Identification of ochratoxin A producing Aspergillus carbonarius and A. niger clade isolated from grapes using the loop-mediated isothermal amplification (LAMP) reaction

Authors


Correspondence

Cesare Gessler, Plant Pathology Group, Institute of Integrative Biology, ETH Zurich, 8092 Zürich, Switzerland. E-mail: cesare.gessler@agrl.ethz.ch

Abstract

Aims

To develop two assays based on the loop-mediated isothermal amplification (LAMP) of DNA for the quick and specific identification of Aspergillus carbonarius and ochratoxigenic strains of the Aspergillus niger clade isolated from grapes.

Methods and Results

Two sets of primers were designed based on the polyketide synthase genes involved or putatively involved in ochratoxin A (OTA) biosynthesis in A. carbonarius and A. niger clade. Hydroxynaphthol blue was used as indirect method to indicate DNA amplification. The limit of detection of both assays was comparable to that of a PCR reaction. Specificities of the reactions were tested using DNA from different black aspergilli isolated from grapes. The two LAMP assays were then used to identify A. carbonarius and ochratoxigenic A. niger and A. awamori grown in pure cultures without a prior DNA extraction.

Conclusions

The two LAMP assays permitted to quickly and specifically identify DNA from OTA-producing black aspergilli, as well as isolates grown in pure culture.

Significance and Impact of the Study

Monitoring vineyards for the presence of OTA-producing strains is part of the measures to minimize the occurrence of OTA in grape products. The two LAMP assays developed here could be potentially used to speed the screening process of vineyards for the presence of OTA-producing black aspergilli.

Introduction

Wine significantly contributes to human exposure to ochratoxin A (OTA), a potent nephrotoxic, carcinogenic, immunotoxic and teratogenic mycotoxin (European Commission 2005; Pfohl-Leszkowicz and Manderville 2007). The maximum level of OTA contamination of wines was limited in 2005 by EU at 2 μg l−1 (European Commission 2005). Responsible for OTA production on grapes and the subsequent contamination of wines are members of the Aspergillus section Nigri, also known as black aspergilli. Aspergillus carbonarius is considered the main OTA-producing species in grapes (Cabañes et al. 2002). In fact, most of the strains of this species are able to produce OTA (Perrone et al. 2007). Other ochratoxigenic black aspergilli occurring on grapes belong to the A. niger clade, which has been recently shown to be composed of two sibling species, A. niger and A. awamori (Perrone et al. 2007, 2011). However, only 2–20% of the strains of these species can produce OTA (Perrone et al. 2007, 2011). Other nonochratoxigenic black aspergilli recovered from grapes in European vineyards usually belong to the species A. tubingensis and A. uvarum (Perrone et al. 2007; Varga et al. 2011). As part of the strategy to minimize the risk of OTA contamination of grapes and grape products, the incidence of OTA-producing black aspergilli on grapes should be routinely monitored (Hocking et al. 2007; Visconti et al. 2008). Black aspergilli present on grapes can be isolated and identified on agar media and tested for OTA production in vitro (Bau et al. 2004; Battilani et al. 2006; Bejaoui et al. 2006; Hocking et al. 2007). Alternatively to these classical methods, which are time-consuming and require expertise, molecular tools allow a rapid, sensitive and specific detection of the pathogens. In the last years, several primers and real-time PCR-based assays have been developed for the quick and specific identification of black aspergilli isolated from grapes (Mulè et al. 2006; Atoui et al. 2007; Susca et al. 2007). Due to the lack of information on the genes involved in OTA biosynthesis pathway, these assays targeted house-keeping genes such as calmodulin or polyketide synthase (PKS) genes involved in not-yet characterized biosynthesis pathways. However, these genes are not suitable to be used as molecular target to detect specifically ochratoxigenic strains of the A. niger clade. The targeting of OTA biosynthesis genes is expected to guarantee a higher specificity in respect to methods based on species-specific primers (Niessen et al. 2005). Recently, Storari et al. (2010) cloned a PKS gene fragment from A. carbonarius having a strong amino acid identity with a PKS involved in OTA biosynthesis in A. ochraceus and A. westerdijkiae (O'Callaghan et al. 2003; Bacha et al. 2009). Moreover, the presence of another PKS (an15g07920) possessing a similar high amino acid identity to that of A. ochraceus was shown to be associated with the ability to produce OTA in A. niger isolates (Pel et al. 2007; Storari et al. 2010; Castellá and Cabañes 2011). These two PKS genes could be targeted to specifically detect ochratoxigenic strains of A. carbonarius and of the A. niger clade in the field and in foodstuff. In fact, Castellá and Cabañes (2011) used an15g07920 as molecular target in a real-time PCR system to detect ochratoxigenic A. niger strains on maize. Recently, Yamada et al. (2011) confirmed the role of an15g07920 in the biosynthesis of OTA using a knockout mutant.

