TY - JOUR
AU - De Vuyst, Luc
AB - Abstract The yeast species composition of 12 cocoa bean fermentations carried out in Brazil, Ecuador, Ivory Coast and Malaysia was investigated culture-independently. Denaturing gradient gel electrophoresis of 26S rRNA gene fragments, obtained through polymerase chain reaction with universal eukaryotic primers, was carried out with two different commercial apparatus (the DCode and CBS systems). In general, this molecular method allowed a rapid monitoring of the yeast species prevailing during fermentation. Under similar and optimal denaturing gradient gel electrophoresis conditions, the CBS system allowed a better separated band pattern than the DCode system and an unambiguous detection of the prevailing species present in the fermentation samples. The most frequent yeast species were Hanseniaspora sp., followed by Pichia kudriavzevii and Saccharomyces cerevisiae, independent of the origin of the cocoa. This indicates a restricted yeast species composition of the cocoa bean fermentation process. Exceptionally, the Ivorian cocoa bean box fermentation samples showed a wider yeast species composition, with Hyphopichia burtonii and Meyerozyma caribbica among the main representatives. Yeasts were not detected in the samples when the temperature inside the fermenting cocoa pulp-bean mass reached values higher than 45 °C or under early acetic acid production conditions. DGGE, yeast, cocoa bean fermentation Introduction Fermented, dry cocoa beans are the basic raw material for chocolate production (Beckett, 2009). The raw cocoa originates as seeds in fruit pods of the tree, Theobroma cacao L., and has to undergo a natural fermentation and drying process to develop essential colour and flavour precursors (Schwan & Wheals, 2004; De Vuyst et al., 2010). The latter compounds are transformed during cocoa bean roasting and chocolate manufacturing into the final characteristic chocolate flavour (Afoakwa et al., 2008; Camu et al., 2008a; Beckett, 2009). The fermentation step plays an important role for the final quality of the cocoa beans and, consequently, for chocolate produced thereof. As it remains a spontaneous process, cocoa bean fermentation is affected by the surrounding environment and local fermentation practices (Camu et al., 2007; Thompson et al., 2007; De Vuyst et al., 2010; Papalexandratou et al., 2011). In general, the sterile carbohydrate-rich cocoa pulp that surrounds the cocoa beans in the fruit pods is rapidly contaminated by microorganisms present in the environment after opening of the pods (Nielsen et al., 2005, 2007; Camu et al., 2007, 2008b). Fermentation methods depend on the cocoa-producing region and include heaps (Ghana, Ivory Coast), trays (Ghana, Ivory Coast), boxes (Brazil, Malaysia), baskets (Ivory Coast) and platforms (Ecuador) (Wood & Lass, 2001). During fermentation, microbial successions occur as the microenvironment (temperature, pH, oxygen availability) changes (Schwan & Wheals, 2004; De Vuyst et al., 2010). The main microbial groups that occur during a cocoa bean fermentation process are yeasts, lactic acid bacteria (LAB) and acetic acid bacteria (AAB). Yeasts are mainly responsible for ethanol production out of glucose (from sucrose) during the early stages of the fermentation, when the cocoa pulp-bean mass is anaerobic with low pH and temperature (Ardhana & Fleet, 2003; Camu et al., 2007, 2008b). Also, yeasts are the major contributors to pectinolysis for pulp removal, enabling air ingress in the cocoa pulp-bean mass (Schwan et al., 1995; Schwan & Wheals, 2004). LAB produce mainly lactic acid and/or mannitol out of citric acid and/or fructose under micro-aerophilic conditions, enabling a slight pH increase of the fermenting cocoa pulp-bean mass (Camu et al., 2007, 2008b). AAB appear at the later stage of fermentation, when more oxygen penetrates the fermenting cocoa pulp-bean mass, and they mainly oxidize the ethanol produced by yeasts into acetic acid (Camu et al., 2007, 2008b). The most frequently isolated yeast species are Saccharomyces cerevisiae, Hanseniaspora guilliermondii (anamorph Kloeckera apis), Hanseniaspora opuntiae, Pichia fermentans, Pichia kluyveri, Pichia kudriavzevii (formerly Issatchenkia orientalis, anamorph Candida krusei), Pichia membranifaciens and Wickerhamomyces anomalus (formerly Pichia anomala) (Schwan et al., 1995; Ardhana & Fleet, 2003; Schwan & Wheals, 2004; Jespersen et al., 2005; Nielsen et al., 2005, 2007; Lagunes-Gálvez et al., 2007; Leal et al., 2008; Daniel et al., 2009). Based on a combination of molecular analyses and morphological and physiological data, Daniel et al. (2009) have identified P. kudriavzevii, S. cerevisiae and H. opuntiae as the main yeast species involved in Ghanaian cocoa bean heap fermentations, confirming data from earlier studies (Jespersen et al., 2005; Nielsen et al., 2005, 2007). Furthermore, they have reported on several species that were not found in cocoa bean fermentation before, such as Candida carpophila, Candida orthopsilosis, Kodamaea ohmeri, Meyerozyma caribbica and Saccharomycodes ludwigii (Daniel et al., 2009). Also, the new species Candida halmiae and Candida awuaii have been reported recently (Nielsen et al., 2010). Most of the studies dealing with yeast identification are based on phenotypic characterization (Carr et al., 1979; Ardhana & Fleet, 2003), followed by molecular identification of the picked up isolates, often based on sequencing of the D1/D2 domain of the 26S rRNA gene (Jespersen et al., 2005; Nielsen et al., 2005; Daniel et al., 2009). PCR denaturing gradient gel electrophoresis (PCR-DGGE) has been proven to be an easy tool for monitoring the total yeast species composition in, for instance, cocoa bean (Nielsen et al., 2007), wine (Cocolin et al., 2000; Stringini et al., 2009) and sourdough (Meroth et al., 2003) fermentation samples. Using universal eukaryotic primers, the D1/D2 domain of the 26S rRNA gene is usually targeted, but other regions such as the 18S rRNA gene may be used as well (Beh et al., 2006). It has been shown that this kind of molecular approach overcomes the limitations of culture methods and reveals species that might fail to produce colonies on selected agar media. Hence, a more accurate representation of the composition of yeast species in food fermentation ecosystems may be obtained (Giraffa, 2004). However, a few drawbacks of this technique have been reported. One drawback of several yeast studies using DGGE is the multiple banding patterns displayed by single strains (Mills et al., 2002; Masoud et al., 2004; Prakitchaiwattana et al., 2004; Garofalo et al., 2008; Ongol & Asano, 2009). Also, some studies mention the presence of single-stranded DNA (ssDNA) artefacts in the DGGE patterns, although the actual reason of their appearance has not been defined (Cocolin et al., 2000; Masoud et al., 2004). One study has reported on the influence of the DGGE apparatus on the bacterial and yeast species composition of the ecosystem studied (Ascher et al., 2010). The aim of the present study was to rapidly monitor the species composition of yeasts involved in traditional spontaneous cocoa bean fermentations carried out in different cocoa-producing regions, namely Brazil, Ecuador, Ivory Coast and Malaysia, through 26S rRNA gene-PCR-DGGE, with a focus on the impact of the different origins of the cocoa and methods of fermentation (in particular