6 (2006) 529–537www.elsevier.com/locate/jmicmeth

Journal of Microbiological Methods 6

Detection of Fusarium culmorum in wheat by a surface plasmonresonance-based DNA sensor

Francesco Zezza ⁎, Michelangelo Pascale, Giuseppina Mulè, Angelo Visconti

Institute of Sciences of Food Production (ISPA), National Research Council, Via G. Amendola 122/O, 70126 Bari, Italy

Received 31 August 2005; received in revised form 3 February 2006; accepted 7 February 2006Available online 24 March 2006

Abstract

A surface plasmon resonance (SPR) sensor based on DNA hybridization has been developed for the detection of Fusariumculmorum, a fungal pathogen of cereals. A 0.57 kbp DNA fragment of F. culmorum was amplified by specific primers and a 25-meroligonucleotide probe was selected within the sequence of the PCR amplicon. After biotinilation, the probe was immobilized on astreptavidin sensor chip and tested for biospecific interaction with PCR products of F. culmorum. The effect of denaturating agents(formamide and urea) and ionic strength (NaCl) on hybridization efficiency of double-stranded PCR products with the immobilizedprobe and the specificity of the probe were investigated. The SPR biosensor was successfully used for the detection of F. culmorumin culture material of different strains and in naturally infected wheat samples. Tested on fungal cultures, it showed a goodselectivity for F. culmorum against other species of either Fusarium or other fungal genera. A background signal was observed inwheat samples strictly depending on the DNA amount of the testing matrix. Testing 30ng of durum wheat DNA the detection limitwas 0.06pg of F. culmorum DNA. The developed PCR-SPR assay allowed to detect F. culmorum with sensitivity and specificityhigher than gel-electrophoresis analysis.© 2006 Elsevier B.V. All rights reserved.

Keywords: DNA biosensor; Fusarium culmorum; Surface plasmon resonance; PCR; Wheat

1. Introduction

Fusarium culmorum is a common pathogen of wheatand other small cereal grains involved in the etiology ofFusarium head blight, a worldwide disease causing totalor partial ear premature senescence with consequentreduction of both crop yields (up to 70%) and grainquality (Bai and Shaner, 1994; Bottalico and Perrone,2002; Parry et al., 1995). In addition, F. culmorum canproduce various toxic secondary metabolites, i.e.

⁎ Corresponding author. Tel.: +39 080 5929 763; fax: +39 080 5929374.

E-mail address: francesco.zezza@ispa.cnr.it (F. Zezza).

0167-7012/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.mimet.2006.02.003

deoxynivalenol and other trichothecenes and zearale-none, which are harmful to humans and livestock(Logrieco et al., 2003). Rapid methods for earlydetection of this fungal pathogen in cereals are thereforehighly required in order to plan appropriate and timeddisease control measures and to avoid losses in bothquantity and quality of a crop.

The conventional approach to fungal detection ininfected plants involves isolation of the fungus intoaxenic culture or culturing on selective media followedby identification based upon morphological character-istics (Nirenberg, 1981). These techniques are time-consuming, require taxonomical expertise and canbe applied only to viable mycelium. Moreover,

530 F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

identification within closely related fungal species, suchas Fusarium spp., requires considerable experience(Marasas et al., 1984). Recently, PCR-based assays forspecies-specific detection of several Fusarium speciesin cereals have been reported (Niessen and Vogel, 1997,1998; Schilling et al., 1996; Doohan et al., 1998;Nicholson et al., 1998; Parry and Nicholson, 1996;Edwards et al., 2001), and commercial kits are currentlyavailable for the detection of several fungal pathogensfrom seeds and other plant tissues, includingF. culmorum(Neogen Europe Ltd., http://www.neogeneurope.com).Measurement of PCR products obtained by conventionalPCR methods is performed by gel electrophoresis thatallows discrimination of different DNA fragments bytheir molecular weight. Despite these methods are simpleand relatively inexpensive, they use mutagenic orpotentially mutagenic labeling reagents (such as ethi-dium bromide or SYBR Green™), and do not providesequence information. Moreover, PCR could providebyproducts arising from the amplification of non-specificsequences with the same length of the desired fragment,leading to false positive results. Therefore, a hybridiza-tion assay is strongly recommended to confirm theselectivity of the reaction. Southern blotting is a well-established method for the detection of specific DNAsequences bymeans of a hybridization step; nevertheless,it is not recommended for routine analyses since it istime-consuming, not always reproducible and involvesradioactive reagents (Southern, 1975).

Surface plasmon resonance (SPR)-based DNAbiosensors have been shown to be rapid, label-freeand selective tools for detection of PCR products(Sawata et al., 1999; Giakoumaki et al., 2003). Thesebiosensors measure refractive index changes thatoccur on a metal film providing a signal that ispositively correlated with the mass density changes onthe metal surface (Jönsson et al., 1991). Byimmobilizing a biological recognition molecule (li-gand) on the sensing surface, SPR biosensors providea tool for real-time biospecific interaction analysis.The choice of a specific single-stranded DNA ligandshould allow rapid detection of the target DNA byhybridization with higher selectivity with respect togel-electrophoresis analysis. SPR technology, com-bined with PCR amplification, has been used for thedetection of genetically modified maize and soybean(Feriotto et al., 2002, 2003; Mariotti et al., 2002), ofEscherichia coli and HIV-1 in human samples, andfor the diagnosis of human pathologies induced bygenetic mutations (Nilsson et al., 1997; Feriotto et al.,1999, 2001, 2004; Kai et al., 2000; Bianchi et al.,1997).

