• The function of fungal volatiles in fungal–plant interactions is poorly understood. The aim here was to address this lack of knowledge, focusing on truffles, ectomycorrhizal fungi that are highly appreciated for their aroma.
• The effect of volatiles released by truffles was tested on Arabidopsis thaliana in a closed chamber bioassay. The volatiles produced by Tuber melanosporum, Tuber indicum and Tuber borchii fruiting bodies inhibited A. thaliana in terms of root length and cotyledon leaf size, and in some cases induced a bleaching of the seedlings, thus indicating toxicity. Ten synthetic volatiles were tested in a similar way. The strongest inhibitory effect was observed with C8 molecules such as 1-octen-3-ol, an alcohol with a typical ‘fungal smell’.
• Two of these C8 compounds were further tested to investigate their mechanism of action. 1-Octen-3-ol and trans-2-octenal induced an oxidative burst (hydrogen peroxide, H2O2) in the A. thaliana leaves as well as a strong increase in the activities of three reactive oxygen species (ROS)-scavenging enzymes.
• These results demonstrate that fungal volatiles inhibit the development of A. thaliana and modify its oxidative metabolism. Even though limited to laboratory observations, these results indicate the presence of a hitherto unknown function of fungal volatiles as molecules that mediate fungal–plant interactions.
Volatile organic compound (VOC)-mediated interactions among plant–plant, plant–insect and bacteria–plant have been frequently documented (De Moraes et al., 2001; Dicke et al., 2003; Ryu et al., 2003; Kappers et al., 2005; Schnee et al., 2006). Despite the importance of fungi in nature, in terms of biomass (Müller & Loeffler, 1976), very little attention has been paid to the role of their VOCs. It was decided to study the effect on plants of VOCs emitted by truffles, hypogeous ectomycorrhizal fungi that are highly appreciated for their aroma (Mello et al., 2006). More than 100 VOCs have recently been identified in both fruiting bodies and mycelia (Splivallo et al., 2007) and it has been hypothesized that they serve as specific signal molecules to host plants (Menotta et al., 2004). It is also believed they are involved in the formation of a ‘burnt’ area – a zone around the host tree where herbaceous plants are inhibited, which has been observed with some species such as T. melanosporum (Pacioni, 1991).
In order to investigate the potential phytotoxic effect of truffle VOCs, Arabidopsis thaliana was used as a model of a nonhost plant. The experimental design consisted of a chamber in which the seedlings were exposed to a single VOC or to a natural VOC blend from the fruiting bodies of three species. The effects on the seedlings were recorded at the morphological (germination rate, root length, leaf surface area and bleaching) and biochemical (H2O2 and reactive oxygen species scavenging enzymes) levels.
Materials and methods
Truffle fruiting bodies
The fruiting bodies belonging to three species –Tuber melanosporum Vittad, Tuber borchii Vittad. (both species from Piedmont in northern Italy) and Tuber indicum Cooke & Masee Grevillea (from the Yunnan and Sichuan Provinces, China) – were used. In short, 13 pieces of T. melanosporum were collected in 2004, five pieces of T. borchii were collected in 2004 and 10 pieces in 2006 and five pieces of T. indicum in 2004.
Morphological identification of one or two fruiting bodies taken at random for each species was confirmed through PCR (ITS4) according to Murat et al. (2004) (accession number for T. indicum, AM 406672).
The following compounds were purchased from Sigma-Aldrich (Milan, Italy): 3-methyl-1-butanol (= 99%); 1-hexanol (= 99%), trans-2-octenal (= 94%), 2-phenylethanol (= 99%), 1-octen-3-ol (= 98%) and 3-octanol (99%) (both racemic of R and S enantiomers); 3-octanone (= 97%); dimethyl sulfide (> 99%), 3-methylsulfanylpropanol, 2,3-butanediol (= 99%) (racemic mixture of RR and SS isomers). In terms of the TIC peak area of the GC-MS, the first eight compounds accounted for 79 ± 20% (± standard deviation) for T. borchii; 52 ± 15% for T. melanosporum; and 71 ± 17% for T. indicum. The same samples that were analyzed through GC-MS were used for the experiments presented in Figs 1(a)–(d).