heap, box and platform). Materials and methods Samples Fermentation samples were available from 12 field experiments set up in Ivory Coast (main crop of 2006; one farm; heap and box), Brazil [main crop of 2006 (two farms; box) and main crop of 2007 (two farms; box)], Ecuador (main crop of 2008; two farms; box and platform) and Malaysia (main crop of 2010; one farm, two plantations; box). During fermentation, samples of 500 g were taken every 6–12 h (and after mixing in the case of box fermentations) from the same depth, but in different points of the fermenting cocoa pulp-bean mass (c. 40 cm from the upper surface). Freshly taken samples were transported to the laboratory for immediate plating on malt extract agar (MEA; Oxoid, Basingstoke, UK) supplemented with 100 mg g−1 oxytetracycline for yeast enumeration. Then, all fermentation samples were temporarily stored at −20 °C before transport on dry ice to Belgium for further culture-independent analysis. Yeast counts were obtained after incubation of the agar media at 37 °C for 1–3 days. This incubation temperature was selected because of the high temperature inside the fermenting cocoa pulp-bean mass, where the samples were withdrawn from. Fermentation method, duration of fermentation, mixing/spreading points during fermentation, initial and final values of pH of and temperature inside the fermenting cocoa pulp-bean mass and initial and maximum yeast counts are summarized in Table 1. 1 Cocoa bean fermentations from which samples were available. The origin, fermentation practices and changes in physical parameters and yeast counts during fermentation are given Fermentation parameter      T (°C)  pH  Yeast counts (CFU g−1)  Origin of the cocoa bean fermentation  Duration (h)  Mixing/spreading point (h)  Ti  Tf  pHi  pHf  (CFU g−1)i  (CFU g−1)max; h  Ivory Coast (main crop 2006; 1 farm)  Heap (H)  150  –  24.0  39.6  N/A  N/A  N/A  N/A  Box (B)  150  24; 48  24.0  43.3  N/A  N/A  N/A  N/A  Brazil (main crop 2006; 2 farms)  Box (F1)  144  –  26.9  42.2  2.6  3.8  N/A  N/A  Box (F2)  144  –  30.4  48.2  2.8  4.6  N/A  N/A  Brazil (main crop 2007; 2 farms)  Box (B1)  144  54; 76; 96; 120  25.6  48.5  3.5  4.3  7.09  7.16; 12  Box (B2)  144  48; 72; 96; 120  25.6  47.6  3.4  4.3  5.48  7.22; 12  Ecuador (main crop 2008; 2 farms)  Platform (P1; cocoa from farm 1)  96  50; 72 (spreading for 3 h)  25.9  46.3  3.9  4.5  4.85  7.88; 72  Box (I1; cocoa from farm 1)  96  24; 72  28.7  46.5  3.7  4.2  5.53  7.77; 36  Platform (P2; cocoa from farm 2)  96  54; 72 (spreading for 3 h)  26.6  49.0  3.4  3.7  3.70  7.77; 24  Box (I2; cocoa from farm 2)  96  24; 72  25.7  47.9  3.7  4.1  4.28  7.65; 42  Malaysia (main crop 2010; 2 plantations)  Box (M1)  120  48; 96  28.5  43.9  3.9  4.4  6.28  7.73; 12  Box (M2)  120  48; 96  31.8  42.7  3.9  4.2  5.29  7.24; 36  Fermentation parameter      T (°C)  pH  Yeast counts (CFU g−1)  Origin of the cocoa bean fermentation  Duration (h)  Mixing/spreading point (h)  Ti  Tf  pHi  pHf  (CFU g−1)i  (CFU g−1)max; h  Ivory Coast (main crop 2006; 1 farm)  Heap (H)  150  –  24.0  39.6  N/A  N/A  N/A  N/A  Box (B)  150  24; 48  24.0  43.3  N/A  N/A  N/A  N/A  Brazil (main crop 2006; 2 farms)  Box (F1)  144  –  26.9  42.2  2.6  3.8  N/A  N/A  Box (F2)  144  –  30.4  48.2  2.8  4.6  N/A  N/A  Brazil (main crop 2007; 2 farms)  Box (B1)  144  54; 76; 96; 120  25.6  48.5  3.5  4.3  7.09  7.16; 12  Box (B2)  144  48; 72; 96; 120  25.6  47.6  3.4  4.3  5.48  7.22; 12  Ecuador (main crop 2008; 2 farms)  Platform (P1; cocoa from farm 1)  96  50; 72 (spreading for 3 h)  25.9  46.3  3.9  4.5  4.85  7.88; 72  Box (I1; cocoa from farm 1)  96  24; 72  28.7  46.5  3.7  4.2  5.53  7.77; 36  Platform (P2; cocoa from farm 2)  96  54; 72 (spreading for 3 h)  26.6  49.0  3.4  3.7  3.70  7.77; 24  Box (I2; cocoa from farm 2)  96  24; 72  25.7  47.9  3.7  4.1  4.28  7.65; 42  Malaysia (main crop 2010; 2 plantations)  Box (M1)  120  48; 96  28.5  43.9  3.9  4.4  6.28  7.73; 12  Box (M2)  120  48; 96  31.8  42.7  3.9  4.2  5.29  7.24; 36  For each fermentation, the following parameters are listed: temperature (T) [initial (Ti) and final (Tf) temperature] and pH [initial (pHi) and final (pHf) pH value] inside the fermenting cocoa pulp-bean mass. Yeast counts [colony forming units (CFU) g−1] were obtained by plating on malt extract agar supplemented with 100 mg g−1 oxytetracycline (37 °C; 1–3 days): at the beginning (CFU g−1)i of the fermentation and at their maximum (CFU g−1)max during the fermentation. N/A, not available. Cocoa bean fermentations carried out without separation of healthy and infected pods and no regular mixing. Cocoa bean fermentations carried out with organic cocoa. View Large 1 Cocoa bean fermentations from which samples were available. The origin, fermentation practices and changes in physical parameters and yeast counts during fermentation are given Fermentation parameter      T (°C)  pH  Yeast counts (CFU g−1)  Origin of the cocoa bean fermentation  Duration (h)  Mixing/spreading point (h)  Ti  Tf  pHi  pHf  (CFU g−1)i  (CFU g−1)max; h  Ivory Coast (main crop 2006; 1 farm)  Heap (H)  150  –  24.0  39.6  N/A  N/A  N/A  N/A  Box (B)  150  24; 48  24.0  43.3  N/A  N/A  N/A  N/A  Brazil (main crop 2006; 2 farms)  Box (F1)  144  –  26.9  42.2  2.6  3.8  N/A  N/A  Box (F2)  144  –  30.4  48.2  2.8  4.6  N/A  N/A  Brazil (main crop 2007; 2 farms)  Box (B1)  144  54; 76; 96; 120  25.6  48.5  3.5  4.3  7.09  7.16; 12  Box (B2)  144  48; 72; 96; 120  25.6  47.6  3.4  4.3  5.48  7.22; 12  Ecuador (main crop 2008; 2 farms)  Platform (P1; cocoa from farm 1)  96  50; 72 (spreading for 3 h)  25.9  46.3  3.9  4.5  4.85  7.88; 72  Box (I1; cocoa from farm 1)  96  24; 72  28.7  46.5  3.7  4.2  5.53  7.77; 36  Platform (P2; cocoa from farm 2)  96  54; 72 (spreading for 3 h)  26.6  49.0  3.4  3.7  3.70  7.77; 24  Box (I2; cocoa from farm 2)  96  24; 72  25.7  47.9  3.7  4.1  4.28  7.65; 42  Malaysia (main crop 2010; 2 plantations)  Box (M1)  120  48; 96  28.5  43.9  3.9  4.4  6.28  7.73; 12  Box (M2)  120  48; 96  31.8  42.7  3.9  4.2  5.29  7.24; 36  Fermentation parameter      T (°C)  pH  Yeast counts (CFU g−1)  Origin of the cocoa bean fermentation  Duration (h)  Mixing/spreading point (h)  Ti  Tf  pHi  pHf  (CFU g−1)i  (CFU g−1)max; h  Ivory Coast (main crop 2006; 1 farm)  Heap (H)  150  –  24.0  39.6  N/A  N/A  N/A  N/A  Box (B)  150  24; 48  24.0  43.3  N/A  N/A  N/A  N/A  Brazil (main crop 2006; 2 farms)  Box (F1)  144  –  26.9  42.2  2.6  3.8  N/A  N/A  Box (F2)  144  –  30.4  48.2  2.8  4.6  N/A  N/A  Brazil (main crop 2007; 2 farms)  Box (B1)  144  54; 76; 96; 120  25.6  48.5  3.5  4.3  7.09  7.16; 12  Box (B2)  144  48; 72; 96; 120  25.6  47.6  3.4  4.3  5.48  7.22; 12  Ecuador (main crop 2008; 2 farms)  Platform (P1; cocoa from farm 1)  96  50; 72 (spreading for 3 h)  25.9  46.3  3.9  4.5  4.85  7.88; 72  Box (I1; cocoa from farm 1)  96  24; 72  28.7  46.5  3.7  4.2  5.53  7.77; 36  Platform (P2; cocoa from farm 2)  96  54; 72 (spreading for 3 h)  26.6  49.0  3.4  3.7  3.70  7.77; 24  Box (I2; cocoa from farm 2)  96  24; 72  25.7  47.9  3.7  4.1  4.28  7.65; 42  Malaysia (main crop 2010; 2 plantations)  Box (M1)  120  48; 96  28.5  43.9  3.9  4.4  6.28  7.73; 12  Box (M2)  120  48; 96  31.8  42.7  3.9  4.2  5.29  7.24; 36  For each fermentation, the following parameters are listed: temperature (T) [initial (Ti) and final (Tf) temperature] and