Quantitative molecular diagnostic assays of grain washes for Claviceps purpurea are correlated with visual determinations of ergot contamination

Abstract

We examined the epiphytic microbiome of cereal grain using the universal barcode chaperonin-60 (cpn60). Microbial community profiling of seed washes containing DNA extracts prepared from field-grown cereal grain detected sequences from a fungus identified only to Class Sordariomycetes. To identify the fungal sequence and to improve the reference database, we determined cpn60 sequences from field-collected and reference strains of the ergot fungus, Claviceps purpurea. These data allowed us to identify this fungal sequence as deriving from C. purpurea, and suggested that C. purpurea DNA is readily detectable on agricultural commodities, including those for which ergot was not identified as a grading factor. To get a sense of the prevalence and level of C. purpurea DNA in cereal grains, we developed a quantitative PCR assay based on the fungal internal transcribed spacer (ITS) and applied it to 137 samples from the 2014 crop year. The amount of Claviceps DNA quantified correlated strongly with the proportion of ergot sclerotia identified in each grain lot, although there was evidence that non-target organisms were responsible for some false positives with the ITS-based assay. We therefore developed a cpn60-targeted loop-mediated isothermal amplification assay and applied it to the same grain wash samples. The time to positive displayed a significant, inverse correlation to ergot levels determined by visual ratings. These results indicate that both laboratory-based and field-adaptable molecular diagnostic assays can be used to detect and quantify pathogen load in bulk commodities using cereal grain washes.

About the Authors:

Alexia Comte

Affiliation: Agriculture and Agri-Food Canada Saskatoon Research and Development Centre, Saskatoon, Saskatchewan, Canada

Tom Gräfenhan

Affiliation: Grain Research Laboratory, Canadian Grain Commission, Winnipeg, Manitoba, Canada

Matthew G. Links

Affiliations Agriculture and Agri-Food Canada Saskatoon Research and Development Centre, Saskatoon, Saskatchewan, Canada, Department of Computer Science, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Sean M. Hemmingsen

Affiliations National Research Council Canada, Saskatoon, Saskatchewan, Canada, Department of Microbiology and Immunology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Tim J. Dumonceaux

* E-mail: [email protected]

Affiliations Agriculture and Agri-Food Canada Saskatoon Research and Development Centre, Saskatoon, Saskatchewan, Canada, Department of Veterinary Microbiology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

ORCID http://orcid.org/0000-0001-5165-0343

Introduction

Claviceps purpurea (Fr.) Tul. (ergot) is a fungus that infects cereal crops such as rye (Secale cereale), wheat (Triticum aestivum), and durum (Triticum durum). The life cycle of this fungal pathogen includes the formation of fungal sclerotial bodies in place of normal seeds [1]. The production by C. purpurea of toxic and hallucinogenic alkaloids including ergometrine, ergotamine tartrate, and lysergic acid diethylamide causes a disease known as ergotism upon consumption of grain products affected by C. purpurea [1,2]. While cases of humans affected by ergot are rare today, ergot alkaloid contamination represents a serious problem in agriculture since grazing cattle are highly susceptible. Therefore, ongoing surveillance and knowledge of baseline levels of ergot contamination in grain products destined for export is critical to support cereal producers and exporters, who are at economic risk from ergot contamination of their crops.

Since mycotoxins produced by C. purpurea are the source of concern related to ergot contamination, several methods have been developed for the detection and determination of ergot alkaloids in grains, grasses, feeds, and foods. These include planar solid phase extraction [3], enzyme-linked immunosorbent assay [4] and liquid chromatography-tandem mass spectrometry [5]. Alternatively, since ergot produces dark purple sclerotia that take the place of seed in an infected plant, harvested grain can be easily graded for ergot contamination by determining the proportion (weight basis) of sclerotia by visual inspection. While this is a simple, low-cost method, it can be time consuming and labor intensive and may miss small sclerotial bodies. Using this approach, and acknowledging that ergot is ubiquitous in the environment, tolerance levels have been set for ergot sclerotia in grain products that affect prices for both producers and exporters [2]. However, recent reports of Egyptian authorities shifting towards more stringent guidelines and regulations for ergot sclerotia in grain shipments [6] has emphasized the need for producers and exporters to be knowledgeable about the microorganisms associated with these commodities and for tolerance levels to be set and respected. This applies equally to pathogens besides ergot that are not as easily detected but could be a source of trade disputes or used as a criterion for novel grading standards.

Molecular diagnostic methods offer a means to detect and quantify microorganism (including pathogen, symbiont, and commensal) DNA on plant material pre- and post-harvest, with the potential to provide a “molecular grade” for a grain products based on the presence and/or abundance of particular microorganisms. Moreover, molecular-based approaches may be more accurate and sensitive depending on the sample size of plant material that is tested. Seed washing followed by deep sequencing of the chaperonin-60 (cpn60) molecular barcode [7–9] can provide a detailed picture of the bacterial and fungal microbiota of these environments, along with quantitative data that correlates to biological activity [10]. Alternatively, specific microorganism-targeted molecular diagnostics can be used to quantify bacterial or fungal taxonomic markers in these same environments [10].

