Background of the invention Cheese ripening is a time-consuming process involving complex and well-baianced reactions between glycolysis, proteolysis and lipolysis of milk components. Whatever cheese is considered, bacterial enzymes play a major role in this process. The proteases, peptidases, esterases, and the enzymes of amino acid catabolism in the lactic acid bacteria involved in these processes have been extensively studied (Law, J. and Haandrikman, A., 1997, International Dairy Journal 7 : 1-11).

It is well known that changing the bacterial enzyme content in cheese directly affects the rate of cheese ripening and its final flavour. One of the most efficient ways to increase the bacterial enzyme pool in cheese curd, without altering the primary starter and the type of cheese, is to add whole lactic acid bacteria, which are unable to grow and to produce a significant quantity of lactic acid, but which can still deliver active ripening enzymes during cheese ageing. These starter cultures are normally weakened and are called"attenuated". The preparation of attenuated starter cultures can be achieved by various treatments like heating, freezing, spray drying, freeze-drying, fragilisation using lysozyme or solvents, or by the selection of lactose negative mutants.

The first preparation and use of attenuated starter cultures as a cheese additive was reported in 1975 by Pettersson and Sjostrom (Pettersson, H. and Sjöström, G., 1975, Journal of Dairy Research 42,313-326). At that time, their main purpose was to accelerate proteolysis and shorten ripening time, which was of great economic interest. Today, in most developed countries, milk is of such high hygienic quality that the problem is not only to shorten the ripening time but also to improve the flavour.

The most widely studied method for the preparation of attenuated starter cultures is heat-treatment. This attenuation method is based on sub-lethal heat-

treatment of a bacterial cell suspension. The attenuated cells are then added to cheese mil long with the primary starter culture. It was shown that heat-treatment destroyed or delayed the acidification ability of the cells for several hours. Therefore the attenuated starter cultures do not significantly modify the acidification course of the curd, and subsequently the fat/moisture, and pH of the cheese. Most often, the loss of lactic acid production closely follows the loss of viability (Pettersson and Sjöström, 1975, ibid.).

In CA 2072 159 a method was disclosed to accelerate the maturation of cheddar cheese by using live and heat-attenuated L. casei.

Although there is still little agreement on which heat-treatment procedure is best, in most cases proteolysis and lipolysis were increased, leading to a reduced ripening time and improved flavour (bitterness reduction).

However, a critical point for heat attenuation remains in that the correct temperature/time combination must be defined so that acidification is eliminated or delayed without excessively denaturing potential ripening enzymes (e. g. proteases, peptidases, esterases, or others). Furthermore, the very narrow difference between adequate and excessive heat treatments remains a major practical problem (Castaneda, R., Vassal, L., Gripon, J. and Rousseau, M., 1990, Netherland Milk and Dairy Journal 44,49-52).

Alternative methods for attenuation have also been described, for example attenuation by freezing-thawing (Johnson, J. A. C., Etzel, M. R., Chen, C. M. and Johnson, M. E., 1995, Journal of Dairy Science 78,769-776), by lysozyme treatment (EP 02 46 163), by the use of solvents or by spray-and freeze-drying (Johnson, J. A.

C. and Etzel, M. R, 1995, Journal of Dairy Science 78, 761-768).

Despite promising results in cheese, the industrial use of attenuated starters is not widespread. The main reason for this is the empirical character of the attenuation methods which have been proposed so far.

The present invention seeks to solve some of the problems encountered in prior art processes, by providing a novel process for the preparation of attenuated starter cultures which is reproducible and well-balanced. Furthermore, the process of the invention is rapid and simple and it can be carried out easily and at low cost on an industrial scale.

Summary of the invention The present invention discloses a process for the preparation of an attenuated starter culture, which process comprises exposing cells of at least one strain in a starter culture to microwave radiation. Attenuated starter cultures of the invention can be used in, for example, production of cheese, cheese analogues or cheese- derived products. Attenuated starter cultures of the invention contain cells that have been treated with microwave radiation.

