International Journal of Food Microbiology xxx (2008) xxx-xxx

FOOD-04415; No of Pages 7

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j ourna l homepage: www.e lsev ie r.com/ locate / i j foodmicro

ARTICLE IN PRESS

Application of whole genome amplification and quantitative PCR for detection andquantification of spoilage yeasts in orange juice

Angelique Renard, Perla Gómez di Marco, Marcos Egea-Cortines, Julia Weiss ⁎Agricultural Science and Technology Department, Genetics, Institute of Plant Biotechnology, Universidad Politécnica de Cartagena, 30203 Cartagena, Spain

⁎ Corresponding author. Tel.: +34 986 325705; fax: +3E-mail address: julia.weiss@upct.es (J. Weiss).

0168-1605/$ – see front matter © 2008 Elsevier B.V. Aldoi:10.1016/j.ijfoodmicro.2008.05.021

Please cite this article as: Renard, A., et al., Aof spoilage yeasts in orange juice, Internati

A B S T R A C T

A R T I C L E I N F O

Article history:

Small cell numbers in com Received 20 August 2007Received in revised form 16 May 2008Accepted 17 May 2008Available online xxxx

Keywords:Saccharomyces cerevisiaeHanseniaspora uvarumWhole genome amplificationQuantitative PCRMelting point analysisStrand displacement amplification

plex food matrices and undefined PCR inhibitors often limit detection andidentification of DNA species by molecular techniques. Thus in many industrial situations enrichmentgrowths are performed. However, growth speed of different species in complex microbial mixtures in definedmedia is in most cases different, thus final results do not always reflect the starting situation. We tested DNA-strand displacement whole genome amplification as a possible substitute of enrichment growth. Using wholegenome amplification on orange juice contaminated with Saccharomyces cerevisiae, we lowered detectionlevel from 106 down to 102 cfu/ml. Whole genome amplification showed to be linear (R=0.992) and therelative yeast DNA copy number compared to other DNA templates did not change thus allowing quantitativeestimation of initial contamination by quantitative PCR. Using melting point analysis, we were able todistinguish between the PCR products of the 5.8S-ITS region, obtained with universal primers from purecultures of S. cerevisiae and Hanseniaspora uvarum, two major spoilage yeasts in orange juice and formingpart of wine microbiota during fermentation. However, in mixed-contaminated samples, identification ofboth species was hampered by preferential appearance of the melting peak coinciding with H. uvarum,except when S. cerevisiae was the dominating species. Application of whole genome amplification did notprevent the preferential detection of H. uvarum. This handicap was resolved by applying an enrichmentprocedure up to saturation after which the melting peak of both species could clearly be identified.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Yeasts play a central role in the spoilage of a wide range of foodsand beverages, especially those with a high acidity like orange juice,since they can resist extreme conditions better than bacteria and cangrow at refrigeration temperatures (Arias et al., 2002; Wyatt et al.,1995). Food-spoilage yeasts include many different genera andspecies, such as Candida, Pichia, Rhodotorula, Torulopsis, Saccharo-myces, Zygossacharomyces, Hansenula and Trichosporon (Souza et al.,2007; Loureiro and Malfeito-Ferreira, 2003). Among the importantspoilage yeasts in fresh or pasteurised orange juice are Saccharomycescerevisiae and Hanseniaspora uvarum (Arias et al., 2002). Yeastsgenerally are not a toxic risk for humans, except for someopportunistic pathogens like Candida albicans or Cryptococcus neofor-mans, but they affect negatively juice's flavour, turbidity and odour(Fleet, 2007; Tournas et al., 2006).

Most commercially processed juices are heated to 75–85 °C for 1–3min (Ingallinera et al., 2005) or twice to 99 °C for 15 swith intermittentcooling (Ros-Chumillas et al., 2007) in order to obtain safe micro-biological conditions. Although juice spoilage can be minimized by

4 968 325433.

l rights reserved.

pplication of whole genomeonal Journal of Food Microb

applying sanitary measures and low-temperature storage, microbio-logical spoilage problems occur (Arias et al., 2002). The detection andenumeration of microorganisms by traditional plating techniquesusing differential or selective media are accurate, discriminative andcan detect lowest contamination levels but is time demanding with 3–7 days in the case of yeasts (Loureiro, 2002; de Boer and Beumer,1999;Pascual Anderson and Calderon y Pascual, 1999; van der Vossen andHofstra, 1996).