The loop-mediated isothermal amplification (LAMP), developed in 2000 by Notomi et al., is an alternative method for molecular detection of pathogenic organisms, which do not require gel electrophoresis to separate and visualize the products or the expensive real-time machines. LAMP is a simple and rapid nucleic acid amplification method, which leads to a great output (up to 109 copies of target DNA) within 1 h under isothermal conditions (Notomi et al. 2000). The output of the LAMP can be visualized in different ways, including gel electrophoresis, turbidity due to the precipitation of magnesium pyrophosphate, the addition of complexometric dyes, which change their colour upon complexation with divalent metal ions or by adding DNA intercalating dyes (Tomita et al. 2008; reviewed by Goto et al. 2009). Hydroxynaphthol blue (HNB), a metal ion indicator, was recently used by Goto et al. (2009) to detect DNA amplification in the LAMP reaction. In case of a positive LAMP output, the Mg2+ present in the solution complexes and precipitates with pyrophosphate formed during DNA polymerization, leading to a colour change of the solution from violet to sky blue. This method has the important advantage of not requiring the opening of the reaction tubes, reducing the risk of cross-contaminations (Goto et al. 2009). The LAMP method has been applied for the detection of several human (reviewed by Parida et al. 2008), animal (Cardoso et al. 2010) and plant pathogens (Kuan et al. 2010). The first LAMP reaction for the detection of a mycotoxin-producing fungus, Fusarium graminearum, was reported recently (Niessen and Vogel 2010).

Aim of this work is the development of two LAMP-based assays targeting acpks14 and an15g07920 for the rapid and specific identification of A. carbonarius and ochratoxigenic strains of the A. niger clade isolated from grapes. HNB was used as indirect method to indicate successful amplification of the target DNA. The two LAMP assays were used to identify A. carbonarius and ochratoxigenic A. niger and A. awamori strains grown in pure agar cultures without the need of a prior DNA extraction step.

Materials and methods

Fungal strains, DNA extraction and OTA production

Fungal isolates used in this study are listed in Table 1. Reference Aspergillus strains were obtained from the ITEM (http://server.ispa.cnr.it/ITEM/Collection/) and the CBS (http://www.cbs.knaw.nl/databases/) culture collections. Further tested black aspergilli were isolated from vineyards in Northern Italy (Trentino region) in September 2008, 2009 and 2010 (Storari et al. 2012). For DNA extraction, black aspergilli were grown in 10 ml potato dextrose broth (PDB; Difco, Becton, Dickinson and Company, Sparks, MD, USA) at 24°C on a horizontal shaker (150 rpm). After 7–10 days of incubation, mycelia were collected, lyophilized and ground using glass beads. DNA extraction was carried out using the DNeasy® Plant Mini kit (Qiagen, Hilden, Germany). Species determination of field isolates was performed by sequencing a fragment of the calmodulin gene using CL1 and CL2A primers (O'Donnell et al. 2000). Amplicons were sequenced using the ABI PRISM BigDye Terminator v3.0 ready reaction cycle sequencing kit (Applied Biosystems, USA). The sequences were aligned using ClustalW (BioEdit software version 7.0.0; Hall, 1999). The alignment output was entered in MEGA software, version 4.0 (Tamura et al. 2007) to generate phylogenetic trees (bootstrap analysis using neighbour-joining with 1000 replicates and default settings). Calmodulin sequences of the reference black aspergilli strains obtained from the NCBI database (http://www.ncbi.nlm.nih.gov/genbank/) were included in the phylogenetic analysis (Perrone et al. 2011; Sørensen et al. 2011; Varga et al. 2011). Aspergillus niger, A. awamori and A. carbonarius isolates were tested for OTA production using the agar plug method of Bragulat et al. (2001) with minor modifications. Briefly, fungal isolates were incubated on CYA (Pitt and Hocking 1997) and YES agar (Bragulat et al. 2001) for 7 days at 24°C. An agar plug (6 mm diameter) was then cut near the centre of the colony and extracted with 0·5 ml methanol for 1 h. The extracts were filtered (Millex-HV Filter (0·45 μm, PVDF, 13 mm; Millipore, Billerica, MA, USA) directly into HPLC vials. The HPLC system (Jasco, Gross-Umstadt, Germany) was equipped with Jasco PU-980 pumps, a Jasco AS-2055 PLUS sampling system, a C18 column (Spherisorb ODS 2, 250 × 4·6 mm, 5 μm; Waters, Milford, MA, USA) and a Jasco 2020 PLUS fluorescence detector (excitation 330 nm, emission 460 nm; gain:1000; response STD). The system was controlled using Jasco Chrompass software. The mobile phase consisted of 57% acetonitrile, 41% water and 2% acetic acid (isocratic) pumped at 1·0 ml min−1. The injection volume was 20 μl. The retention time of the OTA standard (10 μg ml−1 in acetonitrile; analytical standard; Sigma-Aldrich, Buchs, Switzerland) was around 6 min. The detection limit of the HPLC apparatus was about 0·01 μg g−1.