In this work, we hypothesize that DNA from grain-associated ergot bodies of Claviceps purpurea is detectable using a simple seed wash and that quantitative molecular diagnostic assays are correlated with visual determinations of ergot contamination. We assessed the molecular-based detection of Claviceps purpurea through microbiota profiling and by specific quantitative diagnostics. We have provided a proof-of-principle for molecular grading by examining harvested grain products for the presence of C. purpurea. The results provide a framework for the development of molecular diagnostic tools to provide a “molecular grade” for ergot or other pathogens via the detection and quantification of target DNA from grain samples, with implications for trade in agricultural commodities and suitability of cereal grain for food and feed especially at the on-farm level.

Materials and methods

Grain sources

Grain samples were collected from the harvest sample program of the Canadian Grain Commission; therefore, no specific permissions were required. These studies did not involve any endangered or protected species. Ten samples of cereal grain from various locations that were downgraded for different factors (Table 1) were initially chosen for microbiome profiling using chaperonin-60 (cpn60) amplicon sequencing [11]. Subsequently, a series of 141 harvest samples from the 2014 crop year (Fig 1 and Table 2) that were rated for ergot contamination (weight/weight proportion of ergot sclerotia in the seed sample) were selected for detection and quantification of ergot DNA by the molecular diagnostic methods described below. Most grain samples were either wheat or durum, while a small number of non-cereal grain samples were also included. Samples were stored separately in plastic bags at room temperature before they were processed.

[Figure omitted. See PDF.]

Fig 1. Ergot sclerotia observed in sample 9129 (Rye).

Sclerotia are indicated by arrows. This sample had an ergot severity rating of 0.294% on a percentage weight basis (Table 2). Scale bar indicates 1 cm.

https://doi.org/10.1371/journal.pone.0173495.g001

[Figure omitted. See PDF.]

Table 1. Seed samples selected for microbiome profiling using cpn60.

https://doi.org/10.1371/journal.pone.0173495.t001

[Figure omitted. See PDF.]

Table 2. Quantification of C. purpurea DNA in grain wash samples using ITS-targeted ddPCR and cpn60-targeted LAMP.

https://doi.org/10.1371/journal.pone.0173495.t002

DNA extraction from grain-associated epiphytic microbiota and profiling the microbial communities

DNA was extracted from grain washes as described [10]. Briefly, 25 g subsamples of grain were soaked in 45 ml of buffered peptone water containing 0.05% Triton X-100 (Sigma, St. Louis, MO) in a 250 ml Erlenmeyer flask at room temperature with shaking (150 rpm) for 1 hour. The liquid fractions were centrifuged at 4000 × g for 15 minutes and the supernatant discarded. Pellets were resuspended in 200 μl of TE buffer and subjected to DNA extraction using a previously described bead-beating protocol [12]. DNA was quantified using a Quant-IT DNA quantification kit and Qubit fluorometer (Invitrogen, Burlington, Ontario). To account for the possible presence of PCR inhibitors in the extract, a dilution series of one of the grain wash samples was prepared and each dilution used as a template for cpn60 universal PCR as described [10]. The dilution that provided the strongest bands upon agarose gel electrophoresis (1:5) was used as template for cpn60 amplicon generation and sequencing. Purified, concentrated amplicons from all grain samples were pooled on an equimolar basis prior to emPCR adaptor ligation and pyrosequencing using Titanium chemistry (Roche) as described previously [8,13]. For quantification of ergot in grain using molecular diagnostic assays, DNA was prepared in a similar manner except that 10 g of grain were used and the final solution was further purified using Agencourt AMPure XP beads (Beckman-Coulter) at 1:1 (v/v) bead:DNA solution according to the manufacturer’s recommendations. DNA samples were eluted from the beads with 30 μl of 10 mM Tris-Cl pH 8.0.

Determination of taxonomic marker gene sequences from C. purpurea and ergot sclerotia

