The present invention relates to a process to prepare these stabilized wines.

The presence of tartaric salts, potassium hydrogen tartrate (KHT) and calcium tartrate (CaT) is one of the major causes of instability of wines. Tartaric acid is the main organic acid produced by the grape berry during its development. It is solubilised as potassium and calcium salts into grape musts during the processing of berries. During the fermentation, tartrate salts are preserved by the yeast but their solubility decreases with the increase of ethanol concentration due to the fermentation of sugars.

In young wines, potassium hydrogen tartrate is always present in supersaturating concentrations and crystallises spontaneously. After bottling of wines, the KHT instability may become a major commercial problem due to the unpredictable character of crystallisation and the presence of unattractive crystals in a bottle is often considered as an inferior quality by consumers.

Physical treatments can be used prior to bottling of the wine to prevent crystallisation of tartrate salts. These treatments consist in promoting the crystallisation by cooling the wine at -4°C or in elimination of the potassium and tartaric ions by electrodialysis or by the use of ionexchange resins. However, these time and energy consuming processes are supposed to alter the colloidal equilibrium of wines.

The alternative to physical treatments of wines is to use additives which prevent the nucleation and/or the growth of KHT crystals. An example of additives that prevent the nucleation of crystals are mannoproteins, as described in WO 96/13571 and in WO 97/49794. However, since they have no effect on the growth of added crystals, they cannot guarantee the complete stabilization of wine. Consequently, treated wines are in fact unstable. In addition, the average mannoprotein concentration in wine is not sufficient to guarantee complete prevention of crystallisation.

Carboxymethyl cellulose and meta-tartaric acid belong to the group of additives which inhibit the growth of KHT crystals. Unfortunately, carboxymethyl cellulose has not been accepted by the wine community due to its presumed negative organoleptic effect on treated wines and meta-tartaric acid is unstable at the pH of wine and at the temperature at which wine is stored. Over time, the meta-tartaric acid will hydrolyse and its protective effect disappears. Therefore, its use is restricted to low quality wines for quick consumption. Another drawback is that ideally an additive should be a natural component of wine. This is definitely not the case with meta-tartaric acid or carboxymethyl cellulose.

A. Vernhet et al (Am. J. Enol. Vitic. (1999) 50, 391-397) describe a study in which several polysaccharides and phonetic acids are identified as natural components of white wine that inhibit the crystallization of tartrate salts. V. Gerbaud et al (J. Int. Sci. Vigne Vin. (1997) 31, 65-83) describe the inhibitor effect of mannoproteins and polyphenols for the crystallization of potassium hydrogen tartrate in white and red wines. V. Moine-Ledoux & D. Dubordieu J. Sci. Food Agric. (1999) 79, 537-543 describe the use of mannoproteins extracted from yeast cell walls for stabilizing aged white wine.

Natural additives which are active on both nucleation and growth rate of crystals are by far the best guarantee for long-term stability. Until now, no such additives have been available.

The present invention relates to the use of an oligonucleotide or polynucleotide containing preparation to produce stabilized wine, to oligonucleotide or polynucleotide containing preparation for stabilizing wine and to a process for preparing such preparations as defined in the appended claims.

The present invention relates to a process for the production of stabilized wine. This process comprises the addition of an oligonucleotide or polynucleotide containing preparation to wine. In the present context, stabilized wine is wine in which the crystallization of salts of tartaric acid is prevented or retarded, either as a result of the prevention or retardation of the nucleation process or the inhibition or retardation of the growth of said crystals or both. Unstable wine is characterized by rapid crystallization of salts of tartaric acid, which is usually of an unpredictable behaviour.

Surprisingly, it was found that the addition of an oligonucleotide or polynucleotide containing preparation has a stabilizing effect on wine by preventing both the nucleation as well as the growth of KHT crystals.

In the present context, an oligonucleotide consists of 2-10 nucleotides, a polynucleotide consists of more than 10 nucleotides. The sugar moiety in the oligonucleotide or polynucleotide may be ribose, as in RNA, or deoxyribose, as in DNA.