Table 1. Black aspergillus strains tested in the two loop-mediated isothermal amplification assays using extracted DNA or mycelium scraped from pure cultures in agar
SpeciesIsolatesDNAMycelium
AcPKSAnPKSAcPKSAnPKS
A. carbonarius (OTA+)ITEM 5005 + +
A. carbonarius (OTA+)ITEM 5010 + +
A. carbonarius (OTA+)ITEM 5012 + +
A. carbonarius (OTA+)10_27 + +
A. awamori (OTA+)ITEM 7096 + +
A. awamori (OTA+)ITEM 7097 + +
A. awamori (OTA+)ITEM 7098 + +
A. awamori(OTA)09_23ntnt
A. awamori (OTA)10_20
A. awamori (OTA)10_35ntnt
A. awamori (OTA)10_48
A. niger (OTA+)08_42 + +
A. niger (OTA)08_8
A. niger (OTA)08_21
A. niger (OTA)08_37
A. niger (OTA)08_57
A. niger (OTA)08_80
A. tubingensis ITEM 4496
A. tubingensis ITEM 5014ntnt
A. tubingensis CBS 119552ntnt
A. tubingensis CBS 119553ntnt
A. tubingensis 08_34
A. tubingensis 08_54
A. tubingensis 10_1
A. tubingensis 10_9
A. tubingensis 10_14
A. tubingensis 10_40
A. tubingensis 10_57
A. uvarum ITEM 4843
A. uvarum 08_3
A. uvarum 08_14
A. uvarum 08_18
A. uvarum 08_22
A. uvarum 08_33
A. uvarum 10_6
A. uvarum 10_17
A. uvarum 10_39
A. uvarum 10_53
A. ibericus CBS 121593ntnt
A. brasiliensis CBS 101740ntnt
A. acidus CBS 564·65ntnt
A. vadensis CBS 11365ntnt
A. japonicus CBS 114·51ntnt
A. aculeatus CBS 172·66ntnt
A. ellipticus CBS 707·79ntnt
A. heteromorphus CBS 117·55ntnt
A. homomorphus CBS 101889ntnt
A. piperis CBS 112811ntnt
A. costaricaensis CBS 115574ntnt
A. aculeatinus CBS 121060ntnt

LAMP reaction

Loop-mediated isothermal amplification primers for the detection of A. carbonarius (Acpks primer set) and ochratoxigenic strains of the A. niger clade (Anpks primer set) were designed using the PrimerExplorer V3 software (Eiken Chemical Co., Ltd., Tokyo, Japan; http://primerexplorer.jp./elamp3.0.0/index.html) on the interdomain sequences between the KS and AT domains of the PKS genes acpks14 (Accession number GU991531) for A. carbonarius and an15g07920 for A. niger and A. awamori (Table 2).

