Trichothecene mycotoxin genotypes of Fusarium graminearum sensu stricto and Fusarium meridionale in wheat from southern Brazil

Authors

  • L. B. Scoz,

    1. Departmento de Fitossanidade, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 7712, 91540000, Porto Alegre, RS, Brazil; and
    Search for more papers by this author
  • P. Astolfi,

    1. Departmento de Fitossanidade, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 7712, 91540000, Porto Alegre, RS, Brazil; and
    Search for more papers by this author
  • D. S. Reartes,

    1. Departmento de Fitossanidade, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 7712, 91540000, Porto Alegre, RS, Brazil; and
    Search for more papers by this author
  • D. G. Schmale III,

    1. Department of Plant Pathology, Physiology, and Weed Science, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA
    Search for more papers by this author
  • M. G. Moraes,

    1. Departmento de Fitossanidade, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 7712, 91540000, Porto Alegre, RS, Brazil; and
    Search for more papers by this author
  • E. M. Del Ponte

    Corresponding author
    1. Departmento de Fitossanidade, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 7712, 91540000, Porto Alegre, RS, Brazil; and
    Search for more papers by this author

*E-mail: emerson.delponte@ufrgs.br

Abstract

A total of 82 fungal isolates was obtained from wheat kernel samples affected by fusarium head blight collected from 20 locations in southern Brazil. Polymerase chain reaction (PCR) assays were used to characterize trichothecene mycotoxin genotypes [deoxynivalenol (DON), nivalenol (NIV) and two acetylated derivatives of DON]. To identify isolates that producing DON and NIV, portions of the Tri13 gene were amplified. To identify 3-acetyl-deoxynivalenol (3-ADON) and 15-acetyl-deoxynivalenol (15-ADON) genotypes, portions of Tri3 and Tri12 were amplified. Nearly all of the isolates studied (76/82) were of the DON/15-ADON genotype. Six of the isolates were of the NIV genotype. The DON/3-ADON genotype was not observed. Portions of three genes were sequenced from representative isolates of the NIV and DON/15-ADON genotypes and compared with sequences from curated reference isolates of Fusarium in GenBank. blast queries for individual gene sequences and pairwise comparisons of percentage identity and percentage divergence based on 1676 bp of concatenated DNA sequence suggested that the isolates representing the DON/15-ADON genotype were Fusarium graminearum sensu stricto and the isolates representing the NIV genotype were Fusarium meridionale. This is the first detailed report of trichothecene mycotoxin genotypes of F. graminearum and F. meridionale in Brazil.

Introduction

Fusarium head blight (FHB) is a re-emerging disease of increasing concern to wheat and other small grain crops with a devastating impact worldwide (McMullen et al., 1997; Goswami & Kistler, 2004). In Brazil, FHB epidemics have become more frequent since the 1990s and often result in significant yield losses (Panisson et al., 2003). The principal causal agents of FHB worldwide are members of the Fusarium graminearum species complex (hereafter referred to as the Fg complex), representing at least nine phylogenetically distinct species (O'Donnell et al., 2004). Previous reports in Brazil suggested that F. graminearum sensu stricto (teleomorph Gibberella zeae) was the principal causal agent of FHB in wheat (Angelotti et al., 2006).

PCR assays have been developed to differentiate F. graminearum isolates from those of related Fusarium species (Schilling et al., 1996; Nicholson et al., 1998; Jurado et al., 2005) and other molecular tools have been used to characterize the population structure of members of the Fg complex from different geographical areas that could relate to differences in pathogenicity and/or mycotoxin potential (Nicholson et al., 2003; O'Donnell et al., 2004; Schmale et al., 2006a).