The oligonucleotide or polynucleotide in the preparation, or the complete preparation, may be obtained from organisms, including plants, or be synthesized, e.g. by a nucleotide synthesizer. Suitable organisms include but are not limited to microorganisms, fish and mammals. Preferably, the oligonucleotide or polynucleotide is obtained from a microorganism. This includes but is not limited to bacteria, fungi and yeast. Examples of suitable yeasts are Brettanomyces, Saccharomyces and Kluyveromyces, such as for example Brettanomyces bruxellensis, Saccharomyces cerevisiae and Kluyveromyces lactis. For instance, heat-inactivated yeast may be used to prepare the oligonucleotides or polynucleotides.

Oligonucleotides and polynucleotides may be prepared according to methods known in the art. These methods comprise the culturing and harvesting of the cells, disruption of the cells in order to liberate the cell contents and isolation of the oligonucleotide or polynucleotide containing fraction by precipitation techniques, chromatography and the like.

Both RNA and DNA may be used in oligonucleotide or polynucleotide containing preparations. The RNA and DNA oligonucleotides or polynucleotides may be in any form, be it single stranded or double stranded. Suitable RNA and DNA molecules have a molecular weight between 0.4 and 500 kDa. If RNA is used, preferably RNA with a molecular weight between 3 and 200 kDa, more preferably between 10 and 100 kDa is used. If DNA is used, preferably DNA with a molecular weight between 1 and 340 kDa, more preferably DNA with a molecular weight between 1 and 100 kDa is used. RNA and DNA may be used separately, or in combination. A well known method in the art for determining the molecular weight of the oligonucleotide or polynucleotide fraction is size exclusion chromatography using a column that is calibrated with protein molecular weight standards.

The oligonucleotide or polynucleotide containing preparation according to the invention may consist exclusively or almost exclusively of nucleotides, e.g. for more than 80%. However, it is not necessary to use completely purified or almost completely purified nucleotides, any proportion of nucleotides between 0.1 and 100% of the preparation may be used in a preparation according to the invention. For instance, it is also possible to use a crude preparation obtained from e.g. yeast, which may contain 0.1 to 50% nucleic acids, or even 0.1 to 20% nucleic acids. If the nucleotides are not completely or almost completely purified, the nucleotide-containing preparation preferably contains 0.1-15% nucleotides, more preferably 1-15% nucleotides (all % on a dry matter weight basis). Other suitable constituents of the preparation apart from the oligonucleotides or polynucleotides are e.g. yeast cell wall components, such as for instance mannoproteins, or parts thereof.

The oligonucleotide or polynucleotide containing preparation may be added to the grape must or to wine. It is preferably added to the wine during ageing, i.e. after fermentation but before bottling. The invention is extremely suitable for white wines and rosé wines.

The oligonucleotide or polynucleotide containing preparation is added in such amounts that a stabilizing effect is achieved. In general, one will try to work with as low as possible amounts. Good results are possible if so much preparation is added that a final nucleic acid concentration in the wine is reached of 1 to 400 mg per liter wine. Preferably, so much is added that a concentration of 1 to 200 mg per liter wine is reached, most preferably 1 to 100 mg per liter wine is added. Any insolubles which develop may subsequently be removed using standard techniques. The skilled person will understand that the amount added will also depend on the addition or presence of e.g. other stabilizers.

The nucleation and crystal growth can be measured and quantified by the following methods (Moutounet et al. In : Actualités OEnologiques 1999 Vieme Symposium International d'Oenologie de Bordeaux (Lonvaud-Funel ed.)

Method 1, indicative of crystal nucleation, measures the time of appearance of crystals in the wine when stored at -4°C. A visual inspection is performed daily and the time necessary to detect the appearance of crystals (T_crys) is expressed in number of days. Unstable wine has a T_crys that can vary between 0.5 and 15 days. Stabilized wine is characterized by a T_crys^{stabilized wine}/T_crys^{unstable wine} of >2, preferably >5, more preferably > 10 and most preferably >40.

Method 2, indicative of crystal growth, measures the Degree of Tartaric inhibition (DTI) of the wine. Hereto, wines are stirred at -4°C and the initial conductivity is measured. Subsequently, calibrated crystals of KHT are added and the conductivity is then measured after a stable value has been reached. The DTI is defined as the percentage decrease of the initial conductivity. Stabilized wine is characterized by a DTI^{stabilizedwine}/DTI^{unstable wine} of <0.8, preferably <0.5 and more preferably <0.30.

