UNIFORMLY MOIST CHEESE

This application claims benefit of the filing dates of United States

Provisional Application Serial No. 60/712,621, filed August 30, 2005, and United States Provisional Application Serial No. 60/775,049, filed February 20, 2006, the contents of which applications are incorporated herein in their entireties.

Field of the Invention

The invention relates to methods for making blocks of cheese, where the cheese has improved moisture content and composition. For example, the cheese is uniformly moist throughout even large blocks of cheese. The cheese also has an increased moisture content to prevent drying, improve shelf life and reduce manufacturing wastes and costs.

Background of the Invention

In the United States, Cheddar cheese was traditionally produced in 18 kg (40 Ib) blocks. In a highly cost-competitive market, more automated and efficient means of handling large quantities of cheese in rapidly expanding cheese factories were developed to control costs. Thus, in the late 1970s and early 1980s, the first 290 kg (640 Ib) block Cheddar production lines were put into production. One 290 kg block replaced sixteen 18 kg blocks. The 290 kg block system reduced labor and handling costs, on-the-job lifting injuries, intermediate packaging costs, and trim loss when blocks were converted to the exact weight pieces needed for retail marketing.

However, although the handling of 290 kg blocks of cheese with forklifts was efficient and easy, the cooling of the cheese in these large blocks immedi- ately after manufacture was more difficult. Thus, as the 290 kg block systems became common in the industry, it became apparent that the cheese within the 290 kg blocks had variations in both composition and cheese quality. For example, in 1988, Reinbold et al. (J. Dairy Sci. 71: 1499-1506) observed that

after 7 days of cooling a 290 kg block of cheese, moisture had traveled from areas of high to low temperature. Reinbold et al. also observed that after 24 hours of cooling, the curd had not completely fused and was still porous.

Barbano et al. conducted systemic studies on 290 kg blocks of cheese and observed that a moisture gradient of about 5% existed from the inside to the outside of the cheese block. J. AOAC Intl. 84: 613-19 (2001). Thus the center of 290 kg blocks of cheese was significantly drier than the outside. Moisture was apparently wicking from the interior to the exterior during cooling of the cheese blocks, leading to irregularities and non-uniformities in cheese composition and quality. Smaller portions of cheese cut for retail sale from these 290 kg blocks were sometimes too wet, or too dry, depending upon what part of the block the retail portion was taken.

Hence, a problem exists in the cheese industry that threatens to undermine the efficient 290 kg block process routinely used for making cheese.

Summary of the Invention

The present invention provides a new approach to making cheese that avoids the wicking, drying and moisture retention problems of existing procedures. The present methods provide a uniformly moist block of cheese with uniform composition and quality. The cheeses produced by the methods of the invention have excellent flavor, melt very well and can be produced to retain more moisture than existing cheeses. The methods of the invention are simple, and require less rennet and less salt than existing procedures.

The methods of the invention relate to controlling the pH of the cheese making process to optimize the partitioning of minerals and proteins between curd and whey, and between the matrix and water phase within curd particles.

Thus, the present invention involves a method for reducing water migration in cheese that includes reducing the pH of pasteurized milk used for making the cheese to apH of about 5.6 to about 6.2, before adding cheese- making starter cultures. The milk can be warmed to a temperature of about 85 ⁰ F to about 100 ⁰ F after the pH is adjusted and starter bacterial cultures can then be added to ripen and begin the cheese-making process. In some embodiments,

the milk is acidified to a pH of about pH 5.80 to about 5.85 when the milk is at a temperature of about 88 ⁰ F to about 95 ⁰ F.

Milk typically has a pH of about 6.6 to about 6.7. Lowering the pH of milk helps the cheese making process and improves the cheese product in a variety of ways. For example, instead of being tightly bound to protein, calcium tends to migrate into the soluble phase and becomes available to rennet, an enzyme required in a later stage of the cheese making process. Moreover, bacterial cultures used to initiate the cheese making process actually grow better under low oxygen conditions, and use of carbon dioxide to acidify the milk tends to drive some of the oxygen out of solution. Such low oxygen and high carbon dioxide levels optimize growth of cheese-making bacteria and inhibit growth of undesirable microorganisms that might otherwise contaminate the cheese- making process. Acidification is also believed to move proteins such as casein into the water phase. An increased protein content in the soluble phase helps to hold water so that the cheese has a higher, more uniform moisture content. Such a uniform increased moisture content helps the cheese to resist drying, promotes a longer shelf life and reduces cheese waste and manufacturing costs. A higher protein content in the soluble phase also helps the cheese to retain salt, not only reducing the amount of salt needed but also reducing salt run-off and the need to safely dispose of salt waste.

Thus, an improved cheese product is produced using the methods of the invention. The improved cheese of the invention is uniformly moist, melts smoothly, has excellent flavor, has somewhat less fat (e.g. 5% to 10% less fat) than cheese made without acidification, and has more calcium and casein in a soluble phase of the cheese than does a cheese made without acidification.

Description of the Figures

FIG. 1 graphically illustrates cheese pH versus time (minutes) during cheese-making for cheeses made from control milk (O) and milk to which CO ₂ has been added (D).

FIG. 2 shows cheese pH as a function of position within slab (n =3), position 1 = bottom 2.54 cm of cheese slab and position 7 = top 2.54 cm of

cheese slab. Control (0), average pH value over 3 weeks and CCVtreated (O), average pH value over 3 weeks.

FIG. 3 shows cheese moisture as a function of position within the cheese slab (n=3), position 1 = bottom 2.54 cm of cheese slab and position 7 = top 2.54 cm of cheese slab. Control (D), average moisture over 3 weeks and CO ₂ -treated (O) ₅ average moisture over 3 weeks.

FIG. 4 graphically illustrates the mean (n = 3) temporal pH pattern of control (squares) and CO ₂ -treated (circles) cultures during cheese making. Samples at times 0 and 45 min were milk, samples from 100 to 140 min were whey, and samples after 140 min were curd.

FIG. 5 graphically illustrates the mean (n = 3) CO ₂ content of control and Cθ ₂ -treatment cheeses during 6 months of aging.

FIG. 6 graphically illustrates the mean (n = 3) pH of the control and CO ₂ - treated cheeses during 6 months of aging. FIG. 7 graphically illustrates the mean (n = 3) titratable acidity of the control and CO ₂ -treated cheeses during 6 months of aging.

FIG. 8 graphically illustrates the mean (n = 3) soluble nitrogen as a percentage of total nitrogen (SNPTN) of the control and CO ₂ -treated cheeses during 6 months of aging. Open symbols indicate pH 4.6 SNPTN and closed symbols indicate 12% TCA SNPTN.

FIG. 9 graphically illustrates the mean (n = 3) ratios of c^-casein to para- κ-casein (open symbols) and /3-casem to para-κ-casein (closed symbols) in cheeses during 6 months of aging. Square symbols are used for control cheeses and circular symbols are used for CO ₂ -treated cheeses. FIG. 10 shows the proteins in expressible serum (ES) (25°C) of Cheddar cheese, immediately after overnight pressing (about 16 h), separated by SDS- PAGE. Lanes 1 to 3 contain expressible serum of control cheeses from three cheese makings. Lane 5 is a whole milk reference sample. Lanes 7 to 9 contain expressible serum from CO ₂ -treated cheese from three cheese makings. Protein bands are identified on the gel.

Detailed Description of the Invention

The present invention provides a method for making a flavorful, uniformly moist cheese that includes acidifying the milk used for cheese making just before the cheese making procedure is initiated. In general, a cheese making process involves milk pasteurization, warming the milk to a temperature of about 85 ⁰ F to about 105 ⁰ F, incubating the warmed, pasteurized milk with starter bacterial cultures to ripen the milk, adding rennet to coagulate the ripened milk, cutting the coagulate into curd, healing the curd by stirring the curd/whey, raising the temperature of the curd/whey suspension to about 90 ⁰ F to about 105 ⁰ F, separating the curd from the whey, salting the curd, pressing the curd into blocks and aging the blocks of cheese as needed. The improvement provided by the invention involves acidifying the milk to a pH of about 5.6 to about 6.2, after pasteurizing the milk and before adding cheese making starter cultures. In some embodiments, the pH is adjusted to achieve a pH of about 5.8 to about 6.1 when the milk temperature is about 85 ⁰ F to about 105 ⁰ F. Cheeses made by the present methods are uniformly moist, melt readily, have somewhat less fat, have more calcium and have an excellent flavor. Thus, this improvement helps eliminate waste and improves the uniformity, composition and moisture content of the cheese product.

Acidification

Any acidifying agent can be used including carbon dioxide, vinegar, citric acid, lactic acid and the like. However, while the invention contemplates any acidifying agent, in some embodiments, carbon dioxide is preferred. Carbon dioxide has certain advantages, including the fact that carbon dioxide has essentially no flavor, carbon dioxide can act as an anti-microbial agent and carbon dioxide temporarily modulates the pH of cheese components during key cheese-making steps and then dissipates over time as the carbon dioxide outgases. Hence, carbon dioxide acts as a processing aide and the majority of the carbon dioxide dissipates and does not form a substantial proportion of the final product.

The pH of the milk should be adjusted after pasteurization and after cooling the milk from the pasteurization process. This is done to avoid any

coagulation that may occur as a result of the combination of heating and acidifying the milk. When the milk is acidified at cool temperatures, no significant amount of coagulation should occur. Thus, a cool temperature is used for acidification. Milk is usually pasteurized by heating at 72° C (161°F to 162 ⁰ F) for 15 seconds to destroy potentially harmful bacteria. Milk is then typically cooled to around 3O ⁰ C (86°F). However, for the acidification procedures of the invention, the milk should have a temperature of no greater than about 10 ⁰ C (50 ⁰ F) before acidification is performed. In some embodiments the temperature is kept below about 7 ⁰ C (44°F to 45°F) before acidification is performed. In other embodiments, the temperature is kept below about 4 ⁰ C (39°F to 40 ⁰ F) before acidification is performed.

The pH of the milk should be reduced from the normal milk pH of about 6.6 to 6.7, to apH of about of about 5.6 to about 6.2, after pasteurization and before adding cheese making starter cultures. In some embodiments, the pH of the milk can be reduced to an initial pH of about 5.7 to about 6.1 , or a pH of about 5.85 to about 6.05, or a pH of about 5.9 to about 6.0 before addition of cheese-making cultures.

The pH can vary somewhat with temperature. Because the milk will be incubated with the starter bacterial culture at about 85 ⁰ F to about 105 ⁰ F, or at a temperature of about 9O ⁰ F to 100 ⁰ F, or at a temperature of about 88 ⁰ F to about 95 ⁰ F, the pH should be measured, adjusted and/or calculated at this temperature. Hence, an initial pH of about 5.8 to about 6.2, or apH of about 5.9 to about 6.0 at about 85 ⁰ F to about 105 ⁰ F is desired. In some embodiments, the temperature is about 90 ⁰ F to about 100 ⁰ F and the pH is about 5.9 to about 6.0. When carbon dioxide is used to reduce the pH of milk, approximately

1000 ppm to about 2000 ppm carbon dioxide are used. In some embodiments, approximately 1300 ppm to about 1900 ppm carbon dioxide are used, or approximately 1400 ppm to about 1800 ppm carbon dioxide are used to achieve the desired pH.

Starter Cultures

Starter bacterial cultures are used to ripen and begin the cheese making process. The starter cultures contain lactic acid producing bacteria is to help

soui' the milk and to convert lactose into lactic acid. This helps in the coagulation process. In addition, the starter cultures also have a beneficial effect on the eventual quality, taste and consistency of the cheese.

Starter cultures typically include live cultures of lactic acid bacteria such as, for example, Streptococcus thermophilics and Lactococcus cremoris bacteria. These bacteria naturally produce lactic acid and naturally lower the pH of the ripening milk used during cheese making. The methods of the invention accelerate the pH lowering process and facilitate bacterial action. Use of carbon dioxide as the acidifying agent minimizes oxygen content in the milk culture, further enhancing bacterial action.

Any available cheese making starter cultures can be used with the methods of the invention. For example, commercially available cheese making starter cultures such as 911 DVS pellets (Chr. Hansen Inc., Milwaukee, WI) can be employed. Ripening by starter cultures can be done for about 30 minutes to about 90 minutes at a temperature of about 85 ⁰ F to about 100 ⁰ F. During this process the pH will typically remain at about 5.6 to about 6.2.

Coagulation

The ripened milk is coagulated by the addition of rennet. The active ingredient of rennet is the enzyme, chymosin (also known as rennin). Any available rennet can be used in the invention. One source of rennet is the stomach of slaughtered newly-born calves. Vegetarian cheeses are manufactured using rennet from either fungal or bacterial sources. Advances in genetic engineering processes have made recombinantly produced chymosin available. Any of these rennet types can be employed in the invention. For example, rennet can be obtained commercially Chymax Extra from Chr. Hansen Inc. (Milwaukee, WI).

The amount of rennet employed can be reduced when employing the methods of the invention. In general, when the pre-acidification methods of the invention are used, the amount of rennet employed can be about one-third to about two-thirds of the rennet used for making cheese without pre-acidification. In some embodiments, the amount of rennet employed after pre-acidification is about one-half that used when no acidification is performed. Hence, the

methods of the invention can be less expensive than currently available methods because lower amounts of rennin can be used.

The exact amount of rennet employed depends upon the activity of the enzyme. When using Chymax Extra, about 0.05 to about 0.2 milliliters per kilogram of milk can be used; in other embodiments about 0.1 mL/kg of milk is used. The temperature employed during coagulation with rennet can vary. In some embodiments, the ripened milk is incubated at a temperature somewhere between room temperature and body temperature, for example, at about 20°C (68°F) to about 37 ⁰ C (98.6°F), or at about 29 ⁰ C (84.2°F) to about 33 ⁰ C (91.4°F). The coagulation time can also vary. In general, the milk is allowed to coagulate for about 10 minutes to about 40 minutes, or about 15 minutes to about 30 minutes, or about 20 minutes to about 25 minutes.

Curd Production After treatment with rennet, the coagulum is cut to form curds of desired size. The curds and whey are not stirred for a short while, for example, about 2 min to about 10 min to allow the curds to heal, ha some embodiments, the curds and whey are not stirred for about 5 min to heal the curds. After healing, the curds and whey are stirred gently without added heat for 10 min. Note that the curds will float when carbon dioxide is used as the acidifying agent. Hence, cheese vats may need to be adapted to allow for curds that float rather than sink.

