This study was carried out on 172 samples of winter wheat. The samples consisted of various cultivars that had been randomly collected from farmers’ fields in different areas of Bavaria, South Germany. The objectives of this study were detecting the presence of Tri-5 gene producing fungus that generates trichothecene mycotoxins, especially Deoxynivalenol (DON), by using conventional qualitative PCR; determining the correlation between the presence of Tri-5 gene and DON content; evaluating the Fusarium graminearum and Fusarium culmorum infection by Real-Time PCR and estimating the correlation between DON content and the severity of F. graminearum and F. culmorum contamination. This study showed that 86% of all infected samples had a Tri-5 gene and amplified a single 544bp fragment associated with a detectable amount of DON (ranged from 10 to 2990 μg kg-1). This study demonstrated that F. graminearum is the predominant species associated with Fusarium head blight (FHB) and was considered as the predominant trichothecene producer that associated with FHB since there was a highly significant correlation (R2=0.7) between DON and F. graminearum DNA content, compared to a weak correlation (R2=0.03) between DON and DNA content of F. culmorum infected wheat kernels.
Keywords: Wheat, Fusarium head blight, Tri-5 gene, Trichothecene DON, Conventional, Real-time PCR
Fusarium head blight (FHB) of small grains was first described over a century ago and was considered as a major threat to wheat and barley during the early years of last century . Head blight or scab of wheat caused epidemics in many wheat area worldwide [2,3]. The International Maize and Wheat Improvement Centre (CIMMYT) have considered FHB as a major factor limiting wheat production in many parts of the world . FHB is also known as “tombstone” kernels of wheat because of the chalky and lifeless appearance of the infected kernels . It has the capacity to destroy a potentially high-yielding crop within few weeks . FHB is a significant disease of small-grain cereals throughout Europe , United States , Canada , South America , Asia  and Australia . FHB was identified more than 120 years ago, in 1884, in England. The United States Department of Agriculture ranked FHB as the worst plant disease to appear since the 1950's . It has increased worldwide  and it was considered as a major threat to wheat and barley during the early years of the twentieth century [13,14].
FHB is caused by a number of different fungal species of the genus Fusarium (Fusarium spp). However, F. avenaceum, F. culmorum, F. graminearum (teleomorph, Gibberella zeae), F. poae, and Microdochium nivale (teleomorph, Monographella nivalis) are the species which are most commonly associated with the FHB disease .
The threat posed by Fusarium spp. is multifaceted. It causes yield and quality losses due to sterility of the florets and formation of discolouration, which reduces kernel size and losses light weight kernels . In addition, grain quality factors such as protein content and germination can be severely affected by the pathogen . Several Fusarium species which cause FHB are able to produce trichothecene mycotoxin. F. culmorum, F. graminearum, and F. poae produce type B trichothecenes such as nivalenol (NIV), deoxnivalenol (DON), and fusaenon-X , while other species are not . DON is the predominant Trichothecenes found in Europe and North America . Trichothecene produced by this fungus pose a serious hazard to human and animal health  because these toxic materials are potent inhibitors of eukaryotic protein biosynthesis [20,21]. Acute adverse effects of the toxin in animals causes weight loss and feeding refusal in non-ruminant livestock, high rates of abortion, diarrhoea, emesis, alimentary haemorrhagy and contact dermatitis . Human ingestion of grain contaminated with F. graminearum has been associated with alimentary toxicity as well as illness characterized by nausea, vomiting, anorexia, and convulsions . Trichothecenes are also powerful modulators of human immune function and may promote neoplasms, cause autoimmune disease, or have long-term effects on resistance to infectious disease by altering immune response [24,25].
Several genes of Fusarium are involved in the biosynthesis of trichothecene and most of them are localized in a Tri gene cluster. The Tri-5 gene encodes the enzyme trichodiene synthase , which catalyzes the first step in the trichothecene biosynthetic pathway in trichothecene-producing strains of Fusarium species. The development of Tri-5 gene specific primers has allowed trichothecene-producing Fusarium spp. to be distinguished from nonproducing species using PCR-based assays . The nucleotides sequence of the Tri-5 gene has been characterized in several Fusarium species [28,29].