Method 3, indicative of KHT crystal nucleation, measures the true, dissolved tartaric acid concentration. An accurate volume of the wine is transferred into a glass vial, and mixed with the same accurate volume of D₂O containing a precisely known concentration of maleic acid. The ¹H NMR spectrum is run with conditions of full relaxation, and the integral of the internal standard (maliec acid) is compared with the integral of tartaric acid. In this way the dissolved tartaric acid concentration can be determined with very high precision and accuracy.

Wines Stabilized by the process according to the invention are characterized by a T_crys^{stabilized wine}/T_crys^{unstable wine} of > 2, preferably > 5, more preferably > 10 and most preferably > 40 and by a DTI^{stabilized wine}/DTI^{unstable wine} of < 0.8, preferably <0.5 and more preferably <0.3 and are biologically stabilized. In this context, biologically stabilized means that the stabilization is achieved by the addition of an additive which is generally considered as a natural component of wine, e.g. because it is derived from yeast or grapes. Examples of such biological additives are mannoprotein, DNA and RNA.

The time of the appearance of crystals (T_crys), T_crys^{stabilized wine}/T_crys^{unstable wine} the degree of tartaric inhibition (DTI) of the wine and tartaric acid concentrations were measured as described above.

The concentration of nucleic acids was determined by dissolving a well known amount of the sample in D₂O, the nucleic acids were hydrolyzed by means of RNase and DNase, and the concentration of mononucleotides was measured by 600 MHz NMR after the addition of a suitable internal standard, such as maleic acid.

In all experiments an unstable white wine, Chardonnay from harvest 2000, was used. The concentration of tartaric acid at the onset of this study was 1.92 g/l. All quantitative experiments were carried out by means of NMR as described above. 0.500 ml of wine was added to 0.500 ml of a stock solution of D₂O, containing approximately 5 g/l EDTA and 5.0637 g/l maleic acid disodium salt as an internal standard for NMR measurements.

To test the effect of DNA on the stabilization of wine, herring sperm DNA (Sigma) was purchased. Small volumes were added to 10 ml of unstable white wine, to achieve final concentrations of 0.4, 0.2, 0.1 and 0.05 g/I. The samples were stored for 1 hour at +4°C. After the removal of insolubles by centrifugation at 3000 rpm for 30 min., the samples were stored at -4°C. Results are presented below and show that DNA has a stabilizing effect on this unstable white wine, which wine showed crystals after only 16 hours at -4°C. Days Property evaluated Control stored at -4°C Control stored at ambient T 0.4 g/l DNA -4°C 0.2 g/l DNA -4°C 0.1 g/l 0.2 DNA -4°C 0.05 g/l DNA -4°C 4 Crystals + - - - - - Tartaric acid (g/l) 1.04 1.93 1.91 2.03 1.94 1.96 8 Crystals + - - - - - 11 Crystals + - - - - - 14 Crystals + - - - - - Tartaric and (g/l) 0.89 2.06 1.91 1.95 1.96 1.99 25 Crystals + - - - - - Tartaric acid (g/l) 0.85 1.94 1.89 1.96 1.93 1.96

The DNA was also used for tests with DNA fractions of different molecular weight. 250 mg of the DNA were dissolved in H₂O and passed over an ultra-filtration membrane (MW cut-off 10 kDa), to the retentate water was added twice for complete removal of lower molecular weight components. The retentate was freeze dried, and a yield of 165 mg was obtained, this fraction was labeled DNA-LS (large size). The permeate was passed over a membrane with MW cut-off of 1 kDa To the retentate water was added two times, and the retentate was freeze-dried. The yield was 76 mg. This product was labeled DNA-MS (medium size). The permeate was also freeze-dried, and a yield of 9 mg was obtained. This product was labeled DNA-SS (small size). The low molecular weight fraction was not used for these experiments, because the yield was too low.

