Soybean Pod Blight and Root Rot Caused by Lineages of the Fusarium graminearum and the Production of Mycotoxins

Queima da vagem e podridão de raízes da soja causadas por linhagens de Fusarium graminearum e produção de micotoxinas Amostras de sementes de soja (Glycine max) produzidas no Sul do Brasil estavam infetadas com Fusarium graminearum. Para determinar se membros desse complexo de espécies eram patogênicos à soja, seis isolados do fungo, originados de sementes soja, foram adicionados ao solo numa densidade de 103 macroconídios/ml ou vagens foram inoculadas individualmente com 104 macroconídios/ml. Plântulas crescidas em solo infestado desenvolveram pequenas lesões necróticas na coroa e na parte superior das raízes. Vagens inoculadas com conídios desenvolveram lesões necróticas marrom-escuro grandes (> 1 cm). Vagens jovens inoculadas com o fungo secaram e caíram da planta. Isolados de F. graminearum recuperados das lesões das coroas, raízes e vagens das plantas de soja foram identificadas como pertencentes aos grupos de compatibilidade 1, 2 ou 8, através da análise da seqüência de DNA do gene EF1-alfa, quando comparados com as linhagens conhecidas. Dois isolados de F. graminearum dos Estados Unidos, pertencentes ao grupo de compatibilidade 7, causaram sintomas semelhantes em soja. Testes das micotoxinas produzidas em soja e trigo (Triticum aestivum) indicaram que a maioria dos isolados Brasileiros produz Nivalenol, como a principal toxina dos tricotecenos, ao invés de Deoxinivalenol. Em adição, os isolados dos grupos 2 e 8 produziram uma nova toxina, o 3,AcetilNivalenol. RESUMO Surveys of soybean (Glycine max) seed grown in South Brazil revealed infection with Fusarium graminearum. To determine if members of this complex were pathogenic to soybean, six strains derived from soybean were added to soil at a rate of 103 macroconidia/ ml or individual pods were inoculated with 104 macroconidia/ml. Seedlings grown in infested soil developed small necrotic lesions in the crown and upper roots. Pods inoculated with conidia developed large (>1 cm), dark brown, necrotic lesions. Younger pods inoculated with the fungus blighted and dropped from the plant. Strains of the F. graminearum complex recovered from lesions on the crown, roots and pods of soybean plants were identified as lineage 1, 2 or 8 by obtaining the DNA sequence from the EF1-alpha gene and comparing it to strains of the known lineage. Two strains of F. graminearum lineage 7 from the U.S. caused similar symptoms of the disease on soybean. Mycotoxin tests on soybean and wheat (Triticum aestivum) indicate that most Brazilian strains produce nivalenol as the major trichothecene mycotoxin rather than deoxynivalenol. In addition, strains from lineages 2 and 8 produce the novel trichothecene, 3-acetylnivalenol. Additional keywords: Glycine max, Gibberella zeae, pathogenicity, Deoxinivalenol, Nivalenol, compatibility groups. 492 Fitopatol. bras. 29(5), set out 2004 INTRODUCTION Members of the Fusarium graminearum Schwabe complex (hereafter referred to as the Fg complex) are as pathogens of cereal crops found throughout the world. During the 1990’s, members of this complex were largely responsible for the highly destructive Fusarium head blight (scab) epidemics of wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.) in North and South America, Asia and Europe (Dubin et al., 1997). The re-emergence of Fusarium head blight as an important constraint to small grain production may be attributable, at least in part, to the increase of lowor no-till cropping systems worldwide. These cultural practices leave crop residue on the soil surface where it may be colonized by saprophytic or potentially pathogenic fungi. Fusarium graminearum has been shown to readily colonize crop debris left behind by wheat, corn (Zea mays L.) and other rotation crops (Cook, 1984; Wicklow et al., 1987; Fernandez, 1991; Miller et al., 1998). Soybean [Glycine max (L.) Merril] is often used in rotation with wheat and other cereal crops. Soybean crop residues in fields under conservation tillage have been found to be heavily colonized by F. graminearum (Wicklow et al., 1987; Fernandez & Fernandes, 1990; Baird et al., 1997). Although members of this species complex also reportedly colonize living soybean stems (Harrington et al., 2000) and seeds (Osorio & McGee, 1992; Jacobsen et al., 1995), many authors consider members of the F. graminearum complex to A soybean pod blight and root rot caused by... 493 Fitopatol. bras. 29(5), set out 2004 NRRL # / CDL # 1 Host Cultivar Country State 2 Lineage 31230 = 01-105 Soybean Bragg Brazil RS 2 31321 = 01-156 Soybean BRS66 Brazil RS 8 31322 = 01-157 Soybean BRS66 Brazil RS 2 31323 = 01-158 Soybean BRS66 Brazil RS 1 31351 = 01-186 Wheat BRS192 Brazil 8 31354 = 01-189 Soybean FT Abyara Brazil 2 31355 = 01-190 Soybean FT Abyara Brazil 8 29169 Wheat Unknown USA Kansas 7 31084 Corn Unknown USA Michigan 7 Paraná Paraná Paraná 1 Numbers given to strains by culture collections at the National Center for Agricultural Utilization Research (NRRL) or the Cereal Disease Laboratory (CDL). 