Scientific Publications - Work Done by Microbiology Reader Bioscreen C

IDF Symposium on Cheese, Prague, Czech Republic, March 21–25, 2004, Ripening, Characterization & Technology, organized by International Dairy Federation, Prague Institute of Chemical Technology, Czech and Moravian Dairy Association, Czech Committee of IDF

ABSTRACTS

Spray drying in the cheese industry

J. Písecký

Niro A/S, Denmark

The paper is dealing with the application of spray drying in, and in the relation to the cheese industry. In the introduction the technology of producing the cheese powder is briefly described and also the production of the skim milk powder suitable for later production of cheese. Also the technology of whey processing to various dry products is briefly discussed. The main subject of the paper is the description of a new process, called TIXOTHERM™, developed recently by Niro, for the processing of permeate, produced as a byproduct from the ultrafiltration of whey, into a non-hygroscopic powder. After evaporation to 60% TS the permeate concentrate is

subjected to a three step process consisting of concentration to 86% in the Rosinaire™ paddle dryer, holding, stabilization and curing in a screw conveyor with two augers, and finally drying and cooling in a combined back-mix/plug-flow fluid bed. In comparison with the traditional processes the TIXOTHERM™ provides great savings of energy (about 30%) and building costs (up to 75%).

Cheese Authenticity and Traceability: an Analytical Challenge 

L. Pillonel*, J.O. Bosset

Agroscope Liebefeld-Posieux (ALP), Berne, Switzerland

laurent.pillonel [ AT ] alp.admin.ch

Emmental or Swiss cheese is one of the most known cheese variety worldwide. It is produced in many industrialised countries at a very different price. To fight or prevent fraudulent labelling of this cheese type, pertinent tools for geographic origin assignment are needed. For this purpose, cheese samples from the main European manufacturers (n = 15 for Austria, n = 12 for Finland, n = 63 for France, n = 23 for Germany and n = 70 for Switzerland) were collected, corresponding to 110 winter and 73 summer samples.

From a preliminary study, a series of promising analytical methods were applied to these 183 samples to measure the following parameters: total nitrogen, water soluble nitrogen, 12 % TCA soluble nitrogen, pH-value, volatile short chain acids, sodium, chloride, organic acids, microorganisms, trace elements (Cu, Zn, Mg) and stable isotope2 ratios (ä2H, ä13C, ä15N, ä34S). In a first step, analysis of variance (ANOVA) was applied with the the two factors geographic origin and the season of production. Significant differences between the regions of origin were found for all the parameters mentioned above (p = 0.001). The criteria acetate, propionate, capronate, water soluble nitrogen, 12 % TCA soluble nitrogen, pyruvate, obligate heterofermentative Lb. (OHL), Zn and ä13C showed significant

differences according to the season in certain regions. In a second step, the data was treated by multivariate statistical analysis. Various pattern recognition algorithms (discriminant analysis, general linear model (GLM), neural network) were compared to classify the Emmental samples according to their geographic origin whereas the main accent was put on the differentiation of the Swiss Emmental from those manufactured in other countries. The former could be 100% correctly identified combining the following parameters by pairwise comparison in a GLM model: volatile short chain acids, total nitrogen, pH-value, D-lactate, succinate, L-leucin-aminopeptidase, pyruvate, OHL, ä2H, ä13C, ä15N and ä34S.

Keywords: Cheese Authenticity, Emmental, multivariate statistic 5

Species Identification in Cheese Varieties Using Electrophoretic, Chromatographic and PCR Techniques 

H.K. Mayer

BOKU - University of Natural Resources and Applied Life Sciences, Vienna, Austria

helmut.mayer [ AT ] boku.ac.at

In Europe, cheese varieties made from ewes’ and goats’ milk are of considerable economic importance as a result of widespread acceptance of traditional cheeses. However, the substitution of cows’ milk for ewes’, goats’ and buffaloes’ milk is a fraudulent practice in the dairy industry.

In Austria, mixed cheese varieties are produced either from cows’ and ewes’ milk or from cows’ and goats’ milk; in both cases the maximum percentage of cows’ milk allowed is 49%. In addition, some cheeses are manufactured from cows’, ewes’ and goats’ milk

(minimum 10% of each species). Therefore, food analysts are challenged not only by milk species identification (e.g., the qualitative detection of cows’ milk), but also by the need for quantitative determination of cows’, ewes’ and goats’ milk in mixtures.

Using standard mixtures of milk as well as model cheeses (Camembert, Tilsit, Kashkaval) of different ages, quantitative determination of the percentage of cows’ milk in mixed-milk cheeses was performed by isoelectric focusing (IEF) and subsequent densitometric evaluation of ã-caseins (Cn). Because ewes’ and goats’ milk cannot be distinguished by this method, bovine, ovine and caprine milks were differentiated using IEF and cation-exchange high performance liquid chromatography (HPLC) of para-ê-Cn. Species identification by polymerase chain reaction (PCR) using species-specific primers was applied to compare the detection limits of protein- and

DNA-based methods, respectively. The analytical procedures used were appropriate for the qualitative detection of cows’ milk in cheese of different species. However, quantitative results in adulteration control have to be understood as approximate values.

Keywords: species identification, cheese, electrophoresis, HPLC, PCR 6

Effect of the Use of Three Different Lamb Paste Rennets on Lipolysis of the PDO Pecorino Romano Cheese

M. Addis*, G. Piredda, M. Pes, A. Pirisi

Istituto Zootecnico e Caseario per la Sardegna, I-07040 Olmedo, Italia

maraddis [ AT ] tiscali.it

Two homemade and one industrial lamb paste rennets were used to produce of PDO Pecorino Romano cheese. The homemade rennets came from only suckling (A) or suckling and allowed to graze (B) 30 day-old lambs, slaughtered 6 h after suckling. The third rennet (C) was available commercially. The aim of the work was to investigate the enzymatic differences of the rennets and to study their influence on lipolysis in PDO Pecorino Romano cheese.

Chymosin activity was higher in the homemade rennets (88% and 83% of total clotting activity respectively for rennets A and B) than in the commercial rennet (64%). Both rennet A and B also had higher lipolytic activity (5.60 and 6.17 LFU/g of rennet respectively) than did rennet C (1.78 LFU/g of rennet).

Fifteen cheesemaking trials of PDO Pecorino Romano cheese were carried out in a pilot plant, five replicates for each rennet. Cheese produced with rennet B showed greater lipolysis and different amounts of free fatty acids. Total free fatty acids content was significantly different (P < 0.001) in the cheeses (B>A>C). Short chain free fatty acid content showed a similar trend. The level of medium and long chain free fatty acids was very close in A and B cheeses, but was significantly lower (P < 0.001) in C cheese.

The percentage of short chain free fatty acids to total fatty acids was higher in the cheese produced with rennets A and B than in those produced with rennet C (P < 0.01), whereas rennet C had the highest percentage of both medium and long chain free fatty acids

(P < 0.01). Since the lipolytic activity of the rennet strongly influenced the characteristics of the cheese, both the diet of the lambs and the slaughtering conditions should be controlled in order to produce a rennet which can preserve the traditional characteristics of the PDO Pecorino Romano cheese.

Keywords: lamb paste rennet, PDO Pecorino Romano, lipolysis

Characterisation of Sicilian Cheeses by Flavour Analysis Using SPME-GC/MS

M. Ziino, A. Verzera, C. Condurso*, D. Giuffrida, V. Romeo, M. Zappalà

University of Messina, Department of Organic and Biological Chemistry, Italy ccondurso [ AT ] pharma.unime.it

The flavour composition is an important marker for cheese characterisation; the contribution of short and medium free fatty acids is determining; other substances such as alcohols, chetones, aldehydes, etc. are important, too.

Cheeses possess highly characteristic aromas, indicating the presence of various components mainly dependent, as it is known, on the milk used, production technology, etc. Several researchers studied the flavour of cheeses; unfortunately, they used solvent extraction, which needs long time for the sample preparation, gives modification of the components present, leads to artefacts and disagreeing results for the same type of cheese.

The flavour of the samples analysed were extracted by SPME, a new easy extraction technique, and analysed by HRGC/MS; typical chromatograms for each cheese analysed were obtained (Fig.1). The method, here proposed, provided simple procedure for solvent free sample preparation and no sample manipulation is required; it is quite cheap, moreover, it needs only three hours for the flavour extraction and analysis.

The analyses were carried out on typical Sicilian cheeses such as Provola dei Nebrodi (plastic curd cheese) and Pecorino (sheep milk cheese).

Each cheese will be divided in two samples and each sample, in order to prove method repeatability will be analysed three times. The absolute peak area obtained for the same components determined during the three different analyses were tabulated and the RSD was calculated. Fifty components were identified in the cheese analysed (free fatty acids, esters, alcohols, chetones, etc.) by spectral data, linear retention indices, injection of standards and literature data.

The Provola dei Nebrodi samples were collected over one year and, the flavour composition was studied from the manufacturing and up to 90 days of ripening.

Finally, the identified marker, or the typical chromatogram obtained (Fig.1) were utilised to characterise each Sicilian cheese; differences were observed, moreover, during the ripening.

Keywords: Sicilian cheeses, flavour composition, characterisation, SPME-GC/MS Figure 1 – SPME-HRGC /MS chromatogram of a sample of Provola dei Nebrodi

More than Moustrap – Measuring the Sensory Profile of Cheese

D. D. Muir

Charis Innovative Food Services Ltd, Ayr, Scotland

donald.muir [ AT ] charisfoods.co.uk

Natural products that ripen over an extended period like cheese and wine pose a difficult problem to the sensory scientist. At best, a single examination of the cheese will provide a snapshot and does not comment on the time course of the ripening product. To establish a comprehensive and balanced view of the cheese, it is essential to determine the sensory profile at different stages of maturation. There are four essential elements of sensory profiling and these are independent of the product to be assessed:

1. the measuring instrument, i.e. the panel of assessors; 2. the properties to be measured – the sensory vocabulary;

3. the measuring protocol, i.e. the method of measurement;

4. the results, i.e. interpretation of the data in a user-friendly way.

The relative importance of these four pillars of sensory profiling will be illustrated by examples based on cheese. Particular attention will be paid to ensuring the stability of the sensory panel over an extended period of assessment.

Keywords: sensory vocabulary, assessors 9

The Influence of the Fat Phase on Sensory Characteristics in a Cheese Imitation

S. Karlsson1*, J. Alander1, M. Modig1, S. Ekstedt2, B.F. Nilsson3

1Karlshamns AB, Sweden; 2Institute for Food and Biotechnology, Sweden; 3Skånemejerier, Sweden stefan.karlsson [ AT ] karlshamns.se

The type and amount of fat and protein, together with the size distribution of the fat droplets in the milk used in cheese production, will influence the microstructure of the final product. In addition, also the sensory characteristics of the cheese will be changed. Since the sensory characteristics of a food product are important, the control of them is of interest.

In this work the type of vegetable fat used, the protein blend, and the homogenising pressure in the production of the imitation milk used were varied and sensory characteristics evaluated. The different vegetable fats have a similar fatty acid composition but differ in solid fat content as well as in crystallisation behaviour. The factors were varied according to an experimental design while the sensory attributes were analysed by a trained sensory panel. In addition, the cheese microstructure was analysed with different techniques and at different magnifications using light microscopy (LM), confocal light scanning microscopy (CLSM), and transmission electron microscopy (TEM). All analyses were performed after an appropriate ripening.

The results show a correlation between the factors and the sensory characteristics, the most important factors being the type of fat and the protein blend. A careful choice of vegetable fat for cheese production can improve the sensory characteristics of the product. Also the impact of the protein blend on the texture of the cheese can be counteracted by the choice of fat.

Keywords: vegetable cheese, sensory analysis, microstructure, microscopy 10

Defining U.S. Cheddar Cheese Flavor Using an Anchored Sensory Language

M.A. Drake1*, M.D. Keziah1, M.E. Carunchia-Whetstine1, P.D. Gerard2

1Dept. Food Science, Southeast Dairy Foods Research Center, North Carolina State University, USA; 2Experimental Statistics Unit, Mississippi State University, USA

mdrake [ AT ] unity.ncsu.edu

Cheese flavor represents a crucial aspect for cheese acceptability for direct consumption or ingredient applications. Descriptive sensory analysis of cheese flavor using a defined and anchored sensory language provides a powerful tool for research and marketing communication. Recent work in the U.S. has resulted in a defined and anchored descriptive sensory language for Cheddar cheese flavor. The

language has been applied nationally and internationally to define and describe Cheddar cheese flavor. Studies have shown that regional differences in flavor exist in many European cheeses. The objectives of this study were to determine if regional differences exist in flavor of U.S. Cheddar cheeses.

Cheddar cheeses were collected as industrial 18 kg blocks three times per year for two years. Cheeses were received by overnight carrier less than 1 month after production and subsequently aged at 7 °C. Twelve companies representing nineteen industrial locations in four regions of the U.S. participated in the study. Cheeses were aged for 6, 9, 12 and 18 months. At each timepoint, cheese flavor was characterized by descriptive sensory analysis using the defined and anchored Cheddar cheese flavor language. Trained panelists (n=15) evaluated each cheese in duplicate using appropriate descriptive sensory analysis techniques. Data were analyzed using both univariate and multivariate statistical analysis.