pH [initial (pHi) and final (pHf) pH value] inside the fermenting cocoa pulp-bean mass. Yeast counts [colony forming units (CFU) g−1] were obtained by plating on malt extract agar supplemented with 100 mg g−1 oxytetracycline (37 °C; 1–3 days): at the beginning (CFU g−1)i of the fermentation and at their maximum (CFU g−1)max during the fermentation. N/A, not available. Cocoa bean fermentations carried out without separation of healthy and infected pods and no regular mixing. Cocoa bean fermentations carried out with organic cocoa. View Large DNA extraction DNA was directly extracted from the fermentation samples, as described previously (Camu et al., 2007). The original protocol was as follows. Twenty grams of frozen cocoa beans plus pulp samples were washed twice with 70 mL of saline solution [0.85% (w/v) NaCl]. The combined fluid (±120 mL) was removed by decanting and subsequently centrifuged at 170 g at 4 °C for 5 min to remove large particles. The supernatant was filtered through a coffee filter and the filtrate was centrifuged at 8000 g at 4 °C for 20 min to pellet the cells, which were subsequently frozen at −20 °C overnight. The thawed pellet was washed in 1 mL of TES buffer [6.7% (w/v) sucrose, 50 mM Tris–HCl, pH 8.0, 1 mM EDTA] and resuspended in 300 μL of STET buffer [8% (w/v) sucrose, 5% (w/v) Triton X-100, 50 mM Tris–HCl, pH 8.0, 50 mM EDTA]. Seventy-five microlitres of lysis buffer [TES containing 1330 U mL−1 of mutanolysin (Sigma-Aldrich, St. Louis, MO) and 100 mg mL−1 of lysozyme (Sigma-Aldrich)] and 100 μL of a solution of proteinase K [2.5 mg mL−1 of TE buffer (10 mM Tris–HCl; 1 mM EDTA, pH 8.0); VWR International, Darmstadt, Germany] were added for cell lysis, and the suspension was incubated at 37 °C for 1 h. After the addition of 40 μL preheated (37 °C) 20% (w/v) sodium dodecyl sulphate in TE buffer and a pinch of glass beads with a diameter of 150–212 μm (Sigma-Aldrich), cells were vortexed for 60 s and incubated at 37 °C for 10 min, followed by a 10-min incubation at 65 °C. One-hundred microlitres of TE buffer were added, and the lysate was extracted with 1 volume of phenol–chloroform–isoamyl alcohol (49 : 49 : 1) (Sigma-Aldrich) for 30 s. Phases were separated by microcentrifugation (18 900 g for 5 min at 4 °C) using Phase Lock Gel tubes (Eppendorf AG, Hamburg, Germany). To get rid of cocoa pulp compounds potentially inhibiting PCR, such as polysaccharides, proteins, enzymes and polyphenols, a NucleoSpin column (Macherey Nagel GmbH, Düren, Germany) was used for further purification of the DNA-containing aqueous phase, following the manufacturer's instructions. The final DNA samples (50 ng μL−1) were stored at −20 °C until further use. For a few samples, an extra enzymatic treatment with lyticase (20 mg g−1; Sigma-Aldrich) and lysing enzymes from Trichoderma harzianum (40 mg g−1; Sigma-Aldrich) was performed for cell lysis to optimize the DNA extraction of yeasts (Meroth et al., 2003). However, as no different DGGE patterns were found, most experiments were carried out according to the original protocol. To extract DNA from yeast cultures to construct an identification ladder, the protocol of Gevers et al. (2001) was used with minor modifications (Camu et al., 2007). Briefly, total genomic DNA was extracted from stock cultures that were propagated twice in YG medium (glucose, 20 g L−1; yeast extract, 5 g L−1) following the phenol–chloroform–isoamyl alcohol method. The enzymatic treatment for the DNA extraction consisted of a final concentration of 5 U μL−1 of lysozyme (VWR International) and 0.8 U μL−1 of mutanolysine (Sigma-Aldrich), suspended in TES buffer. The final DNA samples were diluted in TE buffer to a concentration of 50 ng μL−1 and stored at −20 °C until further use. 26S rRNA gene-PCR-DGGE The yeast 26S rRNA gene fragment (250 bp) was amplified using the eukaryotic universal primers NL1 (5′-GCA TAT CAA TAA GCG GAG GAA AAG-3′), containing a GC-clamp (5′-CGC CCG CCG CGC GCG GCG GGC GGG GCG GGG-3′) at the 5′ end and LS2 (5′-ATT CCC AAA CAA CTC GAC TC-3′) (Cocolin et al., 2000; Nielsen et al., 2007). The PCR mixture and conditions were as reported by Vasilopoulos et al. (2008) and Nielsen et al. (2007), respectively. To optimize the monitoring of yeast species through 26S rRNA gene-PCR-DGGE, two commercial DGGE apparatus were used: the DCode system (gels of 16 cm × 16 cm × 0.1 cm; Bio-Rad, Hercules, CA) and the CBS Scientific system (gels of 17.7 cm × 22 cm × 0.1 cm; CBS Scientific Company, San Diego, CA). For both systems, electrophoresis of the PCR amplicons was carried out in 1.0× TAE buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.0) at 70 V for 16 h at a constant temperature of 60 °C. For the separation of the PCR amplicons, an optimization of the denaturing gradients (25–60%, 35–70% and 35–60%) was performed. A 35–60% denaturing gradient was optimal (data not shown). A reference ladder was included in all gels for normalization of the gels for numerical analysis with BioNumerics version 5.1 software (Applied Maths, Sint-Martens-Latem, Belgium) as well as for preliminary identification of DGGE bands in the lanes of the fermentation samples. This ladder was constructed by running PCR amplicons from DNA of several yeast strains on DGGE gels of the DCode DGGE system. All these yeasts were isolates from the Ecuadorian cocoa bean fermentation samples and were identified previously through a multi-gene sequencing approach, including D1/D2 LSU, ITS and ACT1 gene sequencing. The accession numbers are given below. Using this ladder, in combination with DGGE band sequencing, the yeast species composition and community dynamics of all fermentation samples were analysed. For numerical analysis of DGGE band patterns, the band-based Dice coefficient was used to discriminate among them concerning the influence of mixing, duration of the fermentation or fermentation method. DGGE band sequencing DGGE bands of interest were excised from the gels with a sterile blade, mixed with 40 μL of sterile water and put at 4 °C for 24 h to let the DNA diffuse out of the bands. Then, the bands were vigorously vortexed for 15–20 min and microcentrifuged (1000 g) for 15 s to collect the aqueous DNA solution. Three microlitres of this solution were used for a PCR assay with the same primers as mentioned above, excluding the GC-clamp, under the same PCR conditions. The PCR products were purified using the Wizard Plus SV Minipreps DNA Purification System (Promega, Madison, WI) and sequenced in a commercial facility (VIB, Antwerp, Belgium). Searches in the GenBank database were performed with the Blast (Basic Local Alignment Search Tool) program to determine the closest known relatives of the partial 26S rRNA gene sequences (http://www.ncbi.nlm.nih.gov/BLAST). Accession numbers are indicated in parentheses at the appropriate places. Results Comparison of two commercial DGGE systems for yeast species composition monitoring of cocoa bean fermentation samples Reference yeast strains The PCR amplicons of DNA from the Ecuadorian cocoa reference yeast strains migrated differently in the polyacrylamide gels of the two commercial systems and their overall patterns differed at species level (Fig. 1). The DCode system gave multiple banding patterns for almost all strains tested (Fig. 1a). All S. cerevisiae strains gave the same pattern, composed of three bands with the DCode DGGE system, which showed 