Background of the Invention

The present invention relates to the production of frozen products, particularly foodstuffs and other biological products.

Hitherto, in order to achieve satisfactory freezing of foodstuffs and other biological products, freezing operations have had to be operated at temperatures well below those at which such products food would theoretically be expected to freeze in order to avoid problems of super¬ cooling. Such super-cooling can occur in both intra-and extra-cellular moisture present in foodstuffs and other biological products and leads to a failure of the product to freeze satisfactorily. While the degree of super-cooling may vary substantially from batch to batch, the risk that super¬ cooling may occur has led food freezing practitioners to operate plants at temperatures as low as - 0°C. in order to avoid such risks. Furthermore, in such operations, foods or other biological products to be frozen are typically kept at such low temperatures for excessive periods of time to ensure that problems resulting from super-cooling are avoided. Not only does the requirement to use very low temperatures for prolonged periods result in very high energy costs, but also a further problem that arises from super-cooling is a tendency for ice crystals in the frozen foodstuff to grow to a large size thereby diminishing the organoleptic and textural properties of the frozen food. Similarly, the thaw drip loss (normally regarded as a measure of the quality of the frozen

product) may increase if excessive super-cooling occurs during freezing. Conventional practice has been to try to avoid ice crystal growth resulting from super-cooling by effecting quick freezing of products. Alternatively, use has been made of additives such as emulsifiers in ice cream or cryoprotectants such as glycerol or sugar in other products. These techniques, however, require cooling to very low temperatures with the consequence that such techniques have high energy requirements.

The effects of certain ice-nucleating active agents produced by various bacteria on the nuclβation of ice crystals have recently attracted interest, primarily because of their potential for producing artificial snow for ski slopes and because of their role in frost damage of growing plants. In particular, the ice nucleating properties of Pseudomonas syrinαae have been widely reported. (See, for example, Maki, et al Applied Microbiology, September 1974, p. 456). In recent years, these properties have been investigated with particular emphasis on the role played by Pseudomonas syrinαae or a protein derived therefrom in causing frost damage to growing plants (see for example indow, Plant Dise s , March 1983, p. 327). Considerable press attention has been given to the allowability of experiments in the open environment involving mutants of Pseudomonas syrinσae that have been mutated to eliminate the gene that is responsible for the production of ice-nucleating protein. It is hoped by allowing these mutants to compete with Pseudomonas syrinαae that are naturally present on plants that the population of ice nucleating Pseudomonas syrinαae may be reduced (see, for example, Lindow, et al Phytopathology Vol. 76, No. 10, p. 1069 Abstract 95 (1986).

Similar ice nucleating properties have been reported for several other microorganisms including Erwinia. herbicola (see, for example, Lindow, et al Phytopathology 21 1097-1106 (1983)) and Kozloff, et al J. Bacteriology Jan. 1983 p. 222- 231) , Pseudomonas fluorescens (see, for example, Phelps, et al

J. Bacteriology Aug. 1986 p. 496-502)) and Corotto, et al (EMBO Journal ^' 231-236 (1986)) and Xanthonon a Campestris (Deriβ and Schaail, Phytopathology 2i (10) p. 1117 (1986), Pseydofflonflg viridflava (Paulin, et al Proc 4th Int. Conf on Plant Pathogenic Bacteria Vol. 2 INRλ Beaucoaz* France 1978 Vol. 2 p. 725-731 and Anderson and Aβhworth, Plant Phyβiol Vol. 80 pages 956-960 (1986)).

A number of investigations of the specific genes or proteins produced by them that are responsible for ice nucleation have been reported. Tor example. Green and Warner in a letter to Nature 3_12 p. 645 (1985) describe the determination of the sequence of the ice nucleation gene from pseudomopas syrinαae which they called inaZ. They noted that this contains several repeats of a sequence reiteration with the consensus repeat having the sequence

GCCGGTTATGGCAGCACGCTGACC, the gene having a total size of 4458. They also investigated whether deletion of fragments of inaZ affected the ability of the gene to produce an ice nucleating protein by inserting such modified genes into £ Coli. It was found that in the case of deletions that did not result in a fra eshift, ice nucleation properties were retained in a number of cases.

