Following inoculation of the base of soft wheat seedlings with Fusarium culmorum, disease symptoms typical of crown rot developed at the stem base and extended up to the third node by plant maturity. Fungus was isolated from all tissues exhibiting symptoms but not from symptomless tissues. Histopathological analysis revealed that the fungus was present mainly in the parenchymatic cells of the stem base and colonized the tissues via apoplastic and symplastic pathways. Host response in advance of pathogen colonization was observed. At maturity, plants were divided into sections from the inoculated area to the head. Heads were also separated into grain, rachis and chaff components. Colonization by the fungus was assessed by isolation from surface-sterilized segments and quantified by real-time PCR. Disease symptoms and the fungus were detected up to the third node, while deoxynivalenol (DON) was present in all stem segments and heads. Within the head, the DON concentration was higher in the rachis than in the chaff and grain components. These results demonstrate that F. culmorum can extensively colonize stem tissues but not reach the head by the time of plant maturity. In contrast, DON was detected in tissues beyond those colonized by the fungus, translocating to the head where, although accumulating mainly in the rachis, significant quantities accumulated in the grain. These findings indicate that there is a potential threat of contamination of grain with DON where severe crown rot is present in a crop.
Fusarium culmorum, F. graminearum and other Fusarium species, such as F. pseudograminearum (Dyer et al., 2009) are the principal cause of crown rot (CR), producing necrotic lesions in these tissues. CR is a serious problem for wheat production in many parts of the world (Hogg et al., 2007), in particular where dry climatic conditions are present and conservation agricultural practices are used (Smiley et al., 2005). Although CR can cause very high yield losses in all the wheat growing areas of the world (Chakraborty et al., 2006), no highly CR-resistant varieties of wheat are yet available and cultural management strategies are only partially effective and are not reliable for controlling damage caused by CR in those regions where it is a chronic disease (Paulitz et al., 2002).
Crown rot infection generally occurs in the region of the emerging shoot, crown and stem base from an inoculum of conidia and mycelium on residual stubble (Southwell et al., 2003). The histopathology of the initial infection process of F. culmorum on wheat seedlings showed that the pathogen sequentially penetrated the stem base leaf sheaths. In response to this the host reaction is thought to include systemic signalling, resulting in the expression of defence-associated genes in leaf sheaths in advance of those infected by the fungus (Beccari et al., 2011). A study conducted in the related species F. graminearum by Stephens et al. (2008) showed that CR disease development involves distinct phases of colonization associated with well defined fungal gene expression.
Once Fusarium spp. infection is established in the stem base and crown, several reports suggest that fungal systemic colonization by F. culmorum and/or F. graminearum can occur into distal parts of the stem. While many studies conclude that neither species can colonize heads by this route (Purss, 1971; Burgess et al., 1975; Snijders, 1990; Clement & Parry, 1998; Gèlisse et al., 2006), others have reported that the pathogens can colonize as far as, and including, the head (Mudge et al., 2006; Poels et al., 2006). One of the most complete studies was conducted by Mudge et al. (2006) who isolated F. graminearum and F. pseudograminearum from flag leaf node and head tissue following inoculation of the stem base. Fusarium graminearum colonized the central stem region in association with the pith parenchyma by intercellular growth, and hyphae were also observed in the lumen. Very little colonization of the vascular tissues occurred at the infection site and no hyphae were observed in the vascular bundles in the non-inoculated upper internodes (Mudge et al., 2006).
In addition, F. culmorum, together with F. graminearum, are the main pathogens responsible for fusarium head blight (FHB), a globally important wheat disease. It is well known that infection of wheat heads by F. culmorum reduces grain quality through the production of mycotoxins such as the trichothecene deoxynivalenol (DON), making grain unsafe for human and livestock consumption (Champeil et al., 2004). Epidemics of FHB caused by F. culmorum have been reported in many countries, but these are usually localized to regions that have a high incidence of rainfall during anthesis and grain filling and where maize or sorghum have been grown in rotation with wheat (Southwell et al., 2003). Crop management and agrochemical measures are only partially effective in controlling FHB and, similar to the situation for CR, fully resistant varieties are not yet available (Buerstmayr et al., 2009).