In this work, the use of SPR combined with PCRamplification for the detection of F. culmorum in purefungal cultures as well as in wheat samples is reported.A SPR-based biosensor was developed using a 25-merDNA probe that was selected from a 0.57 kbp PCRproduct specific for F. culmorum and immobilized onthe surface of a sensor chip. The biosensor was tested forbiospecific interactions with PCR products of F.culmorum, and an optimized protocol for SPR detectionwas developed after studying the effect of ionic strengthof the testing solution and of different denaturing agentson the hybridization signal.

2. Materials and methods

2.1. Fungal strains and culture conditions

Fungal strains used in the present study were obtainedfrom the ITEM collection (CNR-Institute of Sciences ofFood Production, Bari, Italy, http://www.ispa.cnr.it/Collection/). They include: 20 strains of Fusariumculmorum, 4 strains of Fusarium graminearum, 2 strainsof Fusarium proliferatum, and 1 strain of Fusariumavenaceum, Fusarium poae, Fusarium cerealis, Fusar-ium verticillioides, Fusarium subglutinans, Fusariumsambucinum, Fusarium acuminatum, Fusarium oxy-sporum, Aspergillus ochraceus, Aspergillus carbonar-ius, Aspergillus niger and Penicillium expansum. Strainswere cultured at 25 °C on potato-dextrose-agar (PDA)(Difco Laboratories Incorporated, Sparks, MD, USA).Mycelium was scraped from the surface of 7-day-oldcolonies on PDA, inoculated into 100 mL sterile liquidmedia (40g glucose, 5g peptone, 3g yeast extract, 3gmalt extract and water up to 1L) and incubated at 25 °Cfor 48h on a rotary shaker (110rpm). The mycelium wasthen collected by filtration throughWhatman No. 1 filterpaper and lyophilized.

2.2. DNA extraction and quantification

Genomic DNA from fungal isolates was extractedusing a conventional phenol/chloroform/isoamyl alcoholprotocol (Sambrook et al., 1989). Briefly, lyophilizedmycelium (30mg) was finely ground and treated with600μL of extraction buffer containing 200 mM Tris–HCl pH 7.8, 20 mMEDTA, 150 mMNaCl, 2% SDS, 1%β-mercaptoethanol and 100μg/mL proteinase K inwater. After incubation for 90 min at 65 °C, shakingfrequently by inversion, DNA purification was per-formed by adding 700μL of phenol/chloroform/iso-amylic alcohol (25:24:1, v/v) and by centrifuging at10,000×g for 20min at 4 °C. The aqueous phase was

531F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

collected in a vial and purified again following the sameprocedure with 700μL of chloroform/isoamylic alcohol(24:1, v/v). After addition of 0.1 volumes of 3M sodiumacetate at pH 5.2 and 2.5 volumes of 95% ethanol, theDNA solution was left for 1 h at −20 °C and centrifugedat 12,000×g for 20min at 4 °C to allow DNAprecipitation. The pellet was washed twice with 1mLof 70% cold ethanol and recovered by centrifugation(12,000×g, 4 °C for 10min). Precipitated DNAwas driedunder vacuum, brought up with 100 μL sterile water andleft overnight at 4 °C to allow complete dissolving.

Genomic DNA from wheat grain samples wasextracted and purified according to the method describedby Wilson et al. (2004), with minor modifications.Briefly, DNA from 4g of milled wheat grains wasextracted in 20mL CTAB buffer (22mM cetyltrimethy-lammonium bromide, 34mM sarkosyl, 137mM sorbitol,22mM EDTA, 1% cross-linked polyvinylpyrrolidone,1.5M NaCl) for 1h at 65 °C, shaking frequently. One-third volume of 5M potassium acetate was added alongwith one-third volume of chloroform/isoamyl alcohol(24:1), mixed and held at−20 °C for 20min. Themixturewas then centrifuged at 1900×g for 15 min; 600μL of theaqueous phase was added to two volumes of absoluteethanol, mixed and held at −20 °C for 20min. The DNAwas precipitated by centrifugation at 850×g for 10min.The pelletwaswashed twicewith 1mL70% (v/v) ethanoland recovered by centrifugation (850×g for 5min) aftereach washing step. The precipitated DNA was driedunder vacuum, brought up with 300μL sterile water andleft overnight at 4 °C to allow complete dissolving.

Concentrations of extracted DNAwere determined byelectrophoresis on 0.7% agarose gel and comparisonwith DNA standard markers (MassRuler™ DNALadder, High Range, ready-to-use, Fermentas Interna-tional Inc., Burlington, Ontario, Canada). Gels werestained with ethidium bromide, visualized under UVlight on a ChemiDoc system (Bio-Rad Laboratories S.r.l., Segrate, Italy) and analysed using a Bio-Rad QuantityOne software.

