All of the new primer pairs amplified a unique band of the expected size with the genomic DNA from the control strains F. verticillioides (CML 767) or F. subglutinans (UnB 379). The lowest amount of genomic DNA of the control isolates that could generate a visible band in a conventional agarose gel stained with ethidium bromide was 50 pg (0.05 ng) for the primer pair FV-F1/FV-R and 500 pg (0.5 ng) for the other new primer pairs (Figure 1). This sensitivity is similar as that of other primers described in the literature [13,14,25].

Regarding to the specificity analysis with the G. fujikuroi species, the primer pair FV-F1/FV-R targeting the gaoB gene from F. verticillioides amplified a DNA fragment of the expected size from the genomic DNA from F. verticillioides, but also from F. thapsinum and F. nygamai, two species phylogenetically related to F. verticillioides [2] (Figure 2A). The primer pair FV-F2/FV-R amplified a DNA fragment only from the F. verticillioides genomic DNA (Figure 2B). The two FS primer sets targeting the gaoB gene of F. subglutinans have amplified a DNA fragment of the expected size from the genomic DNA from the F. subglutinans strains used as control (UnB 379 and CML 772) and from F. konzum (Figure 2E,F), a species phylogenetic related to F. subglutinans [1]. The primer pair VER1/VER2 described by Mulè et al. [14] and targeting the calmodulin gene had specificity similar to the primer pair FV-F1/FV-R (Figure 2C). The primer set SUB1/SUB2 described by Mulè et al. [14] had a slightly different specificity of the FS primers, amplifying a DNA fragment from the genomic DNA from F. subglutinans and also from F. thapsinum (Figure 2G). The specificity of these primer pairs described by Mulè et al. [14] to the G. fujikuroi isolates had not been tested before.

All designed primer pairs were unable to amplify any DNA fragment, specific or nonspecific, from the genomic DNA of several species of the Fusarium genera and from several other fungal genera that could be associated with maize seeds (Table 1, Figure 3). In addition, the FV-F2/FV-R and FS primer pairs were also unable to recognize any target in the genomic DNA of several fungi that were isolated from maize seeds in this work as Penicillium purpurogenum, Penicillium digitatum, Penicillium sp., Stenocarpella sp., Curvularia sp., Cladosporium sp., Phlebiopsis sp., Pichia sp., Rhyzopus sp., Phoma sp., Thrichoderma sp., Epicocum sp., Cladosporium sp., Irpex sp., Microdochium nivale, Epicocum sp., Mucor sp., Fusarium graminearum, Aspergillus oryzae, Aspergillus candidus, Aspergillus flavus, and Aspergillus niger.

Concerning the primer sensitivity testing against Fusarium isolates found in maize seeds, the two FV primer pairs were able to amplify a DNA fragment from all 47 F. verticillioides isolates obtained from maize seeds (Table 2). The FS primer pairs were also able to molecularly identify the obtained F. subglutinans isolates. The DNA of other found Fusarium species was not recognized by the new primer pairs (Table 2). The genomic DNA of the isolate RV27-2 resulted in positive PCR reactions for the primer pairs FV-F1/FV-R and VER1/VER2 [14], indicating that it could be F. verticillioides, F. nygamai, or F. thapsinum. As it resulted in a negative PCR reaction with the more specific FV-F2/FV-R primer pair, it could not be F. verticillioides. In addition, it also resulted in a negative PCR reaction with the SUB1/SUB2 [14] primer pair, demonstrating that it is neither F. thapsinum nor F. subglutinans. Thus we have concluded that this isolate is F. nygamai, what was confirmed in the Fusarium-ID analysis. The identification of the isolate RV 27-2 as F. nygamai points out that the combination of individual PCR reactions with the primers FV-F2/FV-R (this work), FV-F1/FV-R (this work), and SUB1/SUB2 [14] can be used for the identification of F. nygamai. The information that the FV and FS primer pairs do not recognize the gaoB gene of F. andiyazi (Table 2) enhances the data about their specificity.