Fusarium culmorum, along with F. graminearum and F. pseudograminearum, produce the mycotoxin DON in planta when inciting FHB disease. The biosynthesis of trichothecene mycotoxins is determined by multiple enzymes, many of which are encoded on a gene cluster. One of the most intensively studied of these genes is TRI5 that encodes trichodiene synthase, the enzyme responsible for the first committed step in the synthesis of trichothecenes (McCormick, 2003). For F. graminearum, DON, and thus TRI5, contribute to the aggressiveness of the fungus in FHB (Jansen et al., 2005).
The production and role of DON in CR during the colonization of stem tissue by F. graminearum and F. pseudograminearum have been reported by Mudge et al. (2006). The authors showed that expression of TRI5 and synthesis of DON were induced during infection of the stem base but that DON production was not required to cause CR symptoms. However, DON was demonstrated to have a role in stem colonization by the fungus with the DON-producing wildtype strain being recovered from grain at a higher rate than the DON-minus strain following inoculation of the stem base. These findings are supported by the observations of others (Poels et al., 2006; Moretti, 2008) who also detected DON within the head following seed inoculation with F. graminearum.
Evidence suggests that trichothecenes can be translocated in the plant tissues in advance of fungal growth. For example, in maize, DON appeared in an area of the stalk in the absence of fungus (Young & Miller, 1985). Snijders & Kretching (1992) concluded that DON was translocated from chaff to young kernels prior to F. culmorum colonization but provided no evidence for this. In addition, Kang & Buchenauer (2002) detected DON in wheat heads a few cells in advance of the F. culmorum hyphae. For this reason, in addition to the production of cell-wall degrading enzymes, the secretion of trichothecene mycotoxins constitutes an important factor in FHB development by F. culmorum (Kang & Buchenauer, 2002).
In adult plants, DON produced during infection of heads (FHB) seems to circulate in the phloem, with the concentration following a descending gradient from the rachis, through lemmas and grains to the peduncle. Photosynthates from glumes and lemma are translocated preferentially to the nearest kernel via the phloem as a bulk flow of solution. Because DON is water-soluble it may also be distributed throughout the chaff and kernel in the same fashion (Sinha & Savard, 1997). In addition, the flower parts of wheat heads, the rachis and the peduncle below the point of infection, contain larger amounts of DON than the region above it (Savard et al., 2000).
To the authors’ knowledge, there are no reports on the potential for systemic translocation of DON to the head following infection of the crown or stem base by F. culmorum. It is important to determine the degree of systemic fungal colonization and/or DON movement from stem base to determine whether CR caused by F. culmorum could be an additional source of contamination of grain with DON, with potential repercussions for risk assessments of mycotoxin contamination of grain in the absence of FHB and on the management of the cultivation practices.
This study reports on aspects of symptom development, fungal colonization, DON translocation and DON distribution within the head following inoculation of the stem base of wheat seedlings by F. culmorum.
Materials and methods
Plant and fungal material
Surface-sterilized seed of Genio, a commercial cultivar of hexaploid wheat (Triticum aestivum) susceptible to CR, was used in all experiments. Two virulent strains of F. culmorum DON-producing chemotypes, Fu5 and GFP1, the latter constitutively expressing green fluorescent protein, were cultured at 20°C on half-strength V8 juice agar in 9 cm Petri dishes for producing mycelial inoculum.
Stem base material and inoculation procedure
Surface-sterilized seed (one per pot) were grown in 12 cm2 pots filled with sterilized substrate (1:1 peat:sand). Pots were grown at 22°C under 16 h of photoperiod and a 3 cm long PVC collar (3 mm internal diameter) was placed over each emerging coleoptile. Collars were split along their entire length to allow expansion with growth of the plant. When the second leaf was fully expanded, seedlings were inoculated as described by Simpson et al. (2000). In brief, 7-day-old colonies of the fungus grown on half-strength V8 agar were macerated to obtain a homogenate. Plants were inoculated by placing the homogenate (350 μL) into the space between seedlings and PVC collars. Half-strength V8 agar macerate was used as a control treatment for uninoculated plants.