FHB is of particular concern because grain may be contaminated with trichothecene mycotoxins, such as deoxynivalenol (DON) and nivalenol (NIV) (Joffe, 1986). Unlike large-scale efforts for monitoring mycotoxins in other regions of the world (D'Mello et al., 1999; Bottalico & Perrone, 2002), little is known about the occurrence, distribution and impact of trichothecene mycotoxins in commercial small grains in Brazil (Furlong et al., 1995; Oliveira et al., 2002; Birck et al., 2006). Recent studies in Brazil suggest a consistent threat of mycotoxin contamination in small grains. Calori-Domingues et al. (2007) showed high levels of DON contamination in 94% (47/50) of wheat samples collected in several locations in southern Brazil. Additional work is needed to monitor trichothecene levels in food and feed, and to establish a series of regional food safety guidelines.

Both DON and NIV have been implicated in a number of human and animal mycotoxicoses (Bryden, 2007), and also are acutely phytotoxic, acting as “virulence” factors in wheat (Maier et al., 2006). Isolates of F. graminearum that produce DON may also produce two acetylated derivatives of DON, 3-acetyl-deoxynivalenol (3-ADON) and 15-acetyl-deoxynivalenol (15-ADON) (Miller et al., 1991). DON-producing isolates of F. graminearum appear to occur more frequently than NIV-producing isolates in Europe (Waalwijk et al., 2003; Jennings et al., 2004a), North America (Ward et al., 2002), South America (Ramirez et al., 2006) and China (Ji et al., 2007). In Japan, however, NIV-producing isolates appear to predominate (Suga et al., 2008).

PCR assays have been developed to rapidly assess trichothecene mycotoxin genotypes of F. graminearum. Such assays rely on the amplification of portions of genes that code for key enzymes involved in trichothecene biosynthesis (Lee et al., 2001; Ward et al., 2002; Chandler et al., 2003; Jennings et al., 2004a; Quarta et al., 2006). These assays are attractive because they are rapid, relatively inexpensive, and may be used to accurately predict the types of trichothecene mycotoxins produced by these isolates in culture (Ward et al., 2002). Based on previous observations of trichothecene mycotoxin genotypes of members of the Fg complex in other areas of the world (Jennings et al., 2004a; Quarta et al., 2006; Ji et al., 2007), it was hypothesized here that (i) DON/3-ADON, DON/15-ADON and NIV genotypes of members of the Fg complex are present in wheat fields in southern Brazil, and (ii) the frequency of these trichothecene genotypes varies within and among wheat fields and geographic locations in southern Brazil. These hypotheses were tested by collecting and purifying isolates tentatively identified as F. graminearum from 20 different locations in southern Brazil, and implementing a series of PCR assays to assess trichothecene mycotoxin genotypes in these populations. Portions of three genes from representative isolates were also sequenced to assist in their identification within the Fg complex. The specific objective of the work was to determine trichothecene mycotoxin genotypes (3-ADON, 15-ADON or NIV) of F. graminearum from wheat crops in southern Brazil.

Materials and methods

Wheat sampling and culture conditions

Grain samples were obtained from commercial wheat fields located across the states of Rio Grande do Sul (RS), Santa Catarina (SC) and Paraná (PR), Brazil, during the 2006 growing season. Samples weighing 200–500 g, obtained after harvesting operations, were shipped in paper bags to the laboratory. Grain samples were labelled and stored at 5°C in the laboratory. Kernel samples from all locations were visually inspected to separate out those showing symptoms and signs of FHB: a rough, wilted appearance, ranging in colour from pink to soft-grey or light-brown (McMullen et al., 1997). Affected kernels were transferred to water agar media amended with 50 µg ampicillin mL−1, and incubated at 28°C for 5 days with a 12-h photoperiod under cool white fluorescent lamps.

Fungal isolation and DNA extraction

Fungal colonies growing from the damaged kernels were examined under the microscope and those showing typical F. graminearum morphology (Nelson et al., 1983) were single-spored and maintained as spore suspensions in 15% glycerol at −80°C. Isolates were deposited in the culture collection of the Department of Plant Protection, Federal University of Rio Grande do Sul, Brazil. All isolates were grown on potato dextrose agar (PDA) plates for 5–6 days and genomic DNA was extracted from aerial mycelium using a protocol described by Doyle & Doyle (1987).

