Freezing Quality of High Moisture Dairy Products and Dairy Waste Utilization A Thesis Presented for the Master of Science Degree The University of Tennessee, Knoxville Ripley Vaughan December 2024 ii ACKNOWLEDGEMENTS I’d like to first thank my major advisor, Dr. Toni Wang, and committee members Dr. Vermont Dia, and Dr. Liz Eckelkamp for providing guidance and support throughout my graduate program. I’d like to make a special thanks for Dr. Wang for playing an instrumental role in my growth as a researcher. Her hands-on guidance and mentorship has led me to a point I did not think I was capable of two years ago. I’d also like to acknowledge the University of Tennessee Food Science Faculty and Staff for assisting me in all other aspects other than research and providing an encouraging and friendly environment to work in with the mindset to always pursue learning. I’d like to next acknowledge the USDA Southeast Dairy Innovation Initiative for funding of my projects. I’d also like to thank Sweetwater Valley Farm for providing the whey for my project and being very accommodating and helpful with all my questions and requests for many weeks. Lastly, I’d like to thank my friends and family who I truly could not have done this without. They gave me support and encouragement to continue to be resilient throughout my program. iii OVERALL ABSTRACT Freezing food products is a common method used to extend shelf life but products with high-moisture content, such as sour cream and yogurt, or even low-moisture mozzarella are unable to be frozen due to a decrease in textural quality. The research objective of the first chapter of the thesis is to create a food grade antifreeze agent and evaluate its effect against freeze induced damage in sour cream and yogurt. Whey protein isolate and soy protein isolate were complexed with locust bean gum and lambda carrageenan and their hydrolysates. The complexes were tested for ice recrystallization inhibition (IRI) activity at pH 7.0 and 4.5 in 20 mM NaCl using splat assay. The freezing induced damage was evaluated through texture analysis and microscopic imaging. The complexes showed increased IRI activity at both pHs compared to the controls. However, once added to the sour cream and yogurt they were shown to be ineffective at protecting against freeze induced damage. The unhydrolyzed and hydrolyzed products resulted in decreased texture, cohesiveness, and consistency in both the sour cream and yogurt. This demonstrated that IRI active complexes in a model system does not always translate to similar effectiveness in the food matrix. The second chapter of the thesis focuses on industrial cheese whey having the potential to be utilized as a source of health beneficial phospholipids, particularly sphingomyelin and phosphatidylserine. Phospholipids are located in the milk fat globule membrane (MFGM) and this component in whey can be precipitated out using thermocalcic aggregation. The objective was to optimize the extraction conditions of phospholipids from cheese whey by manipulating pH, calcium acetate concentration, and temperature. The optimum conditions of 50 mM, pH of 6.5, and a temperature of 60 ºC iv were chosen by evaluating total lipid yield, phospholipid yield, and protein and salt distributions. The results showed that an optimum phospholipid yield of 92% and distribution, 96% of proteins retained in the supernatant, and a salt recovery of 53% was achieved under these conditions. This project demonstrated a simple and cost-effective method to optimize the extraction of phospholipids from industrial cheese whey. v TABLE OF CONTENTS OVERALL ABSTRACT ................................................................................................... iii GENERAL INTRODUCTION ........................................................................................... 1 High-Moisture Dairy Products and Freeze-Induced Damage in Dairy Products ........... 1 Sour cream and yogurt production and product composition ..................................... 1 Freeze induced damage in dairy products ................................................................... 3 Mechanisms of ice recrystallization............................................................................ 4 Common additives used in dairy products as stabilizers and textural enhancers ....... 6 Why complexing proteins and polysaccharides? ........................................................ 8 Dairy Processing By-Products and Value Capturing ...................................................... 9 Cheese whey as a dairy by-product ............................................................................ 9 Common utilizations of whey ................................................................................... 10 Dairy phospholipid as valuable components ............................................................ 10 Milk fat globule membrane and its distribution and recovery by precipitation ........ 11 Previous extraction techniques and gaps in phospholipid extraction from cheese whey .......................................................................................................................... 13 Thesis objectives and organization ........................................................................... 14 References ..................................................................................................................... 16 CHAPTER I THE EFFECT OF PROTEIN AND POLYSSACHARIDE COMPLEXES ON FREEZE INDUCED DAMAGE IN SOUR CREAM AND YOGURT .................... 26 Abstract ......................................................................................................................... 27 Introduction ....................................................................................................................... 28 Materials and Methods ...................................................................................................... 31 vi Materials for complex preparation and sour cream and yogurt production .............. 31 Complex preparation ..................................................................................................... 32 WPI and SPI hydrolysates created by Alcalase hydrolysis ...................................... 32 Polysaccharide hydrolysates created by cellulase hydrolysis ................................... 33 Protein and polysaccharide complexation through pH cycling ................................ 34 Formation of SPI hydrolysate and lecithin complex................................................. 35 Complex characterization ............................................................................................. 35 Determination of complexe’s IRI activity using Splat Assay ................................... 35 Analysis of particle size of complexes using the Zetasizer ...................................... 35 Confocal microscopic observation for complexation validation .............................. 36 Amphiphilicity of complexes qualitatively evaluated by emulsion stability testing 36 Evaluation of antifreeze activity of complexes in dairy products................................. 37 Sour cream sample preparation and freezing conditions .......................................... 37 Yogurt preparation and freezing conditions ............................................................. 37 Texture analysis to evaluate freeze induced damage ................................................ 38 Qualitative analysis of freeze induced damage by confocal laser scanning microscopy (CLSM) ................................................................................................. 39 Syneresis measured by water separation................................................................... 40 Statistical Analysis .................................................................................................... 40 Results and Discussion ................................................................................................. 40 Degree of hydrolysis of the polysaccharide and protein hydrolysates ..................... 40 Ice recrystallization inhibition (IRI) activity of protein-polysaccharide and hydrolysate complexes .............................................................................................. 43 vii Amphiphilicity of complexes determined by emulsification and stability ............... 45 Particle size analysis of unhydrolyzed and hydrolyzed complexes .......................... 48 Confocal image analysis for complexation validation .............................................. 51 Effect of protein-polysaccharide complexes on freeze induced damage in sour cream and yogurt ..................................................................................................................... 53 Texture analysis to quantify freeze induced damage ................................................ 53 Lacunarity of freeze induced damage ....................................................................... 56 Degree of syneresis measured water displacement quantification. .......................... 59 Qualitative microstructure analysis of freezing induced damage ............................. 61 Conclusion .................................................................................................................... 63 References ..................................................................................................................... 65 CHAPTER II THE OPTIMIZATION OF THE CONCENTRATION OF PHOSPHOLIPIDS FROM CHEESE WHEY .................................................................. 71 Abstract ......................................................................................................................... 72 Introduction ................................................................................................................... 73 Materials and Methods .................................................................................................. 75 Materials ................................................................................................................... 