Table 2. Loop-mediated isothermal amplification primers for the detection of Aspergillus carbonarius (AcPKS) and A. niger (AnPKS) used in the present study
Primer setsPrimersSequence (5′-3′)
A. carbonarius – AcPKSFIPGCTGCAATGCACCCGGTAGTTGAAGACGTGGAGGGCTTCT
BIPAAGTCACGTCAAAAGCCCTGGTGATTCCTTGGGAGGTTGGTC
LFGCTGGCCGAAAAGATCGCTAA
LBCCCCTCTGCTATGAAGTCCG
F3TGGTGGTACGAATGCACAC
B3CGAAATGACAAACAGGCGGT
A. niger – AnPKSFIPCCTGCGCCACCTTCCAAGTGCGATTCGCCCCTCTATGTTG
BIPCTGATCTCAGCCACACTGGCTGATCTTGGGGTTCAAGCTCTG
LFCCAGCACGGATTTTACCGATC
LBTAGAGTCGAAGATGATACCCCCAGT
F3TGCATTAGGTGTTGCCCG
B3AGGTCATCGCGTTGAGGA

The two LAMP reactions were developed following Goto et al. (2009) with minor modifications. The LAMP assays were optimized using DNA (approximately 2 ng per reaction) extracted from different OTA-producing A. carbonarius, A. awamori and A. niger strains (Table 1). Each LAMP assay was carried out in 25 μl reaction mixtures containing the following components: 1·60 μmol l−1 each FIP and BIP, 0·2 μmol l−1 each F3 and B3, 0·8 μmol l−1 each LF and LB (Microsynth, Balgach, Switzerland), 1·2 mmol l−1 each deoxyribonucleotide (dNTP; New England BioLabs, USA), 100 μmol l−1 hydroxynaphthol blue (HNB; Sigma-Aldrich), 8 mmol l−1 MgSO4 (New England BioLabs, Ipswich, MA, USA), 0·320 U μl−1 Bst DNA polymerase (New England BioLabs), 1x ThermoPol buffer (New England BioLabs), 0·8 mol l−1 betaine (Sigma-Aldrich), nuclease-free water (Fermentas, St Leon-Rot, Germany) and 2 μl DNA in solution. As recommended by Varga and James (2006), FIP and BIP primers were HPLC-purified. The reaction mixtures were incubated at 65°C for 60 min in a thermocycler (lid at 99°C); afterwards, the temperature was raised at 80°C for 2 min to terminate the reaction. The LAMP reaction outputs were checked observing the colour change of the HNB from violet towards sky blue in probes containing the target DNA. The master mix setup and the addition of the DNA were carried out in different rooms to reduce the risk of contamination of the master mix.

Limit of detection and specificity of the LAMP assays

To determine the limit of detection, the LAMP assays were carried out using a series of 10-fold dilutions of DNA from 50 ng μl−1 to 0·05 pg μl−1 in autoclaved distilled water, corresponding to 100 ng to 0·1 pg DNA per reaction. DNA used in this test was extracted from A. carbonarius ITEM 5012 and A. awamori ITEM 7096, and it was quantified using NanoDrop® (NanoDrop Technologies, Wilmington, DE, USA). Limits of detection were defined as the lowest DNA amounts that induced recognizable colour change of the HNB dye. The LAMP reactions were carried out in duplicate to confirm the results. The detection limit of the LAMP reaction for A. carbonarius was compared with the PCR assay targeting the calmodulin gene of A. carbonarius developed by Perrone et al. (2004) using the same DNA templates. The cross-reactivity of LAMP primers was tested using approximately 50–100 ng DNA of different strains of black aspergilli isolated from grapes (Table 1). Each reaction was carried out at least twice.

Application of the LAMP assays on mycelium grown on agar

The LAMP reaction was adopted for the identification of pure cultures of A. carbonarius and ochratoxigenic A. niger and A. awamori strains grown on agar. Several black aspergilli isolated from grapes and other reference strains representing almost all species of the Aspergillus section Nigri were tested to evaluate the specificity of the method (Table 1). The isolates were grown on potato dextrose agar (PDA, Difco) at 24°C until the mycelium was clearly visible but still not covered by conidia (2–3 days). A visible piece of mycelium was then scraped from the surface of the agar with a sterile pipette tip, mixed with 50 μl of sterile water and incubated for 10 min at 95°C. The LAMP reaction was carried out with 2 μl of the mycelium mix. Each strain was tested at least twice.

