| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Unité Mixte de Recherche Génie et de Microbiologie des Procédés Alimentaires (UMR G.M.P.A.), F-78 850 Thiverval-Grignon, France
Corresponding author: M.-N. LecClercq-Perlat; e-mail: perlat{at}grignon.inra.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: deacidified cheese medium • Debaryomyces hansenii • Kluyveromyces marxianus • volatile compound composition
Abbreviation key: DCM = deacidified cheese medium, Dh = Debaryomyces hansenii, Km = Kluyveromyces marxianus, YEGC = yeast extract glucose chloramphenicol
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
To better understand the impact of yeasts on cheese flavor, many authors have proposed to use a cheese model medium made with curd and water. Depending on the properties of the different strains present, yeasts may have a positive (increase in pH, etc.) or a negative (unpleasant taste, etc.) effect on ripening (Fleet, 1990; Eliskases-Lechner and Ginzinger, 1995; Tzanetakis et al., 1998; Martin et al., 1999; Petersen et al., 2002; Carreira et al., 2002). The predominant species of yeasts in the cheese rind are Kluyveromyces lactis or K. marxianus and Debaryomyces hansenii (Eliskases-Lechner and Ginzinger, 1995; Welthagen and Viljoen, 1999). Kluyveromyces spp. produces large amounts of alcohols, aldehydes, esters, and terpenes when cultured alone in cheese slurries (Martin et al., 2001). Ethyl acetate synthesis by K. fragilis is catalyzed by both an esterase (a constitutive enzyme) and alcohol-acetyl transferase (an inducible one) (Kallel-Mhiri and Miclo, 1993). Kluyveromyces lactis produced a variety of esters, ethyl acetate being the major one (Arfi et al., 2002). The production of ethyl and butyl acetates during ripening of a Camembert cheese followed the growth of this yeast (Leclercq-Perlat et al., 2003b). From a D. hansenii strain, an esterase, which was able to hydrolyze the esters, was isolated (Besancon et al., 1995). It may be a reason why this yeast cannot produce large quantities of esters. To our knowledge, the production of aroma from D. hansenii has not been analyzed in detail in isolation from interactions with some other cheese microorganisms. In their former study (Leclercq-Perlat et al., 2003a), the color of Brevibacterium linens biofilms was found to be dependent on the deacidification yeast used. It was equally observed that the odor of a deacidified cheese medium (DCM) was dependent on the yeast used. Debaryomyces hansenii spawned a typical and strong wine-faded rose odor, while K. marxianus generated an ethereal fruity (apple and pineapple) aroma, which was close to the one characteristic of young Camembert in the ripening chamber. For these reasons, it was decided to focus attention on the volatile compound composition of the initial slurry (reference), the DCMDh or the DCMKm.
In this study of a Munster cheese curd, the production of volatile compounds by pure cultures of 2 yeasts (D. hansenii and K. marxianus) was considered. It was measured by dynamic headspace-gas chromatography-mass spectrometry analysis. This analytical approach involved extraction, separation, identification, and quantification of the extracted volatile compounds. It has been successfully applied to dairy products (Imhof and Bosset, 1991; Engels et al., 1997). However, dynamic headspace conditions differed between studies, depending on the studied food products. Consequently, those conditions, as well as a sample preparation, were first optimized to obtain good sensitivity and repeatability of quantification (Berger et al., 1999; Leclercq-Perlat et al., 2003b). Using this technique, a better understanding of the role of yeasts in aroma production on cheese slurries was possible.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Preparation of K. marxianus and D. hansenii Cheese Starters
The starter of each yeast strain was prepared as described by Leclercq-Perlat et al. (2003a).
Cheese Medium Preparation
Muenster cheeses used to prepare the slurries were obtained from the industrial factory, Coopérative Agricole Laitière (BP2; 54450 Blâmont, France). Kept on the first day of cheese making, they were not salted and were not seeded with ripening flora. They were frozen at –80°C to prevent enzymatic reactions and auto-oxidation and kept frozen for less than 6 mo.
After thawing, 240 g of cheese was ground in a mortar and transferred to a 1-L flask containing 160 mL of sterile water (120°C, 20 min). This preparation was sterilized at 100°C for 10 min to kill the lactic bacteria. When its temperature was close to 60°C, it was homogenized with an Ultra-Turrax T25 grinder (Janke & Kunkel, IKA-Labortechnik, Staufel, Germany) (1 min, 24,000 rpm). After cooling down to 30°C, one of the yeast starters (106 cfu/mL) was inoculated. After a second homogenization (Ultra-Turrax, 40 s, 13,500 rpm), 300 mL of this medium was transferred to a 1-L conical flask corked with a cellulose cork that was previously sterilized (120°C, 20 min). Then, this medium (initial pH close to 4.8) was incubated in darkness at 25°C while being shaken (150 rpm) until reaching pH 5.8. Five independent trials were carried out to test the repeatability of this preparation.