75 Local dairy facility for whey sampling ..................................................................... 75 Evaluation of the effect of calcium acetate concentration on MFGM precipitation . 76 Evaluate the effect of pH on MFGM precipitation ................................................... 76 Evaluation of the effect of temperature on MFGM precipitation ............................. 77 Extraction of the lipids from calcium acetate precipitated pellet for quantification and yield determination............................................................................................. 77 viii Lipid extraction from whey for total lipid quantification ......................................... 77 Protein quantification of supernatant and MFGM pellet using the bicinchoninic acid (BCA) method ........................................................................................................... 78 Moisture analysis of MFGM pellet to quantify salt recovery under optimum conditions .................................................................................................................. 78 31P-NMR quantification of phospholipids and yield calculation .............................. 79 Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis (SDS-PAGE) for protein characterization and distribution under optimum conditions ....................... 79 Statistical analysis ..................................................................................................... 80 Results and Discussion ................................................................................................. 80 Effect of calcium acetate concentration on MFGM precipitation and protein distribution ................................................................................................................ 80 Effect of pH on MFGM precipitation and protein distribution ................................. 85 The effect of temperature on MFGM precipitation .................................................. 87 Composition of MFGM pellet under optimum conditions of precipitation .............. 91 Characterization of protein partition under optimum conditions .............................. 91 The effect of whey storage on MFGM precipitation and lipid yield ........................ 94 Effect of drainage stage or time whey is collected from vat ..................................... 94 Conclusion .................................................................................................................... 95 References ..................................................................................................................... 97 OVERALL CONCLUSION ........................................................................................... 101 VITA ............................................................................................................................... 103 ix LIST OF TABLES Table 1: Composition comparison of sour cream and yogurt on as-is basis* .................... 2 Table 2: Composition comparison of whey and beta stream. * ........................................ 14 Table 3: Reducing sugar content after 48 hr cellulase hydrolysis of LBG and LC samples (of 10 mg/mL). .......................................................................................................... 41 Table 4: Native protein and hydrolysate MW distribution determined by HPLC-SEC. .. 42 Table 5: IRI activity of unhydrolyzed and hydrolyzed complexes tested at 2% in 20 mM NaCl at pH 4.5 and 7.0.............................................................................................. 44 Table 6: Particle size of protein and polysaccharides dispersions (1 mg/mL) at pH 4.5 and 7.0 in 20 mM NaCl. .................................................................................................. 49 Table 7: Texture analysis from a single penetration test of frozen-thawed sour cream with unhydrolyzed and hydrolyzed protein-polysaccharide complexes added at 3% weight as is. ............................................................................................................... 54 Table 8: Texture analysis from a single penetration test of frozen-thawed yogurt with unhydrolyzed and hydrolyzed protein-polysaccharide complexes added at 3% weight as is. ............................................................................................................... 55 Table 9: Phospholipid composition of MFGM pellets under different calcium acetate concentration, pH, and temperature treatments ........................................................ 84 Table 10: Whey composition (as-is) determined by FOSS Milkoscan*. .......................... 96 x LIST OF FIGURES Figure 1: Oil in water emulsions (94% water, 5% oil, 1% (w/v) complexes) at pH 7.0 after 10 minutes prepared by homogenizing at 10,000 rpm for 3 minutes. U- indicates unhydrolyzed and H- is hydrolyzed. Negative control is only water and oil and the positive control is oil and water with Tween 80 (1%). ................................ 46 Figure 2: Top panel (A and B); Particle size distribution of A) unhydrolyzed WPI and LBG complex and controls and B) hydrolyzed WPI and LBG complex and controls at pH 7 in 20 mM NaCl. Bottom panel (C and D); Particle size distribution of hydrolyzed SPI and lecithin complex and controls at C) pH 4.5 and D) pH 7 in 20 mM NaCl. ................................................................................................................. 50 Figure 3: Confocal images of protein and polysaccharide complexes at 5% (w/v). Green indicates protein and red indicates polysaccharide. In each set of pictures, the bottom left image is the two wavelengths overlayed. U-indicates unhydrolyzed and H- indicates hydrolyzed. Samples visualized at 63x objective. ..................................... 52 Figure 4: Lacunarity, using a sliding box test, of frozen-thawed sour cream (top) and frozen-thawed yogurt (bottom) with unhydrolyzed and hydrolyzed complexes and controls. Standard deviation is between two replicates that each had three images evaluated. Different letters indicate significantly different treatment outcomes (P<0.05). ................................................................................................................... 57 Figure 5: Microscopic image of sour cream before (left) and after (right) freezing. ........ 59 Figure 6: Syneresis, expressed as water (%) relative to the total volume, of sour cream (top) frozen with hydrolyzed complexes and yogurt (bottom) frozen with unhydrolyzed and hydrolyzed complexes. Water was measured after 10 mL of xi sample settled overnight. Standard deviation was between two replicates. Different letters indicate significantly different samples (P<0.05). ......................................... 60 Figure 7: Confocal images of fresh and frozen yogurt with unhydrolyzed (U-) and hydrolyzed (H-) complexes and controls added at 3% as-is weight basis. Samples visualized at 10x objective. Red indicates fat and green indicates protein. .............. 62 Figure 8: Confocal images of fresh and frozen sour cream with hydrolyzed complexes added at 3% weight as is. Samples visualized under 10x objective. Red indicates fat and green indicates protein. ...................................................................................... 64 Figure 9: The effect of calcium acetate concentration on total lipid yield, phospholipid (PL) yield, and PL concentration relative to total sample lipid (top) and protein distribution (bottom). MFGM precipitation performed at pH 6.5 and 55ºC. Standard deviation is between three trials from three different batches of whey. Different mean comparison letters indicate significantly different salt concentration treatments (P<0.05). ................................................................................................................... 81 Figure 10: The effect of pH with 25 mM calcium acetate and 55 ºC on total lipid yield, PL yield, and PL concentration relative to the total lipid in the sample (top) and protein distribution (bottom). Standard deviation is between one replicate from two different batches of whey. Different mean comparison letters indicate significantly different treatments within the same line (P<0.05). .................................................. 86 Figure 11: The effect of temperature at 25 mM calcium acetate and pH 6.5 on total lipid yield, PL yield, and PL concentration relative to total lipid (top) and protein distribution (bottom). Standard deviation was between two replicates from two different batches of whey. ......................................................................................... 88 xii Figure 12: Mass distribution of MFGM pellet and supernatant under optimum conditions of calcium acetate concentration of 50 mM, pH 6.5, and temperature of 60 ºC. Standard deviation is between three replicates from a single batch of whey. .......... 92 Figure 13: SDS-PAGE of protein distribution in the supernatant and MFGM pellet under optimum precipitation conditions of a calcium acetate concentration of 50 mM, pH 6.5, and temperature of 60 ºC. Protein marker is on the left with molecular weight (kDa). ........................................................................................................................ 