After a brief resting period, the temperature is increased a few degrees over a 15 minute time period, to a temperature of about 35°C (95 °F) to about 40-41 ⁰ C (105 ⁰ F), or about 37.8 ⁰ C (100 ⁰ F). At the end of this heating process, some whey can be removed, cooled to about 4-5 ⁰ C (40 ⁰ F) and saved for later use. The remaining curds and whey are continuously stirred at a temperature of about 35 ⁰ C (95 ⁰ F) to about 40 ⁰ C to 41 ⁰ C (105 ⁰ F), or at about 37.8 ⁰ C (100 ⁰ F) until the curd pH reaches about 5.2 to about 6.2, or about 5.4 to about 6.0. When the curd pH is about 5.2 to about 6.2 (or in some embodiments, about 5.4 to about 5.8), cool the curd and whey to a temperature of about 83 ⁰ F or lower by addition of the cold whey that was previously removed at the end of the heating process. The addition of this cold whey (or cold UF permeate) helps

to cool the curd/whey suspension. Cooling the curd in liquid (e.g. whey) until the target pH of about 5.2 to 6.2 is reached, causes the curd to absorb moisture and increases final cheese moisture. Addition of cooled whey is preferred over addition of water because whey is more flavorful. Also, this is a good use for the whey and reduces waste.

This combination of curd pH and temperature of the whey is an important step in moisture control and pH control. When the pH of the cooled whey/curd suspension is appropriate (about pH 5.2 to pH 6.0, or about 5.6), the whey is drained off. The curds can then be salted. Three applications of salt can be made with mixing between applications. The curd is drained and salted at a pH and temperature low enough to achieve the desired final moisture and pH. The combination of low temperature and salt slows down the culture so that the pH does not go too low. Proper combinations of lower pH's and lower temperature' s can be used to achieve a higher cheese moisture without having the final pH of the cheese go too low. Note that less salt can be employed in the cheeses of the invention because the salt is retained better by the curds produced according to the methods of the invention. After salting, the curds can then be pressed into blocks or otherwise processed for aging, storage and/or sale. In many embodiments, the cheese can be packaged for immediate use or frozen for later use. If the cheese is frozen, the desired final pH of the cheese after freezing and thawing is in the range of about 5.1 to about 5.4, with a final percent moisture of greater than 53%. Depending on the combination of pH and temperature used prior to salting, cheese moistures of over 60% can be achieved. Such high moisture cheeses are useful because they remain moist for longer periods of time (thereby avoiding drying), they tend to be lower fat (more water, less fat) and they are less expensive to produce.

If a shredded cheese is desired, the block cheese can be cut or shredded into small particles. Most cheese-making procedures require an aging period before the cheese can be shredded. However, the present methods can eliminate this aging step. If desired, a hydrophobic surface coating can be sprayed onto the particles to help separate the particles and modify the melting characteristics of the shredded cheese. Shredded cheeses made by the process of the invention

can be used on top of a variety of frozen and non-frozen food products including pasta, chicken dishes, veal dishes, vegetables and the like. Because the cheeses of the invention are so moist, there is no need to add starch or another agent that improves moisture retention. Cheeses produced by the method of the invention can have a fat on dry basis (FDB) value of about 40% with a moisture content of 53 to 54%, salt content 1.5 to 1.7, pH of about 5.2. Such cheeses can be packaged and sold as cheese blocks, or as shredded cheese. The cheeses of the invention can also be used in a variety of food products.

Benefits of Using Carbon Dioxide

Use of carbon dioxide has several beneficial effects upon the cheese- making process and upon the ultimate cheese product. For example, carbon dioxide causes a shift in equilibrium of calcium in the milk moving some of the calcium that is bound to casein into the whey. This shift in calcium enhances the milk coagulation action of rennet and permits less rennet to be used (about 50% less). The removal of bound calcium from casein by carbon dioxide early in the cheese-making process is also important for achieving excellent meltability of the cheese without aging the blocks at refrigeration temperature for several days before shredding (and freezing, if desired).

The removal of bound calcium from casein early in the process due the action of CO ₂ causes changes the casein matrix structure of the curd particles. When this curd is salted and cooled, more casein moves from the matrix and dissolves in the water phase of the cheese than when CO ₂ is not used. The casein that has moved into the water phase binds water and allows the cheese to achieve a higher moisture content while still holding the water (without starch) during melting or baking. Thus, no starch or other material need be added to cheeses of the invention to improve the meltability and water retention of the cheese. Moreover, curd produced by the present methods binds water takes up salt more efficiently during the salting process. This reduces salt loss, salt waste and the cost of making cheese.

Summary of a Preferred Cheese-Making Procedure

1. Pasteurize the skim (or standardized milk) milk, cool the milk and inject CO ₂ into milk. The injection of carbon dioxide is generally performed at a milk temperature of 4O ⁰ F, but some variation in temperature is permitted

2. The milk is heated to 93 to 95 ⁰ F. The level of CO ₂ in the milk needs to be sufficient to produce a milk pH of about 5.90 to 6.0 at a temperature 90 to 100 ⁰ F.

3. Add starter culture (for example, mixed cultures of Streptococcus thermophilis and Lactococcus cremoris) and ripen 45 minutes.

4. At the end of ripening, add rennet. The amount of rennet used can be about 50% of that used when no CO ₂ is used in the cheese making. Let the milk coagulate (about 20 to 25 minutes) and then cut it.

5. After cutting, allow a 5 min curd heal and then stir gently and heat from about 95 ⁰ F to 101 ⁰ F in 15 min. If cheese vats are used, they maybe adapted to allow for curd that floats instead of sinks.

6. Remove some hot whey at the end of heating and cool this whey to 4O ⁰ F. Continue stirring the curd and the remaining whey at 95 ⁰ F to 101 ⁰ F until the curd pH reaches 5.8. 7. At a curd pH of 5.4- 5.8, cool the curd from 95 ⁰ F to 100 ⁰ F by addition of cold whey to reach a temperature of about 83 ⁰ F or lower. The combination of curd pH and temperature are used to control the final pH and moisture of the cheese.

8. At a curd pH of 5.6 or less, drain the whey. Add salt to the curd in 3 applications with about 5 minutes of mixing between applications. The approximate time from rennet addition to salting is about 110 to about 1600 minutes.

9. Immediately put the salted curd into hoops and press for 1 hour at 40 psi.

10. After 1 hour remove the block cheese from the hoop, place it into a plastic bag and vacuum seal it.

11. Place the block (packaged in plastic) into ice water to rapidly cool it. The block cheese is ready to be shredded and/or quick frozen the next day.

12. The cheese can be thawed and packaged directly with or without a hydrophobic surface spray to enhance meltability. When the procedures of the invention are used, the aging of cheese can effectively be eliminated. Moreover, cheese made by the procedures of the invention can be frozen and thawed before use without separation of fat from water, loss of moisture or adverse effects upon the txture, taste and melting properties of the cheese. If the cheese will be frozen immediately after it is made, the desired final pH of the cheese after freezing and thawing is in the range from 5.1 to 5.4, the final moisture >53%, and the final pH < 5.4. The combination of pH and temperature used prior to salting can influence the final pH and moisture content. Final cheese moistures of over 60% can be achieved using the methods of the invention.

Types of Cheese AU types of cheese can be made by the present methods. For example,

American, Cheddar, Monterey Jack, mozzarella, Muenster, Swiss, and the like can be made by the present methods. In some embodiment, all types of Cheddar cheese and mozzarella cheeses are made by the present methods.

Moreover, low-fat cheeses can be made by the methods of the invention. As illustrated herein, use of the pre- acidification methods of the invention leads to a cheese product that has a lower fat content than do similar methods that do not employ pre-acidification. Thus, for example, the fat content of the cheeses produced according to the invention can have about 2% to about 50% less fat, or about 3% to about 30% less fat, or about 4% to about 20% less fat, or about 5% to about 10% less fat than cheeses made without pre-acidification. Fat was reduced mostly during draining of whey.

Cheese Composition, Flavor and Texture

The composition of the cheeses produced by the methods of the invention is improved in several respects. First, the cheese is uniformly moist. Second, the moisture content of the present cheeses is somewhat increased, for example, from about 50% up to about 60%. In some embodiments, the cheeses of the invention have a moisture content of about 53% to about 54%. Such an increase

in moisture prevents the cheese drying out and improves the economics of cheese production. Third, the cheese does not bleed moisture. It is believed that the moisture in the cheese is retained by the higher content of protein in the soluble phase of the cheese. The flavor of cheeses produced by the methods of the invention is not adversely affected by the pre-acidification procedure and is generally improved by the increased moisture, improved salt retention and improved melting characteristics of the cheese. The intensity of cheese flavor results primarily from the action of enzymes, starter organisms, and nonstarter culture bacteria on intact casein and the degradation products of casein and amino acids. Aston and Creamer (1986) have reported that a sub-fraction of the water-soluble fraction containing most of the salt, free methionine, and free leucine contributed the most to the flavor of the water soluble fraction. By retaining and stabilizing the moisture naturally produced by the cheese, the flavor of the cheeses produced by the present methods is improved.

The texture of cheeses is typically due to the proteolytic breakdown of the casein matrix and possibly to changes in casein-water-calcium interactions as a function of aging. For example, Cheddar cheese texture starts out rubbery and corky, but rapidly changes to a softer more smooth texture as proteolysis continues during aging.

The methods of the invention improve the texture of cheese in several respects. Addition of an acidifying agent shifts the equilibrium of calcium from being bound to casein to being in solution within the whey. The removal of bound calcium from casein also shifts the insoluble-soluble casein equilibrium towards solubility. Thus the matrix structure of casein particles changes. This change in the casein matrix improves the melting characteristics of cheese without the need for extensive aging.

Thus texture of cheese made by the procedures of the invention is moister, smoother and dissipates even more quickly in the mouth than cheese made without use of acidifying agents.

The invention will now be illustrated by the following non-limiting Examples.

EXAMPLE l: Comparison of Procedures for Controlling Moisture Migration in Cheese

In this Example, three approaches to reducing moisture migration in 290 kg blocks were evaluated: (1) addition of hydrocolloids (e.g., starch, cellulose) to milk during cheese making, (2) addition of denatured whey proteins (e.g., Simplesse®) to milk during cheese making, and (3) addition of CO ₂ to milk to shift casein monomers and calcium from the micelles into the water phase of the milk, prior to rennet addition, to bind water and reduce moisture mobility in the cheese. Some of the materials in the first two approaches increased the moisture content of the cheese, but did not make a major reduction in moisture migration during cooling of the cheese. The addition of CO ₂ to milk was very effective at reducing moisture migration in cheese during cooling and the results of that work are reported below.

Materials and Methods Milk Carbonation

The milk carbonation system was a countercurrent stainless steel tubular heat exchanger (internal diameter = 0.5 cm) circulated with 0 to 1°C water. The carbonation system consisted of four units. They were sequentially from inlet to exit: (1) a milk feed reservoir; (2) a peristaltic feed pump (Amicon LP-I pump, Beverly, MA with a Cole-Palmer Masterflex® 7015-81 pump head, Vernon

Hills, IL) which fed the milk (4 ⁰ C) into the carbonation system at a flow rate of 900 ml/min; (3) a CO ₂ injection port, which was a stainless steel tube (internal diameter = 0.08 cm) inserted through a tee-fitting perpendicular to the milk flow; and (4) a holding section in which the milk was kept at 2 to 3 °C. Throughout the system, several temperature probes were inserted inline through tee-fittings to monitor milk temperature. CO ₂ (beverage grade) was injected inline into cold skim milk as it entered the carbonation system. The CO ₂ injection port was connected to a CO ₂ tank [CO ₂ line pressure = 55 psi (380 Kpa)] and the flow rate ofCO ₂ (1110 ml/min) was controlled with a flow meter. The CO ₂ flow rate was determined in a preliminary experiment to achieve a target concentration of CO ₂ in milk of approximately 3000 ppm. Carbonation of pasteurized skim milk was done at 2 to 3 ⁰ C.

Cheese Making Procedure

Reduced-fat Cheddar cheese was made by transferring either.215 kg of 4 ⁰ C pasteurized carbonated, or 215 kg of noncarbonated, skim milk to a cheese vat (model 4MX; Kusel Equipment Co., Watertown, WI). The casein to fat ratio was standardized by adding non-carbonated pasteurized heavy cream

(approximately 40% fat) to achieve a 50% fat reduction (compared to full-fat Cheddar cheese) in the final cheese. A stirred-curd cheesemaking procedure was used, as previously described in Olabi and Barbano, J. Dairy Sci. 85: (2002). The amount of chymosin and starter added per unit weight of milk was the same for carbonated and noncarbonated milk and the time from chymosin addition to cutting was 30 min. The coagulum was cut with 1.2 cm wire cheese knives and then allowed to heal for 5 min. Healing was followed by slow agitation for 10 min at 31 ⁰ C. The temperature was increased from 31 to 33°C and from 33 to 37°C in two 15 min intervals. The curd plus whey was stirred at 37 ⁰ C until a whey pH of 6.2 was achieved. Because the pH of the whey produced from carbonated milk was already lower than 6.2 (mean of three cheese-making trials was 6.01), the whey was drained, as soon as the second phase of cooking (at 37°C) ended.

At this point most of the whey was drained and the remaining curd- whey mixture was stirred at 37°C until a curd pH of 5.8 was attained. Next, additional whey was removed and cold water (about 5 ⁰ C) was added until a curd temperature of 28°C was achieved. Stirring was continued until a curd pH of 5.6 was achieved. Next, the mixture of whey plus water was drained from the curd, the curd was weighed, and salt was added (total of 2.7% wt/wt) in three equal portions with a 5-min interval between applications. The curd plus salt was mixed for about 1 min after each salt addition. The salted curd was put into an 18 kg stainless steel Wilson hoop and pressed, using a hydraulic A-Frame press (Model APVS, Kusel Equipment Co., Watertown, WI), at 10 psi (70 kPa) for 30 min followed by 60 psi (420 kPa) for 4.5 h at room temperature.

Sampling

Immediately after the 18 kg blocks of cheese (one bock for each treatment) were removed from the press they were cut and sampled in a walk-in

climate control room at 27 ± 0.5°C. The internal temperature of the blocks of cheese was approximately 25°C. Each 18 kg pressed block was cut in half at the middle and then three 17.78 x 7.62 x 2.54 cm (length, width, thickness) slabs of cheese were obtained from the center of each block. These slabs were immediately vacuum packaged (Multi Vac model 160; Koch, Kansas City, MO) in 3-mil polyethylene bags (25.4 x 45.7 cm, standard barrier; Koch, Kansas City, MO). In addition, two slabs of 28.6 x 14.0 x 2.54 cm (length, width, thickness) were cut from the 18 kg cheese block and vacuum packaged, as described above. The two slabs were cooled immediately and stored at 4°C. One slab was used for the cheese composition analysis and the other slab was used for the expressible serum analysis.