Techniques based on PCR are another option for yeast detection.One approach lies in the amplification of the rRNA internal transcribedspacer (ITS) region including the 5.8 rRNA gene. Differences in the ITSregion can be used to discriminate between yeast species, since theyshowhigh interspecific polymorphism (LasHeras-Vazquez et al., 2003;White et al., 1990; Fleet, 2007), which can be identified either byelectrophoretic size discrimination in conventional PCR or RFLP (LasHeras-Vazquez et al., 2003). An alternative is real-time or quantitativePCR (qPCR), either in a lightcycler with fluorescent double strandspecific SYBR GREEN I dye based on melting point analysis (Casey andDobson, 2004; Phister and Mills, 2003; Phister et al., 2007) or bysequence-specific TaqMan analysis (Haugland et al., 2002). Whenusing melting point analysis, it needs to be considered that amplifica-tion with universal primers from multitemplate samples can lead tothe formation of heteroduplex molecules (Kanagawa, 2003) and theirmelting might diverge from those of homoduplex molecules.

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

2 A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

Pre-detection enrichment is sometimes unavoidable due to theproblematic of recovering small cell numbers from complex foodmatrices and the existence of unidentified PCR inhibitors (Mukho-padhyay and Mukhopadhyay, 2007; Li and Mustapha, 2004). Forexample, the detection of spoilage wine yeasts present on the surfaceof grapes requires enrichment in order to overcome detection limitsby molecular methods (Renouf and Lonvaud-Funel, 2007). Amplifica-tion of the ITS region directly from S. cerevisiae contaminated orangejuice without preenrichment is possible and can be improved by atechnique using simple glass bead disruption of yeast cells within thejuice in combination with silica absorption of DNA (Ros-Chumillaset al., 2007). The detection level using this protocol lies at 104–103 cfu/ml of juice samples at convenient sample volumes of 2 ml eliminatingall possible PCR inhibiting substances in the DNA extraction. The juicepasteurisation process in combination with the DNA degradativeeffect of juice acidity was shown to eliminate any naked DNA so thatamplification products are considered to derive from recontamination(Ros-Chumillas et al., 2007; Weiss et al., 2007).

Yeasts hardly occur as single cultures or single microorganisms infoods and beverages with the exception of highly processed productsand in the case of spoilage outbreaks (Fleet, 2007). Differential growthof microorganisms on a given media is commonplace in microbialecology. Indeed it has been shown in wine fermentation, thatinoculation with active dry yeast strains changes wild strain diversitycompared to non-inoculated wine, possibly due to perturbation of thestrain equilibrium, favouring the development of minority wild strains(Querol et al., 1992).

It is known from bacterial contamination that detection limitsdepend on the specific food-product and food-spoilage organisms.After a 24-hour enrichment period, Li and Mustapha (2004) detectedthe three pathogens E. coli O157:H7, Salmonella and Shigella at aninitial inoculation level of 8×10−1/g or ml in apple cider, cantaloupe,lettuce, tomato and watermelon and 8×101/g in alfalfa (Li andMustapha 2004). Mukhopadhyay andMukhopadhyay (2007) detected101 cfu simultaneously from E. coli O157:H7 and Listeria monocyto-genes in preenriched, artificially inoculated broth (Mukhopadhyay andMukhopadhyay, 2007). In post-enriched broth, only E. coli has beenamplified across all levels indicating that unequal growth leads to adisparity in the amount of template DNA in the multiplex PCR anddisappearance of signal DNA from the underrepresented species. AqPCR assay developed to detect Hanseniaspora species in mustand wine with Hanseniaspora specific primers, was reported not tobe impacted by high concentrations of S. cerevisiae until levels ofHanseniaspora are below 100 cfu/ml (Phister et al., 2007).

Reasons for PCR constraints may be due to unequal accessibility oftemplate DNA, dissimilar efficiency in the formation of primer andtemplate hybrids, different polymerization efficiency as well asuneven substrate exhaustion in combination with decreasing primerconcentrations and increasing primer competition (Crosby andCriddle, 2007). Thus if a beverage, like wine or orange juice, containsseveral yeast species, not all of them might be detectable in adiscriminative manner by PCR, not even after enrichment.

Whole genome amplification (WGA) based on DNA-strand displace-ment is an option to amplify DNAwhen the type and amount of sampleavailable is a limiting factor, either due to degradation or low copynumber. The DNA polymerase of the phage Phi29 (Φ29) replicates thegenome exponentially in the form of multiple displacement amplifica-tion, creating microgram quantities from sub-nanogram template DNAlevels. A commercially available kit is GenomiPhi™. Phi29 DNApolymerase was also used for partial genome sequencing of a soilarchaeon from single cell environmental isolates (Kvist et al., 2007).