93 1 GENERAL INTRODUCTION High-Moisture Dairy Products and Freeze-Induced Damage in Dairy Products Sour cream and yogurt production and product composition Fermented high-moisture dairy products like yogurt, Greek yogurt, kefir, and sour cream have become increasingly popular due to them being an excellent source of calcium and protein as well as nourishing the bacterial microbiota in the gastrointestinal tract (Bourrie et al., 2016, García-Burgos et al., 2020). Whole milk yogurt is made through the fermentation of milk with Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus and is made up of 85.3% water, 3.83% protein, 4.48% fat, and 5.57% carbohydrates (21CFR131.200, 2023; USDA 2019). The salt used in yogurt is sodium chloride and it is typically added at 0.04% (USDA, 2019). Yogurt production consists of three main steps: modification of milk composition, pasteurization and homogenization, and fermentation. Depending on the desired final fat content, yogurt is modified by either separating fat out through the use of centrifuges or by adding milk fat, allowing yogurts to be created with fat contents ranging from 1-10% (McHugh, 2015). In addition, the total solids content of the milk is increased to about 16% by either water evaporation or through the addition of milk powder, whey protein powder concentrates, and casein powders (McHugh, 2015). The increase in total solids content helps produce a firmer and more stable yogurt. Pasteurization is performed to eliminate harmful or unwanted microorganisms and denatures the whey protein, contributing to a thicker texture (McHugh, 2015). Homogenization, which is typically carried out between 1-20 MPa, disrupts the milk fat globules and decreases their particle size from 1-8 m to 0.3-0.8 m resulting in a creamier and 2 more stable product (Thiebaud et al., 2003). The milk is then cooled to 40-44 ºC before adding the culture for fermentation (McHugh, 2015). Per the Code of Federal Regulations, the final pH of the yogurt must be below 4.6 and typically ranges between 4.0-4.5 (21CFR131.200, 2023, McHugh 2015). According to the USDA (2001) specifications, yogurt should be firm like a custard, smooth, and homogeneous. Sour cream’s composition is similar to that of yogurt with it being 72.60% water, 3.07% protein, 18.0% fat, and 5.56% carbohydrates (USDA 2022). It also has a similar salt content of 0.05% and sodium chloride is typically used (USDA, 2022). Sour cream is made through the fermentation of cream, with a fat content between 18-20%, using a variety of lactic acid bacteria (Niamsiri and Batt, 2009). The process of sour cream production is similar to that of yogurt as it goes through pasteurization, homogenization, and fermentation (Niamsiri and Batt, 2009). The fermentation temperature of sour cream is lower than yogurt, at about 22-24 ºC and the length of fermentation is longer, usually between 14-16 hours (Costello, 2009). The final pH is also acidic at 4.5. According to the USDA (2000) specifications, sour cream should be thick and smooth, should be able to be formed into a mound with a spoon, and free of lumps or graininess. A comparison of yogurt and sour cream’s composition is shown in Table 1. Table 1: Composition comparison of sour cream and yogurt on as-is basis* Composition on as-is basis Water (%) Fat (%) Protein (%) Carbohydrates (%) Salt (%) pH Yogurt 85.30 4.48 3.83 5.57 0.05 < 4.6 Sour Cream 72.60 18.00 3.07 5.56 0.04 <4.6 *(USDA 2019, 2022) 3 Freeze induced damage in dairy products Freezing is a common method for extending a food product’s shelf-life and slowing down various chemical reactions and microbial activities. However, freezing can significantly decrease product quality, especially products with a high-moisture content, as a result of cryo- concentration, ice recrystallization and growth, and protein dehydration, conformational changes, and aggregation (Diefes et al., 1993, Alinovi et al., 2021, Sun et al., 2022). This kind of freeze induced damage has mostly been seen in low and high-moisture mozzarella (Diefes et al., 1993, Reid and Yan, 2004, Kuo and Gunasekaran, 2009, Alinovi and Mucchetti, 2020). Ice recrystallization and growth in dairy products during freezing can cause leakage of fat from ruptured milk fat globules resulting in large serum channels in the protein matrix (Alinovi et al., 2020). Such damage can cause coalescence, graininess or coarse texture and syneresis in dairy products (Alinovi et al., 2020, Tribst et al., 2020). Most research investigating freeze induced damage in dairy products has been conducted using ice cream and low and high-moisture mozzarella (Kuo and Gunasekaran, 2009, Damodaran and Wang, 2017, Alinovi and Mucchetti, 2020, Alinovi et al., 2021, Reeder et al., 2023). In ice cream, ice recrystallization can result in a coarser texture due to the larger ice crystal sizes (Hartel, 1996, Adapa et al., 2000, Soukoulis and Fisk, 2016) and the protein matrix in mozzarella can become more reticular and porous most likely due to protein dehydration (Kuo and Gunasekaran, 2009). This change in protein structure can increase meltability and decrease stretchability of the mozzarella (Kuo and Gunasekaran, 2009). Freezing can also cause a less cohesive and more rigid structure (Alinovi and Mucchetti, 2020). Deterioration of the protein matrix is one of main contributors to the texture quality loss observed in high-moisture dairy products such as mozzarella. During freezing, the protein matrix 4 can physically break due to dehydration of the proteins and the expansion of water when it crystalizes into ice (Diefes et al., 1993, Kuo and Gunasekaran, 2009). This local dehydration can prevent the protein’s ability to rebind to water (Kuo and Gunasekaran, 2009). As bound water is released from the proteins during freezing it causes an increase in unbound water resulting in syneresis (Alinovi et al., 2020). The migrated water can occupy large serum channels formed from the compressed protein matrix (Alinovi et al., 2020). These mechanisms of damage to the food matrix in dairy products during freezing can have a negative effect on textural attributes. Ice growth can cause protein dehydration resulting in a more porous structure (Graiver et al., 2004, Kuo and Gunasekaran, 2009). The development of large serum channels and a more porous structure from the dehydration and compaction of the protein matrix can also result in the accumulation of fat globules in these pockets (Kuo and Gunasekaran, 2009, Alinovi et al., 2020). This can cause decreased firmness and cohesiveness (To et al., 2020). This freeze induced damage can cause dairy products texture to become mealy or grainy after freezing (Kasprzak et al., 1994). A grainy mouthfeel is caused by protein aggregation which occurs when the protein becomes denatured. Mealiness can also be caused by the separation of moisture from the protein network (Webb and Arbuckle, 1977, Kasprzak et al., 1994). Mechanisms of ice recrystallization Mechanisms of ice recrystallization include Ostwald ripening which is the growth of larger ice crystals at the expense of smaller ones, accretion which is the merging of two ice crystals, and isomass rounding where rougher, less smooth crystals tend to change to smoother surfaces (Hartel, 1998, Budke and Koop, 2006, Midya and Bandyopadhyay, 2024). Ice recrystallization typically occurs during transportation of frozen foods due to fluctuating 5 temperatures. In recent years, supply chain issues have created the need to develop molecules with ice recrystallization inhibition (IRI) activity that can be used in frozen foods to prevent damage and lengthen shelf life. Some of these IRI active molecules include naturally occurring antifreeze proteins found in polar fish, small synthetic molecules like synthetic polyvinyl alcohol (PVA), plant-based protein and polysaccharide hydrolysates, and stabilizers (Knight et al., 1984, Budke and Koop, 2006, Kamińska-Dwórznicka et al., 2015, Fomich et al., 2023, Sun et al., 2023, Yuan et al., 2024). There are a few theories as to how these IRI active molecules help prevent the growth of ice crystals. One molecular attribute that has been found to have increased IRI activity is amphiphilicity (Graham et al., 2018). Facially amphiphilic molecules can also have enhanced IRI activity due to the hydrophobic motif interacting or facing the ice plane and the hydrophilic motif facing and interacting with the local unfrozen water (Biggs et al., 2017, Graham et al., 2018). Amino acid composition has also been shown to impact IRI activity and one study highlighted the importance of peptides with a repeating glycine-proline-and any other amino acid sequence (Damodaran and Wang, 2017, Luo et al., 2023). Antifreeze proteins with alpha helix secondary structure have also shown improved IRI activity due to its more stable conformation (Mochizuki and Molinero, 2018, Yuan et al., 2024). Molecular weight has shown to be an important characteristic in both proteins and polysaccharides (Fomich et al., 2023, Sun et al., 2023, Yuan et al., 2024). Smaller molecular weight molecules could allow for more mobility at the ice water interface, however Fomich et al. (2023) has found that in some cases there seems to be a need for both large and small peptides in order to have increased IRI activity. There is another theory that hypothesizes that when these molecules interact or bind with the ice surface they disrupt the water in the quasi-layer creating a curvature that makes it unfavorable for more 6 water to bind to, therefore reducing the growth of ice crystals (Weng et al., 2018, Midya and Bandyopadhyay, 2024). Common additives used in dairy products as stabilizers and textural enhancers Hydrocolloids are commonly added to dairy products to help stabilize, thicken, and form milk gels (Corredig et al., 2011, Yousefi and Jafari, 2019). Hydrocolloids can also be used to help prevent instabilities such as wheying-off or syneresis and preventing the flocculation of casein micelles while also extending its shelf life (Yousefi and Jafari, 2019). Some frequently used polysaccharides in dairy products include guar gum, carrageenan, locust bean gum, xanthan gum, and carboxymethyl cellulose (Yousefi and Jafari, 2019). Some gums are often used in combination because of their synergistic effects, such as locust bean gum and carrageenan or xanthan gum, providing ice cream with a creamier mouth feel (Kulkarni and Shaw, 2015). Tamarind seed polysaccharides has also been shown to prevent ice recrystallization in ice cream mixes (Sun et al., 2024). Hydrocolloids, such as carrageenan or carboxy methyl cellulose that possess a negative charge are able to interact with the positively charged areas of casein micelles providing a more stable network while neutral hydrocolloids, i.e. locust bean gum, can increase the viscosity of the continuous phase, reducing syneresis (Arab et al., 2023). Proteins and polysaccharides are commonly added individually to dairy products to help improve their textural properties, but they are not used as complexes. Proteins and polysaccharides have been shown to form functional complexes through electrostatic interaction (Schmitt et al., 1998). Oppositely charged biopolymers can form complexes through these associative electrostatic interactions (Corredig et al., 2011). The formation of these complexes is dependent on many factors including pH, ionic strength, temperature, and biopolymer ratio (Corredig et al., 2011). 