Arabidopsis thaliana Heinh. wild type (Col-0) was used for the bioassays. The seeds were surface-sterilized and vernalized overnight before each experiment. The growth chamber conditions were 23°C during the day, 21°C during the night and a 16 h photoperiod under 80 µmol m−2 s−1 light.
To check the effects of volatiles on a host plant, seeds of Cistus incanus L. were tested in a similar way.
Bioassays with truffle fruiting bodies
The truffle sample (1 g) comprised either one whole fresh fruiting body (n = 10 for T. borchii Fig. 1e) or small pieces pooled from frozen fruiting bodies of the same species (n = 5 for T. borchii and T. indicum; n = 13 for T. melanosporum; Figs 1a–d, Supplementary material, Fig. S1).
The sample (unsterilized to avoid aroma alteration) was placed on one side of a compartmented Petri dish (diameter 9.0 cm) and packed in sterile cotton, allowing the truffle VOCs to diffuse inside the Petri dish. A. thaliana or C. incanus seeds were placed on the other side of the Petri dish on filter paper that was wetted with distilled water and were allowed to germinate for 63 h or 12 d. The control consisted of a Petri dish containing no fruiting body (cotton only). The seedlings were mounted on glass slides and photographed for cotyledon leaf surface and root length determination.
Bioassays with synthetic VOCs
Square Petri dishes (12 × 12 cm) were filled with 80 ml MS-agar (plant agar 8.0 g l−1, Murashige & Skoog Basal Salt mixture (MS) 4.4 g l−1 and sucrose 15.0 g l−1, pH 5.8–6.0). Half of the agar was removed with a scalpel by cutting it parallel to the lower side of the Petri dish. Fifteen A. thaliana seeds were placed on the agar parallel to the direction of the cut (just below it), and a piece of sterile cotton was placed onto the right top corner of the Petri dish.
A small hole was made in the plastic with an incandescent iron pin exactly on top of the cotton, and 30 µl VOC was injected into the cotton with a GC syringe, either pure or diluted in cold CH2Cl2 to reach concentrations of 130, 13, 1.3 and 0.13 ppm per volume (per vol., defined as the quantity (µl) of VOC injected into the cotton over the internal volume (l) of the Petri dish) for each single VOC inside the Petri dish. After injection, the hole was immediately closed with a Teflon septum. CH2Cl2 or H2O was used as a control at 130 ppm per vol. The Petri dishes were placed vertically in the growth chamber. The VOCs were either injected before germination (day 0) or just afterwards (day 4). Each Petri dish was scanned after 12 d to quantify the germination, the root length of the germinated seedlings, and the leaf color (quantified visually according to the color scale in Fig. 2b).
In a second set of experiments carried out to determine the leaf surface area, 10 A thaliana seeds were placed on MS-agar on one side of a compartmented Petri dish (diameter 9.0 cm), a small piece of cotton was placed on the other side of the Petri dish, and the VOCs (10 µl) were injected in the same manner as previously described. The seedlings were photographed from above after 12 d to quantify the leaf surface area.
H2O2 quantification and ROS-scavenging enzyme activity
Arabidopsis thaliana seedlings were grown in square Petri dishes (12 × 12 cm) as already described. The VOCs were added 3 wk after germination. The seedlings were collected at 15 min, 2, 24 and 48 h after the VOCs had been added (of 10 germinated seedlings per Petri dish, five were pooled for enzymatic extraction and three were used for H2O2 determination).
H2O2 visualization was performed with an Amplex Red kit (Invitrogen, Molecular Probes, Eugene, OR, USA) by immersing the whole seedlings for 20 min in the solution prepared according to the manufacturer's specifications. The leaves were detached from the seedlings, placed on a microscope slide and immediately visualized/digitalized for the localization of H2O2 using a Leica TCS SP2 confocal microscope, equipped with a long distance ×20 dry objective (HC PL Fluotar 20.0 × 0.5 Dry). The He/Ne laser band of 543 nm (67% intensity) was used to excite the Amplex Red and an emission window was set at 585–610 nm for Amplex Red fluorescence acquisition. Image analysis for H2O2 quantification was performed with ImageJ (http://rsb.info.nih.gov/ij/).