Sclerotia of ergot (Table 3) were obtained from field-grown samples from Manitoba, Canada. Sclerotia were crushed in liquid nitrogen using a mortar and pestle and the powder was subjected to a bead beating protocol as previously described [12], or using a DNeasy Plant DNA mini kit (Qiagen). The final volume of DNA solution was 200 μl. The DNA was used as template for amplification of the fungal internal transcribed spacer (ITS) fungal barcode using primers ITS4/ITS5 as previously described [14]. PCR products of ~625 bp were cloned into pGEM-T Easy (Promega) according to the manufacturer’s recommendations, and individual colonies were selected from each sample for sequencing. To amplify the Cpn60-encoding gene [15], PCR primers targeting fungal cpn60 sequences (S1 Table) were used to generate PCR products from sample erg0256 (Table 3). PCR for cpn60 used 1x Taq buffer (Invitrogen); 2.5 mM MgCl2; 500 nM each dNTP; 400 nM each primer; and 1 U Taq DNA polymerase (Invitrogen). The sequences of these individual amplicons spanned regions upstream and downstream of the cpn60 universal target (cpn60 UT) [15], but contained a gap. To determine the complete sequence of the C. purpurea cpn60 UT, primers based on these sequences were designed to amplify a product of 741 bp that contained the entire cpn60 UT (S1 Table; concentrations of all components in PCR as described above). These 741 bp amplicons were generated from all samples and were cloned into pGEM-T Easy. After sequence determination from individual clones, the 555-bp UT region was manually extracted and used for analysis. To determine the reference cpn60 and ITS sequences of C. purpurea, reference strain 714 was obtained from Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ). The organism was cultured on 10 cm petri dishes containing YpSs medium (4 g/L yeast extract, 15 g/L soluble starch, 0.5 g/L MgSO4 · 7 H2O, 1.0 g/L KH2PO4, and 15 g/L agar) overlayed with a sterile 0.1 μm polycarbonate membrane filter (Sterlitech). After 14 d at 25°C, a colony of approximately 2 cm diameter was retrieved from the plates using sterile forceps and crushed in liquid nitrogen. Aliquots of the powder (~100 mg each) were used for DNA extraction with a DNeasy Plant Mini kit (Qiagen). DNA so obtained was used as a template for PCR with ITS primers [14] and with primers D0282/D0283 (S1 Table) to determine the reference ITS and cpn60 UT sequences, respectively.

[Figure omitted. See PDF.]

Table 3. ITS and cpn60 clone diversity observed in sclerotia sourced from Manitoba, Canada.

https://doi.org/10.1371/journal.pone.0173495.t003

Quantitative PCR for C. purpurea based on ITS sequences

A representative ITS sequence obtained from an ergot sclerotium (erg0256; see Table 3) was used as a basis for hydrolysis probe assay design using Beacon Designer v.7.90 (Premier Biosoft, Palo Alto, CA) (S1 Table). Amplification primers and the hydrolysis probe were purchased from Integrated DNA Technologies (Coralville, IA). To determine the PCR efficiency, a set of standards was prepared using ITS-containing plasmid DNA. DNA was prepared using a miniprep kit (Qiagen), and the DNA was linearized using PstI. The concentration of linearized plasmid DNA was determined in triplicate using a Qubit instrument (BR kit, Life Technologies). The mean concentration (ng/μl) was converted to copies/μl using an approximate molecular weight of 650 g/mol per base pair. This solution was diluted to provide 107–101 copies per assay and used as control templates in qPCR. Reactions used SsoFast Universal probes supermix (Bio-Rad, Mississauga, ON, Canada) in a 20 μl final volume with 300 nM of each primer and 200 nM of probe. Amplification was carried out using a CFX96 real-time system with a C1000 base (Bio-Rad) and reactions were quantified using BioRad CFX manager software (v.3.1). The slope of the line resulting from plotting threshold cycle (Cq) values vs. log10 copy number was used to determine PCR efficiency according to E = 10(-1/slope), where 2.0 is theoretical [16]. To obtain quantification results that were independent of standards, the assay was adapted to the droplet digital PCR (ddPCR) format. Reaction conditions for ddPCR were first optimized using gradient PCR (54–65°C). Reactions used ddPCR supermix for probes (Bio-Rad) and had 900 nM of each primer and 250 nM of hydrolysis probe in a 20 μl reaction volume. The accuracy of ddPCR quantification was examined using a dilution series of known copy numbers (standard curve prepared as described above). To quantify C. purpurea in intact grain wash extracts, template DNA prepared as described above was digested using EcoRI (37°C, 60 min, then 85°C, 5 min) and 2 μl of the digested DNA was added to the ddPCR mixture. Emulsions were prepared prior to amplification using a QX100 droplet generator (Bio-Rad), and amplifications were done using a C1000 Touch thermocyler (BioRad). After amplification, positive and negative droplets were quantified using a QX100 droplet reader (Bio-Rad) and the proportion of negative droplets was converted to copies per well using QuantaSoft v.1.6.6 (Bio-Rad). Results reported by QuantaSoft were converted to copy number/g grain extracted by correcting for sample preparation. In cases where very high counts were observed, samples were diluted accordingly.

Detection of C. purpurea DNA in seed washes using Loop-Mediated Isothermal DNA Amplification (LAMP) based on cpn60

Amplification primers for LAMP (S1 Table) that targeted cpn60 of C. purpurea were designed using LAMP Designer v. 1.12 (Premier Biosoft, Palo Alto, CA). LAMP conditions were as described for detection using calcein [17]. The same primers were also used for amplification and detection using Isothermal Mastermix (Prolab Diagnostics, Richmond Hill, ON, Canada), which features proprietary detection chemistry that also facilitates the determination of product annealing temperature. For both detection chemistries, a temperature of 63°C was used for amplification. Reactions were monitored in real time using a Genie II or Genie III instrument (OptiGene, Horsham, UK) and the time to positive (Tp) was reported by the instrument.

