OUP user menu

Sublethal concentrations of azoles induce tri transcript levels and trichothecene production in Fusarium graminearum

Tomasz Kulik, Maciej Łojko, Marika Jestoi, Juliusz Perkowski
DOI: http://dx.doi.org/10.1111/j.1574-6968.2012.02637.x 58-67 First published online: 1 October 2012

Abstract

The effect of sublethal concentrations (below the recommended field doses) of propiconazole and tebuconazole on the amount of tri transcripts and accumulation of trichothecenes by three Fusarium graminearum isolates of 3ADON, 15ADON, and NIV chemotypes was examined on yeast extract sucrose agar (YES) medium. RT-qPCR analyses showed higher tri4,tri5, and tri11 transcript levels in cultures of all three F. graminearum isolates supplemented with sublethal concentrations of azoles as compared to those in nontreated control, although the fold changes in the amount of tri transcripts differed according to the type of azole used. Mycotoxin analysis revealed higher increase in trichothecene accumulation in most of the tebuconazole-treated samples of all chemotypes tested. A huge increase in all trichothecene compounds was revealed in samples of all F. graminearum isolates treated with 5 mg L−1 of tebuconazole. An inducing effect of azoles on trichothecene accumulation in the grain was confirmed in an in planta experiment; however, the results obtained were inconsistent. A higher amount of trichothecenes and fungal DNA was quantitated in two grain samples treated with sublethal propiconazole concentrations. In contrast, no significant increase in trichothecene levels was revealed in grain samples treated with sublethal concentrations of tebuconazole.

Keywords
  • Fusarium graminearum
  • azoles
  • tri transcripts
  • trichothecenes

Introduction

The Fusarium graminearum (teleomorph Gibberella zeae) species complex is one of the most important causal agents of Fusarium head blight (FHB) of wheat and other cereals worldwide (Ward et al., 2008). Fusarium graminearum contaminates the grain with high levels of type B trichothecenes: deoxynivalenol (DON) and nivalenol (NIV) and their acetyl derivatives. Contamination of plant products with these toxins poses a significant risk to food safety and animal health (Foroud & Eudes, 2009).

Three major F. graminearum chemotypes: 3-acetyldeoxynivalenol (3ADON), 15-acetyldeoxynivalenol (15ADON), and NIV synthesize specific trichothecene compounds (Wang et al., 2011). 3ADON chemotype synthesizes DON and 3ADON, 15ADON chemotype produces DON and 15ADON, while NIV chemotype produces NIV and 4ANIV (4-acetylnivalenol Wang et al., 2011). However, it has been documented that some isolates from one defined chemotype are able to produce mycotoxins from other chemotypes in considerable amounts (Ward et al., 2002 Mugrabi de Kuppler et al., 2011). In F. graminearum, the enzymes catalyzing the biochemical reactions which result in formation of trichothecenes are encoded by tri genes (Foroud & Eudes, 2009). Polymorphism of tri sequences contributes to the trichothecene chemotypes. NIV synthesis is determined by the expression of both tri7 and tri13 genes, while in DON chemotypes, tri13 and tri7 are nonfunctional as a result of multiple insertions and deletions (Lee et al., 2002). The sequence differences resulting in differential activity of tri8 are a key determinant of the 3ADON and 15ADON chemotypes in F. graminearum (Alexander et al., 2011).

Besides its genetic background, mycotoxin production has received considerable attention in analyses of external factors affecting trichothecene production within Fusarium. It has been demonstrated that regulation of mycotoxin biosynthesis occurs primarily at a transcriptional level (Proctor et al., 1999 Marín et al., 2010). Estimating relative transcript abundances by RT-qPCR allows for precise identification of factors regulating the biosynthesis of mycotoxins in Fusarium (Merhej et al., 2011). The impact of abiotic factors such as temperature (Schmidt-Heydt et al., 2008 Marín et al., 2010), osmotic potential (Marín et al., 2010), and pH (Merhej et al., 2010) on tri transcript levels and trichothecene accumulation in media has been examined. Moreover, several reports have indicated that different substrates (Jiao et al., 2008 Gardiner et al., 2009) and signaling molecules (Ponts et al., 2007) regulate mycotoxin production in Fusarium. Limited studies have identified the impact of anthropogenic factors such as fungicides on trichothecene biosynthesis within Fusarium, especially at a transcriptional level (Covarelli et al., 2004 Ochiai et al., 2007).