2.3. PCR amplification

Specific primers, Fc01F (5′-ATG GTG AAC TCGTCG TGG C-3′) and Fc01R (5′-CCC TTC TTA CGCCAATCTCG-3′), were used to amplify a 0.57kbp DNAfragment selected from F. culmorum (Nicholson et al.,1998). Primers were synthesized by MWG Biotech(MWG Biotech, Ebersberg, Germany). PCR amplifica-tions were carried out on 100μL solution containingdifferent amounts of fungal DNA (from 0.04pg to10ng) or 30ng of total DNA extracted from wheat

samples. The reaction mixture consisted of 100μM ofeach deoxynucleotide triphosphate (dNTP), 100nM ofeach oligonucleotide primer and 2 U of Hotmaster™ TaqDNA polymerase (Eppendorf AG, Hamburg, Germany)in Hotmaster™ Taq buffer with Mg2+ (Eppendorf AG,Hamburg, Germany). Amplifications were performedwith a GeneAmp PCR System 9700 thermal cycler(Perkin-Elmer, Wellesley, MA, USA) set up as follows:denaturation at 94 °C for 2min, then 40cycles of 94 °Cfor 50 s, 57 °C for 50 s and 72 °C for 1min, and a finalextension at 72 °C for 7min followed by cooling to 4 °Cuntil recovery of samples. Tubes without DNA templatewere included in each experiment and used as blankcontrol samples. Screening of PCR products wereperformed by gel electrophoresis on 2% agarose gel,stained with ethidium bromide, visualized using a UVtransilluminator and compared with a 100kb DNAladder (New England Biolabs Inc., Ipswich, MA, USA).In order to allow concentration and removal of primersand unincorporated nucleotides, PCR products werepurified using Montage™ PCR centrifugal filter devicesprior to SPR analysis (Millipore, Billerica, MA, USA)according to the manufacturer's protocol and recoveredin sterile water.

2.4. SPR probe design

The 0.57kbp PCR product from F. culmorum wassequenced using BigDye™ Terminator Cycle Sequenc-ing Ready Reaction Kit for both strands (AppliedBiosystems, Foster City, CA, USA), and the sequenceswere aligned with MT Navigator software (AppliedBiosystems). From the sequence of this fragment, a 25-mer target probe (Fc01P) was designed using the PrimerExpress program Version 2.0 (Applied Biosystems).The probe specificity was preliminarily assessedfollowing BLAST (www.ncbi.nlm.nih.gov/BLAST/)and FASTA (www2.ebi.ac.uk/fasta3/) database searchesfor DNA sequences. In addition, a 25-mer referenceprobe (Ref01P) was selected with no significantsimilarity with the target PCR product (assessed byBLAST http://www.ncbi.nlm.nih.gov/blast/bl2seq/bl2.html). The base sequences of the probes were: 5′-CTTTAC CCC TCT GTTACTAAA CTAT-3′ (Fc01P) and5′-ATC TAA AAT GCT TCC AAA CAG CGG C-3′(Ref01P). 5′ biotin-modified probes were synthesizedby MWG Biotech.

2.5. SPR analysis

SPR analyses were performed by a Biacore® Xapparatus (Biacore International AB, Uppsala, Sweden),

532 F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

using Biacore SA sensor chips (carboxymethylateddextran matrix pre-immobilized with streptavidin) forall experiments. PBS-E300 buffer solution (20mMNa2HPO4 [pH 7.4], 0.1mM EDTA, 300mM NaCl)was used as running buffer for immobilization of probeson sensor chips and for hybridization experiments. Thepresence of two flow cells on the sensor chip allowedthe immobilization of Fc01P on flow cell 1 (FC1), asprobe for target DNA detection, and Ref01P on flow cell2 (FC2), as a reference probe intended to investigatenon-specific interactions. Prior to immobilization,sensor chips were conditioned with three consecutive1-min injections of a solution containing 1M NaCl and50mM NaOH. The biotinylated probes were immobi-lized by injecting 100μl of a 1μM probe solution inPBS-E300, by means of biotin–streptavidin interactionuntil saturation of the surface (Persson et al., 1997).

In order to establish optimum hybridization condi-tions, PCR product solutions containing 20mMNa2HPO4 [pH 7.4], 0.1mM EDTA, NaCl at concentra-tion from 0 to 600mM, and denaturing agents (urea orformamide) at concentration from 0% to 30% weresubjected to SPR analysis. A PBS-E150 solution(20mM Na2HPO4 [pH 7.4], 0.1mM EDTA, 150mMNaCl) was used as reference experimental condition.Optimized SPR measurements of PCR products wereperformed by injecting 50μl of a solution prepared byadding 10μL formamide to 40μL of purified PCRproduct and then adjusting the total volume to 100μLwith 2X PBS-E300 buffer. Prior to SPR analysis, PCRproducts were subjected to a denaturation step at 95 °Cfor 5min followed by 1min cooling on ice.

A real-time baseline correction was automaticallyperformed during each measurement by subtracting thereference signal recorded in FC2 from the target signalrecorded in FC1. The analytical signal was representedby the resonance unit (RU) shift measured during sampleinjection, i.e. the difference between the RU valuerecorded after washing the sensor surface followinghybridization reaction and the baseline RU value.

All SPR measurements were carried out at 25 °C.The flow rate was 5μL/min and the injection time was10min. Each measurement was repeated three timesunless otherwise reported. Regeneration of the sensorsurface was performed by short injections (up to 1min)of 2–50mM NaOH until complete recovery of thebaseline RU value on both flow cells.