Assay parameters

The performance characteristics of the molecular diagnostic assays were determined according to established standards [18,19]. Analytical specificity was examined by using DNA isolated from a panel of fungi that grow in association with one or more of the field crops typically grown in Canada, including Alternaria spp, Fusarium spp., Stemphylium sp., Rhizoctonia sp., Plectosphaerella sp., Leptosphaeria spp., and Verticillium sp. To determine the limit of detection (LOD) of the LAMP assay, a series of 6 serial dilutions of C. purpurea DNA was added to grain wash DNA that was determined to lack detectable ergot DNA and a total of 70 replicates were analyzed using probit (SPSS). The weight of DNA (ng) added to each assay was converted to genome equivalents using a genome size of 32.1 Mbp [20] and 650 g/mol per base pair. The LOD was defined as the number of C. purpurea genome equivalents that yielded positive results 95% of the time [18]. The linearity of the LAMP assay was examined by polynomial regression analysis of the Tp determined over a wider range of dilutions and including 3–14 replicates in each dilution. Intra-assay precision was determined by calculating the coefficient of variation of the Tp determined at each of three levels (near the LOD, at twice the LOD, and at 20 times the LOD). Finally, assay sensitivity and specificity were determined by scoring the numbers of positive and negative results obtained using visual inspection as a gold standard and calculating these parameters as described [19].

Results

Microbial community profiling of downgraded grain

The epiphytic microbiota of grain that had been downgraded for various factors including Fusarium, mildew, ergot, and midge damage [11] were profiled. DNA extracts representing seed epiphytic microorganisms were characterized by sequencing PCR amplicons using pyrosequencing. A total of 355715 reads produced 3609 assembled, unique cpn60 UT sequences (operational taxonomic units, OTU) after processing with mPUMA [21]. In addition to the bacterial OTU that were observed, sequences similar to Sordariomycetes such as Cylindrocarpon, Magnaporthe, Fusarium, and Verticillium spp were observed but they had relatively low sequence identities to known strains (~85%). Two OTU in particular, OTU02794 and OTU03634, clustered with fungal sequences from cpnDB but appeared to occupy a gap in the reference database. These fungal OTU were found at low levels in several of the datasets (Table 1), but they could not be further identified due to the lack of reference data.

Identification of fungal OTU as C. purpurea

BLAST analysis showed that the ITS sequences that were determined from 5 ergot sclerotia (Table 3) were closely related to previously reported ITS sequences from C. purpurea, with > 99% identity observed (data not shown). This suggested that these sclerotia consisted primarily of C. purpurea DNA, but as these were environmental samples it was assumed that DNA from other organisms was also present. Examining the cpn60 sequences generated from these sclerotia revealed that these sequences clustered closely with the sequence obtained from the reference strain of C. purpurea (DSMZ 714), and with a cpn60 sequence that was extracted from a genome sequence of C. purpurea [20] (Fig 2). Thus, the cpn60 sequences determined from the ergot sclerotia indeed corresponded to the sequence of C. purpurea cpn60. In addition, the previously unidentified OTU from the microbiome profiling of the downgraded grain lots were thereby identified as deriving from C. purpurea (Fig 2).

[Figure omitted. See PDF.]

Fig 2. Phylogenetic analysis of cpn60 UT sequences derived from microbial profiling and ergot sclerotia compared to reference sequences.

Sequences are prefixed by cpnDB ID number (www.cpndb.ca) and GenBank accession numbers (www.ncbi.nlm.nih.gov) are provided in parentheses where available. The tree was calculated using the Maximum Likelihood method based on the Tamura-Nei model [22] using MEGA6 [23]. The tree was bootstrapped (100 iterations) and numbers next to the nodes indicate the percentage of trees in which the associated taxa clustered together. Branch lengths correspond to the number of substitutions per site.

https://doi.org/10.1371/journal.pone.0173495.g002

Ergot sclerotia contain many copies of cpn60 and ITS

Each of the five ergot sclerotia contained 3–10 distinct copies of ITS (Table 3), all of which had high sequence identities with previously reported ITS sequences for C. purpurea [24]. The cpn60 sequences revealed an even higher heterogeneity in most samples, with 2–16 distinct sequences observed (Table 3, S1 Fig). Like the ITS sequences, the sclerotia cpn60 sequences were closely related to sequences determined from the C. purpurea reference strain (DSMZ 714) and the C. purpurea genome. To determine the likely copy number of cpn60 and ITS within the C. purpurea genome, we used representative sequences from ergot sclerotia to query the genome sequence by BLAST. Using a scelerotium-derived ITS sequence as query for blastn, only a single hit was observed in the C. purpurea genome (GenBank CAGA00000000.1), with an e-value set at 1000 (data not shown). Similarly, using tblastx, only a single match was observed in the C. purpurea genome using the translated amino acid sequence of cpn60 as query (data not shown).