Corotti, et al in The EMBO Journal 5_ p. 231-236 (1986) describe a DNA fragment of 75 kb obtained from Pseudomonas fluorescena that is capable of imparting ice nucleating activity to E Coli. They designated their gene inaW. They also investigated the activity of inaW mutants and determined that insertions into a particular 3.9 kb sequence had particular effect on the activity of the gene. They concluded that the product of the gene, postulated to be a protein of about 180 kd molecular weight, was necessary to confer an ice nucleating (INA ⁺ ) phenotype on ^~ coli.

Kozloff, et al in J. Bacteriology 153 p. 222-231 (1983) hypothesize that the ice nucleating activity of pseu omonas syrlnσae and £ herbicola stems from the presence of nucleating sites on their cell walls. Their results

iΛdicatβ that Pseudomonas syrinαae probably have 4-8 sites per cell and £ herbicola 2 sites per cell.

Kozloff, et al in Science 226 845-846 (1984) suggest that a lipid phosphatidylinositol is present as well as a protein at ice nucleating sites or Pseudomonas syrlnαae and Erwinia herbicola. current thinking, however, now doubts this hypothesis.

Wobler, et al in Proc. Natl. Acad. Sci. USA 8J.7256- 7260 (1986) report the isolation of an ice nucleation protein from £. coli that have been transformed with a plasmid containing the inaZ gene obtained form Pseudomonas syrinαae. The protein (p 153) has an apparent molecular weight of 153 kd. The amino acid content of pl53 corresponded closely with that predicted as the product of inaZ and the opening sequence of pl53 (Met-Asn-Leu-Asp-Lyβ-Ala-Leu-Val-Leu-) corresponded exactly with the sequence that would be predicted to be coded by the inaZ gene. They further reported that a pl80 protein obtained by a similar technique using the inaW gene obtained from Pseudomonas fluorescens apparently had similar sequences and structures to pl53 obtained using inaZ. They suggest that the INA ⁺ phenotype is conferred on both Pseudomonas syrinαae and Pseudomonas fluorescens by such proteins and that these proteins act as templates for ice nucleation even in the absence of the phospholipids Kozloff indicated to be necessary.

In Biophysical Journal 19.293a (1986) Wolber and Warner report that in the ice nucleating protein obtained from Pseudomonas syrinαae. the majority of the sequence consists of interlaced 8, 16 and 48 amino acid repeats. The secondary structure is apparently a B-sheet structure for repeated sequences punctuated by 5-6 turns per 48 amino acid unit. It is suggested that the protein folds into a regular structure built up from the 48 amino acid repeat and that this structure presents hydrogen bonding side chains that mimic the ice lattice.

Phelps, et al in J. Bacteriology 167 p. 496 (1986)

indicate the proteinaceous nature of the ice nucleating material obtained from Erwinia herbicola. reporting that cell free ice nucleating agents could be obtained from outer membrane residues and that such agents could induce nucleation of water at temperatures in the range -2 to -10°C. Makino in Ann Phytopath Soc. Japan IS. 452-457 (1982) reports that ice nucleation activity was found in Pseudomonas arσinalis but to a lesser extent than in Pseudomonas syrinαae.

In addition to studies on ice nucleating bacteria as noted above, studies have also been made of ice nucleation having a biogenic origin in other organisms. Thus, Duman and Horwath in Ann. Rev. Physiol. 4_5 261-70 (1983) review the role of ice nucleation proteins in the hemolymph of certain beetles in imparting freeze-tolerance to the beetles by inhibiting super-cooling and ensuring freezing of extra-cellular fluid at fairly high temperatures thereby reducing the risk of intra- cellular freezing. Such proteins are reported to exist in Vespula. (which have 3-6 ice nucleation proteins) and Dendroides (having 1-2 ice nucleation proteins) . The nature of the ice nucleating protein produced by Vespula maculala was investigated by Duman, et al in J. Comparative Physiology B 15 79-83 (1984) .

Ice nucleating agents have also been reported as being produced by Heterocapsia niei (a marine dinoflagellate) by Fall and Schnell in J. Marine Research Vol. 43 pages 257- 265 (1985).

Each of the above cited articles is incorporated herein by reference.