After the 17.78 x 7.62 x 2.54 cm cheese slabs were vacuum packaged, each bag was marked with horizontal lines to identify each 2.54 cm position. The different positions were numbered from the bottom (1) to the top (7). The slabs were attached to the suspension wires on a temperature gradient apparatus designed to cause moisture migration upwards from position 1 to position 7. The apparatus consisted of a water bath and a rotating cylinder designed to gradually raise vacuum packaged slabs of cheese out of the 27 ⁰ C water into the 3°C air over a period of 36 h. The 3 ⁰ C slabs of cheese were removed from the apparatus after 36 h and cut with a knife into 2.54 cm pieces by position. The cheese from each position was ground in a blender (model 31BL92; Waring, New Hartford, CT) and placed into two 50 ml plastic snap-lid vials (leaving no head space) and held at 4°C after blending. The pH of cheese from each position within each slab was measured at 23 ⁰ C as a single measurement within 2.25 h of grinding and the moisture content of the cheese within each position was measured in triplicate within 24 h after sampling.

Chemical Analyses Titratable acidity of the milk and whey was determined as described in

Marshall (1992) Chemical and physical methods. Pages 433-529 [method number 15.3.A] in Standard Methods for the Examination of Dairy Products

(16 ^th ed.) American Public Health Association, Washington, DC. The pH of the

milk, whey and cheese was determined using pH model HA405, Mettler Toledo electrode, Columbus, OH and Accumet pH meter, model 915; Fisher Scientific, Fair Lawn, NJ) during cheese manufacturing. The electrode was immersed in 3 M KCL storage solution at 38 ⁰ C between pH measurements to improve response speed and stability. Reference solutions (Fisher Scientific, Fair Lawn, NJ) for pH 4 (SB 101-500) and pH 7 (SB 107-500) were used at 38°C for calibration of the pH meter. The actual pH of the reference buffers was calculated for 38 ⁰ C based on the temperature coefficients recommended by the buffer manufacturer.

After cheese manufacture, the fat content of milk was determined by the Babcock method for milk (Association of Official Analytical Chemists, Methods of Analysis (17 ^th ed. 2000); method number 33.2.27; 989.04) and whey by skim milk Babcock test [(Marshall, 1992); method number 15.8. B] modified for use at 48 ⁰ C, instead of 58 ⁰ C for tempering and reading the fat columns (Lynch et al., JAOACI. 80. 845-859 (1997)). Total nitrogen (TN) for the milk and whey was determined by Kj eldahl ((Association of Official Analytical Chemists, 2000); method number 33.2.11; 991.20). Non-protein nitrogen in milk and whey was determined by Kj eldahl ((Association of Official Analytical Chemists, 2000); method number 33.2.12; 991.21). Noncasein nitrogen (NCN) was determined by Kj eldahl ((Association of Official Analytical Chemists, 2000); method number 33.2.64; 998.05). The casein content was calculated as TN minus NCN multiplied by 6.38. Calcium was determined by an atomic absorption spectroscopy procedure of Brooks et al. (Atomic absorption Newsletter 9(4): 93- 94 (1970)), as modified by Metzger et al. (J. Dairy Sci. 83:648-58 (2000)). The ppm CO ₂ content of the milk was determined using an infrared gas analysis method described by Ma et al. J. Dairy Sci. 84: 1959-68 (2001). AU analyses were performed in duplicate.

To prepare cheese for analysis of moisture, fat, protein, salt, calcium, and pH, the cheese from the 28.6 x 14.0 x 2.54 cm slab was ground in a blender (model 31BL92; Waring, New Hartford, CT) to a particle size of 2 to 3 mm. The ground cheese particles were packed into 50 ml plastic snap-lid vials (leaving no head space) and immediately placed in a 4°C refrigerator. Fat content was determined by the Babcock method (Marshall (1992) Chemical and physical methods. Pages 433-529 in Standard Methods for the Examination of Dairy

Products (16 ^m ed.). American Public Health Association, Washington, DC; method number 15.8.A). To determine fat content 9 g of cheese was used with 12 ml of distilled water at 100 ⁰ C. Total nitrogen was determined by Kjeldahl using a 1 g sample size. See, Association of Official Analytical Chemists (2000) Official Methods of Analysis. 17 ^th ed. AOAC, Gaithersburg, MD (method number 33.2.11; 991.20). Salt content was determined by the Volhard test ((Marshall, 1992); method number 15.5.B.); and moisture was determined gravimetrically, by drying 2 g of cheese at 100°C in a forced air oven (model OV-490A-2; Blue M, Blue Island, IL) for 24 h)(Association of Official Analytical Chemists, 2000); method number 33.2.44; 990.20). Calcium was determined by an atomic absorption spectroscopy procedure of Brooks et al. (Atomic Absorption Newsletter 9(4): 93-94 (1970) and as modified by Metzger et al. (J. Dairy Sci. 83: 648-58 (2000). Cheese pH was measured at 23 ⁰ C. The analyses were done in duplicate for pH, salt, and calcium, in triplicate for fat and total nitrogen, and in quadruplicate for moisture. The amount (g/100 g of cheese) of expressible serum was determined at d 2 of refrigerated storage, as described by Guo and Kindstedt (J. Dairy Sci. 78:2099-2107 (1995)) with one modification. The cheese serum was removed by centrifugation at 25,000 x g (maximum force) instead of 12,500 (maximum force).

Experimental Design and Statistical Analysis

Two vats of reduced-fat Cheddar cheese, one from non-carbonated and one from a carbonated portion of the same milk, were made side by side in each of 3 wk. One 18 kg block of reduced fat Cheddar cheese was produced for each treatment in each of the three weeks. On each week, three cheese slabs for each treatment were removed from each 18 kg block, as described by Olabi and Barbano, J. Dairy Sci. 85: (2002). The cheese slabs were put on the apparatus using a temperature gradient designed to move moisture upward (id.), hi the ANOVA model, treatment was a category variable and position was a continuous variable, while cheese making week (i.e., batch of milk) was blocked as a fixed effect. The position variable was transformed to make the data set orthogonal. Therefore, positions 1, 2, 3, 4, 5, 6 and 7 were coded as -3, -2, -1, 0, +1, +2, and +3, respectively, as the input data for the position variable in the

ANOVA. The interaction term between treatment and cheese making week and position was used as the error term for the main effects. The PROC GLM procedure of SAS was used for all data analyses (SAS version 8.02, 1999-2001).

RESULTS

Milk Composition and CO ₂ Content

No significant difference (P > 0.05) in fat, protein, casein, and NPN content of the standardized milk with and without added CO ₂ was detected (Table 1). Table 1. Mean (n=3) milk, whey, and cheese composition for control and

CO ₂ treatments. i

Treatment

Sample type CONTROL CO ₂ LSD

Milk

Fat, % 1.23 1.23 NS

Protein, % 3.06 3.01 NS

Casein, % 2.23 2.20 NS

NPN ¹ , % 0.20 0.21 NS

CO ₂ at 4°C ² , ppm 139 ^B 3175 ^a 800

CO ₂ at 31 ⁰ C ³ , ppm 110 ^B 1721 ^a 523 pH at 31 ⁰ C 6.63 ^A 6.01 ^b 0.05

TA ⁴ at 31°C 0.16 ^B 0.42 ^a 0.09

Whey

Fat, % 0.05 0.06 NS

Protein, % 0.80 0.79 NS

NPN ¹ , % 0.27 0.26 NS

Cheese

Moisture, % 47.72 ^b 53.48 ^a 3.01

Fat, % 15.17 ^a 13.61 ^b 0.64

FDB, % 29.00 29.26 NS

Protein, % 30.59 ^a 26.64 ^b 1.69

PDB, % 58.51 ^a 57.27 ^b 0.78

M/P ⁵ 1.56 ^b 2.01 ^a 0.22

Salt, % 2.04 ^b 2.40 ^a 0.28

SM ⁶ , % 4.27 4.50 NS pH 5.08 5.09 NS

^a ' means within row not having a common superscript differ (p < 0.05).

¹ NPN expressed as nonprotein nitrogen times 6.38.

² CO ₂ at 4°C = CO ₂ content of pasteurized carbonated skim milk.

³ CO ₂ at 31 ⁰ C = CO ₂ content of vat milk collected just before the addition

of starter culture. ⁴ TA = titratable acidity. ⁵ M/P = moisture to protein ratio. ⁶ S/M = salt concentration in moisture.

The titratable acidity of the milk with added CO ₂ WaS very high (Table 1) and reflects the degree of interaction of CO ₂ with water to form carbonic acid in the milk. The system for addition carbon dioxide to skim milk used in this study increased (P <0.05) the level of CO ₂ in the 4 ⁰ C skim milk prior to cheese making and achieved our target of approximately 3000 ppm of CO ₂ (Table 1). After addition of cream, that did not contain CO ₂ , to the skim milk, the mixture was stirred and heated from 4 ⁰ C in an open cheese vat to 31 ⁰ C prior to cheese making. During this process some CO ₂ was lost from the milk. The CO ₂ content of the standardized milk at the point of starter culture addition at 31 ⁰ C had decreased from about 3000 ppm to about 1721 ppm (Table 1).

Addition of CO ₂ to milk caused a decrease in milk pH from 6.63 to 6.01 at 38 ⁰ C (Table 1), as was expected from the results of Ma et al., J. Dairy Sci. 84:1959-1968 (2001). Ma et al. (2001) also reported a corresponding decrease in milk freezing point as the concentration of CO ₂ in milk increased and pH decreased. The decrease in freezing point reported by Ma et al. (2001) is caused by the combined effects of the dissociation products of carbonic acid and the shift in equilibrium from bound calcium phosphate in the casein micelles to soluble calcium phosphate in the serum phase of milk. De La Fuente et al. (J. Food Prot. 61:66-72 (1998)) reported that addition of HCl, lactic acid, and CO ₂ to milk to produce a reduction of milk pH to 6.1, all produced an increase in the proportion of soluble calcium in cow's milk and a similar effect was reported for sheep and goat milks. The increase in soluble calcium content, caused by CO ₂ addition to milk, would be expected to influence milk coagulation properties. Lowering the pH of milk also has an effect on micelle structure. As the pH of milk is lowered, a considerable portion of the micellar casein solubilizes. Dalgleish and Law, J. Dairy Res. 55:529-538 (1988); Singh et al., J. Dairy Sci. 79:1340-1346 (1996); van Hooydonk et al., Neth. Milk Dairy J. 40:281-296. (1986). In the case of the milk containing CO ₂ prior to cheesemaking, this means that a higher level of soluble casein should have been present in the serum phase of milk prior to rennet addition. It was not clear if these nonmicellar

caseins would be trapped in the rennet coagulation or if they will be lost in the whey.

Cheesemaldng The curd was visibly firmer at 30 min and started to coagulate sooner after rennet addition for the milk with added CO ₂ than for milk without CO ₂ . Okigbo et al. ((1985) J. Dairy Sci. 68:3135-3142), found that reducing milk pH prior to rennet addition increased the speed of milk coagulation with rennet. Other workers have reported decreased coagulation time. (Calvo et al. (1993) J. Food Prot. 56: 1073-1076; De Ia Fuente, et al. (1998) J. Food Prot. 61:66-72). Further work indicates that addition of CO ₂ may reduce the amount of rennet used in cheese making (McCarney, et al. (1995) Milchwissenschaft 50:670-674; Montilla et al. (1995) Z. Lebensm Unters Forsch. 200.289-292; St-Gelais et al. (1997) Milchwissenschaft 52: (11) 614-618). The decision of when to drain the whey in our cheese making was based on the pH of the whey at the end of the cooking step. The milk with added CO ₂ had a lower whey pH than the target whey pH than generally used for draining in the inventor's cheesemaking procedure and therefore the whey was drained immediately after the final cook temperature of 37 ⁰ C was achieved. As a result, the time from starter addition to whey draining was shorter (P < 0.05) for milk with added CO ₂ , but the time from whey draining to the time when a curd pH of 5.6 was achieved (i.e., salt addition) was longer (P < 0.05) for the milk with added CO ₂ . This produced a total make time from starter additional to salt addition that was not different (P > 0.05) for milk with or without added CO ₂ (Table 2). St-Gelais et al. (1997) reported a 30 min reduction in total cheese making time due to the preacidification of milk to pH 6.55 with either CO ₂ or lactic acid. This effect was not observed in the experiments reported herein.

Table 2. Mean make times (n =3), (min.) for control and CO ₂ treatments.

Cheesemaking step Treatment

CONTROL CO ₂ LSD

Starter addition to whey draining 164 ^a 128 ^b 2.9

Whey draining to salt addition 83 ^b 129 ^a 10.8

Starter addition to salt addition 247 256 NS ^a> means within column not having a common superscript differ (p < 0.05).

The profile of pH change with time during cheese making for milk with added CO ₂ was quite different than for milk without added CO ₂ , as shown in FIG. 2. Milk pH starts out low (about 6.0) and decreased very slowly for the first 200 min of the cheese making process for the milk containing approximately 1700 ppm of CO ₂ (FIG. 2). During this time, lactic acid was being produced by the starter culture and CO ₂ was being lost from the curd plus whey, with the net effect being a small and slow decrease in pH from about 6.0 to 5.8 over a period of about 200 min. This difference in pH profile during the cheese making process had a major impact on the calcium content of the whey and the cheese.

Whey and Cheese Composition

There were no significant differences in the fat, protein, or NPN contents of the whey from cheese making between control and milk with CO ₂ (Table 1). If the addition of CO ₂ did cause casein monomers to move out of the casein micelles along with some calcium phosphate prior to rennet addition, the presence of these caseins outside the micelles did not result in an increase in protein content of the whey.

The control cheese had similar composition to reduced-fat Cheddar cheese produced in other research studies (Johnson et al. (2001) J. Dairy Sci. 84: 1027-1033; Fenelon et al. (1999) J. Dairy Sci. 82:10-22; Chen et al. (1998) J. Dairy Sci. 81:2791-2797; Metzger and Mistry (1995) J. Dairy Sci. 78:1883- 1895). The moisture content of the cheese made with CO ₂ added to the milk was almost 5% higher (P < 0.05) than the reduced fat Cheddar made from the same milk without CO ₂ . McCarney et al. ((1995) Milchwissenschaft 50:670-674), also reported that moisture content of their Cheddar cheese made from milk containing CO ₂ was higher despite modifications to the manufacturing process to reduce moisture levels. However, they did not provide any moisture data in their report, so it is not possible to determine the magnitude of the effect they observed. Johnson et al. ((2001) J. Dairy Sci. 84: 1027-1033) reported that a firmer coagulum at cutting favors higher moisture content. In our study the coagulum at cutting for milk with added CO ₂ was much firmer than without added CO ₂ and the final moisture content of the cheese was higher. As

expected, the higher moisture content of the cheese in this study caused the concentration of both fat and protein (on a wet basis) to be lower (P < 0.05) due to dilution. The fat on dry basis (FDB) in the cheese was not influenced by the use of CO ₂ , but the protein on a dry basis (PDB) was slight lower (P < 0.05) for cheese made from milk with added CO ₂ . Further work with measurement of cheese yield would be needed to determine if the lower PDB is an indication of slightly higher protein loss in salt whey during pressing. The higher moisture content of the cheese produced with CO ₂ added to the milk results in a much higher moisture to protein ratio and would tend to produce a softer cheese (Table 1). The absolute concentration of salt in the cheese made from milk with added CO ₂ was higher (P < 0.05) than cheese made from milk without added CO ₂ , but no difference in the salt concentration in the water phase of the cheese was detected (P > 0.05). The final pH values of the cheeses were almost identical (Table 1). As shown in Table 3 , there was a maj or influence of the addition of CO ₂ to the milk on the calcium content of the whey and cheese.

