FOOD MICROBIOLOGY Food Microbiology 24 (2007) 393–402 Rhizopus oligosporus and yeast co-cultivation during barley tempeh fermentation—Nutritional impact and real-time PCR quantification of fungal growth dynamics Xin Mei Feng a, , Volkmar Passoth a , Charlotte Eklund-Jonsson b , Marie Larsson Alminger b , Johan Schnu¨rer a a Department of Microbiology, Swedish University of Agricultural Sciences (SLU), Box 7025, SE-750 07 Uppsala, Sweden b Department of Chemical and Biological Engineering, Food Science, Chalmers University of Technology, SE-412 96 Go¨teborg, Sweden Received 3 March 2006; received in revised form 2 June 2006; accepted 23 June 2006 Available online 30 August 2006 Abstract Barley tempeh was produced by fermenting barley kernels with Rhizopus oligosporus. The potential of the yeasts Saccharomyces cerevisiae (three strains), S. boulardii (one strain), Pichia anomala (one strain) and Kluyveromyces lactis (one strain) to grow together with R. oligosporus during barley tempeh fermentation was evaluated. All yeast strains grew during the fermentation and even during cold storage of tempeh (Po0:01). The growth of yeasts slightly increased the ergosterol contents, but did not influence amino acid contents and compositions, and did not reduce phytate contents. Slight increases of vitamins B 6 and niacinamide, and slight decreases of B 1 and biotin were observed. Quantification of fungal growth is difficult during mixed species fermentations because ergosterol is found in all fungal species, and colony-forming-unit (cfu) estimations are not reliable for R. oligosporus and other sporulating fungi. Therefore, we developed a quantitative real-time PCR method for individually quantifying S. cerevisiae and R. oligosporus growth in barley tempeh. The PCR results were highly correlated with the ergosterol content of R. oligosporus and with the number of cfu of S. cerevisiae. Thus, real-time PCR is a rapid and selective method to quantify yeasts and R. oligosporus during mixed species fermentation of inhomogenous substrate such as barley tempeh. r 2006 Elsevier Ltd. All rights reserved. Keywords: Yeast; Rhizopus oligosporus; Nutrition; Barley tempeh; Real-time PCR 1. Introduction Tempeh is a traditional Indonesian food prepared by fermentation of dehulled and cooked soybeans with moulds of the genus Rhizopus (mainly Rhizopus oligos- porus) to a compact and sliceable cake (Shama and Hall, 1991). Tempeh can also be produced from cereal grains, e.g. barley, through fermentation with R. oligosporus (Berg et al., 2001; Feng et al., 2005). Phytate (myo-inositol hexakisphosphate or IP6) is the principal storage form of phosphorus in cereal grains and other plant materials (Reddy et al., 1982). Minerals, such as iron, magnesium and zinc, are strongly bound to phytate (Hallberg et al., 1989; Sandstro¨m and Sandberg, 1992; Bohn et al., 2004). Humans and other monogastric animals lack enzymes in the gastrointestinal tract to degrade phytate, and a diet rich in phytate leads to a considerably impaired absorption of dietary minerals (Hallberg et al., 1989; Sandstro¨m and Sandberg, 1992; Bohn et al., 2004). Therefore, phytate is regarded as a nutritional inhibitor. During the fermenta- tion process, R. oligosporus produces phytase that degrades phytate, and consequently increases the availability of minerals in barley (Eklund-Jonsson et al., 2006). R. oligosporus also produces ergosterol (provitamin D 2 ) and some vitamins (Wiesel et al., 1997). Yeasts such as Saccharomyces cerevisiae have been used as a single-cell protein source for animals and humans (Daghir and Sell, 1982; Litchfield, 1983). They can produce savoury flavours such as glutamic acid, peptides and ARTICLE IN PRESS www.elsevier.com/locate/fm 0740-0020/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.fm.2006.06.007 Corresponding author. Tel.: +46 18 673212; fax: +46 18 673392. E-mail address: [email protected] (X.M. Feng).
10
Embed
Rhizopus oligosporus and yeast co-cultivation during barley tempeh fermentation—Nutritional impact and real-time PCR quantification of fungal growth dynamics
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ARTICLE IN PRESS
FOODMICROBIOLOGY
0740-0020/$ - se
doi:10.1016/j.fm
�CorrespondE-mail addr
Food Microbiology 24 (2007) 393–402
www.elsevier.com/locate/fm
Rhizopus oligosporus and yeast co-cultivation during barley tempehfermentation—Nutritional impact and real-time PCR quantification
of fungal growth dynamics
Xin Mei Fenga,�, Volkmar Passotha, Charlotte Eklund-Jonssonb,Marie Larsson Almingerb, Johan Schnurera
aDepartment of Microbiology, Swedish University of Agricultural Sciences (SLU), Box 7025, SE-750 07 Uppsala, SwedenbDepartment of Chemical and Biological Engineering, Food Science, Chalmers University of Technology, SE-412 96 Goteborg, Sweden
Received 3 March 2006; received in revised form 2 June 2006; accepted 23 June 2006
Available online 30 August 2006
Abstract
Barley tempeh was produced by fermenting barley kernels with Rhizopus oligosporus. The potential of the yeasts Saccharomyces
cerevisiae (three strains), S. boulardii (one strain), Pichia anomala (one strain) and Kluyveromyces lactis (one strain) to grow together with
R. oligosporus during barley tempeh fermentation was evaluated. All yeast strains grew during the fermentation and even during cold
storage of tempeh (Po0:01). The growth of yeasts slightly increased the ergosterol contents, but did not influence amino acid contents
and compositions, and did not reduce phytate contents. Slight increases of vitamins B6 and niacinamide, and slight decreases of B1 and
biotin were observed. Quantification of fungal growth is difficult during mixed species fermentations because ergosterol is found in all
fungal species, and colony-forming-unit (cfu) estimations are not reliable for R. oligosporus and other sporulating fungi. Therefore, we
developed a quantitative real-time PCR method for individually quantifying S. cerevisiae and R. oligosporus growth in barley tempeh.
