aBoth the fungal and bacterial isolates listed are tolerant to 10 mM catechol.

Fig 1.

Microbial isolates tolerant to iron chelator selected on PDA plate containing catechol. The value on top of each column is the number of microbes that survived and developed a colony on the selection plate.

Hyphae of PS1-7 developed a flat white mycelium on PDA plates that, after a 4-day incubation at 25°C, formed light green conidia in a concentric circle within the mycelium. Conidiophores were hyaline and smooth walled, and the phialide was lageniform or ampulliform, with a length of 8 to 10 μm and a base of 2 to 3 μm. Conidia were ovate or obovate with a diameter of 3 to 4 μm, with a rough surface. The conidial zone turned dark green with an amber background on PDA plates after a 14-day incubation. These morphological characteristics agreed with those of Trichoderma sp., particularly T. virens. By combining these observations with the sequence of an amplicon derived from its ITS (accession no. AB744653), this fungal isolate, PS1-7, was tentatively identified as a strain of T. virens.

We observed suppression of rice seedling blight symptom development in rice seedlings infested with B. plantarii when infested seedlings were inoculated at germination with a conidial suspension of T. virens PS1-7. In control plants not inoculated with T. virens PS1-7 conidia, typical symptoms of the disease, such as chlorosis, stunting, and wilting, were obvious. In contrast, the symptoms of plants inoculated with the conidial suspension were obviously remedied and were followed by significantly improved growth (Fig. 2A and B). No difference in the growth of seedlings inoculated only with PS1-7 (Fig. 2C, blank 1) or those inoculated with neither B. plantarii nor PS1-7 (Fig. 2D, blank 2) was observed (Fig. 2C and D).

Fig 2.

Biocontrol efficacy of T. virens PS1-7 on growth of rice seedlings inoculated with B. plantarii. Shown are comparisons of stem and root growth among control rice seedlings inoculated with B. plantarii (A), treated rice seedlings inoculated with both B. plantarii and T. virens PS1-7 (B), rice seedlings inoculated with T. virens PS1-7 only (blank 1) (C), and rice seedlings without any inoculation (blank 2) (D). Rice seedlings grown for 5 days at 25°C under a 12-h photoperiod (A to D) were harvested, the lengths aboveground (stem growth) and of the root (root growth) of each seedling were recorded (right panel). Values are means ± standard deviations (SD [shown by error bars]) (n = 40). *, P < 0.001 by Student's t test.

Identification and quantification of carot-4-en-9,10-diol as a chemical stress-responsive metabolite of T. virens PS1-7.

A distinctive metabolite was detected as a chemical stress-responsive metabolite in the culture medium of T. virens PS1-7 grown in the presence of 0.5 mM catechol (see Fig. S1 in the supplemental material). Its chemical structure and relative configuration (rel.) were elucidated to be (rel. 1S,7R,9R,10S)-carot-4-en-9,10-diol (see Table S1 and Fig. S2 to S4 in the supplemental material), which is identical to those of a carotane-type sesquiterpene, CAF-603, isolated from another strain of Trichoderma virens (24).

Upon exposure to 0.5 mM catechol, carot-4-en-9,10-diol production in PS1-7 was maintained from 12 h to 72 h and reached a maximum of 45 μM at 72 h, which was approximately 2.5-fold the level in the control (Fig. 3A and B). However, a culturing time longer than 72 h led to a loss of carot-4-en-9,10-diol production in mycelial cultures of T. virens PS1-7. Upon exposure to 0.1, 0.5, and 1 mM catechol, carot-4-en-9,10-diol production in T. virens PS1-7 was maintained up to 72 h, and its maximum level was enhanced in a dose-dependent manner by catechol (Fig. 3A). Upon exposure to tropolone, the metabolic response of T. virens PS1-7 was a dose-dependent enhancement of carot-4-en-9,10-diol production (Fig. 3C), similar to what was observed when treated with catechol.

Fig 3.

Time course of carot-4-en-9,10-diol production and quenching in culture media of T. virens PS1-7. (A) Carot-4-en-9,10-diol was analyzed quantitatively for culture medium inoculated with T. virens PS1-7 containing catechol at 0, 0.1, 0.5, and 1 mM in triplicate at 24, 48, and 72 h. (B) Carot-4-en-9,10-diol production was quantified from cultures of T. virens PS1-7 in PDB containing 0.5 mM catechol (▲) and in PDB without catechol (●). (C) Carot-4-en-9,10-diol was analyzed quantitatively for culture medium inoculated with T. virens PS1-7 containing tropolone at 0, 0.05, 0.1, and 0.2 mM in triplicate at 24, 48, and 72 h. (D) Dynamics of supplemented tropolone (initial concentration 0.2 mM) in PDB inoculated with T. virens PS1-7 (▲) or left uninoculated (●). Values are means ± SD (shown by error bars) (n = 3).

The similar response of T. virens PS1-7 to chemical stress from catechol or tropolone (Fig. 3C) seemed to be attributable to their iron-chelating properties. As for catechol (see Fig. S1 in the supplemental material), degradation of tropolone in T. virens PS1-7-cultured PDB was also observed by HPLC analysis (Fig. 3D). This unique capacity of T. virens PS1-7 toward tropolone seems to be highly related to the biocontrol efficacy of this fungus.