The combination of the primer pairs FV-F2/FV-R and FS-F1/FS-R in only one PCR reaction could amplify a 370 bp DNA fragment from the F. verticillioides genomic DNA and a 649 bp DNA fragment from F. subglutinans genomic DNA (Figure 4A), what is consistent with the amplicon sizes obtained with the primer pairs used individually (Figure 1). In addition, this established reaction could also amplify DNA fragments of the expected size from genomic DNA extracted from fungi growing on and on the medium around the maize seeds that were artificially or naturally contaminated with F. verticillioides or F. subglutinans, or both species (Figure 4A). The naturally contaminated maize seed lots chosen for this analysis were: MGA 10 from which F. verticillioides was isolated, PG 1 from which F. subglutinans was isolated, and RV 27 from which neither of these fungi was initially isolated. The results obtained (Figure 4A) indicate the presence of F. subglutinans in the PG 1 sample and the presence of F. verticillioides in all three naturally contaminated seed samples tested, what is in accordance with the literature that reports the presence of F. verticillioides in 100% of the maize seeds in Brazil [26]. Möller et al. [13] have also set up a multiplex PCR to detect F. verticillioides and F. subglutinans simultaneously, but there are no data on the specificity of these primers with the genomic DNA of F. circinatum, F. konzum, and F. andiyazi. The present multiplex PCR reaction represents a significant advance for the simultaneous molecular detection of F. verticillioides and F. subglutinans.

When the primer pair FS-F2/FS-R was used in a multiplex reaction with the primer pair SUB1/SUB2 [14], two DNA fragments were amplified from the genomic DNA of F. subglutinans: one of 370 bp and one of 631 bp (Figure 4B). In addition, this set of primers could amplify a single DNA fragment of 370 bp when the genomic DNA from F. konzum was present in the reaction or a single DNA fragment of 631 bp when the genomic DNA from F. thapsinum was used in the reaction (Figure 4B). This reaction represents an improvement to the described reactions designed to detect F. subglutinans. The primers developed by Möller et al. [13] to identify F. subglutinans were not tested for all G. fujikuroi isolates, and the primers designed by Zheng and Ploetz [19] were specific for F. subglutinans and F. nygamai. In addition, there are few data in the literature about reactions to identify F. thapsinum, which is an important sorghum pathogen.

The isolation of DNA directly from the seeds, without previous culture to obtain mycelia, was also tried but resulted in no DNA amplification in the multiplex PCR reaction to detect F. verticillioides and F. subglutinans. This could have happened because the amount of fungal cells in the contaminated seeds is too low or because of the presence of inhibitors in the extracted DNA. To test for the presence of inhibitors in the DNA extracted from maize, the established multiplex PCR reaction to simultaneously detect F. verticillioides and F. subglutinans was performed with the lowest amount of fungi DNA that could result in amplification spiked with increasing amounts of genomic DNA extracted from healthy maize seeds. The results shown in Figure 5 evidenced that maize DNA only did not inhibit the reaction when present in a low amount.