The PCR results were highly correlated with the ergosterol content of R. oligosporus and with the number of cfu of S. cerevisiae. Thus,
real-time PCR is a rapid and selective method to quantify yeasts and R. oligosporus during mixed species fermentation of inhomogenous
Tempeh is a traditional Indonesian food prepared byfermentation of dehulled and cooked soybeans withmoulds of the genus Rhizopus (mainly Rhizopus oligos-
porus) to a compact and sliceable cake (Shama and Hall,1991). Tempeh can also be produced from cereal grains,e.g. barley, through fermentation with R. oligosporus (Berget al., 2001; Feng et al., 2005). Phytate (myo-inositolhexakisphosphate or IP6) is the principal storage form ofphosphorus in cereal grains and other plant materials(Reddy et al., 1982). Minerals, such as iron, magnesiumand zinc, are strongly bound to phytate (Hallberg et al.,
e front matter r 2006 Elsevier Ltd. All rights reserved.
.2006.06.007
ing author. Tel.: +4618 673212; fax: +46 18 673392.
1989; Sandstrom and Sandberg, 1992; Bohn et al., 2004).Humans and other monogastric animals lack enzymes inthe gastrointestinal tract to degrade phytate, and a diet richin phytate leads to a considerably impaired absorption ofdietary minerals (Hallberg et al., 1989; Sandstrom andSandberg, 1992; Bohn et al., 2004). Therefore, phytateis regarded as a nutritional inhibitor. During the fermenta-tion process, R. oligosporus produces phytase that degradesphytate, and consequently increases the availability ofminerals in barley (Eklund-Jonsson et al., 2006). R.
oligosporus also produces ergosterol (provitamin D2) andsome vitamins (Wiesel et al., 1997).Yeasts such as Saccharomyces cerevisiae have been used
as a single-cell protein source for animals and humans(Daghir and Sell, 1982; Litchfield, 1983). They can producesavoury flavours such as glutamic acid, peptides and
ARTICLE IN PRESSX.M. Feng et al. / Food Microbiology 24 (2007) 393–402394
ribonucleotides from the degradation of cell protein andRNA (Walker, 1999). They can also produce phytase todegrade phytate (Andlid et al., 2004; Passoth et al., 2006),ergosterol (Tan et al., 2003) and different vitamins(Diplock et al., 1961; Fleet, 1990). Yeasts can inhibit otherfungi (Druvefors et al., 2002; Passoth and Schnurer, 2003),and recently, it has been found that yeasts can also inhibitpathogenic bacteria (Goerges et al., 2006; Leverentz et al.,2006).
Yeasts can grow spontaneously in soybean tempeh(Steinkraus et al., 1983; Samson et al., 1987). However,their role in the tempeh fermentation is not well under-stood. Many yeast species, e.g. Trichosporon beigelii,Clavispora (Candida) lusitaniae, Candida maltosa, Candida
curvata, Hansenula anomola) are found in soybean tempeh(Samson et al., 1987). Among them, T. beigelii was themost frequently detected yeast species. Several reports havesuggested that T. beigelii can cause mycosis in humans andanimals (Gonzalez et al., 2001; Khandpur and Reddy,2002; Youker et al., 2003). This yeast is widely spread insoil and water (Han et al., 2000), and might be only anincident contaminant in tempeh. Introducing known food-grade yeasts to tempeh might improve the nutritionalquality and simultaneously reduce the risk of unwantedyeasts, moulds and bacteria. However, the yeasts intro-duced to tempeh must not restrict the growth ofR. oligosporus.
Previously, we have quantified R. oligosporus growth bymeasuring hyphal lengths and ergosterol content whenlactic acid bacteria were introduced into barley tempeh(Feng et al., 2005). Direct microscopy determinations ofhyphal lengths allows differential quantification of mycelialand yeast biomass during co-cultivation experiments(Bjornberg and Schnurer, 1993). However, measuring ofhyphal length is a time-consuming method with a largeexperimental error (Feng et al., 2005). Ergosterol, provi-tamin D2, is produced by both yeasts and moulds (Pasanenet al., 1999), and thus cannot be used to separatelydetermine the growth of yeasts and R. oligosporus in aco-cultivation system. Traditionally fungi are quantified ascolony-forming-units (cfu) on selective substrates. How-ever, for spore forming filamentous fungi such as R.
oligosporus, this method is not ideal (Schnurer, 1993).Quantitative real-time PCR, using specific primers, mightovercome the problems of the conventional methods, andhas been used for rapid quantification of yeasts and mouldscontaminating yogurt and pasteurized food products(Bleve et al., 2003).
In this study, we investigated whether yeasts can grow inbarley tempeh, and estimated the effects of yeast growth onthe nutritional characters of barley tempeh, such ascontents of amino acids, ergosterol and vitamins, as wellas on phytate reduction. To determine the effect of co-
cultivation with yeasts on R. oligosporus growth duringbarley tempeh fermentation, we developed a selective real-time PCR quantification method based on yeast andR. oligosporus total DNA directly extracted from barleytempeh.