Cheddar cheeses were diverse in flavor profile throughout aging (p<0.05). Principal component and cluster analysis revealed some flavor similarities throughout cheese aging for locations in the Northeastern United States. Other regions did not have specific consistent flavor similarities and specific flavor profiles were instead associated with particular companies.

Keywords: Cheddar cheese, cheese flavor, regional flavor, sensory analysis 11

Changes of Sensory Characteristics during Ripening of the Blue Cheese Niva

J. Pokorný*, J. Dostálová, D. Šabata, J. Piaszczynska

Department of Food Chemistry and Analysis, Institute of Chemical Technology, Prague, Czech Republic Jan.Pokorny [ AT ] vscht.cz

Blue cheese samples were prepared from cow’s milk, and were ripened in presence of Penicillium roqueforti mould. A set of 82 samples was tested by sensory analysis using 28 sensory characteristics; both samples of Niva and Niva mixed with butter were examined. The sensory analysis was carried out according to the international standards ISO. Six characteristics of appearance possessed the best reproducible properties, followed by textural data (four characteristics). The texture determined by the sensory analysis correlated with data obtained using Instron apparatus. Direct flavour assessment of Niva (eight descriptors) was more precise than that of cheese samples of Niva ground with butter (ten characteristics). The overall flavour quality and estimation of cheese ripeness were used as the only hedonic descriptors. The optimum flavour and texture quality preceded the full ripeness, which is typical for Czech consumers, also in other brands of imported cheese. Sensory characteristics were found closer to the overall quality than chemical criteria. Relations between sensory characteristics were mostly linear, but sometimes semilogarithmic or double logarithmic, however, close to linear. The multivariate analysis, such as the principal component analysis, canonical analysis, cluster analysis or similar methods, was a good indicator of changes occurring during the ripening, and the results obtained correlated well with the degree of ripeness and the flavour quality.

Keywords: Sensory, ripening, Niva, blue cheese

Evaluation of Cheese Quality and Ripening Characteristic by Spectroscopy and Chemometrics

J. Sørensen

Arla Foods amba, Innovation Centre Nr. Vium, Denmark

john.sorensen [ AT ] arlafoods.com

Quantitative Near Infrared (NIR) and Mid Infrared (MID) spectrometry are established techniques in the dairy industry. Both methods have proved to be excellent in quality control mainly due to precision and to the fact that no or limited sample preparation is necessary, hence the analysis is very simple and fast. These qualities have made NIR and MID superior techniques in in-, on- or at-line monitoring and control.

Compared to NIR and MIR fluorescence spectrometry is a less widespread method, primarily used in research laboratories. But recently the use of this technique has been reported in a number of non-trivial applications in cheese.

As a consequence of the complexity of the spectra, multivariate mathematics (chemometrics) is normally applied in order to extract useful information.

For example a set of spectra can be analysed by Principal Component Analysis (PCA) to find the main phenomena in the set, or the spectra, in combination with one or more quality values, can be used to build predictive Partial Least Squares (PLS) regression models. But also modern multi-way chemometrics, e.g. Parallel Factor Analysis (PARAFAC) has been applied in studies of cheese quality. The concentrations of constituents such as water, protein, fat and carbohydrate can be determined using classical methods for

calibration and validation, but food properties (e.g. textural and flavour) can also be monitored on a continuous basis. Protein breakdown has an obvious role in determining the texture and background flavour in all matured cheese varieties. Proteolysis in cheese is usually described by changes in peptide patterns (analysed e.g. by reversed phase liquid chromatography). It is demonstrated that latent structures in data from peptide analysis can be predicted from spectroscopy data.

Keywords: Infrared, Fluorescence, Chemometrics, Cheese 13

Application of FT-NIR and FT-IR Spectroscopy to Study the Shelf-life of Crescenza Cheese

T.M.P. Cattaneo1*, C. Giardina1, N. Sinelli2, M. Riva2, R. Giangiacomo1

1Istituto Sperimentale Lattiero Caseario, Italy; 2University of Milan, Department of Food Science & Technology, Italy tcattaneo [ AT ] ilclodi.it

Crescenza is a fresh cheese in which structural and chemical modifications can occur during the shelf-life, due to intense enzymatic activity. The severity of these modifications is related to the transport and storage conditions, as determined through measurements of colorimetric and structural variations of the cheese paste, and the development of aromatic compounds influencing sensory evaluation. Crescenza „freshness” is generally associated with low acidity, limited proteolysis, and no bitter taste.

In this study, the FT-NIR and FT-IR spectroscopy was applied in order to evaluate the shelf-life period to maintain Crescenza „freshness”. Two different types of Crescenza cheese, stored at 10 ±1°C for the declared shelf-life period (20 days), were analysed at different times.

Spectral data were collected in two spectral ranges, from 12000 cm-1 to 4400 cm-1, by using a FT-NIR spectrometer with an optic fibre working in diffuse reflectance, and from 4000 cm-1 to 700 cm-1, by using a FT-IR spectrometer with an ATR cell. Measurements were carried out in replicates using the original packets. A qualitative statistical software was used for data processing.

Chemometrics applied to FT-NIR and FT-IR spectroscopy was able to detect the decrease of Crescenza „freshness” and to determine a critical day during the shelf-life period. These results are in agreement with those reported both in previous studies and in physicalchemical and chemical tests, where different analitycal techniques were applied. The principal absorption bands involved were determined. Physical, chemical, and microbiological factors affect the product „freshness” having some negative influence on the peculiarity of cheese from production to consumption. Therefore, the main advantage of using the IR spectroscopic techniques is to rapidly draw a profile of the product related to its total quality.

Keywords: spectroscopy, chemometrics, shelf-life, Crescenza cheese

Prediction of Quality Parameters and Age in Cheddar-Type Cheese by Near Infrared Reflectance Spectroscopy and Multivariate Data Analysis

G. Downey1*, D.J. O’Callaghan2, V. Howard2, T.P. Guinee2, E.M. Sheehan3, C.M. Delahunty3

1TEAGASC, The National Food Centre, Ireland; 2TEAGASC, National Dairy Production Research Centre, Co. Cork, Ireland; 3Department of Nutritional Sciences, University College Cork, Ireland

gdowney [ AT ] nfc.teagasc.ie

Cheese quality assessment includes chemical and sensory testing, procedures unsuitable for routine use. Near infrared (NIR) spectroscopy is a rapid, easy-to-use, and non-destructive analytical technique. Little research on its use for prediction of sensory characteristics or age of cheese has been published.

Cheddar cheese samples were manufactured and aged for 9 months at 4 °C. At time periods of 1, 2, 4, 6 and 9 months, aliquots were subject to sensory and NIR spectroscopic analysis. Calibrations for a number of sensory parameters and age were developed using partial least squares (PLS) regression against spectral data using a number of wavelength ranges and data pre-treatments. Best models used spectral data between 750 and 1098 nm in all cases.

Textural attributes „rubbery” and „crumbly” were modelled well. Rubbery texture (range = 6.4–34.9) was predicted with a standard error of cross-validation (SECV) of 3.8 and a correlation coefficient (R) of 0.76 (6 PLS loadings). Crumbly texture (range 1.6–20.9 units) was predicted best by an 8 loading model (SECV = 2.2 units; R = 0.87).

Age was best predicted using spectral data pre-treated by scatter correction and a 2nd derivative. A SECV of 0.90 months was obtained with a correlation coefficient of 0.95 (8 loadings).

These results suggest a role for NIR spectroscopy in the off-line quality assessment of Cheddar cheese. Keywords: Cheddar cheese; near infrared spectroscopy; chemometrics

Electronic Nose: New Tool in Modelling the Ripening of Danish Blue Cheese

J. Trihaas1*, T. van den Tempel2, P. Væggemose Nielsen1 1BioCentrum-DTU, Technical University of Denmark; 2Chr. Hansen A/S, Denmark

jtr [ AT ] biocentrum.dtu.dk

Objective aroma assessments for monitoring the ripening process of Danish blue cheese are of great interest in the industry. Instrumental techniques and sensory panels are generally characterised by increased capital cost and require skilful personnel. An E-nose system might offer a number of advantages such as less skills required, low capital cost, automated artificial intelligence etc.

Danish blue-veined cheese is a traditional semi-soft cheese produced from heat-treated cow’s milk. Ripening of this type of cheese depends on complex microbial interactions involving the primary lactic starter culture, the secondary starter culture, Penicillium roqueforti, mould contaminants and yeast.

It is well established that during the ripening process of Danish blue cheese the growth of P. roqueforti contributes extensively to the unique flavour and texture of Danish blue cheese. The characteristic flavour and taste of blue cheese are closely connected to the ripening stage and can be directly related to the changes in appearance, to aroma formation due to lipid degradation and oxidation and to texture changes due to proteolysis.

The possibility of implementing an electronic nose system in order to evaluate the ripening stage of Danish blue cheeses (5 to 12 weeks after brining) was investigated and results were considered very promising.

In this experiment electronic nose technology has been applied directly on representative cheese samples in order to get an objective characteristic of the cheese aroma profile at the different stages of ripening. These results were evaluated and compared to sensorial analysis performed by a trained panel and aroma analysis examined by means of sampling by diffusion from the headspace of the cheese and gas chromatography-mass spectrometry (GC-MS) carried out on the same cheeses.

Keywords: Danish blue cheese, ripening, e-nose, GC-MS, sensory 16

Investigation at the Molecular Level of Soft Cheeses Quality and Ripening by Infrared and Fluorescence Spectroscopies and Chemometrics – Relations with Rheology Properties

A. Kulmyrzaev1, É. Dufour1*, Y. Noël2, E.M. Qannari3 , G. Mazerolles2

1ENITA Clermont-Ferrand, France; 2INRA-SRTAL Poligny, France; 3ENITIAA / INRA Nantes, France dufour [ AT ] enitac.fr

Classical techniques such as chemistry, biochemistry and rheology are usually used for the characterisation of cheeses. Recently, new investigations have been successfully attempted by coupling infrared and fluorescence spectroscopies. Chemometric methods such as common components and specific weights analysis (CCSWA) have been recently used to 1) describe in a simple and synthetic manner the overall information collected, 2) extract and use the relevant information related to the structural changes of the studied component, 3) provide tools for a molecular interpretation of the results and 4) investigate the relations between structures and rheology properties. Experimental soft cheeses, i.e., 12 traditional and 12 stabilised, made using typical making schemes were manufactured at the INRA pilot-scale dairy-plant. The gross chemical compositions of the cheeses were made according to an experimental design: the dry matters and the fat/dry matters of the stabilised soft cheeses were fixed to 44% & 48%, and 51% & 55%, respectively, whereas the values for traditional soft cheeses were 38% & 42%, and 46% & 50%, respectively. Cheese physico-chemical parameters were determined. Spectra were recorded directly on samples. The emission spectra of tryptophan residues (305-400 nm) were recorded with the excitation

wavelength set at 290 nm and the excitation spectra of vitamin A (250–350 nm) were recorded with the emission wavelength set at 410 nm. Infrared spectra were recorded between 3000 and 900 cm-1 at a resolution of 4 cm-1. The rheology method used was the uniaxial compression test at constant displacement rate. Cheese samples were analysed before salting and after 30 days of ripening by the different methods mentioned above. CCSWA was applied to the seven data sets.

The results showed that the common component 1 discriminating young and ripened cheeses explained 95%, 91%, 73.6% and 51% of the inertia of 900–1500 cm-1 infrared region, 2800–3000 cm-1 infrared region, 1500-1700 cm-1 infrared region and rheology data, respectively. It was also shown that common component 3 discriminating cheeses as a function of the technology explained 54%, 26.4% and 11% of the inertia of vitamin A spectra, chemical data and rheology data, respectively. The spectral patterns associated with the common components allowed molecular interpretations of the discrimination.

Keywords: soft cheese, fluorescence, infrared, chemometric 17

Characterisation of Cheese Variety and Maturity on the Basis of Proteolysis

C. Coker1*, R. Crawford1, V.L. Crow1, T. Dodds1, T. Fayerman1, S. Gregory1, C. Honoré1, K. Johnston1, H. Singh2,

P.J. Watkinson1, N. White1, L. Creamer1 1Fonterra Research Centre, New Zealand; 2Institute of Food, Nutrition and Human Health, Massey University, New Zealand

christina.coker [ AT ] fonterra.com The main features of an acceptable cheese are desirable texture and flavour in the context of a particular cheese type or variety. These features can be assessed by trained sensory panels and there is interest in using instrumental methods because there can be inter-panel differences in sensory assessment. Both flavour and texture are considered to be strongly influenced by the extent and type of proteolysis that has occurred in the particular cheese under examination. In the past, methods such as the proportion of the protein that is watersoluble, alcohol-soluble or soluble in dilute trichloroacetic acid have been used to examine proteolysis. Alternatives such as polyacrylamide gel electrophoresis in a urea buffer, size-exclusion chromatography or reverse phase high performance liquid chromatography (RPHPLC) have also been used. RP-HPLC methods can readily resolve up to 100 peaks and some peaks contain several different peptides.