Among the fungicides used, the application of azoles during wheat anthesis is a primary method for management of FHB (Paul et al., 2010). These compounds block the ergosterol biosynthesis pathway by inhibiting the sterol 14-α-demethylase encoded by the CYP51 gene (Liu et al., 2010). Azoles have been shown to be effective in reducing FHB symptoms and DON content in wheat, although the effectiveness between azole compounds varies (Paul et al., 2010). On the other hand, unsatisfactory effects of this group of fungicides against Fusarium spp. have also been documented (Mesterházy et al., 2011). It seems that the effectiveness of azoles in reducing Fusarium biomass and mycotoxin content is strongly dependent on different factors, particularly disease severity, the resistance level of the wheat cultivar, and the spraying technology (Paul et al., 2010 Mesterházy et al., 2011). The optimal concentration of fungicides in plant tissues is essential for effective control of fungal pathogens in the field. However, azoles appear to be only partially systemic in wheat and do not translocate well from leaves to heads or inside heads (Mauler-Machnik & Zahn, 1994). Several reports have indicated the inducing effect of sublethal concentrations of azoles on trichothecene biosynthesis within the F. graminearum complex. Ochiai et al. (2007) showed that sublethal concentrations of tebuconazole induce tri5 transcript level in genetically engineered Fusarium asiaticum, which results in increased production of NIV-type trichothecenes. In another study, Becher et al. (2010) showed that in vitro adaptation of the F. graminearum strain to a sublethal dose of tebuconazole resulted in the recovering of morphologically distinguishable azole-resistant phenotypes that produced higher levels of NIV (Becher et al., 2010). Recent studies of Audenaert et al. (2010) showed that sublethal concentrations of prothioconazole induce hydrogen peroxide in F. graminearum, which results in increased accumulation of DON. Interestingly, an inducing effect of azoles on tri transcript levels and trichothecene biosynthesis has not been found in closely related Fusarium culmorum (Covarelli et al., 2004). In this study, the effect of sublethal concentrations of propiconazole and tebuconazole on tri transcript levels and the accumulation of trichothecenes was investigated. The term sublethal is understood to mean concentrations below the recommended field doses. Three F. graminearum field isolates identified preliminary by qPCR assays as potential 3ADON, 15ADON, and NIV producers were used. In an in vitro assay, fungal isolates were grown on yeast extract sucrose agar (YES) medium with sublethal concentrations of azoles. RT-qPCR analyses were performed using highly sensitive TaqMan technology. In addition, trichothecene content was determined. In an in planta assay, the effect of sublethal levels of azoles on trichothecene levels and fungal DNA in grain samples harvested from artificially inoculated wheat heads was analyzed. This work underlines the risk of enhanced trichothecene production by F. graminearum under low concentrations of azoles.

Materials and methods

Fusarium graminearum field isolates

Three F. graminearum field isolates were used in this study: DDPP1002T (3ADON chemotype), DDPP1001T (15ADON chemotype), and DDPP0357 (NIV chemotype). The isolates were isolated from Fusarium-damaged kernels from two wheat fields located in northern Poland. Both DDPP1002T and DDPP1001T isolates were isolated in 2010, while isolate DDPP0357 was recovered in 2003. The isolates were identified to the species level using a qPCR assay developed by Waalwijk et al. (2004). Their potential to produce 3ADON, 15ADON, and NIV was preliminary assessed using three qPCR assays specific for 3ADON, 15ADON, and NIV producers (Kulik, 2011). The isolates are available at the Department of Diagnostics and Plant Pathophysiology, University of Warmia and Mazury in Olsztyn. Isolates are stored as mycelium/spore suspensions in 15% glycerol at − 25 °C.

Medium and culture conditions

YES agar medium (yeast extract 20 g L−1, sucrose 150 g L−1, MgSO4.7H2O 0.5 g L−1, agar 20 g L−1) recommended for secondary metabolite analysis was used. Propiconazole and tebuconazole (Sigma-Aldrich, Germany) were dissolved in 0.65 mL of acetone and then added to autoclaved YES medium to obtain the final concentrations: 0.25 mg L−1, 0.5 mg L−1, 2.5 mg L−1, and 5 mg L−1. Recommended field doses of both azoles completely inhibited fungal growth on the media. The control sample was supplemented with an identical volume of acetone. Experiments were performed on Petri plates (Ø 80 mm). Petri plates containing 10 mL of YES medium were inoculated with fungal hyphae with a sterile tip and incubated at 25 °C in darkness. For each condition, plates (in triplicate) were incubated at 25 °C for 4 days.