2. Materials and methods
2.1. Microbial strains
R. oligosporus ATCC 64063 (Strain number J401,maintained in the culture collection of the Department ofMicrobiology, Swedish University of Agricultural Sciences,Uppsala, Sweden) is an efficient strain for barley tempehfermentation (Berg et al., 2001). Fungal spores weremaintained on sterile silica grains in Eppendorf tubes at2 1C. Six yeast strains from three species were used in thisstudy (Table 1). S. cerevisiae (S. boulardii) and Kluyver-
omyces lactis are regarded as food grade, or even probioticsspecies (Breunig and Steensma, 2003; Galvao et al., 2005),and may be especially appropriate for tempeh production.Pichia anomala J121 was selected as it has strong growthability and inhibitory effects on growth of several mouldson cereal grains (Druvefors et al., 2002).
2.2. Preparation of barley tempeh
To produce inocula for tempeh fermentation, R.
oligosporus spores were prepared by inoculating three tofive grains of spore-containing silica onto a 100ml MEAagar (Oxoid) slant in a 200ml bottle with a looselytightened lid (for air access) and incubating at 30 1C. After4–5 days, spores were harvested and washed three timeswith NaCl solution (9 g/L) to remove the residual nutrientsof the medium. The spore concentration was assessed witha haemocytometer and the spore suspension was stored at2 1C (Feng et al., 2005). Yeast strains were grown in 5ml ofYPD broth (10 g yeast extract, 20 g peptone,20 g glucose L�1) for 24 h. Hundred microlitres of the pre-culture was inoculated into a new 5ml YPD broth andincubated for 24 h again. Yeast cells were washed twicewith NaCl solution, and re-suspended in 1ml of NaClsolution. The concentration of this cell suspension wasapprox 108 cells/ml as estimated with a haemocytometer.Commercially available pearled whole barley kernels
(Gourmet Korn, a mixture of different cultivars) wereobtained from Kungsornen AB, Jarna, Sweden. Barley(300 g) was soaked in 500ml tap water containing 0.12Mlactic acid (VWR international AB, Stockholm) for 6 h atroom temperature. After draining, the barley was boiledfor 10min in 750ml tap water supplemented with 6 g NaCl.The drained and cooled (42 1C) barley was inoculated withR. oligosporus at approx 104 spores/g wet weight and yeastsat approx 104 cfu/g. Barley inoculated with R. oligosporus
at approx 104 spores/g alone was used as a control. Theinoculated barley (60 g wet weight) was packed intodisposable, sterile petridishes (9� 1.5 cm), which were
Species Culture no.a Strain nos.b Origin References (selected)
Saccharomyces cerevisiae J122 ATCC 9763; CBS 2978;
CBS 5900; DSM 1333;
NRRL Y-567
Distiller’s yeast, wild-type
strain
(Birol et al., 1998)
Saccharomyces cerevisiae J543 Baker’s yeast, Svenska
Jastfabrik AB, Sollentuna,
Sweden. Isolated 2004
Saccharomyces cerevisiae J191 Baker’s yeast, Svenska
Jastfabrik AB, Sollentuna,
Sweden
(Petersson and Schnurer,
1995)
Saccharomyces boulardii
(subspecies of S. cerevisiae)
J551 Precosa, AstraZeneca AB (Kuhle and Jespersen, 2003;
Kotowska et al., 2005)
Pichia anomala J121 CBS 100487 From grains in airtight
storage.
(Bjornberg and Schnurer,
1993)
Kluyveromyces lactis J469 ATCC8585; CBS2359;
DSM70799; NRRLY-
1140
Isolated from creamery, USA (Steensma et al., 1988; Picos
et al., 1993)
aCulture no. in the Department of Microbiology, Swedish University of Agricultural Sciences, Uppsala, Sweden.bATCC, American Type Culture Collection, Manassas, VA, USA; CBS, Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands; DSM,
Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany; NRRL, ARS Culture Collection, Northern Regional
Research Laboratory, National Center for Agricultural Utilization Research, Peoria, IL, USA.
individually wrapped in plastic bags and incubated at 35 1Cin a ventilated growth cabinet (Binder, Tillquist instrumentAB, Stockholm) (Feng et al., 2005).
2.3. Analysis of yeast growth during tempeh fermentation
and storage
Twenty-gram-portions of barley tempeh were asepticallyremoved from each petridish, mixed with 180ml of sterilepeptone water (2 g/L in distilled water), homogenized in astomacher (Stomacher 400, Seward, Sweden) at normalspeed for 2min, and 10-fold diluted in sterile peptonewater. Then 0.1ml of adequate dilution was surface-spreadon DG18 agar (Oxoid) plates containing chloramphenicol(Sigma) (1 g/L) to prevent bacterial growth. The agar plateswere incubated at 25 1C, and the cfu were counted after 3–4days. Meanwhile, the pH values of the tempeh cakes weremeasured in the first dilution.