Application of multivariate statistical methods enables the differentiation of a group of cheese types into different categories on the basis of the changes in the intensities of the RP-HPLC or SE-HPLC peaks. However, when these techniques are used to analyse a diverse range of New Zealand cheese of one type, maturity trends are evident but there is considerable overlap between cheese from different manufacturers and differentiation on the basis of manufacturer is not possible. The data can be used to develop maturity indices. Keywords: characterisation, maturity, proteolysis

Monitoring of Proteolysis Phenomena Using Antibodies Specifically Directed Against the Enzyme Cleavage Site on its Substrate

D. Dupont*, O. Rolet-Repecaud, D. Senocq

INRA – Unité de Recherches en Technologie et Analyses Laitières – Poligny, France

dupont [ AT ] poligny.inra.fr

Proteolysis is a generic biochemical process that plays a major role in most of the biological activities. Proteolysis can be monitored by direct quantification of the protease activity, using specific substrates that, when hydrolyzed by the enzyme, will produce an absorbance, a fluorescent or a luminescent signal. Other techniques consist in the monitoring of the appearance of some peptides, known as markers of the enzymatic activity. Finally, the hydrolysis of peptide bonds can be directly studied using high technology methodologies, such as atomic force microscopy.

An alternative strategy for monitoring this biochemical process is proposed here. This approach is based on the production of rabbit polyclonal antibodies against synthetic peptides mimicking the cleavage site on the substrate of the enzyme responsible for proteolysis. As long as the molecule’s cleavage site is intact, the antibody will bind to the protein. However, after cleavage of the peptide bond by the protease, the antibody will no longer be able to recognize the substrate.

Hydrolysis of â-casein by plasmin, the main indigenous protease of milk, was chosen as a model for this study. Six synthetic peptides which included or bordered the plasmin-sensitive bonds Lys28-Lys29, Lys48-Ileu49, Lys105-His106 and Lys107-Glu108 were produced and administered to rabbits. Five out of 6 antisera cross-reacted with â-casein and were used in five indirect ELISAs. Hydrolysis of sites Lys28-Lys29, Lys48-Ileu49, Lys105-His106 and Lys107-Glu108 was monitored with these ELISAs throughout the ripening of Swiss-type cheese allowing to determine which sites are preferentially cleaved during cheese maturation.

Keywords: antibody, ELISA, plasmin, proteolysis 19

Applications of High Pressure Processing in Cheese Manufacture and Ripening

A.L. Kelly1*, T.P. Guinee2, T.P. Beresford2

1Department of Food and Nutritional Sciences, University College, Cork, Ireland; 2Dairy Products Research Centre, Moorepark, Fermoy, Ireland

a.kelly [ AT ] ucc.ie

High pressure (HP) processing is a recently-introduced food processing technology, with many applications for food products including fruit juices and shellfish. It may also have applications either in treatment of milk prior to cheesemaking, or in treatment of cheese directly, as reported herein. HP treatment of Cheddar cheese at 50 MPa for 3 d accelerated ripening through increased proteolysis of ás1-casein, particularly for very young (< 1 week-old) cheese; however, the magnitude of such acceleration was small. Manipulation of processing conditions (pressure and duration of treatment) increased the degree of acceleration, but the overall effect remained quite small. HP treatment of immature Mozzarella cheese significantly affected its functional properties, accelerating development of desirable cheese characteristics for applications such as pizza toppings. The magnitude of the effect of HP on Mozzarella cheese was somewhat variable, and not all factors influencing this variability are apparent. Of ripening enzymes, plasmin was highly resistant to HP-induced inactivation in cheese, while chymosin was inactivated at > 400 MPa. HP treatment of milk for cheesemaking has several interesting effects; growth of starter bacteria is accelerated, and cheese yield may be increased through incorporation of HP-denatured whey proteins without increasing curd moisture content. Overall, the effects of HP treatment on cheese itself are variety-dependent, and may be related to changes in physicochemical conditions or proteolysis; the benefit of HP treatment, relative to the cost of equipment, requires careful consideration. HP treatment of milk for cheesemaking may be of more immediate potential applicability, although further aspects of the effects of HP on cheese milk remain to be clarified.

Keywords: high pressure, cheese, cheesemaking, ripening

Rheological and Calcium Equilibrium Changes during Ripening of Cheddar Cheese

J.A. Lucey1*, R. Mishra1, A. Hassan1, M.E. Johnson2

University of Wisconsin-Madison, 1Department of Food Science, 2Wisconsin Center for Dairy Research, USA jalucey [ AT ] facstaff.wisc.edu

Most studies on the ripening of Cheddar cheese have focused on proteolysis and textural properties. It is well known that the Ca concentration influences the textural properties of cheese. We believe that the amount of Ca still within the casein particles of Cheddar cheese, i.e., insoluble Ca, plays a key role in modulating cheese texture and that changes from insoluble to soluble Ca occur during ripening that contribute to textural changes. Standard full-fat Cheddar cheese was manufactured and ripened for up to 9-mo. Rheological properties of cheese were determined using dynamic low-amplitude oscillatory rheology over the temperature range 5 to 80 °C. Changes in the Ca equilibrium were determined using acid-base buffering of cheese homogenates or analysis of cheese serum (juice) expressed by hydraulic pressure. Proportion of insoluble Ca as a % of total Ca decreased from ~70 to 50–60% between d 1 and 9-mo; most changes occurred within the first 4 wk. Storage modulus (G’) of cheese decreased during heating. G’ values at low temperature (e.g. 5 °C) were higher in older cheese (e.g. 9-mo) than young cheese. G’ values at high temperature (e.g. 80 °C) decreased rapidly during ripening. At low temperatures (e.g. 5 °C) the loss tangent values of cheeses were similar during ripening. At temperatures > 40 °C the loss tangent increased to reach a maximum at temperatures ~70 °C in young cheese and lower maximum temperatures were observed with ripening. Maximum loss tangent values increased substantially during the first 3 weeks and then slightly decreased or showed little change. The increased loss tangent behaviour at high temperatures in young cheese coincided with increased melt and flow. These results indicate that Ca equilibrium changes also contribute to the initial changes in cheese texture since they occur concomitantly with well known initial proteolytic events.

Keywords: rheology, calcium, melt, cheese ripening.

Biochemistry of Cheese Flavor Development: New Insights from the Genetics and Physiology of Lactic Acid Bacteria

J.L. Steele1*, J.R. Broadbent2

1University of Wisconsin-Madison, Department of Food Science, USA;

2Utah State University, Logan, Department of Nutrition and Food Sciences, USA

jlsteele [ AT ] wisc.edu Microorganisms and enzymes (endogenous, added, and microbial) in the cheese matrix gradually convert milk carbohydrates, citrate, proteins, and lipids in a manner that ultimately yields the desired final product. Lactic acid bacteria (LAB) were linked to cheese flavor development more than 100 years ago, however the variation and complexity in the microbiota and enzyme content of cheese has

hindered efforts to evaluate the relative contributions of specific microorganisms and enzymes. Although many details regarding the contribution of LAB to flavor must still be elucidated, current knowledge indicates that LAB influence cheese flavor development via several key mechanisms. These include: 1) hydrolysis of proteins into peptides and free amino acids; 2) metabolism of free amino acids; and 3) hydrolysis of milk lipids into free fatty acids and their conversion to esters.

This presentation will include both an overview of the biochemistry of LAB’s contribution to cheese flavor and our recent geneticsbased approach for understanding how Lactobacillus helveticus CNRZ32 influences cheese flavor development. We recently assembled

a draft-quality (4X) genome sequence of L. helveticus CNRZ32. Genomic analysis identified genes encoding numerous additional proteolytic enzymes, including proteinases, endopeptidases, aminopeptidases, dipeptidases, and tripeptidases. Also, numerous genes encoding putative amino acid catabolic pathways were identified. Of particular interest due to their potential role in debittering bacterialripened cheeses are the three additional endopeptidases identified in the draft genomic sequence. The genes encoding these three

endopeptidases, designated pepO3, pepF and pepE2, have been characterized and their role in the hydrolysis of model bitter peptides evaluated. Currently, we are utilizing DNA microarrays to characterize expression of the proteolytic and amino acid catabolic systems of L. helveticus CNRZ32.

Keywords: lactic acid bacteria, genetics, physiology, cheese flavor

Characterisation of a decarboxylase involved in the formation of the potent flavour component 3-methylbutanal

B.A. Smit, W.J.M. Engels*, G. Smit

NIZO Food Research, Department of Flavour, Nutrition and Ingredients, The Netherlands

wim.engels [ AT ] nizo.nl

Flavour is one of the most important attributes of cheese. The breakdown of milk components during ripening gives rise to a series of volatile and non-volatile compounds which contribute to cheese flavour. The enzymatic degradation of amino acids by, starter and nonstarter, micro-organisms is an important step in the production of volatile flavour compounds, such as sulphur compounds and compounds

derived from branched-chain and aromatic amino acids. In our study, various micro organisms, used in cheese making, were analysed for their potential to convert leucine into flavour components, most notably 3-methylbutanal. A large variation between, but also within, species was observed for various enzyme

activities which are involved in leucine conversion. In particular, á-keto acid decarboxylase activity, leading to 3-methylbutanal, was found to be present in only few of the strains tested. It is proposed that this activity is rate controlling in the leucine conversion pathway leading to the flavour compound 3-methylbutanal.

Current research is focussed on the purification and characterisation of á-keto acid decarboxylase enzymes. Based on homology studies, several decarboxylase genes have been identified in L. lactis genomes. Their overexpression, however, generally does not result in enhanced 3-methylbutanal production. A gene coding for a á-keto acid decarboxylating enzyme recently has been identified in L. lactis ssp. cremoris B1157, using a fast method for screening of knock-out mutants. After construction of an overproducing strain high levels of production of 3-methylbutanal were measured.

Detailed knowledge of the á-keto acid decarboxylating enzyme and of other essential enzyme activities needed for flavour formation, is important for selection of new starter cultures. Such new cultures will provide novel possibilities regarding cheese flavour diversification and accelerated cheese ripening.

Keywords: flavour, leucine, decarboxylation, 3-methylbutanal

Pathways for á-ketoglutarate Formation in Lactic Acid Bacteria and Their Role in Amino Acid Catabolism

C. Tanous*, L. Rijnen, E. Chambellon, A. Gori, M. Yvon

INRA, Unité de Biochimie et Structure des Protéines, France

catherine.tanous [ AT ] jouy.inra.fr

In lactic acid bacteria (LAB), the production of á-ketoglutarate (á-KG), required as amino group acceptor in the amino acid transamination, is often a limiting factor for the conversion of amino acids to aroma compounds in cheese. The presence of an á-KG formation

pathway from components present in cheese or milk (citrate, glutamate, pyruvate) appears thus essential for aroma formation from amino acids. Theorically, three main á-KG formation pathways exist in bacteria. The first utilises glutamate and is catalysed by a glutamate dehydrogenase (GDH). The second pathway utilises citrate or pyruvate and glutamate, and necessitates citrate lyase or pyruvate carboxylase and aspartate transaminase. Finally, the third pathway also utilises citrate and requires the action of aconitase (ACN) and isocitrate dehydrogenase (IDH). The utilisation of citrate from the medium also requires the presence of a citrate permease.

The objective of the study was to search for and select LAB strains exhibiting the enzymatic activities necessary for each of these pathways, in order to evaluate their impact on amino acid catabolism.

For the first pathway, we detected a weak NADP-dependent GDH activity in several LAB species, while a NAD-dependent GDH activity was only observed in a few strains of S. thermophilus and L. lactis. L. lactis subsp. diacetylactis possesses all the activities required for the second pathway. Finally none of the tested LAB strains exhibited the activities necessary for the third one. Only those strains exhibiting the activities necessary for the first or the second á-KG production pathway catabolised amino acids in a medium containing citrate and/or glutamate. These results show that functional pathways for á-KG production exist in LAB and the presence of such a pathway is a major criteria for the selection of strains capable of catabolising amino acids. However it remains to verify if the selected strains can produce aroma compounds in cheese.

Keywords: amino acid catabolism, lactic acid bacteria, á-ketoglutarate, aroma compounds 24

Esterases of Lactic Acid Bacteria and Cheese Flavour

R. Holland

Fonterra Palmerston North, Private Bag 11029, Palmerston North, New Zealand

ross.holland [ AT ]  fonterra.com

Esterases of lactic acid bacteria (LAB) are a diverse collection of enzymes grouped by their hydrolytic activity on chromogenic ester substrates. Often this activity bears no relation to the true nature of the enzyme activity and has little or no relevance to the impact of the enzyme on milk fat substrates. There are two ways in which esterases might impact on flavour generation from milk fat substrates in cheese: hydrolysis of milk fat glycerides to release free fatty acids, and synthesis of ester flavour compounds. To explore the potential for esterases of LAB to impact on flavour development during cheese ripening, we have compared esterases from Lactococcus lactis, Streptococcus thermophilus and Lactobacillus rhamnosus in a number of different reaction systems, ranging from in vitro assays to experimental cheese systems.