Table 3. Mean (n=3) milk, whey, and cheese calcium levels (g/100g) for control and CO ₂ treatments.

Sample type Treatment

CONTROL CO ₂ LSD

Milk 0.110 0.109 NS

Whey 0.046 ^b 0.066 ^a 0.005

Cheese 0.789 ^a 0.508 ^b 0.102

¹ CaZP 2.58 ^a 1.90 ^b 0.251

^{a> b} means within row not having a common superscript differ (p < 0.05).

¹ CaZP = calcium as a percentage of protein in cheese.

The calcium content (Table 3) of the whey was higher (P < 0.05) and the calcium content of the cheese was lower (P < 0.05) from milk with added CO ₂ . The increase in calcium content of the whey that was observed is consistent with the shift in bound to soluble calcium in milk during acidification as reported by Le Graet and Brule ((1993) Lait 73:51-60). hi another study, Metzger et al. ((2000) J. Dairy Sci. 83: 648-658) acidified milk to pH 6.0 prior to low fat Mozzarella cheese making with acetic and citric acid and reported higher concentrations of calcium in whey and lower concentrations of calcium in cheese. Metzger et al.

((2001) J. Dairy Sci. 84: 1348-1356) reported that reducing the pH of milk prior to cheese making reduced the hardness, initial apparent viscosity, and expressible serum of low fat Mozzarella cheese. While no texture measurements were done on the reduced fat Cheddar cheese made in the current study, the higher moisture content and lower calcium content for the cheese made from milk with added CO ₂ produced a cheese that was visibly softer than the cheese made without CO ₂ added to the milk.

Expressible serum content of cheese can be used to reflect the status of water mobility within the structure of the cheese. The mobility of water within the cheese structure could be very important in moisture migration during cooling of 290 kg blocks of Cheddar cheese. The expressible serum content of the reduced fat Cheddar cheese produced in the current study was measured on d 2 after cheese making and was found to be much lower (P < 0.05) in the cheese that was produced from milk with added CO ₂ (Table ).

Table 4. Mean (n=3) cheese expressible serum at day 2 (g/100g cheese) for control and CO ₂ treatments.

Sample type Treatment

Cheese CONTROL CO ₂ LSD 7.487 ^a 0.931 ^b 3.66

^a> means within row not having a common superscript differ (p < 0.05).

This low expressible serum content may have an impact on water mobility within large blocks of Cheddar cheese during cooling. The results provided herein are consistent with the results of Metzger et al. ((2001) J. Dairy Sci. 84: 1348-1356), who reported that reduction in milk pH prior to cheese making produced Mozzarella cheese with a lower amount of expressible serum. It is normal for expressible serum content of cheese to decrease with time during refrigerated storage (Guo and Kindstedt (1995) J. Dairy Sci. 78:2099-2107) and after about 2 wk of storage of Mozzarella cheese, the amount of expressible serum is near zero. Guo et al. ((1997) J. Dairy Sci. 80:3092-3098) also reported that the amount of expressible serum in the structure of cheese immediately after manufacture was reduced by the addition of salt. These changes in the mobility of water can influence, or are at least related to, functional properties such as meltability of cheese. McMahon et al. ((1999) J. Dairy Sci. 82:1361-1369)

described water partitioning in Mozzarella cheese and its relationship to cheese meltability. They found that the amount of chemically bound water in cheese remained constant during refrigerated storage of cheese but that there was an increase in entrapped water and a decrease in expressible water with increasing age of the cheese and that these changes were related to time dependent changes in cheese functionality. Assuming that a similar behavior happens in Cheddar cheese, the amount of water bound to protein probably remained constant in our cheese but the CO ₂ treatment caused a large decrease in the expressible water (i.e., easily movable) and an increase in the entrapped water within the structure of the reduced fat Cheddar cheese. This increase in entrapped water may have an influence on moisture migration within the cheese during cooling.

Temperature Induced Changes in Cheese pH and Moisture During Cooling Cheese pH. Slabs of cheese were placed in an apparatus designed to move moisture upward from position 1 to position 7 within a slab of cheese. Position 7 was the cold end position of the cheese and as the cheese is gradually raised out of the warm (27 ⁰ C) water over a period of 36 h, moisture migrates against the force of gravity from the bottom to the top of the cheese slab. The temperature gradient in the slab of cheese may also produce a gradient of pH within the slab of cheese. There was a trend (P = 0.065) for cheese pH (Table 5) to increase from the bottom to the top of the slab of cheese but the total range of difference in pH (i.e., 4.88 to 5.03) was small. No direct influence OfCO ₂ on cheese pH was detected (Table 5), but there was a position by CO ₂ treatment (P < 0.01) interaction (Table 5).

Table 5. ANOVA (degrees of freedom, means squares, probabilities (in parenthesis) and R ² ) for CO ₂ treatment, week of cheese making, and position within cheese slab on percent moisture and pH of cheese exposed to a temperature gradient during cooling.

Factor df Moisture pH Variable type

TR ¹ ϊ 328.94 0.0004 Category

(pO.Ol) (0.749)

W ² 2 23.30 0.0068 Fixed

(pO.Ol) (pO.Ol)

P ³ 1 98.36 0.0373 Continuous (p<0.01) (0.065)

TR x W 2 7.00 0.0028 (pO.Ol) (po.oi)

P x W 2 0.03 0.0002 (0.547) (0.495)

P x TR 1 51.12 0.0131 (pO.Ol) (pO.Ol)

P x TR x W 2 0.57 0.0027

(po.oi) (pO.Ol)

Error (for P, 30 0.045 0.0003 PxTR, PxW, TRxW, and PxTRxW)

R ¹ 0.998 0.900

¹ TR = Treatment. ² W = Week. ³ P = Position.

Generally, there was almost no change in pH with position in the slab for the control and there was a slight trend for pH to increase from the bottom to the top of the slab for the cheese made from milk with CO ₂ . The area of the slab of cheese that was cooled the fastest (i.e., position 7) had the highest pH and the area of the slab that was cooled the slowest (i.e., position 1) had the lowest pH. Cheese moisture. There was a large impact (Table 5) of CO ₂ treatment (P < 0.01) and a significant position by CO ₂ treatment interaction on moisture migration in reduced fat Cheddar cheese (FIG. 4). The range of moisture migration across the seven positions within the slab of cheese for the control was about 7.47%, while the moisture migration in the cheese made from milk containing CO ₂ was approximately 1.3%. Clearly, the use of CO ₂ reduced (ca. 80% reduction) the mobility of water during the cooling of the cheese and prevented the majority of moisture migration. This result is even more impressive when one considers the fact that the moisture content of the reduced fat Cheddar made from the milk with added CO ₂ contained 5.76% more moisture

(Table 1) and had a significantly higher moisture to protein ratio than the control

(Table 1). The reduction in the temperature induced moisture migration is consistent with the large reduction in expressible serum in the cheese that was produced by the addition of CO ₂ to the milk prior to cheese making. Thus, it appears that addition of a sufficient quantity of CO ₂ to milk to reduce the pH to about 6.0 prior to cheese making, would significantly reduce moisture migration and the associated quality defects in 290 kg blocks of Cheddar cheese.

In some instances, the large increase in moisture due to CO ₂ addition to milk may not be desirable from a product characteristic point of view. During this study, no attempt was made to modify the cheese making procedure to produce cheese with lower moisture content from CO ₂ containing milk, but traditional modifications of the cheese making procedure should be able to bring the moisture content of cheese made from milk containing CO ₂ closer to that of cheese made from milk without CO ₂ . However, if it was possible to carry a higher moisture level in the cheese and achieve acceptable cheese quality, then this would provide some yield benefit. This may be more feasible for higher moisture Cheddar varieties that are not aged for long periods of time.

Thus, addition of hydrocolloids and denatured whey proteins to milk prior to cheese making increased cheese moisture content but did not reduce temperature induced moisture migration in Cheddar cheese during cooling. Addition of CO ₂ to milk prior to cheese making made a major (ca. 80%) reduction in temperature induced moisture migration in reduced-fat Cheddar cheese during cooling. The use OfCO ₂ also caused an increase (ca. 5%) hi cheese moisture in reduced-fat Cheddar cheese. Despite the higher moisture content, the temperature induced movement of moisture was still dramatically smaller for the reduced-fat Cheddar cheese made with CO ₂ added to the milk. The CO ₂ produced a reduction in milk pH to about 5.9 to 6.0 and caused a movement of calcium out of casein micelles prior to rennet addition. The use of CO ₂ in the milk did not interfere with normal acid production by the starter culture and the final cheese pH was not influence by the addition of CO ₂ .

EXAMPLE 2: CO ₂ Cheese Making Procedures

Materials and Methods Experimental Design and Statistical Analysis

Two 18 kg blocks of full-fat Cheddar cheese were. One block was made from milk with added CO ₂ and another control block was made from milk without added CO ₂ . Cheese was made on three different days (using a different batch of milk each day) over a one-week period. A one-way ANOVA was used to determine if there was an impact of CO ₂ on the composition and yield of the cheese. The least significant difference (P < 0.05) test was used to compare treatment means if the F-test for the statistical model was significant (P < 0.05). The PROC GLM procedure of SAS was used for all data analysis (SAS version 8.02, 1999 - 2001, SAS Institute Inc., Gary, NC) Milk Processing and Cheese Manufacture

Milk processing and cheese manufacture were completed on the same day. Each day, raw whole bovine milk was received from the Cornell University dairy farm. The raw whole milk was pasteurized with a plate heat exchanger at 72°C and a holding time of 15 sec. The pasteurized milk was then cooled to 4°C utilizing the regeneration and cooling sections of the system. Carbonated milk was collected after about 250 kg of pasteurized milk was collected for the control treatment. A stainless steel sparger (7 pan) was inserted in-line after the cooling section of the pasteurization system. An additional 8 m of 2.54-cm diameter stainless steel pipe, with a sanitary conical-seat flow controlling valve at the end, were added after the point of CO ₂ injection to allow 15 s of holding time for CO ₂ incorporation into the cold pasteurized milk. A pressure gauge was located in-line at the point of injection and another just before the flow control valve. Carbonation conditions (CO ₂ flow at 1.13 m ³ /h with 172 kPaback pressure at the flow control valve) were equilibrated with water (16.6 L/min) before milk was started through the pasteurization system. Milk was carbonated to approximately 1650 ppm to achieve a CO ₂ level of about 1600 ppm in the cheese vat, which produced a milk pH of 5.9 at 31°C. The pH of 5.9 is between the two pH levels reported by Metzger et al. (2001) where increased soluble nitrogen levels were observed.

All weights for cheese making including milk, whey, salt whey, cheese, and samples were weighed to the nearest gram (Model PE24, Mettler Instrument Co., Highstone, NJ). After pasteurization, approximately 240 kg of milk was weighed and placed into the control vat. The second vat was filled with about 240 kg of pasteurized and carbonated milk. Cheese making for these two vats was conducted simultaneously. The milk in each vat was heated to 31°C, with agitation (Model 4MX; Kusel Equipment Co., Watertown, WI). The milk was ripened for 45 min at 31 ⁰ C after the starter culture (911 DVS pellets, Chr. Hansen Inc., Milwaukee, WI) was added at a concentration of 0.27 g/kg of milk. When ripening was complete, annatto color (AFC WOS 550, Rhodia Inc.,

Madison WI) was added (0.0033 mL/kg of milk) to each vat. The ripened milk, 31 ⁰ C, was coagulated with double strength chymosin (0.1 mL/kg of milk; Chymax Extra, Chr. Hansen Inc., Milwaukee, WI). The chymosin was diluted, in 200 mL of water processed by reverse osmosis, immediately before addition to the milk. After 30 min, the coagulum was cut (1.2 cm wire knives) and the curds and whey were not stirred for 5 min. After 5 min, the curds and whey were stirred gently without added heat for 10 min. The temperature was increased from 31 to 33°C over 15 min and then from 33 to 38°C over an additional 15 min. The curds and whey were continuously stirred and a temperature of 38 ⁰ C maintained until the target whey draining pH of 6.35 was attained. When the whey was drained, the curds were piled and allowed to knit together for 15 min. The large slab of curd was cut into two smaller slabs then turned. The two curd slabs were stacked after 15 min. Curd slabs were maintained at 38°C, piled two high, and turned over every 15 min throughout the Cheddaring process. Curd slabs were milled when the curd pH reached 5.30. Salt was added at 2.7% of the curd weight. The salt was divided equally into three portions. The milled curds were dusted with a small amount of the first portion of salt, then stirred for 2 min and allowed to sit for 10 min. The remainder of the first portion of salt was then added, the curds stirred, and then the curds were allowed to sit for 10 min. The curds were salted with the two other portions of salt in 10 min intervals. The salted, milled curds were placed in a 18 kg capacity stainless steel Wilson hoop and pressed in an A-frame press (Model AFVS, Kusel Equipment Co., Watertown, WI) for 30 min at 70 kPa. Pressing was continued overnight, about

17 h, at 420 IcPa. The cheese blocks were vacuum packaged and placed in a 4 ⁰ C cooler for 24 h before being placed in a cooler set at 6°C for aging.

Sampling and Sample Preparation Sampling. Raw whole milk at 4°C was mixed and sampled immediately before pasteurization. Pasteurized control and CO ₂ treated milks were collected after heating in the cheese vat to 31°C prior to starter addition. The whey collected from the start of curd draining to the end of draining was placed in a separate vat for each treatment and sampled for CO ₂ analysis. Additional whey collected throughout Cheddaring was added to the vats containing the whey. When all the whey from each vat was collected, the whey was heated to 38°C and mixed to assure uniform composition before a sample was taken for compositional analysis and used in mass-balance calculations. Salt whey was collected and weighed separately after milling at the vat and mixed with the press whey, which was weighed. Press whey was collected during pressing by placing the hooped curds in large 8-mil plastic bags (model number S-5851, Uline, Waukegan, IL). Hot water was run on the outside of the bags to liquefy fat that may have solidified on the inside surface of the bag during pressing and all of the whey was removed from the bags. A 1-cm thick by 28 cm by 19 cm cross sectional slice from the center of the rectangular 18-kg block of cheese was removed immediately after the block was removed from the press. This slice of cheese was used for compositional analysis and was vacuum- packaged and cooled to 4 C prior to analysis.