This work evaluated the applicability of a whole genomeamplification procedure using Phi29 DNA polymerase in order todetect low contamination levels of spoilage organisms directly fromthe food matrix ‘commercial orange juice’ that was shown to containPCR inhibiting substances (Ros-Chumillas et al., 2007) by increasing

Please cite this article as: Renard, A., et al., Application of whole genomeof spoilage yeasts in orange juice, International Journal of Food Microb

genomic DNA quantity. We also assessed, whether melting pointanalysis in qPCR with a universal primer pair for the 5.8SrDNA andadjacent ITS regions is applicable for detecting and identifyingspoilage yeasts in mixed contaminations with and without enrich-ment culture and if simultaneous detection could be improved byapplication of whole genome amplification techniques.

2. Materials and methods

2.1. Yeast strains, culture medium and conditions

The adenin-auxotroph S. cerevisiae strain HF7c, developed byFeilotter et al. (1994) was used (Feilotter et al., 1994). The H. uvarumyeast culture (Nr. CECT 1444) was obtained from the ‘ColecciónEspañola de Cultivos Tipo (CECT)’, Valencia, Spain. Adenin auxotrophyof HF7c leads to red color upon overgrowth (Bender and Pringle,1991),thus allowing to steadily control purity of our yeast strains. Yeastcultures were grown in YPD broth (1% (w/v) yeast extract, 2%(w/v)peptone, 2% (w/v) dextrose) at 30 °C. Cell concentration wasdetermined spectrophotometrically at an optical density (OD) of600 nm. Number of colony forming units (cfu) in serial dilutions wasconfirmed through duplicate counts on YPD agar (2%) plates afterincubation at 30 °C for 3 days. Culturesweremaintained at 4 °C on YPDagar (2%) plates.

2.2. Serial dilutions in YPD and orange juice

Yeasts were grown to an OD600 of 1.0. Serial dilutions of S.cerevisiae between 106 and 100 cfu/ml were done in orange juice. Forthe enrichment culture experiment, we inoculated YPD broth with103 cfu/ml of S. cerevisiae and H. uvarum. Mixed inoculation in orangejuice consisted of 106 cfu/ml of either yeast species combined withserial dilutions between 100 and 106 cfu/ml of the second yeastspecies. We used commercially available 100% orange juice with pulp,minimally pasteurised and bottled in tetrapack (Hacendado, Merca-dona white label).

2.3. Protocols for yeast DNA isolation from orange juice and YPD

Yeast DNA extractions from 1 ml samples of YPD and orange juicewere performed as described (Ros-Chumillas et al., 2007) using 400 μlyeast extraction buffer (2% Triton X-100, 1% sodium dodecyl sulfate,100 mM NaCl, 10 mM Tris–HCl, pH 8.0, 10 mM EDTA) (Las Heras-Vazquez et al., 2003) in combination with cell disruption techniquebased on acid washed glass beads (Sigma, 150–212 μm), a conicalgrinder exactly fitting to the tube and a strong vortex. DNA waspurified by adding phenol–chloroform–isoamylalcohol (25:24:1),followed by isopropanol precipitation. DNA was finally eluted into20 μl H2O and 2 μl was used for conventional and qPCR.

2.4. PCR protocols

The 5.8S-ITS region was amplified by conventional or qPCR usingthe primers ITS1 (5′-TCCGTAGGTGAACCTGCGg-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) described by White et al. (1990). PCRwas performed in a final volume of 25 μl containing 2 μl of theextracted DNA, 0.08 mM each of dATP, dCTP, dGTP and dTTP, 1.5 mMMgCl2, 1× PCR buffer and 1 U GoTaq Flexi DNA polymerase (Promega,Madison, WI, USA). Absolute DNA concentrations in the samples takendepended on DNA source, cell titer and whole genome amplificationefficiency. The amplification was carried out as follows: initialdenaturation at 94 °C for 5 min, 30 cycles of 94 °C for 10 s, 55.3 °Cfor 30 s and 72 °C for 60 s, and a final extension at 72 °C for 5 min.Amplification was performed in a GeneAmp PCR System 9700Thermocycler from PE Applied Biosystems. qPCR was performed ona Rotor Gene 2000 Real time cycler from Corbett Research (Sydney,

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

Fig. 2. Standard quantification curve plotting the Ct values for the yeast DNA extractionsagainst yeast concentration in cfu/ml.