2.6. Statistical analysis

Data were processed by variance analysis (ANOVA,Tukey's test) at P=0.05 to indicate statistically

significant differences between means using the Sigma-Stat for Windows version 3.0 statistical software(Sigma, St. Louis, MO, USA). Sigmoidal fit of datawas performed using Origin® version 6.0 (OriginLabCorporation, Northampton, MA, USA).

The following equation ΔRULOD=ΔRUblank +3×SDblank was used to calculate the RU shift associatedwith the minimum amount of DNA detectable by theSPR biosensor. ΔRUblank and SDblank were the RU shiftobtained from blank control sample injection and therelevant standard deviation; these values were respec-tively obtained from the lower asymptote of thesigmoidal fit and the relevant SD calculated byunweighted least-square method.

3. Results and discussion

The immobilization procedure, based on streptavi-din–biotin binding, resulted in signal shifts of 1622±35RU (n=7) on FC1 for Fc01P probe (target DNAdetection) and 1568±64 RU (n=7) on FC2 for Ref01Pprobe (reference). The relative standard deviations(RSDs) were lower than 4%, showing a high repeatabil-ity of the immobilization procedure.

In order to allow hybridization with the immobilizedprobe, symmetric PCR products of F. culmorum,consisting of double-stranded DNA (dsDNA) fragments,were subjected to a denaturation step by thermaltreatment prior to injection. This procedure could giverise to possible re-annealing of denatured strands withformation of distinctive secondary structures, whichcould hinder the hybridization reaction with theimmobilized probe (Sawata et al., 1997). In order toavoid this problem and improve sensitivity of the SPRbiosensor, denaturation and detection of PCR productswere carried out in the presence of different amounts ofdenaturing agents (urea or formamide) and different salt(NaCl) concentrations, and the results were comparedwith those obtained under reference experimentalconditions, i.e. with PCR products of F. culmorumdissolved in PBS-E150. In Table 1, the sensitivity gain ofthe SPR signal shift measured in the presence of variousconcentrations of formamide and NaCl with respect tothe signal shift obtained after injection of PCR productsunder reference experimental conditions is reported (i.e.testing ΔRU value/reference ΔRU value). Resultsindicate that, at low formamide concentrations, 5% and10% (v/v), formation of inter- and intra-molecularhydrogen bonds between DNA strands was reducedallowing a more efficient hybridization of the targetDNAwith the immobilized probe (up to about threefoldwhen 10% formamide was used). Higher concentrations

0.01 0.1 1 10 100 1000

0

100

200

300

Sig

nal

sh

ift

(RU

)

[F. culmorum PCR product] (ng/µL)

Fig. 1. Correlation between RU shift and concentration of F. culmorumPCR product in the range 0.05–200ng/μL. The curve represents thesigmoidal fit of data (correlation coefficient r=0.9996). All valueswere recorded in the optimum experimental conditions (20mMNa2HPO4 [pH 7.4], 0.1mM EDTA, 300mM NaCl, 10% formamide).Each value is the mean of three measurements; error bars represent therelevant standard deviations.

Table 1Effect of formamide and NaCl concentrations on SPR response of asolution containing 11ng/μL of F. culmorum (ITEM 627) PCRamplification product

Formamide [% (v/v)] [NaCl] (mM) Signal gain⁎ SD

0 0 0.09a 0.040 150 1.00c 0.100 300 1.16c,d 0.030 450 1.23d 0.060 600 0.47b 0.095 150 2.53f 0.0510 0 0.10a 0.0310 150 2.92f 0.1110 300 3.92g 0.0810 450 2.56f 0.1220 150 1.76e 0.0630 150 1.04c,d 0.07

⁎ Values represent the average signal gain (three measurements,50μL injected) with respect to reference experimental conditions(PBS-E150), i.e. the ratio between RU shift measured in theexperimental conditions and RU shift measured in the referenceconditions. Values followed by the same letter are not significantlydifferent, P<0.05 (Tukey's test).

533F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

of formamide (20% and 30%) caused the inhibition ofboth inter-strands and intra-strand interactions andminorannealing of DNA with the probe. These results are inagreement with previous studies (Sambrook et al., 1989;Sawata et al., 1999). Similar results were observed byadding different amounts of urea (data not shown),although a lower sensitivity was observed in allexperiments (best result: signal gain of 2.6 with 10%urea) making formamide a preferable choice as denatur-ing agent. The addition of NaCl to the testing solutionincreased the overall salt concentration and provided ashielding effect acting on electrostatic interactionsaccording to the Debye–Huckel theory. This led toweakening of repulsive forces between negativelycharged DNA strands in the solution and betweenDNA in solution and the sensor surface. Negativecharges on the sensor surface were due to both phosphategroups of the oligonucleotide probes and free carboxyl-groups on the dextran matrix. A low sensitivity wasobserved when no NaCl was added to the testingsolution, showing that ionic shielding effect is necessaryto allow target DNA to reach the sensor surface andhybridize with the probe. The increase of salt concen-tration led to an improved hybridization efficiency up toa maximum value, above which the re-annealing ofdenatured strands prevailed on target–ligand interaction.In all cases, injection of blank control solutions, obtainedfrom PCR carried out without DNA template, resulted insignal shifts similar to the instrumental noise, thusconfirming that the real-time baseline correction (sub-traction of the FC2 signal) is a suitable method for non-

specific binding correction in the above reportedconditions. This indicated that a further correction step,e.g. the subtraction of blank sample curve, as reported byothers authors (Rich and Myszka, 2000), was notnecessary. The use of 10% formamide and 300mMNaCl in the testing solution allowed a nearly fourfoldincrease of the hybridization efficiency with respect toreference conditions (Table 1).