100% identity with S. cerevisiae strain SS1-1 (HM123752) after Blast analysis (Fig. 1a). A better separation of the bands and a distinct DGGE image were obtained with the CBS system; in addition, no multiple banding patterns occurred, except for P. kudriavzevii (Fig. 1b). Therefore, the final selection of strains per species for the construction of an identification ladder was based on their migration in gels of the CBS DGGE system (Fig. 1b), which were (according to the direction of electrophoresis): H. opuntiae Y120 (FR870033), Candida tropicalis Y49 (FR870028), Torulaspora delbrueckii Y7 (FR870025), Kluyveromyces marxianus Y40 (FR870027), S. cerevisiae D652 (representative GenBank accession number HM123752) and P. kudriavzevii D695 (representative GenBank accession number HM191632). However, concerning Hanseniaspora, no distinction could be made between the species H. opuntiae, H. guilliermondii and Hanseniaspora uvarum, as these three species yielded the same DGGE band because there are no differences in the nucleotide sequences of the corresponding 26S rRNA gene fragments. Although the two DNA bands of P. kudriavzevii were close to each other, they did not interfere with bands of other yeast strains in the ladder. Also, an intense band was present in almost all DCode DGGE patterns in the lower part of the gels, interfering with the detection of species migrating almost at the same height of this band, such as P. kudriavzevii (Fig. 1a). This band was not visible in the gel of the CBS System, indicating differences in gradient formation between the two DGGE systems. 1 View largeDownload slide DGGE patterns of identified cocoa yeast strains obtained with the DCode DGGE system (a). Strains selected for the ladder (L) construction, optimized using the CBS Scientific DGGE system (b). The lanes represent: (1) Pichia kudriavzeviiD695 (representative GenBank accession number HM191632); (2) Pichia kudriavzeviiD642 (representative GenBank accession number HM191632); (3) Saccharomyces cerevisiaeD652 (representative GenBank accession number HM123752); (4) Saccharomyces cerevisiaeD650 (representative GenBank accession number HM123752); (5) Saccharomyces cerevisiaeD534 (representative GenBank accession number HM123752); (6) Torulaspora delbrueckiiY7 (FR870025); (7) Candida sorborivorans-like Y54 (FR870030); (8) Pichia kluyveriY57 (FR870031); (9) Kluyveromyces marxianusY40 (FR870037); (10) Candida tropicalisY49 (FR870028); (11) Hanseniaspora opuntiaeY120 (FR870033); and (12) Rhodotorula minutaY50 (FR870029). 1 View largeDownload slide DGGE patterns of identified cocoa yeast strains obtained with the DCode DGGE system (a). Strains selected for the ladder (L) construction, optimized using the CBS Scientific DGGE system (b). The lanes represent: (1) Pichia kudriavzeviiD695 (representative GenBank accession number HM191632); (2) Pichia kudriavzeviiD642 (representative GenBank accession number HM191632); (3) Saccharomyces cerevisiaeD652 (representative GenBank accession number HM123752); (4) Saccharomyces cerevisiaeD650 (representative GenBank accession number HM123752); (5) Saccharomyces cerevisiaeD534 (representative GenBank accession number HM123752); (6) Torulaspora delbrueckiiY7 (FR870025); (7) Candida sorborivorans-like Y54 (FR870030); (8) Pichia kluyveriY57 (FR870031); (9) Kluyveromyces marxianusY40 (FR870037); (10) Candida tropicalisY49 (FR870028); (11) Hanseniaspora opuntiaeY120 (FR870033); and (12) Rhodotorula minutaY50 (FR870029). Fermentation samples Both DGGE systems allowed fast monitoring of the yeast species composition of the cocoa bean fermentation samples studied. However, an important difference in detected yeast species between the two systems was found, as illustrated for the Ivorian cocoa bean box fermentation samples (Fig. 2). As mentioned above, a fuzzy band was formed in the lower part of the DCode system gels in all samples, making accurate detection and successful sequencing of co-migrating yeast DNA bands difficult or impossible, in particular, those corresponding to S. cerevisiae and P. kudriavzevii. 2 View largeDownload slide Yeast DGGE patterns of the Ivorian cocoa bean box fermentation samples using the DCode DGGE system (a) and the CBS Scientific DGGE system (b). The lane numbers represent the time (h) of sampling during fermentation. Lane L represents the ladder of the reference yeast strains. 2 View largeDownload slide Yeast DGGE patterns of the Ivorian cocoa bean box fermentation samples using the DCode DGGE system (a) and the CBS Scientific DGGE system (b). The lane numbers represent the time (h) of sampling during fermentation. Lane L represents the ladder of the reference yeast strains. Yeast species composition of cocoa bean fermentation samples Comparing the semi-quantitative PCR-DGGE data with culture-dependent data (Table 2), Hanseniaspora sp. (100% identity with H. opuntiae, H. uvarum and H. guilliermondii) was prevailing in all fermentation samples, independent of their origin and fermentation method (Fig. 3). Based on the presence of DGGE bands corresponding with P. kudriavzevii and/or S. cerevisiae, these yeast species were the second most common ones found in several fermentation samples of most fermentations (Fig. 3). This indicated a restricted yeast species composition involved in the cocoa bean fermentation processes (Fig. 3). An exception was the Ivorian box fermentation that showed a wider yeast species composition, with Hyphopichia burtonii and Meyerozyma caribbica among the main representatives (Fig. 3a). 2 Overall results of the yeast species detected in each sample, including culture-dependent data from previous studies Cocoa-producing region  Culture-dependent analysis  Culture-independent analysis  Reference  Ivory Coast  Not available  Hanseniaspora sp.  This study  Hyphopichia burtonii Meyerozyma caribbica  Pichia kudriavzevii  Pichia veronae/fabianni  Saccharomyces cerevisiae  Brazil  Not available  Candida jaroonii/friedrichii  This study  Hanseniaspora sp.  Hanseniaspora vineae  Pichia kudriavzevii  Saccharomyces cerevisiae  Wickerhamomyces anomalus  Ecuador  Candida sorbosivorans-like  –  This study; Z. Papalexandratou, G. Falony, E. Romanens, J.C. Jimenez, F. Amores, H.M. Daniel, L. De Vuyst, unpublished results  Candida tropicalis  –  –  Debaryomyces sp./Candida sp.  Hanseniaspora opuntiae  Hanseniaspora sp.  –  Issatchenkia tericola  Kluyveromyces marxianus  –  Pichia kluyveri  –  Pichia kudriavzevii  Pichia kudriavzevii  Pichia manshurica  –  Rhodotorula minuta  –  Saccharomyces cerevisiae  –  Torulaspora delbrueckii  –  Malaysia  Not available  Hanseniaspora sp.  This study  Pichia kudriavzevii  Saccharomyces cerevisiae  Torulaspora delbrueckii  Cocoa-producing region  Culture-dependent analysis  Culture-independent analysis  Reference  Ivory Coast  Not available  Hanseniaspora sp.  This study  Hyphopichia burtonii Meyerozyma caribbica  Pichia kudriavzevii  Pichia veronae/fabianni  Saccharomyces cerevisiae  Brazil  Not available  Candida jaroonii/friedrichii  This study  Hanseniaspora sp.  Hanseniaspora vineae  Pichia kudriavzevii  Saccharomyces cerevisiae  Wickerhamomyces anomalus  Ecuador  Candida sorbosivorans-like  –  This study; Z. Papalexandratou, G. Falony, E. Romanens, J.C. Jimenez, F. Amores, H.M. Daniel, L. De Vuyst, unpublished results  Candida tropicalis  –  –  Debaryomyces sp./Candida sp.  Hanseniaspora opuntiae  Hanseniaspora sp.  –  Issatchenkia tericola  Kluyveromyces marxianus  –  Pichia kluyveri  –  Pichia kudriavzevii  Pichia kudriavzevii  Pichia manshurica  –  Rhodotorula minuta  –  Saccharomyces cerevisiae  –  Torulaspora delbrueckii  –  Malaysia  Not available  Hanseniaspora sp.  This study  Pichia kudriavzevii  Saccharomyces cerevisiae  Torulaspora delbrueckii  View Large 2 Overall results of the yeast species detected in each sample, including culture-dependent data from previous studies Cocoa-producing region  Culture-dependent analysis  Culture-independent analysis  Reference  Ivory Coast  Not available  Hanseniaspora sp.  This study  Hyphopichia burtonii Meyerozyma caribbica  Pichia kudriavzevii  Pichia veronae/fabianni  Saccharomyces cerevisiae  Brazil  Not available  Candida jaroonii/friedrichii  This study  Hanseniaspora sp.  Hanseniaspora vineae  Pichia kudriavzevii  Saccharomyces cerevisiae  Wickerhamomyces anomalus  Ecuador  Candida sorbosivorans-like  –  This study; Z. Papalexandratou, G. Falony, E. Romanens, J.C. Jimenez, F. Amores, H.M. Daniel, L. De Vuyst, unpublished results  Candida tropicalis  –  –  Debaryomyces sp./Candida sp.  Hanseniaspora opuntiae  Hanseniaspora sp.  –  Issatchenkia tericola  Kluyveromyces marxianus  –  Pichia kluyveri  –  Pichia kudriavzevii  Pichia kudriavzevii  Pichia manshurica  –  Rhodotorula minuta  –  Saccharomyces cerevisiae  –  Torulaspora delbrueckii  –  Malaysia  Not available  Hanseniaspora sp.  This study  Pichia kudriavzevii  Saccharomyces cerevisiae  Torulaspora delbrueckii  Cocoa-producing region  Culture-dependent analysis  Culture-independent analysis  Reference  Ivory Coast  Not available  Hanseniaspora sp.  This study  Hyphopichia burtonii Meyerozyma caribbica  Pichia kudriavzevii  Pichia veronae/fabianni  Saccharomyces cerevisiae  Brazil  Not available  Candida jaroonii/friedrichii  This study  Hanseniaspora sp.  Hanseniaspora vineae  Pichia kudriavzevii  Saccharomyces cerevisiae  Wickerhamomyces anomalus  Ecuador  Candida sorbosivorans-like  –  This study; Z. Papalexandratou, G. Falony, E. Romanens, J.C. Jimenez, F. Amores, H.M. Daniel, L. De Vuyst, unpublished results  Candida tropicalis  –  –  Debaryomyces sp./Candida sp.  Hanseniaspora opuntiae  Hanseniaspora sp.  –  Issatchenkia tericola  Kluyveromyces marxianus  –  Pichia kluyveri  –  Pichia kudriavzevii  Pichia kudriavzevii  Pichia manshurica  –  Rhodotorula minuta  –  Saccharomyces cerevisiae  –  Torulaspora delbrueckii  –  Malaysia  Not available  Hanseniaspora sp.  This study  Pichia kudriavzevii  Saccharomyces cerevisiae  Torulaspora delbrueckii  View Large Pichia kudriavzevii was present at the beginning of the Ivorian heap fermentation and was succeeded by Hanseniaspora sp. after 48 h of fermentation (Fig. 3a). In the case of the two Brazilian cocoa bean box fermentations (F1 and F2), Hanseniaspora sp. was present in parallel with P. kudriavzevii during the first 36 and 54 h of fermentations F1 and F2 respectively (Fig. 3b). However, P. kudriavzevii was absent in the two Brazilian cocoa bean box fermentations (B1 and B2) (Fig. 3c). In Ecuador, all box and platform fermentations showed the same yeast species composition; Hanseniaspora sp. and P. kudriavzevii were both present, although the latter species appeared, in some cases, as a band of low intensity (Fig. 3d and e). During the Malaysian box fermentations, S. cerevisiae was present during the whole fermentation process, together with Hanseniaspora sp. (Fig. 3f). 3 View largeDownload slide Yeast DGGE patterns of cocoa bean fermentation samples obtained with the CBS Scientific DGGE system. (a) Ivorian box and heap fermentations carried out in 2006; (b) Brazilian box fermentations carried out in 2006; (c) Brazilian box fermentations carried out in 2007; (d) Ecuadorian platform fermentations carried out in 2008; (e) Ecuadorian box fermentations carried out in 2008; (f) Malaysian box fermentations carried out in 2010. The lane numbers represent the time (h) of sampling during fermentation. Lane L represents the ladder of the reference yeast strains. The closest relatives of the fragments sequenced (% of identical nucleotides compared with sequences retrieved from the GenBank database and representative Accession No. between parentheses) were: (i) Hanseniaspora opuntiae/uvarum/guilliermondii (100%; FM180549, AB438210); (ii) Wickerhamomyces anomalus (100%; FJ515235); (iii) Hyphopichia burtonii (100%; GQ222349); (iv) Pichia veronae/fabianni (100%; AB436465, EF550321); (v) Meyerozyma caribbica (100%; AB557838); (vi) Debaryomyces sp./Candida sp. (100%; FJ432624, AY520416); (vii) Candida jaroonii/friedrichii (100%; AB292057); (viii) Hanseniaspora vineae (100%; HM191667); (ix) Saccharomyces cerevisiae (100%; HM123752); (x) Issatchenkia terricola (100%; EF550233); and (xi) Pichia kudriavzevii (100%; HM191632). 3 View largeDownload slide Yeast DGGE patterns of cocoa bean fermentation samples obtained with the CBS Scientific DGGE system. (a) Ivorian box and heap fermentations carried out in 2006; (b) Brazilian box fermentations carried out in 2006; (c) Brazilian box fermentations carried out in 2007; (d) Ecuadorian platform fermentations carried out in 2008; (e) Ecuadorian box fermentations carried out in 2008; (f) Malaysian box fermentations carried out in 2010. The lane numbers represent the time (h) of sampling during fermentation. Lane L represents the ladder of the reference yeast strains. The closest relatives of the fragments sequenced (% of identical nucleotides compared with sequences retrieved from the GenBank database and representative Accession No. between parentheses) were: (i) Hanseniaspora opuntiae/uvarum/guilliermondii (100%; FM180549, AB438210); (ii) Wickerhamomyces anomalus (100%; FJ515235); (iii) Hyphopichia burtonii (100%; GQ222349); (iv) Pichia veronae/fabianni (100%; AB436465, EF550321); (v) Meyerozyma caribbica (100%; AB557838); (vi) Debaryomyces sp./Candida sp. (100%; FJ432624, AY520416); (vii) Candida jaroonii/friedrichii (100%; AB292057); (viii) Hanseniaspora vineae (100%; HM191667); (ix) Saccharomyces cerevisiae (100%; HM123752); (x) Issatchenkia terricola (100%; EF550233); and (xi) Pichia kudriavzevii (100%; HM191632). A few single bands were detected during the first hours of some fermentation, such as those corresponding with Wickerhamomyces anomalus, Debaryomyces sp. (100% identity with Debaryomyces hansenii, Debaryomyces nepalensis and Debaryomyces mycophilus), Candida jaroonii/friedrichii and Hanseniaspora vineae, perhaps accidental contaminants from the environment (Fig. 3c and d). Although all fermentations lasted for 4–6 days, yeast bands were not always found at the end of fermentation. In contrast, in the case of the Malaysian box fermentations, yeasts were detected during the whole fermentation process; in particular, Hanseniaspora sp. was still found after 4 days of fermentation (Fig. 3f). Based on numerical analysis of all DGGE band patterns, no significant influence of fermentation practices, such as mixing of the fermenting cocoa pulp-bean mass or duration of the fermentation or fermentation methods (heap, box, platform) on the yeast species composition was seen (Fig. 4). However, yeasts were only detected for a short time in the case of fermentations characterized by practices such as no separation of healthy and infected pods, no mixing of the fermenting cocoa pulp-bean mass or spreading of the fermenting beans during fermentation (Fig. 3b and d), indicating atypical processing. 