3A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

Australia) using the SYBR GREEN I dye ‘SYBR Premix Ex Taq™’ (TakaraBIO INC, Japan). The PCR conditions were an initial denaturation of5 min at 95 °C followed by 40 cycles with a scheme of 95 °C 30 s, 55 °C30 s, 72 °C 30 s, a read at 83 °C for 15 s and finally a melting pointanalysis starting at 60 °C till 94 °C with reads every 0.5 °C for 15 s. Datawere analysed using the Rotor-Gene Analysis Software, developing astandard quantification curve based on Ct values which show the cyclenumber at which the fluorescence crosses a threshold line defined aspoint of logarithmic increase in fluorescence (http://pathmicro.med.sc.edu/pcr/realtime-home.htm).

PCR amplification products were analysed together with amolecular weight ladder by electrophoresis on a 2% agarose gelcontaining ethidium bromide in Tris-acetate–EDTA buffer (TAE:40 mM Tris-acetate, 1 mM EDTA).

2.5. Whole genome amplification with GenomiPhi

Whole genome amplification was performed with the ‘illustraGenomiPhi V2 DNA Amplification Kit’ (GE Healthcare UK, Buckin-ghamshire, UK) according to themanufacturer's instructions. One μl ofyeast DNA from the various dilutions and from H2O or orange juice asnegative controls was added to 9 μl of GenomiPhi sample bufferfollowed by denaturation at 95 °C for 3 min and cooling on ice. Ten μlof reaction buffer containing dNTPs, random hexamers and Phi29 DNApolymerase were added and reactions were incubated at 30 °C for1.5 h. The polymerase was then inactivated by incubation at 65 °C for10 min. Reaction products were diluted 1:10 and 2 μl of this dilutionwas used in qPCR.

3. Results

3.1. Improved detection of yeast contamination by whole genomeamplification

Yeast DNA can be detected by conventional PCR directly fromorange juice without enrichment procedure at a concentration of106 cfu/ml resulting in a PCR product for the 5.8S-ITS regionwith a sizeof 880 bp (Ros-Chumillas et al., 2007; Las Heras-Vazquez et al., 2003).It is desirable to lower the detection limit, which we tried hereperforming a whole genome amplification step before PCR detection.We applied the whole genome amplification (WGA) kit ‘GenomiPhi’directly to the crude yeast DNA extract of 1 ml orange juice in order toinvestigate whether yeast DNA is amplified in a background of foreigngenomic DNA, thus improving detection levels. After extraction wediluted DNA in 20 μl H2O. Lower volumes were not feasible due to thehigh gelatinous character of DNA pellets. One μl of this DNA extractwas used for WGA and subsequent PCR reactions were done using 2 μl

Fig. 1. PCR product of 5.8S-ITS region (A) before and (B) after WGA. Yeast DNA wasextracted from 1 ml orange juice containing 100–106 cfu of S. cerevisiae. DNA wasdissolved in 20 μl H2O and 1 μl was used for GenomiPhi amplification. 1 μl of a 1/10dilutionwas used for subsequent PCR. Cgm = GenomiPhi kit applied to H2O; C− = negativePCR control.

Please cite this article as: Renard, A., et al., Application of whole genomeof spoilage yeasts in orange juice, International Journal of Food Microb

of 1/10 dilutions since PCR reactions directly from the WGA-reaction,even after precipitation, proofed inferior (data not shown). BeforeWGA, a product was detected in qPCR based on melting peakobservation and agarose gel electrophoresis (Fig. 1A) down to105 cfu and after WGA for as little as 102 cfu (Fig. 1B) When plottingthe Ct values for the serial yeast dilutions against concentration ofyeast in cfu/ml, a linear trend was observed (Fig. 2) suggesting thatGenomiPhi amplification is proportional to DNA concentration andthat the standard curve can be used to quantify initial yeastcontamination levels.

3.2. Conventional PCR versus DNA melting point analysis to detect yeastspecies

Although amplicons of the ITS regions from different microorgan-isms can sometimes directly be distinguished on agarose gels,differentiation between PCR products of similar size might befacilitated by melting point analysis (Manchado-Rojo et al., 2008).We amplified the 5.8S-ITS region from H. uvarum and S. cerevisiaeboth by conventional and real-time PCR in order to evaluatedifferentiation capacities. Product size can be directly distinguishedusing agarose gel electrophoresis with an amplicon size of 750 bp inH. uvarum and 880 bp in S. cerevisiae. Melting points in qPCR lie at

Fig. 3. A) PCR product of 5.8S-ITS region from H. uvarum and S. cerevisiae afterconventional PCR. B) Melting peak analysis identifying the specific melting temperaturefor H. uvarum and S. cerevisiae using primers ITS1 and ITS4.