Results

LAMP reaction, specificity and limit of detection

A colour change from violet towards sky blue of the LAMP reaction indicated a successful amplification of the target DNA (Fig. 1). The two LAMP primer sets (Acpks and Anpks) gave positive reactions for the DNA of all A. carbonarius and ochratoxigenic A. niger and A. awamori strains used in this study (Table 1). Serial dilutions of A. carbonarius and A. awamori DNA were used to determine the limit of detection of the two LAMP assays. A positive LAMP reaction was observed for both assays for DNA quantities ranging from 0·1 to 100 ng per reaction. Figure 1 shows, as an example, the amplification of the serial dilutions of A. carbonarius DNA by LAMP reaction. The limit of detection of both reactions is between 0·1 and 0·01 ng; in fact, 0·01 ng of DNA was detected only in one of two repetitions for both assays. The colour intensity does not depend on the DNA concentration as far as it remained above the limit of detection of the method. No intermediate colours between violet and blue resulted. A PCR reaction using specific primers for A. carbonarius calmodulin gene was carried out using the same DNA solutions. The PCR assay showed a slightly visible amplification also for 1 pg. The specificity of the two LAMP assays was tested using DNA from different black aspergilli (Table 1). No amplification of nontarget DNA was observed, with the exception of few false positive in one of the two LAMP repetitions for some tested fungal isolates. However, amplification for these DNA was not observed when the LAMP reaction was repeated a third time, suggesting that the first observed colour changes were due to cross-contaminations between samples.

Figure 1.

Amplification of different amounts of Aspergillus carbonarius DNA by loop-mediated isothermal amplification (LAMP) and PCR reaction. The limit of detection of the LAMP assay is between 0·1 and 0·01 ng DNA (C = negative control, M = 100-bp marker).

Application of the LAMP assays: identification of Aspergillus carbonarius and ochratoxigenic Aspergillus niger and Aspergillus awamori in pure cultures

The two LAMP reactions were used as a quick tool to identify pure cultures of A. carbonarius and ochratoxigenic A. niger and A. awamori strains grown on agar plates. Different black aspergilli isolated from grapes and other reference strains belonging to different species of the Aspergillus section Nigri were tested with both primer sets (Table 1). The method could detect all tested A. carbonarius and ochratoxigenic A. niger and A. awamori strains. To obtain the detection of OTA-producing black aspergilli, a visible piece of mycelium had to be scraped from the agar plate between the first and third day of incubation of the isolates. The nontarget black Aspergillus species, including different nonochratoxigenic A. niger and A. awamori isolates, were not detected with the exception of few false positives. Also in this case, the repetitions of the reaction gave negative outcomes, indicating that the false positives were probably due to cross-contamination with fungal material.