The use of bacteria having an INA ⁺ phenotype fErwinia ananas) to cause texturization of certain proteinaceous foods has been suggested by Arai and Watenabe in Agricultural and Biological Chemistry 5_0_, 169-175 (1986). They added cells to aqueous dispersions or hydrogels of proteins and polysaccharides to convert bulk water into directional ice crystals at temperatures between -5 and 0°C. to form ariisotropically textured products. Such products were

produced using raw egg white, bovine blood, soybean curd, milk curd, aqueous dispersions or slurries of soybean protein isolate, hydrogβl of agar, corn starch paste and hydrogelβ of glucomannan and calcium glucomannan. Such products are slowly cooled to -5°C. in an air bath in the presence of the bacterial cells. The frozen products were then vacuum freeze dried at -30°C. before setting by steaming to form flake-like textures. By slicing the products at right angles to the plane of the flakes, a textured product may be obtained.

When used herein the term ice nucleating phenotype (INA ⁺ ) means an ability to induce nucleation of super-cooled water at a temperature above -20°C. For example, P. syrinσae PV pisi has a nucleation temperature of -2.9°C. according to Makino (Ann Phytopath Soc. Japan Vol. 48 pages 452-457 (1982)).

The term "non-toxic microorganism" when used herein means microorganism having no adverse effect on humans when ingested in the amounts that are likely to result from the use of such microorganism to assist in ice nucleation.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide a means for freezing foodstuffs and other biological products that are frozen to preserve them, which reduces energy requirements for freezing by applying to said foodstuff or other biological product a non-toxic microorganism having an INA ⁺ phenotype or an ice nucleating agent derived therefrom or a functionally equivalent agent so as to enable freezing to be effected without excessive super-cooling.

It is a further object of the invention to reduce the time taken to effect freezing typically of foodstuffs or other biological products.

It is now thought that in freezing foodstuffs and other biological products it is desirable to minimize the time that samples being frozen are held in the range 0 to -5°C. since this is a temperature range in which many destructive enzymes are particularly active. It is an object of the

present invention to increase the consistency of freeze times in freezing foodstuffs and other biological products so as to minimize the risk of prolonged exposure to these temperatures.

It is a further object of the present invention to provide a means for improving the quality of frozen food, such as fish or meat fillets by avoiding super-cooling to a degree that intra-cellular freezing occurs within muscle tissue of such foodstuffs.

A further advantage of the present invention is that it may improve the stability of the frozen product in cold storage where temperature fluctuations may cause deterioration of product quality, thereby improving the shelf life of such product and may allow for freezer to operate at a higher temperature.

Accordingly, from one aspect, the present invention provides a means for freezing solid foodstuffs and other biological products which comprises applying thereto a non- toxic microorganism having an INA ⁺ phenotype or an ice nucleating agent derived from said microorganism or a functionally equivalent agent thereof and lowering the temperature of said solid foodstuff or other biological product to effect freezing thereof.

From a second aspect, the present invention provides a means for providing frozen foodstuffs or biological products which comprises applying thereto a non-toxic microorganisms having an INA ⁺ phenotype or an ice nucleating agent derived from said microorganism or a functionally equivalent agent thereto and lowering the temperature of said foodstuff or biological product to effect freezing at atmospheric pressure and thereafter placing said foodstuff or biological product into cold storage without any further processing steps.

Without wishing to be bound by any theory, I believe that the ice nucleating agent produced by microorganisms having an INA ⁺ phenotype is probably a protein. BRIEF EXPLANATION OF THE DRAWINGS Figure 1 is a representation of an idealized

freezing curve that is referred to for explanatory purposes.

Figure 2 shows the freezing curves for untreated fish muscle and fish muscle treatment according to the present invention.