Extraction of total RNA and preparation of cDNA

The total RNA was extracted from 4-day-old cultures from three F. graminearum field isolates grown on YES medium with or without supplementation of the tested azole. Two biological replications were prepared for each condition independently in time. Mycelium (350 mg) was ground in liquid nitrogen with mortar and pestle. Total RNA was extracted using a Quick-RNA™ MiniPrep kit (Zymo Research) following the manufacturer recommendations. Total RNA was reverse-transcribed using the SuperScript® VILO™ cDNA Synthesis Kit (Invitrogen). Reverse transcription was performed immediately after RNA extraction with a Mastercycler ep gradient (Eppendorf AG, Germany) with the thermal cycling conditions recommended by the manufacturer (Invitrogen). cDNA samples were stored at − 25 °C for RT-qPCR analysis.

RT-qPCR and data analyses

To design primer/probe sets for RT-qPCR analyses, the F. graminearum sequence data of ef1α, tri4, tri5, and tri11 published in the NCBI database were aligned with geneious pro 4.0.0 (Drummond et al., 2011). To prevent amplification of genomic DNA, at least one primer and/or probe from each set of primers/probes was designed on exon–intron boundaries using primer express 3.0 (Applied Biosystems, Foster City Table 1).

View this table:
Table 1

List of primers and probes designed for RT-qPCR analyses

Target transcriptPrimer/TaqMan probe namePrimer/TaqMan probe sequence
ef1αef11
ef12
ef1α probe
TCGACAAGCGAACCATCGA
CCCAGGCGTACTTGAAGGAA
VIC-CGAGAAGGAAGCCGC-MGB
tri4tri41
tri42
tri4probe
TGCATGAAATAGGTGGACTGAGA
AACTTGAAGTACAAGGAGCATGTCA
FAM-ATGGGAGTTCCTTTAGGG-MGB
tri5tri51
tri52
tri5probe
AACGAGCACTTTCCCAACGT
ATCCAACATCCCTCAAAAAAGTC
FAM-TCATTGAACCTTATCCGTAGCA-MGB
tri11tri111
tri112
tri11probe
CCAGCATCATGCGCATCTC
AATCGGACCACGGAATTGTATT
FAM-CGTAGGCAAGGTTCATA-MGB

Probes, conjugated with an MGB group, were labeled at the 5′-end with FAM, while the ef1α probe was labeled at the 5′-end with VIC. All primers were synthesized by Genomed (Warsaw, Poland), while MGB probes were ordered from ABI PRISM Primers and TaqMan Probe Synthesis Service. Duplex RT-qPCR reaction conditions were used for each tri transcript, including the ef1α reference control in the fast PCR protocol: 95 °C for 20 s (95 °C for 3 s, 60 °C for 55 s) × 36.

The RT-qPCR reagents were optimized as follows: 2 µL cDNA, 10 µL H2O, 1.8 µM of each ef1α primer, 0.5 µM of ef1α probe, 6 µM of either the tri4, tri5 or tri11 primers and 1.7 µM of either the tri4, tri5 or tri11 probe, and 5 µL Real-Time 2 × PCR Master Mix Probe (A&A Biotechnology, Gdynia, Poland). Tubes containing 1.25 mL of Real-Time 2 × PCR Master Mix Probe were mixed with 20 µL of ROX 50 × (A&A Biotechnology) before TaqMan analysis. Real-Time 2 × PCR Master Mix Probe is composed from 1 U µL−1 Taq DNA polymerase, reaction buffer (2 ×), MgCl2 (10 mM), and dNTP mix (0.5 mM each).