The whole seedlings (20–70 mg FW) were crushed in liquid N2 and extracted at 4°C according to Maffei et al. (2006) for the enzymatic assays. After extraction, the supernatant was directly used for the enzyme assays. The superoxide dismutase (SOD), catalase (CAT) and guaiacol peroxidase (PX) activities were measured according to Maffei et al. (2006). The activities were calculated in µKat mg−1 protein before calculating the ratio (AVOC/Acontrol).
The soluble protein concentration was measured according to Bradford (1976), using bovine serum albumine as a standard.
Data analysis and statistics
At least four replicates were performed for each treatment for the morphological data collection (one replicate = 10 or 15 seeds per Petri dish), three for the enzymatic data (one replicate = five seedlings per Petri dish), and three for H2O2 quantification (one replicate = three seedlings per Petri dish). For the preliminary experiments on C. incanus, seven replicates for each treatment were performed (one replicate = 10 seeds per Petri dish). The data were checked for normality and homogeneity of variance, and analyzed according to the Mann–Whitney test or with anova (Tukey or Dunett post-hoc tests) (SPSS Software).
Effect of the natural aromas of truffles
In the first set of experiments, the effect of the natural aroma of T. melanosporum, T. indicum and T. borchii on A. thaliana seedlings was investigated. By placing 1 g of fruiting body in a compartmented round Petri dish, the VOCs could diffuse inside the Petri dish and interact with the seedlings in the second compartment. The effects on A. thaliana, observed after 12 d, included inhibited root development (Fig. 1a, d), a strong bleaching of the cotyledon leaves (Fig. 1b) and a reduced surface of the cotyledon leaves (T. borchii, Fig. 1c). Similar experiments performed on C. incanus, a host plant, led to comparable results (Fig. S1a, b), and for this reason only A. thaliana was used in the subsequent experiments.
Since truffles are prone to fermentative processes within 12 d, another set of experiments was conducted over a much shorter period during which the truffle aroma remained unaltered (Splivallo et al., 2007). The bioassay was repeated with 10 whole fresh fruiting T. borchii bodies over 63 h. Root inhibition was observed in every single fruiting body (Fig. 1e), confirming that the inhibitory effect is the result of the VOCs released by the fresh truffles.
Effect of synthetic VOCs
Ten VOCs were tested on A. thaliana. These were chosen because 3-methyl-1-butanol, 1-hexanol, 1-octen-3-ol, 3-octanol, 3-octanone, trans-2-octenal, and 2-phenylethanol represented the main VOCs that are emitted by truffles (Splivallo et al., 2007); dimethyl sulfide and 3-methylsulfanylpropanol are characteristic of truffle aromas (Mauriello et al., 2004; Buzzini et al., 2005); and 2,3-butanediol has plant growth promotion properties (Ryu et al., 2003). The quantities of the main VOCs released by the samples are summarized in Table 1.
Table 1. Volatile organic compounds (VOCs) of three truffle species and their inhibition threshold on Arabidopsis thaliana grown on agar
Fruiting bodies (ppm per vol.)
Inhibition threshold of single VOCs on agar (ppm per vol.)
Single VOCs were tested on A. thaliana seeds at four concentrations: 0.13, 1.3, 13 and 130 ppm per volume (vol.) Fruiting bodies, average concentration reached after 63 h by the volatiles of three truffle species in a 9.0-cm-diameter Petri dish; inhibition threshold, lowest concentration at which a single volatile significantly inhibited germination, primary root elongation or cotyledon leaf size compared with water and solvent (CH2Cl2) controls (P < 0.05, anova, post-hoc Tukey or Dunett, or Mann–Whitney tests –n = 4–6 repetitions (Petri dishes) per treatment and per concentration with 10 or 15 seeds per Petri dish).