4 View largeDownload slide Dendrogram derived from the cluster analysis of the 26S rRNA gene-PCR-DGGE patterns of the yeast communities associated with cocoa bean fermentation samples from Ivory Coast, Brazil, Ecuador and Malaysia, based on the Dice coefficient of similarity (weighted) and obtained with the UPGMA clustering algorithm. 4 View largeDownload slide Dendrogram derived from the cluster analysis of the 26S rRNA gene-PCR-DGGE patterns of the yeast communities associated with cocoa bean fermentation samples from Ivory Coast, Brazil, Ecuador and Malaysia, based on the Dice coefficient of similarity (weighted) and obtained with the UPGMA clustering algorithm. Discussion During the last decade, the culture-independent PCR-DGGE method has been used frequently to monitor the microbial species composition and community dynamics during food fermentation processes; however, its detection limit is rather high (103–104 CFU g−1, depending on the species and/or strain) and data are semi-quantitative (Ercolini, 2004; Giraffa, 2004). In the present study, 26S rRNA gene-PCR-DGGE was used for monitoring the yeast community dynamics and species composition during cocoa bean fermentations carried out in several cocoa-producing regions. In general, DGGE allowed a fast detection of the main yeast species involved in the process, representing at least 90% of the total yeast communities, as has been seen for detailed studies on the bacterial species composition revealed by plating, PCR-DGGE and rDNA libraries (Garcia-Armisen et al., 2010). However, the technique was not free of drawbacks, mainly in the case of the DCode apparatus, as sometimes multiple banding patterns were formed for a single species. The reasons for this appearance are not well understood. In general, it may reflect artefacts of the PCR assays using primers with GC clamps and DNA denaturation kinetics during electrophoresis (Beh et al., 2006; Rettedal et al., 2010). Also, multiple bands could arise from nucleotide variations among multiple rDNA copies within a single strain or could indicate the presence of different strains within a species (Beh et al., 2006). Artefacts of double bands of PCR amplicons of several species could be eliminated by extension of the final elongation step of the PCR assays (Janse et al., 2004). Finally, the presence of a fuzzy band in the lower part of the DCode gels, which affects accurate sequencing of excised DGGE bands, has been ascribed to ssDNA artefacts that are not influenced differentially by the gradient (Heuer et al., 1997; Cocolin et al., 2000). Most of these drawbacks were overcome when using the CBS apparatus, allowing a better separation of the yeast bands in the reference ladder as well as distinct yeast DGGE patterns for all cocoa bean fermentations studied. Comparison of the two apparatus revealed few differences in gradient formation, perhaps because of the different dimensions of the gels. The fact that the DGGE apparatus can affect DGGE-based yeast and bacterial community structure analysis has already been reported, in the case of soil samples (Ascher et al., 2010). Concerning yeast species composition, Hanseniaspora sp. was the predominating yeast species in all fermentations of the present study. For the Brazilian box fermentations (B1 and B2), carried out with organic cocoa under optimal fermentation practices such as selection of healthy pods solely, use of washed equipment and regular mixing during fermentation (Garcia-Armisen et al., 2010; Papalexandratou et al., 2011), Hanseniaspora sp. was the only species present throughout the fermentation process, thus suggesting its possible importance for the success of the fermentation. Distinction between Hanseniaspora opuntiae, H. uvarum and H. guilliermondii is only possible by multi-gene sequence analysis (Cadez et al., 2003, 2006; Daniel et al., 2009). However, the results of the culture-dependent study of Daniel et al. (2009) that used accurate molecular identifications suggests that H. opuntiae is the predominant species during cocoa bean fermentation (Daniel et al., 2009). In the study of Nielsen et al. (2007), 247 yeast isolates were analysed revealing the dominance of H. guilliermondii and P. membranifaciens. In the study of Daniel et al. (2009), H. opuntiae prevails during the initial phase of Ghanaian cocoa bean heap fermentations, based on the analysis of 90 isolates from seven fermentation heaps. The ethanol tolerance up to 5–10% (v/v) of H. opuntiae explains its dominance during cocoa bean fermentation, wherein ethanol never exceeds 2.5% (v/v) (Camu et al., 2007, 2008a, b; Daniel et al., 2009). Originally, H. opuntiae was isolated from Opuntia ficus-indica rot in Hawaii (Cadez et al., 2003). During the present study and based on the samples analysed, P. kudriavzevii and/or S. cerevisiae were the second most common yeast species found for most fermentations. The former species was less prevalent during the Ecuadorian cocoa bean fermentations, whereas the latter species prevailed during the Malaysian box fermentations. In general, the most frequently isolated cocoa yeast species is S. cerevisiae (Schwan & Wheals, 2004; De Vuyst et al., 2010). Pichia kudriavzevii has been reported as an acid- and ethanol-tolerant yeast species (Okuma et al., 1986; Daniel et al., 2009). It ferments reducing carbohydrates at a slower rate than S. cerevisiae, but it is able to use malic acid efficiently, in contrast to S. cerevisiae (Kim et al., 2008). Although P. kudriavzevii partially reduces citric acid, cocoa-originating H. opuntiae and S. cerevisiae isolates do not assimilate citric acid, indicating a possible (minor) role of P. kudriavzevii in citric acid conversion besides LAB (Camu et al., 2007, 2008a, b; Daniel et al., 2009). Meyerozyma caribbica and H. burtonii represented the main yeast species involved in the Ivorian box fermentations and the latter had not been yet documented from cocoa bean fermentations before. Meyerozyma caribbica was originally isolated from sugar cane in Cuba (Vaughan-Martini et al., 2005). Hyphopichia burtonii was originally isolated from murcha, a natural starter culture for the production of chunga/jand, a fermented liquor from Nepal, and appears to possess amylolytic activity (Takeuchi et al., 2006). It has been isolated from tempeh, a fermented soy product as well (Aldsworth et al., 2009). Pichia kudriavzevii, S. cerevisiae and H. opuntiae represent the main yeast species of seven Ghanaian cocoa bean heap fermentations, as revealed through M13-PCR fingerprinting of yeast isolates and multiple locus gene sequencing in the study of Daniel et al. (2009). Application of 26S rRNA gene-PCR-DGGE on Ghanaian heap and tray fermentation samples with the DCode apparatus has revealed similar results in the study of Nielsen et al. (2007), whereby H. guilliermondii was the prevalent band followed by P. membranifaciens and P. kudriavzevii. Although H. guilliermondii and P. membranifaciens were not found during the present study, the former species in the study of Nielsen et al. (2007) should have been identified as H. opuntiae (Daniel et al., 2009), based on the elongation factor-1α and ACT1 gene sequences (Cadez et al., 2003, 2006). The absence of P. membranifaciens may be ascribed to the dominance of H. opuntiae, as substrate competition between these