2.4. Ergosterol analysis
Freeze-dried and milled samples (0.4–0.5 g) were ex-tracted in 10ml glass tubes with screw caps. Five-acholesterol (50 ml of 2mg/ml) was used as an internalstandard. Each sample was mixed with 5ml of ethanol/methanol (60/40), five potassium hydroxide pellets (ap-proximately 0.5 g) and 0.2ml of pyrogallol solution (10 g/Lin ethanol). Saponification and extraction were made at80 1C for 90min. The samples were shaken vigorouslyevery 10min. After cooling 4ml toluene was mixed with thesample. After addition of 2.5ml distilled water the sampleswere mixed and centrifuged. The supernatants weretransferred to new tubes, air-dried and dissolved in 0.5mltoluene. Two hundred microlitres of the solution were
transferred to GC vials. The samples were analysedquantitatively with GC HP-5890 coupled with an HP-7673 Auto sampler (Hewlett Packard, Waldbronn,Germany). Two microlitres of the samples was auto-injected. The injector temperature was 280 1C and thecolumn (HP Ultra-1, 50m�0.32mm df 0.52 mm) wasprogrammed from 260 to 320 1C with gradual increase at10 1C/min, keeping at the final temperature for 15min. Thedetector temperature was 320 1C.The results were evaluated with Borwin Chromatogra-
phy software (Le Fontanil, France).
2.5. Analysis of phytate (myo-inositol hexakisphosphate)
Freeze-dried and milled samples (0.5 g) were extracted bystirring with 0.5M HCl (10ml) for 3 h and then centri-fuged. The supernatant was frozen at �20 1C overnight,thawed and centrifuged through an ultracentrifugal filterdevice (Micron YM-30, Millipore, Bedford MA, USA).Phytate (myo-inositol hexakisphosphate or IP6) was
quantified by high performance ion chromatography(HPIC) (Carlsson et al., 2001). Samples were run induplicates and phytate content was determined asmmol phytate g�1 dry matter.
2.6. Analysis of amino acid and vitamins
Amino acids and vitamins (vitamins B1, B6, Niacina-mide and biotin) were analysed by AnalyCen Nordic AB,Lidkoping, Sweden. Amino acids were analysed accordingto standard SS-EN ISO 13903:2005, except tryptophanethat was analysed according to EU standard (Eu Dir 2000/45/EG part C). Vitamin B1 was analysed according to vanden Berg et al. (1996), B6 and niacin amide were estimated
ARTICLE IN PRESSX.M. Feng et al. / Food Microbiology 24 (2007) 393–402396
by microbiological assay (Freed, 1966; Hashim and Fields,1979), and biotin was analysed according to Indyk et al.(2000) and Quant Biotin Kit User’s guide (EditionDecember 2000, BiaCore, Uppsala, Sweden).
2.7. Extraction of R. oligosporus and S. cerevisiae total
DNA from barley tempeh
R. oligosporus and S. cerevisiae J543 were inoculated atapprox 104 spores (cfu)/g wet barley in pure- or co-cultureexperiments. Triplicate samples were taken after 0, 9, 12, 17and 20 h incubation at 35 1C. To extract total DNA fromthe barley tempeh samples, 1 g of frozen barley tempeh(wet weight) from each sample was added with 2ml of lysisbuffer (50mM Tris-Cl pH 8.0; 50mM EDTA; 3% SDS;2% of b-mercaptoethanol), 2 ng of EcoR1-linearizedpUC19 plasmid with ampR gene (as an external standardfor the subsequent real time PCR) and 2 g acid-washedglass beads (0.45–0.5mm in diameter, Sigma). The sampleswere vortexed for 5� 1min at maximum speed (Vortex-genie 2), and kept for 1min on ice in between. Subse-quently, they were incubated in a water bath at 65 1C for10min with gentle mixing after 5min. One volume (2ml) ofphenol/chloroform (1/1) was added and the solution wasmixed by inverting the tube for few times. The mixture wascentrifuged and 1ml of the upper aqueous layer wastransferred to a new Eppendorf tube. The phenol/chloro-form treatment was repeated until no white precipitateoccurred in the upper aqueous layer. Subsequently, theDNA was precipitated by adding 0.1 volume of 3M NaAc(pH 5.2) and 1 volume of ice-cold 96% EtOH. The pelletwas washed with 100 ml of 70% EtOH, air-dried anddissolved in 200 ml of TE buffer (10mM Tris-Cl, pH 8.0and 1mM EDTA).