The fungal isolates used in this work are listed in Table 1. To obtain a Fusarium collection from maize seeds, corn spikes showing signs of rot were collected from January to October of 2009, in fields in several producing regions of Brazil (Figure 6). The seeds of each spike were denominated as a lot and were kept in paper bags at 4 °C after insecticide treatment. A total of 57 maize seed lots were collected and analyzed. Six maize seeds of each lot, in duplicate, were disinfected by one min incubation in a solution containing 0.2% active chlorine, washed in sterile distilled water, and inoculated in a 10 mm diameter dish containing Malaquite Green Agar-MGA [27] supplemented with 350,000 UI/L of penicillin and 145 UI/L of streptomycin. The seeds plated in the medium were incubated for four to five days at 25 °C, with a photoperiod of 12 h. After germination, mycelia and conidia of peach or violet colored colonies were transferred to Petri dishes containing Carnation Leaf-piece Agar-CLA [1], which were incubated for a period of seven days at 25 °C, with a photoperiod of 12 h. A well-colonized carnation leaf fragment in the CLA culture was used for monosporic isolation as described in Nelson et al. [28]. The isolates morphologic characteristics were analyzed in micro culture performed in a one-cm³ block of Spezieller Nährstoffarmer Agar-SNA [1]. The isolates with cultural and morphological characteristics of Fusarium were cultivated for DNA extraction and molecular identification with the specific primers described by Mulè et al. [14]. These primers were chosen because they target a protein coding gene as the primers designed in this work. The genomic DNA of Fusarium isolates that were not identified as F. verticillioides or F. subglutinans were used in PCR reactions with specific primers for F. proliferatum [14]. These fungi were also analyzed using the methodology described in Geiser et al. [29] in which a portion of the elongation factor α was amplified through PCR, purified with the ExoSap-IT Kit (GE HealthCare, USA), and sequenced in the Center for the Human Genome Studies (CEGH) in the University of São Paulo (USP), Brazil. For the isolate identification, the obtained sequences were compared with sequences deposited in the data bank Fusarium-ID [29]. All obtained isolates are being maintained in the laboratory collection on PDA and SNA media with trimestral passages and in SNA medium under mineral oil.

To obtain fungi other than F. verticillioides or F. subglutinans from maize seeds, to be used in the primers specificity analysis, six maize seeds of each lot, in triplicate, were disinfected as described above and inoculated in a 10 mm diameter dish containing PDA, pH 4.5 with lactic acid [6], supplemented with penicillin and streptomycin, as described above. The seeds plated in the medium were incubated for five to seven days at 25 °C, with a photoperiod of 12 h. The fungi growing on and on the medium around the maize seeds with culture characteristics different from G. fujikuroi were submitted to monosporic isolation [28]. The classification of Aspergillus and Penicillium isolates was performed as described in Pitt and Hocking [30] with cultures in specific media. The molecular identification of the isolates of other genera was performed by the amplification of an rRNA gene fragment with the ITS4/5 primers [31], purification of the amplified DNA fragment with the PureLink™ PCR purification kit (Invitrogen, USA), sequencing in the CEGH (USP, Brazil) and comparison with sequences deposited in data banks. Specific identification of F. graminearum was performed in PCR reactions with the GOF/R specific primers [21]. All isolates obtained are being maintained in the laboratory collection on PDA with trimestral passages and on PDA medium under mineral oil.

Initially, the gaoB genes from F. verticillioides and F. subglutinans [24] (GenBank AN HM069186 and HM069185, respectively) and the gaoA gene from F. austroamericanum [20] (GenBank AN M86819) were aligned with the ClustalW program. Low similarity regions were chosen for primer design. Four pairs of primers were designed; two directed to the gaoB gene from F. verticillioides (FV-F1/FV-R and FV-F2/FV-R) and two other for the gaoB gene from F. subglutinans (FS-F1/FS-R and FS-F2/FS-R) to amplify a DNA fragment of 649 bp and 370 bp, respectively (Figure 7).

The primer sensitivity was tested by a ten-fold serial dilution of the control isolates genomic DNA in the established PCR reactions. The primer specificity was tested in PCR reactions using genomic DNA from G. fujikuroi species isolates, other Fusarium spp., other fungal genera, and fungi isolated from maize seeds.