Quantitative molecular diagnostic assays of grain washes were correlated with visual assessments of ergot load

The qPCR assay that was designed to detect the ITS sequence of C. purpurea was highly efficient (E = 2.04; r2>0.999), and detected as few as 10 copies of target DNA per reaction (Fig 3A). Similarly, the ddPCR-adapted version of the assay was highly accurate; the number of copies added to each assay was reported correctly at levels of input ITS copy numbers ranging from 10 to 105 copies (Fig 3B).

[Figure omitted. See PDF.]

Fig 3. ITS-targeted qPCR assay linearity assessed by standard curve (A) or ddPCR calibration curve (B).

https://doi.org/10.1371/journal.pone.0173495.g003

The ITS-targeted ddPCR assay was positive with all grain wash samples, including 37 unrated/negative samples and four samples from canola, which is not a host for C. purpurea (Table 2). Although the ddPCR assay did not yield a signal with the reference fungi used to determine analytical specificity, this observation suggested that the assay may generate a signal with nontarget grain-associated microorganisms. Nevertheless, a positive, highly significant correlation was observed between the number of C. purpurea genomes detected by ddPCR and ergot severity of a percentage weight basis (Table 4).

[Figure omitted. See PDF.]

Table 4. Spearman rank correlation between ergot severity (% weight basis) and molecular quantification of ergot DNA in grain wash templates.

https://doi.org/10.1371/journal.pone.0173495.t004

The cpn60-targeted LAMP assay in both detection formats was apparently specific, as none of the fungal isolates examined provided evidence of cross-reactivity (data not shown). In addition, the samples that had been profiled by cpn60 sequencing were examined using LAMP and none of the samples that lacked C. purpurea reads were positive in the LAMP assay (Table 1). The LOD of the LAMP assay with Isothermal detection chemistry was approximately 75 genome equivalents. The Tp of the assay in both formats showed a strong, inverse correlation to input template amount, and the Isothermal detection format was positive in less than 10 minutes at higher template amounts (Fig 4). The slopes of the two detection chemistries were quite different, with calcein detection featuring a steeper curve and much slower detection compared to the Isothermal detection chemistry (Fig 4). The LAMP assay also featured a reasonable intra-assay variability at the three input levels examined (Table 5).

[Figure omitted. See PDF.]

Fig 4. cpn60-targeted LAMP assay linearity assessed by expressing Tp related to C. purpurea genome copies using the two LAMP detection systems evaluated in this study.

The equations for each curve are: y = -9.03x+96.98 (calcein detection) and y = -1.90x+19.44 (isothermal detection). The correlation coefficients (r2) are 0.99 (calcein detection) and 0.95 (isothermal detection).

https://doi.org/10.1371/journal.pone.0173495.g004

[Figure omitted. See PDF.]

Table 5. Intra-assay reproducibility of the cpn60-targeted LAMP (isothermal detection).

https://doi.org/10.1371/journal.pone.0173495.t005

Unlike the ITS-targeted ddPCR assay, the LAMP assay targeting cpn60 was not positive in all of the grain wash samples analyzed. With calcein detection, 16 of the 141 samples, including the four canola grain wash samples, tested negative (Table 2). However, using the apparently more sensitive Isothermal detection format (Fig 4), only 5 samples (canola -7; canola-10; buckwheat-201; wheat-2875; and rye-8094) tested negative. To determine if samples providing discordant results contained amplifiable C. purpurea DNA, the calcein-negative grain wash templates were amplified using primers D0282-D0283 (S1 Table). Amplicons were generated for samples canola-9, canola-10, wheat-9952, and wheat-9985. The sequences of the amplicons indicated that the wheat samples indeed contained C. purpurea genomic DNA, while the canola samples yielded amplicons that did not correspond to C. purpurea (data not shown). These observations indicate that the LAMP assay in both formats, as well as the PCR assay, can generate false positive and false negative results, although in the case of the Isothermal detection format the false positive results were very late and could be avoided by reducing the assay time.

Despite this, the LAMP assay (calcein detection) generated quantitative data that correlated inversely with ergot severity, with a statistically significant p-value (Table 4). Moreover, the sensitivity of the LAMP assay using visual rating as a gold standard was very high (0.97 using calcein detection or 0.99 using Isothermal detection), with only 3/100 (calcein) or 1/100 (Isothermal) positive samples generating a negative result with LAMP (Table 6). This is consistent with a low false negative rate, or type I error [19]. Conversely, the specificity of the LAMP assay was apparently low in both formats, suggesting a high false positive rate (type II error). With Isothermal detection, more of the grain wash samples tested positive, consistent with the increased analytical sensitivity of the assay in this format (Fig 4). Like the calcein detection format, the Tp of the LAMP assay with Isothermal detection chemistry was inversely, significantly correlated to ergot levels determined using visual inspection (Table 4).

[Figure omitted. See PDF.]