2.8. Real-time PCR
The selected target genes for specific detection were theR. oligosporus chs1 gene (encoding for a chitin synthase,EMBL accession number D10159), the S. cerevisiae PDA1
gene (encoding for the E1-alpha subunit of the pyruvatedehydrogenase complex, EMBL accession numberX71664), and the ampR gene of the pUC19-plasmid(encoding a beta-lactamase, EMBL-accession numberM77789). Primers for real time PCR were designed using
Table 2
Characteristics of primers for real-time PCR quantification of DNA
Gene Primer pairs Concentra
Roligchreal 1F 50-GTCCCCTTCTTAAGCCCAATG-30 160
Roligchreal 1 R 50-GCGACCACCAGGCTTTGTAC-30 160
ScPDArealF1 50-GAAAGCCGTTCTGGCTGAAT-30 160
ScPDArealR1 50-GCATGGAACCACCCTTACCA-30 160
AmprealtimeF1 50-CCGGCAACAATTAATAGACTGGAT-30 40
AmprealtimeR1 50-GGGCCGAGCGCAGAA-30 40
aThe primer melting temperature (Tm) was calculated by the Primer Expres
the Primer Expresss Software v2.0 (Applied Biosystems).The sequences, amplicon length and Tm of the specificprimers are given in Table 2.The primers were tested at different concentrations for
non-specific DNA-amplification (dimer-formation) by run-ning real time PCR without any templates. For thesubsequent analysis, primers were used at the maximumconcentrations where no non-specific amplifications oc-curred (Table 2). The primers were also tested withtemplate DNA from other microorganisms used in thisstudy, and no non-specific amplification was observed.The amplification efficiencies (E) of the respective DNA
amplicons were empirically determined from the amplifica-tion of serial dilutions of the respective target DNAaccording to Wilhelm and Pingoud (2003):
E ¼ 10½�1=slope�,
where the slope was calculated from the standard curve ofDNA concentrations vs. the CT values. The CT value isdefined as the cycle number at which the fluorescencesurpasses the noise threshold.To produce target DNA for the amplification efficiency
determination, parts of the chs1 and PDA1-gene sequenceswere amplified from respective genomic DNA using theprimer pairs Roligch1F1410 (50-gttttagctgccatgggtgt-30)and Roligch1R1952 (50-gaccaaagacggactccaaa-30), andPDA306up (50-atcacmtcttayagatg-30) and PDA849d (50-gacatrgartgrccacc-30), respectively. For ampR DNA,EcoR1-linearized pUC19-plasmid was used.To quantify S. cerevisiae and R. oligosporus DNA
extracted from tempeh, real-time PCR was performed withthe ABI Prisms 7000 Sequence Detection System fromApplied Biosystems using SYBR green I as detective dye.The reaction mixture contained 1 ml of sample DNA, theappropriate primer concentrations (Table 2), 12.5 ml of2� SYBRs Green Master Mix (see user’s manual) andsterile ultrafiltered dH2O to a final volume of 25 ml. Eachcombination of sample DNA and primer pairs wasanalysed in triplicates. The PCR was carried out by usingthe default reaction parameters.The log of the initial concentration of any amplified
DNA molecule is proportional to E�CT, where E is theamplification efficiency, and CT is the cycle number forsurpassing the noise threshold (see above). Thus, the real-time PCR data are expressed as relative DNA copy
ARTICLE IN PRESSX.M. Feng et al. / Food Microbiology 24 (2007) 393–402 397
numbers (R) of the sample gene copy number related to theexternal standard gene (ampR) copy number according tothe equation:
R ¼E
CTðStandardÞStandard
ECTðSampleÞSample
.
2.9. Statistical analysis
Differences in treatments were analysed using the One orTwo Way Repeated Measure ANOVA procedures inSigmaPlot 9.0 (Systat Software Inc., Richmond, USA).The Tukey test was used for post hoc pairwise comparisonsof levels.
3. Results
3.1. Growth of yeasts during co-cultivation with
R. oligosporus in barley tempeh
R. oligosporus was inoculated together with each yeaststrain (Table 1). The number of yeasts was determined ascfu immediately after inoculation, after 20 h incubation at35 1C, and after storage of the 20 h-fermented tempeh cakesfor 7 weeks at 2 1C. All yeasts inoculated at approx 4.0 logunits increased significantly (Po0:01) during the barleytempeh fermentation and even during tempeh storage(Table 3). During 20 h fermentation, the cfu of P. anomala
J121 increased by 3.4 log units, while the other five strainsincreased between 2.6 and 2.9 log units. During storage oftempeh at 2 1C, the cfu numbers of K. lactis J469 andP. anomala J121 increased by 2.2 and 1.3 log units,respectively, while the four S. cerevisiae/S. boulardii strainsincreased between 0.6 and 0.8 log units. The pH increasedslightly from 4.6 immediately after inoculation to 4.8 afterfermentation in all treatments, including the control. Thisimplies that the growth of yeasts does not markedlyinfluence the pH values. The growth of yeasts at thisinoculation levels did not visibly affect the growth of R.
oligosporus. All tempeh cakes, whether co-cultivated with
Table 3
Growth of yeasts during barley tempeh fermentation and during tempeh
cold storage
Strains Fermentation Storage at 2 1C
0ha 20 ha 7 weeksb
S. cerevisiae J 122 4.070.4 6.970.1 7.770.0
S. cerevisiae J 543 4.070.4 6.970.1 7.570.0
S cerevisiae J 191 4.170.4 6.970.1 7.670.0
S. boulardii J 551 4.170.4 6.970.1 7.770.4
P. anomala J 121 4.070.4 7.470.2 8.770.0
K. lactis J 469 4.270.5 6.870.2 9.070.0
Data are given as log cfu/g dry tempeh (mean7SD, an ¼ 4, and
mean7deviation from mean values, bn ¼ 2).
yeasts or not, had firm texture and were sliceable, exhibitedsweet and fruity odour, and had a creamy white colour.
3.2. Ergosterol and vitamin production
Ergosterol is a component of the fungal cell membrane,both in filamentous fungi and in yeasts. Thus, increasedergosterol content was expected for R. oligosporus-yeast co-fermentations. Tempeh fermented with R. oligosporus hada significant (Po0:001) increase in the ergosterol contentscompared to non-fermented barley (Table 4). Co-cultiva-tion with different yeasts further increased the ergosterolcontent by 7.1–17.5 mg/g dry tempeh (Table 4). However,the statistical significance of these differences could not beproven. Calculating from the ergosterol contents and thecfu numbers estimated in barley inoculated with S.
cerevisiae J543 alone, the ergosterol content of individualcells of S. cerevisiae J543 was 3.4–6.6� 10�7 mg/cfu. Thus,the expected increase in the ergosterol content in tempehco-fermented with this yeast strain was in the range of2.7–5.2 mg/g barley, which is close to the experimentallydetermined value (9.9 mg/g) (Table 4). This, together withthe fact that a similar increase was observed for all yeastco-fermentations, indicates that the detected increases ofergosterol content in barley are due to the growth of yeasts.The concentrations of some vitamins were also surveyed
for barley tempeh fermented with R. oligosporus and S.
cerevisiae J191 for 20 h. S. cerevisiae co-cultivationincreased vitamins B6 and niacinamide from 0.7 to 0.9and 53.0 to 69.1mg/kg dry tempeh, respectively. Incontrast, vitamin B1 decreased from 0.5 to 0.2mg/kg drytempeh and biotin from 117.5 to 94.5 mg/kg dry tempeh.