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

Fig. 5. A) Melting peaks for the 5.8S-ITS region amplified from mixed-contaminatedYPD medium during enrichment culture. Yeast DNA was extracted from 1 ml samplesafter 0 h, OD600=0; 9 h, OD600=0.162 and 24 h, OD600=2.503. B) Agarose gelelectrophoresis of the real-time PCR products after 0, 9 and 24 h. C− = negative controlH2O.

Fig. 4. A) Agarose gel electrophoresis of real-time PCR products from 5.8S-ITS region. a–f: 106 cfu H. uvarum co-inoculated with 101, 102, 103, 104, 105, and 106 S. cerevisiae; g–l: 106 cfuS. cerevisiae co-inoculated with 101, 102, 103, 104, 105, and 106 H. uvarum. B) Melting peaks for the 5.8S-ITS region amplified from mixed-contaminated orange juice. Yeast DNA wasextracted from 1 ml juice artificially co-inoculated with different concentrations of H. uvarum and S. cerevisiae; control = H2O.

4 A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

83.6 °C forH. uvarum and 85.4 °C for S. cerevisiae (Fig. 3A and B). Risingthe annealing temperature for primer hybridisation from 55 °C, theoptimal Tm calculated for primers in conventional PCR, to 62 °C,promoted the appearance of clear melting peaks probably by reducingmispriming events (data not shown). Differentiation of the two yeastspecies both on agarose gel and bymelting peak analysis allowed us todirectly compare and evaluate these two analysis methods.

3.3. Detection of mixed contamination in orange juice by DNA meltingpoint analysis

In order to estimate the feasibility of melting point analysis todetect mixed contaminations, we mimicked unequilibrated contam-ination levels of microorganisms. We inoculated orange juice withincreasing concentrations (101 cfu–106 cfu/ml) of one yeast speciesand a constant high level (106 cfu/ml) of the other. Agarose gelelectrophoresis of qPCR products (Fig. 4A) showed that in mixedcontaminations with unequal concentrations, the 5.8S-ITS productfrom S. cerevisiae is detected at a concentration of 106 cfu and that ofH. uvarum at 105 cfu. Melting peak analysis reflected these observa-tions (Fig. 4B) although detection of both yeasts at higher concentra-tion was difficult since the peak coinciding with the Tm of H. uvarumappeared preferentially.

3.4. Genotyping yeast species in mixed-contaminated enrichment cultureby DNA melting point analysis

We inoculated YPDbrothwith 103 cfu ofH. uvarum and S. cerevisiae,a concentration that was shown to be detectable by conventional PCR(Ros-Chumillas et al., 2007). At time 0, the 5.8S-ITS product of bothyeast species appeared on the gel (Fig. 5B). During enrichment growth,between 6 and 9 h, the dominant band on agarose gel was that ofS. cerevisiae (representatively shown for 9 h) suggesting that in YPD

Please cite this article as: Renard, A., et al., Application of whole genome amplification and quantitative PCR for detection and quantificationof spoilage yeasts in orange juice, International Journal of Food Microbiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

Fig. 6. Agarose gel electrophoresis of real-time PCR products form 5.8S-ITS region after WGA A) a:106 cfu H. uvarum; b–h: 106 cfu H. uvarum co-inoculated with 100, 101,102, 103, 104,105, and 106 S. cerevisiae; B) a: 106 cfu S. cerevisiae; b–h:106 cfu S. cerevisiae co-inoculated with 101, 102, 103, 104, 105, and 106 H. uvarum; i: GenomiPhi kit applied to H2O, j: negativecontrol, PCR with H2O.

5A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

medium, S. cerevisiae overgrows H. uvarum. This assumption wasconfirmed by plate assay showing 50% more red colonies of S.cerevisiae if plated after 9 h of growth in liquid broth (data notshown). At saturation after 24 h, amplification products for bothyeastscan be detected, but S. cerevisiae consistently gave stronger signals.Melting peak analysis confirmed the observations from agarose gel(Fig. 5A), although the prominent peak at low concentrations (0 h)coincided with the Tm of H. uvarum.