Analysis of Volatile Compounds Present in Cheese Suspensions (Reference and DCM)
Gas chromatography and mass spectrometry measurements.
The analysis of volatile compound composition was performed as described by Leclercq-Perlat et al. (2003b). The volatile compounds were extracted with a dynamic headspace analyzer (purge and trap concentrator, model 7695A, Hewlett-Packard, Avondale, PA). Each sample tube was connected to the apparatus and heated at 60°C for 10 min. Then it was purged with helium (99% of purity) at a flow rate of 30 mL/min for 15 min. The volatile compounds were extracted by adsorption onto a porous-polymer-adsorbent Tenax trap column (Teckmar Inc., Cincinnati, OH) at room temperature, which was then heated to 225°C for 2 min to desorb them. The compounds eluted were directly transferred at 150°C to the head of a capillary column where they were cryofocused at –150°C. Water was removed by condensation with the in-line system MCS (Hewlett-Packard) between the Tenax trap and the cryofocusing system.
The condensed compounds were separated by GC (model 6890, Hewlett-Packard) by heating the interface to 180°C for 1 min. They were automatically injected (splitless) into a nonpolar capillary column (HP-5MS; 30m x 0.25 mm and film thickness 0.25µm) at a helium flow rate of 1.6 mL/min. The oven temperature was kept at 4°C for 8 min. Then, temperature increases were programmed, from 5 to 50°C at 3°C/min, from 50 to 100°C at 5°C/min, and to the final temperature of 150°C at 10°C/min. The oven stayed at this temperature for 5 min. The GC column was connected directly to a mass selective detector (HP6890/5972). The specifications of the method used were summarized by Berger et al. (1999). The electron impact energy was set at 70 eV, and data were collected in the range of 29 to 300 atomic mass units at a scan rate of 1.68 scans/s.
Before analysis, each cheese medium sample was diluted to 1/100 with (4°C) cold Milli-Q water (Millipore system) and its pH after dilution was 5.80 ± 0.05. Prior to headspace analysis, this procedure maintained the same conditions of volatile suspension for all the samples (Mariaca and Bosset, 1997).
This analysis was carried out on 3 independent cultureS of the same type. Each culture was analyzed in triplicate by GC/MS. The mean of the abundances obtained from each of the 3 cultures was called "mean abundance."
Selection and quantification of some volatile compounds produced during deacidification.
The volatiles produced in the cheese samples were identified by GC comparison with the one of the pure compounds (retention time and mass spectra). The concentrations of 12 compounds reflected on the main flavor pathways were determined using a calibration curve made with the pure compounds. These compounds were purchased from Aldrich (Sigma-Aldrich chemistry SARL, Saint Quentin Fallavier, France) except isoamyl acetate and ethyl acetate from Fischer Scientific (Elancourt, France) and ethyl acetate from VWR (Fontenay sous Bois, France), all with the highest purity available. Known amounts of these compounds were added to a reference cheese medium at various concentrations (0, 0.25, 0.50, 0.75, 1.0, 2.0, and 5.0 µL/L). For each of these concentrations, triplicate samples of reference cheese medium were prepared and submitted to headspace analysis 3 times. The determination of calibration curves was carried out as described by Leclercq-Perlat et al. (2003b).
This analysis was carried out on 5 reference as well as 5 DCMKm and 5 DCMDh independent suspensions. Each sample was analyzed in triplicate by GC/MS to determine the culture concentration. The mean concentration was the arithmetic average of 5 measurements (one per culture).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
) was used to estimate the influence of volatile compound production by the yeast. Volatile compounds isolated in cheese slurries may be ranked into 3 classes: 1) the ones that appeared in all tested media, 2) the ones produced by K. marxianus, and 3) the ones generated by D. hansenii.
Volatile Compound Changes in DCM
Volatile compounds present whatever the media tested.
Table 1
shows the main volatile compounds, the flavor description from bibliography (Sable and Cottenceau, 1999; Cristiani and Monnet, 2001), the average abundance (A) and the associated variation coefficient for each yeast as well as the ratio and the yeast involved in its production. The main volatile compounds isolated from DCM may be classified into 2 subclasses:
|
Volatile compounds in relation to the deacidification yeast used.
The quantity of the other components appeared negligible in the reference cheese medium. In the case of DCM, the other components were either apparent or negligible according to the yeast used. For the yeast involved in their production, Table 2 shows the volatile compounds, whose quantity changed considerably during deacidification, as well as their flavor description from bibliography (Sable and Cottenceau, 1999; Cristiani and Monnet, 2001), their mean abundance, the calculated variation coefficient, and their threshold. If a volatile component solution to perform the quantification was easily found, the concentration present in the medium was quantified and its value is also shown in table 2.
|
The ones produced by K. marxianus were essentially esters with fruity odor (pineapple or banana). Except for ethyl butanoate and butyl acetate, the ester concentrations were higher than their thresholds. This fact explained that the global odor of this DCM was ester-like and fruity. One compound (ethyl acetate) was produced in large amounts.