Giakoumaki et al. (2003) reported that thermaldenaturation with or without formamide was noteffective to generate reproducible and significant SPRsignals when PCR products were tested, due to theformation of secondary structures. Our results suggestthat different experimental conditions lead to differentlevels of dsDNA denaturation and target–probe hybrid-ization, and different optimum conditions can beidentified for specific probe–amplicon interactions.The optimization of salt and denaturing agent concen-trations is therefore strongly recommended when settingup new SPR biosensors for amplified DNA detectiontaking into account the peculiar intra- and inter-molecular interactions of the probe and the target PCRproduct.

Fig. 1 shows the SPR calibration curve obtained withthe 0.57 kbp PCR product of F. culmorum in the range0.05–200ng/μL. A sigmoidal curve was the best fit forthe experimental data with correlation coefficient of0.9996. A good linearity was observed between 0.25and 2.5ng/μL (r=0.9954). Thermal denaturation treat-ment, combined with a thorough control of formamideand salt concentration of the testing solution, resulted tobe an effective approach for SPR detection of the

0.01 0.1 1 10 100 1000

0

100

200

300Durum wheat DNA

0 ng 30 ng150 ng

sig

nal

sh

ift

(RU

)

F. culmorum DNA amount (pg)

Fig. 2. Correlation between RU shift and the amount of startingF. culmorum genomic DNA template (ITEM 627) In PCR, atdifferent concentration of durum wheat DNA. Curves representthe sigmoidal fit of data (durum wheat DNA: 0 ng (solid line),correlation coefficient r=0.9992; 30ng (dashed line), r=0.9989;150ng (dotted line), r=0.9965). All values were recorded inoptimum experimental conditions (20mM Na2HPO4 [pH 7.4],0.1mM EDTA, 300mM NaCl, 10% formamide).

534 F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

0.57kbp F. culmorum amplicon with good repeatability(SD<10 RU) and detection limit (LOD=0.25ng/μL).Values <1 RU (similar to the instrumental noise) wereobserved for measurements of blank control samples(PCR mixture without DNA template) indicating theabsence of non-specific interactions of the PCR reagentswith the sensing surface.

In order to evaluate the matrix effect that could affectthe SPR analysis of F. culmorum in wheat samples, PCRamplifications of different amount of genomic DNA ofF. culmorum (ITEM 627), ranging from 0pg to 800pg,were performed in the presence of different amounts (0,30 or 150ng) of DNA of durum wheat buds grownunder sterile conditions and the relevant PCR productswere detected by SPR biosensor (Fig. 2). Positive SPRshifts of 8.0±1.5 RU (n=3) and 14.1±6.7 RU (n=3)were respectively observed when 30ng and 150ng ofdurum wheat DNA were amplified in the absence offungal DNA template, indicating a background signal(baseline drift) due to the wheat matrix. The subtraction

Fig. 3. Gel electrophoresis of the PCR product from different amount of sta(3) 0.1pg, (4) 0.25pg, (5) 0.5pg, (6) 1pg, (7) 2pg, (8) 5pg, (9) 40pg, (10)DNA ladder. Similar results were obtained in the presence of 30ng of du

of corresponding blank RU shift to assess the SPRresponses relevant to F. culmorum PCR product (Richand Myszka, 2000) led to reliable results (similar tothose obtained with pure culture) only for sample with30ng durum wheat DNA. In the presence of 150ngdurum wheat DNA, RU shifts (corrected for blankcontribution) were significantly lower with respect tothe corresponding values recorded with pure F.culmorum DNA, indicating an inhibitory effect of thematrix on PCR amplification.

Detection limits of 0.05pg (corresponding to a signalshift of 10.1 RU) and 0.06pg (corresponding to a signalshift without correction for blank of 17.4RU) of F.culmorum DNA template were observed in the absenceand in the presence of 30ng of durum wheat DNA,respectively. Matrix effects could then be consideredneglectable when 30ng of total DNA extracted fromwheat samples were used for PCR amplification. Due toboth the higher LOD (0.53pg) and the observedinhibitory effects in the presence of 150ng of durumwheat DNA, 30ng of wheat DNA template was used forfurther experiments in this paper and is recommendedfor SPR detection of F. culmorum DNA in wheatsamples.

High repeatability was observed by analyzing purefungal culture materials (triplicate experiments) result-ing in RSD lower than 8% for all tested concentrationsexcept for limit of detection (RSD=25%). Detectionlimit of SPR analysis was over one order of magnitudelower than the minimum amount of starting fungal DNA(1pg) necessary to produce a visible band on gelelectrophoresis (Fig. 3). The application of the methodfor quantification of fungal DNA in wheat resulted in arelatively poor reproducibility of the measurements,with RSD values up to 30% recorded in the presence of30ng durum wheat DNA. Amplification efficiency ofPCR on DNA extracted from plant material is known tobe non-reproducible due to the presence of variousinhibitory compounds acting on Taq DNA polymerase,such as phenolics and polysaccharides (Wilson, 1997).Moreover, due to the exponential growth of DNA during

rting genomic DNA template (ITEM 627): (1) 0.025pg, (2) 0.05pg,80pg, (11) 800pg. (B) Negative control (blank sample), (M) 100bprum wheat DNA (data not shown).