PCR assays for trichothecene mycotoxin genotypes

Trichothecene mycotoxin genotypes of F. graminearum were identified using a suite of primers directed to portions of genes that are predictive of isolates producing DON, NIV, 3-ADON and 15-ADON. DON and NIV genotypes were identified using a PCR assay to amplify portions of the Tri13 gene (Chandler et al., 2003). DON, 3-ADON, 15-ADON and NIV genotypes were identified using a multiplex PCR assay to amplify portions of Tri3 and Tri12. With the Tri3 primer set, PCR products of 243, 610 and 840 bp are produced for 3-ADON, 15-ADON and NIV genotypes, respectively (Ward et al., 2002; Gale et al., 2005). With the Tri12 primer set, PCR products of 410, 670 and 840 bp are produced for 3-ADON, 15-ADON and NIV genotypes, respectively (Ward et al., 2002; Gale et al., 2005). All information on primer sequences and amplicon sizes for respective trichothecene mycotoxin genotypes is summarized in Table 1.

Table 1.  Primer identification, sequences and expected amplicon sizes for trichothecene mycotoxin genotypes of Fusarium observed in this study
Primer designationPrimer sequenceTarget geneAmplicon (bp)Trichothecene mycotoxin genotypeReference
Tri13FCATCATGAGACTTGTKCRAGTTTGGGTri13282DONChandler et al. (2003)
Tri13DONRGCTAGATCGATTGTTGCATTGAG    
Tri13NIVFCCAAATCCGAAAACCGCAG 312NIVChandler et al. (2003)
Tri13RTTGAAAGCTCCAATGTCGTG    
3CONTGGCAAAGACTGGTTCACTri32433-ADONWard et al. (2002)
3NAGTGCACAGAATATACGAGC 61015-ADON 
3D15AACTGACCCAAGCTGCCATC 840NIV 
3D3ACGCATTGGCTAACACATG    
12CONCATGAGCATGGTGATGTCTri124103-ADONWard et al. (2002)
12NFTCTCCTCGTTGTATCTGG 67015-ADON 
12–15FTACAGCGGTCGCAACTTC 840NIV 
12–3FCTTTGGCAAGCCCGTGCA    

PCR conditions

PCR assays were conducted using 10–20 ng of fungal DNA in a total volume of 25 µL containing 1·5 mm MgCl2, 2U Taq DNA polymerase and 200 mm dNTPs. Primers are listed in Table 1. PCR amplification of Tri3, Tri12 and Tri13 consisted of an initial step at 94°C for 10 min, followed by two cycles of 94°C for 30 s, 59°C for 30 s and 72°C for 30 s. The annealing temperature was stepped down every two cycles to 58, 56, 54, 53, 52 and 51°C, then 50°C for 21 cycles, with a final step at 72°C for 10 min. Resulting PCR products were separated by gel electrophoresis, stained with EtBr at a final concentration of 0·5 µg mL−1 and visualized under UV light.

Sequence-assisted identification of species

Portions of three genes [phosphate permease (PHO), putative reductase (RED) and ammonia ligase (URA) (O'Donnell et al., 2004)] were amplified and bi-directionally sequenced from four Brazilian isolates representing the NIV genotype (61901, 62503, 62701 and 64305) and three representing the DON/15-ADON genotype (62001, 62902 and 63604) (GenBank Accession Nos., FJ183406–FJ183426). Primer pairs were as follows: PHO1 and PHO6, RED1d and RED2, and URA7 and URA10 (O'Donnell et al., 2004). Consensus sequences were aligned and trimmed in seqman pro (version 7·1·0, DNASTAR, Inc.). PHO, RED and URA sequences from curated NRRL reference isolates of Fusarium were downloaded from GenBank. Sequences from the Brazilian isolates and GenBank accessions were aligned in megalign (version 7·1·0, DNASTAR, Inc.) using the clustalv method (Higgins & Sharp, 1989). Pairwise comparisons of percentage identity and percentage divergence were investigated for sequences from portions of individual genes and for concatenated sequences.