Induction of conidiation in PS1-7 with carot-4-en-9,10-diol.

Enhanced conidiation and maturation of conidia by supplemented catechol were clearly shown as increased numbers of conidial rings in T. virens PS1-7 mycelia grown on PDA plates (Fig. 4). Conidiation of PS1-7 mycelia grown on nutrient-poor PDA plates (1/4× or 1/10× PDB), which is inhibited under these nutrient conditions (see Fig. S5 in the supplemental material), was restored by supplementation with 10 μM carot-4-en-9,10-diol (Fig. 5). In addition, formation of conidiophores and maturation of conidia were accelerated by supplementation with 10 μM carot-4-en-9,10-diol. Based on microscopic observation, conidiophores and conidia clearly formed in T. virens PS1-7 hyphae growing near paper discs loaded with 11.9 μg carot-4-en-9,10-diol under nutrient-poor conditions (e.g., 1/10× PDA). Differentiation from hyphae to conidiophores was inhibited in T. virens PS1-7 hyphae in the control area (see Fig. S6 in the supplemental material).

Fig 4.

Conidiation of T. virens PS1-7 mycelia grown on PDA plates with exposure to catechol. Incubation was at 25°C for 5 days for all plates tested. Enhanced conidiation and maturation of conidia were observed as the morphological response of T. virens PS1-7 to exposure to 0.2 mM (center) and 1.0 mM (right) catechol added to a 1× PDA plate compared with the response of the control (without catechol addition).

Fig 5.

Effect of carot-4-en-9,10-diol on conidiation of T. virens PS1-7 mycelia on PDA plates. (A, B, and C) PDA plates that contain 10 mM carot-4-en-9,10-diol; (D, E, and F) PDA plates without carot-4-en-9,10-diol. Conidiation of T. virens PS1-7 mycelia grown on PDA plates supplemented with 10 μM carot-4-en-9,10-diol (A, B, and C) was compared with conidiation on unsupplemented plates (D, E, and F). Plates A and D contained 10× diluted PD broth solidified with 1.5% agar, while plates B and E contained the same mixture but 4× diluted. Red arrowheads indicate conidial rings on the mycelia. Incubation was at 25°C for 5 days.

DISCUSSION

The rhizosphere, a characteristic habitat for microorganisms, is a natural source of functional microorganisms that are highly associated with host plant defense or are competitive or antagonistic against soilborne phytopathogenic microorganisms (25–27). Due to long-term survival in the rhizosphere microenvironment, the microbial community that is dominant and cooperative in the host root system within the complex food web (28) is regarded as an appropriate source of biopesticidal agents effective in host protection. T. virens PS1-7 was isolated from the rice rhizosphere as a catechol-resistant fungus and showed great promise as a biocontrol agent in controlling rice seedling blight.

During investigation of the metabolic traits of T. virens PS1-7 exhibited in response to antifungal iron chelators in its culture medium, we found that this fungus can degrade the virulence factor tropolone as well as it can catechol (Fig. 3D). Some strains of the genera Trichoderma and Penicillium resistant to tropolone and catechol have shown potent degrading capacity toward tropolone and catechol (29, 30). Decomposition of xenobiotics by the genus Trichoderma was first reported in Trichoderma viride by Baarschers et al. in 1986 (31). A large portion of toxic xenobiotics, such as endosulfan (32), cyanide pollutants (33), dichlorvos (34), and even heavy metals (35), are removed from polluted environments by the genus Trichoderma. The potent capacity of the genus to degrade various xenobiotics is due to both extracellular and intracellular metabolic pathways catalyzed by both specific and nonspecific enzymatic systems. This cellular detoxification sustains the survival and growth of this fungus (36, 37).

In addition, we found that T. virens PS1-7 did not show the ability to degrade either catechol or tropolone under carbon-poor conditions (unpublished data). Hence, it seems that T. virens PS1-7 metabolizes catechol and tropolone to detoxify them to facilitate its survival and growth rather than to utilize such chemicals as a carbon source. Whether such a tropolone-degrading ability is necessary for a biocontrol agent to control rice seedling blight is unclear. The use of T. virens in biocontrol of pathogenic fungal disease is mainly based on its antagonism against pathogens via extracellular enzymes causative of mycoparasitism, secretion of antagonistic secondary metabolites, and cooperation with the host plant by induction of systemic resistance and promotion of growth (38, 39). T. virens PS1-7 did not significantly affect the growth of rice seedlings (Fig. 2C) in our experiments. Hence, the mechanism of efficacy of PS1-7 in biocontrol of rice seedling blight is either antagonism against B. plantarii or degradation of tropolone.