The hydrolase potential of esterases is the best understood activity of LAB esterases, because this is the marker activity for the enzyme group. Tributyrin esterase (estA) from Lactococcus is most active on monoacylglycerols, and shows some activity on diacylglycerols and tributyrin substrates. The enzyme shows a clear preference for glyceryl esters of short chain fatty acids. An esterase from Streptococcus thermophilus also has peptidase activity and, based on gene sequence data, is identifed as X-prolyldipeptidyl aminopeptidase (pepX). This illustrates the plurality of enzyme activity that can occur with esterases. The major esterase (estAA7) from

Lactobacillus rhamnosus HN001 appears to be yet another type of esterase. The ability of esterases to synthesise short chain ester compounds is less well understood. Ester synthesis in cheese and other dairy systems has long been regarded as a 2-step process in which free fatty acids are released by hydrolysis of milk fat and esterification of free fatty acids with an alcohol then follows. However, we have shown that esterase enzymes of a number of lactic acid bacteria form flavourful esters in a single-step transesterification reaction in which ethanol is the acceptor. Tributyrin esterase from Lactococcus and esterases from Streptococcus produce significant amounts of ethyl esters when glyceride substrates are presented together with ethanol. We are now able to direct cheese flavour development by directing esterase activity in cheese through control of milk fat substrates, the cheese environment, and esterase levels.

Keywords: esterase, fatty acids, esters, LAB

Comparison of Purge and Trap and Solid Phase Micro Extraction Techniques for the study of Volatile Organic Compounds in Three PDO European Cheeses

S. Mallia1, E. Fernández-García2*, H. Schlichtherle-Cerny3, J.O. Bosset3

1Consorzio Ricerca Filiera Lattiero-Casearia, Ragusa, Italy; 2Instituto Nacional de Investigacion y Tecnologia Agraria

y Alimentaria, Spain; 3Agroscope Liebefeld-Posieux (ALP), Switzerland Estrella.Fernandez-Garcia [ AT ] alp.admin.ch

Eighteen loaves of three PDO cheeses (six loaves of each), namely Ragusano (Italy), Manchego (Spain) and Gruyère (Switzerland) were studied for volatile organic compounds (VOC) using Solid Phase Micro Extraction (SPME) and Purge and Trap (P&T) extraction,

coupled to Gas Chromatography-Mass Spectrometry/Flame Ionisation Detection/ Olfactometry (GC-MS/FID/O). SPME was carried out with a three-phase StableFlex fibre (Divinyl benzene/Carboxen/Poli dimethyl siloxane). P&T extraction was performed with a Tekmar 2100TM equipped with a tenax trap.

Gas Chromatography was performed in an HP 5890 GC equipped with a DB-Wax column, with simultaneous FID and MS detection. The same GC apparatus was used for the olfactometric detection, by exchanging the line from the FID to the olfactometric port. In total, 102 and 84 VOC were found by using SPME and P&T respectively. Differences were significant for the abundances of most VOC between the three different cheese varieties using the same extraction method. Differences between the extraction methods were also significant for the abundances of most VOC within the same cheese variety.

SPME showed a better extraction efficiency for volatile fatty acids, aromatic compounds and, in general, for the higher molecular weight (MW) methyl ketones and esters. P & T showed a better extraction efficiency for hydrocarbons, alcohols, diacetyl and, in general, for the lower MW methyl ketones and esters.

Using SPME, Gruyère cheeses revealed higher abundances of 2,6-dimethyl pyrazine, n-alkanals and branched chain acids compared to the two other cheese types. Manchego cheeses showed higher abundances of 2-methyl ketones and aromatic compounds. Ragusano

cheeses showed higher abundances of ethyl esters and linear fatty acids. Using the P&T method, Gruyère cheeses showed higher abundances of alkenes, diacetyl and branched chain alkenols compared to the other cheese varieties. Manchego cheeses revealed higher abundances of alkanes, other alcohols, low MW ketones and propyl esters. Ragusano cheeses again showed higher abundances of ethyl esters, butyl esters and linear fatty acids.

By GC-O the three cheese varieties studied revealed potent aroma compounds, such as diacetyl (buttery odour), esters (fruity notes), methional (potato-like odour), butyric acid (cheesy odour), and 3-methyl-1-butyric acid (cheesy odour).

Keywords: Purge and Trap, Solid Phase Micro Extraction, Volatile Organic Compounds, PDO European Cheese 26

Cultures for the Ripening of Smear Cheeses

W. Bockelmann

Bundesanstalt für Milchforschung, Institut für Mikrobiologie, Germany

bockelmann [ AT ] bafm.de

During the past five years several projects at the Federal Dairy Research Centre, Kiel, have been dedicated to the development of surface cultures for smear cheeses. The typical surface microflora of semi-soft cheeses (Tilsit-type), soft cheeses (Limburg-, Romadour-type) and acid curd cheeses (Harzer Roller-type) was determined. The minimal composition of the surface starter cultures which allows rapid smear development was determined for these varieties.

Cultures for semi-soft cheeses should contain Debaryomyces hansenii, Staphylococcus equorum, Corynebacterium casei, Microbacterium gubbeenense (or Arthrobacter nicotianae), and Brevibacterium linens. Yeasts and staphylococci can be inoculated into the cheese brines at a level of 104–106 cfu/ml instead of smearing them on the cheese surface. The smear should contain the other bacteria at a concentration of at least 107 cfu/ml for each species which allows fast deacidification of the cheese surface and smear development. After seven days, the surface pH should be >pH 7 and the smear >109 cfu/cm2. Yeasts and staphylococci reach maximum numbers in the first week, Microbacterium or Arthrobacter in the second week. The percentage of B. linens in the total cell counts is usually low, with a maximum reached after 2–4 weeks of ripening. Highest cell counts in all stages of ripening are always observed for corynebacteria. Surface cultures for soft cheeses should contain Geotrichum candidum which is responsible for the typical white streaks and patches on the orange coloured cheeses and for typical aroma development. G. candidum should be inoculated into the cheese milk at 102 cfu/ml. The presence of Corynebacterium casei is not essential for typical soft cheese ripening. The use of S. equorum, not regularly found on the surface of commercial soft cheeses, accelerates deacidification and smear development without detectable influence on aroma development. A high percentage or Microbacterium or Arthrobacter is usually found (near 100% for some cheeses) with no or low

counts of B. linens. For soft cheeses, it can be assumed that interactions of Microbacterium or Arthrobacter with G. candidum are responsible for typical aroma development.

The ripening of acid-curd cheeses is completely different because they are produced from quarg (dry mass >30%). Two yeasts, Kluyveromyces marxianus and Candida krusei, should be inoculated into the milk used for quarg production. The quarg should be stored for at least five days at about 15°C to allow the development of the yeast flora. Staphylococci seem to be essential for cheese ripening. S. equorum can be added to the quarg with ripening salts to replace the non food-grade S. saprophyticus which is always present on acidcurd cheeses. Suitable corynebacteria for spraying of cheeses are Brevibacterium linens and Corynebacterium variabile.

Smear cheese ripening with defined surface cultures without the typical old-young-smearing was used successfully many times on laboratory scale. At present, defined cultures have been tested on a pilot scale. The introduction of defined cultures could help to reduce the otherwise inevitable levels of contaminating enterobacteria, enterococci, saprophytic staphylococci, moulds and pathogens. Keywords: food safety, smear cheeses, surface cultures, Corynebacterium, Brevibacterium, Microbacterium, Arthrobacter

J.L.W. Rademaker1, L. Rijnen2*, M. Peinhopf1, W. Noordman2, G. Smit2 NIZO Food Research, 1Department of Processing, Quality & Safety; 2Department of Flavour,

Nutrition and Ingredients, Ede, The Netherlands liesbeth.rijnen [ AT ] nizo.nl

The surface microflora of smear cheeses strongly influences the appearance, flavour and texture, as well as the ripening time of the cheese. The composition of the microflora of three Tilsit- type smear-ripened cheeses was studied after one, two, four and eight weeks of ripening. Cheese was treated with an undefined smear starter mix or with defined smear starter mixes. Defined mixes comprised strains of Microbacterium barkeri, Corynebacterium casei, Corynebacterium variabilis, Staphylococcus sciuri, Staphylococcus equorum and Brevibacterium linens.

The microflora on the cheese surface at each time point was analyzed using Terminal Restriction Fragment Length Polymorphism (TRFLP) analysis of the partial small subunit ribosomal gene.

Starter strains were present throughout ripening. Additional strains, apparently originating from the environment, were found. Maximal diversity was observed between weeks two and four, followed by a decrease towards the end of ripening. Most of the strains

reached a maximum numbers in week four and had almost disappeared after eight weeks. However, the Corynebacterium species tended to increase in relative numbers rapidly towards the end and were the dominant bacterial genus on the surface of the fully ripened cheese. Observations were in agreement with those from culture-based studies.

T-RFLP-analysis was found to be an excellent tool for the rapid and specific assessment of population composition and dynamics that facilitates the reconstruction of starter mixes.

Keywords: smear cultures, T-RFLP analysis, population dynamics, cheese

 H.E. Spinnler1*, F. Irlinger2, M.-N. Leclercq-Perlat2, P. Bonnarme2

1Institut National Agronomique-Paris Grignon; 2 INRA - LGMPA Thiverval-Grignon, France spinnler [ AT ] grignon.inra.fr

The microbial mixtures found in soft cheeses are particularly complex. Their addition to milk or to the rind washing solutions should lead to the generation of various properties, such as the limitation of undesirable flora, texture, colour and flavour generation.

To understand the generation of these desired functionalities, it seemed reasonable to identify the biological activities used, by the different species (strains) present on the cheese surface, to produce the desired functions. This screening step can be done on a model curd medium quite efficiently for the yeasts and for the bacteria.

There is at the moment no method other than the empirical expertise for the formulation of the microbial mixtures able to produce the desired functionalities. Which rule could help to combine them? Some examples, such as the combination of Geotrichum candidum with Penicillium camemberti were quite encouraging, because the various typicalities of the pure cultures lead to mixed cultures on curd with the same variety of typicalities and finally the same was observe in cheeses. However, this approach always failed with bacteria. In associations with different yeasts, the weigh of the yeast flavour was much higher than the one of bacteria. The colour of the bacterial strain was dependant on the yeast used for curd deacidification.

In order to amplify the activity of bacterial strains having having rather strong flavouring properties, they were associated with bacteria having efficient hydrolytic properties. The flavour differences observed in pure cultures were not always observed in mixed cultures on the same biologically deacidified curd medium. Consequently, the success of an approach by addition of flora having specific functions seems to be quite random.

Consequently we changed of approach using the combination of a larger number of bacteria and tried to find subgroup of bacteria able to produce the major properties research. The use of molecular methods (PCR-SSCP) permitted to follow the growth of the main species growing among the whole mixture. The first results of this approach will be discussed.

Keywords: flavour, colour, ecosystems, microbial formulation 29

S. Das1*, R.J. Bennett1, V.L. Crow2, R. Holland2, G.J. Manderson3

1Institute of Food, Nutrition and Human Health and 3 Institute of Technology and Engineering, Massey University, Palmerston North, New Zealand; 2Fonterra Research Centre, Palmerston North, New Zealand

S.Das [ AT ] massey.ac.nz

The objective of this study was to manufacture a cheese with novel flavour and with an elevated level of conjugated linoleic acid (CLA) for enhanced nutrition. Our earlier experiments with Cheddar demonstrated that yeast lipases were active in the cheese and selectively released unsaturated long-chain fatty acids (FFA) from milk fat. Also, three strains of propionibacteria were shown in our laboratory to convert free linoleic acid to CLA. By using yeast lipase and propionibacteria together, we attempted to generate CLA in the ripening cheese.

Two yeasts, one strain of Geotrichum candidum and one strain of Yarrowia lipolytica, were used as anaerobic adjuncts in a dry-salted, washed-curd cheese. The yeasts were grown under optimum conditions for lipase production. Two non-starter lactic acid bacteria (NSLAB) were used as adjuncts: Lactobacillus fermentum to produce ethanol and thus promote ester formation to mask the undesirable flavours of fatty acids, and a strain of L. rhamnosus to control the adventitious NSLAB flora. Three strains of Propionibacterium were added to three different 375L cheese vats along with starter and the other adjuncts. The resulting blocks of cheese were matured at 20 °C. The proportions of long-chain unsaturated FFA’s (oleic, linoleic and CLA) were higher in the yeast cheeses throughout ripening than in cheese with no added yeast. This indicates that the yeast lipase(s) selectively hydrolysed unsaturated long-chain fatty acids from the milk fat. Higher concentrations of ethyl esters were detected in yeast cheeses. The proportion of total CLA was the same in both control and experimental cheese over the four months ripening period. We conclude that free linoleic acid was produced by the yeast lipase(s) but there was no conversion to CLA by the isomerase of the propionibacteria, probably due to unfavourable environmental conditions in the cheese.