All PCR amplifications were carried out in a 7500 Fast Real-Time PCR System (Applied Biosystems) with a final volume of 17 µL. The threshold value was 0.1 and 0.05 for tri and ef1α transcripts, respectively. Each qPCR reaction was prepared in at least six replicates. The amplification efficiency of each duplex assay was determined based on five fivefold dilutions of the cDNA template. The PCR efficiencies obtained were as follows: 99.3%, (R2 = 0.947, slope = − 3.339, Y-inter = 27.418) for ef1α, 99.7% (R2 = 0.957, slope − 3.329, Y-inter = 23.226) for tri4, 97% (R2 = 0.929, slope − 3.394, Y-inter = 27.245) for tri5, and 94.5% (R2 = 0.975, slope − 3.462, Y-inter = 25.246) for tri11. In this study, the relative quantitation of tri targets was normalized to an ef1α reference gene. Ef1α was found to be constitutively expressed in F. culmorum (Covarelli et al., 2004) and F. graminearum (Lysøe et al., 2009). The Cq values of the target tri4, tri5, tri11 and reference ef1α gene were compared to those in control and treated samples and normalized relative to the Cq values obtained for the reference ef1α gene using the Relative Expression Software Tool 2009 (rest). The mathematical model used accounts for differences in efficiencies for the reference gene and the target gene and for the mean Cq deviation between the control and treated conditions (Pfaffl et al., 2002). The expression ratio results were tested for significance by running a Pair Wise Reallocation Randomisation Test© with a P value of 0.001 using the rest 2009 software (Pfaffl et al., 2002).

Inoculation of wheat heads and application of azoles in planta

Fusarium graminearum isolates were kept on potato dextrose agar medium at 25 °C for 14 days. To promote sporulation, a cycle of 12-h darkness and 12-h daylight was applied. Ultraviolet light (UV) was not applied to prevent introduction of potential UV mutations into the field. Approximately 3000 winter wheat heads (var. Wydma) per plot (6 m2) were spray-inoculated with a Titan 16 hand-sprayer (Marolex, Poland) at flowering, with a mixture of three F. graminearum isolates as described previously by Suchowilska et al. (2010). Three days after inoculation, plants were sprayed with different dilutions of propiconazole (Bumper 250 EC Makhteshim Agan Industries Ltd, Israel) and tebuconazole (Orius 250 EW Makhteshim Agan Industries Ltd) starting from a field dose. The field dose of each fungicide differed according to manufacturer instructions and was 125 g ha−1 (1250 mg L−1) and 250 g ha−1 (1500 mg L−1) of propiconazole and tebuconazole, respectively. Fungicide spraying was repeated after 14 days to strengthen the effect of azoles on Fusarium isolates. In the positive control group, wheat plants were inoculated with fungal biomass without fungicide spraying. In the negative control group, wheat plants were not inoculated with fungal biomass without fungicide treatment.

In addition, 25 wheat heads from each plot we collected 24 h after the first fungicide treatment for azole quantitation. Azoles were assessed using gas chromatography (GC Łozowicka et al., 2009, 2011) at the Plant Protection Institute, National Research Institute in Białystok. GC analysis was performed with an Agilent (Waldbronn, Germany) model 7890 A gas chromatograph equipped with electron capture (ECD) and nitrogen-phosphorus (NPD) with a non-polar column HP-5 ((5%-Phenyl)-methylpolysiloxane 30 m × 0.32 mm and film thickness 0.50 µm) and Chemstation chromatography manager data acquisition and processing system (Hewlette-Packard, version A.10.2). For confirmation of residues, a mid-polarity column HP-35 ((35%-Phenyl)-methylpolysiloxane 30 m × 0.32 mm and film thickness 0.50 µm) was used. The operating conditions were as follows: for detectors — injector temperature, 210 °C carrier gas, helium at a flow-rate of 3.0 mL min−1 detector temperature, 300 °C (ECD and NPD) make up gas, nitrogen at a flow-rate of 57 mL min−1 (ECD) and 8 mL min−1 (NPD), hydrogen 3.0 mL min−1, air 60 mL min−1 for oven-initial temperature, 120 °C increase to 190 °C at 16 °C min−1, then to 230 °C at 8 °C min−1 and finally to 285 °C at 18 °C min−1 and hold 10 min (ECD and NPD). The volume of final sample extract injected at 210 °C in splitless mode (purge off time 2 min) was 2 mL. Total time of analysis: 20.43 min.

Preparation of cell lysates and quantitation of Fusarium genotypes from wheat grain by qPCR

The amounts of trichothecene genotypes (3ADON, 15ADON, and NIV) were quantitated in pooled samples by three qPCR assays specific to 3ADON, 15ADON, and NIV producers within F. culmorum/F. graminearum (Kulik, 2011). Each pooled sample (100 g) contained kernels pooled from c. 500 wheat heads per sample. Kernels were ground to a fine powder for 5 min in A11 basic analytical mill (IKA, Germany). Preparation of cell lysates from 0.1 g of grounded kernels was made in triplicate from each sample as previously described (Kulik, 2011). Each qPCR reaction was prepared in at least three repetitions.