The invention also provides: -An attenuated starter culture obtainable by a process of the invention for the preparation of an attenuated starter culture; -An attenuated starter culture which has a viability of less than 30% and an enzyme activity of more than 50% that of a non-attenuated starter culture; -A process for the preparation of a cheese, a cheese analogue as a cheese- derived product which process comprises contacting a cheese, a cheese analogue or a cheese-derived product or a precursor thereof with an attenuated starter culture of the invention; and Use of an attenuated starter culture of the invention in the preparation of a cheese, a cheese analogue or a cheese-derived product. Optionally, the starter culture can be contacted with untreated cells following exposure to microwave radiation.

In a particular embodiment, the starter culture can be exposed to from 1000 to 30000 Joule of microwave radiation. Preferably cells can be treated for from 5 to 30 seconds at a power of from 200 to 1000 Watts (Joule/s), or most preferably for 15 to 20 seconds at from 300 to 650 Watts per 10ml of an aqueous composition for example a suspension containing zl09 cells/ml.

Preferably, attenuated starter cultures comprise cells that have been treated with microwave radiation in such a way that the viability of the attenuated starter culture is less than 20%, more preferably from between 1 and 10%, as compared to untreated starter cultures when measured in CFU/ml. Furthermore, preferably the killed cells are not lysed. Typically the same time, the treated cells retain their enzyme activity. In general, more than 50%, more preferably more than 90% of the enzyme activity of the attenuated starter culture is maintained. The activity of one enzyme, in particular preferably peptidase, present in the attenuated starter culture is more than 90% as compared to the activity in untreated cells, as can be

determined via the method indicated in Example 2. This peptidase activity is used as indication for the whole enzyme activity present in treated cells.

Ripening enzymes can be chosen from the group containing: proteases, peptidases, decarboxylases, amine oxidases, alcohol dehydrogenases, amino acid dehydrogenases, amino acid oxidases, amino transferases or transaminases.

Attenuated starter cultures can contain cells belonging to one or more of the strains selected from the group of species consisting of Lactobacillus, Lactococcus, Streptococcus, Corynebacterium, Bifidobacterium, Micrococci and yeasts.

An attenuated starter culture of the invention may be in a concentrated form and optimally may also be in a frozen form. Such attenuated starter cultures may be prepared by carrying out the process of the invention for the preparation of an attenuated starter culture and subsequently concentrating the resulting attenuated starter culture, optionally followed by freezing.

Detailed description of the invention In present invention a process for the preparation of an attenuated starter culture using microwave radiation is described.

The use of microwaves is well known in the art for killing cells in food preparations. From these studies it has been concluded that the exposure of cells to microwave radiation results in inactivation of microorganisms. While an inhibitory effect of microwave radiation on the viability of microorganisms is established, the question how microwave radiation brings about that effect has been a matter of discussion. It is most likely that the effect is of a thermal character. However, experiments performed by Rosenberg and co-workers revealed that microwave treatment resulted in a loss of viability at a rate which was higher than that obtained with conventional heating alone (Rosenberg, U. and Sinell, H. J., 1989, Zentralbl. Hyg. Umweltmed. 188 : 271-283). This suggests the co-existence of a so-called"athermal effect"during microwave treatment as well. This hypothesis is strengthened by the fact that recent studies have shown that microwave treatment at reduced temperatures (below 40°C) also results in decreased viability of microorganisms (Kozempel, M. F., et al. J. food Prot., 1998, 61 : 582-585). This is likely to be due to the fact that bacterial DNA, RNA and proteins absorb microwaves of specific frequencies which results in genetic and protein disorders (Webb, S. J., Nature, 1969,222: 1199-1200).

Taking the thermal and thermal effect together it can be concluded that microwave treatment can not be seen simply as an equivalent to heat treatment. On

the contrary, its destructive action on cells does not favour the use of microwaves as a method to generate attenuated starter cultures with high intact enzyme activity.