535F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

PCR, minor differences in amplification efficiencyresult in large differences in amplicon yields, leadingto poor reproducibility.

It should be noted that nucleotide changes ordeletion/insertion rearrangements may also existamong isolates of the same species and may determinethe failure of both the amplification and the sequencespecific recognition. In order to assess the presence, inthe region from which the present primer pair and probewere chosen, of genetic variability that can compromisethe reliability of the method leading to false-negativeresults, the PCR-SPR assay was used for the detection of

Table 2Comparison of SPR signals and gel-electrophoresis analysis of PCR productsFusarium, Aspergillus and Penicillium

Species Accession (ITEM) Origin SPR

F. culmorum 627 Yugoslavia 276.628 Yugoslavia 275.5991 Spain 295.5996 Spain 273.6221 Poland 298.6229 Poland 291.6234 Poland 282.6249 Poland 277.6255 Poland 267.6260 Poland 295.6264 Poland 268.6267 Poland 273.6268 Poland 258.6269 Poland 264.6270 Poland 263.6625 Holland 258.6627 Holland 272.6628 Holland 294.6629 Holland 265.6630 Holland 289.

F. graminearum 633 Yugoslavia 18.4432 Yugoslavia 5.4433 Yugoslavia −5.4440 Yugoslavia 12.

F. cerealis 666 Finland −3.F. poae 3725 Poland 4.F. verticillioides 3993 USA 5.F. proliferatum 1725 Italy −1.

2620 Slovakia −2.F. avenaceum 3409 Finland 1.F. subglutinans 3922 USA 11.F. sambucinum 846 Germany 0.F. acuminatum 797 South Africa 12.F. oxysporum 3867 Italy 5.A. ochraceus 4537 Portugal 3.A. niger 4736 Italy 4.A. carbonarius 4864 Italy 5.P. expansum 7011 Italy −3.a +, amplified product of expected size; −, no amplification or non-specifib n.d.: not determined, single measurement.c Below the threshold value of 10.1 RU.

20 F. culmorum strains from fungal cultures. All strainswere successfully identified by the PCR-SPR assay,with PCR products giving RU shifts ranging from 258.4to 298.6RU and repeatability values (RSD) from 3% to14% (n=3). Statistical analysis of variance (Tukey'stest, P=0.05) resulted in no significant differencesbetween different strains. Results were in agreementwith gel-electrophoresis analysis, thus demonstratingthat no polymorphism, detrimental to the amplificationor hybridization steps, occurs.

In order to test the specificity of the primer–probecombination, the assay was tested against 10 Fusarium

of different strains of F. culmorum and other fungal species belonging to

analysis (ΔRU) RSD (n=3) Electrophoretic analysis a

8 5.1 +8 3.2 +3 5.3 +5 3.7 +6 6.5 +7 8.1 +3 5.6 +7 14.0 +7 4.7 +2 12.7 +5 3.9 +1 4.0 +4 4.4 +3 3.8 +8 4.2 +8 7.4 +6 3.3 +1 5.4 +3 6.5 +6 4.3 +4 n.d. b −8 c n.d. b −6 c n.d. b −4 n.d. b −7 c n.d. b −4 c n.d. b −6 c n.d. b −2 c n.d. b −9 c n.d. b −1 c n.d. b −4 n.d. b −8 c n.d. b −3 n.d. b − (0.8kbp)5 c n.d. b −4 c n.d. b −9 c n.d. b −4 c n.d. b −4 c n.d. b −

c amplification (product size indicated in brackets).

536 F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

spp., other than F. culmorum, known to occur aspathogens or saprophytes on wheat, and against fourfungal species belonging to the genera Aspergillus andPenicillium. Results are reported in Table 2. No PCRproducts were detected by gel-electrophoresis analysisfor all tested strains, except for F. acuminatum showinga bright 0.8kbp fragment. SPR signals lower than5.6RU were observed for all species with theexception of two strains of F. graminearum, one ofF. subglutinans and one of F. acuminatum that gaveRU shifts ranging from 11.4 to 18.4RU (higher thanΔRULOD for F. culmorum on culture material)indicating the presence of non-specific interaction ofthe relevant PCR products with the sensor surface.Nevertheless, these values corresponded to 0.05–0.10pg of F. culmorum DNA as template for PCRamplification, an amount over five orders of magnitudelower than those used in these experiments, i.e. 10ngof fungal genomic DNA. In the same experimentalconditions, F. culmorum strains gave RU shifts morethan 20-fold higher than other fungal strains and fallenin the upper plateau of the calibration curve,corresponding to a F. culmorum DNA amount higherthan 800pg. This indicated a good selectivity of thebiosensor assay for F. culmorum.

