Texas maize samples from the 2011 harvest were obtained by the Office of the Texas State Chemist, following the One Sample Strategy (Office of the Texas State Chemist, 2012). Briefly, 2.2 kg maize samples per truck were collected at approved grain elevators, probing maize loads 7-to-9 times with a 72 cm spiral hand probe using representative sampling patterns described previously (Federal Crop Insurance Corporation, 2010). Once collected, the entire 2.2 kg sample was ground in a Romer mill (Romer Labs) until at least 70% of the particles pass through a No. 20 wire woven mesh sieve. Using a spatula, the ground sample was then stirred for 30 s to ensure a homogenous blend and 50 g subsamples were weighed, placed and marked in plastic bags. Bags were stored at −20°C until further studies were conducted.

A minimum of three bags of pulverized maize per county were randomly chosen when sufficient samples were available. Pulverized maize (100 mg) was transferred to a tube containing sterile water (1 ml) and vortexed for 10 s. We further diluted the sample by transferring 10 μl of the solution into a new tube containing 1 ml of sterile water. The diluted sample (50 μl) was then spread onto petri dishes containing 0.2× potato dextrose agar (PDA) medium amended with ampicillin and streptomycin (50 mg/ml) and incubated for four days at room temperature. This process was repeated with each pulverized maize bag. Once fungal growth was observed, individual colonies were isolated, subcultured as a single microconidium, transferred to individual petri dishes and incubated at room temperature. Isolates exhibiting the standard F. verticillioides phenotype (Leslie and Summerell, 2006) were further analyzed. At least three independent isolates per county, when possible, were used in this study. F. verticillioides wild-type strains 7600 (MAT1-1) and 7598 (MAT1-2) (Fungal Genetics Stock Center) as well as our isolated F. verticillioides strains were stored in 30% glycerol (v/v) at −80°C in this study.

Yeast extract peptone dextrose (YEPD) broth (5 ml) was poured into a petri dish (60×15 mm) and inoculated with an agar plug (0.5 cm in diameter) containing fungal mycelium. Plates were incubated at room temperature without agitation for 5 days. Once time lapsed, the broth was vacuumed out and the fungal tissue was flash frozen in liquid nitrogen. DNA was extracted using the OmniPrep kit (G-Biosciences) following the manufacturer’s recommended protocol. DNA concentrations were determined using a NanoDrop 2000 system (Thermo Fisher Scientific) and DNA quality was visually inspected by gel electrophoresis. DNA working solutions (150 ng/μl) were used for all PCR in this study unless otherwise stated.

All PCR analyses were performed using a 25 μl final aliquot volume of the amfiEco Taq DNA Polymerase kit (Gendepot) following the manufacturer’s recommendation in either a GeneAmp PCR system 9700 (Applied Biosystems). Amplification of mating type genes (MAT1-1 and MAT1-2) as well as the FUM1 gene was carried out using primer pairs listed in Table 1. All three sets of PCR were carried out at 52°C annealing temperature and 1 min extension for 35 cycles. RAPD markers were produced with a single primer (Table 1) with an annealing temperature of 35°C and 2 min extension. The final concentration of all primers was 15 pmol. Electrophoresis of standard PCR amplicons and RAPD products was carried out in standard 1% and 1.5% agarose gels in 0.5× TBE buffer, respectively, with subsequent ethidium bromide staining/destaining and photography under UV light. DNA samples from wild-type strains, F. verticillioides 7600 and 7598, were used as controls.

The RAPD profile of each isolate was obtained by comparing their individual marker pattern to the patterns of all the samples across the population produced with the same primer. Variations in band intensity of such markers was disregarded and a binomial system, where 1 is positive and 0 is negative, was used to score for the presence or absence of markers generated with a given primer. Once the OPA-01, OPB-01 and UBC-18-generated profiles of all the samples were scored, the 3 profiles per isolate were consolidated into one in an OPA-OPB-UBC fashion. Using the DendromUPGMA utility (http://genomes.urv.cat/UPGMA/) described by Garcia-vallvé et al. (1999), a similarity matrix of the RAPD profiles was computed using Dice similarity coefficient, 100 bootstrap replicates and cluster analysis was performed using an unweighted pair-group method using an arithmetic averaging (UPGMA) algorithm. A dendrogram was then drawn using Phylogeny (http://www.phylogeny.fr) (Dereeper et al., 2008). Geographical maps of the State of Texas were generated using the diymaps utility (http://diymaps.net/tx.htm).