The main objectives of this study were detecting the presence of Tri-5 gene producing fungus, which encodes the key enzyme in trichothecene production, especially DON, by using conventional PCR; determining the correlation between the presence of Tri-5 gene and the DON content, which was analysed by the chemist Dr. Puttner. J. Lepschy; evaluating the amount of F. graminearum and F. culmorum infection through Real-Time PCR assay; investigating the relationship between DON content and the degree of F. graminearum and F. culmorum contamination and determining the aggressiveness of FHB towards plant host.
Fungal reference material
50 ng of extracted DNA from F. graminearum isolates were applied in tenfold serial dilutions (10-1 to 10-4) as a quantitative standard in Real-Time PCR (RT-PCR) using a F. graminearum specific Taqman® hybridization probe for beta-tubulin gene. In parallel, 50 ng of DNA of F. culmorum strains were used in tenfold serial dilution also as a quantitative standard for RT-PCR using SYBR Green®1.
At harvest time, 172 winter wheat ears samples of various cultivars have been randomly collected from farmers’ fields in different areas of Bavaria, South Germany. Directly after harvest, samples were sent to the Institute of Plant Protection, LfL and preserved at -20°C.
DNA of infected wheat kernels was extracted by homogenising 10 mg of dried kernels in a mixer with the presence of 1 ml DNA extraction buffer (2% CTAB, 1.4 M NaCl, 100 mM Tris, 20 mM Na- EDTA, and 1% PVP-40). The mixture was vortexed and the flow was transferred to microcentrifuge tubes. 1 ml chloroform/isoamylalkohol (24:1) was added, well mixed and spined at 5000×g for 10 minute at 20°C. The aqueous phase containing DNA molecules was transferred into 2 ml fresh tubes where 100 μl Na-acetat (3M, pH 5.2) and 1 ml isopropanol (-20°C) were added and mixed by inverting the tubes many times. Tubes were placed in a freezer (-20°C) for at least 1 hour. For each sample, the lysate mixture was transferred to SV Minicolumn placed in 1.5 ml tubes and spined at 16.000×g for 1 minute at 4°C. The supernatant was discarded, and the SV Minicolumns were washed with ethanol many times as described by (30). Finally, 50 μl of distilled sterile water was added directly into SV Minicolumn which was placed in 1.5 ml microcentrifuge tube and incubated at room temperature for 5 minutes then spined at 16.000×g for 2 minutes at 4°C to collect the eluted DNA .
Two Tri-5 specific primers have been used to detect the presence of Tri-5 gene in F. spp. infected wheat kernels. 172 wheat samples were tested with a sample of F. graminearum used as a positive control. The sequences of these primers are: forward primer Tri-5 F: (5’-AGCGACTACAGGCTTCCCTC-3’) and reverse primer Tri-5 R: (5’-AAACCATCCAGTTCTCCATCTG-3’). These primers were derived from the conserved region of Tri-5 gene in Fusarium spp. Tri-5 primers (Tr5F and Tr5R) amplified a single 544bp fragment in both DNA extracted of F. graminearum and F. culmorum and DNA of infected wheat grains. The total volume of reaction master mix was 22.7 μl. The PCR amplification was performed using (2.7 μl of 25 μg ml-1) of both fungal F. graminearum DNA and DNA from 10 mg dry weight of wheat material, 0.5 μM of each of the Tri-5-specific primers, 0.8 mM concentration of nucleotides dNTPs, 0.5 unit of Taq polymerase and 2.27 μl of PCR buffer with 1.5 mM MgCl2. The PCR negative control was a reaction master mix with 2.7 μl of distilled water instead of DNA template. Cycler programme was set as the following: one cycle at 95°C for 75 s then 32 cycles of 94°C for 20 s, 62°C for 17 s, 72°C for 45 s, and a final cycle at 72°C for 4 min and 15s. DNA banding were revealed by electrophoresis at 90 v on 2% agarose gels in 1x Tris-acetate EDTA buffer (TAE) (where 50X TAE contained: 2M Tris, 1M Acetic Acid, and 0.1M Na-EDTA x 2H2O at ?H 8.0) using ethidium bromide staining (30 μg of a ethidium bromide for 100 ml of 1X TAE buffer) and photographed under UV light using a camera and a photo print image visualizer.