Results

Sampling

Trichothene mycotoxin genotypes were determined for a total of 82 single-spored isolates tentatively identified as F. graminearum collected from 20 different municipalities in the three southern states of Brazil. The median of the number of isolates per location was 2·5 (Table 2). Isolates were selected based on the number of received samples per municipality and the incidence of the fungus in the kernels from each sample (data not shown), which led to a variable number of isolates per location. In some locations, individual isolates originated from both the same or different field samples, but always from different infected kernels.

Table 2.  Trichothecene mycotoxin genotypes of Fusarium graminearum sensu stricto and Fusarium meridionale in three southern states in Brazil. Isolates were obtained from fields in Rio Grande do Sul (RS), Santa Catarina (SC) and Paraná (PR)
StateNo. of locationsNo. of isolatesIsolates/locationaDON/15-ADONDON/3-ADONNIV
  • a

    Median (minimum–maximum numbers).

RS14542·5 (1–14)5103
SC 3112 (1–8)1001
PR 3177 (1–9)1502
Total20822·5 (1–14)7606

Trichothecene genotypes

The Tri13 PCR assay demonstrated that nearly all (93%, 76/82) of the isolates were of the DON genotype, with the remainder (7%, 6/82) of the NIV genotype. These genotypes were further characterized using Tri3 and Tri12 multiplex PCR assays. These assays demonstrated that all of the DON genotypes were also of the 15-ADON genotype; 610- and 640-bp products were observed for the Tri3 and Tri12 assays, respectively (Fig. 2). No DON/3-ADON genotypes were observed.

Figure 2.

PCR products for Fusarium trichothecene mycotoxin genotypes. Singleplex PCR assays with primers Tri13NIVF/Tri13R and Tri13F/Tri13DONR were used to determine (a) NIV (312 bp) and (b) DON (282 bp) genotypes, respectively. (c) A multiplex PCR assay for Tri3 was used to determine 3-ADON (243 bp), 15-ADON (610 bp) and NIV genotypes (840 bp). (d) A multiplex PCR assay for Tri12 was used to determine 3-ADON (410 bp), 15-ADON (670 bp) and NIV (840 bp) genotypes. PCR products for 13 Brazilian isolates and control isolates are presented. The following isolates are represented in lanes 1–13: 1, 61901; 2, 60101; 3, 60801; 4, 60901; 5, 62701; 6, 61001; 7, 61101; 8, 61201; 9, 62401; 10, 61902; 11, NIV control; 12, 15-ADON control; and 13, 3-ADON control. The three control isolates of F. graminearum (lanes 11–13) were from the USA. C, absence of genomic DNA (negative control). M, 100-bp mass ladder.

Species identification

Preliminary data from PCR assays using the Fg16F/R primer pair suggested that the populations might contain two different species within the Fg complex: all of the isolates of the DON/15-ADON genotype produced a PCR amplicon of 450 bp, and all of the isolates of the NIV genotype produced a PCR amplicon of 490 bp (data not shown; Nicholson et al., 1998). To assist in the identification of the Brazilian isolates to species, portions of the PHO, RED and URA genes from isolates representing both NIV and DON/15-ADON genotypes were amplified and bi-directionally sequenced. blast queries for individual gene sequences and pairwise comparisons of percentage identity based on 1676 bp of concatenated DNA sequence showed that (i) representative Brazilian isolates of the DON/15-ADON genotype were on average 98·7% similar to reference isolates of F. graminearum sensu stricto (Table 3), and (ii) representative Brazilian isolates of the NIV genotype were on average 98·43% similar to reference isolates of F. meridionale (Table 3). Separate sequence alignments for PHO, RED and URA showed consistent clustering of NIV genotypes with F. meridionale (NRRL isolates 28436, 28723 and 29010) and DON/15-ADON genotypes with F. graminearum (NRRL isolates 28336, 5883 and 13383) (data not shown). Alignments of sequences from Brazilian isolates alone showed clear separation of NIV and DON/15-ADON genotypes into two separate clades for individual genes and for concatenated sequences from all three genes (data not shown).