Keywords: Adjunct, CLA, Lipase, Yeast

B.V. Thage1*, M.L. Broe2, M.H. Petersen1, M.A. Petersen2, M. Bennedsen1, Y. Ardö2

1Chr. Hansen A/S, Hørsholm, Denmark; 2The Centre for Advanced Food Studies, Department of Dairy and Food Science, The Royal Veterinary and Agricultural University, Frederiksberg C, Denmark

BirgitteVedel.Thage [ AT ] dk.chr-hansen.com

Aminotransferase (AT) activity is the first step in the formation of aroma compounds from amino acids in cheese mediated by lactic acid bacteria. The AT activity and specificity for different amino acids are strain dependent for Lactococcus, as well as Lactobacillus. The influence on flavour formation in cheese of three Lb. paracasei strains (CHCC 2115, 4256, and 5583) with different AT profiles was investigated in cheesemaking trails. All strains had AT activity on Asp and Met, but especially on branched-chain amino acids (BcAA); CHCC 2115 had 5 to 10 times higher BcAA activity than the other strains, and CHCC 5583 had activity mainly on Leu. All strains had activity on aromatic amino acids (ArAA), but the strain with very high BcAA-AT activity did not have higher activity than the other strains against ArAA. Reduced-fat, semi-hard, round-eyed cheese was made from cows’ milk using mesophilic starter. Three cheese vats inoculated with starter and a Lactobacillus strain, and a control cheese vat inoculated with starter only (4 vats) were made during one day and repeated three times.

The cheese manufacturing procedure was not influenced by the Lactobacillus adjunct and similar cheese was made from all vats. Unfortunately, clostridia spores were present and developed slightly in all cheeses but not with higher impact than significant differences in flavour profiles were detected between cheeses made with different Lactobacillus strains. In all cheeses made with CHCC 4256, significantly more flavour compounds were produced from BcAA, and they were also preferred for tasting aromatic and sweet with no bitterness compared to the other cheeses. Cheeses with CHCC 4256 also contained considerably more diacetyl and acetoin than the others, and especially the control cheeses were low in those compounds. Diacetyl and acetoin might have been produced from Asp (and Asn).

Keywords: cheese, Lactobacillus paracasei, Aminotransferases, aroma formation

D. Rouch, M.L. Britz*

The University of Melbourne, Department of Agriculture and Food Systems, Victoria, Australia

m.britz [ AT ] landfood.unimelb.edu.au

Non-starter lactic acid bacteria (NSLAB) occur in Cheddar cheese and contribute to the development of flavour during extended periods of maturation. This paper describes some of the physiological and metabolic traits of two Lb. casei strains that occurred either transiently or more persistently in samples taken from Australian cheddar cheeses over 63 weeks, with emphasis on metabolites formed under starvation conditions.

The two strains studied were considered to be related biotypes on the basis of genetic analysis, including pulse-field gel electrophoresis patterns but could be differentiated on the basis of carbohydrate metabolism profiles (ability to utilise tagatose), survival in low pH environments and protein profiles. In a minimal medium based on „cheese juice”, both strains showed requirements for eight amino acids and several vitamins. This medium was used to determine the metabolites formed under different conditions of nutrient supplementation for periods of up to 60 days. Both strains failed to grow in the minimal medium containing all growth requirements if a carbohydrate source was not provided. Trials were conducted on cells in minimal medium containing tryptone plus additional specific growth

requirements in the presence of either ample (1%) or growth-limiting (0.2%) levels of lactose. In some trials, specific amino acids were provided at higher concentrations to determine their impact on metabolite formation. Under starvation conditions (0.2% lactose), peptide consumption occurred as determined from changes in HPLC fingerprints, whereas peptide profiles for cells incubated in the presence of higher concentrations of lactose did not change. The range of metabolites seen during lactose starvation also differed and this was particularly noticeable when the medium was supplemented with high concentrations of specific amino acids. The amount of new end products made decreased when lower concentrations of amino acids were supplied, indicating a direct link between the products and the presence of specific amino acids.

Keywords: Lb. casei, starvation, amino acids

Exploiting Potentials of Bacterial Ecosystems in Cheese

C. Peláez*, T. Requena

Department of Dairy Science and Technology, Instituto del Frío (CSIC), Madrid, Spain

cpelaez [ AT ] if.csic.es

Cheese ripening is a complex process that is largely influenced by the microorganisms present in the cheese matrix. The diversity of cheese microbiota contributes greatly to the complexity of the cheese ripening process, which is of crucial importance for development of the unique organoleptic characteristics of each traditional cheese variety. Interactions between microorganisms, such as microbial competition, microbial cooperation and antagonistic effects, are of prime importance for the survival, growth and activity of microorganisms in cheese. Bacteriocins are antimicrobials produced by lactic acid bacteria that enable them to dominate the cheese ecosystem. Their potential as biopreservatives in cheese is well known. On the other hand, much less attention has been paid to the possible use of some bacteriocins to promote lysis of other lactic acid bacteria and hence contribute to cheese proteolysis and flavour development. This presentation will review some of the strategies proposed to achieve control of the microbiota present in the cheese ecosystem as a means of ensuring high quality and safety standards in the production of cheeses. In particular, strategies such as selection of defined starter/adjunct systems that include strains with interesting, enhanced or complementary enzymatic activities, and the use of bacteriocins to modulate the rate of bacterial lysis during cheese ripening will be discussed.

Keywords: microorganism, ecosystem, cheese ripening 33

Fermentation of Carbohydrates from Cheese Sources by Non-Starter Lactic Acid Bacteria isolated from Semi-Hard Danish Cheese

K. Adamberg1,2*, M. Antonsson1, F.K. Vogensen1, E.W. Nielsen1, P.L. Møller1, S. Kask1,2, Y. Ardö1

1 The Centre for Advanced Food Studies, Department of Dairy and Food Science, The Royal Veterinary and Agricultural University, Frederiksberg C, Denmark; 2 Tallinn Technical University, Department of Food Processing, Tallinn, Estonia

kaarel [ AT ] kbfi.ee

Carbohydrate fermentation by non-starter lactic acid bacteria (NSLAB) isolated from Danish semi-hard cheese was studied using

BioScreen C equipment. Of 45 isolates studied, 39 were identified as Lb. paracasei/casei, 2 as Lb. curvatus and 4 as Lb. danicus using ITS-PCR. Carbohydrate fermentation patterns were analysed by API-50CH. A specially designed carbohydrate restricted media (CRM) added individual carbohydrates (lactose, glucose, galactose, mannose, ribose, N-acetylglucosamine, N-acetylgalactosamine, N-acetylneuraminic acid and N-acetylmuramic acid) was used. Carbohydrates were selected as potential sources that could be found in cheese

(free sugars, sugars from milk-fat globule membrane (MFGM), glycomacropeptide (GMP) or lysed cells). Results showed that Lb. paracasei were able to grow on the main milk carbohydrates from MFGM and GMP as well as from

peptidoglycan of bacterial cell walls but ribose was not a preferred carbohydrate. The highest growth rate was observed on Nacetylglucosamine (0.32–0.56 h-1) and lowest on ribose (0.12–0.23 h-1) if fermented. Nearly half of the ribose-positive strains according

to API-50CH were not able to use ribose in BioScreen tests. These strains grew also slowly on CRM-ribose agar plates (pH 5.8, 2.5 % NaCl) incubated anaerobically at 12 °C. Half of the Lb. paracasei/casei strains were able to use N-acetylgalactosamine and 30 % Nacetylmuramic acid. None of the tested strains fermented N-acetylneuraminic acid.

On the contrary, Lb. danicus indicated better growth on carbohydrates from dead cells than from most of the milk carbohydrates of MFGM or GMP in cheese. Lb. danicus strains grew at 30 °C only in ribose CRM, but at a lower temperature (24 °C) also in CRM with N-acetylglucosamine, glucose, lactose or galactose and the fastest growth was observed on ribose (0.23–0.27 h-1).

Keywords: non-starter lactic acid bacteria, cheese carbohydrate fermentation 34

Antimicrobial Activity of Genetically and Physiologically Characterized Lactobacillus Isolated from Semi-Hard Cheeses

P. Christiansen*, P. L. Møller, S. Kask, M. H. Petersen, M. Petersen, M. Antonsson, K. Adamberg, D. F. Jensen,

E. Waagner Nielsen, F. K. Vogensen, Y. Ardö The Centre for Advanced Food Studies, Department of Dairy and Food Science,

The Royal Veterinary and Agricultural University, Frederiksberg C, Denmark pch [ AT ] kvl.dk

Non-starter Lactobacillus dominate the microbial flora in semi-hard cheese during the main part of ripening and they may affect cheese quality by competing with detrimental bacteria using nutrients or producing antimicrobial substances. In this study, about 1000 isolates from Danish semi-hard cheeses of different varieties, age and origin were analysed by ITS-PCR profiles. Approximately half of the isolates were shown to be Lactobacillus. The Lactobacillus isolates were classified using Pulse Field Gel Electrophoresis (PFGE), resulting in 135 different PFGE profiles or Lactobacillus strains. Sugar fermentation patterns for Lactobacillus isolates with different PFGE profiles were tested by API 50 CHL analysis. Screening for anticlostridial activity of the Lactobacillus strains against 15 isolates of Clostridium from milk, cheese and silage was performed by using agar well diffusion assay. Lytic activities of the Lactobacillus strains were tested against autoclaved Lactococcus in rich MRS-medium and in carbohydrate-restricted medium.

Of 147 Lactobacillus strains screened for their ability to inhibit Clostridium, one third had antimicrobial effect. Nine strains had a broad anticlostridial effect inhibiting 5-11 Clostridium. 17% of the Lactobacillus strains had the ability to lyse autoclaved Lactococcus cells, most of them only in carbohydrate-restricted medium. These strains were typically not the same as those that had anticlostridial activities demonstrating flexibility in competition strategy. The results indicate that anticlostridial non-starter Lactobacillus could be useful as protective cultures against growth of Clostridium during ripening of semi-hard cheeses.

Keywords: Cheese, Lactobacillus, Antimicrobial activity, Clostridium, Lytic activity 35

Diversity of Dominant Microflora in Comté Cheese

F. Berthier, A. Depouilly, F. Dufrene, R. Palme, E. Beuvier*

INRA, Station de Recherches en Technologie et Analyses Laitières, France beuvier [ AT ] poligny.inra.fr

Comté cheese is a variety of hard cooked cheese with an „Appellation d’Origine Protégée” (PDO), manufactured from raw milk and complex starter cultures. Ripened cheeses exhibit different sensorial characteristics according to the dairies and the period at which they are manufactured. The mechanisms generating such a large variety remain largely to be elucidated. Microflora plays a major role in the elaboration of the final characteristics, but how does its diversity look like among cheeses and generate the diversity of final characteristics? Lactic and propionic acid bacteria of the inner part of the cheese represent the most active microflora in Comté cheeses with respect to the elaboration of their final characteristics. About 1600 isolates of lactic and propionic acid bacteria were collected at nine stages of manufacture and ripening (including raw milk and starter cultures) from two Comté cheeses exhibiting different sensorial properties. They were identified at species and strain levels by PCR techniques. In each cheese and for each species several strains were isolated, with one or two largely dominant. Although the same genus and species were identified in each cheese – 6 and 11 different respectively –, patterns and dynamics of dominant strains were cheese specific. The microflora was thus complex in each cheese in term of number of different species and of different dynamics.

The results confirmed previous reports on lactic acid bacteria in Comté cheeses. The strains originated either from milks or starter cultures, which can only partially explain their dynamics. But interactions between strains in environmental conditions that vary as ripening progressed should be considered in the future.

Keywords: Comté cheese, lactic acid bacteria, propionic acid bacteria, diversity 36

Characterization of the Microflora in Scandinavian Cheese During Ripening

H. M. Østlie*, L. Eliassen, S. Skeie

Agricultural University of Norway, Department of Chemistry, Biotechnology and Food Science, Norway hilde.ostlie [ AT ] inf.nlh.no

One of the most popular cheeses in Norway is the semi hard cheese named Norvegia and in Sweden the semi hard cheese named Präst. Norvegia and Präst are both cheese varieties made with a mesophilic DL-starter including strains of Lactococcus and Leuconostoc. The commercial manufacture of Norvegia and Präst are controlled processes that utilizes pasteurized milk in dedicated plants under hygienic conditions. However, despite the application of different precautions during cheese-making a secondary non-starter bacterial population develops in the curd. Non-starter lactic acid bacteria (NSLAB) is a collective term used to describe the adventitious LAB flora capable of growth under the selective conditions of ripening cheese. The objective of this study was to evaluate the diversity of the microflora in Norvegia and Präst cheese according to different dairies and different ripening stages.

Microbiological sampling of Norvegia and Präst cheese from three Norwegian and three Swedish manufactures, respectively, was done during ripening. The evolution of total bacterial counts, lactococci, lactobacilli, enterococci, presumptive leuconostoc and pediococci was investigated after 30, 90, 180 and 270 days of ripening. Isolates (275) of non-starter lactic acid bacteria (NSLAB) from 9 Norvegia and 12 Präst cheeses after 90, 180 and 270 days of ripening were examined. The isolates were tested by physiological and biochemical assays, species-specific PCR and 16S rDNA sequencing.