In all experiments, at harvest, the main tillers were divided into four segments: C-2, inoculated area between crown and 2nd node; 2–3, proximal region, between 2nd and 3rd node; 4-H, last internode, between 4th node and head; and H, head.
Experiment 1: Histological analysis
Twenty-one randomly distributed plants were sampled 50 days post-inoculation (dpi) of the stem base with the GFP1 isolate of F. culmorum and CR visual disease symptoms were assessed along the length of the stem of each wheat plant. The percentages of plants showing symptoms were assessed in each segment. Subsequently the plants were observed under bright field stereomicroscopy and confocal laser scanning microscopy (CLSM) to examine fungal colonization in the stem base. Whole, longitudinal or horizontal sectioned wheat stem bases were initially observed under a stereomicroscope and then the material transferred to a Zeiss LSM 510 META CLSM using ×10 water immersion or ×20 oil immersion objective lenses. Spectral data were collected by excitation with 488 nm argon laser and using a GFP specific filter (505–555 nm) and autofluorescence detection wavelength of 590–660 nm. The same procedure was used for assessing uninoculated samples.
Experiment 2: Assessment of colonization by fungal isolation and DON translocation
At maturity (90 dpi with Fu5 strain), after symptom observation as described above, 21 randomly distributed plants were sampled and segments of 5 cm were removed from stem regions as previously described. Segments were surface sterilized with an ethanol/sodium hypochlorite solution (82:10:8 sterile deionized H2O:EtOH 96%:NaClO 8% v/v) for at least 1 min, rinsed, air-dried, then longitudinally cut in half and placed cut side down onto PDA (potato dextrose agar) amended with 14 mL L−1 10% solution of tartaric acid to inhibit bacterial growth. Plated tissues were incubated at 20°C for 7 days and the frequency (%) of recovery of the fungus recorded. The same procedure was also adopted for uninoculated control plants.
For analysis of DON from the stem base to the head, 36 randomly distributed plants at 90 dpi, after symptom observation, were divided into three groups (three replications) of 12 plants each and each plant divided into four segments (5 cm) as described in Experiment 1. All tissues were finely ground by Ultraturrax (IKA). DON was extracted by the addition of water (1:4 w/v) to the samples and vigorously shaken for 3 min. After centrifugation at 600 g for 3 min, the supernatant was passed through filter paper (Whatman No. 1). DON was quantified in the filtrates using an Agraquant DON assay 0·25/5·0 ELISA kit (Romer Labs) according to the manufacturer’s directions, and the analysis was repeated twice. The same procedure was adopted for the uninoculated control plants. t-tests were used to reveal statistically significant differences (P ≤0·05).
Experiments 3 and 4: Detection and quantification of DON and F. culmorum by real-time PCR
At maturity (90 dpi) 30 randomly distributed plants, after symptom observation, were divided into three groups of 10 plants each (three replications) and divided into sections as previously described. This material was immediately freeze-dried and then finely ground under liquid nitrogen. Each sample was divided into two, and one subsample was used for real-time PCR and the other for DON analysis. Real-time PCR analysis was carried out as described previously (Beccari et al., 2011) using specific primers for detection and quantification of F. culmorum (C51ENDF 5′-AACTGAATTGATCGCAAGC-3′ and C51ENDR 5′-CCCTTCTTACGCCAATCTC-3′) and elongation factor 1 (EF1α) primers for quantification of wheat DNA (Beccari et al., 2011). Fusarium culmorum quantification limits ranged from 0·04 to 40 ng of fungal DNA. Because of the non-independence of mean and variance, data relative to these experiments were log10-transformed and t-tests were performed to reveal statistically significant differences (P ≤0·05) between samples. The whole analysis was conducted in two different experiments.
Experiment 5: DON measurements in head tissues
For analysis of DON in head tissues, 30 randomly distributed plants, grown as part of Experiment 1, were sampled at 90 dpi, divided into three groups of 10 plants each (three replications), and three heads per plant (the principal and two secondary tillers) were collected, immediately freeze-dried, and separated into rachis, chaff and grain components. The analyses were conducted as previously described for Experiment 2. The same procedure was adopted for the uninoculated control plants. t-tests were used to reveal statistically significant differences (P ≤0·05).