Table 6. Sensitivity and specificity of the C. purpurea cpn60-targeted LAMP assay compared to visual rating (gold standard).

https://doi.org/10.1371/journal.pone.0173495.t006

Discussion

Claviceps purpurea is a pathogen of grasses and cereals that co-evolved with its host in the Cretaceous period, at least 100 million years ago [25]. The pathogen is therefore expected to be ubiquitous in the environment; this fact, combined with the danger posed to humans and animals associated with the consumption of ergot alkaloids, makes ongoing surveillance necessary to protect cereal grain and end products from ergot contamination. While laboratory-based methods can quantify ergot alkaloids effectively, the level of ergot contamination in harvested grain can be determined visually by picking and weighing sclerotia from representative subsamples (e.g. Fig 1) of most cereals. One exception to this is canary seed, which is not typically rated due to seed size, but is also not normally considered for human consumption. In Canada, canary seed is not considered an official grain, and the industry is free to establish its own quality criteria. However, a recently registered hairless variety of canary seed is intended for human consumption (http://www.hc-sc.gc.ca/fn-an/gmf-agm/appro/canary-seed-lang-graine-alpiste-decision-eng.php) [26], so it is expected that there will be some need for ergot determination on canary seed in the near future. This would likely be done on ground material rather than grain washes.

The detection of ergot and other pathogen DNA in grain wash samples can be readily accomplished using sequencing methods, which can provide a profile of the bacterial and fungal microbiome associated with the grain lot under analysis [10]. This method has the major advantage of being non-targeted; rather than querying a sample for the presence of a particular pathogen, microbiome profiling can determine if the sample under analysis contains DNA from any potential pathogen of concern. However, despite all of the improvements in sequencing technology, screening hundreds of samples in this way would still be a difficult and rather expensive undertaking. Moreover, the results can be somewhat ambiguous. For example, a nucleotide sequence identity or read abundance cutoff may need to be established for various pathogens to determine an actionable quarantine pathogen detection threshold. The consequences of such decisions can be important, especially regarding grade, end use markets and trade measures. Finally, as we have shown, the success of a microbiome profiling method for detecting pathogen DNA is limited by the breadth of the reference database, since a sequence can only be identified by comparison to known reference sequences.

We determined the microbial profile of a range of cereal grain samples that had been downgraded for various reasons, including ergot contamination. Reads from sequencing datasets that were initially unidentified were found to correspond to C. purpurea cpn60, which emphasizes the importance of continuously enhancing reference databases. Despite the fact that reads corresponding to C. purpurea were detected in the samples profiled using cpn60 universal PCR, purified DNA from ergot sclerotia as well as from the reference strain of C. purpurea failed to generate a cpn60 amplicon using these same primers; a modified set of universal PCR primers targeted to fungal cpn60 was required for this purpose (S1 Table). These observations are consistent with our previous work with Alternaria and Fusarium [10], and may be explained in part by the difference in primer/template ratios between complex template and single-template PCR. Nevertheless, we successfully amplified and sequenced cpn60 and ITS from ergot sclerotia, and the detection of multiple related copies of both genes suggests that a single sclerotium contained a cluster of related strains of C. purpurea rather than a single strain. This is supported by the fact that a single copy of both genes was found in the genome of C. purpurea.

We have investigated the feasibility of applying targeted molecular diagnostic assays to the detection of C. purpurea DNA from grain samples. We used unground grain samples in the current study, which leaves ergot sclerotia intact. Despite this, we were able to detect C. purpurea DNA in grain washes of samples that were known to be contaminated with ergot. It is unknown if grinding the seed samples, which may release more C. purpurea DNA from contaminating sclerotia but would also release large quantities of host DNA and potentially interfering starch, which may have facilitated or hampered the detection. Nevertheless, the ITS-targeted ddPCR assay generated quantitative results that were highly correlated to ergot severity as determined by visual rating, but all samples tested positive. This made the calculation of assay sensitivity and specificity compared to visual rating impossible and suggests that the ddPCR assay we described suffered from a low analytical specificity, which is observed when nontarget analytes generate a signal [18]. In contrast, the cpn60-targeted LAMP assay was apparently discriminatory for C. purpurea DNA, since neither the nontarget genomic DNA templates nor the C. purpurea-negative samples profiled by sequencing (Table 1) generated a signal. This is not surprising, since LAMP is thought generally to feature a higher analytical specificity than PCR [17]. The LAMP assay is also rapid, inexpensive, and adaptable to use outside of the laboratory environment; these advantages make it analogous to visual-based rating. Moreover, the assay we have described generated quantitative data that correlated strongly and inversely with ergot severity, irrespective of the detection chemistry used.

The cpn60-LAMP assay featured a very low false negative rate compared to visual rating; in other words, virtually all samples in which ergot sclerotia were visually observed tested positive. The apparently high false positive rate for LAMP could be attributed to the fact that some samples like canary seed were not rated but given a score of 0 for ergot. Alternatively, this could be attributed to molecular detection being more sensitive than visual inspection. In addition, the ubiquity of Claviceps spores or sclerotia and their dispersion through wind, insects, and mechanical means [27], combined with the relative stability of DNA, could lead to target detection in the absence of observable disease. However, the Brassica napus seed lots we investigated tested negative for ergot by the cpn60-targeted LAMP (calcein detection), suggesting that C. purpurea DNA is not detected on the seeds of plants that are non-hosts. It is possible that C. purpurea DNA could be detected on the seeds of other canola samples, since cross-contamination between infected cereals and canola seeds could occur during growth or transport.