In Norvegia cheese, Leuconostoc spp., most probably from the starter, and the NSLAB specie Lactobacillus paracasei dominated among the isolates after 90 days of ripening , however, after longer ripening Lb. paracasei dominated. In Präst cheese, the predominant NSLAB specie was Lb. paracasei. The development and evolution of the microflora in Norvegia and Präst varied according to dairy and ripening time.

Keywords: non-starter lactic acid bacteria (NSLAB), cheese, ripening, identification

Starter Lysis in Cheese: an Essential but Complex Phenomenon

S. Lortal1*, M.P. Chapot-Chartier2

1INRA-STLO, Rennes, France; 2INRA-UBSP, Jouy en Josas, France

lortal [ AT ] labtechno.roazhon.inra.fr

Bacterial autolysis can be defined as the breakdown of a cell due to the hydrolysis of the peptidoglycan network. Endogenous cell wall associated enzymes, autolysins, or prophage encoded enzymes, endolysins, can induce this «suicide» phenomenon. Bacterial lysis has been observed in many species, but the mechanisms are, in most cases, not fully elucidated. Following conclusive experimental data obtained in the last decade, dairy starters lysis is now regarded as an essential pre-requisite in cheese maturation, since intracellular starter enzymes, in particular peptidases, can then play their major role.

In this review, we will present:

i) Current knowledge regarding starter lysis in cheese and, in particular, the different ways used to detect and follow lysis in situ ii) Current knowledge regarding the mechanisms involved, focussing mainly, but not exclusively, on the model most investigated: Lactococcus lactis. Recent advances concerning the molecular characterization of peptidoglycan hydrolases will be summarized

(sequence, structure, regulation) as well as current knowledge about the relationship between lysogeny and lysis. iii) The impact of lysis on the ripening indices of several cheeses and the different ways to control or induce increased lysis in situ. Briefly, the increase of free amino acids due to early lysis has been well documented in the literature, as well as the reduction of bitterness by hydrolysis of large hydrophobic peptides. However, the impact of lysis on the generation of flavour compounds by the subsequent catabolism of these amino acids needs to be clarified.

In conclusion, lysis of dairy starters has been well documented over the last few years, but many questions remain to be answered: origin of strain dependancy? natural induction of lysis in cheese? occurrence in cheeses other than pressed cheeses (what about lysis of surface flora in soft cheeses? in blue veined cheeses?…), the direct relationship between lysis and amino acid catabolism or lipolysis (very few reports)? and the outcome of intracellular compounds released through lysis like other enzymes, minerals, DNA constituants, pyruvate, co-factors… Do they have any role in the growth of other species or in cheese ripening reactions?

Keywords: starter lysis, cheese 38

Proteomics used to analyse enzymes in cheese

V. Gagnaire*, M.A. Manso, G. Jan, J. Léonil

Institut National de Recherche Agronomique, Laboratoire de Recherches et de Technologie Laitière, France gagnaire [ AT ] rennes.inra.fr

Bacterial ecosystems confer to cheeses organoleptic characteristics, during ripening, not only through the living cells (acidification of milk, production of organic acid compounds), but also through the action of proteins released into cheese as the result of cell lysis, especially different groups of enzymes. A relevant control of cheese quality would require a deeper understanding of the cell machinery responsible for the enzymatic reactions as they take place in situ.

The use of proteomic tools allows a global and dynamic view of the proteins expressed by the bacteria, i.e. their proteome, under various conditions of culture. Since an increasing number of bacterial genomes, notably those from lactic acid bacteria (lactococci, lactobacilli) and propionibacteria have been sequenced, and great progresses in protein sequencing have been performed, it is now possible to use proteomics to screen the enzymes in complex food systems, such as cheese.

For this purpose, the release of enzymes into Emmental cheese aqueous phase was investigated using a methodology based on: size exclusion chromatography, 2D electrophoresis and MALDI-TOF mass spectrometry and/or de novo sequencing by nanoscale LC-ESIMS/

MS tandem mass spectrometry. This methodology allowed us to identify five functional groups of proteins involved in: 1) proteolysis, 2) glycolysis, 3) stress response, 4) DNA and RNA repair and 5) oxido-reduction. These results revealed stress responses

triggered by thermophilic lactic acid bacteria and Propionibacterium strains at the end of Emmental ripening. Such a study offers other potential applications. The bacterial enzymatic battery could be screened in situ in various types of cheese or dairy products, to study the enzyme release and even enzyme activity throughout technological processes. In turn, metabolic pathways involved in organoleptic characteristic elaboration or in technological stress resistance could be orientated.

Keywords: cheese, proteomic, stress, proteolytic and glycolytic enzymes 39

Lactic Acid Bacteria Peptidase Activities: Model System to Investigate the Role in Grana Cheese Ripening

E. Neviani1*, M.E. Fornasari2, F. Gardini3, G. Mucchetti2, M. Gatti2

1 Università di Parma, Dipartimento di Genetica Evoluzione Antropologia, Italy; 2Istituto Sperimentale Lattiero Caseario di Lodi, Italy; 3Università di Bologna, Dipartimento di Protezione e Valorizzazione Agroalimentare, Italy

erasmo.neviani [ AT ] unipr.it

Cheese ripening is essentially characterised by enzyme activity accounting for the transformation of the curd’s principal components into more simple substances. Firstly as a well defined cellular entity and secondly as enzymes released after cell lysis, lactic acid bacteria (LAB) are the main cause of this process of change. The intervention of a complex bacterial enzyme-based proteolytic system was supposed to ensure a progressive degradation of the casein-derived peptides into free amino acids and other compounds participating to flavour formation during the time. At present time this biochemical event is only deduced by the evidence of simultaneous presence and evolutions in cheese of shortened casein peptides and LAB peptidase activities in a stable form. As enzymatic activities could be induced or inhibited by different technological parameters of processing or ripening conditions, the presence of enzymatic activities was not enough to demonstrate the activity in cheese paste.

The aim of this work was to investigate in which phases of ripening of Grana cheese, LAB peptidase activities present in cheese extract could be involved in casein degradation. The ability to efficiently hydrolize N-terminal proteins residues was investigated varying pH of substrates, incubation temperatures and salt concentration to elaborate models useful to understand when these activities play a specific role in subsequently moments of cheese ripening.

Results evidenced that the different LAB peptidases are strongly conditioned by cheese processing and ripening parameters. Enzymes act at different time during ripening. The different condition present in the different cheese paste zones during the first period of ripening also conditioned enzymatic activity.

Keywords: peptidase activities, LAB, Grana cheese, model system

Role of Starter Lysis in Cheese Production

W. Meijer1*, F. Kingma1, S. van Schalkwijk1, H. Brandsma2, J. Hugenholtz1

1NIZO Food Research, Department of Flavour, Nutrition and Ingredients, The Netherlands; 2CSK Food Enrichment, PO Box 225, 8901 BA, Leeuwarden, The Netherlands

wilco.meijer [ AT ] nizo.nl

Lysis of lactic acid bacteria plays a crucial role in dairy fermentations. The bacterial lysis ensures the involvement of the intracellular starter enzymes in the gradual process of cheese flavour formation. At NIZO food research we are investigating the lysis process in the following way: i) The comparison of highly stable and unstable strains in cheese manufacture to establish the role of starter lysis in the various stages of cheese ripening, and: ii) The role of growth conditions on the starter cell permeability and on the efficiency of starter lysis and its implication for cheese manufacture.

200 Lactococcus lactis strains were compared in cell stability and sensitivity of the cell walls for the lytic enzyme mutanolysin. In general, rapidly lysing strains showed high sensitivity of the cell walls for mutanolysin, while highly stable strains were more resistant. In cheese experiments, the rapidly lysing strains generated much more flavour components than the highly stable strains. Special attention was given to the accumulation and removal of the bitter-flavour in cheese. A direct correlation was seen between stability of cell(s)(walls) and the accumulation of bitter flavour in cheese. A range of specific debittering, flavour enhancing cultures, e.g. L. lactis NIZO B33 and AB28 with a high lysis efficiency, have been isolated and developed into adjunct starter cultures.

The effect of growth conditions on lysis efficiency was studied in a continuous culture. A highly labile strain (L. lactis AB28) and a highly stable strain (L. lactis AB8) were both cultivated at different growth conditions. In the labile strain, the cells, and the cell wall composition, became increasingly stable at lower growth rates, while the stable strain was not visibly effected by these changes. In addition, effects of growth temperature and pH on cell permeability were observed for both strains. Most of the observed changes in cell stability and/or permeability could be correlated with changes in the cell wall composition, in particular the peptidoglycan structure. These results indicate that the lysis process can indeed be controlled by choosing suitable growth conditions.

Keywords: lysis, cheese ripening, starter cultures 41

Mechanisms of Incorporation and Release of Enzymes into Cheese During Ripening

M.G. Wilkinson

Department of Life Sciences, University of Limerick, Castletroy, Limerick, Ireland

Martin.Wilkinson [ AT ] ul.ie

This article will review current knowledge regarding the means by which enzymes are incorporated and released into cheese during ripening. Cheese ripening is catalysed by enzymes from: milk, coagulant, starter Lactic Acid Bacteria (LAB) and non-starter Lactic Acid Bacteria (NSLAB). Attempts to accelerate flavour development have also lead to the addition to cheese of exogenous commercial enzyme preparations including adjunct and attenuated LAB and NSLAB strains. The possible entry points for enzymes into cheese are via: milk, wash water, dry salt or brine. However, exogenous enzymes are generally added to cheesemilk or with dry salt to the curd. The

contribution of enzymes present in cheese during ripening depends on: (1) the degree to which they entrapped in the curd (2) their distribution within the curd and (3) their activity under the physico-chemical conditions existing in the particular variety. Most indigenous milk enzymes are associated with the caseins or milkfat and partition with curd during cheesemaking providing well distributed residual enzyme activity. Little information exists on the mechanism of entrapment of LAB or NSLAB within the curd but it is believed they become physically entrapped during cheesemaking and localise at the interface between milkfat gobules and the casein matrix. Release of intracellular enzymes from starter LAB and NSLAB occurs through cell permeabilization or autolysis in a strain-dependent manner. In the case of exogenous enzyme preparations, the choice of entry point greatly affects both the extent of enzyme incorporation and distribution in the curd. For exogenous enzymes added to milk, considerable losses are experienced at whey drainage, hence to improve the extent of incorporation, various enzyme encapsulation techniques have been developed and these are also reviewed.

Keywords: enzymes, incorporation, release, encapsulation. 42

Acceleration of Proteolysis in Cheese using Streptokinase, a Plasminogen Activator

V.K. Upadhyay1, M.J. Sousa1, P. Ravn2, H. Israelsen2, A.L. Kelly1, P.L.H. McSweeney1*

1Department of Food and Nutritional Sciences, University College, Cork, Ireland; 2Biotechnological Institute, Hørsholm, Denmark p.mcsweeney [ AT ] ucc.ie

In milk, there are higher levels of plasminogen, the inactive precursor of plasmin than of the active proteinase. Streptokinase is an extracellular protein produced by the mastitis pathogen Streptococcus uberis which can activate plasminogen to plasmin. The aim of this study was to activate plasminogen to plasmin using a semi-purified preparation of streptokinase, or as starter a strain of Lactococcus (PRA270) which had been genetically modified to produce streptokinase, and to study effects of increased plasmin activity on proteolysis during cheese ripening. Cheddar cheese was manufactured from milk supplemented with a semi-purified preparation of streptokinase at 0.1%, 0.25 % or 0.50 % (v/v) and, in separate trials, miniature Cheddar-type cheeses were made using strain PRA270 as starter. Both approaches resulted in activation of plasminogen, with most occurring on the day of manufacture, hence increasing plasmin activity in experimental cheeses ~ 3 fold compared to control cheeses. Increased plasmin activity in the experimental cheeses resulted in accelerated proteolysis as indicated by higher levels of pH 4.6-soluble N, accelerated hydrolysis of â-casein and concomitant increases in concentrations of ã-caseins. Reversed-phase HPLC showed increased production of hydrophobic peptides in cheeses with higher plasmin activity; however, levels of free amino acids were unaffected. In conclusion, use of streptokinase, either as an exogenous semipurified preparation or produced by the starter culture, activated bovine plasminogen to plasmin, which accelerated proteolysis in

Cheddar cheese during ripening. Keywords: Plasmin, streptokinase, genetically-modified starter, proteolysis, cheese

Injection of Starter Bacteria and Ripening Enzymes into a Cheese Matrix

K.T. Andersen1*, J. S. Madsen1,2 1Lact Innovation Aps, Denmark; 2The Royal Veterinary and Agricultural University,

Department of Dairy and Food Science, Denmark kim [ AT ] st-clemens.dk

Adding starter bacteria and rennet to milk makes cheese, and during maturation the cheese characteristics develops. However, adding starter bacteria, adjunct ripening bacteria or enzymes directly to the cheese matrix improves ripening control and may even accelerate ripening (patent WO0180655, 2001). In addition this method reduces the loss of ripening agents in the whey and for that reason enhances the quality of the whey.