Experiment 1: Microscopic observation of the colonization process on the wheat stem base
To observe histopathological aspects of the F. culmorum development on the wheat stem base during CR, stereomicroscope and CLSM observations were conducted. At 50 dpi all the inoculated plants showed CR symptoms in the C-2 segment, while only 10% of the plants showed symptoms in segment 2–3 (Fig. 1a). No symptoms were observed in the 4-H or H segments or in the uninoculated controls.
CLSM observations (50 dpi) permitted a preliminary visualization of the fungal presence in the plants. Horizontal sections of each wheat plant segment were assessed and the fungus was detected in C-2 in 80% of the observed plants (Fig. 1b). In a significant proportion of plants, no fungus could be observed in segments exhibiting necrotic symptoms (20% of plants in C-2 and 2–3 segments). No symptoms were observed in 4-H or H segments and no fungus was detected in these segments (Fig. 1). Similarly, no symptoms were observed, nor any fungus detected, in any control plants.
Longitudinal sections of the C-2 segments exhibited the typical CR brown discoloration in the external part of the node and of the stem wall surrounding the pith cavity or the node constriction (Fig. 2a). CLSM observations of this area revealed the presence of hyphae in the external edge of the node constriction (Fig. 2b). The horizontal sections of the C-2 segments revealed the presence of brown, discoloured tissue in the external area of the stem base including the epidermal cells and the first layers of the parenchyma cells (Fig. 2c). In this area, CLSM analysis confirmed the presence of fungal hyphae (Fig. 2d) with fungal colonization that proceeded via intracellular (Fig. 2e) and extracellular (Fig. 2f) growth. In some C-2 and 2–3 sections, at the front of symptom development (Fig. 2g), CLSM could not detect any fungus in either the epidermis or first cell layers of the parenchyma tissues exhibiting typical brown discoloration (Fig. 2h). No symptoms were observed in any of the 4-H and H segments or any of the uninoculated control samples. Similarly, CLSM analysis of these tissues did not reveal the presence of the fungus.
Experiment 2: Fungal colonization of wheat and DON production in distal non-inoculated tissues following stem base inoculation
The above experiment indicated that symptom development occurred slightly ahead of fungal colonization. The second experiment was designed to further investigate this association and establish whether this was mirrored by the distribution of DON in tissues following inoculation of the stem base with F. culmorum. Under the present conditions, isolate Fu5 of F. culmorum was able to induce CR symptoms in soft wheat cv. Genio. In fact, by 90 dpi of the stem base, all the plants showed the typical CR necrotic lesions on both the stem base and the crown up to the second node, extending approximately up to the third node in 35% of plants (Fig. 3a). No symptoms were present in the 4-H segments, in the heads or in any of the uninoculated plants.
The pathogen was reisolated from the C-2 (100%) and 2–3 (60%) but not from the 4-H or H segments of the inoculated plants (Fig. 3b). No F. culmorum colonies were isolated from any part of the uninoculated plants.
DON was detected in all the segments of inoculated plants (Fig. 3c) including those segments that exhibited no symptoms or fungal presence. Segments C-2 and 2–3 had the highest DON accumulation, with averages of 64 and 84 mg kg−1, respectively. The amount of DON in segment 4-H was significantly lower than that in the segment immediately below (2–3) (12 mg kg−1) but higher than that in the head (1·7 mg kg−1). The presence of DON in the 4-H and H sections indicated that the mycotoxin translocated to distal tissues ahead of fungal colonization.
Experiments 3 and 4: Fungal quantification and DON distribution along wheat plants following CR development
Experiment 2 indicated that DON could translocate ahead of fungal colonization, as assessed by isolation of the fungus from stem sections. The highly sensitive technique of real-time PCR was used to exclude the possibility that low levels of fungal colonization had occurred despite the inability to isolate fungus from the tissues. The real-time PCR quantification of F. culmorum measured the distribution of the fungus in the non-inoculated stem and head tissues following inoculation of the stem base. The pathogen was detected in the C-2 and 2–3 sections (3·57 and 1·57 ng of fungal DNA per 100 ng of total DNA, respectively) but not from the 4-H or H sections of the inoculated plants (Fig. 4a). No F. culmorum was detected in any part of the uninoculated plants.