GC-MS analysis of trichothecenes from YES medium and grain samples

The levels of DON, 3ADON, 15ADON, NIV, and 4ANIV in an in vitro experiment were determined in 10 pooled samples by GC-MS analysis as previously described by Perkowski et al. (2003) (Table 2 and Table 3). Each pooled sample contained lyophilized fungal biomass pooled from seven replicates (7 Petri plates per variant). Each pooled sample was analyzed once. The detection limit for analyzed toxins was 0.001 mg kg−1. The levels of DON, 3ADON, 15ADON, and NIV were determined in pooled grain samples by GC-MS as previously described by Eskola et al. (2001). Each pooled sample (100 g) contained kernels pooled from c. 500 wheat heads per sample. Each sample was analyzed once. The limits of quantitation were 0.01 mg kg−1 for DON and its derivatives and 0.03 mg kg−1 for NIV.

View this table:
Table 2

Changes in transcript abundances of tri4, tri5, and tri11 and trichothecene accumulation by F. graminearum chemotypes grown on YES medium under different propiconazole concentrations

Tri gene up-regulationTrichothecene concentration (mg kg−1)
Propiconazole level (mg L−1)tri4tri5tri11DON3ADON15ADONNIV4ANIV
3ADON chemotype
N.T.C.0.91.6
0.257 ± 12 ± 25 ± 11.15.2
0.59 ± 24 ± 27 ± 127.6
2.510 ± 419 ± 116 ± 126.5
59 ± 220 ± 17 ± 42.58.4
15ADON chemotype
N.T.C.2.20.60.6
0.25N.D.N.D.N.D.0.30.10.1
0.5N.D.N.D.N.D.0.40.10.2
2.511.5 ± 910.5 ± 62.5 ± 210.40.5
510.5 ± 66 ± 14 ± 20.80.40.5
NIV chemotype
N.T.C.0.040.04
0.251 ± 14 ± 44 ± 419.850.3
0.5355 ± 320N.D.3.5 ± 20.40.7
2.5419 ± 395112 ± 1038 ± 40.080.3
549 ± 46N.D.14 ± 140.040.1
  • Fold-change values represent up-regulation in F. graminearum isolates treated with azole tested as compared to nontreated control. N.D., not different at P(H1) = 0.001; ±, standard error of two biological replicates; N.T.C., nontreated control; –, not detected.

View this table:
Table 3

Changes in transcript abundances of tri4, tri5 and tri11 and trichothecene accumulation by F. graminearum chemotypes grown on YES medium under different tebuconazole concentrations

Tri gene up-regulationTrichothecene concentration (mg kg−1)
Tebuconazole level (mg L−1)tri4tri5tri11DON3ADON15ADONNIV4ANIV
3ADON chemotype
N.T.C.0.91.6
0.2513 ± 117.5 ± 2.510 ± 3525
0.520 ± 222.5 ± 0.558 ± 381047
2.526 ± 258 ± 73 ± 20.85
54 ± 32 ± 0.53 ± 224182
15ADON chemotype
N.T.C.2.20.60.6
0.25N.D.N.D.N.D.1.80.50.6
0.58.5 ± 0.56 ± 323 ± 223557.3
2.512 ± 111 ± 7N.D.152.58.7
542 ± 40201 ± 19936 ± 161691363308
NIV chemotype
N.T.C.0.040.04
0.252.5 ± 0.545 ± 393 ± 20.20.3
0.54 ± 387 ± 3712 ± 110.10.2
2.5N.D.N.D.N.D.0.20.7
5N.D.30 ± 1112 ± 103451296
  • Fold-change values represent up-regulation in F. graminearum isolates treated with azole tested as compared to nontreated control. N.D., not different at P(H1) = 0.001; ±, standard error of two biological replicates; N.T.C., nontreated control; –, not detected.

Statistical analyses

The relationships between the quantitated DNA and trichothecene concentrations in grain samples were determined by Pearson's correlation analysis using statistica software (Data Analysis Software System, version 6.1 StatSoft Inc., 2003, http://www.statsoft.com).