The ones generated by D. hansenii were essentially methylketones with fruity, floral (rose), moldy, cheesy, or wine odors plus 2-phenylethanol with rose and wine odors. The methylketone concentrations were much lower than their thresholds. Therefore, their impact on the global flavor of the DCM was low and even negligible. The concentration of 2-phenylethanol was the highest. It explains the global odor of faded roses.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
In DCMKm, the main volatile component was ethyl acetate, which resulted from ethanol and lactose metabolisms. The ethanol was very abundant in the medium after deacidification whatever the yeast used (results not shown). The results were in accordance with those of Arfi et al. (2002) and Kallel-Mhiri et al. (1993). Indeed, in a synthetic medium, Kluyveromyces strains produce a variety of esters, with ethyl acetate being the major one. It was been shown that ethanol, lactose, and O2 were the direct precursors for the ethyl acetate synthesis that was affected by the carbon/nitrogen ratio and iron when K. fragilis was used (Kallel-Mhiri and Miclo, 1993). The ethanol and oxygen concentrations were not limited during K. marxianus growth but the lactose one strongly decreased. A peak of ethyl acetate was produced during the growth of K. lactis. It has been observed in a synthetic medium incubated at 25°C (Arfi et al., 2002) or in a model cheese ripened in industrial conditions (Leclercq-Perlat et al., 2003b). The isoamyl acetate also seemed to be important. It is admitted that this ester is issued from degradation of lactose (formation of Acetyl-coA) and of Leucine through the formation of isoamyl alcohol (Erhlich pathway). Kluyveromyces lactis was more efficient in releasing amino acids but its efficiency varied depending on the strains tested (Klein et al., 2002).
In DCM Dh, the main volatile components were methylketones and 2-phenylethanol. The methylketones were issued from ß-oxidation metabolisms of FFA (Molimard and Spinnler, 1996). Their concentrations were very low. The 2 main factors involved in this production may be: 1) the presence of lactose and lactate as easy assimilation sources of nutriments and 2) the FFA concentration. Lactose and lactate were favorable to the growth of D. hansenii and its growth was exponential (Leclercq-Perlat et al., 2003a). They were the source of lipid catabolism repression. The low concentration of FFA (results not shown) may result from this repression, and it appeared as a decisive factor for the low production of methylketones. The concentration of esters in this DCM was negligible. This result was in accordance with those of Arfi et al. (2002). Ethyl acetate or other ester productions were not detected in synthetic medium cultures of D. hansenii. The main volatile component was 2-phenylethanol, its measured concentration being close to 188 µL/L at 62 ± 4 h. It came from the degradation of phenylalanine through the Erhlich pathway (Lee and Richard, 1984). The phenylalanine could be produced by: 1) lactic acid bacteria and/or 2) the yeast. In fact, the nonproteic nitrogen was not negligible (close to 3.5% TN) (Leclercq-Perlat et al., 2003a). Debaryomyces hansenii was less efficient in releasing amino acids than Kluyveromyces strains (Klein et al., 2002). This could explain that the ester production efficiency observed in DCMKm was higher than the one obtained in DCMDh. This is coincident with the fact that cheese medium deacidified by D. hansenii showed the lowest ester concentrations, suggesting lesser release of peptidases into the cheese mass. According to Klein et al. (2002), peptidase release was related to cell lysis. Nevertheless, we did not investigate yeast peptidase release and cell yeast lysis, so further research on these subjects may be needed.
In conclusion, depending on the yeast cultivated, 2 classes of volatile compounds were found in DCM. Esters were essentially produced by K. marxianus and developed fruity odors. Their concentrations were generally higher than their thresholds, explaining the DCMKm global fruity odor. Methylketones were essentially generated by D. hansenii with a lower impact on the global odor. 2-Phenylethanol also produced by D. hansenii had a higher impact on global odor of this medium, which remained faded roses. In fact, the carbon and nitrogen substrates (Leclercq-Perlat et al., 2003a) and volatile compound analyses permitted the demonstration that DCM properties were dependent on the yeast used. Its aroma compositions were directly related to the deacidification yeast as well.
Although the Muenster curd used was the same, whatever the yeast tested, the deacidification medium was completely different. The influence of the deacidification yeast used on the B. linens biofilm color observed previously (Leclercq-Perlat et al., 2003a) was probably due to the differences of cheese composition observed after deacidification.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication May 21, 2003. Accepted for publication September 25, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. De Freitas, N. Pinon, J.-L. Berdague, P. Tournayre, S. Lortal, and A. Thierry Kluyveromyces lactis but Not Pichia fermentans Used as Adjunct Culture Modifies the Olfactory Profiles of Cantalet Cheese J Dairy Sci, February 1, 2008; 91(2): 531 - 543. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