The DON analysis conducted on tissues coming from the same material as used for real-time PCR revealed that DON was present in all tissues of the plant (Fig. 4c) with a decreasing concentration up the tiller of 32·2, 4, 3·1 and 2·5 mg kg−1 in sections C-2, 2–3, 4-H and H, respectively. The accumulation of DON in C-2 was significantly higher than that in H sections. No DON was detected in any parts of the uninoculated plants.
The repeated experiment (Experiment 4) revealed a very similar pattern of both F. culmorum biomass and DON distribution by 90 dpi. The pathogen was detected in the C-2 and 2–3 (3·63 and 1·62 ng of fungal DNA per 100 ng of total DNA, respectively) but not from the 4-H and H segments of the inoculated plants (Fig. 4b). DON was present in all tiller tissues of the plant (Fig. 4d) with average concentrations of 12·6, 4·2, 2·6 and 1·2 mg kg−1 in sections C-2, 2–3, 4-H and H, respectively. No F. culmorum or DON was detected in any part of the uninoculated plants.
Experiment 5: DON distribution in the head components following stem base infection
The above experiments revealed that DON translocated ahead of fungal colonization of the stem to reach the head. This trial was designed to assess the distribution of DON within the head components of rachis, chaff and grain. Although no fungus was detected in the distal parts of the plants, at 90 dpi DON was present in the head, with a non-homogeneous distribution in the different tissues (Fig. 5). DON concentrations were highest in the rachis (18·5 mg kg−1) and significantly lower in the chaff (4·5 mg kg−1) and in the grain (1·2 mg kg−1).
This study investigated symptom development, pathogen colonization and translocation of DON through stem tissues following infection of the stem base of wheat by F. culmorum.
The colonization process of F. culmorum in the stem base of wheat seedlings initiates by progression of the fungus through the layers of the stem base leaf sheaths towards the stem (Beccari et al., 2011). The present study indicates that the pathogen enters the stem via the point of attachment of the leaf sheath to the stem base (Fig. 2a,b). No direct penetration from the leaf sheath to the lignified stem base was observed. A similar pathway of stem colonization has also been described for F. graminearum by Stephens et al. (2008).
Once inside the parenchyma tissue of the stem, F. culmorum colonized the tissues by both intercellular and intracellular growth (Fig. 2d–f). The former mode of colonization provides a possible means for the systemic growth of the fungus through the stem, with intercellular colonization taking place behind the advancing hyphal front. This hypothesis is supported by the absence of F. culmorum fungal structures in the vascular bundles in the inoculated or non-inoculated areas, as has also been observed for F. graminearum (Mudge et al., 2006).
Fusarium culmorum colonization of the stem was determined by isolation onto a nutrient substrate. This method employed natural amplification (growth) to detect the fungal presence and was used successfully by Mudge et al. (2006) who detected the presence of F. graminearum and F. pseudograminearum in wheat tillers following stem base inoculation with these two pathogens. In contrast, in the present study F. culmorum was isolated up to the third node but not from the last internode or from head tissues.
In addition, the highly sensitive amplification technique of real-time PCR was used to detect F. culmorum and quantify fungal biomass in tissues. This method confirmed that the fungus, following stem base inoculation, colonized the stem and reached the central portion of the tiller but did not reach the head, supporting the findings from fungal isolation. These results are broadly in accordance with previous results obtained for the same pathogen by other authors (Purss, 1971; Snijders, 1990; Clement & Parry, 1998). The results here confirm that F. culmorum can systemically grow along the stem, leading to infection in higher stem internodes. However, like the previous reports, no evidence was found that systemic fungal growth could lead to infection of the heads (Purss, 1971; Snijders, 1990; Clement & Parry, 1998). These findings contrast with those for the related species F. graminearum and F. pseudograminearum where the fungi were shown to colonize the flag leaf node and heads following stem base inoculation (Mudge et al., 2006). While differences in environmental factors and relative aggressiveness are most probably the basis for the observed differences in the ability of these species to reach the heads, it is conceivable that this reflects underlying differences in the rate of stem colonization by these species. Additional comparative studies are required to investigate this possibility in more detail.