Quantification of F. species by Real-Time PCR
Quantification of F. graminearum by RT-PCR using a TaqMan probe assay: Two primers, derived from the consensus beta-tubulin sequence which is associated with head blight in wheat, were used for F. graminearum quantification. The forward primer FGtubf: (5′-GTCTCGACAGCAATGGTGTT-3′) and reverse primer FGtubr: (5′-GCTTGTGTTTTTCGTGGCAGT-3′) specifically amplified a 111 bp fragment of the beta-tubulin gene of F. graminearum which was quantified by the TaqMan probe FGtubTM (FAM- 5′ACAACGGAACGGCACCTCTGAGCTCCAGC3′-TAMRA). PCR was monitored on a Real-Time 7000 Sequence Detection System. PCR Master Mix contained: Hot Start Taq DNA polymerase and PCR buffer specifically adopted for quantitative PCR analysis using species-specific probes. The total volume of master mix reaction (23 μl) contained: optimal primer concentrations 0.3 μM of FGtubf and FGtubr primers, 1 x PCR buffer, 50 ng of template wheat DNA samples, and 0.2 μM of TaqMan probe, and 50 ng of F. graminearum dilution template DNA as a standard curve. There were four series of diluted standard curves, with 1:10 fold of dilution factor of F. graminearum DNA. The number of cycles in the PCR was set at 40, as the 40th cycle represented the extrapolated threshold cycle for a reaction with a theoretical single copy of the template DNA. PCR program was as the following: 95°C for 15 min, 40 cycles of 95°C for 15 s and 67°C for 1 min. All reactions were performed in triplicates. PCR efficiency was calculated from threshold cycles of the standard dilution curve.
Quantitation of F. culmorum by Real-Time PCR using DNA binding dye assay: Two specific primers were used for the detection of F. culmorum by amplifying 140bp fragment of F. culmorum. Forward primer sequence Fc03: (5’-TTCTTGCTAGGGTTGAGGATG-3’) and reverse primer sequence Fc02: (5’-GACCTTGACTTTGAGCTTCTTG-3’) were specifically amplified a 140bp fragment of F. culmorum genome that was quantified by the DNA binding dye, SYBR® Green 1. The SYBR Green 1 assay is similar to that of TaqMan assay except the presence of an intercalating agent such as fluorescent dye SYBR Green 1 instead of fluorescent probes TaqMan.
Detection of Tri-5 gene
The Tri-5 specific PCR assay could provide a screening tool for detection of trichothecene-producing Fusarium species in plant tissues. In this study, 172 DNA samples of infected wheat kernels were analyzed using Tri-5 gene primers in a PCR reaction to detect trichothecene producing Fusarium species. The separation of PCR products on agarose gels showed that 86% of DNA samples possessed a unique fragment of 544bp representing part of the Tri-5 gene while 14% of samples didn’t show any amplification product (Figure 1).
The same wheat samples were used to estimate the content of DON in their tissues determined by HPLC (results provided by the chemist Dr. puttner J. Lepschy, LfL). The quantity of DON varied between the samples leading to regroup the samples into 3 categories according to their DON content.
The comparison between the amplification products produced with Tri-5 gene primers and the DON contents demonstrated that the samples free of Tri-5 gene products (14% of all samples) showed either absence of DON in their tissues (in 7% of all samples) or detected an amount of DON ranged from 11-226 μg kg-1 (in 7% of all samples).
For all samples possessing 544bp DNA fragments (86% of all samples), an amount of DON was detected in their tissues. Approximately 76% of them (59% of all samples) were considered DON high-producing strains (101-2990 μg kg-1) and possessing an intensive DNA band on agarose gels, while 14% of them (27% of all samples) were considered DON low-producing strains (10-99 μg kg-1) and showing a faint DNA band at 544bp.