2 RS = Rio Grande do Sul. TABLE 1 Strains of Fusarium graminearum used in this study be nonpathogenic to soybean (Chamberlain, 1972; Fernandez & Fernandes, 1990; Garcia-Romera et al., 1998; Miller et al., 1998). Published reports of disease on soybean caused by the F. graminearum complex are equivocal, circumstantial and/ or contradictory. Anderson et al. (1988) reported isolation of these fungi from diseased and stunted soybean in Ontario, Canada at a fairly high frequency (18%). However, they were also isolated at a similar rate (14%) from plants showing no symptoms. Fusarium graminearum has been considered a secondary colonist of soybean seed damaged by other fungi or by freezing (Osorio & McGee, 1992; Jacobsen et al., 1995; Ward et al., 2002). Although Agarwal (1976) reported Fusarium root rot of soybean caused by F. graminearum in India, the description of the pathogen, especially the lack of a homothallic sexual stage, may implicate the fungus now known as F. pseudograminearum O’Donnell & T.Aoki (Aoki & O ́Donnell, 1999), [formerly known as F. graminearum group 1 (Francis & Burgess, 1977)] as the actual pathogen in that report. Other reports appear to suggest that the F. graminearum complex (teleomorph = Gibberella zeae), is nonpathogenic to soybean even at high inoculum levels (Garcia-Romera et al., 1998), or that it may even protect the plant against other root diseases (Chamberlain, 1972). Soybean production in South Brazil involves no-till cultivation and double cropping rotation with wheat, barley or oats (Avena sativa L.). Large inoculum levels of F. graminearum are present under these conditions and we have observed that members of the F. graminearum complex appear to be affecting plants previously considered nonhosts. Oats, for example, now are susceptible to scab disease and perithecia may even form on maturing panicles in the field (J. Martinelli, personal observation). Recent surveys of soybean seed grown in South Brazil revealed infection by members of the Fg complex. Seed lots varied in the percentage of infected seed, ranging from 0 ca. 20%. Therefore, the objective of this study was to determine whether strains of the F. graminearum complex obtained from soybean seeds are pathogenic to soybean and to determine their genetic lineage as well as their ability to produce toxins. An abstract of this work was published previously (Martinelli et al., 2001). MATERIALS AND METHODS Origin and identification of the isolates Brazilian soybean seeds were surface-disinfested with 1.0% sodium hypochlorite for 2 min and plated on 2% water agar or one-quarter strength potato dextrose agar (PDA Difco; Detroit, MI). After eight days, conidia were examined from six colonies growing independently from the seeds and determined by morphology (Aoki & O ́Donnell, 1999) to be members of the F. graminearum complex. Monosporic cultures were grown on mung bean agar (Evans et al., 2000) to obtain inoculum for subsequent pathogenicity experiments. For comparison of pathogenicity, three additional strains were used: one isolated from wheat in Brazil, one from wheat (NRRL 29169) and corn (NRRL 31084) in the US. To determine the lineage of the Brazilian F. graminearum complex (Table 1) isolates obtained from soybean and wheat, a portion of the gene encoding translation elongation factor EF1-alpha was amplified and the DNA sequences were compared to those previously published from strains of known lineage (O’Donnell et al., 2000). In that work, allelic genealogies were constructed from DNA sequence of six single-copy nuclear genes from 27 strains of F. graminearum selected to represent the global genetic diversity of this pathogen. With the exception of one strain, all six genealogies recovered the same seven biogeographically structured lineages suggesting they represent phylogenetically distinct species among which gene flow has been very limited historically. Test with seedlings and adult plants To determine if strains of the F. graminearum complex are capable of causing disease symptoms on soybean, macroconidia of the six strains isolated from soybean were used to challenge seedlings and pods on adult plants of the