For the DNA extractions, an approximately one-cm³ fragment of a monosporic culture in inclined PDA was smashed and shaken in 5 mL of distilled water. Two mL of the obtained suspension were used as inoculum in Erlenmeyer flasks of 125 mL containing 25 mL of potato dextrose medium. These flasks were incubated for five days at 25 °C. The mycelia were collected by filtration in sterile gauze. The mycelia was macerated in a mortar with liquid nitrogen, and transferred to microcentrifuge tubes. The DNA was extracted from the macerated mycelia using the protocol described by Koenig et al. [32]. The DNA obtained was quantified by measuring the absorbance at 260 nm and/or by fluorometry using the Qubit Quantitation Fluorometer and the Quant-it™ dsDNA HS Assay Kit (Invitrogen, USA).

The PCR with the primer pairs designed in this work was performed in a total volume of 25 μL of the following mixture: 1X enzyme buffer (20 mM Tris, pH 8.4, and 50 mM KCl), 1.5 mM MgCl₂, 1.5 U of Platinum^® Taq DNA polymerase (Invitrogen, USA), 0.2 mM of each dNTP, 25 pmol of each primer (FW and RV), and 50 ng of the genomic DNA. The reaction consisted of 25 cycles of 1.5 min at 94 °C, 1.5 min at the recommended annealing temperature for each pair of primers (Table 3), and 2 min at 72 °C in a Techne Thermocycler (England). Before the cycles, the samples were heated for 5 min at 94 °C, and after the cycles, the samples were incubated for 10 min at 72 °C. Negative controls (no DNA template) were used in each experiment to test for the presence of DNA contamination of reagents and reaction mixtures. Ten microliters of the PCR reaction were analyzed in an agarose gel containing ethidium bromide (0.25 μg/mL). Molecular weight markers were the 100 bp markers from Invitrogen (USA). The PCR products were visualized and photographed under UV light. All genomic DNA that resulted in nonamplification with the other used primers was tested for PCR amplification quality with the primers ITS4/5 [31]. All the PCR with the other primer pairs used in this work was performed as above with the exception with the primer pair ef1/ef2 [29] that was performed in a total volume of 25 μL with 50 ng of genomic DNA template and following the authors’ methodology.

A first multiplex PCR reaction to simultaneously detect F. verticillioides and F. subglutinans was established with the primer pairs FS-F1/FS-R and FV-F2/FV-R. Another multiplex PCR reaction to identify F. subglutinans was established with the primer pairs FS-F2/FS-R (this work) and SUB1/SUB2 [14]. Genomic DNA from F. konzum and F. thapsinum gave a positive signal in this last multiplex PCR reaction as well.

The reaction established to simultaneously detect F. verticillioides and F. subglutinans was also tested with genomic DNA extracted directly from artificially or naturally contaminated seeds or from fungi growing on artificially or naturally contaminated maize seeds using the protocol described by Koenig et al. [32]. To artificially contaminate maize seeds, healthy maize seeds were disinfected as described above and incubated for 5 min in 5 mL of a suspension of 10⁴ spores/mL of F. verticillioides, or F. subglutinans, and of F. verticillioides together with 10⁴ spores/mL of F. subglutinans. The artificially contaminated seeds were inoculated in Petri dishes containing MGA-penicillin-streptomycin medium [27], prepared as described above. These seeds were incubated for five days at 25 °C on the medium, with a period of 12 h of light. Naturally infected maize seeds (Ponta Grossa-1, Rio Verde 27, and Maringá-10) showing signs of rot were disinfected and incubated in the same way. The mycelia growing on the seeds and on the agar around the seeds were scraped and used for the DNA extraction that was conducted with the method described by Koenig et al. [32]. To test if maize DNA could inhibit the reaction, maize DNA was also extracted from health maize seeds using the protocol described by Koenig et al. [32] and this DNA was added in multiplex PCR reactions with DNA from the reference isolates of F. verticillioides and F. subglutinans.

The multiplex reactions were performed as shown above, except that they contained 25 pmol of each primer. The annealing temperatures are listed in Table 3. The amplification of specific fragments was analyzed by running 10 μL of the PCR reaction on a 1.5% agarose gel that was stained with ethidium bromide (0.25 μg/mL).