Results

In vitro supplementation of sublethal concentrations of propiconazole and tebuconazole increases tri4, tri5 and tri11 transcript levels within F. graminearum chemotypes

The quantitation of transcripts of tri4, tri5, and tri11 genes located in the 12-gene core tri cluster (Brown et al., 2004) was evaluated using TaqMan probes. The proposed trichothecene biosynthetic pathway in Fusarium has been presented in Foroud & Eudes (2009). The analyzed genes encode the first steps of the type B trichothecene biosynthesis pathway and are representative of the initial flux of the biosynthetic pathway. Table 2 and Table 3 show fold-change values representing tri up-regulation in F. graminearum isolates treated with azoles as compared to nontreated control.

The tri transcript levels were always higher in cultures supplemented with sublethal concentrations of azoles, although in some cases, fold-change values were not significantly altered [P(H1) = 0.001]. Among the tri transcripts analyzed of all studied isolates, the amount of tri4 transcript was the highest during the culture process followed by tri11 and tri5 (data not shown). It should be noted that the tri transcript levels in nontreated samples differed among the tested DON and NIV chemotypes. The tri transcript levels of DON chemotypes were at a similar and higher level than in the NIV chemotype (data not shown).

Notably, the tri transcript levels seemed to be related to the type of azole used. Within DON chemotypes, the amount of tri transcripts treated with tebuconazole was higher compared to samples treated with propiconazole however, such a relation was not clear for the NIV chemotype (Table 2 and Table 3).

In vitro supplementation of sublethal concentrations of propiconazole and tebuconazole affects differential accumulation of trichothecenes within F. graminearum chemotypes

In an independent experiment, the levels of trichothecenes (DON, 3ADON, 15ADON, NIV, and 4ANIV) were determined in 14-day-old cultures supplemented or not with different concentrations of azoles Table 2 and Table 3). Isolate 1002T, identified with qPCR assay as 3ADON genotype, accumulated DON and higher amounts of 3ADON. Isolate 1001T, determined to be of the 15ADON genotype, produced DON and lower amounts of 3ADON and 15ADON. Isolate 0357, predicted with qPCR assay as an NIV producer, accumulated NIV, 4ANIV. For 3ADON chemotype, an increase in DON and 3ADON was revealed in samples treated with all sublethal concentrations of propiconazole. However, all samples of 15ADON chemotype exhibited decreased accumulation of trichothecenes as compared to N.T.C. A considerable increase in NIV and 4ANIV was reported in NIV chemotype treated with all sublethal levels of propiconazole. An increase in trichothecene accumulation was revealed in most of the tebuconazole-treated samples of all chemotypes. Notably, a huge increase in all trichothecene compounds was revealed in samples of all chemotypes treated with 5 mg L−1 of tebuconazole.

Sublethal concentrations of azoles as potential factors affecting fungal biomass and trichothecene levels in grain

In an in planta experiment, fungal DNA and trichothecene accumulation were assessed in grain samples collected from wheat heads treated with different concentrations of azoles tested (Table 4). A higher amount of 3ADON DNA was quantitated with qPCR in the sample treated with 125 mg L−1 of propiconazole. Correspondingly, the highest levels of DON were detected in this sample. Two samples treated with 125 and 5 mg L−1 of propiconazole showed a higher amount of NIV DNA. Similarly, the highest level of NIV was detected in these samples. In samples treated with tebuconazole, an increase in 3ADON DNA as compared to the positive control was found in a sample treated with 3 mg L−1 of tebuconazole, although the increase was not significant. In this sample, the highest levels of DON/3ADON were detected.

View this table:
Table 4

In planta effect of sublethal levels of propiconazole and tebuconazole on the amount of fungal DNA and trichothecene accumulation in grain samples