American soybean varieties Glacier, Parker and Lambert. A macroconidial suspension (10 ml at 103 spores/ml) was added to pots (5x5x5 cm) with vermiculite containing 11day-old plants or to pots (13 cm height, 8.5 cm diameter at base and 12.5 cm diameter on top) with soil (pasteurized mix of field soil, sand, peat moss, composted manure 6:6:5:2) immediately prior to sowing seeds. Roots of seedlings in vermiculite pots J.A. Martinelli et al. 494 Fitopatol. bras. 29(5), set out 2004 were wounded with a knife just before inoculation whereas seedlings grown in soil were not wounded. For the latter, five holes per pot were made with a pencil by pushing it from the soil surface to the bottom of the pot after which the spore suspension (10 ml at 103 spores/ml) was added into each hole. One seed was then placed on top of each hole and they were then covered with a 3 cm layer of soil. Each isolate was added to four pots per variety containing either four (vermiculite) or five (soil) seedlings. Each soybean variety was tested for both seedling experiments, and for both experiments bedding material treated with water was used as a negative control. After addition of inoculum, seedlings were grown for 4 weeks after which the roots and crowns were examined. To test for the ability of the strains to cause disease on developing soybean seeds and pods, seeds of the same three varieties were planted in larger pots (15.5 cm height with a 10 cm diameter at the base and a 15 cm diameter at the top), one plant per pot, five pots per treatment and grown to the adult stage. At the beginning of seed formation (R5 stage), pods were inoculated with the strains by injecting 0.1 ml of a spore suspension containing ca.103 conidia/ml. To determine if the fungus could invade beyond the point of inoculation in most cases the pods were inoculated on the mid-lateral surface by injecting spores into a central carpal containing a developing seed. Plants inoculated with sterile water were included as a negative control. Recovery of isolates Seedlings were removed from pots four weeks after inoculation to test for fungal growth from the roots. Roots were excised, washed, surface-disinfested and plated on onequarter strength PDA. A similar procedure for identifying colonization of above ground tissue was used for inoculated pods and seed harvested from adult plants. Testing soybean isolates on wheat All Brazilian strains also were inoculated on wheat heads at anthesis to determine if the isolates from soybeans could cause disease on wheat. A drop (ca. 0.05 ml) of spore suspension from each isolate containing 103 spores/ml was inoculated into either a central spikelet of one head or on four other heads. The same spore suspensions were sprayed over the entire surface of the inflorescence until runoff. After inoculation the plants were kept in a dew chamber for 72 h and then transferred to the greenhouse to be evaluated ten days later for symptoms of Fusarium head blight. Gas chromatography/mass spectrometry (GC/MS) analysis of toxins Samples of pods showing symptoms of disease caused by Fusarium spp. seven days after inoculation were assayed as were soybean seeds and pods, and wheat heads ten days after inoculation. Each was analyzed for trichothecene toxins. Since the inoculated wheat did not produce any seed, entire heads were harvested. All samples were ground to a fine powder for toxin analysis according to the method described by Mirocha et al. (1998) with modifications. Briefly, 1 g of wheat heads, 2 g of soybean pods or 3 g of soybean seeds were extracted with 10 ml, 16 ml and 16 ml of acetonitrile:water (84:16 v/v), respectively, in 50-ml centrifuge tubes for 1 h. An aliquot of 1 ml of the extract after column cleanup was placed into a 1⁄2 dram vial and evaporated to dryness under nitrogen. The sample was derivatized with 20 μl of trimethylsilyl ether (TMS) reagent (TMSI/TMCS 100:1, Sigma) and diluted with 200 μl of isooctane. The GC/MS sample analysis was carried out on a Shimadzu QP5000 Gas chromatography/mass spectrometry system with a Shimadzu AOC-1400 autosampler and AOC-17 autoinjector (Shimadzu, Kyoto, Japan). Quantitative analysis was run in selected ion monitoring (SIM) mode. Ions at m/z 235.10, 259.10, 295.10, 377.10 and 392.15 were monitored. A full scan GC/MS analysis (scan range m/z 50 to m/z 600) was carried out for toxin identification of the wheat head samples. The GC column was a J&W DB-5ms, 0.25 μm film thickness, 250 μm i.d. and 30 μm in length. The GC temperature program included an initial temperature of 150 °C for 1 min, a ramp to 280 °C at 25 °C/ min, and an isothermal for 5 min.