Sample preparation. Liquid samples were placed in 59-mL snap lid vials and either analyzed fresh or stored frozen at -40°C. Frozen liquid samples were thawed in a microwave oven in a manner that kept the sample temperature below 10 C. Cheese slices were cut into 2-cm pieces then ground (Model 31BL92, Waring, New Hartford, CT) in 2 to 3 mm pieces and packed into 59-mL snap lid vials (Capital Vial, Inc., Fultonville, NY) with no head space and either analyzed fresh or held frozen at -40°C before analysis. Frozen cheese samples were thawed overnight at 4 ⁰ C prior to analysis.

Standard Plate, Coliform, and Somatic Cell Counts

Standard plate and total coliform counts of pasteurized whole milks were determined by standard methods (Marshall, 1992; 6.2 and 7.8). Somatic Cell Counts (SCC) of raw whole milk (AOAC 2000; 17.13.01, 978.26) were determined using a fluorimetric method (Milk-Scan Combi 4000, Integrated Milk Testing; A/S N. Foss Electric Hillerpd, Denmark) by a New York State licensed commercial laboratory (Dairy One, Ithaca, NY). Chemical Analyses

Milk, whey, and salt whey composition. Fat, total salt (TS), total nitrogen (TN), nonprotein nitrogen (NPN), noncasein nitrogen (NCN) content of the milk, whey, and salt whey were determined using ether extraction (AOAC, 2000; 33.2.26, 989.05, forced air oven drying (AOAC, 2000; 33.2.44, 990.20), Kjeldahl (AOAC, 2000; 33.2.11, 991.20), Kjeldahl (AOAC, 2000; 33.2.12, 991.21), Kjeldahl (AOAC, 2000; 33.2.64, 998.05), respectively. Crude protein (CP) was calculated by multiplying total nitrogen by 6.38. The calcium content was determined using atomic absorption (Metzger et al, 2000). CO ₂ content of the milk and whey was determined (Ma et al., 2001) using a CO ₂ analyzer (MOCON Pac Check 650, MOCON, Minneapolis, MN). The Volhard method (Marshall, 1992; 15.5.B) was used to determine the salt content in the salt whey, using a 0.5-g test portion. Milk, whey and salt whey compositions were determined in triplicate with the exception of calcium and CO ₂ , which were determined in duplicate.

Cheese composition andpH. Fat content was determined using the Babcock method (Marshall (1992) at 15.8.A). Cheese moisture was determined gravimetrically by drying 2 g of cheese in a forced-air oven at IOOEC for 24 h (AOAC, 2000; 33.2.44, 990.20) using 2 g of cheese. Salt content was determined using the Volhard method (Marshall, 1992; 15.5.B). The Kjeldahl method employingl g of cheese, was used to determine total nitrogen (Lynch et al., 2002). The cheese calcium content was determined by atomic absorption (Metzger et al., 2000).

Cheese pH was measured using a Xerolyt combination electrode (Model HA405; Mettler Toledo, Columbus, OH) and an Accumet pH meter (Model AR 25, Fisher Scientific, Pittsburgh, PA) after tempering to 23°C. All analyses were

carried out in duplicate except total nitrogen, moisture, and fat which were performed in quadruplicate.

Component Recoveries Fat, crude protein, calcium, total milk solids, and added salt recoveries were determined by multiplying the weights (determined to the nearest g) of milk, whey, salt whey, and cheese by the compositions determined by chemical analysis then dividing by the total weight of either fat, crude protein, calcium, total milk solids, or added salt and multiplying by 100. Total milk solids recovery calculations did not include salt in the salt whey or in the cheese. If the mean actual total unadjusted recoveries between treatments for the component were not significantly different (P > 0.05), then the recoveries were adjusted by dividing the actual recoveries by the mean total recovery for each day of cheese making and multiplying by 100.

Yield and Yield Efficiency

Actual cheese yields were calculated by ((cheese weight + curd sample weight) / (milk weight - milk sample weight) x 100. Moisture and salt adjusted yield was calculated accordingly (actual yield x (100 - (cheese moisture content + cheese salt content))) / (100 - (37 + 1.5)). The moisture and salt adjustment allows for comparison between treatments. Cheese yield efficiency was calculated by dividing the adjusted yield by the theoretical yield and multiplying by 100. Both Van Slyke and Barbano theoretical cheese yield formulas (Neocleous et. al., 2002) were used for a cheese yield efficiency calculation. The Van Slyke cheese yield formula for Cheddar cheese was calculated according to the following formula: yield = (((0.93 ^χ percent fat in milk) + (percent casein in the milk - 0.1)) x 1.09) / (1 - (target cheese moisture / 100)). The Barbano formula for Cheddar cheese differs from the Van Slyke in that the nonfat solids of the whey were used to determine the nonfat whey solids retained in the water phase of the cheese (Barbano, 1996). The Barbano formula is useful when manufacturing a preacidified cheese because it can compensate for the loss of calcium into the whey. Theoretical Cheddar cheese yield using the Barbano formula was calculated using the, following formula: yield = (A + B + C) / (1 -

((target cheese moisture + target cheese salt) / 10O)) where A = (0.93 x percent fat in milk), B = (percent casein in milk - 0.1) x (calcium phosphate retention factor), C = ((((A + B) / (1 - (actual cheese moisture percent / 100))) - (A + B)) x (percent nonfat whey solids / 100)) x (solute exclusion factor). The calcium phosphate retention factors used in this study for the control and CO ₂ treatments were 1.092 and 1.082, respectively. The same calcium phosphate retention factor for the control theoretical yield calculation was used by Neocleous et al. (2002). The lower calcium phosphate retention factor used for CO ₂ treatment theoretical yield calculation was obtained by plotting calcium retention factor data (acetic acid treatments) of Metzger et al. (2000) and using the second order polynomial equation (solute exclusion factor = - 0.0333(x ² ) + 0.4333(x) - 0.316) to compute a calcium retention factor to use for the milks (mean milk pH of 5.93) with added CO ₂ , where x = milk pH. The solute exclusion factor of 0.6941 used by Neocleous et al. (2002) was also used in this study for both the control and added CO ₂ theoretical yield formulas.

Results Milk Composition and Quality

Milk composition data are presented in Table 6. The mean casein as a percentage of true protein (TP) was about 83%, which in turn gave rise to a mean casein to fat ratio of 0.66. The pH of the milk at 31 ⁰ C before CO ₂ addition (Table 6) was normal for good quality milk. The standard plate counts of the control and preacidified milks of 440 and 1150 cfu/mL were low and not significantly different. The coliform count was < 10 cfu/mL for both treatments. The mean SCC of the milks used each day for cheese making are shown in Table 6.

Table 6. Pasteurized control milk composition and pH for each day of cheese making.

Day of cheese making

Component 1 2 3

Fat, % 3.44 3.42 3.35

Grade Protein, % 2.90 2.92 2.93

Non-Casein Protein ¹ , % 0.66 0.67 0.66

Nonprotein Nitrogen ² , % 0.20 0.18 0.19

TS, % 11.88 11.75 11.74

Casein/Total Protein ³ , % 82.96 82.26 82.99

Casein/fat, % 0.65 0.66 0.68

Calcium, % 0.105 0.106 0.106

PH ⁴ 6.63 6.68 6.64

SCC, per rnL 268,000 260,000 226,000

¹ NCN = noncasein nitrogen x 6.38. ² NPN = NPN x 6.38. ³ CN/TP = (((CP-NCN) / ((TN - NPN) x 6.38)) x 100).

⁴ Measured at 31°C.

The CO ₂ levels of the control and CO ₂ treated milks were significantly different after CO ₂ addition (Table 7). The target milk pH of about 5.9 was achieved (Table 8) by the addition of about 1600 ppm of CO ₂ (Table 7).

Table 7. Mean (n = 3) carbon dioxide content (ppm) of the control and CO ₂ treated milk and whey.

Control CO ₂ treated LSD SEM ppm

Milk 88 ^b 1615 ^a 77 19.7

Whey at draining 85 ^b 1000 ^a 115 29.4 ^a> Means within a row that do not share a common superscript differ (P < 0.05).

LSD = least significant difference; SEM = standard error mean.

Table 8. Mean (n = 3) milk, whey, and curd pH during cheese making.

Control CO ₂ LSD SEM

Milk 6.65 ^a 5.93 ^b 0.054 0.014

Coagulant addition ¹ 6.57 ^a 5.93 ^b 0.052 0.013

Drain ² 6.35 ^a 5.96 ^b 0.074 0.019

Mill ³ 5.30 5.30 NS 0.000 ^a> Means within a row that do not share a common superscript differ (P < sample;

³ Curd sample;

LSD = least significant difference; SEM = standard error mean. Effect of CO ₂ on Cheese Making

The mean (n = 3) temporal patterns of pH for the control and CO ₂ treatment during cheese making are shown in FIG. 4. The milk pH before starter culture addition, at coagulant addition, and of the whey at draining was higher (P < 0.05) for the control than the CO ₂ treatment (Table 8). The slope of the control pH curve from 0 to 120 min (FIG. 4) was negative whereas the CO ₂ treatment pH was constant over the same period. The downward slope of the control pH curve was expected because the starter culture was producing lactic acid during the 45 min of ripening. Although the starter culture growing in the milk of the CO ₂ treatment was producing lactic acid, the pH of the whey did not change much from the initial pH of the milk until after about 130 min into cheese making because the milk was also losing CO ₂ . The time from coagulant addition to whey draining (Table 9) was shorter for the CO ₂ treatment (P < 0.05). Moreover, if CO ₂ produced the milk pH decrease usually caused by lactic acid during ripening, then the lactic acid content of the sweet whey collected at draining would be reduced. The CO ₂ remaining in the whey could be removed with a vacuum chamber. This might improve the quality of whey products in certain applications. The separate impacts of lactic acid and CO ₂ on the pH observed for the CO ₂ treatment (FIG. 4) cannot be determined from the data collected. The total make time was shorter with the milk preacidified with CO ₂

(Table 9) mostly due to the decrease in cooking time before whey draining. The

target whey draining pH of 6.35 at the end of cooking was surpassed by the CO ₂ treatment when the curd reached the final cooking temperature (38°C), while the control required an additional 10 min of stir-out time at 38 C before the target pH was achieved (Table 9). The curds from the CO ₂ treatment and control were both milled at pH 5.3. There was no significant difference in time from drain to mill between the control and CO ₂ treatment (Table 9). A shorter (P < 0.05) total make time of the CO ₂ treatment was also reported by St-Gelais et al. (1997) using CO ₂ to preacidify milk to pH 6.56 prior to cheese making, however, they reported a similar time from cut to whey draining and a 30 min shorter Cheddaring time.

Table 9. Mean (n = 3) Cheddar cheese make times for control and CO ₂ treated cheeses.

Control CO ₂ LSD SEM mmutcb

Coagulant addition to drain 88 ^a 78 ^b 2.6 0.67

Drain to mill 88 83 NS 1.72

Total 176 ^a 161 ^b 8.6 2.20

Stir out 10 ^a 0 ^b 0.0 0.00 ^a> Means within a row that do not share a common superscript differ (P < 0.05). The effect of CO ₂ was visually apparent during cheese making.

Preacidification with CO ₂ produced a firmer coagulum that could be detected by touch and by the resistance of the coagulum when cut with wire knives. Other investigators have observed more rapid coagulation (Montilla et al., 1995; St- Gelais et al., 1997) and Van Slyke et al. (1903) observed that the rennetability of milk that had been heat treated above 85°C was restored with the addition of CO ₂ to milk prior to cheese making. After cutting, the curds of the CO ₂ treated material floated to the surface with increasing temperature during cooking whereas the curds of the control settled to the bottom if stirring was not constant. Floating curds may require a change in procedure if a portion of the whey is normally drained from the cheese vat through an outlet located about half-way between the surface of the whey and the bottom of the vat (i.e., predraining). In the case of floating curds, predraining could be accomplished by draining a

portion of the whey through the outlet located at the bottom of the cheese vat. If horizontally stirred cheese vats were used to manufacture cheese from milk containing CO ₂ , the thickness of the floating curd mass may be an issue that needs to be investigated with regard to curd integrity and fines. The fact that curds can be made to float in this process provides an opportunity to think about a different design of cheese vat and curd handling system that could reduce curd shattering.

Whey, Salt Whey, and Cheese Composition Whey and salt whey composition. A major portion of the CO ₂ added to the milk was removed with the whey at draining (Table 7). Means and composition differences of whey and salt whey due to CO ₂ treatment are reported in Table 10. CO ₂ treatment resulted in a higher (P < 0.05) fat content in the whey and salt whey. The calcium content was higher (P < 0.05) in the whey from the CO ₂ treatment and calcium content in the salt whey was lower (P < 0.05) than the control. There was no significant difference in whey CP content but there was a slight increase in the CP content of the salt whey. There was substantially less salt in the salt whey of the CO ₂ treatment than the control (Table 10). St-Gelais et al. (1997) reported higher (P < 0.05) fat content in the control whey versus the CO ₂ treatment (0.54 vs. 0.35%, respectively). The level of fat in whey reported by St-Gelais et al. (1997) would normally be associated with much lower fat recovery in cheese (Barbano and Sherbon, 1984) and does not seem consistent with the high fat recovery in the cheese (91.93 and 98.54%, respectively) reported in the same paper. The same authors did not detect a significant difference in the ash content of the whey, and salt whey composition was not reported.

Table 10. Mean (n = 3) whey, salt whey, and cheese composition.

Component Control CO ₂ treated LSD SEM percent

Whey

Fat 0.27 ^b 0.37 ^a 0.058 0.015

CP ¹ 0.85 0.86 NS 0.020

NPN ² 0.25 0.25 NS 0.005

TP ³ 0.60 0.60 NS 0.019

TS 6.71 ^b 6.87 ^a 0.098 0.025

Calcium 0.05 ^b 0.07 ^a 0.002 0.0004

Salt whey

Fat 1.28 ^b 4.07 ^a 1.174 0.299

CP 1.08 ^b 1.10° 0.023 0.006

NPN 0.40 0.40 NS 0.008

TP 0.68 ^b 0.71 ^a 0.024 0.006

TS 16.15 15.73 NS 0.210

Salt 8.73 ^a 6.73 ^b 0.856 0.218

Calcium 0.21 ^a 0.16 ^b 0.023 0.006

Cheese

Fat 34.19 ^a 32.83 ^b 1.024 0.261

FDB ⁴ 54.12 ^a 52.67 ^b 0.859 0.219

CP 23.72 24.02 NS 0.213

PDB ⁵ 37.49 38.09 NS 0.423

Moisture 36.84 37.67 NS 0.285

MNFS ⁶ 55.97 56.08 NS 0.246

Salt 1.44 ^b 2.24 ^a 0.119 0.030

Salt-in-the-

3.92 ^b 5.96 ^a 0.385 0.098 moisture

Calcium 0.69 ^a 0.52 ^b 0.035 0.009

⁶ MNFS = moisture in the nonfat substance.