DETAILED DESCRIPTION OF THE INVENTION

Non-toxic microorganisms having a phenotype INA ⁺ include natural bacteria having such properties, for example, Erwinia ananas. Erwinla herbicola. pgeudoπonag svrinσae and pseudomonas fluorescens. They also include other non-toxic bacteria that are naturally present in foodstuff, for example, species of Lactobacillus such as Lactobacillua bulσarlcus and lactobacillus acldophllus and Streptococcus βuch as Streptococcus lactus and Streptococcus thermophilus that have been transformed by plasmidβ containing InaZ or inaW genes derived from Pseudomonas svrinαae or Pseudomonas fluorescens or the ice nucleating gene of one of the other microorganisms having an INA ⁺ phenotype listed above or by an ice-nucleating sub unit thereof or a synthetic equivalent of such a gene or sub unit. Suitable means for producing βuch plas ids are described, for example, by Green and Warner (c-p. clt) and Wobler (ς~. £_______.) . Introduction of such plasmids into suitable bacteria and cloning of such bacteria may be effected by standard methods of genetic engineering. It is also possible to introduce such genes into eukaryotic organisms such as yeasts. Many yeasts are present in foodstuffs and other biological products naturally. Such yeasts include various Saccharomyces such as Saccharomyces cerevislae. Saccharomyces cartsbeiαensis. Saccharomyces rosei. Saccharomvces rouxli _f and Saccharomvces uvarum. Introduction of genes to impart an INA ⁺ phenotype to such yeasts may be accomplished by means of standard operations such as use of a plasmid shuttle vector containing the gene giving an INA ⁺ phenotype or by use of an integrative DNA transformation involving removal of yeast cell wall, addition of a DNA fragment containing a gene coding for an INA ⁺ phenotype in the presence of polyethylene glycol and then regenerating the cell wall.

Microorganis s having an INA ⁺ phenotype may be employed live, as may often be the case with microorganism that are commonly present in food or dead. In the case when the microorganisms are killed prior to use, for example, by a heat treatment, care should be taken to avoid conditions that are so severe as to denature and deactivate the ice nucleating agent. For example, heating a suspension of the microorganism to a temperature at or near boiling may be appropriate.

Alternatively, ice nucleating protein itself may be used in the process of the present invention. Such proteins may be isolated from any of the microorganisms described above or indeed from other organisms that have been transformed to import an INA ⁺ phenotype, βuch as £_ _» . coli aβ long aβ care iβ taken to isolate the desired protein prior to use. In βuch techniques, the gene coding for the INA ⁺ phenotype iβ cloned and expressed in a suitable microorganism. The production of protein pl53 in £*. Coli is described in Wolber, et al Proc. Natl. Acad. Sci. USA 8.7256-7260 (1986).

Recovery of the ice nucleating agents from the microorganism in which they have been produced may be accomplished by a variety of means. For example, in cases where the ice nucleating agent is expressed from the cell in substantial amounts recovery can be effected by centrifugation or filtration of the cell mass followed by dilation of the resulting pellet with a buffer (such as HEPES or phosphate) followed by refiltration and high speed centrifugation. In other cases, disruption of the cell may be required, for example, by ultra sonication, pressure shearing in a pressure bomb, solid shearing or enzymatic digestion using an enzyme such as a lyβozyme. Following disruption of the cell, the cell mass may be centrifuged to remove undierupted cells and the broken cell membrane containing the ice nucleating agent recovered. This material may then be centrifuged and washed with a neutral buffer prior to fractionation to recover the ice nucleating agent. Such fractionation may be carried out using standard techniques, testing each fraction obtained to

determine the presence of the ice nucleating agent. In appropriate cases more than one fractionation, possibly using different techniques may be required. Suitable techniques include

(a) Liquid-liquid extraction using organic solvents with varying polarity. This will remove leββ polar fractionβ and leave the proteinβ in the aqueous phaβe.

(b) Gel filtration chromatography - e.g. Sephadex G-200 and other molecular sieve packings will separate proteinβ based on their molecular size. Addition ef sodium dodecyl sulphate (SDS) may be necessary to denature the tertiary and/or quaternary structure of proteins (if not already done) . If SDS iβ used, dialysiβ of the resulting fractions will be necessary to remove the SDS.

(c) Ion exchange chromatography will separate the proteins based on their ionic charge. This technique can alβo be used for concentration, as dilute βolutionβ of a protein can be loaded onto a column until the column iβ saturated, and then an eluting ion will strip the column of the protein.