, germination completely blocked; –, no inhibition at the tested concentrations; BD, below detection; NQ, not quantified.
The different letters indicate statistically different results in terms of the total VOC contents (P = 0.05, Mann–Whitney test).
A first screening, with VOCs added to the Petri dishes before germination, was performed at four concentrations: 130, 13, 1.3 and 0.13 ppm per vol. With the exception of 1-octen-3-ol, which inhibited the cotyledon leaf development at 1.3 ppm per vol. (compared with the H2O and CH2Cl2 controls), the 10 tested VOCs had no significant effect on morphology and germination at 1.3 and 0.13 ppm per vol. (Table 1). Some inhibition was observed for most of the VOCs at 13 and/or 130 ppm per vol. (Table 1) and these two concentrations were therefore chosen to further investigate the action mechanism of those VOCs on A. thaliana.
In order to check whether the reduced root length was the result of retarded germination, a second set of experiments was performed in which VOCs were added to the seedlings 4 d after germination. Bleaching of the cotyledon leaves was observed at 130 ppm per vol. for seven out of the 10 tested compounds (Fig. 2a, b). At 13 ppm per vol., bleaching was only visible for three C8 VOCs (trans-2-octenal, 3-octanol, 1-octen-3-ol), which also inhibited root growth (Fig. 2c).
H2O2 and ROS-scavenging enzymes
Since bleaching is often associated with oxidative stresses (Cao et al., 2006), the induction of ROS and the production of ROS-scavenging enzymes in A. thaliana were investigated for the VOCs that induced the strongest bleaching (trans-2-octenal and 1-octen-3-ol). Three-week-old A. thaliana seedlings were used for the enzymatic extraction and assays. H2O2 was visualized and quantified in cotyledon leaves with a fluorescent probe (Amplex Red) 24 h after the VOC treatment. An H2O2 burst was detected above all in the leaf parenchyma tissues after treatment with 1-octen-3-ol 130 ppm per vol. and trans-2-octenal 13 ppm per vol., reaching, on average, three times (1.5 µm) the concentration observed in the control seedlings (0.5 µm) (Fig. 3a–c). The specific activities were also monitored of three enzymes involved in H2O2 production from the dismutation of the superoxide radical (SOD; EC 184.108.40.206., Halliwell, 2006); H2O2 removal (CAT; EC 220.127.116.11: 2H2O2→ 2H2O + O2); and removal of H2O2 by oxidizing it with the cosubstrate guaiacol (PX; EC 18.104.22.168, Halliwell, 2006). The enzymatic activities were determined at 15 min, and 2, 24 and 48 h after the VOC treatment. In order to emphasize the changes vs the control (CH2Cl2), the results were expressed as the activity ratio (AVOC/Acontrol). Increases in enzymatic activities were observed within 2 h of the VOC treatment with 1-octen-3-ol 13 ppm per vol. and trans-2-octenal 130 ppm per vol., but peaked (up to 100-fold increases in some cases) at 24 or 48 h for 1-octen-3-ol 130 ppm per vol. and trans-2-octenal 13 and 130 ppm per vol. (Fig. 4a–c). Bleaching was also monitored for the four time points of the enzymatic assays and correlated to the strong increase in the enzymatic activities of the SOD, CAT and PX observed after 24 and 48 h (Fig. 4d).
In this paper, we demonstrate the phytotoxic action of the natural aroma of truffles on the development of C. incanus and A. thaliana, which is mirrored by root inhibition and leaf bleaching. The fact that both the host and nonhost plants were inhibited by volatiles released by the fruiting bodies suggests that these volatiles might not serve as premycorrhizal signals, but rather as phytotoxic compounds. Far from natural conditions, the experimental design we used is known to maximize the probability of response of the plant (Baldwin et al., 2006). Nevertheless, truffle VOCs could also be phytotoxic in nature for two reasons. First, in natural conditions, fruiting bodies release volatiles continuously and probably for a much longer period than the timescale used in our experiments. Second, since truffle fruiting bodies are located 10–20 cm underground, the VOCs will slowly diffuse into the soil, forming a plume with a concentration gradient, as was demonstrated in the case of a sesquiterpene released from the roots of corn seedlings (Rasmann et al., 2005).