7 The charges of the biopolymers change based on the pH and therefore can affect their complexation (Igartúa et al., 2022). If proteins and polysaccharides possess the same charge, for instance above the proteins isoelectric point, there will be repulsive forces. On the other hand, if the pH is below the proteins isoelectric point and the protein is positively charged then it will be attracted to the anionic polysaccharide and have associative behavior. Since one of the main mechanisms of complexation is through electrostatic interaction, ionic strength can also affect this. Salts can screen the charges of the molecules and reduce the electrostatic interactions (Niu et al., 2014). A study by Niu et al. (2014) saw that coacervate formation of ovalbumin and gum Arabic was not significantly affected when the NaCl concentration was below 10 mM, but as the salt concentration increased they saw a significant effect on the coacervation formation (Niu et al., 2014). They attributed this to the competitive nature of Na+ ions with ovalbumin for binding sites on the gum Arabic (Niu et al., 2014). Biopolymer ratio and concentration can also have an effect on complexation behavior. A study by Benichou et al. (2007) evaluated the behavior of whey protein and xanthan gum complexes formed at different concentrations. They found that at low protein concentrations (less than 4% whey protein isolate) and low polysaccharide concentrations (less than 0.5% xanthan gum) the two polymers were co-soluble because of the sufficient distance between interacting molecules (Benichou et al., 2007). As the concentration increases until a certain point, conjugated biopolymers are formed with new physiochemical properties (Benichou et al., 2007). A ratio of 5:1 WPI and xanthan gum fell within the region where these two molecules are co-soluble and can interact to form these conjugated biopolymers. At higher protein- polysaccharide concentrations (WPI above 10% and xanthan gum above 0.5%) two phases, each rich in one biopolymer were formed (Benichou et al., 2007). 8 Why complexing proteins and polysaccharides? Proteins and polysaccharides are commonly complexed together because the functional properties of the complexes are enhanced than when the biopolymers are used on their own (Schmitt et al., 1998). These complexes can have enhanced viscosity, water-holding capacity, and foam and emulsion stability (Schmitt et al., 1998). Protein and polysaccharides also have improved gelling which in turns increases water holding capacity. In a study done by Le and Turgeon (2015), they saw that electrostatically formed gels with -lactoglobulin and xanthan gum had increased water holding capacity. The aggregation of protein along the polysaccharide chain allowed for better water retention of the gel (Le and Turgeon, 2015). Another advantage to complexing proteins and polysaccharides is their improved amphiphilicity which provides better foam and emulsion stability (Schmitt et al., 1998). Proteins can lower interfacial tension through repulsive steric forces or the formation of a gel-like surface, both of which are further improved when polysaccharides are complexed onto the protein. The complexes can also behave as dispersed particles and resist flow in the system increasing its viscosity (Schmitt et al., 1998). There have also been recent studies that show enhanced IRI activity due to synergistic affects between antifreeze proteins (AFP) and complexed molecules, such as with wheat flour and starches (Monalisa et al., 2021, 2023). It has been hypothesized that the hydrophilic nature of starches allows for increased interaction with the water molecules which allows the AFP to better interact with the ice surface and prevent ice crystal growth (Monalisa et al., 2021). With the current literature there is a gap of research that is focused on the freezing quality of high-moisture dairy products, such as sour cream and yogurt. With the significant deterioration of textural quality due to freezing, this leaves an opportunity for research into an antifreeze additive that can be used to prevent this. Proteins and polysaccharides and their 9 hydrolysates have been used individually as IRI active molecules and stabilizing additives in food products, but they have not been used as complexes in dairy products. Improved functional properties such as amphiphilicity and water holding capacity are also thought to be enhanced when proteins and polysaccharides are complexes justifying their uses as potential antifreeze additives for maintaining the textural quality of sour cream and yogurt during freezing. Dairy Processing By-Products and Value Capturing Cheese whey as a dairy by-product With increasing demand and consumption of dairy products, the dairy processing industry generates significant waste that requires further processing to reduce environmental stress (Pires et al., 2021). There are many by-products that come from dairy manufacturing including buttermilk from butter making, whey from cheese making, and -stream from the production of anhydrous milk fat. According to USDA’s 2023 Dairy Products Summary the U.S produced 938 million pounds of dry whey which is equivalent to approximately 6.7 million tons of liquid whey. Cheese whey is one of the largest sources of dairy waste and because of its high biological oxygen demand and chemical oxygen demand, it can threaten aquatic life in the environment if it is not properly treated before disposal (Pires et al., 2021, Sar et al., 2022). Whey, however, is a nutrient rich waste source consisting mainly of lactose, protein, and fat that can be utilized as value added food ingredients such as whey protein powders and whey based beverages (Chavan et al., 2015). Whey can be classified as either sweet whey, which has a pH between 6-7, and acid whey, which has a pH of less than 5, depending on the method of coagulation used during cheese production (Siso, 1996, Argenta and Scheer, 2020). Sweet whey is from the production of cheese that uses rennet, a microbial enzyme consisting of chymosin; acid whey is produced from 10 cheese that is coagulated using lactic acid producing bacteria (Siso, 1996, Argenta and Scheer, 2020). Whey is 93% water and 7% solids where 70-72% of the total solids is made up of lactose, followed by ~13% protein, ~5% lipids, and the remaining being minerals (Ryan and Walsh, 2016). The main proteins in whey are β-lactoglobulin, α-lactalbumin, bovine serum albumin (BSA), and immunoglobulins (Ryan and Walsh, 2016). Common utilizations of whey Traditionally, whey has been used as animal feed, but one of the main uses for whey is for the production of whey powders, whey protein concentrates, and whey protein isolates (Ryan and Walsh, 2016, Çelik, 2020). These are all made from slightly different processing techniques. Whey powder production includes pretreatments to remove remnant curd particles and fat before being pasteurized (Çelik, 2020). The whey is then concentrated to a solids content of 40-60% through evaporation (Çelik, 2020). The whey concentrate is then dried, typically through spray drying, to obtain a powdered final product (Çelik, 2020). Whey protein concentrate is a product that has at least 25% protein and is produced through precipitation, filtration, and dialysis (Çelik, 2020). Ultra-filtration can be used to further concentrate the whey and produce products with up to 80% protein (Çelik, 2020). Whey protein isolate is made similarly to these other products and typically includes ion exchange chromatography to demineralize the whey and ultra-filtration to further concentrate the protein to at least 90% (Çelik, 2020). Dairy phospholipid as valuable components In addition to protein, there are other components of dairy by-products that have high value, such as phospholipids. Phospholipids are amphiphilic molecules which consist of two hydrophobic acyl chains and a hydrophilic head. Phospholipids can be classified as glycerophospholipids or sphingolipids which consists of a glycerol or sphingosine backbone, 11 respectively. Phospholipids serve as a vital part in biological membranes as they form a lipid bilayer and contribute to the cell membranes fluidity. There are different sources of phospholipids including soybeans, eggs, dairy, marine organisms, and bovine brain (Weihrauch and Son, 1983, Burling and Graverholt, 2008, Küllenberg et al., 2012). The recent interest in phospholipids stems from their multiple health benefits. Dairy sources have become a popular interest because of their high amount of sphingomyelin and phosphatidylserine compared to other sources like soy and egg (Burling and Graverholt, 2008). Sphingomyelin has been shown to reduce the uptake of LDL cholesterol in mice and humans which can help prevent cardiovascular disease (Conway et al., 2013, Millar et al., 2020). There have also been studies, mainly in mice, that sphingomyelin helps reduce gut microbiota imbalance (Norris et al., 2016, Millar et al., 2020). Phosphatidylserine has shown to improve ADHD symptoms and short term auditory memory in children (Hirayama et al., 2014). Phosphatidylserine has also shown to improve cognitive performance including memory and mood in elderly patients with Alzheimer’s and dementia (Moré et al., 2014, Ma et al., 2022). The ability to extract these phospholipids would allow for their use as a health promoting additive in foods or nutraceuticals. One of the main phospholipids in human breast milk and colostrum is sphingomyelin (Verardo, Gómez-Caravaca