We made semi-hard cheeses from milk acidified to pH 6.4 with lactic acid. Just before final pressing we used a needle device to inject a suspension of starter (Lactococcus lactis) into the cheese matrix. Adjunct bacteria (Lactococcus and Lactobacillus species) and enzyme (serine protease), respectively was injected to accelerate ripening. We evaluated ripening by pH 4.6 soluble N analysis, casein degradation by Capillary Electrophoresis (CE), and peptide profiles by reversed phase High Performance Liquid Chromatography (rp-HPLC). For comparison we analysed a commercial nine-month-old Cheddar cheese.

Cheese pH decreased to 5.1–5.3, and no marks were detected from the needle penetration. Cheese containing starter bacteria and adjunct bacteria approached similar level of pH 4.6 soluble N (22.1 %) after 3 weeks as in a 9 months old Cheddar cheese (21.2 %). However, cheeses with enzyme increased soluble N to 35–45 % of total N. CE revealed, that áS1-and â-caseins were totally degraded in the latter cheeses, while more than 50 % of casein was present in the cheeses without ripening enzyme. Total peak area in the HPLC chromatograms reflected the comprehensive casein degradation, as enzyme treated cheese showed the highest area. However, hydrophilic peptides accounted relatively more to total peptide profile in cheese with adjunct bacteria, indicating release of peptidases to the cheese matrix.

Full-scale cheese trials with combination of ripening enzyme and adjuncts are in progress.

Keywords: Technology, Adjunct, Incorporation, Peptides

Overproduction of Cystathionine â-Lyase in Lactococcus lactis Affects Flavour Development in Gouda Cheese

J.E.T. van Hylckama Vlieg1, R. van Kranenburg1, P. Bruinenberg2*

1NIZO Food Research, The Netherlands; 2Campina Innovation, The Netherlands

p.bruinenberg [ AT ] campina.com An attractive approach to accelerate flavour development in Gouda cheese is to overproduce in L. lactis cystathionine â-lyase (CBL), an enzyme involved in conversion of methionine into volatile sulphur compounds present in several cheese types. A comparative in vitro expression study showed that the CBL enzyme harboured the highest activity towards methionine in comparison with other predicted lactococcal C-S lyases. Fifty-fold overproduction of specific CBL-activity was achieved by controlled expression of the metC gene, encoding CBL, of L. lactis B78 using the food-grade nisin-inducible expression system (NICETM). Cheese graders awarded an additional sulphur/thermophilic flavour to Gouda cheese prepared with the CBL-overproducing strain as an adjunct. In an attempt to accelerate flavour development we manufactured Gouda cheese which, together with the CBL-overproducer, also contained a L. lactis strain

producing the lysin and holin proteins of lactococcal bacteriophage US3. Production of lysin and holin leads to in trans lysis of surrounding bacteria during cheese ripening, thereby increasing the amount of released CBL in the cheese matrix two-fold. However, sensory analysis showed no significant effect of enhanced cell lysis on cheese flavour development. A possible explanation for this result is that part of CBL present in cheese extracts is found to be inactive, probably due to dilution of its cofactor pyridoxal-5’-phosphate into the cheese matrix upon cell lysis.

Keywords: cystathionine â-lyase, food-grade overproduction, cheese trials, sulfur flavour 45

Approaches to Maximising Recovery of Milk Constituents During Cheese Manufacture

J.M. Banks

Hannah Research Institute, Ayr, Scotland

j.banks [ AT ] hannah.ac.uk

This review will consider milk compositional and processing factors influencing the efficiency of recovery of milk constituents in the manufacture of rennet curd cheeses. Compositional factors will include genetic polymorphs of milk proteins and other factors influencing curd formation which impact on the retention of fat and protein during cheese manufacture. Processing variables influencing cheese yield in routine manufacture will be considered. Novel technologies such as high pressure, microfiltration and ultrafiltration and their potential use in maximising protein or fat recovery in curd will be considered together with the effects of these processes on the syneretic properties of curds and the ripening characteristics of the resultant cheese.

Keywords: Cheese yield, manufacture, genetic polymorphs

Use of Cold Ultrafiltered Retentates for Standardization of Milks for Pizza Cheese: Impact on Yield and Functionality

S. Govindasamy Lucey1*, J.J. Jaeggi1, M.E. Johnson1, J.A. Lucey2

University of Wisconsin-Madison, 1Wisconsin Center for Dairy Research, 2Department of Food Science, USA rani [ AT ] cdr.wisc.edu

A washed, stirred-curd cheese was manufactured from two types of UF fortified milks: high-solids (HS; 15.4% TS, 4.5% casein), and medium-solids (MS; 13.5% TS, 3.6% casein). Cheesemilks (casein:fat ratio ~ 1.0) were obtained by blending cold processed UF

retentate (27.5% TS, 8.3% casein, 12.2% fat) with partially-skimmed milk and UF (skim milk) retentate (11.3% TS, 3.9% casein, 0.2% fat). Control cheese was made with partially-skimmed milk (11.3% TS, 2.5% casein). Coagulation at 34 °C was monitored by dynamic low-amplitude oscillatory rheology (DLAOR). Cheese functionality was assessed using DLAOR and when baked on a pizza. Gels made from UF-fortified milks had similar clotting times and were faster (16–17 vs 22 min) than control milk. Shear stress values of gels were 62, 50 and 25 Pa for HS, MS and control milks, respectively. MS cheese had lower moisture contents (45.6%) than control (47.6%) or HS cheeses (47.7%). Fat recoveries in cheeses were higher in MS cheeses (92.51 ± 0.86) than control cheeses (90.26 ± 1.35) but lower in HS cheeses (88.57 ± 0.83). Nitrogen recoveries were lower in control cheeses than in UF-fortified cheeses. Loss tangent curves (at temperatures > 40 °C) shifted higher as cheese aged up to a month and decreased with further ripening. Temperature for the loss tangent decreased with age up to 2 mo for all three cheeses types, thereafter it did not change much. Loss tangent values were slightly lower in UF-fortified cheeses. TCA-soluble nitrogen levels were similar in all three cheeses. When the cheeses were baked on pizzas, only slight differences were noted between the cheeses. In conclusion, standardization of milk with cold UF retentates can be used to increase cheesemaking productivity without adversely affecting cheese functionality.

Keywords: UF, non- pasta-filata pizza cheese, cheese yield, cheese functionality 47

Carbohydrate-Based Fat Replacers and Functionality of Low Fat Mozzarella Cheese

R.K. Bhaskaracharya, B. Zisu, N.P. Shah*

School of Molecular Sciences, Victoria University, Melbourne, Australia Nagendra.Shah [ AT ] vu.edu.au

Reducing the fat content in Mozzarella cheese results in a lower ratio of moisture to protein leading to reduction in functionality of cheeses. Reduction in fat content also results in an inferior quality cheese, specifically poor melt and stretch and the cheese becomes harder with rubbery texture. Carbohydrate-based fat replacers has been used to increasing the moisture content and consequently improving the functional properties of the cheese. Skim milk Mozzarella cheeses (<3% fat) were made using two maltodextrin based (Maltrin M1 and M2) and a modified potato starch based (StaSlim S1) fat replacers. A control batch was made with skim milk without any fat replacers. The texture characteristics of the cheeses were measured with an Instron Universal Testing Machine and the microstructure examined using scanning electron microscopy. The moisture contents of the skim milk and Maltrin-based cheeses were similar,

while StaSlim S1 based cheeses showed lower moisture levels. The protein contents of the cheeses made with fat replacers were lower than the control cheeses. The Maltrin based cheeses were lower in hardness than the control or StaSlim based cheeses and the hardness decreased during storage, while the StaSlim-based cheeses became harder during storage. In general, the cheeses made using fat replacers showed less cohesiveness and springiness than the control cheese, while adhesiveness increased. The Maltrin-based cheeses showed significantly less gumminess and chewiness, however, these values for StaSlim-based cheeses increased during storage. Incorporation of fat replacers resulted in an increased openness in cheeses, and large serum channels (up to 0.1mm diameter) were seen. The Maltrinbased cheeses showed more openness than the StaSlim-based cheeses. However, carbohydrate-based fat replacers do not behave like fat, particularly when heated. The pizza bake performance for such cheeses showed poor results in regards to melting and fusion of shredded cheese used as pizza topping. High degree of surface scorching was also observed.

Keywords: Low fat, Mozzarella cheese, fat replacer 48

Economic Aspects of Cheese Making, Including Comparisons to High Value Whey Products

R.H. Peters1*, M. Doré2

1Agropur Cooperative, Granby, Canada; 2Manufacturing Services Parmalat, Canada

rpeters [ AT ] agropur.com

Cheese making inevitably entails the production of whey. The economics of processing this liquid fraction is for a great deal dependant on capital investments which are much more dependant on scale of economics, than cheese making itself. Small cheese plants with daily milk throughput of approximately less than 100 000 kg cannot economically justify the capital for water removal equipment. For small plants which have to convert the whey to dry product, this can be done by pre-concentrating with a reverse osmosis unit or a small plate evaporator and drying on a double roller dryer. the economics are evaluated at several price levels. At the upper scale of cheese plant size at 2 to 3 million kg/d of milk, the investment for whey processing is about half the total investment. Returns on investment are calculated for several, electricity and natural gas prices and at varying whey powder prices. Increased investment for further processing into whey protein concentrate and dried whey solubles or lactose is evaluated at several price levels.

Keywords: Economic aspects, cheese making, whey products

P001 Particular Treatments to Achieve Typical Flavours in Two Peculiar Semi-Hard Italian Cheeses

C. Corradini*, N. Innocente

Department of Food Science, University of Udine, Italy

Cesare.Corradini [ AT ] DSA.Uniud.it

In the North East of Italy small amounts of few characteristic cheeses are produced with typical flavours that can be achieved only by particular treatments. For instance, the Asino cheese is a traditional dairy product of some areas of Friuli Venezia Giulia, obtained by normal cheeses (soft or semi-hard) that, after a normal salting, are dipping in a special diluite salting brine, mixed with whey, milk and milk cream, called „salmuerie”. So Asino cheeses acquire their own typical flavour. This research contribute to the description of the biochemical processes thank to which the Asino cheese has the peculiar sensorial characteristics that make this cheese different from similar cheeses ripened in warehouses without dipping in „salmuerie”. Particularly, in all the „salmuerie” analyzed there was a fair presence of short chain free volatile fatty acid that remains almost steady for all the time in which the cheese is dipped into. Instead the free volatile fatty acids in cheese starts from values very low and then rises and reaches a balance with the salmuerie values. An other interesting product that we are considered in this research was a cheese dipping, after ripening, in wine or must and marc. We have showed that components of wine migrate to cheese by an exchange process that gives to this product typical taste and smell. Actually, after dipping in wine was found fairly amounts of ethyl acetate component, that give to cheese characteristics floral flavours.

Keywords: semi-hard cheeses, salting, biochemical processes, flavour 

P002 The Effect of Moisture and Ripening Time on Model Cheese Textural Properties and Proteolysis

P.J. Watkinson*, C. Coker, C. Dodds, S. Hewson, B. Kuhn-Sherlock, N. White

Fonterra Research Centre, Fonterra Palmerston North, New Zealand philip.watkinson [ AT ] fonterra.com

Uniaxial compression was used to measure the effect of moisture and ripening time on model cheese texture. This allowed a universal approach to cheese characterisation, because properties that are well defined in rheology and method protocols used in the international community were utilised.

Knowledge of the textural changes induced by moisture can assist the development of new cheeses of different moisture contents and defined textural attributes. Two sets of model cheeses with moisture ranges of 34–39 % (pH 5.51–5.54) and 40–48 % (pH 5.76–5.80) were made using the acidulant glucono-ä-lactone to minimise pH changes within a set of model cheeses.

The rheological properties predicted by moisture (averaged over 2–184 days of ripening time) changed as follows. As the moisture increased from 40 to 48 %, fracture strain (longness) increased 2.1 fold, fracture stress (firmness) decreased to 54 %, modulus (stiffness) decreased to 30 % and adhesion area (adhesiveness) increased 44 fold, compared with the value at 40 % moisture. Similarly, from 34 to 39 % moisture, fracture strain increased 1.9 fold, fracture stress decreased to 81 %, modulus decreased to 73 % (with slightly non-linear trends) and adhesion area gave non-linear trends. The cheddaring temperature (which induced moisture changes) had an effect (P < 0.05) that depended on the ripening time for fracture strain, modulus and adhesion area (at 40–48 % moisture) and for fracture strain and adhesion area (at 34–39 % moisture).

ás1-Casein breakdown was faster at higher moisture, probably because chymosin was more active (and cheddaring temperature had an effect [P < 0.0005] that depended on the ripening time for 40–48 % moisture cheeses). â-Casein breakdown was either faster for higher moisture in the 34–39 % moisture cheeses or similar at all moistures in the 40–48 % moisture cheeses. Moisture had no significant influence on non-protein nitrogen per total nitrogen levels (cheddaring temperature had no effect [P > 0.05]).