3.3. Phytate content of tempeh made by co-fermentation of
R. oligosporus and yeasts
One of the aims of barley tempeh fermentation is toreduce the content of phytate. Yeast can produce phytase,and it was anticipated that the growth of R. oligosporus
together with yeasts would reduce phytate further. Barleytempeh fermented with R. oligosporus alone significantly
Table 4
Ergosterol and phytate contents of barley and barley tempeh fermented
with R. oligosporus J401 alone or co-inoculated with S. cerevisiae J122,
J543 or J191, S. boulardii J551, P. anomala J121 or K. lactis J469
Mean7SD, an ¼ 6, and mean7deviation from mean value, bn ¼ 2.
ARTICLE IN PRESSX.M. Feng et al. / Food Microbiology 24 (2007) 393–402398
(Po0:001) reduced the phytate contents by 28.3%compared to non-fermented barley (Table 4). However,yeast co-fermentations with R. oligosporus did not sig-nificantly reduce the phytate contents further compared toR. oligosporus fermentation alone (Table 4).
3.4. Total amino acids content in barley after
co-fermentation of R. oligosporus and yeasts
Amino acid contents were determined for tempeh cakesfermented by R. oligosporus alone, or co-cultivated witheither S. cerevisiae J122, P. anomala J121 or K. lactis
J469 (Table 5). R. oligosporus fermentation alone orco-fermentation with J122, J121 and J469 did notconsiderably change the total and essential amino acidscontents compared to non-fermented barley (0 h, J401)(Table 5).
3.5. Development of a rapid DNA extraction method for
yeast and R. oligosporus on barley tempeh
Extraction of microbial DNA from an inhomogenoussystem like tempeh is difficult due to the big particles ofbarley kernels. Moreover, ground barley kernels willrelease a number of different compounds that may affectPCR amplification (Rossen et al., 1992). Considering that
Table 5
Amino acid composition of barley and barley tempeh fermented with R. oligo
K. lactis J469
Item 0h 20h
J401a J401a
Cystine 2.870.0 2.870.0
Methionine� 1.870.0 1.870.0
Aspartic acid 5.770.2 5.770.2
Threonine� 3.670.1 3.570.0
Serine 5.070.2 4.770.1
Glutamic acid 25.670.4 22.672.6
Proline 14.170.0 13.370.8
Glycine 3.970.0 3.870.1
Alanine 3.970.0 4.370.2
Valine� 7.070.1 6.370.6
Isoleucine� 4.670.0 4.570.1
Leucine� 8.170.0 7.770.1
Tyrosine (calculated) 3.970.0 3.870.1
Phenylalanine� 6.670.1 6.470.2
Histidine� 2.570.0 2.470.1
Ornitine o0.2 o0.2
Lysine� 3.270.0 3.570.0
Arginine 5.170.0 4.870.0
Hydroxiproline o0.2 o0.2
Tryptophane 1.570.0 1.470.0
Ammonia 1.870.0 1.870.0
Essential amino acids 37.470.1 36.071.3
Total amino acids 108.870.9 103.075.5
Data are given as g amino acid(s)/kg dry matter. Values represent mean valu
measuring tolerance given by the analysing company is 8% for each amino ac�Essential amino acid.
most of the mycelia (Varzakas, 1998) and the yeast cells arelikely in the outside of the kernels, we avoided to usemethods including grinding of barley kernels. Instead, wetried an alkalic lysis-boiling (Olsson et al., 1999) and alysing buffer-heating method (Bilkova and Kralova, 1999).The alkalic lysis-boiling method is efficient to extract DNAfrom spores and hyphae of filamentous fungi (Olsson et al.,1999), but in our hands it did not work for yeasts, asevaluated by PCR amplification using specific primers. Incontrast, the lysing buffer-heating method was efficientin extracting DNA from yeast, but not from R. oligosporus
in barley tempeh. Using the lysis buffer method as startingpoint, we added a mechanical cell disruption step byvortexing the barley samples with glass beads. Thismodification resulted in a successful isolation of DNAfrom both R. oligosporus and S. cerevisiae. Some para-meters were further optimized. Incubating time at 65 1Cwas reduced from 25 to 10min without loosing extractionefficiency. Vortexing barley samples for 5min after addingof the glass beads gave maximal extraction of DNA fromyeast cells, while 2–10min of vortexing did not influencethe DNA extraction efficiency of R. oligosporus. This wasverified by estimating the relative DNA copy numbers withreal-time PCR. Therefore, vortexing for 5min, followed byincubation for 10min at 65 1C was used for the DNAextraction.