Previous results revealed the presence of dark brown tissue discoloration in the wheat stem base leaf sheaths in the absence of the fungus and in advance of fungal colonization by F. culmorum (Beccari et al., 2011). Similarly, the present study observed the presence of typical CR brown discoloration of stem tiller tissues ahead of fungal colonization indicating that the host is responding in advance of the presence of the fungus (Fig. 2g,h). The nature of the components responsible for triggering this response is not known, but DON may be involved in this process as it has been shown to induce reactive oxygen species (ROS) production and programmed cell death (PCD) in wheat (Desmond et al., 2008). Further investigations are required to better define the role of DON in the tissues showing symptoms ahead of the fungus. The appearance of necrosis ahead of colonization by F. culmorum contrasts with findings for F. graminearum and F. pseudograminearum where the fungi were observed to colonize stem and head tissues without inducing symptoms (Mudge et al., 2006). It is tempting to suggest that the inability of F. culmorum to colonize wheat stems to reach the head by the time of plant maturity may be linked to the host response occurring in advance of fungal colonization. This contrasts with the situation for F. graminearum and F. pseudograminearum where symptom development is somehow suppressed by these fungi, potentially facilitating more rapid progress through the stem.
In the current experiments, although the fungus was not able to grow systemically beyond the 3rd node, DON synthesized by the pathogen was detected in the last internode and even in the heads. Fusarium culmorum, like F. graminearum and F. pseudograminearum, produces DON in vitro and in planta not only when inciting FHB but also when inciting CR. The accumulation of DON in wheat heads has been observed following infection of the stem base by F. graminearum and F. pseudograminearum but in these reports the fungus was also present in these tissues (Mudge et al., 2006; Poels et al., 2006). A preliminary report indicated that DON can be found in tissues of durum wheat ahead of colonization by F. graminearum (Moretti, 2008). However, to the authors’ knowledge, the present report is the first to demonstrate the translocation of DON to the head well in advance of colonization following stem base infection by F. culmorum. Combined with results from the present study, the available data suggest that, following infection of the stem base by F. culmorum, F. pseudograminearum or F. graminearum, DON can accumulate in wheat heads making CR disease an additional potential source of grain contamination. The translocation of DON from the stem base to the head provides an explanation for the findings of previous studies where DON was found in grain from field samples in the absence of any detectable fungus (Xu et al., 2008).
The combined data from the present study show that, in the experimental conditions here, DON was present in wheat tissues in the absence of fungal colonization. Similar findings were made in maize by Young & Miller (1985) who detected DON in an area of the maize stalk in the absence of fungus after ear inoculation. Translocation of DON within wheat heads during FHB infection was proposed by Snijders & Kretching (1992) who hypothesized that DON is transported from the chaff to the young kernels prior to fungal colonization. Photosynthates from glumes and lemma are translocated preferentially to the nearest kernels and enter the kernel via the phloem (Heyne, 1987), and Snijders & Kretching (1992) suggested that, as it is water soluble, DON might also be similarly distributed throughout the chaff and kernel by this means. This is supported by the observed translocation of DON in xylem and phloem sieve tubes ahead of fungal colonization of wheat heads following inoculation with F. culmorum (Kang & Buchenauer, 2002).
The distribution of DON in the heads following translocation from CR at the stem base is very similar to that found in the heads following FHB infection. The concentration of DON in wheat heads described a descending gradient from the rachis, through the lemmas and grains to the peduncle following FHB infection (Sinha & Savard, 1997). Similarly, this study observed that the rachis contained higher concentration of DON compared to the chaff (lemma and palea tissues) and grain tissues, describing a similar distribution to that observed for FHB. The mechanisms responsible for the apparent preferential accumulation of DON in rachis tissues are unknown.
In conclusion, these investigations provide novel insights into interactions between soft wheat and F. culmorum following CR infection. Most significantly, the repeated observation of the translocation of DON from stem base to the head following infection by F. culmorum indicates that CR infection may pose an additional route for the contamination of wheat and wheat-derived foodstuffs by Fusarium mycotoxins.