Line 1, 16 standards 100 bp. Lines 2, 3 F. graminearum DNA used as positive control. Lines (4-15) some tested wheat samples: lines (5-7) free of DNA band, Lines (4, 8-11) faint bands, and lines (12-15) bright bands at 544bp.
Quantification of F. graminearum and F. culmorum using Real-Time PCR
The specific primers and TaqMan hybridization probe targeting the beta-tubulin gene amplifies DNAs from F. graminearum infected wheat. Based on determination of threshold cycle (Ct-values) in individual samples and known DNA standards during Real-Time PCR, amounts of target DNA present in the samples were calculated. The amplification of standard dilution curves of F. graminearum in Real- Time PCR gave linear and reliable results (R2 values were between 0.997 and 0.989). The concentration of F. graminearum DNA ranged from 0.04 to 4945 μg kg-1 (mean of triplicate samples ranged from 9.19E-07 to 1.29E-01 μg ml-1) of dry weight wheat kernels (Figure 2), while, F. culmorum DNA content was limited and ranged between 0.04 and 39.22 μg kg-1 (mean of triplicate samples ranged from 8.00E-07 to 7.13E-04 μg ml-1) of dry weight wheat kernels over all Bavaria (Figure 3).
By comparing the results of Tri-5 specific PCR assay, DON quantification by HPLC technique, and RT-PCR DNA quantification of F. graminearum and F. culmorum for the same wheat samples, we found that the same Tri-5 DNA intensive bright samples that associated with the largest amounts of DON were contained also the highest amount of F. graminearum DNA (4945-100.82 μg kg-1) and trace amounts of F. culmorum DNA (39.22 to 10.67 μg kg-1). Alternatively, faint Tri-5 DNA samples which had a trace amount of DON (10-99 μg kg-1) showed trace detectable amounts of both F. graminearum DNA ranged from 0.04 to 47.15 μg kg-1 and F. culmorum DNA ranged from 0.04 to 1.99 μg kg-1. In case of absence of Tri-5 gene products which associated with free DON contents (7% of all samples), there was trace amounts of both F. graminearum and F. culmorum DNA (0.72 and 0.78 μg kg-1, respectively). While in case of absence of Tri-5 gene products on agarose gels but with presence of DON in their tissues (the other 7%), there was trace detectable amounts of both F. graminearum DNA (0.66-1.83 μg kg-1) and F. culmorum DNA (0.08-39.22 μg kg-1).
Correlation between F. graminearum DNA and DON content
The plot of F. graminearum DNA content in 172 wheat samples, determined by Real-Time TaqMan probe PCR and the DON content in their tissues determined by HPLC, showed a strong positive linear correlation between both parameters. Correlation coefficient was 0.725 (Figure 4). Moreover, the regression analysis of all data sets indicated a strong and highly significant correlation (p < 0.05) between DON contents in the plant tissues and F. graminearum DNA contents in wheat samples, and the regression equation was (y=0.4896x + 79.784; R2=0.7252).
Correlation between F. culmorum DNA and DON content
The plot of DON content against F. culmorum DNA standard curve (Figure 5) showed a slight correlation between DON and F. culmorum DNA, whereas the linear correlation coefficient was ≈ 0.2 which is very far from +1. The regression analysis of all data set showed a weak correlation (p < 0.01) between F. culmorum DNA and DON content, and the regression equation was (y=8.9319x + 117.51; R2=0.0353).
Moreover, comparison of the results of TaqMan Real-Time PCR for F. graminearum with analysis of DON content for the same samples showed that 73% of F. graminearum presence was associated with DON production. On the contrary, the results of SYBR Green 1 Real- Time PCR for F. culmorum with DON content showed that 56% of F. culmorum incidence was associated with DON production. However, in spite of the noticeable infections of wheat grains with F. culmorum 56%, its DNA content was low (0.04- 39.22 μg kg-1) compared to that of F. graminearum (0.04 to 4945 μg kg-1 Figure 6).
On the other hand, although the F. graminearum DNA was not detected in 20% of samples, but DON was found (10-320 μg kg-1) and F. culmorum DNA was detected in some of these samples (2-39 μg kg-1) (Figure 7).