3.5. Detection of mixed contamination in orange juice after WGA by qPCR

After WGA the 5.8S-ITS PCR product from H. uvarum in mixedcontamination with constant high concentrations of S. cerevisiae canbe detected on agarose gel already at a concentration of 104 cfu(Fig. 6B). In mixtures with constant high concentrations of H. uvarum,we could not lower detection level of S. cerevisiae (Fig. 6A) and only a

Fig. 7.Melting peaks for the 5.8S-ITS region amplified frommixed-contaminated orange juiceH. uvarum+106 or 102 cfu S. cerevisiae and B) 106 cfu S. cerevisiae+106 or 102 cfu H. uvarum

Please cite this article as: Renard, A., et al., Application of whole genomeof spoilage yeasts in orange juice, International Journal of Food Microb

melting peak for H. uvarum appeared throughout all concentrations(Fig. 7A). At high concentrations of both yeasts (106 cfu/ml each) thedominating melting peak was that of H. uvarum (Fig. 7B).

4. Discussion

There are two distinct scenarios of yeast detection assays in thefood industry. One describes those cases where yeast represents anunwanted contamination and any trace of contamination has to behandled accordingly. And a second one that happens in fermentationswhere yeasts are always present but the combination during specificfermentation steps is important for the outcome of the process(Phister et al., 2007). In both cases, PCR assays for detection ofcontamination in foods and drinks are often difficult due to poorrecovery of small cell numbers from complex food matrices, low copynumber quantities of DNA and the existence of unidentified PCR

afterWGA. Yeast DNAwas extracted from 1ml juice artificially co-inoculated A) 106 cfu.

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

6 A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

inhibitors (Mukhopadhyay and Mukhopadhyay, 2007; Li and Musta-pha, 2004). Whole genome amplification based on DNA-stranddisplacement with Phi29 Taq polymerase was shown to be able toproduce microgram quantities of DNA starting from sub-nanogramlevels of template DNA (Ballantyne et al., 2007). This studyinvestigated the use of WGA for the detection of low concentrationsof yeast DNA, complexed with DNA of oranges in orange juice.Promising results were obtained, since we were able to detect as littleas 102 cfu/ml after WGA compared to 105 cfu/ml before WGA, thusincreasing detection level by three orders of magnitude. Furthermore,amplification of yeast DNA was proportional to the initial concentra-tion, which means that the relative copy number of yeast DNA is notchanged by the experimental procedure. Thus initial yeast concentra-tions can be calculated using standard curves derived from real-timePCR results after WGA. This method could be valuable in earlydetection of spoilage outbreaks of single yeast species, includingopportunistic pathogens like C. albicans or C. neoformans (Fleet, 2007),in pasteurised beverages, allowing immediate implementation ofintervention measures.

Another important aspect of yeast detection is monitoring thedevelopment of species populations during fermentation since qualityof the product is often linked to original microbial flora (Phister andMills, 2003; Querol et al., 1992). Melting peak analysis of PCR productsfrom the 5.8-ITS region, amplified with unspecific primers and usingfluorescent SYBR GREEN I dye was shown to be applicable inphenotyping a wide range of yeast species (Casey and Dobson,2004). We investigated if this technique is applicable for thedifferentiation between S. cerevisiae and H. uvarum, two importantspoilage yeasts in orange juice. But S. cerevisiae and H. uvarum alsoform part of wine microbia during fermentation and persistentgrowth of H. uvarum throughout wine fermentation is considered tohave a negative impact on wine quality (Phister and Mills, 2003;Querol et al., 1992; Loureiro and Malfeito-Ferreira, 2003). The Tmvalues of PCR products from the 5.8-ITS region, amplified from purecultures, differed by 1.8 °C, thus allowing differentiation by meltingpeak analysis. PCR products could also be identified by electophoresison agarose gel, since fragment size differs by 130 bp.

Our study investigated further if the technique of melting pointanalysis after amplificationwithuniversal primers is applicable to detectand identify S. cerevisiae and H. uvarum from mixed, but differentiallycontaminated orange juice and if this technique is preferable or canreplace gel electrophoresis. We found that S. cerevisiae can be detectedon agarose gels at a concentration of 106 cfu/ml in case of a highbackground contaminationwithH. uvarum, andH. uvarum at 105 cfu/mlin case of a high background contamination with S. cerevisiae, whichcoincides with detection levels from pure cultures.