One g of cracked maize was placed in glass vials, rehydrated in water overnight and then autoclaved. Each vial was then independently inoculated with 1×10⁶ spores of F. verticillioides wild-type strain 7600 and strains Bailey #140B, Floyd #61B, Bailey #141B, Castro #72, Parmer #22B, Lamb #135B, Dallam #70, Milam #217, Moore #237A or Hale #29A isolated from Texas panhandle area counties. Vials were then placed in an incubator for 10 days at 24°C with a 14 h light/10 h dark cycle. Subsequently, extraction of FB1 and ergosterol was accomplished by adding 5 ml of acetonitrile:water (1:1, v/v) and chloroform:methanol (2:1, v/v), respectively, to the vials and incubated at room temperature overnight without agitation. Purification of extracts was performed following protocols described by Shim and Woloshuk (1999) where FB1 crude extracts (acetonitrile:water vials) are passed through equilibrated HyperSep C18 SPE columns (Thermo Fisher Scientific) and subsequently washed with water and 15% acetonitrile and then eluted in 70% acetonitrile. On the other hand, ergosterol crude extracts (chloroform:methanol vials) were passed over Acrodisc 13 mm nylon 0.45 mm filters (Pall Life Sciences). HPLC analyses of FB1 and ergosterol were performed as described elsewhere (Kim et al., 2011; Shim and Woloshuk, 1999). FB1 and ergosterol concentrations were quantified by comparing the peak area to that of FB1 and ergosterol standards (25 ppm), respectively. FB1 levels were then normalized to fungal growth using Kim et al. (2011) equation where FB1 biosynthesis = [Fb1 production (μg/g)/ergosterol contents (μg/g)]×100. Three biological replicates were performed per sample. Standard error was calculated based on the average values of wild type replicates and the average values of the 10 field isolates.

Maize samples from 65 counties, representing 24.8% of the total counties in the state (254 counties), were analyzed for the presence of F. verticillioides. We isolated 164 F. verticillioides strains using standard identification techniques described previously (Leslie and Summerell, 2006). Cultures exhibited white aerial mycelia growing in uniform radial form and developed pigments ranging from light pink to violet gray with age. We isolated at least 3 samples from 50.7% of the counties surveyed while we were able to isolate one or two samples in 49.2% of the counties. Samples from some counties (13.8% of the total) yielded 4 to 7 F. verticillioides isolates.

Out of the 164 fungal isolates tested, 82 contained the MAT1-1 mating type locus while the remaining 82 tested positive for the MAT1-2 locus, therefore presenting an equal mating-type distribution in all isolates tested. All isolates from Dallam, Gray, Hartley, Hidalgo, Hill, Hutchinson, Swisher and Victoria counties corresponded to the MAT1-1 mating type while all those from Hale, Live Oak, Medina, Wharton and Wilson counties were only positive for the MAT1-2 mating type. The remaining 51 counties showed mixed mating-type distribution (Fig. 2b). However, it is important to note that this outcome does not indicate that one mating-type dominates certain counties since we tested a small population sample of F. verticillioides isolates from each county.

All F. verticillioides samples isolated in this study were positive for the FUM1 gene, which encodes a polyketide synthase critical for fumonisin biosynthesis. The FUM1 gene is the central component of the FUM gene cluster and is often used as a diagnostic marker for fumonisin-producing Fusarium species (Shim and Woloshuk, 2001; Shim et al., 2003). This result suggests that all F. verticillioides isolates found in the Texas maize production fields are capable of producing fumonisins under environmental conditions that favor mycotoxin biosynthesis.

Using RAPD techniques, while disregarding variations in band intensity, we identified 27 polymorphic markers using 3 different primers previously utilized for F. verticillioides diagnostics (Desjardins and Plattner, 2000; Moller et al., 1999). Ten markers were generated when using the OPA-01 primer, eleven markers with the OPB-01 primer and six markers with the UBC-18 primer (Fig. 3). Polymorphisms were scored by the presence or absence of a given marker resulting in a binomial DNA fingerprint. We used the unweighted pair group method with arithmetic mean (UPGMA) with the Dice coefficient (Garcia-Vallvé et al., 1999) to analyze the RAPD fingerprints and cluster the samples based on similarity. We then generated a dendrogram which was essential for determining the clustering patterns.

In order to observe the relationship between the RAPD profile data and the geographic point of origin of each sample, we color-coded the clusters of Texas counties into six regions (Fig. 4a), which corresponds to the map of the Texas Health Service regions. Strains in the dendrogram were then color-coded to reflect their region of origin (Fig. 4b). The three samples originating from Burleson County clustered together. This trend was also observed with all samples originating from Dallam, Goliad and Fort Bend Counties (Fig. 4a-2), but neither of the counties grouped together. Clusters of two samples per county were also observed in McLennan, Lamb, Frio, Lamar, Brazoria, Moore, Uvalde, Comal, Bell and Medina counties. Finally, mating type data was added onto the dendrogram (data not shown) and, similar to earlier results, no clear correlation between RAPD profile and geographic location was observed.

Ten F. verticillioides cultures isolated from maize samples from Texas Panhandle counties were randomly selected for in vitro FB1 production assays. Notably, in the 2011 annual report published by the Office of the Texas State Chemist, maize samples collected from this region did not show detectable levels of fumonisins (http://mycotoxinbmps.tamu.edu/mapsupdate.aspx). Thus, we tested whether F. verticillioides strains we isolated from the region’s maize samples were non-toxigenic or they simply failed to produce fumonisins due to unfavorable environmental conditions. Our results showed that FB1 was produced by all samples tested with detection values ranging from 62.313 (Bailey #140B) to 183.312ppm (Dallam #70) while F. verticillioides 7600 produced 50.603ppm. FB1 values were normalized to endogenous ergosterol levels, compounded and graphed (Fig. 4).