et al. 2017). Recovery of these highly sought after phospholipids would allow infant formulas to have increased nutritional components that is more representative of human milk. However, these phospholipids reside in the milk fat globule membrane (MFGM) which makes their extraction challenging. Milk fat globule membrane and its distribution and recovery by precipitation The MFGM is the component in milk which encloses a triglyceride core with a thin outer tri-layer membrane (Dewettinck, Rombaut et al. 2008). The MFGM consists of mainly proteins 12 and lipids; 60% and 40%, respectively (Vanderghem, Bodson et al. 2010). As mentioned previously, phospholipid content varies depending on the source and dairy products such as milk have a higher content of sphingomyelin and phosphatidylserine. The phospholipid concentration relative to fat in dairy by-products is also higher than the raw milk and cream. Raw milk, cream, and butter have a phospholipid concentration relative to fat of 0.98%, 0.45%, and 0.27%, respectively, whereas buttermilk, butter serum, and whey have a phospholipid concentration of 33.05%, 23.66%, and 48.39%, respectively (Rombaut et al., 2006). This shows the phospholipids tendency to partition into the aqueous phase, thus justifying the recovery of phospholipids from cheese whey. The MFGM can be precipitated from dairy by-products, like whey, through thermocalcic aggregation (Rombaut and Dewettinck 2007). Thermocalcic aggregation is a method that uses a calcium salt and moderate heat to precipitate the MFGM particles (Rombaut and Dewettinck 2007). Divalent salts such as calcium, zinc, and magnesium are thought to crosslink MFGM fragments through the phosphate groups of the phospholipids and then precipitate out through hydrophobic interaction of the membrane proteins (Damodaran 2010). In a study done by Price et al. (2020), this method was optimized for the extraction of phospholipids from -stream using calcium and zinc acetate. Both salts showed to be effective in recovering the MFGM and its phospholipids in the -stream, however zinc acetate showed a higher lipid and phospholipid recovery. This study also showed an enrichment of phospholipid concentration relative to total lipid during ethanol extraction of the phospholipids in the pellets from calcium acetate compared to zinc acetate. The geometric interaction is a possible explanation for the enhanced lipid recovery using ethanol from calcium acetate precipitated pellets (Price, Fei et al. 2020). Zinc tends to form a tetracoordinated conformation, whereas calcium forms a hexacoordinated 13 complex (Katz, Glusker et al. 1996, Binder, Arnold et al. 2001). Since the molecules are not as close together in the hexacoordinated system as they are in the tetracoordinated, the ethanol is able to penetrate and extract the phospholipids better (Price, Fei et al. 2020). However, due to concerns over zinc consumption, calcium acetate is a safer alternative and one of the reasons for its use in this study. This type of MFGM precipiatation has been used to improve the ultrafiltration of whey protein concentrates. If the MFGM is not removed it can cause fouling of the membrane (Rombaut and Dewettinck 2007), along with off-flavors and discoloration caused by lipid oxidation and the Maillard reaction (Morr and Ha 1991). Previous extraction techniques and gaps in phospholipid extraction from cheese whey -Stream and whey are both dairy by-products, but their production and composition are slightly different. -Stream is produced through the process of turning cream into anhydrous milk fat (Rajorhia, 2003), whereas whey is the waste stream produced from the coagulation of casein during cheese making. -Stream has a higher protein content, different protein profile, and less lactose than whey as shown in Table 2 (Price, Fei et al. 2020). This difference in composition leaves the opportunity for phospholipid extraction to be optimized on whey, which is a much more abundant waste stream, and the treatment may be used on site of cheesemaking without other filtration operations. Previous research of extraction and isolation of phospholipids from other whey components and concentrated products has been done using filtration, supercritical fluids, and solvents like ethanol (Konrad et al., 2013, Price et al., 2018, Sprick et al., 2019). However, most research involving phospholipid extraction has been performed on processed whey products such as whey protein phospholipid concentrate and not on fresh unprocessed whey. There also has not been to our knowledge a method that uses calcium acetate salt as the precipitation agent for whey. This method of thermocalcic aggregation using calcium 14 Table 2: Composition comparison of whey and beta stream. * Total solids (%) Protein (%dwb) Lipids (% dwb) Lactose (% dwb) Phospholipids (% of fat) Beta Stream 10.0 21.0 14.4 43.9 ~48 Whey 7.0 13.1 5.4 74.9 ~30 * Whey content is from Milkoscan analysis. Beta stream composition is from Rathnakumar, Ortega-Anaya et al. (2021). acetate is a simpler and more cost-effective method of precipitation which would allow for easier implementation into small dairy producer facilities. The use of the least amount of salt and energy for the highest recovery of phospholipid and greatest separation of proteins need to be determined. Thesis objectives and organization The overall objective of this thesis research is to improve dairy product quality and waste stream utilization. The first chapter focuses on the prevention of freeze induced damage in sour cream and yogurt and why there is a need for an antifreeze agent. In this chapter, protein (whey protein isolate and soy protein isolate) and polysaccharides (locust bean gum and lambda carrageenan), unhydrolyzed and hydrolyzed, were complexed together and their effect at preventing freeze induced damage in sour cream and yogurt were evaluated using texture analysis and microscopic image analysis. These complexes were also evaluated for IRI activity in a model system to further justify their use in the food system. A soy protein isolate hydrolysate (95%) and soy lecithin (5%) complex was also evaluated due to its enhanced IRI activity shown in previous work. It was hypothesized that the protein and polysaccharide complexes will prevent freeze induced damage in sour cream and yogurt due to increase amphiphilicity and water holding capacity, which would allow for better interaction with the ice 15 interface and control of local unfrozen water along with preventing the dehydration of proteins. It was also hypothesized that hydrolysate complexes of these biopolymers will have improved antifreeze activity due to their reduced molecular weight, potentially allowing for more mobility at the ice interface and within the food matrix. The second chapter focuses on the optimization of extracting phospholipids from industrial cheese whey. The principle of thermocalcic aggregation was used to precipitate MFGM components for further extraction of phospholipids. This process was optimized by manipulating pH, temperature, and calcium acetate concentration. The optimum conditions were chosen based on total lipid and phospholipid yield and protein distribution. After optimum conditions are found, a mass balance of lipid, protein, and salt was made. It was hypothesized that there would be increased MFGM and phospholipid recovery at increased calcium acetate concentration, pH, and temperature due to increased chemical interactions and the mechanism of precipitation. Enough salt needs to be available to precipitate all the MFGM fragments, at a higher pH the phosphate groups of the phospholipids will be more charged allowing for better interaction with the salt, and at higher temperatures hydrophobic interactions are favored. 16 References Adapa, S., K. A. Schmidt, I. J. Jeon, T. J. Herald and R. A. Flores (2000). "Mechanisms of Ice Crystallization and Recrystallization in Ice Cream: A Review." Food Reviews International 16(3): 259-271. Alinovi, M., M. Corredig, G. Mucchetti and E. Carini (2020). "Water status and dynamics of high-moisture Mozzarella cheese as affected by frozen and refrigerated storage." Food Research International 137: 109415. Arab, M., M. Yousefi, E. Khanniri, M. Azari, V. Ghasemzadeh-Mohammadi and N. Mollakhalili-Meybodi (2023). "A comprehensive review on yogurt syneresis: effect of processing conditions and added additives." J Food Sci Technol 60(6): 1656-1665. Argenta, A. B. and A. D. P. Scheer (2020). "Membrane Separation Processes Applied to Whey: A Review." Food Reviews International 36(5): 499-528. Code of Federal Regulations. 2023. Title 21, part 131 section 131.200 Yogurt. Food and Drug Administation. Accessed on Nov. 20, 2024. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=131.200&SearchTe rm=yogurt Benichou, A., A. Aserin, R. Lutz and N. Garti (2007). "Formation and characterization of amphiphilic conjugates of whey protein isolate (WPI)/xanthan to improve surface activity." Food Hydrocolloids 21(3): 379-391. Biggs, C. I., T. L. Bailey, G. Ben, C. Stubbs, A. Fayter and M. I. Gibson (2017). "Polymer mimics of biomacromolecular antifreezes." Nature Communications 8(1): 1546. Binder, H., K. Arnold, A. S. Ulrich and O. Zschörnig (2001). "Interaction of Zn2+ with phospholipid membranes." Biophysical Chemistry 90(1): 57-74. 