sporus alone or co-inoculated with S. cerevisiae J122, P. anomala J121 or
J401+J122b J401+J121b J401+J469b
2.8 2.8 2.8
1.8 1.8 1.8
5.5 6.0 6.0
3.5 3.7 3.5
4.8 4.8 4.6
26.7 26.7 23.3
14.7 14.1 13.8
3.9 3.9 3.9
4.4 4.6 4.4
6.9 6.7 6.5
4.6 4.6 4.6
8.1 8.1 7.8
4.1 3.9 3.9
6.5 6.7 6.2
2.3 2.5 2.3
o0.2 o0.2 o0.2
3.5 3.7 3.5
4.8 5.1 4.8
o0.2 o0.2 o0.2
1.5 1.4 1.4
1.8 1.8 1.8
37.2 37.8 36.2
110.4 111.1 105.1
es7deviation from mean (n ¼ 2)a or single determinations (n ¼ 1)b. The
3.6. Real-time PCR quantification of R. oligosporus and
S. cerevisiae and correlation to ergosterol contents and cfu
Different primer pairs were designed (Table 2) and theamplification efficiencies were tested. The theoreticalmaximum amplification efficiency (E value) of real-timePCR of one amplification cycle is 2 (Wilhelm and Pingoud,2003). The measured E value was 1.95 for the R.
oligosporus chs1 gene, 1.90 for the S. cerevisiae PDA1 geneand 1.93 for the ampR gene.
The growth of R. oligosporus alone, or together with S.
cerevisiae J543, in barley tempeh was analysed by real-timePCR. For a comparison, ergosterol contents were analysedon all barley tempeh samples fermented by R. oligosporus
alone, and on some samples fermented by R. oligosporus
together with S. cerevisiae. For R. oligosporus grown alone,both real-time PCR values and ergosterol contentsincreased with time (Figs. 1 and 2). However, the ergosterolcontent showed a lag phase in the beginning of thefermentation (Fig. 1), while this lag phase was not detectedwith the real-time PCR method (Fig. 2). Therefore therewas a relatively low correlation (r ¼ 0:83, n ¼ 5, Po0:1)between the ergosterol contents and the real-time PCRdata. Removing the lag phase data values, a high degree ofcorrelation (r ¼ 0:99, n ¼ 4, Po0:01) was found. For R.
oligosporus co-inoculation with yeast, the R. oligosporus
growth was apparently decreased due to high inoculationlevel of S. cerevisiae (more than 4.5 log cfu/g). Theergosterol content in the co-cultivation tempeh (51.7 mg/gdry tempeh) was also lower than that in the pure culturefermentation of R. oligosporus (63.2 mg/g). However, nosignificant differences were observed between the real-timePCR values for R. oligosporus grown alone and growntogether with S. cerevisiae (Fig. 2).
Obviously, the correlation between ergosterol produc-tion and genome copy number in R. oligosporus waschanged during co-cultivation with high number of yeastcells. Because high nucleus copy numbers were found to beconnected with high branching frequency in Aspergillus
nidulans (Lin and Momany, 2004), we assumed a similar
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
-1 4 9 14 19
Log
erg
oste
rol c
onte
nt
Time (h)
Fig. 1. Ergosterol content of barley tempeh fermented by Rhizopus
oligosporus during 20 h at 35 1C (n ¼ 3).
phenomenon in R. oligosporus. To investigate whetherS. cerevisiae can increase branches on the hyphae of R.
oligosporus, we cultivated R. oligosporus and S. cerevisiae
together on malt extract agar (MEA) plates. On the controlplates without S. cerevisiae, R. oligosporus covered thewhole plates and sporulated within 24 h. In contrast, in themixed culture plates, R. oligosporus grew slower and didnot sporulate in the same time period. This slower growthwas accompanied by a high branching frequency of themycelia of R. oligosporus (Fig. 3). Therefore, the high DNAcopy numbers found in R. oligosporus in co-cultivationwith yeast may indicate an increased branching frequencyinduced by the yeast. In this situation, real-time PCRresults might not accurately correlate to the biomassproduction of the mould.The growth of S. cerevisiae cultivated alone, or grown
together with R. oligosporus, was determined with real-timePCR and as cfu. Both techniques gave values that increasedwith time (Figs. 4 and 5). The real-time PCR values werehighly correlated (r ¼ 0:98, n ¼ 9, Po0:001) with the yeastcfu numbers. However, in the real-time PCR analysis of thelate fermentation stage, there was a small but significant(Po0:05) difference between cultivation with and withoutR. oligosporus (Fig. 5), which was not observed in cfudetermination (Fig. 4). This might indicate that real-timePCR quantification was more precise than cfu determina-tion. Our results show that R. oligosporus has no or onlyslight influence on the growth of S. cerevisiae (Figs. 4and 5).
4. Discussion
Our results demonstrated that different yeasts couldgrow together with R. oligosporus during barley tempehfermentation and increase the ergosterol content ofresulting barley tempeh cakes by 12–31%. Nevertheless,the nutritional impact of yeast co-cultivation was lowerthan expected. However, an improvement in nutritionalvalues may be obtained by selecting specific strains with
Fig. 3. Effect of S. cerevisiae J543 on R. oligosporus hyphal morphology (400� ). (A) Control (100 ml of 104 spores/ml of R. oligosporus was spread on
MEA agar plates); (B) R. oligosporus cultivated together with S. cerevisiae J543 (100ml of mixed suspension including 104 spores/ml of R. oligosporus and
105 cells/ml of S. cerevisiae J543 was spread on MEA agar plates).
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
-1 4 9 14 19
Time (h)
Log
cfu
of
S. c
erev
isia
e
Fig. 4. Colony forming units (cfu) of S. cerevisiae J543 grown alone (&)
or together with R. oligosporus (~) in barley during 20 h fermentation at
35 1C (n ¼ 3).