Discussion

Fungal strains belonging to the A. carbonarius species and the A. niger clade are considered responsible for OTA contamination of wines (Perrone et al. 2007; Varga et al. 2011). Continuous survey is necessary to monitor the presence of OTA-producing black aspergilli species in vineyards (Hocking et al. 2007). Because of its simplicity, the LAMP reaction could represent a valuable tool to speed the process of isolation and identification of OTA-producing black aspergilli, which is time-consuming or a valid alternative to the use of expensive instruments such as real-time machines that require a prior DNA extraction step (Mulè et al. 2006; Atoui et al. 2007). In this work, two LAMP assays for the identification of A. carbonarius and ochratoxigenic strains of the A. niger clade were developed. The two primer sets were designed based on the KS-AT interdomain regions of PKS genes involved and putatively involved in OTA biosynthesis in the A. niger clade (an15g07920; Yamada et al. 2011) and A. carbonarius (acpks14; Storari et al. 2010), respectively. According to Mayer et al. (2007), the interdomain regions are less conserved than the PKS domains. This should confer a higher specificity to the LAMP reactions, lowering the risk of amplification of other unwanted PKS genes. Hydroxynaphthol blue (HNB) was used as indirect method for the detection of positive reactions. HNB has been shown not to affect the LAMP reaction and can be thus added to the master mix before the DNA amplification step, avoiding the need of opening the tubes at the end of the reaction (Goto et al. 2009). This is an important advantage to other visualization methods because the LAMP products are an important source of cross-contaminations (Niessen and Vogel 2010). In fact, a few false-positive outcomes were observed also during this study. These were probably due to cross-contaminations with DNA or fungal material because the repetitions of the reactions gave a negative result. Beside cross-contaminations, Lee et al. (2009) suggested that nonspecific amplification could be caused also by low-background DNA and damage to the primers caused by freeze-thaw repetitions. The specificities of the two LAMP assays were tested with DNA of different black aspergilli isolated from grapes. A colour change was observed for DNA of A. carbonarius and ochratoxigenic strains of the A. niger clade. Moreover, primers for A. carbonarius could not amplify DNA of ochratoxigenic strains of the A. niger clade and vice versa, indicating that the two reactions are highly specific. In a reduced number of cases, it has been observed that the colour of positive LAMP reactions was of a darker blue than the usual observed sky blue. However, the positive reactions could be easily distinguished from negative ones. The limit of detection (LOD) for both LAMP assays was between 100 and 10 pg of DNA (corresponding approximately to 3000 and 300 genome equivalents, respectively). A comparable LOD was also found for the PCR reaction developed by Perrone et al. (2004) for A. carbonarius. Different studies reported a LOD for the LAMP reaction lower than that of conventional PCR (Kuboki et al. 2003; Kuan et al. 2010). However, in these studies, the outputs of the reactions were directly visualized on gel electrophoresis, which has been demonstrated to have a lower LOD compared with the detection based on turbidity (Yoshikawa et al. 2004). The LAMP reaction had been shown to be less sensitive to culture media and biological substances than PCR, allowing amplification of the target DNA without a prior purification step. This makes it even more attractive as time- and money-saving detection method (Kaneko et al. 2007). The two LAMP assays developed in the present work were evaluated for the identification of A. carbonarius and ochratoxigenic strains of the A. niger clade grown on agar without the need of DNA purification. Preliminary experiments revealed that the LAMP reactions could not detect the fungal DNA from mycelium directly mixed in the reaction tube; the incubation of the mycelium in water at 95°C was indispensable for a robust detection. With this method, a positive reaction was observed for the OTA-producing strains, whereas no colour change occurred for the other tested grape-associated black aspergilli and also other tested reference strains representing almost all species included in the Aspergillus section Nigri (Table 1). During preliminary experiments, the LAMP reactions failed to work when scraped conidia were used instead of mycelium. This is probably due to the fact that the prior heating step is not sufficient to induce the release of DNA from conidia. To scrape a sufficient quantity of mycelium and, hence, to obtain a robust detection, it is therefore recommended to use young fungal colonies not yet covered by conidia. Black aspergilli classification based on morphological features alone is very difficult or, for species belonging to the so-called A. niger aggregate including the A. niger clade and A. tubingensis, even impossible (Samson et al. 2007; Perrone et al. 2011). Moreover, to identify OTA-producing isolates, black aspergilli are usually tested for OTA production on agar media (Bragulat et al. 2001). A quick molecular method, not needing DNA extraction for the identification of A. carbonarius and ochratoxigenic strains of the A. niger clade like the one proposed here, would facilitate the screening process in surveys of vineyards, which normally involve hundreds of isolates.

In conclusion, this study reports the development of two LAMP assays for the specific detection of A. carbonarius and ochratoxigenic strains of the A. niger clade. Specificity of the two assays was tested using DNA and strains from different species of black aspergilli isolated from grapes and other commodities. The sensitivity of the reactions was comparable to that of conventional PCR. The two assays were used to identify OTA-producing A. carbonarius, A. niger and A. awamori strains grown in pure cultures without a prior DNA extraction step. The results presented here indicate that the LAMP reactions could be potentially applied to fast screen vineyards for the presence of OTA-producing black aspergilli.

Acknowledgements

This work was funded by the Autonomous Province of Trento, project ENVIROCHANGE, Call for proposal Major Projects 2006.

Ancillary