As far as each single VOC is concerned, the quantities released from the fruiting bodies over 63 h were not sufficient to reach the action threshold on agar (Table 1). A single VOC might therefore not be responsible for the observed effect, and the A. thaliana seedlings could also be more sensitive to VOCs when germinated on filter papers (Fig. 1) and not on agar (Fig. 2). VOCs could also act synergistically with each other.
Of the 10 VOCs tested here, it is worth nothing that the most active compounds were the C8 VOCs (Fig. 2) which belong to a large group of fungal volatiles (Chiron & Michelot, 2005), while the plant growth promoting 2,3-butanediol did not inhibit the seedling development or induce bleaching (Fig. 2). This observation is consistent with the fact that 2,3-butanediol only stimulates plant growth in hormone-like concentrations (Ryu et al., 2003). Taken as a whole, the results suggest that fungal VOCs may serve to keep plant development under control.
The second unexpected observation was that truffle VOCs caused a strong bleaching in the A. thaliana leaves. Bleaching has been reported in A. thaliana as a consequence of various stresses, including salt (Achard et al., 2006), viral infection (Deleris et al., 2006) and oxidative bursts induced by the herbicide Paraquat (Cao et al., 2006). The strongest bleaching seen with trans-2-octenal is consistent with the observation that aldehydes containing an α,β-unsaturated carbonyl group are highly reactive compounds that can readily cause damage to plant cells (Almeras et al., 2003).
In our experiments, the degree of bleaching of the seedlings can be considered as a marker of Arabidopsis responsiveness to the VOCs. Bleaching is generally correlated to an increased ROS-scavenging enzyme activity and/or increased H2O2 concentrations (Figs 3, 4). On the one hand, 1-octen-3-ol 130 ppm per vol. and trans-2-octenal 13 ppm per vol. produced an H2O2 burst after 24 h, which coincided with the bleaching of the seedlings and an increase in the SOD and PX activities. The H2O2 burst is consistent with the increase in SOD activity, but the bleaching suggests that the increase in PX was not sufficient to detoxify H2O2. On the other hand, no oxidative burst could be observed with trans-2-octenal 130 ppm per vol. after 24 h (Fig. 3). The sharp increase in CAT at 24 h suggests that H2O2 could be fully scavenged, but the strong bleaching seen at this time point would seem to indicate that an oxidative burst could have taken place earlier.
The full action mechanism of the fungal volatiles tested in this study is still unclear, yet our data demonstrate that they modify the A. thaliana oxidative metabolism to a great extent. Although an oxidative burst is a common feature of both biotic and abiotic stress responses (Dat et al., 2000; Mittler et al., 2004), this is the first time, to our knowledge, that fungal volatiles have been implicated as its elicitor.
We have demonstrated the inhibitory action of truffle volatiles on the development of the nonhost plant A. thaliana in laboratory conditions. The most active volatiles were C8 compounds – described in truffle fruiting bodies and mycelia (Splivallo et al., 2007) but also widespread among higher fungi. These fungal volatiles inhibited the growth of A. thaliana and modified its oxidative metabolism to a great extent. Even though these results are based on laboratory experiments, they raise questions about the role of fungal volatiles in nature. Could these volatiles be the reason for the ‘burnt’ areas that are commonly found in nature around truffle-mycorrhized plants? From a more general point of view, could they serve fungi as a competitive weapon against plants? For now, these questions remain unanswered.
We would like to thank Prof. M. Maffei for his suggestions and comments, Dr Roberta Sirovich for her assistance with the statistical analysis, Dr Alfredo Vizzini for the morphological identification of the truffles, and Dr Claude Murat for his help with the molecular data. The experimental work was supported by the CEBIOVEM Centre of Excellence, IPP-CNR and the CIPE Project B063 ‘Analisi Genetico-Molecolare per la Qualità e Sicurezza del Prodotto Tartufo’. RS was supported by a fellowship from the Fondation pour des Bourses d’Études Italo-Suisses.