species has been reported (Jespersen et al., 2005). Finally, high counts of the species P. kudriavzevii have been found on the surface of pods infected with black pod disease (Jespersen et al., 2005). Concerning yeast community dynamics, no specific temporal distribution was recognized for H. opuntiae, P. kudriavzevii or S. cerevisiae, as all species could be found at the beginning, in the middle or at the end of the cocoa bean fermentation processes, reflecting their adaptation to the cocoa pulp-bean mass ecosystem. Rarely encountered species, such as Debaryomyces sp., H. vineae, C. jaroonii/friedrichii and W. anomalus, only occurred during the initial phase of the fermentation, perhaps because of their low ethanol tolerance. However, Hanseniaspora seems to be more acid- and heat-tolerant, as it could still be found after 4 days of fermentation, as seen in the Malaysian box fermentations (Fig. 3f), when AAB have already produced high amounts of acetic acid. In general, yeasts were not detected in the case of a fermentation temperature above 45 °C or at high acidity. This was for instance the case for the Brazilian box F1 fermentation, wherein AAB dominated the fermentation and produced high amounts of acetic acid early into the fermentation (Garcia-Armisen et al., 2010; Papalexandratou et al., 2011). In conclusion, a restricted yeast species composition was found for different cocoa bean fermentations in different cocoa-producing regions, with H. opuntiae being the prevailing species based on the presence of DGGE bands, followed by P. kudriavzevii and S. cerevisiae. These yeast communities could rapidly be monitored by 26S rRNA gene-PCR-DGGE, not only with respect to their identities but also concerning their temporal dynamics. The latter is much more difficult to perform culture-dependently, unless hundreds of isolates are identified, which is labour-intensive and time-consuming (usually several weeks to months, depending on the number of isolates). The culture-independent analysis reported in the present study was performed on samples from different origins, a comparison that was done for the first time and contributed to the novelty of this study, and the results could be obtained within 2 weeks based on frozen samples. Moreover, they confirmed that all yeast species involved in the fermentation process are cultivable. The CBS Scientific DGGE system appeared to be more reliable to monitor the yeast community dynamics and species composition than the DCode apparatus, as a better band separation occurred for PCR amplicons with a high GC content. The detection of a few prevalent yeast species in all fermentations worldwide will be useful for the development of an ideal yeast starter culture in combination with a bacterial cocktail to allow the control of cocoa bean fermentation processes in the near future. Acknowledgements This research was funded by the Research Council of the Vrije Universiteit Brussel, the Federal Research Policy (Contract C3/00/17), the Fund for Scientific Research-Flanders, the Flemish Institute for the Encouragement of Scientific and Technological Research in the Industry and Barry Callebaut N.V. In particular, we acknowledge the help of Barry Callebaut Belgium (Nicholas Camu and Herwig Bernaert). The cooperation of the local farmers of the ‘Hawaii’ and ‘Leao De Ouro’ plantations (Rodovia Ilhéus/Uruçuca, Bahia, Brazil) is highly appreciated. We thank the Centre National de Recherche Agronomique (Abidjan, Ivory Coast) for their cooperation. The experiments in Ecuador were carried out in collaboration with the Estación Experimental Tropical Pichilingue, Instituto Nacional Autónomo de Investigaciones Agropecuarias (INIAP) (Quevedo, Los Rios, Ecuador). The help of the family Ong for the accomplishment of the Malaysian fermentations (Triang, Pahang) is highly appreciated. Frédéric Ravyts is thanked for the help to optimize the DGGE gradients. References Afoakwa EO Paterson A Fowler M Ryan A ( 2008) Flavor formation and character in cocoa and chocolate: a critical review. Crit Rev Food Sci Nutr  48: 840– 857. Google Scholar CrossRef Search ADS PubMed  Aldsworth T Dodd CER Waites W ( 2009) Food microbiology. Food Science and Technology  ( Campbell-Platt G, ed.), pp. 115– 174. John Wiley & Sons Ltd, Chichester. Ardhana MM Fleet G ( 2003) The microbial ecology of cocoa bean fermentations in Indonesia. Int J Food Microbiol  86: 87– 99. Google Scholar CrossRef Search ADS PubMed  Ascher J Ceccherini MT Chroňáková A Jirout J Borgogni F Elhottová D Šimek M Pietramellara G ( 2010) Evaluation of the denaturing gradient gel electrophoresis apparatus as a parameter influencing soil microbial community fingerprinting. World J Microbiol Biotechnol  26: 1721– 1726. Google Scholar CrossRef Search ADS   Beckett ST ( 2009) Industrial Chocolate Manufacture and Use , 4th edn. John Wiley & Sons, Ltd, Chichester. Beh AL Fleet GH Prakitchaiwattana C Heard GM ( 2006) Evaluation of molecular methods for the analysis of yeasts in foods and beverages. Advances in Mycology  ( Hocking AD Pitt JI Samson RA Thrane U, eds), pp. 69– 106. Springer, New York. Cadez N Poot GA Raspor P Smith MT ( 2003) Hanseniaspora meyeri sp. nov., Hanseniaspora clermontiae sp. nov. and Hanseniaspora opuntiae sp. nov., novel apiculate yeast species. Int J Syst Evol Microbiol  53: 1671– 1680. Google Scholar CrossRef Search ADS PubMed  Cadez N Raspor P Smith MT ( 2006) Phylogenetic placement of Hanseniaspora–Kloeckera species using multigene sequence analysis with taxonomic implications: descriptions of Hanseniaspora pseudoguilliermondii sp. nov. and Hanseniaspora occidentalis var. critica var. nov. Int J Syst Evol Microbiol  56: 1157– 1165. Google Scholar CrossRef Search ADS PubMed  Camu N De Winter T Verbrugghe K Cleenwerck I Vandamme P Takrama JS Vancanneyt M De Vuyst L ( 2007) Dynamics and biodiversity of populations of lactic acid bacteria and acetic acid bacteria involved in spontaneous heap fermentations of cocoa beans in Ghana. Appl Environ Microbiol  73: 1809– 1824. Google Scholar CrossRef Search ADS PubMed  Camu N De Winter T Addo SK Takrama JS Bernaert H De Vuyst L ( 2008a) Fermentation of cocoa beans: influence of microbial activities and polyphenol concentrations on the flavour of chocolate. J Sci Food Agric  88: 2288– 2297. Google Scholar CrossRef Search ADS   Camu N González Á De Winter T Van Schoor A De Bruyne K Vandamme P Takrama JS Addo SK De Vuyst L ( 2008b) Influence of turning and environmental contamination on the dynamics of populations of lactic acid and acetic acid bacteria involved in spontaneous cocoa bean heap fermentation in Ghana. Appl Environ Microbiol  74: 86– 98. Google Scholar CrossRef Search ADS   Carr JG Davies PA Dougan J ( 1979) Cocoa Fermentation in Ghana and Malaysia I . Natural Resources Institute, Chatham/London. Cocolin L Bisson LF Mills DA ( 2000) Direct profiling of the yeast dynamics in wine fermentations. FEMS Microbiol Lett  189: 81– 87. Google Scholar CrossRef Search ADS PubMed  Daniel HM Vrancken G Takrama JF Camu N De Vos P De Vuyst L ( 2009) Yeast diversity of Ghanaian cocoa bean heap fermentations. FEMS Yeast Res  9: 774– 783. Google Scholar CrossRef Search ADS PubMed  De Vuyst L Lefeber T Papalexandratou Z Camu N ( 2010) The functional role of lactic acid bacteria in cocoa bean fermentation. Biotechnology of Lactic Acid Bacteria: Novel Applications  ( Mozzi F Raya RR Vignolo