Table 3.  Pairwise comparisons of percentage identity (above diagonal) and percentage divergence (below diagonal) for representative Brazilian isolates (NIV and DON/15-ADON genotypes) and reference NRRL isolates of Fusarium spp. Comparisons are based on 1676 bp of DNA sequence from portions of three genes. Sequences were concatenated and aligned based on the clustalv alignment method developed by Higgins & Sharp (1989)
 61901 (NIV)62503 (NIV)62701 (NIV)64305 (NIV)62001 (15ADON)62902 (15ADON)63604 (15ADON)F. meridionale NRRL 28436F. meridionale NRRL 28723F. meridionale NRRL 29010F. graminearum ss NRRL 5883F. graminearum ss NRRL 13383F. graminearum ss NRRL 28336F. cerealis NRRL 13721F. culmorum NRRL 25475F. pseudograminearum NRRL 28338
61901 (NIV)***98·9099·5098·6097·1096·7097·2098·6098·7098·7096·7096·7096·7095·9095·4090·90
62503 (NIV)0·20***98·9098·0096·6095·8096·5098·0098·1098·1096·0096·1096·1095·3094·8090·10
62701 (NIV)0·100·20***98·6097·5096·5097·6098·9099·1099·1097·0097·1097·1096·3095·8091·20
64305 (NIV)0·200·300·40***96·4095·8096·5097·8098·0098·0095·9096·0096·0095·2094·7090·00
62001 (15ADON)2·402·402·302·60***98·5099·6097·1097·2097·2099·1099·0099·3096·5095·3090·90
62902 (15ADON)2·802·903·103·101·20***98·6095·8096·0096·0097·8097·8098·0095·3094·3090·30
63604 (15ADON)2·402·502·302·500·101·00***96·9097·1097·1098·9098·9099·1096·3095·2090·80
F. meridionale NRRL 284360·400·400·200·702·203·402·30***99·9099·9097·8097·9097·9097·1096·5092·00
F. meridionale NRRL 287230·200·300·100·502·103·202·100·10***100·0097·9098·0098·0097·2096·7092·20
F. meridionale NRRL 290100·200·300·100·502·103·202·100·100·00***97·9098·0098·0097·2096·7092·20
F. graminearum ss NRRL 58832·402·402·302·600·201·300·202·202·102·10***99·7099·8097·1096·1091·80
F. graminearum ss NRRL 133832·302·402·202·600·301·400·302·202·102·100·30***99·8097·2096·1091·80
F. graminearum NRRL 283362·302·402·202·600·101·100·102·202·102·100·200·20***97·2096·1091·80
F. cerealis NRRL 137212·903·002·903·202·803·802·802·902·702·702·802·702·70***97·2091·90
F. culmorum NRRL 254753·503·603·403·803·905·004·003·403·303·303·903·903·902·80***91·30
F. pseudograminearum NRRL 283387·407·507·407·807·808·407·707·507·307·307·607·707·707·308·20***

Discussion

This is the first detailed report of trichothecene mycotoxin genotypes in populations of F. graminearum sensu stricto and F. meridionale collected from southern Brazil. The majority of the isolates (93%) were of the DON/15-ADON genotype. NIV genotypes were present in the some of the same fields as DON/15-ADON genotypes (Fig. 1). Mixed populations of DON and NIV genotypes in the field were reported previously in populations of F. graminearum (Jennings et al., 2004a) and F. culmorum (Jennings et al., 2004b). A study in China showed that DON/15-ADON genotypes of F. graminearum appeared to be associated with wheat, whereas DON/3-ADON genotypes were frequently associated with barley (Ji et al., 2007). In Japan, of 50 F. graminearum sensu stricto isolates, 35 were of the DON/3-ADON and 15 of the DON/15-ADON genotype (Suga et al., 2008). In the eastern USA, Schmale et al. (2006b) found that over 96% of isolates from atmospheric populations of F. graminearum in New York were of the DON/15-ADON genotype and less than 4% were of the DON/3-ADON genotype; the NIV genotype was not observed in their populations. A similar relationship between DON/15-ADON and DON/3-ADON was observed in field populations from New York, Virginia and North Carolina (Schmale et al., 2006c). In the midwestern USA, field populations of F. graminearum appeared to have higher frequencies of the DON/3-ADON genotype, which may have increased dramatically in recent years (Gale et al., 2005). A dramatic contrast in the American population regarding trichothecene genotypes was observed among FHB-infected heads in Louisiana, which consisted mostly of the NIV genotype (Gale et al., 2005). The results of the present study in southern Brazil differed from those in the USA; the DON/3-ADON genotype was not observed and isolates of the NIV genotype were distributed among mixed field populations.