Freeze-dried DNA with a molecular weight between 1 and 10 kDa was dissolved in water in a concentration of 10 mg/ml. Small volumes were added to wine as described in Example 1a. Days at - 4°C Property evaluated Control stored at -4°C Control Stored at Ambient T 0.4 g/l DNA-MS 0.2 g/l DNA-MS 0.1 g/l DNA-MS 0.05 g/l DNA-MS 4 Crystals + - - - - - Tartaric acid (g/l) 1.04 1.93 1.90 1.95 1.95 1.94 8 Crystals + - - - - - 14 Crystals + - - - - - Tartaric acid (g/l) 0.89 2.06 1.85 1.94 1.97 2.05 25 Crystals + - - - - - Tartaric acid (g/l) 0.85 1.94 1.89 1.93 1.95 1.92

Freeze-dried DNA with a molecular weight > 10 kDa was dissolved in water in a concentration of 10 mg/ml. Small volumes were added to wine as described in Example 1a. Days at - 4°C Property evaluated Control Stored At -4°C Control Stored At Ambient T 0.4 g/l DNA-LS 0.2 g/l DNA-LS 0.1 g/l DNA-LS 0.05 g/l DNA-LS 4 Crystals + - - - - - Tartaric acid (g/l) 1.04 1.93 1.83 1.89 1.93 1.64 8 Crystals + - - - + - 11 Crystals + - - - + - 14 Crystals + - - + + - Tartaric acid (g/l) 0.89 2.06 1.89 1.75 1.09 0.95 25 Crystals + - + + + - Tartaric acid (g/l) 0.85 1.94 1.62 1.25 0.93 0.83

The results of example 1 a-c show that DNA has a stabilizing effect. Excellent results are obtained with DNA of a molecular weight between 1 and 10 kDa.

To further investigate the effect of RNA in the stabilization of wine, RNA was purchased from Sigma (bakers yeast). RNA fractions were prepared as described for DNA in example 1a. Yields were: fraction MW < 1kDa (RNA-SS): 30mg, 1 kDa < MW < 10kDa (RNA-MS): 92 mg, 1 0kDa < MW (RNA-LS): 105 mg.

The freeze dried RNA fraction with molecular weight MW < 1kDa (RNA-SS) was dissolved in water in a concentration of 10 mg/ml. The samples were added to wine as in example 1a. Days at - 4°C property evaluated Control Stored At - 4°C Control Stored At Ambient T 0.3 g/l RNA-SS 0.2 g/l RNA-SS 0.1 g/l RNA-SS 0.05 g/l RNA-SS 4 Crystals + - - - + + tartaric acid (g/l) 1.04 1.93 1.82 1.88 1.13 1.20 8 Crystals + - - + + + 11 crystals + - - + + + 14 crystals + - - + + + tartaric acid (g/l) 0.89 2.06 1.83 1.03 0.88 0.85 25 crystals + - + + + + tartaric acid (g/l) 0.85 1.94 1.30 0.87 0.89 0.84

The freeze dried RNA fraction with molecular weight MW < 1 kDa (RNA-SS) was dissolved in water in a concentration of 10 mg/ml. The samples were added to wine as in example 1a. Days at - 4°C Property evaluated Control Stored At -4°C Control Stored At Ambient T 0.4 g/l RNA-MS 0.2 g/l RNA-MS 0.1 g/l RNA-MS 0.05 g/l RNA-MS 4 Crystals + - - - - + tartaric acid (g/l) 1.04 1.93 1.83 1.90 1.81 1.08 8 Crystals + - - - + + 11 Crystals + - - + + + 14 Crystals + - - + + + Tartaric acid (g/l) 0.89 2.06 1.88 1.29 0.89 0.85 25 Crystals + - + + + + Tartaric acid (g/l) 0.85 1.94 1.50 0.97 0.89 0.81

The freeze dried RNA fraction with molecular weight > 10 kDa was dissolved in water in a concentration of 10 mg/ml. The samples were added to wine as in example 1a. Days at - 4°C Property evaluated Control Stored At -4° C Control Stored At Ambient T 0.4g/l RNA-LS 0.2 g/l RNA-LS 0.1 g/l RNA-LS 0.05 g/l RNA-LS 4 Crystals + - - - - - Tartaric acid (g/l) 1.04 1.93 1.89 1.94 1.98 1.98 8 Crystals + - - - - - 11 Crystals + - - - - - 14 Crystals + - - - - - Tartaric acid (g/l) 0.89 2.06 1.87 1.95 1.97 1.94 25 Crystals + - - - - - Tartaric acid (g/l) 0.85 1.94 1.89 1.91 1.94 1.94

The results of example 2a- c demonstrate that RNA has a stabilizing effect. The best results are obtained with RNA fractions of > 10 kDa.