Using melting peak analysis we were confronted with the problemthat the melting peak of H. uvarum appeared preferentially when bothspecies have identical quantities of 106 cfu/ml, possibly due to anamplification bias between the two DNA sources at this contaminationlevel. We conclude that melting peak analysis is not recommended forsimultaneous detection of various yeast species with universalprimers directly from the food source.

Enrichment culture is a method to increase cell number. Weperformed enrichment in YPD medium, a broth optimized for thecultivation of S. cerevisiae. Growth advantage of S. cerevisiae wasconfirmed by plating aliquots of liquid enrichment cultures on agarplates in 3-hour intervals. During the period of intense S. cerevisiaegrowth, we only observed the amplification product of S. cerevisiae,while at saturation both PCR products could be seen. Equally, themelting peaks for both species could be unequivocally identified onlyat saturation. We conclude that the high concentration of S. cerevisiaetemplate DNA during the first hours of enrichment culture competeswith H. uvarum template DNA for amplification. After growth tosaturation both species can be successfully detected by melting pointanalysis.

Please cite this article as: Renard, A., et al., Application of whole genomeof spoilage yeasts in orange juice, International Journal of Food Microb

Since WGA had proved to increase detection level in S. cerevisiaecontaminated orange juice, we tried to enrich for template DNA ofboth yeast species in mixed, differential contamination by applyingWGA to the respective DNA extracts. We could detect H. uvarum in ahigh S. cerevisiae background on agarose gel at 104 cfu/ml. We couldnot lower detection level for S. cerevisiae in a high H. uvarumbackground and only the melting peak of H. uvarum appearedindicating again an amplification advantage of the smaller H. uvarum5.8S-ITS product, which is even more pronounced due to WGA.

We finally conclude that WGA is a promising method whendetection of very low contamination levels is required. For simulta-neous detection and identification of mixed yeast populations bymeltingpoint analysiswe found enrichment culture favourable toWGA.

Acknowledgements

The authors thank Prof. Dr. Francisco Artés-Calero for his support.A.R. was granted by the Polytech ‘Clermont-Ferrand’, Université BlaisePascal, 63174 Aubière Cedex, France. The Seneca Foundation sup-ported the research.

References

Arias, C.R., Burns, J.K., Friedrich, L.M., Goodrich, R.M., Parish, M.E., 2002. Yeast speciesassociated with orange juice: evaluation of different identification methods.Applied and Environmental Microbiology 68, 1955–1961.

Ballantyne, K.N., van Oorschot, R.A.H., Mitchell, R.J., 2007. Comparison of two wholegenome amplification methods for STR genotyping of LCN and degraded DNAsamples. Forensic Science International 166, 35–41.

Bender, A., Pringle, J.R., 1991. Use of a screen for synthetic lethal andmulticopy suppresseemutants to identify two new genes involved in morphogenesis in Saccharomycescerevisiae. Molecular and Cellular Biology 11, 1295–1305.

Casey, G.D., Dobson, A.D., 2004. Potential of using real-time PCR-based detection ofspoilage yeast in fruit juice—a preliminary study. International Journal of FoodMicrobiology 91, 327–335.

Crosby, L.D., Criddle, C.S., 2007. Gene capture and random amplification for quantitativerecovery of homologous genes. Molecular and Cellular Probes 21, 140–147.

de Boer, E., Beumer, R.R., 1999. Methodology for detection and typing of foodbornemicroorganisms. International Journal of Food Microbiology 50, 119–130.

Feilotter, H.E., Hannon, G.J., Ruddell, C.J., Beach, D., 1994. Construction of an improvedhost strain for two hybrid screening. Nucleic Acids Research 22, 1502–1503.

Fleet, G.H., 2007. Yeasts in foods and beverages: impact on product quality and safety.Current Opinions in Biotechnology 18, 170–175.

Haugland, R.A., Brinkman, N., Vesper, S.J., 2002. Evaluation of rapid DNA extractionmethods for the quantitative detection of fungi using real-time PCR analysis.Journal of Microbiological Methods 50, 319–323.

Ingallinera, B., Barbagallo, R.N., Spagna, G., Palmeri, R., Todaro, A., 2005. Effects ofthermal treatments on pectinesterase activity determined on blood orange juices.Enzyme and microbial technology 36, 258–263.

Kanagawa, T., 2003. Bias and artifacts in multitemplate polymerase chain reactions(PCR). Journal of Bioscience and Bioengineering 96, 317–323.

Kvist, T., Ahring, B.K., Lasken, R.S., Westermann, P., 2007. Specific single-cell isolationand genomic amplification of uncultured microorganisms. Applied Microbiologyand Biotechnology 74, 926–935.