(d) Hydroxylapatite chromatography -

(e) Preparative high pressure liquid chromatography using of the various types of column chromatography mentioned above with the addition of several others, e.g. reverse phase separations, under high pressure. This can result in greater speed and better resolution between fractions.

Foodstuffs that may be frozen by the process of the present invention include muscular foods βuch as fish and meat. For best results, however, the fish or meat should not be in pieces having excessive thickness. For example, the technique of the present invention is more suitable for use on fish fillets or steaks, or scallops of veal or beef steaks or poultry or pork or lamp chops than on a whole tuna fish or a side of beef. Satisfactory results have, however, been obtained when freezing certain types of fish whole (for example, trout and flat fish βuch as sole, flounder and plaice) . Desirably, the flesh (whether it is fish or meat)

βhould be pre-cut to a thickness no greater than 3 to 4 cm, preferably lesβ than 1.5 cm in order to allow the ice nucleating microorganism or agent to penetrate the flesh. As an alternative to cutting the flesh into small pieces, it may be possible to obtain satisfactory results with large pieces into with deep cuts have been made. The process of the present invention is alβo uβeful for freezing comminuted meat or fiβh flesh such as sausages, ground beef and fish cakes. The techniques of the present invention may alβo find use in freezing of berries and other fruits and vegetables such aβ peas and corn that tend to suffer on freezing as a result of intra-cellular freezing. Again, however, best results are obtained on berries rather than, for example, on tomatoes, or other large fruits. The process of the invention may also be used for freezing other solid foods such aβ shell fiβh, food products βuch as pasta and bakery products such aβ cakes, bread and waffles. The invention may also be used to expedite the freezing of preprepared frozen meals such as so-called TV dinners and frozen snacks such as pizzas.

The process of the present invention may also be of use in the production of ice cream and other dairy products and similar products, for example, those based on tofu. Additional products that may be frozen in accordance with the present invention include fruit juice concentrates, frozen fruit bars such as popsicles, soy and other sauces, soups, yogurt and other such foods, for example fruit and vegetable purees. In such cases, in contrast to the process used by Aral and Watanabe described above, the ice nucleating microorganism or protein is mixed with the necessary ingredients and the whole is chilled en masse until it freezes. Such operations are typically carried out under normal atmospheric pressure and the products are not processed in any other way prior to storage.

In addition to such food products, the process of the present invention may also be of use in freezing other products that are typically frozen for storage such as blood

productβ and other biological productβ βuch aβ sperm, ova embryos and other tissue. For example, the present invention may be used to freeze the sperm or ova of farm animals for breeding purposes.

In freezing food products according to the present invention, conventional freezer operations, βuch aβ blaβt or contact plate freezerβ or vacuum freeze drierβ may be used. However, typically they may be operated at temperatures 10 to 20°C. above those normally considered to be appropriate. Thus, satisfactory results may be obtained using the process of the present invention by operating a blast freezer at temperatureβ in the range -5 to -30°C. inβtβad of -20 to -40°C. aβ iβ currently typical. Similarly contact plate freezing may be operated at temperatureβ in the -15 to -20°C. range in contraβt to the -30 to -40°C. range which iβ normally used at present. Operating according to the present invention may also enable satisfactory freezing to be obtained in a reduced period aβ compared to present methods. The time taken will, however, depend upon the thickness of the foods items being frozen and also the packing of these items in the freezer.

The amount of material to be added to assist in ice nucleation according to the present invention will depend upon the nature of the material used and in particular on the temperature at which it will induce ice nucleation in super cooled water.

Since the ice nucleating microorganism or agents effectively act as catalysts for the nucleation of ice, only relatively small quantities are required to be applied in the foodstuff or other biological product that is to be frozen.

As noted above, the number of nucleation sites on a bacterium varies according to species (Pseudomonas syrinσae having 4-8 sites per cell and E. herbicola 2 sites per cell) . Simple experiments will, however, enable one skilled in the art to determine an appropriate amount of bacteria having an INA ⁺ phenotype to be used in a particular case.