Keywords: cheese moisture, texture, rheology, ripening time %&

P003 Authentication of PDO Ovine Cheeses. Identification of the Use of Vegetable Coagulant (Cynara L)

L.B. Roseiro1*, J.A. Gómez-Ruiz2, M. García-Risco2, E. Molina2

1Unidade de Indústrias Lácteas, Portugal; 2Instituto de Fermentaciones Industriales, CSIC, Spain

luisa.roseiro [ AT ] mail2.ineti.pt Some of the traditional ovine cheeses produced in the Iberian Peninsula use the aqueous extracts from dried cardoon flowers (Cynara L.) as coagulant. Some of them carry the label of Protected Designation of Origin (PDO), making compulsory the use of local pure ovine raw ''(

milk and cardoon (usually C. cardunculus L.) as coagulant. However, their increasing consumer demand might lead to the fraudulent addition of milk from other species to extend production and/or the substitution of the vegetable by other conventional coagulant. Detection of milk mixtures in cheese has been done by different methods, but detection of the coagulant origin in cheese is not yet possible. A Capillary Zone Electrophoresis (CZE) method has been applied to a cheese made with a vegetable coagulant, which was compared to other cheeses made with animal rennet and microbial coagulant. The results obtained led us to suggest that CZE of cheese caseins is a suitable, easy to perform and fast method for evaluating the authenticity of cheeses made with Cynara L. A peak, probably arising from â−casein, has only been detected in cheeses coagulated by cardoon and it is not proteolysed during maturation. This peak could be the distinguishing factor of the type of coagulant employed.

Keywords: Authentication, PDO Ovine Cheese, Identification, Vegetable Coagulant, Cynara L Part of this work was presented as a poster at Congrilait, held in Paris from 24th to 27th September 2002. Complete work was submitted to Le Lait in October 2002

P004 Flavonoid Glycosides from the Vegetable Coagulant Cynara L. as Markers for PDO Cheeses (preliminary results)

L.B. Roseiro1*, J.M. Besle2, D. Viala2, J.L. Lamaison3, A. Carnat3, D. Fraisse3

1Unidade de Indústrias Lácteas (UIL)-DTIA–INETI; 2INRA Clermont Ferrand-Theix, France; 3Faculty of Pharmacy, Clermont Ferrand, France

luisa.roseiro [ AT ] mail2.ineti.pt

The majority of the PDO ovine cheeses produced in Portugal use the aqueous extracts from dried cardoon flowers (Cynara cardunculus L.) as coagulant, which is prepared by the cheesemaker just before use. This purple-brownish colour aqueous extract is completely added to the milk, and apart from its proteolytic enzymes (cardosins) nothing is known about other components present therein.

Aqueous extracts from air-dried flowers of cardoon were analysed by HPLC-DAD for the presence of flavonoids, and the results obtained revealed the presence of two major peaks characterised as apigenin-7-O-glycosides, one of them being identified as isorhoifolin (apigenin-7-O-rutinoside).

Model cheeses made from raw cow milk and cardoon, raw cow milk and animal rennet and raw cow milk and animal rennet plus isorhoifolin as standard were made, in order to detect the flavonoids characteristic from the vegetable coagulant in cheese.

HPLC-DAD analysis of curds and respective wheys revealed that the curd obtained with cardoon presented the same two major peaks detected in the coagulant. isorhoifoline was also present in the curd obtained with animal rennet plus added standard, but none of the flavonoids found in the vegetable coagulant were present in the curd obtained with animal rennet. No evidence of these flavonoids in any of the wheys revealed that they are almost completely retained in the curd.

These preliminary results lead us to suggest that the two apigenin-7-O-glycosides characteristic from the vegetable coagulant, specially isorhoifolin, could be used as markers for the authentication of PDO cheeses made with cardoon.

Keywords: cheese authentication, vegetable coagulant, isorhoifolin, authenticity markers &

P005 Ripening Profile of Urfa Cheese (A Turkish White-Brined Cheese) Produced by Traditional Way and by Ultrafiltration Technique

B.H. Özer1*, A.F. Atasoy2, H. Türkoglu1

1Harran University, Faculty of Agriculture, Department of Food Engineering, Sanliurfa, Turkey; 2Harran University, Sanliurfa Vocational School, Department of Food Technology, Sanliurfa, Turkey

bozer [ AT ] e-kolay.tr

The development of ripening in white-brined Turkish cheese (Urfa type) produced by traditional way and by ultrafiltration (UF) technique was studied. The effects of scalding on the development of proteolysis and texture of cheeses were also investigated. Variations in the nitrogeneous fractions and textural characteristics were investigated throughout 90-day storage at 4 °C.

Results obtained showed that the proteolysis developed faster in the UF-cheese than the cheese manufactured by traditional way. The UF sample had higher levels of water soluble nitrogen, proteose-pepton nitrogen, non-protein nitrogen and ripening coefficients than the traditional cheeses. Scalding slowed down the development of proteolysis in both group of cheeses, however, the trend of ripening remained almost unchanged. SDS-P.A.G.E. electrophoretograms showed that the breakdown of á

s - and â-casein was more remarkable in the unscalded UF cheese. SEM pictures revealed that the UF sample had denser microstructure than the traditional cheese and scalding led to even much denser structure. In accordance with microstructure of samples, the textural properties of UF cheeses were higher than that of traditional cheeses. The hardness and springiness of the samples increased rapidly during the first 15–20 days of storage and then ''*

continued to increase with a decreasing velocity during the rest of the storage. The UF cheeses (both scalded and unscalded) had springer bodies than their traditional counterparts.

To conclude, it can be said that by employing UF technique, it may be possible to omit dry salting prior to brining and scalding steps from the manufacturing practices of white-brined Turkish cheese (Urfa type).

Keywords: Turkish white cheese, proteolysis, texture 

P006 Chihuahua Cheese: Microbiological and Physicochemical Properties

C. Figueroa, F. Meda, H. Janacua*

Centro de Investigación en Alimentación y Desarrollo, A. C. Unidad Cuauhtémoc, Chihuahua, Mexico hjanacua [ AT ] cascabel.ciad.mx

Chihuahua Cheese is a semi-hard cheese that is highly desired and it is consumed fresh, usually within two months after manufacture. It is produced by Mennonite communities in the State of Chihuahua, Mexico, and is considered a unique style of cheese, though it is a variation of Cheddar cheese. Daily production is more than 55 000 kg in 23 small cheese factories. Only a small amount of literature has been published on the microbiological content and physicochemical properties. This investigation determined the sanitary quality and physicochemical characteristics of this Chihuahua cheese manufactured from raw milk. Samples of cheese were obtained from 9 cheese factories within a few days of being manufactured during the year 2002-03. The compositional properties of the fresh cheese were similar among all manufacturers. Composition averaged 40 % moisture, 21 % protein, 28 % fat, 1 % salt and the pH was 5.6. All cheeses tested negative for salmonella spp. Fecal coliforms counts ranged from 2 to 8 log10 cfu/g for raw milk cheese. Yeast and mold counts ranged from 3 to 5 log10 cfu/g and Staphylococcus aureus was found in the cheese studied. Establishing the basic chemical, microbiological and physical properties of Chihuahua cheese, is the first step in understanding its unique traits and exploring ways to improve quality and expand consumption.

Keywords: Chihuahua cheese, Mennonite, physicochemical properties

P007 Characterization of Compositional and Microbial Properties of Chihuahua Cheese Manufactured from Raw and Pasteurized milk

F. Meda1, C. Figueroa1, J. Molina1, J. Náñez2, I. García2, A. Orozco1, G. Ávila2, C. Blanco1*, H. Janacua1 Centro de Investigación en Alimentación y Desarrollo, A. C. 1Unidad Cuauhtémoc; 2Unidad Delicias, Chihuahua, Mexico

hjanacua [ AT ] cascabel.ciad.mx Chihuahua Cheese is a fresh and semi-hard cheese traditionally made in the State of Chihuahua, Mexico, and manufactured primarily by the Mennonite community. It is highly desirable, and is considered a unique style of cheese, though it is a variation of Cheddar cheese. Daily production in Chihuahua is more than 130 000 kg in small cheese factories. Little quantitative data has been published on these Cheese. This study characterizes and compares the physicochemical and microbial properties of Chihuahua cheese either from raw or pasteurized milk. Samples of cheese were obtained within a few days of being manufactured from nineteen cheese factories during the year 2002 and 2003 (14 raw and 5 pasteurized milk cheeses). The compositional properties were similar among cheese manufactured from raw and pasteurized milk. Overall composition averaged 41 and 40 % moisture, 21 and 20 % protein, 28 and 29 % fat, for cheeses made from raw and pasteurized milk, respectively. Color characteristics (L, a, b values) for cheeses from raw milk were 81, –2.70, and 16.70, respectively; and 85, –2.40 and 23.70 for pasteurized milk cheeses. Fecal coliforms counts ranged from 2 to 6 log10 cfu/g for raw milk cheeses, and 0.30 to 6 log10 cfu/g for pasteurized milk cheeses. Yeast and molds counts ranged from 3 to 7 log10 cfu/g for raw milk cheeses, and 2 to 6 log10 cfu/g for pasteurized milk cheeses. Staphylococcus aureus was found in raw milk cheeses, and was not found in pasteurized milk cheeses. Both types of cheeses tested negative for Salmonella spp. Further studies is needed to identify and explain which raw and pasteurized milk factors, and processing steps are key factors in producing quality traits of Chihuahua cheese. We urge to use pasteurized milk to avoid the possibility of human diseases due to pathogens in raw milk.

Keywords: Physicochemical characteristics, raw milk, pasteurized milk, Chihuahua cheese '',

P008 Detection of Low Levels of Cow Milk in Buffalo Mozzarella Cheese by Means of IEF and PCR Methods

R. Ghiglietti*, F. Locci, S. Francolino, L. Rossetti, G. Giraffa, G. Mucchetti

Istituto Sperimentale Lattiero Caseario, Lodi, Italy gmucchetti [ AT ] ilclodi.it

Buffalo Mozzarella has to be made exclusively with pure buffalo milk and any milk ingredient from other species has to be absent. According to EEC regulation 213/2001, it is assumed that cows’ milk is present if the apparent cows’ milk casein content of the sample to be analysed is equal to or higher than 1 %. The principle of the EU reference method is based on the different mobility exhibited by isoelectric focusing gel analysis (IEF) of buffaloes’ and cow’s ã2 and ã3 caseins resulting from the activity of plasmin towards the caseins extracted from the cheese.

However, even if cheese-makers claim the exclusive use of buffalo milk, sometimes they have been charged with the presence of cow milk in Buffalo Mozzarella near to the detection limit of the method.

IEF analyses of pure buffalo milk and Mozzarella showed a band with mobility similar to the cow’s ã2 casein. To assess the purity of these samples, a Polymerase Chain Reaction (PCR) technique, that amplifies cytochrome b mitochondrial gene, has been applied. PCR results confirmed the purity of buffalo milk. The sensitivity of the method, applied to detect the addition of cow’s milk to buffaloes’ milk, was higher than 0.05%. When applied to Buffalo Mozzarella cheese obtained from pure buffaloes’ milk or added with 1, 2 or 3 % of cow milk, PCR technique confirmed its reliability, correctly discriminating pure from adulterated samples.

In conclusion, the presence in both, milk and Mozzarella cheese of a band with IEF mobility similar to the cow’s ã2 casein, having a peak area near to that given by samples obtained by addition of 1 % cow milk, it is not necessarily a proof of buffalo milk adulteration. Studies are necessary to identify the nature and the origin of this band.

Keywords: species identification, IEF, PCR, buffalo

P009 Detection of Foreign Fats in Cows’, Ewes’ and Goats’ Milk Cheeses During Ripening

I. Mayo, G. Toledano, T. Requena*, J. Fontecha, M. Juárez

Instituto del Frío (CSIC) Ciudad Universitaria s/n, Madrid, Spain

trequena [ AT ] if.csic.es The triacylglycerides (TAG) composition of different cheeses with a Protected Designation of Origin (PDO) was studied for detection of foreign fat in milk fat. Mahon cheese (51 samples) made using pure cows’ milk and Manchego cheese (51 samples) with pure ewes’ milk were manufactured in either artisanal or industrial cheese factory and matured for 2, 4 and 6 months. Cabrales cheese (12 samples) was industrially made with pure cows’ milk and artisanal Majorero cheese (12 samples) with goats’ milk and ripened for 1, 2 and 3 months. TAG were separated by gas chromatography, using a short capillary column and analyzed according to their carbon number. Multiple regression equations based on TAG composition proposed to detect foreign fats in cows’ milk fat (Precht, 1991; EC, 2001); in ewes’ milk fat (Goudhil et al., 2003); and goats’ milk fat (Fontecha et al., 1998) had been applied.

In Mahon and Manchego cheeses where the lipolysis was feeble, the TAG composition did not change substantially during ripening. The values obtained for the applied formula for cows’ and ewes’ milk respectively maintain its range. Therefore this equations can consider useful for detection of foreign fat in this cheeses during the studied ripening period.

Nevertheless, Cabrales (blue cheese) and Majorero cheese (made with pregastric rennet pastes), showed during the ripening period a extensive lipolysis and changes in the TAG profile. Therefore in these cheeses, the applied formula become inadequate due to that results would indicate the presence of f