Surprisingly, the present invention discloses that a treatment of cells whereby cells are exposed to microwave radiation (also indicated as the"microwave treatment") shows to be very favorable for the attenuation of starter cultures which can be used in, for example, cheese making. The reason for this is that the microwave treatment of present invention appears to have a wider gap between adequate and excessive treatments, as compared to attenuation by conventional heat treatment. Furthermore, this simple and rapid method results in the appearance of cells with low viability and high enzyme activity, is highly reproducible and can be applied to concentrated cell suspensions, making it highly suitable for low cost use and continuous use on an industrial scale.

By the term starter culture we mean a particular strain of bacterium or yeast, or a combination of one or more species or strains from one or more genera, which by growing and metabolising, assist in the manufacture of the product, for example, by formation of taste, flavour and/or texture. For example, dairy starter cultures can contain lactic acid bacteria that grow and metabolise in milk and curd and assist in the production of mature cheese. Also starter cultures can be used in the production of meat and derivatives thereof.

By the term cheese we mean a dairy product belonging to a soft cheese variety such as Camembert, Munster or Livarot, to a semi-hard variety such as Tome de Savoie, Gouda, Edam or Mimolette or to a hard cheese variety such as Cheddar, Gruyere or the Swiss cheeses or Parmesan.

By the term cheese analogue we mean a product composed of dairy proteins such as casein, caseinate, whey protein concentrate coagulated by acidification or by heat-treatment, or a combination of both, to which cheese-flavouring agents are added.

By the term cheese-derived product we mean either a processed cheese, i. e. a preheated mixture of cheeses, butter or other fatty materials, melting salts or any type of cheese curds to which enzymes are added in order to give desired texture- and flavouring properties.

By the term precursor we mean a substance that is present in the process of the invention which will ultimately become a cheese, a cheese analogue or a cheese-derived product. Thus a precursor could be milk, butter or cheese curds for example.

In present invention, attenuated starter cultures can comprise lactic acid bacteria, which are highly affected in their ability to grow (i. e. their viability) and to produce significant quantities of lactic acid but which are still capable of delivering active ripening enzymes during cheese ageing.

Attenuated starter cultures can comprise lactic acid bacteria or any other type of cells that have been treated in such a way that the viability of the attenuated starter culture is less than 20%, more preferably less than 10%, most preferably less than 1%, as compared to an untreated starter culture as measured in CFU/ml.

Preferably more than 50%, more preferably more than 90% of the activity of the ripening enzymes in an attenuated starter culture is retained as compared to an untreated starter culture. Aminopeptidase activity can be used to determine how much activity is retained (see example 2).

Attenuation can be performed on bacterial or fungal starter cultures.

Preferably, Lactobacillus, Lactococcus, Streptococcus, Corynebacterium, Bifidobacterium, Micrococci or yeast species can be used. Most preferably, L. helveticus, L. lactis, L. casei, L. plantarum, L. acidophilus, L. delbrueckii, L. bulgaricus, S. thermophilus, L. lactis ssp. cremoris, K. lactis or B. linens can be used.

Attenuated starter cultures of present invention contain one or several strains from which at least one might be attenuated by treatment with microwave radiation.

Thus, a starter culture may be attenuated by exposure to microwaves and be mixed with manufactured cells and/or starter cultures.

Ripening enzymes are enzymes involved in the formation of taste, texture and/or flavour of cheese, cheese analogues, cheese-derived products and/or processed cheese. Formation of taste, flavour and texture involves complex and well-balanced reactions between glycolysis, proteolysis and lipolysis of the milk components and further degradation of amino acids into amines, keto-acids, aldehydes, ketones, alcools, esters, acid and sulfur compounds by a cascade of enzymatic reactions. Preferred ripening enzymes can be proteases (for example serine-, cysteine-, aspartic-, and/or metallo-endopeptidases), peptidases (for example amino-, carboxy-, di-, dipeptidyl-, tripeptidyl-, serine-type carboxy-, metallocarboxy-, cysteine-type carboxy-, and/or omega peptidases), decarboxylases (for example lysine-, arginine-and/or histidine decarboxylase), amine oxidases, aldehyde dehydrogenases, alcohol dehydrogenases, amino acid dehydrogenases, amino acid oxidases, amino transferases or transaminases.