Severity of F. graminearum and F. culmorum infection over Bavaria
The whole picture of DON contents in 172 infected wheat samples was highlighted in Figures 3-8 in comparsion with the DON established threshold by FDA organization and European Union. This figure showed that only two samples of all harvested wheat kernel lay above the FDA DON advisory level of 1 ppm, and only one sample is higher than the EU threshold of 1.25 ppm.
Identification of trichothecene-producing F. spp. by PCR
Fusarium species are considered as a potential trichotheceneproducing species . The conserved region of Tri-5 gene has been detected in F. culmorum [30,31], F. graminearum , F. poae , F. sporotrichioides , and F. sambucinum . There was a direct relationship between Tri-5 gene expression and the increase in deoxynivalenol production [34,35]. Tri-5 primers were designed from highly conserved regions of the Tri-5 gene of Fusarium spp. . A Tri5-specific PCR assay has been developed to detect trichotheceneproducing Fusarium species in contaminated wheat samples . In our study, we used the qualitative Tri-5 specific PCR assay to detect trichothecene-producing Fusarium species in contaminated wheat kernel samples (172 samples) collected from south Germany. 59% (101 samples) of 172 tested samples were positive in the Tri-5 PCR assay and showed intensive bright DNA bands on agarose gels and were highly infected with one or more of Fusarium species containing Tri-5 gene. Correspondingly, this result was significantly associated with the results of DON content where the same Tri-5 DNA intensive bright samples had simultaneously the highest amount of DON ranged from 101 to 2990 μg kg-1. This result was in accordance with some reports on F. avenaceum isolates that were positive in the Tri-5 PCR assay and produced DON in culture . In addition, the results of F. graminearum and F. culmorum DNA quantification by RT-PCR showed that the same Tri-5 DNA intensive bright samples contained also the highest amount of F. graminearum DNA (4945-100.82 μg kg-1) and F. culmorum DNA (39.22-10.67 μg kg-1). Thus, according to our results we could say that there was a significant direct relationship between the density of Tri-5 DNA bands on agarose gels and DON content. The increasing in DON content was also associated with higher concentration of Fusarium species, particularly, of F. graminearum DNA and slightly with F. culmorum DNA quantified by RT-PCR. In other words, we could say the increased density of Tri-5 DNA bands with the highest amounts of DON were related to the presence of high amounts of F. graminearum DNA rather than F. culmorum. For example, the highest amounts of F. graminearum DNA in some samples (4945, 4255, 1069, 1050 μg kg-1) were associated with the largest amounts of DON (2990, 1174, 981, 728 μg kg-1) while the concentration of F. culmorum for the same samples were very limited (18.81, 0.30, 0.36, 0.00 μg kg-1 respectively Figure 6). On the other hand, faint Tri-5 DNA bands on agarose gels (27% of all tested samples) indicated samples containing a low concentration of Tri-5 gene and as a result, the infection with F.spp. was also low. Accordingly, the amount of DON content in these samples was very low and ranged from 10 to 99 μg kg-1. These results were in accordance with RT-PCR results, where low amounts of F. graminearum DNA (0.04-47.15 μg kg-1) and F. culmorum DNA (0.07-1.99 μg kg-1) were revealed.
Indeed, F.spp-infected wheat kernels that showed negative results in the Tri-5 PCR assay and showing absence of DNA bands on agarose gels should be Tri-5 gene free and should not be infected with F. spp. containing Tri-5 gene and ,accordingly, these samples were DON free. In our results, 7% (12 samples) of all samples were negative in the Tri- 5 PCR assay and didn’t produce DON and they were approximately free of F. graminearum DNA (0.72 μg kg-1) and F. culmorum DNA (0.78 μg kg-1). It has been demonstrated that within the same species and in the same cultural conditions toxin production by Fusarium strains may vary largely. Some strains produce large amount of trichothecene, whereas others produce small or undetectable amount of trichothecene [38-42]. However, there were other 7% of samples in which no Tri-5 DNA was detected on agarose gels but an amount of DON ranged from 11 to 226 μg kg-1, low detectable amounts of F. graminearum DNA ranged from 0.66 to 1.83 μg kg-1 and an amount of F. culmorum DNA (0.08-39.22 μg kg-1) were detected. It might be possible that other genes involved in trichothecene biosynthesis have been identified outside the Tri biosynthetic gene cluster including Tri1  and Tri101  which requires more investigation in our samples. However, it is possible that the pathogenic isolates producing DON in very small amounts could produce other phytotxins instead of DON in the pathogenesis . In our study, the same samples that produced Tri-5 DNA intensive bright bands on agarose gels were containing the highest amount of DON revealed by HPLC analysis and had also the largest amounts of F. graminearum and F. culmorum DNA evaluated by RT-PCR. We conclude from the displayed results that there was a positive relationship between the three techniques used in this study as they provided us with similar results for the same samples.