-5.5
-4.5
-3.5
-2.5
-1.5
-1 4 9 14 19
Time (h)
Log
rel
ativ
e D
NA
cop
ies
of S
. cer
evis
iae
Fig. 5. Real-time PCR quantification of S. cerevisiae J543 grown alone
(&) or together with R. oligosporus (E) in barley during 20 h
fermentation at 35 1C (n ¼ 3).
X.M. Feng et al. / Food Microbiology 24 (2007) 393–402400
specific performances (for instance with high folateproduction).
Inoculation of Pichia anomala at 104 cfu/g inhibitsseveral mycotoxigenic fungal species such as Penicillium
roqueforti (Druvefors et al., 2002), but at this inoculationlevel no inhibition on R. oligosporus growth was observed.We also found that the other tested yeasts did not affect R.
oligosporus growth at this inoculation level. However,yeasts added at levels higher than 104 cfu/g or higher thanthe R. oligosporus inoculation level inhibited the growth ofR. oligosporus. Yeasts, especially non-food grade yeasts,such as T. beigelii, that frequently spontaneously grow insoybean tempeh fermentations (Samson et al., 1987), maybe able to disturb tempeh production and even constitute ahazard for human health. Therefore, it is desirable toprevent the spontaneous growth of yeasts in tempehfermentation. The optimal yeast strains for co-cultivationshould be food-grade and have low inhibitory potentialtowards R. oligosporus, but should also have a highcompetitive ability against other yeasts, moulds or bacteria.Yeast co-cultivation did not further reduce the phytate
content. The maximum phytate reduction obtained in thepresent study was only about one third of the total contentin raw material. Similar results were found by Sutardi andBuckle (1985) who studied phytate reduction in soybeansfermented with different strains of R. oligosporus. How-ever, in a recent study, modifications of the grain surfacehave been found to greatly improve phytate reduction, and97% of the original phytate of whole barley kernels wasdegraded during tempeh fermentation (Eklund-Jonsson etal., 2006). Thus, the non-degraded phytate in our studymay be inaccessible for microbial degradation.The real-time PCR results of S. cerevisiae were highly
(r ¼ 0:98, n ¼ 9, Po0:001) correlated with the cfu numbersof S. cerevisiae. Quantification of yeasts by real-time PCRis much faster than by determination of cfu, and especiallyuseful in mixed culture systems. In contrast, quantificationof filamentous fungi is a challenging task because thefungal colony consists of mycelia and single-celled spores.This makes it difficult to define the filamentous fungalgrowth by biomass, hyphal length, cell numbers or nucleuscopy numbers. Several methods have been established, butevery method has its drawbacks. For instance, the numberof cfu correlates rather with the number of spores thanwith mycelial growth (Schnurer, 1993). Hyphal length
ARTICLE IN PRESSX.M. Feng et al. / Food Microbiology 24 (2007) 393–402 401
determination is time-consuming and prone to substantialexperimental error (Feng et al., 2005), while the ergosterolcontent of the fungal membrane can vary with fungal age,and can be affected by oxygen supply and light (Nout et al.,1987). Furthermore, with the exception of cfu determina-tion with further subculturing none of these methods arespecies-specific. Here, we have described a selective real-time PCR method for determining relative nucleus copynumbers of filamentous fungi in the solid-state tempehfermentation. The real-time PCR results were in generalcorrelated with ergosterol contents of the tempeh. How-ever, the lag phase observed by ergosterol determinationbut not by real-time PCR determination suggests thatDNA replication may occur prior to hyphal growth duringthe early stages of the fungal growth. Lin and Momany(2004) demonstrated that a high proportion of nucleuscopy numbers correlates with high hyphal branches inabnormal hyphal branch mutants of A. nidulans. Ourresults indicate that high nucleus copy numbers in R.
oligosporus may correlate with increased formation ofhyphal branches that were induced by co-cultivation withgrowth inhibitory microorganisms. This suggests that inthis special situation real-time PCR quantification offilamentous fungi should be combined with other quanti-fication methods.
5. Conclusion
A number of yeasts could grow together with R.
oligosporus during barley tempeh fermentation. Theinoculation level of yeasts at 4 log value did not signifi-cantly affect the growth of R. oligosporus, which indicatesthat yeasts can be introduced to barley tempeh to increaseits safety and nutritional values. Real-time PCR determi-nation was demonstrated to be a rapid, reliable andselective method for quantification of fungi in barleytempeh fermentation although it in some cases requiresverification by other methods. It may also be possible toapply in other solid-state food and feed fermentation bychoosing appropriate primer combinations.
Acknowledgements
This project was financed by VINNOVA (SwedishAgency for Innovation Systems). The authors would alsolike to thank M. Drabkova for helpful advices, N.G.Carlsson for help with the analyses of ergosterol andphytate, and Dr. M. Pell for statistical analysis.
References
Andlid, T.A., Veide, J., Sandberg, A.-S., 2004. Metabolism of extracellular
inositol hexaphosphate (phytate) by Saccharomyces cerevisiae. Int.
J. Food Microbiol. 97 (2), 157–169.
Berg, S., Olsson, J., Swanberg, M., Schnurer, J., Eriksson, A., 2001.
Method for the production of fermented cereal food products and
products thereof. World Intellectual Property Organization, Interna-
tional patent application number: PCT/SE02/00357.
Bilkova, K., Kralova, B., 1999. Izolace biomakromolekul, Fakulta
potravinarske a biochemicke technologie. Praha, Vydavatelstvi