Ultrafiltration is often used to fractionate whey before spray drying. Whey proteins, α-lactalbumin in particular, have been implicated in membrane fouling (Tong et al, 1989). In addition to protein, soluble calcium has been found to reduce flux (Ramachandra Rao et al., 1994). With respect to fouling due to protein, there was no significant difference in crude protein content of the whey (Table 10) between the control and the CO ₂ treatment. The calcium content of whey from the CO ₂ treatment (Table 10) was higher (P < 0.05). Experiments can determine if the whey from CO ₂ preacidification differs in ultrafiltration flux from that of typical Cheddar cheese whey. From the current literature it was unclear whether the higher calcium content of the whey from the CO ₂ treatment would impact the process of dehydration during spray drying, the rehydration of whey powder, or functionality of the whey protein concentrate compared to the control.

Cheese composition andpH. No difference (P < 0.05) in cheese crude protein, protein-on-a-dry basis, moisture, or moisture-in-the-nonfat substance was detected between the control and CO ₂ treatment. The fat content of the control cheese was higher (P < 0.05) than the CO ₂ treated cheese (Table 10). The CO ₂ treatment cheese contained less calcium (Table 10) due to the reduced milk pH prior to rennet addition (Table 8). The calcium content of the control was similar to the value of 0.721%, standard error was 10.770, listed in the UDSA National Nutrient Database (USDA, 2003), but the CO ₂ treatment calcium content was lower. Additional experiments can determine if the lower calcium content of the CO ₂ treatment cheese could reduce calcium lactate crystal formation during aging. The control cheese pH, 5.00, was lower (P < 0.05) than the CO ₂ treatment cheese pH, 5.09. The largest difference (P < 0.05) between the control and treatment cheeses was salt content. The control cheese had a salt content of 1.44% compared to 2.24% for the CO ₂ treatment. Thus, the salt-in- the-moisture content for the CO ₂ treatment (5.96%) was higher than the typical value (about 4.6%) for aged Cheddar. This could impact enzymatic changes during aging.

Component Recoveries The actual total recoveries (i.e., accountability) for all components were not influenced by the CO ₂ treatment. Actual total calcium, crude protein, fat, milk solids, and added salt recoveries for cheeses made from milk without and with added CO ₂ were 101.91 and 102.13%, 101.89 and 101.39%, 100.60 and 99.11%, 99.94 and 99.24%, 96.54 and 99.81%, respectively. Therefore, the actual recoveries were adjusted as described in the materials and methods section of this paper (see Table 11).

Table 11. Adjusted mean (n = 3) calcium, fat, CP, milk solids, and added salt recovery in whey, salt whey, and cheese.

Control CO ₂ LSD SEM

^■ pcioci IL

Calcium recovery

Whey 37.27 ^b 54.28 ^a 1.502 0.382

Salt whey 3.55 ^a 2.44 ^b 0.362 0.092

Cheese 59.07 ^a 43.38 ^b 1.669 0.425

Fat recovery

Whey 6.98 ^b 9.79 ^a 1.708 0.435 Salt whey 0.68 ^b 1.89 ^a 0.391 0.100 Cheese 93.08 ^a 87.57 ^b 1.921 0.489

CP recovery Whey 25.48 25.65 NS 0.448 Salt whey 0.66 0.60 NS 0.030 Cheese 74.11 73.51 NS 0.462

Milk solids recovery Whey 50.55 ^b 52.00 ^a 1.132 0.288 Salt whey 1.15 1.24 NS 0.069 Cheese 48.65 ^a 46.41 ^b 1.281 0.326

Salt recovery Salt whey 55.28 ^a 37.62 ^b 4.981 1.269 Cheese 43.04 ^b 64.05 ^a 5.748 1.464 ^a ' ^b Means within a row that do not share a common superscript differ (P < 0.05)

Calcium recovery. More (P # 0.05) calcium was recovered in the whey of the milk treated with CO ₂ (Table 11) than in the control (54.28 vs. 37.27%, respectively). Mean calcium recoveries were higher (P < 0.05) in the control salt whey and cheese compared to the CO ₂ treatment (Table 11). Adding CO ₂ to milk lowers the pH (Table 8) and causes an increase in calcium and phosphate concentrations in the serum phase of milk (Law and Leaver, 1998). The higher milk serum calcium content at coagulant addition was the likely cause for the firmer coagulum of the CO ₂ treatment and the higher calcium content of the CO ₂ treatment whey. The lower calcium recovery in the cheese reduced cheese calcium content from 0.69 to 0.52% (Table 10) was also due to the increased soluble calcium at coagulant addition.

Fat recovery. CO ₂ treatment cheeses had a lower (P < 0.05) fat recovery in the cheese than the control cheeses (Table 11). Almost 10% of the total milk fat was recovered in the whey from the CO ₂ treated milk compared to about 7% in the control. The fat recovery for the control was consistent with the assumption of 93% fat recovery in the Cheddar cheese theoretical yield formulae described in the material and methods section. The mean fat recovery in the salt whey of the CO ₂ treatment was more than twice that of the control (Table 11). Although the difference in fat content between the control and the CO ₂ treatment

was greater for the salt whey than the whey (Table 10), more fat was lost in the whey than in the salt whey, as shown in Table 11.

The lower pH of the milk with CO ₂ added prior to rennet addition changes the rate and firmness of milk coagulation. The coagulation for the milk with added CO ₂ was faster and firmer in this study. Johnson et al. (2001) varied coagulation firmness in a controlled study of composition and yield of 50% reduced fat Cheddar cheese and found an increase in fat loss with increased coagulation time and firmness at cut. Johnson et al. (2001) indicated that a coagulation that is cut soft will lose less fat and serum than a coagulation that is cut firm, if sufficient time is allowed for formation of the skin on the surface of curd particles after cutting and before stirring, hi the present study, the curd for the CO ₂ treatment was firmer than the control, but also was much lower in bound calcium content.

Why was less fat recovered in the CO ₂ treatment cheese than in the control cheese? The milk pH of the CO ₂ treatment (5.9) was closer to the acid pH optimum for chymosin. Higher enzyme activity may lead to excessive casein hydrolysis at coagulation rather than specific action on Nonspecific proteolysis of casein by chymosin would reduce the casein structure's ability to hold fat and higher fat losses would be observed. However, no significant difference in whey crude protein content or crude protein recovery in the cheese was detected between the control and CO ₂ treatment.

It is more likely that the lower calcium recovery, not casein hydrolysis, in the cheese played a role in the lower fat recovery in the cheese of the CO ₂ treatment. Increasing the milk serum calcium level by CO ₂ treatment is similar to adding CaCl ₂ , in that both produce a more firm milk coagulum. The end result of the two methods may be similar, but their impact on cheese composition differs. Adding CaCl ₂ to milk (0.01 to 0.02% w/w) would not be expected to decrease the bound calcium. On the other hand, acidifying milk (i.e., adding CO ₂ to a milk pH of 5.9) would decrease bound calcium and colloidal phosphorus (Law and Leaver, 1998) and increase soluble calcium. Thus, the bound calcium and probably the colloidal phosphorus content of the curd in the present study were lower than if the coagulum would have been formed with added calcium. Also, the bound calcium and phosphorus were probably lower in the CO ₂ treatment

than the control cheese indicated by the higher calcium content in the whey of the CO ₂ treatment. The lower calcium content may have altered the ability of the curd to retain fat during cooking, Cheddaring, salting, and pressing. Further work can identify the exact point in time and the cause for the higher fat loss in the whey when CO ₂ is used to decrease the milk pH to 5.9 for the manufacture of full- fat Cheddar cheese and this will aid in development of strategies to reduce fat loss during manufacture of full- fat Cheddar when CO ₂ is used in cheese making.

Crude Protein and milk solids recovery. No differences (P 30.05) were detected in mean crude protein recoveries between the treatments for the whey, salt whey, or cheese (Table 11). Total milk solids recovery did not include added salt only milk solids. Total milk solids recovery of the CO ₂ treatment was higher in the whey and lower in the cheese. The differences in milk solids recovery were generally consistent with the differences in fat and calcium recovery due to CO ₂ treatment.

Salt recovery. An unexpected result of the current study was the difference (P # 0.05) in salt recovery in the cheese between the control and CO ₂ treatment (Table 11). About 64% of the added salt was recovered in the cheese of the CO ₂ treatment compared to only 43% in the control cheese. The large difference (Table 10) in salt content between control and CO ₂ cheeses occurred even though the curd-salting rate was the same for both treatments. No difference (P > 0.05) was detected in cheese moisture (Table 10), but the CO ₂ treatment caused the salt-in-the-moisture to be one and a half times that of the control. St-Gelais et al. (1997) did not detect a difference in cheese salt content between the control and CO ₂ treatment. The CO ₂ treatment of St-Gelais et al. (1997) not only had a higher milk pH at coagulant addition (6.47), but a higher curd pH at salting (5.46). In our study, the coagulant and salt were added at lower pH values for the CO ₂ treatment than in the work of St-Gelais et al. (1997). It is unclear whether the lower calcium content of the curd at salting was responsible for the high salt uptake of CO ₂ treatment cheese. A marked improvement of added salt retention in Cheddar cheese, like the results shown in Table 11, would reduce salt wastes from a cheese manufacturing facility.

Cheese Yield and Yield Efficiency

Actual and adjusted cheese yields were significantly lower for the CO ₂ treatment (Table 12). The Van Slyke theoretical yield formula predicted the same yield for both the control and the CO ₂ treatment because the milk compositions were the same (Table 12).

Table 12. Mean (n = 3) actual, moisture and salt adjusted,

Van Slyke and Barbano theoretical cheese yields and cheese yield efficiencies. Control CO _j LSD SEM

Yield kg/100 kg

Actual 9.26 9.07 NS 0.052

Adjusted ¹ 9.29 ^a 8.86 ^b 0.302 0.077

Van Slyke 9.21 9.21 NA ² NA

Barbano 9.21 9.18 NA NA

Yield efficiency percent

^{a> b} Means within a row that do not share a common superscript differ (P < 0.05).

¹ Moisture adjusted to 37% and salt to 1.5%.

² NA = not applicable.

Cheese yields predicted by the Barbano theoretical yield formula for the CO ₂ treatment cheeses were lower because of the different calcium phosphate retention factor used for the CO ₂ treatment that allowed for the expected lower retention of calcium phosphate in the cheese. The cheese yield efficiency of the control was 100.9%. The mean fat recovery attained with the control cheese (Table 11) was consistent with the theoretical fat recovery (93%) of the theoretical yield formulae, which indicates that the cheese making methods alone did not create the fat loss observed in the CO ₂ treatment cheese. Preacidification with CO ₂ resulted in a 4.7% lower Van Slyke yield efficiency and a 4.4% lower Barbano yield efficiency than the control. The difference in yield efficiency within the CO ₂ treatment (0.3%) represents the reduction in yield due to mineral loss. The fat loss in the whey caused a greater difference in yield efficiency between the control and treatment than calcium loss.

These results indicate that cheese manufactured from milk acidified to a pH of 5.9 using approximately 1600 ppm of CO ₂ retained less calcium and fat than the control cheese. The higher loss of fat was primarily in the whey at

draining. Preacidification with CO ₂ did not alter the crude protein recovery in the cheese. The CO ₂ treatment resulted in a higher added salt recovery in the cheese and produced a cheese that contained more salt than needed. Considering the higher added salt retention of the CO ₂ treatment, the salt application rate can be lowered to achieve a typical cheese salt content. Cheese yield efficiency of the CO ₂ treated milk was 4.4% lower than the control due to fat loss. The use of CO ₂ led to several beneficial effects including better milk coagulation, reduced rennet use, less need for salt addition and potentially reduced problems with calcium crystal formation. Moreover, the CO ₂ procedure led to production of a reduced fat cheese. While such a reduction in fat is desirable in many instances, fat loss can be eliminated if desired by changes in the manufacturing procedure.

EXAMPLE 3: Less Water Migration and More Uniform Moisture Content Is Observed in Cheese Made from Milk Preacidified with CO ₂

Experimental Design

One 18-kg block of milled-curd Cheddar cheese (35 x 29 x 19 cm) was manufactured per treatment (from milk with added CO ₂ and without added CO ₂ ) on three different days. Pasteurized whole milk was carbonated to approximately 1600 ppm CO ₂ , which resulted in a milk pH at the vat of 5.93 compared to 6.65, at 31 ⁰ C, for the control. Cheese manufacturing conditions were kept constant for the two treatments with the exception the whey from the milk with added CO ₂ was drained at pH 5.96 compared to 6.35 for the control. The addition of CO ₂ decreased the total manufacturing time because of the shorter stir-out time. The usage rates of chymosin and salt were the same for both treatments. Cheeses were pressed overnight (17 h). When the cheeses were removed from the press the temperature in the center of the blocks was about 29°C. A more detailed description of cheese making conditions is described in Example 1. The CO ₂ content, titratable acidity (TA), pH, soluble nitrogen and casein degradation of the cheeses were monitored over 6 mo of aging at 6 ⁰ C. Changes in the water phase (monitored by analysis of expressible serum (ES)) were determined.

Sampling and Sample Preparation

Unsalted milled curd (USMC) and cheese sampling. Unsalted milled curd samples were taken after milling at pH 5.3, placed in plastic bags, and immediately prepared for removal of expressible serum (ES). Cheeses were sampled by removing three cross sections of cheese, with approximate dimensions of 1 cm by 28 cm by 19 cm, from the center of the block immediately after the block was removed from the press. The first cross section was vacuum packaged for compositional analysis. The second cross section was used for the expressible serum procedure. The sides of the third section were trimmed to leave a center piece of about 9 cm by 15 cm, which was vacuum- packaged and used for CO ₂ analysis. After the three slices were removed from the center of the block, the two remaining pieces of the block were placed into a plastic bag and vacuum packaged for further aging. Sampling was done again at approximately 30, 90, and 180 d.

Sample preparation of USMC and cheese. Cheeses and unsalted milled curd were cut into 2-cm pieces, ground (Model 31BL92, Waring, New Hartford, CT) into 2 to 3 mm, and packed into 59-mL snap lid vials leaving no head space and either analyzed fresh or they were frozen at -80°C and held until the time of analysis. Cheese slices for CO ₂ analysis were not ground, but were cut into approximately 3-mm pieces immediately before analysis.