P. syrinαae βuspensionβ exhibit ice nucleating activitieβ in distilled deionized water at concentrations of about 10 ⁷ colony forming unite (c.f.u.) per ml and above. A concentration of 10 ⁷ c.f.u./ml iβ roughly equivalent to 1 mg dry weight of cell/ml. When uβing P. βyrinσae in the present invention, I have found it convenient to use βuspensionβ of bacteria having an aerobic plate count (APC) of 10 ⁵ to 10 ⁸ c.f.u./ml. Concentrationβ of bacteria in a suspension may be determined by absorbance of the βuβpension of light at 520 mm wavelength or other turbidimβtric methods.

I have found that for the following foods, using a bacterial suspension having an aerobic plate count of 10 ⁷ c.f.u/ml, the following rates of application are suitable:

Food Bacteria having Amount

U ± phenotype

Salmon Pseudomonas syringae 0.02-0.2 ml βuspenβion/g fish flesh

Trout Pseudomonas syringae 0.02-0.2 ml βuspenβion/g fish fleβh

If one uses an isolated ice nucleating agent lower amounts of material need to be added to obtain the same concentration of nucleation siteβ. Thus, application of such agents in amounts of at least 0.0005 ag/g of food or biological products, for example 0.0015 to 0.01 mg/g of food or other biological product activities will normally be effective, although higher or lower amounts, for example, up to .1 mg/g may sometimes be useful.

The ice nucleating microorganisms or agents are added to foodstuffs or other biological products in ice nucleating amounts. Typically, for meat, fish and vegetables bacteria will be applied at a rate of at least 10 ^s cfu/g, for example in the range Ϊ0 ⁶ -10 ⁷ and in the production of ice cream or frozen tofu confectionery at a rate of at least 10 ⁵ cfu/ l. If an ice nucleating protein such as pl53 is

used, this may be applied at a rate of from 0.001 to 0.01 mg/g or more.

Said microorganisms or ice nucleating agents may be applied to foodstuffs and other biological products in any conventional way for example, for example, by spraying with aqueous diβperβionβ of the ice nucleating material or, in the caβe of productβ βuch aβ ice cream, fruit juiceβ, pureeβ and the like by mixing the ice nucleating microorganism or agent intimately with the foodstuff that iβ to be frozen.

The process of the present invention iβ illustrated by the following Examples:

GENERAL

Source of Ice Nucleating Active Bacteria - Pseudomonas syrinαae pv pisi was obtained from American Type Culture Collection (ATCC) (Cat. # 11043, Rockvillβ, MD) . This bacteria was chosen aβ it iβ reported to be the most potent of the ice nucleating bacteria. £. syrinαae were revived, check for purity and βtored.

Maintenance of bacteria - Bacteria were maintained on NAG plateβ and brothed and replated on a weekly basis. Three to seven day old (unless otherwise stated) colonies from NAG plates were routinely used as a source of P.. βyrinαae.

Preparation and application of bacteria - The bacteria were βuspended in distilled water and the turbidity of -the suspension adjusted to an absorbance of 1.0. One milliliter aliquots of the bacterial suspensionβ were applied to the samples in the test situations by either adding to the surface of the fish (minced or whole) or by mixing into the fish (minced) . All suspensions were made on the day of use.

Fish used - Fish used in this study were obtained from the Department of Fisheries, Aquaculture and Pathology University of Rhode Island ponds. Atlantic salmon (Salmo salar) or rainbow trout fSalmo σairdnerii were anesthetized using tricain methane sulphonate MS-222 fish anesthetic (Argent Chemical Laboratories Inc. Redmond, Wλ) , killed by

bleeding the artery and then eviscerated and headed. They were stored at 2-4°C. in βealable plaβtic bagβ until use. An alternative source of live rainbow trout was a local supermarket. These were killed by a blow to the head and then similarly stored.

Sample Preparation - Fiβh were filleted and βkinned by hand. Fillets were then prepared for either minced fish or whole fish experiments. For minced fish experiments, each fillet was diced finely using a sharp knife. Then fish muscle was put into a plastic centrifuge tube, and the tube placed into an assembly that allowed reproducible placement of thermocouple probe for subsequent freezing curve determination. For whole fish experiments each fillet was cut, perpendicularly to the backbone, into 4cm pieces. A piece was then placed in an aluminum dish, and the dish likewise placed into an assembly that allowed reproducible placement of thermocouple probes for monitoring of the freezing curve. Both assemblies accommodated 4 samples.