Preferably the microwave treatment is carried out using a microwave or the like. Preferably, a conventional household microwave (oven) is used. For attenuation at an irradiation frequency of 2450 MHz, the microwave power can range from 10- 1800W. More preferably, at a power of between from 300W to 650W, the duration of the treatment can be from 5 to 30 sec.

This treatment is especially suitable in case the starter cultures are present in an aqueous environment e. g. are suspended in water. In case the starter cultures are in dried form (e. g. freeze dried) or in a concentrated state (10 to 40 w% dry matter) the skilled will appreciate the microwave treatment conditions will have to be adjusted to obtain the desired state of killed but not lysed. Most preferably, conditions for the microwave-treatment are chosen which result in at least 90%, more preferably 99% of the cells in the starters culture being killed but not lysed, while in general more than 50%, more preferably 90% of the activity of ripening enzymes is retained. Viability can be measured in, for example, CFU/ml (CFU = Colony Forming Unit) by plating a dilution of the attenuated cells on suitable MRS or YEG agar plates. Lysis can be determined by, for example, conventional methods using microscopy. Activity of ripening enzymes can be determined, for example, as described for peptidase in Example 2. Of course, all attenuation protocols in which a microwave-treatment is used in combination with an alternative attenuation treatment forms part of this invention.

We have found that in the"killed but not lysed"state of the starter culture alcohol formation or acidification were prevented and the flavour related enzymes were still active. So the disadvantages of the starters were reduced because the starter cultures were catabolitically inactive whereas the enzymes present were available for the cheese making process.

Another aspect of present invention provides an attenuated starter culture which consists of an attenuated starter culture comprising cells prepared via the method of the invention which are subsequently concentrated by, for example, centrifugation and optionally deep-freezing or lyophilising (freeze-drying). Also, a method for preparing such attenuated starter cultures, the use of said attenuated starter cultures in cheese, cheese analogue or cheese-derives products making, and a method for making cheese, cheese analogues and/or cheese-derived products with the said attenuated starter cultures are parts of present invention.

The following examples are only to be considered as illustration of the present invention, and are not to be construed as limiting.

Example 1 Killing of starter cultures using microwave radiation Killing activity of a conventional household microwave of 2450 MHz frequency on Lactobacillus helveticus strain CNRZ-32 (available from INRA-collection, Institut National de la Recherche Agronomique, Gouy-en-Josas, France) and Kluyveromyces lactis strain 2166 (ATCC No. 8585) was investigated by exposing the cells to 300 W and 650 W microwaves for different lengths of time. Thus, cells were cultured in 1 litre of appropriate (MRS for L. helveticus and YEG for K. lactis) medium and in the required (37°C, unstirred for L. helveticus ; 30°C, stirred for K. lactis) conditions until an OD650 of 1 and 2 for L. helveticus and K. lactis respectively (about 5X108 to 109 cells/ml) was reduced. Cells were harvested in the early exponential stage under sterile conditions and washed twice with distilled water. Pellets were resuspended in 100 ml of sterile distilled water and divided into portions of 10 ml in test tubes. Test tubes were placed in a water bath at 20°C until treatment.

Microwave treatment was carried out by placing the one glass test tube containing 10 ml of cell suspension in an almost horizontal position onto two beakers of different size. At a power of 650 Watts the following timepoints were measured: 5, 10,15,20 and 25 sec. At a power of 300 Watts, cells were radiated for 15,20, or 30 sec. After treatment test tubes were placed in a water bath again at 20° C for 10 min before determination of the killing rate. Viability was measured by plating the cells onto MRS (a medium for cultivating lactobacilli, the reference is"de Man JC, Rogosa M., and Sharpe E., 1960. A medium for cultivation of lactobacilli. J. appl. Bacteriol., 23,130-135) or YEG (YEG = yeast extract 10g/ ! : bacto pepton 20g/l ; dextrose 20g/l ; pH6.5) plates. Also lysis of cells was monitored by microscopy. Table 1 shows the effect of the different microwave treatments on the viability of the two strains in log CFU/ml. A decrease in viability with a 2 log factor could be obtained for both cell types within 20 sec at a power of 650 W, and within 30 sec at 300 W. Organism Power T=0 t=5 t=10 t=15 t=20 t=30 K. Iactis 650W 8. 7 8. 9 8. 5 5. 6 4. 3 L.helveticus 650W 8. 6 8. 6 8. 6 8. 4 4. 5 K. Iactis 300W 8. 7 8. 6 8. 5 5. 1 L.helveticus 30OW 8. 6--8. 4 7. 7 5. 3