Expressible Serum preparation. Expressible serum from unsalted milled curd and cheese immediately after pressing was collected at 25 ⁰ C as described in Guo and Kindstedt (1995), except that the samples were centrifuged at 23,500 x g. Expressible serum from several centrifuge bottles for each cheese treatment was combined to obtain a enough sample for chemical analyses. Expressible serum was placed in 59-mL snap-top vials and frozen at -8O ⁰ C.

Chemical Analyses Expressible Serum composition. Total nitrogen (TN) content of the expressible serum was determined in duplicate using the Kjeldahl method (AOAC, 2000; 33.2.11, 991.20). Crude protein was calculated by multiplying

the total nitrogen by 6.38. Calcium content was determined in duplicate by atomic absorption (Metzger et al., 2000).

USMC and cheese composition andpH. The fat content was determined by Babcock method (Marshall, 1992; 15.8.A). Moisture was determined gravimetrically by drying in a forced-air oven at 100°C for 24 h (AOAC, 2000; 33.2.44, 990.20) using a 2-g cheese test portion. Salt content was determined using the Volhard method (Marshall, 1992; 15.5.B). The Kjeldahl method (1-g test portion) was used to determine total nitrogen (Lynch et al., 2002) and crude protein was calculated (TN x 6.38). Fat and salt content were not determined for unsalted milled curd. Cheese pH was determined using a Xerolyt combination electrode (Model HA405; Mettler Toledo, Columbus, OH) with an Accumet pH meter (model AR 25, Fisher Scientific, Pittsburgh, PA) after tempering to 23 °C. Titratable acidity (TA) (AOAC, 2000; 33.7.14, 920.124) of the cheese was determined as described by Lau et al. (1991). All analyses were carried out in duplicate except total nitrogen and fat, which were performed in quadruplicate.

CO2 content of Milk and Cheese. A method of standard additions (MOSA) was used to determine the CO ₂ content of cheese by a modification of the method described by Ma et al. (2001) for determining the CO ₂ content of milk. The MOSA was selected because the control cheeses contained a small background level of CO ₂ (thus, no blank matrix was available) and because the technique is especially useful when an analyte (e.g., CO ₂ ) is present in low concentrations near the limit of quantification, which was the case for the control cheeses. In the MOSA, the sample is tested initially and with increasing added amounts of the analyte, essentially creating a calibration curve using the sample itself (Miller, 1991 ; Gonzalez et al., 1999)

For the initial CO ₂ determination, cheese was cut into approximate 3 -mm cubes and 20 ± 0.1 g was weighed into a small, stainless steel blender assembly (catalog number 14-15-18B, mini-sample container, 37-110 mL capacity, Fisher Scientific, Pittsburgh, PA). Then, 20 mL of degassed reverse osmosis purified water and 10 mL of 1 N sulfuric acid were added. The blender was immediately covered with Parafilm M (Pechiney Plastic Packaging, Chicago, IL) and tightly secured with a rubber band. The contents of the jar were blended at low speed for 30 s then for 15 s at 1-min intervals for a total of 5 blends over a 5 min

period. At 15 s after the last blend, a sticky nickel (catalog number 380-035, MOCON, Minneapolis, MN) was placed on the Parafilm M cover. The CO ₂ content in the headspace was determined by sampling with a gas-sampling needle inserted through the sticky nickel, taking care to keep the needle out of cheese slurry. The sampling needle was connected to an infrared CO ₂ analyzer (Pac Check 650, MOCON, Minneapolis, MN) previously calibrated with room air ("O" CO ₂ ) and 99.8% CO ₂ (catalog number 23402, manufactured for Supelco, Bellefonte, PA, by Scott Specialty Gases). A reading of the CO ₂ content (percent CO ₂ ) of the headspace was then taken. After the initial reading was obtained, the same procedure was repeated 5 times using a new 20 g portion of the same cheese sample each time. CO ₂ levels were increased in 30 to 50% increments over the previous reading. CO ₂ was added to the sample by decreasing the amount of degassed water initially added and substituting a corresponding volume of sodium bicarbonate standard solution (0.5 g/100 g, equivalent to ca 2.6 mg CO ₂ Zg or 2600 ppm CO ₂ ) so that the final amount of added standard solution and degassed water totaled 20 mL.

A MOSA linear regression equation (y = mx + b) was constructed from the initial and 5 determinations with added sodium bicarbonate standard solution, where y = instrument reading (% CO ₂ ), m = slope, x = sodium bicarbonate added (expressed as ppm CO ₂ in cheese) and b = intercept. The concentration of CO ₂ in an individual cheese was calculated by extrapolation of the regression equation to y = 0 and using the absolute value of x at y = 0. Visual inspection of the experimental data and the resulting coefficients of determination (R ² ξ 0.99) indicated the resulting regression equations were linear.

Proteolysis

Cheese pH 4.6 and 12% TCA soluble nitrogen as a percentage of total nitrogen (SNPTN) were determined in duplicate as described by Bynum and Barbano (1985). SDS-PAGE was performed as described by Neocleous et al. (2002) except a 7 μL sample (1 g of cheese per 10 mL of sample buffer) was loaded per lane for all cheese samples and a constant 15% concentration acrylamide gel was used. Results of the SDS-PAGE analysis were reported as

the ratio of αfe-casein and /3-casein to para-κ-casein. This was done to normalize the data for small variations in sample loading that can result from sample preparation since para-/c-casein is not hydrolyzed during aging (Nath and Ledford, 1973). The ratios of 04- casein and β- casein to para-κ -casein were used by Lau et al. (1991) and Neocleous et al. (2002) in the casein degradation calculations that were reported by those investigators. However, Lau et al. (1991) and Neocleous et al. (2002) used those ratios to calculate the percentage of casein degraded in the cheeses by using the first day of analysis as 0% of casein degraded. In those studies there was no difference between treatments at time zero. As a result, the data from the first day of analysis was not reported, hi the experiments reported herein, the first day of analysis was very important because of the differences between treatments immediately after the cheeses were removed from the press. During aging Qfe-casein and β- casein were hydrolyzed and their bands on the SDS-PAGE gel became less intense, while para-K- casein remained constant. A decreasing ratio indicates proteolysis of either O ₅ - casein or β- casein.

Expressible serum from cheese and unsalted milled curds were prepared using 0.9 niL of the sample buffer containing dithiothreitol as described by Verdi et al. (1987) and 0.1 mL of expressible serum. SDS was purchased from Sigma- Aldrich Chemical (L-4390; St. Louis, Missouri). A 10 to 20% SDS-PAGE gradient gel (Verdi et al., 1987) was used for expressible serum electrophoresis. Unsalted milled curd expressible serum gels were loaded with 16 μL of sample plus buffer per lane for both the control and CO ₂ treatment. Cheese expressible serum loadings of sample plus buffer for the control and CO ₂ treatment were 8 μL and 4 μL, respectively because the expressible serum from the CO ₂ treatment contained more protein than the control expressible.

The presence of Qferl-casein in the expressible serum of the CO ₂ treatment cheeses was determined by an additional experiment where an O ₈ - casein solution and a milk were separately incubated with chymosin. Samples were then analyzed with our SDS-PAGE procedure. After 1 hour of incubation a large protein band was present below the C^ ₁ -casein in the c^-casein solution treated with chymosin. After two hours that band was more pronounced. A band in the same location was present in the milk sample after incubation with

chymosin. The band appeared below the /3-casein. Since Qfcrl-casein was found by Creamer and Richardson (1974) to be the primary proteolytic product of chymosin action on α^-casein, this unknown band in the expressible serum of the CO ₂ treatment was likely o^-I-casein. The presence of the α^-I-casein band after the jS-casein band in the gels run as described herein differed from those in report of Malin et al. (1995). Protein migration patterns can be different due to different sources of SDS (Swaney et al., 1974). Although Malin et al. (1995) did not report their source of SDS, different sources of SDS were the likely cause of the Gkrl-casein migration differences.

Statistical Analysis

The PROC GLM procedure of SAS was used for all data analysis (SAS version 8.02, 1999 - 2001, SAS Institute Inc., Cary, NC). The least significant difference test (P < 0.05) was used to compare treatment means of the compositional data if the F-test for the statistical model was significant (P < 0.05). One-way ANOVA was used to analyze cheese and USMC composition data. For comparison of the control and CO ₂ treatment at any one sampling period (i.e. 0, 30, 90, and 180 d) a t-test was performed. ANOVA was used to analyze data over the aging period and least-square means are reported in the text for CO ₂ content, titratable acidity, pH, SNPTN, ofe-casein:para-κ -casein, and /3-casein:para-κ -casein of the control and CO ₂ treatment cheeses over the 6 mo of aging. Age was analyzed as a continuous variable. A mathematical transformation of the age variable was necessary to minimize multicollinearity of the linear and quadratic forms of the age variable (Glantz and Slinker, 2001). The transformation of age, age = day of storage at 6°C - ((last testing day - first testing day) / 2), made the data set orthogonal with respect to age. The quadratic term for age and the interaction of age by treatment were included in the statistical model if significant, or if not significant to show that the curvature was not detected, in the case of CO ₂ content during aging (FIG. 5).

RESULTS

USMC and Cheese Composition

No difference (P > 0.05) was detected between the control and CO ₂ treatment for unsalted milled curd moisture and crude protein (Table 13). As expected, the calcium content of the USMC was lower for the CO ₂ treatment (Table 13) because of the lower pH at draining (5.96 vs. 6.35). No difference in CP, PDB, moisture, and MNFS was detected between the control and CO ₂ treatment cheeses (Table 13). The fat content and fat on a dry basis were higher (P < 0.05) for the control cheese. The lower (P < 0.05) calcium content in the CO ₂ treatment cheese was expected, but the higher (P < 0.05) salt content (Table 13) of the CO ₂ treatment cheese was not expected.

Table 13. Mean (n = 3) unsalted milled curd (USMC) and Cheddar cheese composition.

Component Control CO ₂ LSD SEM perccui

USMC

Moisture 46.58 46.35 NS 0.834

Crude Protein 20.29 20.92 NS 0.273

Calcium 0.63 ^a 0.46 ^b 0.063 0.016

Calcium/Crude Protein (x 100) 3.09 ^a 2.20 ^b 0.344 0.088

Cheddar cheese

Fat 34.19 ^a 32.83 ^b 1.024 0.261

FDB ¹ 54.12 ^a 52.67 ^b 0.859 0.219

Crude Protein 23.72 24.02 NS 0.213

PDB ² 37.49 38.09 NS 0.423

Moisture 36.84 37.67 NS 0.285

MNFS ³ 55.97 56.08 NS 0.246

Salt 1.44 ^b 2.24 ^a 0.119 0.030

S alt-in-moisture 3.92 ^b 5.96 ^a 0.385 0.098

Calcium 0.69 ^a 0.52 ^b 0.035 0.009

Calcium/Crude Protein (x 100) 2.91 ^a 2.15 ^b 0.185 0.047 ^a> Means within a row without a common superscript differ (P < 0.05).

¹ FDB = fat on a dry basis.

² PDB = protein on a dry basis.

³ MNFS = moisture in the nonfat substance.

The least squares mean CO ₂ content of the treatment cheese (337 ppm) was higher (P < 0.01) than the control (124 ppm) and did not change during aging (Table 14, FIG. 5). The least squares mean pH of the control, 4.98, was lower (P < 0.01) than the treatment, 5.14. A linear age by treatment interaction was detected as well as a quadratic function of age (P < 0.01, Table 9, FIG. 6).

Table 14. Type III SS for cheese CO ₂ , pH, titratable acidity (TA),soluble nitrogen as a percentage of total nitrogen (SNPTN), and ratios of « _s -casein and /3-casein to para-κ -casein at 0, 30, 90, 180 d of aging.

„ , , 12% a,- β-

Factor df CO ₂ pH TA J _* ™-™ ^TCA CN:para CNrpara

SNPTN -K-CN -K-CN

Treatment ¹ (T) 1 273494 0.15 0.14 30 2 2 0.1

Age (A) 1 3668 < 0.01 0.61 ^* ' 914 ^** 236 ^** 16" 3.3 ^**

A x A 1 3686 0.02" 59 ^** 10" 5"

A x T 1 0.01" 0.03 ^* 1 ^*

Error 19 0.02 3

Error 20 27160 0.13 8 3

Error 21 1.9

R ² 0.91 0.89 0.85 0.99 0.99 0.88 0.64 P < 0.05.

^** P < 0.01. ¹ Treatments are cheeses manufactured from milk with and without added CO ₂

The least squares mean titratable acidity of the control cheese, 1.01%, was higher than (P < 0.01) than the CO ₂ treatment cheese, 0.87%, and was consistent with the difference in pH (FIG. 5 and 6). The titratable acidity, increased as a linear function of cheese age (Table 9, FIG. 4) and there was an age by treatment interaction with the titratable acidity of the control cheese increasing faster with age than the CO ₂ treatment cheese (FIG. 7).

Proteolysis

The CO ₂ treatment had higher (Table 14, P < 0.05) mean levels of pH 4.6 and 12% TCA soluble nitrogen as a percentage of total nitrogen (SNPTN) than the control immediately after pressing, 6.44% versus 4.79% and 2.71%

versus 2.03%, respectively (FIG. 8). During aging, the CO ₂ treatment had a higher (P < 0.01) least squares mean content of pH 4.6 SNPTN, 15.31%, than the control, 13.08%. The CO ₂ treatment also contained more (P < 0.01) 12% TCA SNPTN, 6.85%, than the control, 6.28%, during aging. The pH 4.6 and 12% TCA SNPTN increased in both the control and CO ₂ treatment over the 6 mo aging period (FIG. 7) both as a linear and a quadratic function of age (Table 14, P < 0.01). The level of pH 4.6 and TCA SNPTN were similar to levels reported previously for Cheddar cheese (Lau et al., 1991; Neocleous et al., 2002.) No difference (P > 0.05) in c4-casein:para-κ-casein and /3-casein:ρara-/c- casein ratios were detected between the unsalted milled curds of the control and CO ₂ treatment, data not shown. The difference in the 04-caseπr.para-κ -casein ratio between control and CO ₂ treatment cheeses was more pronounced at 0 d when the cheeses were removed from the press (FIG. 9) than at any other time. When the cheeses were removed from the press the CO ₂ treatment had a lower (P < 0.05) Gk-casein:para-/c-casein ratio than the control 2.48 and 3.87, respectively. The least squares mean c^-casein:para-κ -casein ratio of the CO ₂ treatment cheese, 1.26, was lower (P < 0.01) than the control cheese, 1.88, during 6 mo of aging (FIG. 9). The C4-casein:para-κ -casein ratio changed both as a linear and quadratic function of age and there was a linear age by treatment interaction (Table 14). No significant difference (P > 0.05) in /3-casein:para-κ- casein ratio was detected between the control and CO ₂ treatment immediately out of the press or during aging (FIG. 9, Table 14). The linear function of age was significant (Table 14), because the /3-casein:para~κ-casein ratio decreased in both the control and CO ₂ treatment cheeses during the aging period.

