Fumonisin B 1 and beauvericin accumulation in wheat kernels after seed-borne infection with Fusarium proliferatum

Zhiqing Guo1, Katharina Pfohl2, Petr Karlovsky2, Heinz-Wilhelm Dehne1, Boran Altincicek1 1 Institute of Crop Science and Resource Conservation (INRES-Phytomedicine), Rheinische Friedrich-Wilhelms-University of Bonn, Meckenheimer Allee 166a, D-53115 Bonn, Germany 2 Molecular Phytopathology and Mycotoxin Research, Georg-August-University Göttingen, Grisebachstrasse 6, D-37077 Göttingen, Germany e-mail: altincicek@hotmail.de


Introduction
Fusarium proliferatum is a member of Gibberella fujikuroi species complex which consists of eight mating populations (MPs) and a number of asexual lineages (Leslie et al. 2006), described as morphological species (Nirenberg and O'Donnell 1998).Most members of G. fujikuroi species complex are economically important plant pathogens.Decreased yield as well as diminished quality of plant products due to Fusarium infection cause significant economic losses worldwide (Placinta et al. 1999).F. proliferatum is a member of the complex with a wide host range.The pathogen was predominantly found in maize (Zea mays L.) but it also infects rice (Abbas et al. 1999) and sorghum (Bacon and Nelson 1994) and causes crown, spear and root rot of asparagus (von Bargen et al. 2009), bulb rot in garlic and onion (Stankovic et al. 2007), and fruit rot in date palms (Abdalla et al. 2000).
Grains infected with F. proliferatum accumulate mycotoxins.F. proliferatum and F. verticillioides are the main source contamination with fumonisins, which are toxic and carcinogenic.Fumonisin B 1 is the most abundant fumonisin found in maize.Fumonisins also can be detected in symptomless infected maize kernels (Bacon and Hinton 1996).These mycotoxins inhibit sphinganine/sphingosine N-acyltransferase (ceramide synthase), a key enzyme in the sphingolipid metabolism and triggers apoptosis (Riley et al. 1996).High levels of fumonisins in maize were associated with outbreaks of equine leukoencephalomalacia (Wilson et al. 1990) and swine pulmonary edema (Osweiler et al. 1992).Association between fumonisin exposure and esophageal cancer was reported in South Africa (Rheeder et al. 1992).Besides fumonisins, F. proliferatum was shown to synthesize moniliformin (Marasas et al. 1986), beauvericin (Moretti et al. 1994) and fusaproliferin in maize kernel cultures (Ritieni et al. 1995), and fusarins in mineral medium consisting of ultrapure water (1 l), (NH 4 ) 2 HPO 4 (1 g), KH 2 PO 4 (3 g), MgSO 4 •7H 2 O (0.2 g), NaCl (5 g), sucrose (40 g) and glycerol (10 g; final pH 6.2), which was also satisfactory for moniliformin production (Miller et al. 1995).Besides F. proliferatum, other causal agent pathogens of Fusarium head blight synthesize moniliformin, beauvericin, fusarins, and fusaproliferin in grains.Moniliformin is produced by F. culmorum in rice culture (Scott et al. 1987), by F. avenaceum and F. oxysporum in parboiled rice culture (Abbas et al. 1989).Beauvericin is produced by F. avenaceum in wheat (Logrieco et al. 2002) and by F. poae, F. oxysporum (Logrieco et al. 1998) and F. verticillioides (Leslie et al. 2004) in maize kernels.Fusarin C is produced by major causal agents of Fusarium head blight in maize grains, and synthesized by F. avenaceum in parboiled rice culture (Abbas et al. 1989).F. verticillioides can produce trace levels of fusaproliferin in autoclaved yellow maize kernels (Leslie et al. 2004).Fusaproliferin is not produced by any Fusarium species that has been found in wheat except for F. proliferatum.
Systemic infection of host plants has been reported for several Fusarium species.The entire maize plant can be systemically colonized by F. verticillioides without symptoms.In this way, F. verticillioides can be transmitted via seeds (Munkvold et al. 1997).Bacon and Hinton (1996) reported that maize kernels infected with F. verticillioides (F.moniliforme in their article) served the pathogen as a dissemination vehicle.The fungus grew intercellularly without visual signs of diseases through the whole plant, whereas both intercellular and intracellular hyphae were found in plants with disease symptoms.Plants other than maize can be systemically colonized with F. verticillioides, too (Dastjerdi and Karlovsky 2015).
Infection of wheat kernels with F. proliferatum was reported for the first time from Canada, where the pathogen was found as a rare contaminant of tombstone kernels (Clear and Patrick 1990).In 1992, F. proliferatum was isolated from infected wheat seeds in Mazandaran (Zamani-Zadeh and Forutan 1992).The following study from Canada identified F. proliferatum as the causal agent of black point symptom in wheat (Conner et al. 1996).Spontaneous infection of wheat plants grown in a space shuttle with F. proliferatum (Bishop et al. 1997) motivated a series of physiological studies with little relevance for field pathology but the same laboratory identified F. proliferatum as an opportunistic pathogen of wheat under stress conditions (Kwon et al. 2001).A survey of fumonisins determination in cereals in Spain reported fumonisins in 8 of 17 samples and F. proliferatum infecting 2 of 17 samples naturally (Castella et al. 1999).Field pathologists became interested in F. proliferatum in wheat after contamination of wheat kernels with fumonisins was reported from Italy (Cirillo et al. 2003).In spite of a growing number of studies of F. proliferatum in wheat in recent years (Desjardins et al. 2007, Palacios et al. 2011, Busman et al. 2012, Amato et al. 2015), our understanding of this pathosystem is still limited.
The objective of present study was to assess the ability of F. proliferatum to colonize wheat plants systemically and to compare aggressiveness of strains isolated from different hosts towards wheat.