Freezing Curves - Freezing curves were obtained using a Linβeiβ Model 7040 4 channel chart recorded. Either premadβ Type T copper-constantin thermocouple probes (Model TJ36-CPSS-116G-6, Omega Engineering, Connecticut) or lab made Type T copper-constantin thermocouples, both continually referenced against an ice bath, were used for all temperature measurements. Thermocouple junctions were welded using Helium thermocouple welder. Millivolt readings taken from the chart were converted to temperature (degrees Celsius) by use of a Basic program written by the author. Thermocouple probes were placed into each of 4 samples (minced or whole fish) and the assembly placed in a custom made chest freezer (Scientemp Corp. Adrian, Michigan) . This freezer allowed user selection of temperatures from 0°C. to -20°C. with a ⁺ 0.4°C. fluctuation of the chosen temperature. The inner chamber was double insulated and equipped with a fan for circulation. The lid of the chamber had a 12" square triple plated window to allow direct viewing of the specimen. An interior light was

provided for illumination. Samples were placed in the chamber set at -5°C , -10°C. or -20°C. and the freezing curves monitored at these temperatures .

Definitions - Figure 1 βhowβ an idealized freezing curve shoving actual temperatures in a sample against time when placed in a freezer at -5°C. The following times and temperatures were recorded for each of the 4 channels from the freezing curve:

Freezing Point - Defined aβ the temperature at which freezing occurred.

Nucleation Temperature - Defined aβ the minimum temperature attained by the sample before freezing occurred.

Supercooling - Defined aβ the temperature difference between the freezing point and nucleation temperature of the sample.

Nucleation Times - Defined as the time elapsed from the instant the sample passed through its freezing point temperature after freezing had occurred to the time at which nucleation commenced.

Freezing Times - Defined as the time elapsed from the instant the sample passed through its freezing point temperature to the instant it reached -5°C.

EXAMPLE 1

Minced fish samples prepared as described above were divided into control groups and groups to which Pseudomonas syringae dispersion of a concentration of 10 ⁷ c.f.u./ml was applied. The treated groups were sprayed with about 0.1 ml of dispersion per gram of minced fish. Freezing curves for the treatment and control samples were determined as described above. The results obtained are set out in Table 1.

TABLE 1 - FREEZING CURVES FOR TROUT AND SALMON

Freezing Curve for -5 .

Fiβh Preparation - Minced

Application Method - Surface er or to tube of lOg fiβh

--Time from killing to freezing

Exa ple 2

The procedure of Example 1 waβ repeated uβing eight βampleβ of minced βalmon muβcle, these being subjected to a freezing environment of -5°C. Freezing curves were followed throughout the freezing cycle for these βampleβ. The results (together with standard deviation in parentheseβ) are βet out in Table 2 (total freezing time iβ the time taken to go from • 1°C - the appropriate freezing point of βalmon muβcle - to - 5°C, it therefore includeβ nucleation time) TABLE 2

Sample Super Nucleation Total Freezing Cooling Sc Time (min) Tifte [ la

Control -3 .7 (0.9) 144 .7 (155. 7 ) 332 . 3 ( 159 . 6)

Teβt -0.6 (0.3) 6.1 (0.3) 219.4 (9.9)

Further results obtained in the same way are aβ followβ:

Degree of βupercooling and nucleation timeβ for 8 βampleβ thoroughly diced and mixed βalmon four of which were treated with P. Syrlnoaa aβ in Example 1 and four of which were untreated. The samples were taken from the same fillet and subjected to a -5°C environment (Run 1) . Run 2 represents the same samples having been thawed and refrozen. The results are set out in Table 3.

Sample

RUN RUN

DNF* 5.6 5.6 417 5.2 5.6 42 5.6 4.9 DNF 5.6 5.6 4.8 0.4 0.4 4.2 0.2 0.3 5.4 0.4 0.4 4.8 0.2 0.4

* Does not freeze

Table 4 - Freezing parameters for whole salmon when subjected to a -7°C environment* were determined with and without p. syringae.