Yeast cells of the species Saccharomyces cerevisiae were harvested from the fermentation broth by centrifugation. The concentrated yeast cream is autolysed by adding 2% (w/w) sodium chloride, after which the cream is heated to 52°C for 40 hours. During the autolysis the pH is maintained at 5.4 by using hydrochloric acid and sodium hydroxide. After the autolysis, the remaining insoluble fraction, which consists primarily of the crude cell wall fraction, is separated from the solution by use of a 4-stage disk-stack centrifuge battery. During the separation the solid-fraction is countercurrent washed in order to optimize the yield of the soluble fraction. The separated cell wall fraction with a dry matter content of 12-15% w/w was subjected to the following extraction procedure :

• 1. Cell walls are suspended in water (1:1, v/v) and centrifuged. The resulting "cake" is resuspended to a concentration of 12-15% (w/v) in a solution of 10 mM potassium metabisulfite and 20 mM citrate adjusted to pH 7.0 by the addition of NaOH.
• 2. The cell wall suspension thus obtained is heated at 120°C for 2 hours.
• 3. After cooling, the suspension is centrifuged at 10,000xg for 10 min.
• 4. The resulting supernatant is filtered on a filter press after addition of 2% of Clarcel CBL3 (CECA, France) as filter aid. Residual insolubles are removed by filtration on a one-plate Seitz filter press equipped with a Seitz-EKS filter (0.1 - 0.3 micron cut-off).
• 5. The EKS filtrate is submitted to two successive ultra-filtration steps using two membranes of different molecular weight cut-off. First a 100 kDa membrane (Amicon, USA) is used to remove high molecular weight molecules, the retentate being then discarded. The filtrate obtained from the 100 kDa membrane is then concentrated on a membrane with a cut-off of 3 kDa (Amicon). Diafiltration with water is used to remove salts and low molecular weight contaminants from the sample. The final concentration factor of the retentate is typically 10.
• 6. The final desalted concentrate is lyophilized to give the nucleotide preparation sample MDR1. This preparation is characterized by a contents of 1-1.5% RNA and 1.5-2% DNA.

A freeze dried fraction of MDR1 was dissolved in H₂O to obtain a stock solution of 15 mg/ml. Small volumes of this stock solution were transferred into 10 ml of unstable white wine. In this way final concentrations of MDR1 were obtained of 0.6 g/l, 0.4 g/l and 0.2 g/l. These samples were stored for approximately 1 hour at +4 °C, and all insolubles were removed by centrifugation. Next, the clear supernatants were transferred into a glass vial, and stored at -4 °C. The temperature was kept constant between -2 and -6 °C.The samples were evaluated visually for tartaric acid crystals. The dissolved tartaric acid concentration was determined as described before. Days at -4°C Property evaluated control stored at -4°C Control stored at ambient T 0.6 g/l of MDR1 0.4 g/l of MDR1 0.2 g/l of MDR1 4 Crystals + - - - - Tartaric acid (g/l) 1.04 1.93 1.90 1.92 1.98 8 Crystals + - - - - 11 Crystals + - - - - 14 crystals + - - - - Tartaric acid (g/l) 0.89 2.06 1.91 1.93 1.94 25 Crystals + - - - + Tartaric acid (g/l) 0.85 1.94 1.92 1.94 1.86 These results show that also not completely purified nucleotides stabilize wine.

All nucleotides were removed from MDR1 in the following way: RNase was added to the product, the pH was adjusted to 5.6, and the solution was heated to 65 °C for 5 S hours. The pH was readjusted to 5.6 after 2 hours. The final solution was passed over an ultra-filtration membrane (MW cut-off 2000) to remove all mononucleotides. The retentate was freeze dried and an NMR spectrum was recorded and compared with the NMR spectrum of MDR1. The spectra showed that all nucleotides had been removed.