17 Bourrie, B. C., B. P. Willing and P. D. Cotter (2016). "The microbiota and health promoting characteristics of the fermented beverage kefir." Frontiers in microbiology 7: 196946. Budke, C. and T. Koop (2006). "Ice Recrystallization Inhibition and Molecular Recognition of Ice Faces by Poly(vinyl alcohol)." ChemPhysChem 7(12): 2601-2606. Burling, H. and G. Graverholt (2008). "Milk – A new source for bioactive phospholipids for use in food formulations." Lipid Technology 20(10): 229-231. Çelik, K. (2020). Whey Every Aspect, Tudás Alapítvány. Chavan, R., R. Shraddha, A. Kumar and T. Nalawade (2015). "Whey based beverage: its functionality, formulations, health benefits and applications." Conway, V., P. Couture, C. Richard, S. F. Gauthier, Y. Pouliot and B. Lamarche (2013). "Impact of buttermilk consumption on plasma lipids and surrogate markers of cholesterol homeostasis in men and women." Nutrition, Metabolism and Cardiovascular Diseases 23(12): 1255-1262. Corredig, M., N. Sharafbafi and E. Kristo (2011). "Polysaccharide–protein interactions in dairy matrices, control and design of structures." Food Hydrocolloids 25(8): 1833-1841. Costello, M. J. (2009). Sour Cream and Related Products. The Sensory Evaluation of Dairy Products. S. Clark, M. Costello, M. Drake and F. Bodyfelt. New York, NY, Springer US: 403- 426. Damodaran, S. (2010). "Zinc-Induced Precipitation of Milk Fat Globule Membranes: A Simple Method for the Preparation of Fat-Free Whey Protein Isolate." Journal of Agricultural and Food Chemistry 58(20): 11052-11057. Dewettinck, K., R. Rombaut, N. Thienpont, T. T. Le, K. Messens and J. Van Camp (2008). "Nutritional and technological aspects of milk fat globule membrane material." International Dairy Journal 18(5): 436-457. 18 Diefes, H. A., S. S. H. Rizvi and B. J. A. (1993). "Rheological Behavior of Frozen and Thawed Low-Moisture, Part-Skim Mozzarella Cheese." Journal of Food Science 58(4): 764-769. Fomich, M., V. P. Día, U. I. Premadasa, B. Doughty, H. B. Krishnan and T. Wang (2023). "Ice Recrystallization Inhibition Activity of Soy Protein Hydrolysates." Journal of Agricultural and Food Chemistry 71(30): 11587-11598. García-Burgos, M., J. Moreno-Fernández, M. J. M. Alférez, J. Díaz-Castro and I. López-Aliaga (2020). "New perspectives in fermented dairy products and their health relevance." Journal of Functional Foods 72: 104059. Graham, B., A. E. R. Fayter, J. E. Houston, R. C. Evans and M. I. Gibson (2018). "Facially Amphipathic Glycopolymers Inhibit Ice Recrystallization." Journal of the American Chemical Society 140(17): 5682-5685. Graiver, N. G., N. E. Zaritzky and A. N. Califano (2004). "Viscoelastic Behavior of Refrigerated Frozen Low-moisture Mozzarella Cheese." Journal of Food Science 69(3): FEP123-FEP128. Hartel, R. W. (1998). Mechanisms and kinetics of recrystallization in ice cream. The Properties of Water in Foods ISOPOW 6. D. S. Reid. Boston, MA, Springer US: 287-319. Hirayama, S., K. Terasawa, R. Rabeler, T. Hirayama, T. Inoue, Y. Tatsumi, M. Purpura and R. Jäger (2014). "The effect of phosphatidylserine administration on memory and symptoms of attention-deficit hyperactivity disorder: a randomised, double-blind, placebo-controlled clinical trial." Journal of Human Nutrition and Dietetics 27(s2): 284-291. Igartúa, D. E., D. M. Cabezas and G. G. Palazolo (2022). "Effects of pH, protein:polysaccharide ratio, and NaCl-added concentration on whey protein isolate and soluble soybean polysaccharides electrostatic-complexes formation." Food Chemistry Advances 1: 100123. 19 Kamińska-Dwórznicka, A., M. Matusiak, K. Samborska, D. Witrowa-Rajchert, E. Gondek, E. Jakubczyk and A. Antczak (2015). "The influence of kappa carrageenan and its hydrolysates on the recrystallization process in sorbet." Journal of Food Engineering 167: 162-165. Kasprzak, K., W. L. Wendorff and C. M. Chen (1994). "Freezing Qualities of Cheddar-Type Cheeses Containing Varied Percentages of Fat, Moisture, and Salt." Journal of Dairy Science 77(7): 1771-1782. Katz, A. K., J. P. Glusker, S. A. Beebe and C. W. Bock (1996). "Calcium Ion Coordination:  A Comparison with That of Beryllium, Magnesium, and Zinc." Journal of the American Chemical Society 118(24): 5752-5763. Knight, C. A., A. L. De Vries and L. D. Oolman (1984). "Fish antifreeze protein and the freezing and recrystallization of ice." Nature 308(5956): 295-296. Konrad, G., T. Kleinschmidt and C. Lorenz (2013). "Ultrafiltration of whey buttermilk to obtain a phospholipid concentrate." International Dairy Journal 30(1): 39-44. Kulkarni, V. S. and C. Shaw (2015). Essential Chemistry for Formulators of Semisolid and Liquid Dosages. Saint Louis, UNITED STATES, Elsevier Science & Technology. Küllenberg, D., L. A. Taylor, M. Schneider and U. Massing (2012). "Health effects of dietary phospholipids." Lipids in Health and Disease 11(1): 3. Kuo, M.-I. and S. Gunasekaran (2009). "Effect of freezing and frozen storage on microstructure of Mozzarella and pizza cheeses." LWT - Food Science and Technology 42(1): 9-16. Le, X. T. and S. L. Turgeon (2015). "Textural and waterbinding behaviors of β-lactoglobulin- xanthan gum electrostatic hydrogels in relation to their microstructure." Food Hydrocolloids 49: 216-223. 20 Li, D., Z. Zhu and D.-W. Sun (2018). "Effects of freezing on cell structure of fresh cellular food materials: A review." Trends in Food Science & Technology 75: 46-55. Luo, W., C. Yuan, J. Wu, Y. Liu, F. Wang, X. Li and S. Wang (2023). "Inhibition mechanism of membrane-separated silver carp hydrolysates on ice crystal growth obtained through experiments and molecular dynamics simulation." Food Chemistry 414: 135695. Ma, X., X. Li, W. Wang, M. Zhang, B. Yang and Z. Miao (2022). "Phosphatidylserine, inflammation, and central nervous system diseases." Frontiers in aging neuroscience 14: 975176. McHugh, T. (2015). "How Yogurt is Processed." from https://www.ift.org/news-and- publications/food-technology-magazine/issues/2015/december/columns/processing. Midya, U. S. and S. Bandyopadhyay (2024). "Ice Recrystallization Unveils the Binding Mechanism Operating at a Diffused Interface." The Journal of Physical Chemistry B 128(5): 1170-1178. Millar, C. L., C. Jiang, G. H. Norris, C. Garcia, S. Seibel, L. Anto, J.-Y. Lee and C. N. Blesso (2020). "Cow's milk polar lipids reduce atherogenic lipoprotein cholesterol, modulate gut microbiota and attenuate atherosclerosis development in LDL-receptor knockout mice fed a Western-type diet." The Journal of Nutritional Biochemistry 79: 108351. Miller-Livney, T. and R. W. Hartel (1997). "Ice Recrystallization in Ice Cream: Interactions Between Sweeteners and Stabilizers." Journal of Dairy Science 80(3): 447-456. Mochizuki, K. and V. Molinero (2018). "Antifreeze Glycoproteins Bind Reversibly to Ice via Hydrophobic Groups." Journal of the American Chemical Society 140(14): 4803-4811. Monalisa, K., M. Shibata and T. Hagiwara (2021). "Ice Recrystallization Behavior of Corn Starch/Sucrose Solutions: Effects of Addition of Corn Starch and Antifreeze Protein III." Food Biophysics 16(2): 229-236. https://www.ift.org/news-and-publications/food-technology-magazine/issues/2015/december/columns/processing https://www.ift.org/news-and-publications/food-technology-magazine/issues/2015/december/columns/processing 21 Monalisa, K., M. Shibata and T. Hagiwara (2023). "Ice recrystallization inhibition behavior by wheat flour and its synergy effect with antifreeze protein III." Food Hydrocolloids 143: 108882. Moré, M. I., U. Freitas and D. Rutenberg (2014). "Positive Effects of Soy Lecithin-Derived Phosphatidylserine plus Phosphatidic Acid on Memory, Cognition, Daily Functioning, and Mood in Elderly Patients with Alzheimer’s Disease and Dementia." Advances in Therapy 31(12): 1247-1262. Morr, C. V. and E. Y. W. Ha (1991). "Off-flavors of whey protein concentrates: A literature review." International Dairy Journal 1(1): 1-11. Niamsiri, N. and C. A. Batt (2009). Dairy Products. Encyclopedia of Microbiology (Third Edition). M. Schaechter. Oxford, Academic Press: 34-44. Niu, F., Y. Su, Y. Liu, G. Wang, Y. Zhang and Y. Yang (2014). "Ovalbumin–gum arabic interactions: Effect of pH, temperature, salt, biopolymers ratio and total concentration." Colloids and Surfaces B: Biointerfaces 113: 477-482. Norris, G. H., C. Jiang, J. Ryan, C. M. Porter and C. N. Blesso (2016). "Milk sphingomyelin improves lipid metabolism and alters gut microbiota in high fat diet-fed mice." The Journal of Nutritional Biochemistry 30: 93-101. Panghal, A., R. Patidar, S. Jaglan, N. Chhikara, S. K. Khatkar, Y. Gat and N. Sindhu (2018). "Whey valorization: current options and future scenario – a critical review." Nutrition & Food Science 48(3): 520-535. Pires, A. F., N. G. Marnotes, O. D. Rubio, A. C. Garcia and C. D. Pereira (2021). "Dairy By- Products: A Review on the Valorization of Whey and Second Cheese Whey." Foods 10(5): 1067. 22 Price, N., T. Fei, S. Clark and T. Wang (2020). "Application of zinc and calcium acetate to precipitate milk fat globule membrane components from a dairy by-product." Journal of Dairy Science 103(2): 1303-1314. Rombaut, R., J. V. Camp and K. Dewettinck (2006). "Phospho- and sphingolipid distribution during processing of milk, butter and whey." International Journal of Food Science & Technology 41(4): 435-443. Rombaut, R. and K. Dewettinck (2007). "Thermocalcic aggregation of milk fat globule membrane fragments from acid buttermilk cheese whey." J Dairy Sci 90(6): 2665-2674. Ryan, M. P. and G. Walsh (2016). "The biotechnological potential of whey." Reviews in Environmental Science and Bio/Technology 15(3): 479-498. Sar, T., S. Harirchi, M. Ramezani, G. Bulkan, M. Y. Akbas, A. Pandey and M. J. Taherzadeh (2022). "Potential utilization of dairy industries by-products and wastes through microbial processes: A critical review." Science of The Total Environment 810: 152253. Schingoethe, D. J. (1976). "Whey Utilization in Animal Feeding: A Summary and Evaluation1, 2." Journal of Dairy Science 59(3): 556-570. Schmitt, C., C. Sanchez, S. Desobry-Banon and J. Hardy (1998). "Structure and Technofunctional Properties of Protein-Polysaccharide Complexes: A Review." Critical Reviews in Food Science and Nutrition 38(8): 689-753. Siso, M. I. G. (1996). "The biotechnological utilization of cheese whey: A review." Bioresource Technology 57(1): 1-11. Soukoulis, C. and I. Fisk (2016). "Innovative Ingredients and Emerging Technologies for Controlling Ice Recrystallization, Texture, and Structure Stability in Frozen Dairy Desserts: A Review." Critical Reviews in Food Science and Nutrition 56(15): 2543-2559. 