Amount of fungal DNA in grain samples (ng 0.1 g−1)Amount of trichothecenes in grain samples (mg kg−1)
Amount of azoles used in fungicide spraying (mg L−1; dilution)Amount of azoles in wheat heads (mg kg−1)3ADON15ADONNIVDON3ADON15ADONNIV
Samples treated with different levels of propiconazole
    1250 (field dose)1.04 ± 0.520.95 (D)0.51 (B)0.23 (D)5.30.0190.0490.17
    125 (10 ×)0.15 ± 0.083.06 (A)0.46 (BC)0.87 (A)7.20.0330.0440.4
    5 (50 ×)0.04 ± 0.021.4 (C)0.65 (B)0.46 (B)6.40.0270.0640.35
    25 (100 ×)0.03 ± 0.020.99 (D)0.57 (B)0.26 (CD)4.90.0210.0470.18
    2.5 (500 ×)N.D.0.77 (D)0.47 (C)0.34 (CD)3.70.0150.0440.18
    1.25 (1000 ×)N.D.0.85 (D)0.29 (C)0.39 (BC)3.30.0180.0320.23
Samples treated with different levels of tebuconazole
    1500 (field dose)6.51 ± 3.260.48 (B)0.23 (D)0.31 (A)2.50.010.030.21
    150 (10 ×)0.67 ± 0.340.85 (B)0.7 (BC)0.28 (AB)5.80.0210.0720.27
    30 (50 ×)0.18 ± 0.090.52 (B)0.34 (CD)0.31 (A)2.70.0120.0360.28
    15 (100 ×)0.12 ± 0.060.71 (B)0.43 (CD)0.21 (BC)3.30.0190.0430.15
    3 (500 ×)0.05 ± 0.031.92 (A)0.2 (D)0.21 (C)11.80.0810.0230.63
    1.5 (1000 ×)0.04 ± 0.020.62 (B)0.47 (BCD)0.1 (CD)3.30.0140.0440.12
Positive Control1.84 (A)0.91 (A)0.34 (A)6.10.0370.0830.28
Negative control0.1
  • ±, standard error, different letters (A, B) indicate significant differences, –, not detected.

In this experiment, a high correlation was found between the amount of fungal DNA and trichothecene compounds (Table 5). The lack of a strong relationship between 15ADON DNA and DON could result from the high production of this compound by the 3ADON chemotype.

View this table:
Table 5

Correlation between the amount of fungal DNA and trichothecene compounds in grain samples collected from wheat heads treated with different levels of azoles tested

Samples treated with propiconazoleSamples treated with tebuconazole
Trichothecene compounds
F. graminearum chemotypesDON3ADON15ADONNIVDON3ADON15ADONNIV
3ADON0.680.670.750.78
15ADON0.290.960.0010.98
NIV0.740.91
  • Values represent coefficients of determination (R2) for the relationship between the amount of fungal DNA of three F. graminearum chemotypes and levels of trichothecene compounds in grain samples collected from wheat heads treated with different levels of azoles tested. –, not analyzed.

Discussion

Azoles are widely used fungicides in agriculture (Paul et al., 2010 Mesterházy et al., 2011) and to treat human mycosis (Giavini & Menegol, 2010). Their antifungal activity is based on their ability to inhibit CYP51, a key enzyme in ergosterol biosynthesis (Liu et al., 2010). Azoles have been shown to differ in the control of Fusarium spp., and their unsatisfactory effectiveness may be associated with an insufficient concentration of fungicides in plant tissues (Mesterházy et al., 2011).

In the most recent study, Audenaert et al. (2010) showed that the treatment of F. graminearum with sublethal concentrations of prothioconazole resulted in increased accumulation of DON. It has been further demonstrated that the enhancement of DON production was indicated by the oxidative stress caused by fungicide treatment. Hydrogen peroxide (H2O2) triggers trichothecene biosynthesis in DON chemotypes of F. culmorum/F. graminearum, although NIV chemotypes seem to show higher adaptation to oxidative stress (Ponts et al., 2009). It has been demonstrated that treatment of NIV chemotypes with H2O2 results in decreased accumulation of this toxin (Ponts et al., 2009).

In this study, we showed that treatment of either DON or NIV chemotypes of F. graminearum with sublethal concentrations of azoles results in increased tri transcript levels, which leads to increased accumulation of trichothecenes. This observation is supported by studies of Ochiai et al. (2007) showing that sublethal concentrations of tebuconazole increased tri5 transcript level in genetically engineered F. asiaticum and increased production of NIV-type trichothecenes. Similarly, recent results of Becher et al. (2010) showed that in vitro adaptation of F. graminearum NIV chemotype to sublethal dose of tebuconazole resulted in recovering isolates producing higher levels of NIV.