The SPR assay was applied to the analysis of 18durum wheat samples from field and the resultscompared with gel-electrophoresis analysis (Table 3).SPR responses were not corrected for wheat matrix

Table 3Comparison of SPR signals and gel-electrophoresis analysis of durumwheat samples

Sample # SPR analysis (ΔRU) Electrophoretic analysis a

1 39.2 −2 91.6 +3 254.6 +4 48.1 −5 57.0 − (0.8kbp)6 58.7 − (0.8kbp)7 47.6 − (0.8kbp)8 43.4 −9 32.4 − (0.8kbp)10 34.1 − (0.8kbp)11 20.7 −12 5.8 b −13 3.8 b −14 −0.1 b −15 3.4 b −16 −0.9 b −17 −3.7 b −18 9.1 b −a +, amplified product of expected size; −, no amplification or non-

specific amplification (product size indicated in brackets).b Below the threshold value of 17.4 RU.

contribution to the signal. The SPR biosensor detected F.culmorum contamination in 11 samples, although onlytwo of them, corresponding to the highest SPR signals(91.6 and 254.6RU, respectively), gave a visible band inagarose gel, thus supporting the better sensitivity of thePCR-SPR assay with respect to the conventional PCRmeasurements with gel electrophoresis. Negative resultswere obtained for the remaining seven samples by bothSPR (mean signal shift of 2.5±4.4RU) and electropho-resis assays. A non-specific amplification fragment(0.8kbp) was observed by gel-electrophoresis analysisof five wheat samples. This result suggests a possiblesample contamination with F. acuminatum due to apossible cross-reactivity of the amplification primerswith genomic DNA of this fungal species.

Finally, it is worthwhile to point out that, after eachSPR analysis, the sensor surface could be regeneratedunder alkaline conditions (pH>11) allowing at least75 hybridization cycles with a sensitivity loss lesserthan 15%.

4. Conclusions

In the present work, a SPR-based DNA sensor wasapplied for the first time to the detection of PCR-amplified products of a fungal contaminant in wheat.The use of denaturing agents, together with a thoroughcontrol of the ionic strength and a thermal denaturationtreatment, allowed high hybridization efficiency of thetarget dsDNAwith the probe. Qualitative detection of F.culmorum was performed by using the PCR-SPR assaywith better sensitivity with respect to gel-electrophore-sis, enabling to confirm the amplification selectivity bya hybridization step. Quantification of PCR productswas possible by using the SPR biosensor, although thePCR products could not be accurately correlated withthe amount of starting DNA template due to nonreproducible yields of PCR amplifications. In order toestimate the effective fungal inoculum in plant materi-als, accurate quantification of F. culmorum DNA couldbe performed by using a suitable internal standard and aSPR apparatus, equipped with three or more cells orwith SPR imaging-based arrays, that allows to detectsimultaneously PCR products relevant to the targetDNA, to the internal standard and the non-specificbackground relevant to the matrix interactions. SPRbiosensor combined with PCR amplification representsa new molecular tool for the early detection of F.culmorum that allows to improve the efficacy of controlof Fusarium head blight disease and prevent theconsequent mycotoxin contamination in wheat andother cereals.

537F. Zezza et al. / Journal of Microbiological Methods 66 (2006) 529–537

Acknowledgments

This work was supported by the Italian Ministry ofEducation, University and Research, MIUR Project No.12792 “SINSIAF” (DL 297/99, 27 July 1999).

References

Bai, G.-H., Shaner, G., 1994. Scab of wheat: prospects for control.Plant Dis. 78, 760–766.

Bianchi, N., Rutigliano, C., Tomassetti, M., Feriotto, G., Zorzato, F.,Gambari, R., 1997. Biosensor technology and surface plasmonresonance for real-time detection of HIV-1 genomic sequencesamplified by polymerase chain reaction. Clin. Diagn. Virol. 8,199–208.

Bottalico, A., Perrone, G., 2002. Toxigenic Fusarium species andmycotoxins associated with head blight in small cereals in Europe.Eur. J. Plant Pathol. 108, 611–624.

Doohan, F.M., Parry, D.W., Jenkinson, P., Nicholson, P., 1998. The useof species specific PCR based assays to analyse Fusarium earblight of wheat. Plant Pathol. 47, 197–205.

Edwards, S.G., Pirgozliev, S.R., Hare, M.C., Jenkinson, P., 2001.Quantification of trichothecene-producing Fusarium species inharvested grain by competitive PCR to determine efficacies offungicides against Fusarium head blight of winter wheat. Appl.Environ. Microbiol. 67, 1575–1580.

Feriotto, G., Lucci, M., Bianchi, N., Mischiati, C., Gambari, R., 1999.Detection of the deltaF508 (F508del) mutation of the cysticfibrosis gene by surface plasmon resonance and biosensortechnology. Human Mutat. 13, 390–400.

Feriotto, G., Corradini, R., Sforza, S., Bianchi, N., Mischiati, C.,Marchelli, R., Gambari, R., 2001. Peptide nucleic acid andbiosensors technology for real-time detection of the cystic fibrosisW1282X mutation by surface plasmon resonance. Lab. Invest. 81,1415–1427.

Feriotto, G., Borgatti, M., Mischiati, C., Bianchi, N., Gambari, R.,2002. Biosensor technology and surface plasmon resonance forreal-time detection of genetically modified roundup ready soybeangene sequences. J. Agric. Food Chem. 50, 955–962.

Feriotto, G., Gardenghi, S., Bianchi, N., Gambari, R., 2003.Quantitation of Bt-176 maize genomic sequences by surfaceplasmon resonance-based biospecific interaction analysis ofmultiplex polymerase chain reaction (PCR). J. Agric. FoodChem. 51, 4640–4646.