Las Heras-Vazquez, F.J., Mingorance-Cazorla, L., Clemente-Jimenez, J.M., Rodriguez-Vico, F., 2003. Identification of yeast species from orange fruit and juice by RFLP andsequence analysis of the 5.8S rRNA gene and the two internal transcribed spacers.FEMS Yeast Research 3, 3–9.

Li, Y., Mustapha, A., 2004. Simultaneous detection of Escherichia coli O157:H7, Salmonella,and Shigella in apple cider and produce by a multiplex PCR. Journal of Food Protection67, 27–33.

Loureiro, V., 2002. Spoilage yeasts in foods and beverages: characterization and ecologyfor improved diagnosis and control. Food Research International 33, 247–256.

Loureiro, V., Malfeito-Ferreira, M., 2003. Spoilage yeasts in the wine industry.International Journal of Food Microbiology 86, 23–50.

Manchado-Rojo, M., Weiss, J., Egea-Cortines, M., 2008. Using 23 rDNA to identifycontaminations of E. coli in Agrobacterium tumefaciens cultures. AnalyticalBiochemistry 372, 253–254.

Mukhopadhyay, A., Mukhopadhyay, K., 2007. Novel multiplex PCR approaches for thesimultaneous detection of human pathogens: Escherichia coli 0157:H7 and Listeriamonocytogenes. Journal of Microbiological Methods 68, 193–200.

Pascual Anderson, M.R., Calderon y Pascual, V., 1999. Microbiología Alimentaria. Díaz deSantos, Madrid.

Phister, T.G., Mills, D.A., 2003. Real-time PCR assay for detection and enumeration ofDekkera bruxellensis in wine. Applied and Environmental Microbiology 69,7430–7434.

Phister, T.G., Rawsthorne, H., Joseph, C.M. Lucy, Mills, D.A., 2007. Real-time PCR assay fordetection and enumeration ofHanseniaspora species fromwine and juice. AmericanJournal of Enology and Viticulture 58, 229–233.

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021

7A. Renard et al. / International Journal of Food Microbiology xxx (2008) xxx-xxx

ARTICLE IN PRESS

Querol, A., Barrio, E., Huerta, T., Ramon, D., 1992. Molecular monitoring of winefermentations conducted by active dry yeast strains. Applied and EnvironmentalMicrobiology 58, 2948–2953.

Renouf, V., Lonvaud-Funel, A., 2007. Development of an enrichment medium to detectDekkera/Brettanomyces bruxellensis, a spoilage wine yeast., on the surface of grapeberries. Microbiological Research 162, 154–167.

Ros-Chumillas, M., Egea-Cortines, M., López, A., Weiss, J., 2007. Evaluation of a rapidDNA extractionmethod to detect yeast cells by PCR in orange juice. Food Control 18,33–39.

Souza, E.L., Stamford, T.L.M., Lima, E.O., Trajano, V.N., 2007. Effectiveness of Origanumvulgare L. essential oil to inhibit the growth of food spoiling yeasts. Food Control 18,409–413.

Tournas, V.H., Heeres, J., Burgess, L., 2006. Moulds and yeasts in fruit salads and fruitjuices. Food Microbiology 23, 684–688.

Please cite this article as: Renard, A., et al., Application of whole genomeof spoilage yeasts in orange juice, International Journal of Food Microb

van der Vossen, J.M., Hofstra, H., 1996. DNA based typing, identification and detectionsystems for food spoilage microorganisms: development and implementation.International Journal of Food Microbiology 33, 35–49.

Weiss, J., Ros-Chumillas, M., Pena, L., Egea-Cortines, M., 2007. Effect of storage andprocessing on plasmid, yeast and plant genomic DNA stability in juice fromgenetically modified oranges. Journal of Biotechnolology 128, 194–203.

White, T.J., Burns, T., Lee, S., Taylor, J.W., 1990. Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics. In: Innis, M.A., Gelfand, D.H.,Sninsky, T.J., White, T.J. (Eds.), PCR Protocols: a Guide to Methods and Applications.Academic Press, San Diago, CA, USA, pp. 315–322.

Wyatt, M.K., Parish, M.E., Widmer, W.W., Kimbrough, J., 1995. Characterization of moldgrowth in orange juice. Food Microbiology 12, 347–355.

amplification and quantitative PCR for detection and quantificationiology (2008), doi:10.1016/j.ijfoodmicro.2008.05.021