In the present study, RT-qPCR results did not always parallel the trichothecene accumulation. Three different explanations of this discrepancy are possible. Firstly, the commonly observed low toxin production of F. graminearum in axenic cultures (Gardiner et al., 2009) results in a lack of considerable differences between the treated samples and N.T.C. This was especially evident in the samples of 15ADON chemotype treated with propiconazole. Notably, in these samples, an increase in the amount of tri transcripts was lower than in tebuconazole-treated samples where a higher level of toxins was quantitated. It is tempting to speculate that relatively low tri transcript level in propiconazole-treated samples was the result of less effective induction of H2O2 in the fungus. Ponts et al. (2007) demonstrated that treatment of 15ADON chemotype of F. graminearum with H2O2 resulted in up to 11- and 19-fold increase in tri4 and tri5 transcript levels, respectively. Our results showed that most of the propiconazole-treated samples resulted in a lower tri transcript levels as observed by Ponts et al. (2007), which probably affected low toxin accumulation. Secondly, trichothecene accumulation by azole stress could result from an unknown, additional modulation mechanism which is independent from transcriptional regulation. This hypothesis was suggested by Ponts et al. (2009) who demonstrated differential antioxidant defense responses within F. graminearum strains to H2O2. Thirdly, the discrepancies could also result from variation between the fungal cultures studied. Both RT-qPCR and toxicological analysis were performed on different fungal cultures that might differ at transcriptional levels. We found that despite theoretically identical conditions, the results from two biological replications differed in some cases in the level of tri transcript (data not shown). Such variation could result from partial nutrient deficiency that is exhausted rapidly on agar media (Schmidt-Heydt et al., 2008). Notably, intraculture differences have been observed by Ochiai et al. (2007) who demonstrated differential tri5 transcript levels in fungal hyphae. Moreover, a recent study by Audenaert et al. (2012) demonstrated the increased sensitivity of a tri5 knockout mutant compared to its wild-type parent strain, which indicated that biosynthesis of trichothecenes might also have a physiological meaning.

In an in planta experiment, we analyzed whether treatment of inoculated wheat heads with sublethal azole concentrations could increase fungal DNA and toxin levels in the grain. The presence of azoles in wheat heads was confirmed within 24 h of the first fungicide spraying. The concentrations of azoles differed and values ranged from nondetectable to 1.04 and 6.51 mg kg−1 in samples treated with recommended concentrations of propiconazole and tebuconazole, respectively (Table 4). This indicates that the accumulation of propiconazole and tebuconazole in wheat heads differs and might reflect previously reported differences in the effectiveness between these two azole compounds (Paul et al., 2008). Among the 10 samples treated with diluted concentrations of azoles, an increase in 3ADON DNA was revealed in only one sample treated with 125 mg L−1 of propiconazole. A higher amount of NIV DNA was quantitated in two samples treated with 125 and 5 mg L−1 of propiconazole. Correspondingly, the highest levels of DON/NIV were detected in these samples. No significant increase in trichothecene levels and fungal DNA as compared to a positive control was found in samples treated with tebuconazole. A lack of similar results in the rest of the samples could result from the fact that a complex of factors affects trichothecene biosynthesis by the fungus in the field. Importantly, the impact of these speculated factors was exerted over a relatively long time [wheat heads were sprayed with fungicides (second spraying) 45 days before harvest]. Hallen-Adams et al. (2011) showed that tri5 transcripts can be detectable in plant material over a long time even after the plant tissue had completely senesced. During this time, the process of mycotoxin biosynthesis could be affected by various abiotic and biotic factors. In addition, plant defense mechanisms seem to play a prominent role in regulating trichothecene biosynthesis and biomass growth (Merhej et al., 2011). The influence of these external factors could mask the effect of fungicides used. It seems that specific plant compounds induce trichothecene biosynthesis more effectively. Previous studies of Gardiner et al. (2009) showed that absolute levels of DON induction using H2O2 as revealed by Ponts et al. (2007) appeared low compared to, for example, a > 100-fold increase in DON production with agmatine however, they used different media so it is difficult to compare these results with ours. However, the results of RT-qPCR analyses support this hypothesis. The tri transcript levels observed by Ponts et al. (2007) and in our study are relatively lower compared to the > 1000-fold changes seen after different amine treatment (Gardiner et al., 2009). Taken together, the results described here lead to a better insight into azole stress within F. graminearum chemotypes. We demonstrated that both propiconazole and tebuconazole induce tri transcript levels at sublethal concentrations, which results in differential trichothecene accumulation both in vitro and in planta. Finally, the data obtained here support the hypothesis that the response of Fusarium to azole stress is strain specific.

Acknowledgements

This study was supported by the Polish Ministry of Education and Science, from the Iuventus Plus IP2010 021470 grant. Special thanks to Dr Adam Okorski for statistical analysis.

References

View Abstract