Feriotto, G., Breveglieri, G., Finotti, A., Gardenghi, S., Gambari, R.,2004. Real-time multiplex analysis of four beta-thalassemiamutations employing surface plasmon resonance and biosensortechnology. Lab. Invest. 84, 796–803.

Giakoumaki, E., Minunni, M., Tombelli, S., Tothill, I.E., Mascini, M.,Mogani, P., Buratti, M., 2003. Combination of amplification andpost-amplification strategies to improve optical DNA sensing.Biosens. Bioelectron. 19, 337–344.

Jönsson, U., Fägerstam, L., Ivarsson, B., Johnsson, B., Karlsson, R.,Lundh, K., Löfås, S., Persson, B., Roos, H., Rönnberg, I.,Sjölander, S., Stenberg, E., Ståhlberg, R., Urbaniczky, C., Östlin,H., Malmqvist, M., 1991. Real-time biospecific interactionanalysis using surface plasmon resonance and a sensor chiptechnology. BioTechniques 11, 620–627.

Kai, E., Ikebukuro, K., Hoshina, S., Watanabe, H., Karube, I., 2000.Detection of PCR products of Escherichia coli O157:H7 in human

stool samples using surface plasmon resonance (SPR). FEMSImmunol. Med. Micoloogy 29, 283–288.

Logrieco, A., Bottalico, A., Mulé, G., Moretti, A., Perrone, G., 2003.Epidemiology of toxigenic fungi and their associated mycotoxinsfor some Mediterranean crops. Eur. J. Plant Pathol. 109, 645–667.

Marasas, W.F.O., Nelson, P.E., Toussoun, T.A., 1984. ToxigenicFusarium Species: Identity and Mycotoxicology. PennsylvaniaState University Press, University Park, PA.

Mariotti, E., Minunni, M., Mascini, M., 2002. Surface plasmonresonance biosensor for genetically modified organisms detection.Anal. Chim. Acta. 453, 165–172.

Nicholson, P., Simpson, D.R., Weston, G., Rezanoor, H.N., Lees, A.K., Parry, D.W., Joyce, D., 1998. Detection and quantification ofFusarium culmorum and Fusarium graminearum in cereals usingPCR assays. Physiol. Mol. Plant Pathol. 53, 17–37.

Niessen, L., Vogel, R.F., 1997. Specific identification of Fusariumgraminearum by PCR with GaoA targeted primers. Syst. Appl.Microbiol. 20, 111–123.

Niessen, L., Vogel, R.F., 1998. Group specific PCR-detection ofpotential trichothecene-producing Fusarium-species in pure cul-tures and cereal samples. Syst. Appl. Microbiol. 21, 618–631.

Nilsson, P., Persson, B., Larsson, A., Uhlen, M., Nygren, P.A., 1997.Detection of mutations in PCR products from clinical samples bysurface plasmon resonance. J. Mol. Recognit. 10, 7–17.

Nirenberg, H.I., 1981. A simplified method for identifying Fusariumspp. occurring on wheat. Can. J. Bot. 59, 1599–1609.

Parry, D.W., Nicholson, P., 1996. Development of a PCR assay todetect Fusarium poae in wheat. Plant Pathol. 45, 383–391.

Parry, D.W., Jenkinson, P., McLeod, L., 1995. Fusarium ear blight(scab) in small grain cereals—a review. Plant Pathol. 44, 207–238.

Persson, B., Stenhag, K., Nilsson, P., Larsson, A., Uhlèn, M., Nygren,P.A., 1997. Analysis of oligonucleotide probe affinities usingsurface plasmon resonance: a means for mutational scanning.Anal. Biochem. 246, 34–44.

Rich, R.L., Myszka, D.G., 2000. Advances in surface plasmonresonance biosensor analysis. Curr. Opin. Biotechnol. 11, 54–61.

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: ALaboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press,New York.

Sawata, S., Kai, E., Ikebukuro, K., Iida, T., Honda, T., Karube, I.,1997. Novel detection system of flow injection analysis (1). Theexistence of significant relation between secondary structure ofDNA and sensitivity in signal detection. Nucleic Acids Symp. Ser.37, 247–248.

Sawata, S., Kai, E., Ikebukuro, K., Iida, T., Honda, T., Karube, I.,1999. Application of peptide nucleic acid to the direct detection ofdeoxyribonucleic acid amplified by polymerase chain reaction.Biosens. Bioelectron. 14, 397–404.

Schilling, A., Möller, E., Geiger, H., 1996. Polymerase chain reaction-based assays for species-specific detection of Fusarium culmorum,F. graminearum, and F. avenaceum. Phytopathology 86, 515–522.

Southern, E.M., 1975. Detection of specific sequences among DNAfragments separated by gel electrophoresis. J. Mol. Biol. 98,503–517.

Wilson, I.G., 1997. Inhibition and facilitation of nucleic acidamplification. Appl. Environ. Microbiol. 63, 3741–3751.

Wilson, A., Simpson, D., Chandler, E., Jennings, P., Nicholson, P.,2004. Development of PCR assays for the detection anddifferentiation of Fusarium sporotrichioides and Fusarium lang-sethiae. FEMS Microbiol. Lett. 233, 69–76.