Table 1: shows the viability of Lactobacillus helveticus and Kluyveromyces lactis after microwave radiation applied for different lengths of time (t=0 to t=30 seconds) at different power levels (300W and 650W). The viability is given in CFU/ml.

Example 2 Determination of intracellular peptidase activity after attenuation of starter cultures by microwave treatment The impact of attenuation by microwave treatment on proteolytic activity was systematically followed by measuring the increase in free amino groups during proteolytic degradation of (3-casein by an attenuated cell-free extract. Since the proteolytic activity measurements as described by Doi and co-workers (Doi et a/ 1981, Anal. Biochem. 118,173-184) could not properly be carried out in whole cells, cell-free extracts were prepared, which were subsequently exposed to microwave radiation. It is reasonably expected that the enzyme activity as measured in radiated cell free extracts is an underestimation of the enzyme activity that would be present after radiating whole cells.

Cell free extracts were prepared by disruption of the cells at 1000 Bars in a refrigerated French press. Undisrupted cells and cell debris were removed by centrifugation and the collected supernatant was filter-sterilised. The protein content of the cell free extracts was determined according to Lowry. (Lowry O. H., Rosebrouch N. J., Farr A. L., and Randall R. J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193,265-275) P-Casein was purified from milk by methods known in the art [PCT/FR 91/00506] and subsequently partially hydrolyse using endoproteases trypsin (EC. 3.4.21.4) and chymotrypsin (EC. 3.4.21.1) until a hydrolysis degree of 16% was obtained as measured by free amino-group determination as described by Doi and co-workers (Doi et al. 1981, Anal. Biochem. 118,173-184). The peptide mixture obtained in this way contained 34 different peptides with molecular weights ranging from 373 to 3864 D as determined by mass spectrometry.

Proteolytic degradation of the (3-casein peptide mixture was performed by incubating 1 ml of 0.7 mg/ml beta-casein together with 125 g of cytoplasmic protein of L. helveticus or K. lactis at 24°C in a buffer of 50 mM sodium phosphate at pH 5.7. Samples were taken after 10 or 24 hours, after which the increase in free amino groups was measured as described by Doi [Doi et al. 1981, ibid].

It was thus shown that the microwave treatment had no significant effects on the peptidase activity of K. lactis, whereas the peptidase activity of L. helveticus had dropped with 16%. Furthermore, the release of intracellular constituents during the microwave treatment was determined (not shown) and found to be non-significant. Organism Microwave treatment t=0 hours t=10 hours t=24 hours K. lactis Control 0. 43 1. 73 2. 68 K. lactis 13 sec. at 650W 0. 48 1. 57 2. 52 K. lactis 25 sec. at 300W 0. 47 1. 58 2. 57 K. lactis 30 sec. at 300W 0. 48 1. 71 2. 57 L.helveticus Control 0. 36 2. 38 3. 5 L. helveticus 17 sec. at 650W 0. 35 2. 27 3. 3 L. helveticus 25 sec. at 300W 0. 33 2. 27 3. 39 L. helveticus 30 sec. at 300W 0. 29 1. 71 2. 94

Table 2: Release of free amino groups (in mM) during the kinetic degradation (t=0, t=10 and t=24 hours) of a p-casein hydrolysate by microwave-treated cell free extracts from Lactobacillus helveticus and Kluyveromyces lactis.