Sample Temp.of Super- Nucl. Plateau Slope Freeze Total Freezer cooling Time Time Time Time Time f°Cϊ fPpi fmins) (minsϊ fmins) fmins) tminsϊ

Control -7.0 -3.9 68.9 74.4 41.9 116.3 185.2

(±0.1) (±1.5) (±80.1) (±13.6) (±7.4) (±18.7) (±67.5)

Test -7.0 -0.8 4.6 86.1 43.6 129.7 134.3

(±0.1) (±0.4) (±1.0) (±10.2) (±4.1) (±13.0) (±13.2)

All data are averages (standard deviations in parentheses) of 8 values

-20-

Exa ple 3

The methods of Example 1 were repeated but using fish illets prepared aβ described in the -introduction to the Exampleβ. In thiβ caβe for the teβt samples, the diβperβions were applied at a rate of 0.1 ml dispersion of concentration 10 ⁷ c.f.u./ml per gram of fish.

The results obtained were aβ βet out in Table 6 and shown in Figure 2 for βalmon. - °

Same as above except added P.Syringae after thawing and then refroze

Same lot of muscle tissue as used above, but different sample taken for freezing

-^Time in days from kill to freezing.

Exanplt 4

The ability of P. βyrinαaa to induce ice nucleation in the following liquid foodstuffs waβ teβted by adding 1 ml of a dispersion of P. βyrinσae having a content of 10 ⁷ c.f.u./ml to 20 ml of liquid foodstuff and comparing the nucleation and freezing temperatures to βampleβ treated in the βame way, but to which no bacteria had been added waβ tested and the result set out in Tables 6, 7 and 8.

TABLE 6 - (FREEZER SET AT -6°C.)

Sample With Bacteria Added Control Average Nucleation Average Freezing Average Nucleation Average Freezing T mpt --SΛ Temp. ^C. Te»p, gg_« Temp. 3c.

Whole Milk -1.9 ± 0.44 ^• 0.53 t 0. 5 -6.13 ± . 12 -.63 ± .15

Low Fat Milk -2.8 ± 0.36 ^■ 0.5 -5.07 ± .71 -0.53 ± 0.25

Whipping Cream -0.88 ± 0.095 ^• 0.55 t 0. 1 -5.45 ± 1.83 -0.59 t 0.11

Non-Dairy Creamer -2.85 ± 0.71 -0.86 t 0.09 -6.35 ± 0.3 -1.0 + 0.22

Ice cream

(French vanilla) -5.9 t 0.02 -3.9 I 0. 1 DID NOT FREEZE DID NOT FREEZE t t

TABLE 7 - (FREEZER SET AT -10°C.)

Sample With Bacteria Added Control Average Nucleation Average Freezing Average Nucleation Average Freezing Temp. SC. T-9WP. fig Temp, Sg, T BP. °^_

TABLE 8 - (FREEZER SET OF -16°C.) ample With Bacteria Added Control Average Nucleation Average Freezing Average Nucleation Average Freezing TeBP. £« Te»p, 2QΛ TSύSBjj- Qjx Temp, gc.

rapefruit uice -9.4 -7 -12.2 -6.9 concentrate- ¹ range herbet -8.5 -4.2 -9.15 -4.9 pple uice ^■ 12.8 -9.9 -16.3 -9.9 oncentrate ² *

I t *>.

I . 37.4% sugar . 41.9% βugar

Substantially similar results were obtained when using a dispersion of P. Syringae that had een subjected to microwave heating in a domestic microwave oven for 3 minutes.

Exaaple 5

The ability of P. βyringae to induce ice nucleation in model βolutions typically used in food preparation containing βalt and βtarch waβ tested in the same manner aβ the liquid foodstuffs in Example 4. The reβultβ are βet out in Table 9.

IΔ£LE_2

With Bacteria Added Control Average Nucleation Average Freezing Average Nucleation Average Freezing TenP _« ^Q ς^ Temp. ^' (^ emp. ^Q QΛ Temp. ~C.

-4.9 -0.3 DID NOT FREEZE

-1.8 -0.6 DID NOT FREEZE -4.7 -2.1 DID NOT FREEZE -6.0 -3.6 DID NOT FREEZE -2.53 -1.15 -5.1 1.27

-2.3 -1.4