Fungal strains and wheat plants
Sixteen single spore strains of F. proliferatum isolated from garlic, onion, asparagus, maize, silver grass and a dead larva of Zeuzera pyrina in Germany, France, Syria and Austria were used (Table 1).The strains were taxonomically assigned to F. proliferatum based on micro-morphological features on carnation leaf agar (CLA) (Leslie et al. 2006) and the taxonomical identification was confirmed by species-specific PCR with primers PRO1 and PRO2 (Mulè et al. 2004).

Preparation of inocula
Mung bean medium according to Bai and Shaner (1996) in 50 ml portions in 100 ml Erlenmeyer flasks was autoclaved at 121 °C for 20 min.Three discs of 5 mm diameter from potato dextrose agar (PDA) plates overgrown with F. proliferatum were inoculated into each flask and incubated at 25 °C without light for 4 days with shaking at 150 rpm.Using four-layer cheese cloth for filtering the mycelia, conidia were collected, enumerated in a haemocytometer cell and stored in -80 °C.

Inoculation of wheat plants
Wheat seeds were surface-disinfested for 3 min in 1.2% sodium hypochlorite and rinsed three times with sterile water.50 wheat kernels were incubated in 10 ml of a conidial suspension (10 4 spores ml -1 ) of each F. proliferatum strain for 48 h on a shaker.Inoculated seeds were sowed in twice-autoclaved soil in pots (10 kernels per pot, five replicates per strain).Un-inoculated controls were treated with sterile water.The experiment was carried out twice.
In a greenhouse, the temperature cycled between 26 °C (from 0600 to 2200) and 22 °C (from 2200 to 0600); light of intensity 4000 to 5000 lux was maintaining for 16 h; relative humidity was 50%.Plants in pots were watered every other day till the kernels were ripe.After two more weeks without watering, ears, first flag leaves, and the internodes between last node and ear were collected.

Re-isolation of fungi from wheat plants
The stems (between the ear and the last node of the inoculated plants which were fully asymptomatic) were cut into 2-3 cm sections, surface sterilized with 1.2% sodium hypochlorite for 3 min and rinsed three times in autoclaved distilled water.Stem sections were placed on Czapek-Dox-Iprodione-Dichloran (CZID) medium (Abildgren et al. 1987) and incubated at 25 °C.Un-inoculated controls were handled in the same way.Fungal growth was observed after 3-4 days.Taxonomic identity of the isolates was confirmed by PCR with primers PRO1/PRO2 specific for F. proliferatum (Mulè et al. 2004).
Quantification of fungal DNA in wheat stems, flag leaves and kernels Wheat kernels, flag leaves and stems were dried for 48 h at 45 °C and ground.Total DNA was extracted from 50 mg using a CTAB method (Brandfass and Karlovsky 2008).The pellet was dissolved in 50 µl TE buffer (10 mM Tris/HCl, 1 mM EDTA, pH 8.0).DNA solution was diluted fifty fold prior to PCR analysis.Inhibition assay was carried out to confirm the absence of matrix effects on PCR at this dilution.F. proliferatum DNA in plant tissue was quantified as described previously (Nutz et al. 2011).The lowest standard set as a limit of quantification corresponded to DNA content of 0.935 ng g -1 .
Mycotoxin analysis in kernels 100 mg of ground kernels were extracted with acetonitrile/water (84/16, v/v) as described (Adejumo et al. 2007); samples for beauvericin analysis were not defatted.HPLC separation and MS/MS detection was performed essentially as described (Adejumo et al. 2007) but ion trap 500-MS (Varian, Darmstadt, Germany) was used as a detector.Beauvericin was quantified in a positive ionization mode as a sodium adduct with mass transitions m/z 806 > 645 (quantification), m/z 806 > 545 and m/z 806 > 384 (confirmation).Calibration curves were constructed using analytical standards dissolved in methanol/water (1:1, v/v) with a correction for recovery and matrix effects.The limit of quantification for both beauvericin and fumonisin B 1 was 25 ng g -1 .

Statistical Analyses
Statistical analyses were performed with SPSS 22.0 (IBM SPSS statistics, USA).The Pearson correlation coefficient (2-tailed) between mycotoxins (fumonisin B 1 and beauvericin) production and fungal DNA was calculated.