23 Soukoulis, C., E. Rontogianni and C. Tzia (2010). "Contribution of thermal, rheological and physical measurements to the determination of sensorially perceived quality of ice cream containing bulk sweeteners." Journal of Food Engineering 100(4): 634-641. Sprick, B., Z. Linghu, J. K. Amamcharla, L. E. Metzger and J. S. Smith (2019). "Selective extraction of phospholipids from whey protein phospholipid concentrate using supercritical carbon dioxide and ethanol as a co-solvent." Journal of Dairy Science 102(12): 10855-10866. Sun, X., R. Guo, Y. Kou, H. Song, T. Zhan, J. Wu, L. Song, H. Zhang, F. Xie, J. Wang, Z. Song and Y. Wu (2023). "Inhibition of ice recrystallization by tamarind (Tamarindus indica L.) seed polysaccharide and molecular weight effects." Carbohydrate Polymers 301: 120358. Sun, X., R. Guo, T. Zhan, Y. Kou, X. Ma, H. Song, W. Zhou, L. Song, H. Zhang, F. Xie, C. Yuan, Z. Song and Y. Wu (2024). "Retarding ice recrystallization by tamarind seed polysaccharide: Investigation in ice cream mixes and insights from molecular dynamics simulation." Food Hydrocolloids 149: 109579. Sun, X., Y. Wu, Z. Song and X. Chen (2022). "A review of natural polysaccharides for food cryoprotection: Ice crystals inhibition and cryo-stabilization." Bioactive Carbohydrates and Dietary Fibre 27: 100291. Thiebaud, M., E. Dumay, L. Picart, J. P. Guiraud and J. C. Cheftel (2003). "High-pressure homogenisation of raw bovine milk. Effects on fat globule size distribution and microbial inactivation." International Dairy Journal 13(6): 427-439. To, C. M., L. Vermeir, F. Rebry, B. Kerkaert, P. Van der Meeren and T. P. Guinee (2020). "Impact of freezing on the physicochemical and functional properties of low–moisture part–skim mozzarella." International Dairy Journal 106: 104704. 24 Tribst, A. A. L., L. T. P. Falcade, N. S. Carvalho, M. Cristianini, B. R. d. C. Leite Júnior and M. M. d. Oliveira (2020). "Using physical processes to improve physicochemical and structural characteristics of fresh and frozen/thawed sheep milk." Innovative Food Science & Emerging Technologies 59: 102247. USDA (2000). USDA Specifications for SOur cream and Acidified Sour Cream. U. S. D. o. Agriculture. USDA (2001). USDA Specifications for Yogurt, Nonfat Yogurt and Lowfat Yogurt. U. S. D. o. Agriculture. USDA. (2019). "Yogurt, plain, whole milk." from https://fdc.nal.usda.gov/fdc-app.html#/food- details/171284/nutrients. USDA. (2022). "Cream, sour, full fat." from https://fdc.nal.usda.gov/fdc-app.html#/food- details/2346387/nutrients. Vanderghem, C., P. Bodson, S. Danthine, M. Paquot, C. Deroanne and C. Blecker (2010). "Milk fat globule membrane and buttermilks: from composition to valorization." Base. Verardo, V., A. M. Gómez-Caravaca, D. Arráez-Román and K. Hettinga (2017). "Recent Advances in Phospholipids from Colostrum, Milk and Dairy By-Products." International Journal of Molecular Sciences 18(1): 173. Webb, B. and W. Arbuckle (1977). "Fundamentals of Food Freezing." Fundamentals of Food Freezing., Westport, CT. Weihrauch, J. L. and Y.-S. Son (1983). "Phospholipid content of foods." Journal of the American Oil Chemists’ Society 60(12): 1971-1978. Weng, L., S. L. Stott and M. Toner (2018). "Molecular Dynamics at the Interface between Ice and Poly(vinyl alcohol) and Ice Recrystallization Inhibition." Langmuir 34(17): 5116-5123. https://fdc.nal.usda.gov/fdc-app.html#/food-details/171284/nutrients https://fdc.nal.usda.gov/fdc-app.html#/food-details/171284/nutrients https://fdc.nal.usda.gov/fdc-app.html#/food-details/2346387/nutrients https://fdc.nal.usda.gov/fdc-app.html#/food-details/2346387/nutrients 25 Yousefi, M. and S. M. Jafari (2019). "Recent advances in application of different hydrocolloids in dairy products to improve their techno-functional properties." Trends in Food Science & Technology 88: 468-483. Yuan, Y., V. P. Dia and T. Wang (2024). "The ice recrystallization inhibition activity of wheat glutenin hydrolysates and effect of salt on their activity." Food Hydrocolloids 154: 110153. 26 CHAPTER I THE EFFECT OF PROTEIN AND POLYSSACHARIDE COMPLEXES ON FREEZE INDUCED DAMAGE IN SOUR CREAM AND YOGURT 27 A version of this chapter will be submitted for publication by Ripley Vaughan with co- authors Liz Eckelkamp, Vermont P. Dia, and Tong Wang in the Journal of Dairy Science. Abstract Freezing high-moisture dairy products can cause deterioration of textural quality including decreased firmness and increased graininess due to ice recrystallization and growth and protein dehydration and aggregation. Protein and polysaccharide complexes were made and their effects on preventing freeze induced damage in sour cream and yogurt were investigated. Whey protein isolate (WPI) and soy protein isolate (SPI) were complexed with locust bean gum (LBG) and lambda carrageenan (LC) through a pH cycling method. Hydrolysates of these biopolymers were also complexed to compare their activity against the complex of the unhydrolyzed biopolymers. The ice recrystallization inhibition (IRI) activity of the complexes was evaluated by splat assay at pH 4.5 and 7.0 in 20 mM NaCl and showed that all of the complexes had increased IRI activity compared to the negative control at both pHs (P < 0.05). At pH 7.0, the unhydrolyzed WPI and LBG complex was able to reduce the ice crystal size relative to PEG by 64%, compared to SPI and lecithin complex of 55% reduction. At pH 4.5, the unhydrolyzed SPI and LBG and WPI and LBG complexes performed the best by reducing the ice crystal size, 38% and 35% relative to PEG, respectively. Selected protein and polysaccharide complexes were added to sour cream and yogurt at 3% as-is weight basis before being frozen for 4 days. The results showed that statistically neither of the unhydrolyzed or hydrolyzed complexes were effective at preventing freeze induced damage in sour cream and yogurt. There was a decrease of 32-55% in firmness, 67-85% in cohesiveness, and 32-59% in consistency in the sour cream and yogurt treated with the unhydrolyzed complexes. The hydrolyzed complexes showed a decrease of 32-77% in firmness, 67-92% in cohesiveness, and 32-75% in consistency. There was also an 28 increase in lacunarity in the samples indicating increased graininess most likely due to protein aggregation. However, our treatments did reduce graininess compared to the control. Overall, unhydrolyzed and hydrolyzed protein-polysaccharide complexes showed IRI activity in the model system but were ineffective at preventing freeze induced damage in the dairy system. Key words: Freeze-induced damage; protein-polysaccharide complexes; sour cream; yogurt; ice recrystallization inhibition INTRODUCTION Freeze induced structural damage is a common problem in frozen foods and results in decreased quality and increased food waste. This damage is caused by ice recrystallization and growth, volume expansion, cryo-concentration, and protein dehydration (Diefes et al., 1993, Alinovi et al., 2021, Sun et al., 2022). In high-moisture dairy products, the protein matrix helps structure the food system and contributes to its texture but can physically break due to ice expansion during ice recrystallization (Diefes et al., 1993, Kuo and Gunasekaran, 2009). Ice recrystallization can also cause rupture of the milk fat globule causing fat leakage from the globule (Alinovi et al., 2020). Most research investigating freeze induced damage in dairy products has been conducted using ice cream (Damodaran and Wang, 2017, Reeder et al., 2023) and high and low-moisture mozzarella cheese (Kuo and Gunasekaran, 2009, Alinovi and Mucchetti, 2020, Alinovi et al., 2021). Proteins can become dehydrated during freezing which can reduce the protein’s ability to rebind water (Kuo and Gunasekaran, 2009). As bound water is released from the protein matrix during freezing there is an increase in unbound water resulting in syneresis (Alinovi et al., 2020). Although yogurt and sour cream are not commonly frozen, identifying an additive to help prevent freeze induced damage would be beneficial in the event of 29 supply chain disruption or the need to prolong their shelf life. In addition, these products may serve as a sensitive model to identify antifreeze agents. Many molecules have been investigated for ice recrystallization inhibition (IRI) properties. There is a category of antifreeze proteins (AFP) naturally present in polar fish and plants that are responsible for their survival at low temperatures (Atici and Nalbantoglu, 2003). These proteins have potent IRI activity, but are in low abundance and expensive to obtain (Venketesh and Dayananda, 2008). This led to the need to identify molecules that are more readily available with moderate IRI activity. Synthetic polymers such as polyvinyl alcohol (Budke and Koop, 2006), protein and polysaccharide hydrolysates (Fomich et al., 2023, Sun et al., 2023), and stabilizers have been reported to have IRI activity (Kamińska-Dwórznicka et al., 2015). Proteins and polysaccharides, such as whey protein isolate (WPI), locust bean gum (LBG), carrageenans, and guar gum, are commonly added to dairy products individually to help improve textural properties such as thickening and prevention of syneresis (Corredig et al., 2011, Yousefi and Jafari, 2019). In yogurt specifically, WPI, whole milk powder, and casein powders are commonly added to increase the total solids content and help create a thicker product (McHugh, 2015). Proteins and polysaccharides have been shown to form functional complexes through electrostatic interaction (Schmitt et al., 1998). Functional properties of these biomolecules applied for texture enhancement and stability can be improved when complexed together. Not only do these complexes have potential to retain textural properties during freezing in the dairy systems, they also have potential to be IRI active molecules. A study done by Monalisa et al. (2021) showed an increase in IRI activity when AFP and corn starch were 30 combined and they hypothesized this was due to the starch’s increased interaction with the water molecules allowing the AFP to better interact with the ice surface, preventing ice crystal growth. The increased amphiphilicity of the biopolymers when complexed together could allow for the polysaccharide to interact more with the water and the protein to better interact with the ice crystal surface therefore preventing ice recrystallization (Gaukel et al., 2014). The improved hydration properties from the polysaccharides could also help prevent protein dehydration. There has also been a study that showed a synergistic effect between wheat flour and AFP on ice recrystallization (Monalisa et al., 2023). Wheat flour is a complex substance containing starch, protein, polysaccharides, and lipids. They hypothesized that the complex makeup of wheat flour along with AFP prevent ice recrystallization by synergistically interacting with the quasi-liquid layer (Monalisa et al., 2023). Thus, we hypothesize for this study that the protein and polysaccharide complexes will help protect against freeze induced damage due to their improved amphiphilicity, thus IRI activity, and increased water holding capacity. Hydrolysates of these biomolecules with smaller molecular weights have also shown improved IRI activity compared to larger biomolecules (Fomich et al., 2023, Sun et al., 2023, Ollis et al., 2024, Yuan et al., 2024). Therefore, we also hypothesize that the complexes of hydrolyzed proteins and polysaccharides will have a greater effect than the complexes of unhydrolyzed proteins and polysaccharides due to an increased mobility at the ice interface and within a complex food matrix, providing better protection against freeze induced damage. The objective of this study was to investigate the effect of two proteins (WPI and soy protein isolate (SPI)) and two polysaccharides (LBG and lambda carrageenan (LC)) and their hydrolysate complexes on freeze induced damage in sour cream and yogurt. 