GM, eds), pp. 301– 326. Wiley-Blackwell, Ames, IA. Ercolini D ( 2004) PCR-DGGE fingerprinting: novel strategies for detection of microbes in food. J Microbiol Meth  56: 297– 314. Google Scholar CrossRef Search ADS   Garcia-Armisen T Papalexandratou Z Hendryckx H Camu N Vrancken G De Vuyst L Cornelis P ( 2010) Diversity of the total bacterial community associated with Ghanaian and Brazilian cocoa bean fermentation samples as revealed by a 16 S rRNA gene clone library. Appl Microbiol Biotechnol  87: 2281– 2292. Google Scholar CrossRef Search ADS PubMed  Garofalo C Silvestri G Aquilanti L Clementi F ( 2008) PCR-DGGE analysis of lactic acid bacteria and yeast dynamics during the production processes of three varieties of Panettone. J Appl Microbiol  105: 243– 254. Google Scholar CrossRef Search ADS PubMed  Gevers D Huys G Swings J ( 2001) Applicability of rep-PCR fingerprinting for differentiation of Lactobacillus species. FEMS Microbiol Lett  205: 31– 36. Google Scholar CrossRef Search ADS PubMed  Giraffa G ( 2004) Studying the dynamics of microbial populations during food fermentation – a review. FEMS Microbiol Rev  28: 251– 260. Google Scholar CrossRef Search ADS PubMed  Heuer H Krsek M Baker P Smalla K Wellington EMH ( 1997) Analysis of actinomycete communities by specific amplification of genes encoding 16S rRNA and gel-electrophoretic separation in denaturing gradients. Appl Environ Microbiol  63: 3233– 3241. Google Scholar PubMed  Janse I Bok J Zwart G ( 2004) A simple remedy against artifactual double bands in denaturing gradient gel electrophoresis. J Microbiol Meth  57: 279– 281. Google Scholar CrossRef Search ADS   Jespersen L Nielsen DS Hønholt S Jakobsen M ( 2005) Occurrence and diversity of yeasts involved in fermentation of West African cocoa beans. FEMS Yeast Res  5: 441– 453. Google Scholar CrossRef Search ADS PubMed  Kim DH Hong YA Park HD ( 2008) Co-fermentation of grape must by Issatchenkia orientalis and Saccharomyces cerevisiae reduces the malic acid content in wine. Biotechnol Lett  30: 1633– 1638. Google Scholar CrossRef Search ADS PubMed  Lagunes-Gálvez S Loiseau G Paredes JL Barel M Guiraud JP ( 2007) Study on the microflora and biochemistry of cocoa fermentation in the Dominican Republic. Int J Food Microbiol  114: 124– 130. Google Scholar CrossRef Search ADS PubMed  Leal GA Gomes LH Efraim P Tavares FC Figueira A ( 2008) Fermentation of cocoa (Theobroma cacao L.) seeds with a hybrid Kluyveromyces marxianus strain improved product quality attributes. FEMS Yeast Res  8: 788– 798. Google Scholar CrossRef Search ADS PubMed  Masoud W Cesar LB Jespersen L Jakobsen M ( 2004) Yeast involved in fermentation of Coffea arabica in East Africa determined by genotyping and by direct denaturing gradient gel electrophoresis. Yeast  21: 549– 556. Google Scholar CrossRef Search ADS PubMed  Meroth CB Walter J Hertel C ( 2003) Identification and population dynamics of yeasts in sourdough fermentation processes by PCR-denaturing gradient gel electrophoresis. Appl Environ Microbiol  69: 7453– 7461. Google Scholar CrossRef Search ADS PubMed  Mills DA Johannsen EA Cocolin L ( 2002) Yeast diversity and persistence in Botrytis-affected wine fermentations. Appl Environ Microbiol  68: 4884– 4893. Google Scholar CrossRef Search ADS PubMed  Nielsen DS Hønholt S Tano-Debrah K Jespersen L ( 2005) Yeast populations associated with Ghanaian cocoa fermentations analysed using denaturing gradient gel electrophoresis (DGGE). Yeast  22: 271– 284. Google Scholar CrossRef Search ADS PubMed  Nielsen DS Teniola OD Ban-Koffi L Owusu M Andersson TS Holzapfel WH ( 2007) The microbiology of Ghanaian cocoa fermentations analysed using culture-dependent and culture-independent methods. Int J Food Microbiol  114: 168– 186. Google Scholar CrossRef Search ADS PubMed  Nielsen DS Jakobsen M Jespersen L ( 2010) Candida halmiae sp. nov., Geotrichum ghanense sp. nov. and Candida awuaii sp. nov., isolated from Ghanaian cocoa fermentations. Int J Syst Evol Microbiol  60: 1450– 1465. Google Scholar CrossRef Search ADS PubMed  Okuma Y Endo A Iwasaki H Ito Y Goto S ( 1986) Isolation and properties of ethanol-using yeasts with acid and ethanol tolerance. J Ferment Technol  64: 379– 382. Google Scholar CrossRef Search ADS   Ongol MP Asano K ( 2009) Main microorganisms involved in the fermentation of Ugandan gee. Int J Food Microbiol  133: 286– 291. Google Scholar CrossRef Search ADS PubMed  Papalexandratou Z Camu N Falony G De Vuyst L ( 2011) Comparison of the bacterial species diversity of spontaneous cocoa bean fermentations carried out at selected farms in Ivory Coast and Brazil. Food Microbiol  28: 964– 973. Google Scholar CrossRef Search ADS PubMed  Prakitchaiwattana CJ Fleet GH Heard GM ( 2004) Application and evaluation of denaturing gradient gel electrophoresis to analyse the yeast ecology of wine grapes. FEMS Yeast Res  4: 865– 877. Google Scholar CrossRef Search ADS PubMed  Rettedal EA Clay S Brözel VS ( 2010) GC-clamp primer batches yield 16S rRNA gene amplicon pools with variable GC clamps, affecting denaturing gradient gel electrophoresis profiles. FEMS Microbiol Lett  312: 55– 62. Google Scholar CrossRef Search ADS PubMed  Schwan RF Wheals AE ( 2004) The microbiology of cocoa fermentation and its role in chocolate quality. Crit Rev Food Sci Nutr  44: 205– 221. Google Scholar CrossRef Search ADS PubMed  Schwan RF Rose AH Board RG ( 1995) Microbial fermentation of cocoa beans, with emphasis on enzymatic degradation of the pulp. J Appl Bacteriol (Symp Suppl)  79: 96S– 107S. Stringini M Comitini F Taccari M Ciani M ( 2009) Yeast diversity during tapping and fermentation of palm wine from Cameroon. Food Microbiol  26: 415– 420. Google Scholar CrossRef Search ADS PubMed  Takeuchi A Shimizu-Ibuka A Nishiyama Y Mura K Okada S Tokue C Arai S ( 2006) Purification and characterization of an α-amylase of Pichia burtonii isolated from the traditional starter “murcha” of Nepal. Biosci Biotechnol Biochem  70: 3019– 3024. Google Scholar CrossRef Search ADS PubMed  Thompson SS Miller KB Lopez AS ( 2007) Cocoa and coffee. Food Microbiology: Fundamentals and Frontiers , 2nd edn ( Doyle MP Beuchat LR Montville TJ, eds), pp. 837– 849. ASM Press, Washington, DC. Vasilopoulos C Ravyts F De Maere H De Mey E Paelinck H De Vuyst L Leroy F ( 2008) Evaluation of the spoilage lactic acid bacteria in modified-atmosphere-packaged artisan-type cooked ham using culture-dependent and culture-independent approaches. J Appl Microbiol  104: 1341– 1353. Google Scholar CrossRef Search ADS PubMed  Vaughan-Martini A Kurtzman CP Meyer SA O'Neill EB ( 2005) Two new species in the Pichia guilliermondii clade: Pichia caribbica sp. nov., the ascosporic state of Candida fermentati, and Candida carpophila comb. nov. FEMS Yeast Res  5: 463– 469. Google Scholar CrossRef Search ADS PubMed  Wood GAR Lass RA ( 2001) Cocoa , 4th edn. Longman Group Limited, London. © 2011 Federation of European Microbiological Societies. Published by Blackwell Publishing Ltd. All rights reserved
TI - Assessment of the yeast species composition of cocoa bean fermentations in different cocoa-producing regions using denaturing gradient gel electrophoresis
JF - FEMS Yeast Research
DO - 10.1111/j.1567-1364.2011.00747.x
DA - 2011-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/assessment-of-the-yeast-species-composition-of-cocoa-bean-KMJ61Cdo09
SP - 564
EP - 574
VL - 11
IS - 7
DP - DeepDyve
ER -