The rejection in 2016 of grain imports by Egypt, the world’s largest wheat importer, has clearly demonstrated that importing countries can arbitrarily set standards for any pathogen that have a major socio-economic impact. This suggests that an understanding of the microorganisms associated with agricultural commodities is extremely important for producers, exporters, and importers. Microbial community profiling is one way to approach this, but it has limitations. Pathogen-specific diagnostics can be used to overcome some of these, but they are only capable of querying one or a few organism(s) at a time. In this work, we have chosen ergot to demonstrate these points because it was initially unidentifiable in the sequencing datasets, there is a simple, quantitative visual diagnostic available, and there are recent, serious trade issues associated with ergot contamination. However, not all plant diseases that could contaminate grain can be identified visually or are present externally. The molecular diagnostic methods we have described for the detection of ergot could easily be adapted to other target pathogens.

Supporting information

[Figure omitted. See PDF.]

S1 Table. Primer/Probe sequences and amplification conditions.

https://doi.org/10.1371/journal.pone.0173495.s001

(XLSX)

S2 Table. Distance matrices of the sequences examined from individual sclerotia and reference sequences.

https://doi.org/10.1371/journal.pone.0173495.s002

(XLSX)

S1 Fig. Phylogenetic analysis of cpn60 and ITS sequences amplified from individual sclerotia (Table 3).

https://doi.org/10.1371/journal.pone.0173495.s003

(PPTX)

S1 File. cpn60 nucleotide sequences determined from all sclerotia examined (Table 3).

https://doi.org/10.1371/journal.pone.0173495.s004

(TXT)

S2 File. Cpn60 peptide sequences determined from all sclerotia examined (Table 3).

https://doi.org/10.1371/journal.pone.0173495.s005

(TXT)

S3 File. ITS nucleotide sequences determined from all sclerotia examined (Table 3).

https://doi.org/10.1371/journal.pone.0173495.s006

(TXT)

Acknowledgments

We thank Nick Larkan for providing DNA of fungal isolates for testing assay analytical specificity. We also thank anonymous reviewers for reading and revising the text. The work was supported by Canadian Safety and Security Program (CSSP) funded project CRTI 09-462RD/CSSP 30vv01, and by the Genomics Research and Development Initiative (GRDI) Quarantine and Invasive Species (QIS) project.

Sequence accessions

The sequences determined in this study have been deposited to GenBank with the following accession numbers (cpnDB identification numbers in parentheses): C. purpurea DSMZ strain 714 ITS: KU145585; C. purpurea DSMZ strain 714 cpn60: KU145531 (b27188); ITS sequences obtained from ergot sclerotia: KU145586-KU145622; cpn60 sequences obtained from ergot sclerotia: KU145532- KU145584 (b27189-b27241).

Author Contributions

1. Conceptualization: TD MGL TG SMH.

2. Data curation: TD AC MGL.

3. Formal analysis: TD AC MGL.

4. Funding acquisition: TD MGL.

5. Investigation: TD AC.

6. Methodology: TD AC.

7. Project administration: TD MGL.

8. Resources: TD MGL.

9. Supervision: TD.

10. Validation: TD AC.

11. Writing – original draft: TD AC TG SMH MGL.

12. Writing – review & editing: TD AC TG SMH MGL.

Citation: Comte A, Gräfenhan T, Links MG, Hemmingsen SM, Dumonceaux TJ (2017) Quantitative molecular diagnostic assays of grain washes for Claviceps purpurea are correlated with visual determinations of ergot contamination. PLoS ONE 12(3): e0173495. https://doi.org/10.1371/journal.pone.0173495

References

1. Kendrick B (1992) The Fifth Kingdom. Newburyport, MA: Focus Information Group.

2. Tittlemier SA, Drul D, Roscoe M, McKendry T (2015) Occurrence of ergot and ergot alkaloids in Western Canadian wheat and other cereals. Journal of Agriculture and Food Chemistry 63: 6644–6650.