Quantification of F. graminearum and F. culmorum
Molecular diagnostic of plant pathogenic fungi can be highly specific, very sensitive, and relatively fast . We used in this study a fast and reliable method for the species-specific identification and absolute quantification of F. graminearum and F. culmorum. It is a RT-PCR assay using a TaqMan hybridization probe targeting the betatubulin gene for F. graminearum and SYBR Green 1 for F. culmorum. TaqMan method used in this study because of its sensitivity, selectivity, and reduction of fault signals due to primer-dimer formation  and allowed a fast species-specific identification and quantitation of plant infections by F. graminearum at very early stages where classical microbiological and toxin analysis methods fail to detect the pathogen . The beta-tubulin gene of all non F. graminearum isolates failed to be amplified in the reaction while targeting DNA from all isolates yielded product in the PCR assay . RT-PCR analysis confirmed that F. graminearum was more abundant in the infected grains than F. culmorum since the concentration of F. graminearum DNA ranged from 0.04 to 4945 μg kg-1 while F. culmorum DNA content ranged from 0.04 to 39.22 μg kg-1. Consequently, F. graminearum infections were severe while the severity of F. culmorum infection was not high in wheat kernels. Comparison between the results of TaqMan Real- Time PCR analysis for F. graminearum and DON content showed that F. graminearum is an efficient DON producer where there was high positive significant correlation (R2=0.7) between DON and F. graminearum DNA content. Therefore, F. graminearum was considered as the predominant trichothecene associated with FHB and produced the main part of DON in wheat crop. This is in accordance with former investigations . In contrast, the slight weak correlation (R2=0.03) between DON and F. culmorum DNA content may reflect that F. culmorum was the second important species in the DON producing Fusarium genus. Previous study suggested that F. culmorum along with F. graminearum were consistently the most pathogenic of the Fusarium species infecting cereal ears .
Moreover, the PCR analysis showed that in 20% of total infected samples, DON was found (10-320 μg kg-1) where no F. graminearum DNA was detected and that was linked with only slight content of F. culmorum DNA (0.10- 39.22 μg kg-1 Figure 7). PCR analysis indicated that the presence of other Fusarium species within the field plots may account for the FHB disease and this result was consistent with the observation of ref. . In these cases, DON content is probably attributable to the possibility that FHB infection in the samples is caused by a complex of Fusarium spp. which release DON mycotoxins, and other DON producing Fusarium spp. (like F. pseudograminearum, F. poae and/or F. sporotrichoides) might have been presented, that may require further investigation.
The U.S. Food and Drug Administration (FDA) recommend that DON levels in human foods should not exceed 1 ppm. Higher levels of DON are permitted in feed for poultry and ruminant animals. While the European Community supports the setting of European Union (EU) thresholds of trichothecenes as low as reasonably achievable in order to protect public health. For example, DON levels in human foods should not exceed 1.25 ppm. In general, the aggressiveness of F. graminearum and F. culmorum was relatively low over all Bavaria since the DON content was generally low (10-2990 μg kg-1 Figure 6). Indeed, the aggressiveness of Fusarium was not so high where only 2% of all harvested wheat kernel laid above the FDA DON advisory level of 1 ppm, and only one sample was higher than the EU threshold of 1.25 ppm. Some authors reported that trichothecenes may play an important role in the aggressiveness of fungi towards plant host .