The freeze dried product was added to unstable wine in the same way and concentrations as described in example 3a . Days at -4°C Property evaluated control stored at -4°C Control stored at ambient T 0.6 g/l of sample 0.4 g/l of sample 0.2 g/l of sample 4 Crystals + - - + - Tartaric acid (g/l) 1.04 1.93 1.83 1.39 1.15 8 Crystals + - + + - 11 Crystals + - + + - 14 Crystals + - + + - Tartaric acid (g/l) 0.89 2.06 1.16 0.99 0.90 25 Crystals + - + + - Tartaric acid (g/l) 0.85 1.94 1.00 0.89 0.83 These results show that the nucleotide-free preparation has hardly any stabilizing effect on the wine compared to the nucleotide-containing preparation.

A nucelotide-containing preparation was prepared as described in Example 3. The sample obtained was labelled MR1.

Molecular weight distribution of molecules in MR1 has been determined by size-exclusion chromatography analysis on a TSK G3000SW column. Soluble molecules in MR1 have molecular weights in the range 5-50 kDa.

The dry matter content of the freeze-dried MR1 is 98%. Analyses of the compounds listed below have been performed by standard procedures or as described. Component Percentage (w/w) Total nitrogen 10.8 Phosphate 6.3 Sodium 2.0 Potassium 1.0 Total ashes 3.0 Total carbohydrates 25 Glucose 0.7 Mannose 22.0 Ribose 6.75 Protein 54 Mannoproteins 77 RNA 15 Molecular weight distribution 20-30 kDa

The RNA content was calculated from the ribose content (6.76%) via the respective molecular weights. The molecular weight of a nucleotide in RNA is 339 Da and of ribose is 150 Da. Therefore, the RNA content is 339/150 x 6.75 = 15%. The protein content was calculated from the total nitrogen (10.8%) corrected for the nitrogen present in RNA (2.2%) via (10.8-2.2) x 6.25 = 54%.

Samples of nucleotide preparation MR1 have been added to wines of different tartaric instabilities (i.e. very unstable and unstable).

A 75 g/l solution of MR1 preparation in water has been prepared. Aliquots of 1, 2, 3, and 4 ml have been added to 0.5 liter of wine to reach respective final concentrations of 150, 300, 450 and 600 mg/I. The results are presented below. MR1

(mg/l of wine) T_crys

(days) T_crys^{stabilized wine} /

T_crys^{unstable wine} DTI

(%) DTI^{stabilized wine} /

DTI^{unstable wine} 0 1 23 300 43 43 nd nd 450 64 64 16 0.7 600 90 90 7 0.3

They show that the white wine used is very unstable (KHT crystals can be observed already after 1 day of storage at -4°C). After addition of 300 mg/l MR1, the time necessary to detect the formation of KHT crystals increased to 43 days.

This KHT instability of this white wine is characterized by a DTI of 23%. After addition of 450 and 600 mg/l MR1, the DTI value decreased to 16 and 7% respectively..

The results obtained with the two methods on this very unstable white wine demonstrate that MR1 prevents both the nucleation of KHT crystals and inhibits their growth. The treated wines can be considered as stable.

The results for the unstable white wine are presented below. MR1

(mg/l of wine) T_crys

(days) T_crys^{stabilized wine} /

T_crys^{unstable wine} DTI

(%) DTI^{stabilized wine} /

DTI^{unstable wine} 0 11 16.1 450 >120 >11 4.1 0.25

They show that this white wine is unstable since KHT crystals can be observed after 11 days of storage at - 4°C. After addition of MR1 at 450 mg/l, no crystals could be detected after 120 days of storage at 4°C.

The KHT instability of this white wine is characterized by a DTI of 16.1%. After addition of 450 mg/l of MR1, the DTI value decreased to 4.1 %, showing that MR1 prevents the growth of KHT crystals.

The combined use of the two methods on this unstable white wine indicates that MR1 prevents both the nucleation of KHT crystals and inhibit their growth. Treated wines can be considered as stable.