3. Oellig C, Melde T (2016) Screening for total ergot alkaloids in rye flour by planar solid phase extraction–fluorescence detection and mass spectrometry. Journal of Chromatography A 1441: 126–133. pmid:26947163

4. Flieger M, Wurst M, Shelby R (1997) Ergot alkaloids—Sources, structures and analytical methods. Folia Microbiologica 42: 3–30. pmid:9160999

5. Guo Q, Shao B, Du Z, Zhang J (2016) Simultaneous Determination of 25 Ergot Alkaloids in Cereal Samples by Ultraperformance Liquid Chromatography-Tandem Mass Spectrometry. J Agric Food Chem 64: 7033–7039. pmid:27584949

6. Dahan ME, Knecht E (2016) TIMELINE-Egypt's on again, off again relationship with ergot. Reuters: Reuters.

7. Tian Q, Zhao W, Lu S, Zhu S, Li S (2016) DNA Barcoding for Efficient Species- and Pathovar-Level Identification of the Quarantine Plant Pathogen Xanthomonas. PLoS One 11: e0165995. pmid:27861494

8. Schellenberg J, Links MG, Hill JE, Hemmingsen SM, Peters GA, et al. (2011) Pyrosequencing of chaperonin-60 (cpn60) amplicons as a means of determining microbial community composition. In: Kwon YM, Ricke SC, editors. High-throughput Next Generation Sequencing: Methods and Applications. 1 ed. New York, NY: Humana Press. pp. 143–158.

9. Links MG, Dumonceaux TJ, Hemmingsen SM, Hill JE (2012) The chaperonin-60 universal target is a barcode for bacteria that enables de novo assembly of metagenomic sequence data. PLoS One 7: e49755. pmid:23189159

10. Links MG, Demeke T, Gräfenhan T, Hill JE, Hemmingsen SM, et al. (2014) Simultaneous profiling of seed-associated bacteria and fungi reveals antagonistic interactions between microorganisms within a shared epiphytic microbiome on Triticum and Brassica seeds. New Phytologist 202: 542–553. pmid:24444052

11. Lévesque C, Chen W, Hambleton S, Seifert K, Adam Z, et al. (2015) CRTI 09-462RD—Next generation sequencing, direct detection and genotyping of fungi, bacteria and nematodes in the agri-food system. DRDC-RDDC-2015-C148 DRDC-RDDC-2015-C148. 77 p.

12. Hill JE, Hemmingsen SM, Goldade BG, Dumonceaux TJ, Klassen J, et al. (2005) Comparison of ileum microflora of pigs fed corn-, wheat-, or barley-based diets by chaperonin-60 sequencing and quantitative PCR. ApplEnvironMicrobiol 71: 867–875.

13. Schellenberg J, Links MG, Hill JE, Dumonceaux TJ, Peters GA, et al. (2009) Pyrosequencing of the chaperonin-60 universal target as a tool for determining microbial community composition. Appl Environ Microbiol 75: 2889–2898. pmid:19270139

14. Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge JL, et al. (2012) Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proceedings of the National Academy of Sciences 109: 6241–6246.

15. Hill JE, Penny SL, Crowell KG, Goh SH, Hemmingsen SM (2004) cpnDB: A Chaperonin Sequence Database. Genome Research 14: 1669–1675. pmid:15289485

16. Pfaffl MW (2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29: e45. pmid:11328886

17. Tomita N, Mori Y, Kanda H, Notomi T (2008) Loop-mediated isothermal amplification (LAMP) of gene sequences and simple visual detection of products. Nat Protocols 3: 877–882. pmid:18451795

18. Burd EM (2010) Validation of Laboratory-Developed Molecular Assays for Infectious Diseases. Clinical Microbiology Reviews 23: 550–576. pmid:20610823

19. Banoo S, Bell D, Bossuyt P, Herring A, Mabey D, et al. (2006) Evaluation of diagnostic tests for infectious diseases: general principles. NatRevMicrobiol 4: S21–S31.

20. Schardl CL, Young CA, Hesse U, Amyotte SG, Andreeva K, et al. (2013) Plant-Symbiotic Fungi as Chemical Engineers: Multi-Genome Analysis of the Clavicipitaceae Reveals Dynamics of Alkaloid Loci. PLoS Genet 9: e1003323. pmid:23468653

21. Links MG, Chaban B, Hemmingsen S, Muirhead K, Hill J (2013) mPUMA: a computational approach to microbiota analysis by de novo assembly of operational taxonomic units based on protein-coding barcode sequences. Microbiome 1: 23. pmid:24451012

22. Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol 10: 512–526. pmid:8336541

23. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol 30: 2725–2729. pmid:24132122

24. Comte A, Gräfenhan T, Links MG, Hemmingsen SM, Dumonceaux TJ (2016) Sequence analysis of taxonomic markers cpn60 and ITS from amplified from ergot sclerotia. Data in Brief submitted.

25. Poinar GJ, Alderman S, Wunderlich J (2015) One hundred million year old ergot: psychotropic compounds in the Cretaceous? Palaeodiversity 8: 13–19.

26. Magnuson BA, Patterson CA, Hucl P, Newkirk RW, Ram JI, et al. (2014) Safety assessment of consumption of glabrous canary seed (Phalaris canariensis L.) in rats. Food Chem Toxicol 63: 91–103. pmid:24200856

27. Alderman S, Walenta DL, Hamm PB, Martin RC, Dung JK, et al. (2015) Afternoon ascospore release in Claviceps purpurea optimizes perennial ryegrass infection. Plant Disease 99: 1410–1415.