F. proliferatum strains originating from different plants were capable of infecting wheat systemically
We inoculated wheat seeds with conidia of F. proliferatum and sowed them in soil.The germination rate dropped from 94% for un-inoculated seeds to 84-92% for seeds inoculated with F. proliferatum.No disease symptoms were observed in the inoculated plants.To determine whether plants inoculated with F. proliferatum were colonized systemically, we attempted to re-isolate the pathogen and detect its DNA in stems, flag leaves and harvested kernels.Mycelia growing out from stems of infected wheat plants (Fig. 1a) were confirmed to belong to F. proliferatum by species-specific PCR (Fig. 1b).No mycelia grew out from surface-sterilized stems of un-inoculated plants.
The presence of F. proliferatum DNA in stems, flag leaves and harvested kernels corroborated systemic colonization (Fig. 2).F. proliferatum strains originating from different hosts successfully colonized wheat plants; the extent of colonization among strains from the same host plant differed.Strain F. proliferatum 29-M isolated from maize accumulated the highest amount of fungal DNA both in wheat stems and kernels.The amount of fungal DNA in kernels was lower than in stems and leaves (Fig. 2).The experiment was carried out for the second time with similar results (Fig. 3).

Fumonisin B 1 and beauvericin accumulation
Mycotoxins fumonisin B 1 and beauvericin were found in 1/3 of infected wheat kernels (Fig. 4).The strains causing mycotoxin accumulation originated from different host plants.
We found a strong positive correlation (r = 0.817) between DNA amount and fumonisin B 1 concentration in wheat kernels.A strong correlation (r = 0.871) was also found between F. proliferatum DNA and beauvericin content (Fig. 5).Log 10 DNA ng g -1 replicate 1

Discussion
F. proliferatum strains isolated from all host plants tested (garlic, onions, asparagus, maize, silver grass and dead larvae) were able to systemically infect wheat.The aggressiveness of the strains differed to a large extent.The limited number of strains used did not allow us to conclusively attribute these differences to the origin of the strains but low amount of biomass accumulated by all seven strains isolated from asparagus indicated that the specialization of F. proliferatum to asparagus may be accompanied by a decline of aggressiveness towards wheat.This might also be the case with onion and silver grass; more strains have to be tested to corroborate this hypothesis.
Only F. proliferatum strains that accumulated the largest amounts of DNA in kernels caused detectable contamination of kernels with mycotoxins (Fig. 3).Fumonisin B 1 was only detectable in kernels infected with three F. proliferatum strains out of sixteen strains tested.Interestingly, these three strains originated from different host plants: maize, asparagus, and garlic.Infection of wheat with two other strains resulted in the accumulation of beauvericin but not fumonisin B 1 .The amounts of fumonisin B 1 in infected kernels were much lower than the levels commonly found in maize, supporting the hypothesis of Busman et al. (2012) that wheat kernels are a less favorable substrate for the production of fumonisins than maize.Fumonisin levels in naturally infected wheat were low, too (Desjardins et al. 2007, Palacios et al. 2011, Amato et al. 2015).
F. proliferatum DNA was found in largest amounts in stems, less in leaves and the lowest amounts in kernels.We speculate that F. proliferatum spreads within pith parenchyma (including the vascular system), which offers abundant nutrients and humidity, as shown for F. graminearum and F. culmorum (Guenther andTrail 2005, Mudge et al. 2006).Tight correlation between the amount of F. proliferatum DNA and mycotoxin concentration in kernels (Fig. 5) indicated that mycotoxins were produced in kernels rather than being transported to kernels from the stem with the transpiration stream.This speculation is in line with the results of Winter et al. (2013) on deoxynivalenol, which was produced in large amounts in stems of wheat plants infected with F. culmorum but did not reach kernels due to an anatomical barrier between grain and rachilla designated as "xylem discontinuity".Analysis of fumonisin and beauvericin content in stems of wheat plants infected with F. proliferatum via seeds will be needed to test this hypothesis.We have not observed any symptoms on the inoculated plants except that the leaves near the first node were more chlorotic as compared to control plants.Desjardins et al. (2007) reported kernel black points in wheat inoculated with F. proliferatum but they injected wheat spikelets at early anthesis while we studied seed-borne infection.
Accumulation of mycotoxins in symptomless wheat grain used for human and animal consumption poses a risk to human and animal health.Furthermore, residues of infected wheat plants can serve as source of inoculum of F. proliferatum for susceptible crops in the next season.Systemic infection of wheat with F. proliferatum therefore deserves attention, particularly in crop production systems in which seed coating with fungicides is not practiced.

Fig. 2 .
Fig.2.Fungal DNA of F. proliferatum strains in wheat stems (the internode between the last node and ear), the first flag leaves, and kernels was quantified by real time PCR

Fig. 4 .
Fig. 4. Fumonisin B 1 and beauvericin quantification in colonized wheat kernels infected by different tested F. proliferatum strains with soak-inoculation mature seeds method

Fig. 5 .
Fig. 5. Correlation of mycotoxin accumulation and amount of F. proliferatum DNA in wheat kernels
a Original host is specified by the last character of strain label; G = Garlic; S = Silver grass; O = Onion; M = Maize; A = Asparagus; L = insect larva (dead)