USMC and Cheese Expressible Serum

No difference (P > 0.05) in the amount of unsalted milled curd expressible serum was detected between the control and CO ₂ treatment (Table 8). There was a large decrease in the amount of expressible serum for both the control and CO ₂ treatment due to salting and pressing. After salting and pressing, almost twice the amount of expressible serum could be removed from the control cheese compared to the CO ₂ treatment (Table 8). Expressible serum from the

CO ₂ treated unsalted milled curds had a slightly higher (P < 0.05) CP content than the control. After salting and pressing the crude protein content of the expressible serum from the CO ₂ treatment cheese was much higher than the control (Table 15). The calcium content of the unsalted milled curd and cheese expressible serum from the CO ₂ treatment was lower than the control. Because the crude protein was higher and calcium was lower in the expressible serum of the CO ₂ treatment, the calcium expressed as a percentage of crude protein was much lower (P < 0.05) than the control for both the unsalted milled curd and cheese. Neither the CO ₂ treatment nor the control unsalted milled curd expressible serum contained a detectable amount of casein on an SDS-PAGE gel (data not shown). Casein was found in the CO ₂ treatment cheese expressible serum but not in the control cheese expressible serum (FIG. 9).

As shown in this Example, using milk preacidified with CO ₂ led to decreased water migration and a more uniform cheese moisture content. Moreover, less rennet and salt is needed when milk is preacidified before cheese making. The CO ₂ content of Cheddar cheese manufactured using milk preacidified with CO ₂ was consistently higher during aging than the control cheese. However, such increased CO ₂ content was not detrimental and actually provided certain benefits. Increased CO ₂ content can inhibit microbial growth, thereby increasing cheese shelf life. Hence, using preacidified milk for cheese making provides several benefits.

References

Association of Official Analytical Chemists, International ["AOAC"] 2000. Official Methods of Analysis. 17 ^th ed. AOAC, Gaithersburg, MD.

Aston, J. W. and L. K. Creamer. 1986. Contribution of the components of the water-soluble fraction to the flavour of Cheddar cheese. N. Z. J. Dairy

Sci. and Technol. 21:229-248.

Barbano, D. M. 1996. Mozzarella cheese yield: factors to consider. Pages 29 to 38 in Proc. Wisconsin Cheese Makers Mtg., Ctr for Dairy Res., Univ.

Wisconsin Madison. Barbano, D. M., and J. W. Sherbon. 1984. Cheddar cheese yields in New York.

J. Dairy Sci. 67:1873-1883.

Bynum, D. G., and D. M. Barbano. 1985. Whole milk reverse osmosis retentates for Cheddar cheese manufacture: chemical changes during aging. J. Dairy ScL 68:1-10.

Calvo, M. M., M. A. Montilla, and Agustin Olano. 1993. Rennet - clotting properties and starter activity on milk acidified with carbon dioxide. J.

Food Prot. 56:1073-1076.

Champagne, C. P., D. St-Gelais, and A. de Candolle. 1998. Acidification rates and population ratios of lactic starters in carbonated milk. Lebensm.- Wiss. u. Technol. 31:100-106. Creamer, L. K., R. C. Lawrence, and J. Gilles. 1985. Effect of acidification of cheeses milk on the resultant Cheddar cheese. N. Z. J. Dairy Sci Technol. 20:185-203. Creamer, L. K., and N. F. Olsen. 1982. Rheological evaluation of maturing

Cheddar cheese. J. Food Sci. 47:631-635,646. Creamer, L. K., and B. C. Richardson. 1974. Identification of the primary degradation product of o^-casein in Cheddar cheese. N. Z. J. Dairy Sci. Technol. 9:9-13. Dalgleish, D. G., and A. J. R. Law. 1988. pH-Induced dissociation of bovine

. casein micelles. I. Analysis of liberated casein. J. Dairy Res. 55:529-538. Dalgleish, D. G., and A. J. R. Law. 1989. pH-Induced dissociation of bovine casein micelles. II. Mineral solubilization and its relation to casein release. J. Dairy Res. 56:727-735. de Roos, A. L., R. J. Geurts, and P. Walstra. 2000. The association of chymosin with artificial casein micelles. Inter. Dairy J. 10:225-232. Fox, P. F. 1970. Influence of aggregation on the susceptibility of casein to proteolysis. J. Dairy Res. 37:173-180.

Glantz, S. A., and B. K. Slinker. 2001. MulticoUinearity and what to do about it. Pages 185-187 in Primer of Applied Regression and Analysis of Variance. 2 ^nd ed. McGraw-Hill, hie, New York, NY. Gonzalez, A. G., M. A. Herrador, and A. G. Asuero. 1999. Intra-laboratory testing of method accuracy from recovery assays. Talanta. 48:729-736. Grappin, R., T. C. Rank, N. F. Olson. 1985. Primary proteolysis of cheese during ripening- a review. L.Dairy Sci. 68:531-540.

Guo, M. R., J. A. Gilmore, and P. S. Kindstedt. 1997. Effect of sodium chloride on the serum phase of mozzarella cheese. J. Dairy Sci. 80:3092-3098. Guo, M. R., and P. S. Kindstedt. 1995. Age-related changes in the water phase of mozzarella cheese. J. Dairy Sci. 78:2099-2107. Holmes, D. G., J. W. Duersch, and C. A. Ernstrom. 1977. Distribution of milk clotting enzymes between curd and whey and their survival during

Cheddar cheese making. J. Dairy Sci. 60:862-869. Hotchkiss, J. H. and E. Lee. 1996. Extending shelf- life of dairy products with dissolved carbon dioxide. Eur. Dairy Mag. 8(3):16, 18-19. Johnson, M. E., C. M. Chen, and J. J. Jaeggi. 2001. Effect of rennet coagulation time on composition, yield, and quality of reduced-fat Cheddar cheese. J.

Dairy Sci. 84:1027-1033. King, J. S., and L. A. Mabbitt. 1982. Preservation of raw milk by the addition of carbon dioxide. J. Dairy Res. 49:439-447. King, J. S., and L. A. Mabbitt. 1987. The use of carbon dioxide for the preservation of milk. The use of carbon dioxide for the preservation of milk. Pages 35-43 in Preservatives in the Food, Pharmaceutical, and

Environmental Industries. Technical Series, Society for Applied

Bacteriology. No. 22. Blackwell Scientific Publications, Boston, MA. Lau, K. Y., D. M.. Barbano, and R. R. Rasumssen. 1991. Influence of pasteurization of milk on protein breakdown in Cheddar cheese during aging. J. Dairy Sci. 74:727-740. Law, A. J. R., and J. Leaver. 1998. Effects of acidification and storage of milk on dissociation of bovine casein micelles. J. Agric. Food Chem. 46:5008- 5016.

Lynch, J. M., D. M. Barbano, and J. R. Fleming. 2002. Determination of the total nitrogen content of hard, semihard and processed cheese by the

Kjeldahl method: collaborative study. J. AOAC. 85:445-455. Ma, Y., and D. M. Barbano. 2003a. Effect of temperature of CO ₂ injection on the pH and freezing point of milks and creams. J. Dairy Sci. 86: 1578-1589.

Ma, Y., and D. M. Barbano, and M. Santos. 2003b. Effect of CO ₂ addition to raw milk proteolysis and lipolysis at 4 C. J. Dairy Sci. 86:1616-1631.

Ma, Y., D. M. Barbano, J. H. Hotchkiss, S. Murphy, and J. M. Lynch. 2001.

Impact Of CO ₂ addition to milk on selected analytical testing methods. J.

Dairy Sci. 84:1959-1968.

Malin, E. L., M. H. Tunick, P. W. Smith, and V. H. Holsinger. 1995. Inhibition of proteolysis in mozzarella cheese prepared from homogenized milk.

Pages 237-246 in Chemistry of Structure-Function Relationships in

Cheese. E. L. Malin and M. H. Tunick, ed. Plenum Press, New York,

NY.

Marshall, R. T., ed. 1992. Standard Methods for the Examination of Dairy Products. 16 ^th ed. Am. Publ. Health Assoc, Inc., Washington, DC.

McCarney, T., W. M. A. Mullan, and M. T. Rowe. 1995. Effect of carbonation of milk on Cheddar cheese yield and quality. Milchwissenschaft 50:670-

674.

Metzger, L. E., D. M. Barbano, M. A. Rudan, and P. S. Kindstedt. 2000. Effect of preacidification on low fat Mozzarella cheese. I. Composition and

Yield. J. Dairy Sci. 83:648-658. Metzger, L. E., D. M. Barbano, P. S. Kindstedt, and M. R. Guo. 2001. Effect of milk preacidification on low fat Mozzarella cheese. II. Chemical and functional properties during storage. J. Dairy Sci. 83:1348-1356. Mickelsen, R., and C. A. Ernstrom. 1967. Factors affecting stability of rennin. J.

Dairy Sci. 50:645-650. Miller, J. N. 1991. Basic statistical methods for analytical chemistry. Part 2.

Calibration and regression methods. A review. Analyst. 116:3-14. Montilla, A., M. M. Calvo, and A. Olano. 1995. Manufacture of cheese made from CO ₂ treated milk. Z. Lebensm. Unters. Forsch. 200:289-292.

Mulvihill, D. M., and P. F. Fox. 1977. Proteolysis of o^-casein by chymosin: influence of pH and urea. J. Dairy Res. 553-540. Mistry, V. V., and K. M. Kasperson. 1998. Influence of salt on the quality of reduced fat Cheddar cheese. J. Dairy Sci. 81:1214-1221. Nath, K. R., and R. A. Ledford. 1973. Growth response of Lactobacillus casei variety casei to proteolysis in cheese during ripening. J. Dairy Sci.

56:710-715.

Nelson, B. K., J. M. Lynch, and D. M. Barbano. IN PROGRESS. Impact of preacidification with CO ₂ on Cheddar cheese composition and yield. (IN

PROGRESS - JDS-04-0302).

Neocleous, M., D. M. Barbano, and M. A. Rudan. 2002. Impact of low concentration factor microfiltration on milk component recovery and

Cheddar cheese yield. J. Dairy Sci. 85:2415-2424. Neocleous, M., D. M. Barbano, and M. A. Rudan. 2002. Impact of low concentration factor microfiltration on the composition and aging of

Cheddar cheese. J. Dairy Sci. 85:2425-2437. Ramachandra Rao, H. G., M. J. Lewis, and A. S. Grandison. 1994. Effect of soluble calcium of milk on fouling of ultrafiltration membranes. J. Sci.

Food Agic. 65:249-256. Ramkumar, C, L. K. Creamer, K. A. Johnston, and R. J. Bennett. 1997. Effect of pH and time on the quantity of readily available water within fresh cheese curd. J. Dairy Res. 64:123-134.

Rashed, M. A., N. M. Mehanna, and A. S. Mehanna. 1986. Effect of carbon dioxide on improving the keeping quality of raw milk. J. Soc. Dairy

Technol. 39:62-64.

Ruas-Madiedo, P., J. C. Bada-Gancedo, T. Delgado, M. Gueimonde, and C. G. de los Reyes-Gavilan. 2003. Proteolysis in rennet-coagulated Spanish hard cheeses made form milk preserved by refrigeration and addition of carbon dioxide. J. Dairy Res. 70:115-122. St-Gelais, D., C. P. Champagne, and G. Belanger. 1997. Production of Cheddar cheese using milk acidified with carbon dioxide. Milchwissenschaft. 52:614-618.

Swaney, J. B., G. F. V. Woude, and H. L. Bachrach. 1974. Sodium dodecylsulfate-dependent anomalies in gel electrophoresis: Alterations in the banding patterns of foot-and-mouth disease virus polypeptides. Anal.

Biochem. 58:337-346. Tong, P. S., D. M. Barbano, and W. K. Jordan. 1989. Characterization of proteinaceous membrane foulants from whey ultrafiltration. J. Dairy Sci.

72:1435-1442.

Van Slyke, L. JL, H. A. Harding, and E. B. Hart. 1903. Rennet-enzyme as a factor in cheese-ripening. New York State Agric. Exp. Stn. Bulletin no. 233. Cornell University. Ithaca, NY.

Verdi, R. J., D. M. Barbano, and M. E. DellavaUe. 1987. Variability in true protein, CN, nonprotein nitrogen and proteolysis in high and low somatic cell count milks. J. Dairy Sci. 70:230-242.

U. S. Department of Agriculture, Agricultural Research Service. 2003. USDA National Nutrient Database fro Standard Reference, Release 16. Nutrient Data Laboratory Home Page, http://www.nal.usda.gov/fnic/foodcomp. All patents and publications referenced or mentioned herein are indicative of the levels of skill of those skilled in the art to which the invention pertains, and each such referenced patent or publication is hereby incorporated by reference to the same extent as if it had been incorporated by reference in its entirety individually or set forth herein in its entirety. Applicants reserve the right to physically incorporate into this specification any and all materials and information from any such cited patents or publications.

The specific methods and compositions described herein are representative of preferred embodiments and are exemplary and not intended as limitations on the scope of the invention. Other objects, aspects, and embodiments will occur to those skilled in the art upon consideration of this specification, and are encompassed within the spirit of the invention as defined by the scope of the claims. It will be readily apparent to one skilled in the art that varying substitutions and modifications may be made to the invention disclosed herein without departing from the scope and spirit of the invention. The invention illustratively described herein suitably may be practiced in the absence of any element or elements, or limitation or limitations, which is not specifically disclosed herein as essential. The methods and processes illustratively described herein suitably may be practiced in differing orders of steps, and that they are not necessarily restricted to the orders of steps indicated herein or in the claims. As used herein and in the appended claims, the singular forms "a," "an," and "the" include plural reference unless the context clearly dictates otherwise. Thus, for example, a reference to "an antibody" includes a plurality (for example, a solution of antibodies or a series of antibody

preparations) of such antibodies, and so forth. Under no circumstances may the patent be interpreted to be limited to the specific examples or embodiments or methods specifically disclosed herein. Under no circumstances may the patent be interpreted to be limited by any statement made by any Examiner or any other official or employee of the Patent and Trademark Office unless such statement is specifically and without qualification or reservation expressly adopted in a responsive writing by Applicants.

The terms and expressions that have been employed are used as terms of description and not of limitation, and there is no intent in the use of such terms and expressions to exclude any equivalent of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention as claimed. Thus, it will be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.

The invention has been described broadly and generically herein. Each of the narrower species and subgeneric groupings falling within the generic disclosure also form part of the invention. This includes the generic description of the invention with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

Other embodiments are within the following claims, hi addition, where features or aspects of the invention are described in terms of Markush groups, those skilled in the art will recognize that the invention is also thereby described in terms of any individual member or subgroup of members of the Markush group.