Figure 1.

Geographic distribution of trichothecene mycotoxin genotypes of Fusarium graminearum sensu stricto and Fusarium meridionale from three southern states in Brazil. DON/15-ADON (F. graminearum) and NIV (F. meridionale) genotypes were observed in a sample of 82 isolates collected from 20 locations.

A recent study in Argentina revealed that DON and 3-ADON were the primary trichothecenes produced by F. graminearum grown in rice cultures (Ramirez et al., 2006). Molto et al. (1997) showed that all 27 isolates of F. graminearum from maize in Argentina were able to produce DON, but only seven out of 27 produced 3-ADON and no isolate produced NIV. In contrast, Lori et al. (1992) showed that 17 out of 76 F. graminearum isolates from Argentinean wheat produced NIV and 15 out of 76 produced both DON and NIV. In general, most F. graminearum DON-producers in Argentina seem to be also capable of producing 3-ADON, while most of the Brazilian isolates in the present study were of the DON/15-ADON genotype. Continuous monitoring of South American F. graminearum isolates, especially at broader spatial scales in South America, will be important to understand epidemiological factors that may contribute to the distribution of trichothecene mycotoxin genotypes and if any potential fitness advantages exist among isolates of acetylated forms of DON- and NIV-producing isolates.

The present results contribute to a deeper knowledge of the trichothecene mycotoxin profiles of members of the Fg complex in Brazil. Nicholson et al. (1998) demonstrated a relationship between members of the Fg complex (formerly considered lineages) of F. graminearum and SCAR (sequence-characterized amplified regions) groups. Preliminary data from the present study showed two SCAR groups using the Fg16F/R primer pair (450 bp and 490 bp), and these groups consisted solely of NIV or DON/15-ADON genotypes. Sequencing of portions of three genes confirmed the suspicion of two separate species; blast queries for individual gene sequences and pairwise comparisons of percentage identity and percentage divergence based on 1676 bp of concatenated sequence suggested that the Brazilian isolates representing the DON/15-ADON genotype were F. graminearum sensu stricto and those representing the NIV genotype were F. meridionale. Separate sequence alignments for each of the individual genes showed consistent clustering of NIV genotypes with F. meridionale and DON/15-ADON genotypes with F. graminearum sensu stricto, and alignments of sequences from Brazilian isolates alone showed clear separation of NIV and DON/15-ADON genotypes into two separate clades. Additional studies for establishing relationships between members of the Fg complex and their respective tricthothecene mycotoxin genotypes in Brazil will be necessary to determine their distribution across small scales and to monitor potential shifts in field populations in the future. In addition, it will be necessary to assess trichothecene mycotoxin genotypes of FHB pathogens from other cereals, such as barley and oats. This will be important to ensure that appropriate assays are used to measure mycotoxin contamination in Brazil. Moreover, epidemiological studies will be needed to examine the frequency and distribution of NIV-producers in Brazil and the conditions that favour mycotoxin production. Such information will be crucial for wheat breeders to develop new varieties with reduced mycotoxin potential. PCR assays are an important tool for screening large populations of mycotoxigenic fungi present in grain or processed foods worldwide, and should be considered as part of routine monitoring programmes for mycotoxins that pose a threat to the health of humans and domestic animals.

Acknowledgements

The authors thank Dirceu Gassen and his co-workers with Cooplantio for helping and supporting the survey in commercial fields in southern Brazil. We thank Carolina Deuner, Juliano L. de Almeida and Volmir Marchioro for providing wheat grain samples from experimental plots showing FHB symptoms.

Ancillary