Furthermore, we highlighted enhanced carot-4-en-9,10-diol production as a characteristic metabolic response of T. virens PS1-7 to catechol and tropolone; we regard this sesquiterpene as a metabolite responsive to chemical stress. Upregulation of sesquiterpene biosynthesis has been reported for many sesquiterpene phytoalexins that are generally induced by chemical mediators, such as jasmonic acid, methyl jasmonate, and salicylic acid (40–42). We found that production of carot-4-en-9,10-diol in T. virens PS1-7 was enhanced by an array of compounds acting as iron chelators (Fig. 6; see Fig. S7 in the supplemental material), whereas it showed no response to cinnamic acid, which possesses neither antifungal nor iron-chelating properties. This on one hand indicates that carot-4-en-9,10-diol production in T. virens PS1-7 is an active metabolic response to chemical stress from an antifungal iron chelator and on the other hand indicates that T. virens PS1-7 can sense different antimicrobial iron chelators through a universal sensing system. G protein-coupled receptors and/or nitrogen-sensing receptors spanning the cytoplasmic membrane of hyphal cells of Trichoderma sp. specifically bind extracellular chemical signals and stimuli, which is followed by transmission of the chemical signal to intracellular signaling cascades (36). Hence, T. virens PS1-7 may sense structurally diverse iron chelators by a G protein-coupled receptor, leading to activation of carot-4-en-9,10-diol biosynthesis.

Fig 6.

Production of carot-4-en-9,10-diol in T. virens PS1-7 upon exposure to several types of iron chelators. Production of carot-4-en-9,10-diol in the presence of 0.5 mM an iron chelator was compared with that in the control after a 72-h incubation, except for with EDTA-Na⁺ at used at a concentration of 0.2 mM because the fungal cells did not grow well in its presence at 0.5 mM. (E)-Cinnamic acid was used as a negative control at 0.5 mM. Values are means ± SD (shown by error bars) (n = 6). *, P < 0.0001 by Student's t test.

T. virens PS1-7 also showed a morphological response to catechol, namely, promotion of conidiation and accelerated maturation of the conidia, recognizable by their dark green pigmentation (Fig. 4). Conidiation leads to production and dispersal of propagules during an asexual stage in the life cycle of imperfect fungi (see Fig. S6 in the supplemental material), which are relatively resistant to adverse environments (43), and hence, the promotion of conidiation in response to compounds like catechol likely facilitates the survival of PS1-7. There seems to be a link between the metabolic and morphological responses, as shown by autoregulatory signals acting in concert with environmental cues in regulating a series of morphological events in filamentous fungi (21, 44). The discovery of this chemical stress response sheds light on the link between carot-4-en-9,10-diol production and conidiation (Fig. 4 and 5).

Conidiation, as an outcome of normal cellular differentiation in fungi, is also repressed under nutrient-poor conditions due to the reduction of cell competence to form conidia (45) (see Fig. S5A in the supplemental material). Similarly, radial growth of T. virens PS1-7 mycelium grown on nutrient-poor PDA plates was unaffected (see Fig. S5B), but cellular differentiation recognizable as formation of vertical hyphae and subsequent conidiophore formation was drastically inhibited (Fig. 5; see Fig. S5B). This inhibition of conidiation was restored by supplementation with 10 μM carot-4-en-9,10-diol, and under nutrient-rich conditions, carot-4-en-9,10-diol accelerated conidiophore formation and conidial maturation of T. virens PS1-7. This finding indicates that carot-4-en-9,10-diol is an autoregulatory signal molecule. Similar to our findings, sporogen AO1, an eremophilane-type sesquiterpene produced in Aspergillus oryzae and a Penicillium sp., also has an autoregulatory activity capable of exerting a sporogenic effect (46, 47).

Although the mechanism of autoregulatory signal regulation of this morphological event is not well understood (48), our finding that exogenously added carot-4-en-9,10-diol induces conidiation in T. virens PS1-7 is consistent with a hypothesis proposed by Nemčovič et al. for conidiation of Trichoderma atroviride (49). In their estimation, specific receptors on the plasma membrane for C₈ volatile organic compounds that induce conidiation in T. atroviride transduce the signal from these compounds into conidiation regulation via the mitogen-activated protein and/or G protein signaling pathways (49, 50). A complex autoregulatory system recently discovered in Aspergillus nidulans indicated the importance of log P values (P is partition coefficient) of the signal compounds (51) and suggested that carot-4-en-9,10-diol of T. virens PS1-7 is thus capable of solely exerting autoregulatory function due to its high log P value (3.152; calculated with ACD/Labs software V11.02 in the SciFinder program). Hence, it is possible that carot-4-en-9,10-diol binds to a specific receptor on the hyphal cell membrane that subsequently upregulates genes in the downstream signaling cascade involved in conidiation of T. virens (42).

In conclusion, tropolone-resistant T. virens PS1-7 isolated from the rice rhizosphere is a candidate biocontrol agent for rice seedling blight caused by B. plantarii. In combination with detoxification of antifungal iron chelators, production of carot-4-en-9,10-diol as a metabolic response that mediates conidiation facilitates the adaptation of T. virens PS1-7 to hostile environments. The autoregulatory signal molecule carot-4-en-9,10-diol induced conidiation in PS1-7 mycelia, revealing that this compound triggers chemical stress-mediated conidiation in T. virens.