31 To test these hypotheses, WPI and SPI were complexed with LBG and LC through pH cycling along with their hydrolysates. A SPI and lecithin complex was also evaluated due to its enhanced IRI activity as reported by Fomich et al. (2024). The complexes were first evaluated for IRI activity through splat assay in a model system before being incorporated into sour cream and yogurt. After freezing and thawing, the damage was evaluated through texture analysis and microscopic image analysis. Lastly, further characterization of the complexes was done with particle size analysis, IRI activity, and qualitative emulsion stability testing. MATERIALS AND METHODS Materials for complex preparation and sour cream and yogurt production The proteins used for this project are WPI (90% protein) from Bulk Supplements (Nevada, USA) and SPI ProFam 931 from ADM (Illinois, USA). The polysaccharides are LBG and LC from Modernist’s Pantry (Maine, USA). The enzyme used for protein hydrolysis was Alcalase from Bacillus licheniformis (3.03 Au/mL) from EMD Millipore Corp. (Massachusetts, USA). The cellulase for hydrolysis of the polysaccharides was from MP Biomedicals (Ohio, USA). Fast green dye used to dye protein was from Allied Chemical (New York, USA), Nile Red (99% pure) used to stain fat was from Acros Organics, and rhodamine B used to stain polysaccharides was from Sigma Aldrich (Missouri, USA). Whole milk and organic half-and- half (no added stabilizers) were used to make yogurt and sour cream, respectively. Whole milk powder was purchased from Hoosier Hill Farm (Wisconsin, USA) for use in making the yogurt. Choozit Buttermilk/ Sour Cream culture containing Lactococcus lactis subsp. Lactis, Lactococcus lactis subsp. cremoris, Lactococcus lactis subsp. lactis biovar. diacetylactis, and Leuconostoc mesenteroides subsp. cremoris from Danisco (Wisconsin, USA) was used to make 32 the sour cream. YO-CULT culture containing Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus from Biena (Quebec, Canada) was used to make the yogurt. Complex preparation WPI and SPI hydrolysates created by Alcalase hydrolysis A protein dispersion was made at 5% (w/v) and hydrated overnight with constant stirring. The pH was then adjusted to 8.0 with 1 M NaOH and equilibrated at 55 ºC. Alcalase was added at 0.044 Au/g protein and the WPI and SPI were hydrolyzed for 10 minutes. For the SPI and lecithin complex, the SPI was only hydrolyzed for 2 minutes. The hydrolysate samples were then placed in a boiling water bath for 10 minutes to denature the enzyme and any remaining unhydrolyzed protein. The samples were then cooled to room temperature and centrifuged at 10,000 x g for 10 minutes and the supernatant was collected and lyophilized. High-performance liquid chromatography by size-exclusion principle (HPLC-SEC) was performed to determine the average molecular weight of the WPI and SPI hydrolysates, in duplicates. Samples of native WPI and SPI along with their hydrolysates were prepared at 1 mg/mL concentration in HPLC-grade water and subsequently filtered through 0.45 m nylon membrane filters (GE Healthcare Life Sciences). Filtered samples were then analyzed with 1200 Agilent HPLC (Agilent Technologies, Santa Clara, CA). The HPLC system consisted of an autosampler (G1329A), quaternary pump (G1311A), vacuum degasser (G1322A), a temperature- controlled column oven (G1316A), and a diode array detector (G1315D). The column used for separation was a BioSep-SEC-S2000 column (300x7.80 mm, Phenomenex, Torrance, CA). A flow rate of 1 mL/min was used with a mobile phase consisting of 45% aqueous acetonitrile with 0.1% trifluoroacetic acid. A standard curve was created from the known molecular weight protein standards consisting of albumin, aprotinin, glucagon, bradykinin, glutathione, and 33 glycine. The linear regression equation from the standard curve (log molecular weight vs. time) was used to calculate the molecular weight of the native WPI and SPI and their hydrolysates. Average molecular weight and relative quantity were calculated by finding the percent area of each peak and its corresponding molecular weight compared with the standard curve. Polysaccharide hydrolysates created by cellulase hydrolysis Polysaccharide (LBG and LC) dispersions were made at 1% (w/v) and hydrated overnight with constant stirring. The pH was adjusted to 6.0 with 1 M HCl and equilibrated at 50 ºC. Cellulase was added at 10% weight of the polysaccharides and hydrolyzed for 48 hours with constant shaking. The samples were then placed in a boiling water bath for 10 minutes to denature the enzyme. The hydrolysates were lyophilized individually for use as a control or complexed with protein before lyophilization. The Somogyi-Nelson method with slight modification was used to determine the reducing sugar content in the LBG and LC hydrolysates (Nelson, 1944). Reagent A was prepared by dissolving 12.5 g anhydrous sodium carbonate, 12.5 g sodium potassium tartrate, 10 g sodium bicarbonate, and 100 g anhydrous sodium sulfate in 300 mL of DI water and then brought to a final volume of 500 mL in a volumetric flask. Reagent B was prepared by dissolving 7.5 g cupric sulfate pentahydrate in 40 mL of DI water and brought to a final volume of 50 mL in a volumetric flask. A drop of concentrated sulfuric acid was then added and mixed thoroughly. Reagent C was prepared by dissolving 25 g ammonium molybdate in 450 mL of DI water and mixing with 21 mL of concentrated sulfuric acid. Next, 3 g of disodium hydrogen arsenate heptahydrate was dissolved in 25 mL of water volumetrically and added to the ammonium molybdate solution under stirring. The final volume was brought to 500 mL volumetrically. Reagent D was prepared by combining 25 mL Reagent A with 1 mL Reagent B. A glucose 34 standard in DI water was then prepared with a concentration ranging from 0-100 g/mL. Test tubes were assembled with 1.0 mL of sample and 1.0 mL of Reagent D where they were then placed in a boiling water bath for 20 minutes. The tubes were then cooled in a cool tap water bath for 5 minutes. After cooling, 1.0 mL of reagent C was added to tubes which were then shaken by hand until no more bubbles formed. After letting sit for 20 minutes, samples were diluted to 25 mL volumetrically with DI water and absorbance was measured immediately at 520 nm. A standard curve was then created in Excel using the glucose standard. From the resulting linear regression equation, the content of reducing sugars was calculated for each sample. The degree of hydrolysis was calculated using the following equation. 𝐷𝐻(%) = ( 𝐹𝑖𝑛𝑎𝑙 𝑟𝑒𝑑𝑢𝑐𝑖𝑛𝑔 𝑠𝑢𝑔𝑎𝑟 ( 𝑚𝑔 𝑚𝐿) − 𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝑟𝑒𝑑𝑢𝑐𝑖𝑛𝑔 𝑠𝑢𝑔𝑎𝑟 ( 𝑚𝑔 𝑚𝐿) 𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑝𝑜𝑙𝑦𝑠𝑎𝑐𝑐ℎ𝑎𝑟𝑖𝑑𝑒 ( 𝑚𝑔 𝑚𝐿) ) × 100 Protein and polysaccharide complexation through pH cycling The pH cycling method reported by Li and Zhong (2020) was used to complex the proteins and polysaccharides with slight modifications. Protein dispersions were made at 10% (w/v) and polysaccharide dispersions were made at 1% (w/v) and hydrated overnight. The protein dispersion’s pH was adjusted to 11.3 with 1 M NaOH and then mixed with the polysaccharide dispersion at a 5:1 ratio of dry matter. The pH was adjusted again to 11.3, if necessary. This mixture was then stirred for 1 hour, then the pH was adjusted to 7.0 with 1 M citric acid, and then heated at 50ºC in a water bath with constant stirring for 30 mins. The pH was then adjusted to a final pH of 4.5 with 1 M citric acid and stirred for another hour. These samples were then lyophilized. The hydrolysate complexes were made using the same method as the unhydrolyzed complexes. There was no removal of the un-complexed protein and polysaccharides. These complexes were then lyophilized. 35 Formation of SPI hydrolysate and lecithin complex SPI hydrolysates that were hydrolyzed for 2 minutes by Alcalase were used to prepare a complex with soy lecithin as determined in Fomich et al (2024). In brief, 5% (w/v) dispersions of SPI hydrolysates and lecithin were made and hydrated overnight. Then, calculated volumes were mixed together in a 50 mL tube to create a complex of 95% peptide and 5% lecithin. To induce complexation the tubes were placed in a boiling water bath and vortexed every 2 minutes for 10 minutes. The complexes were then lyophilized before incorporation into the dairy system Complex characterization Determination of complexe’s IRI activity using Splat Assay A standard splat assay procedure was used to quantify ice recrystallization activity of the unhydrolyzed and hydrolyzed protein-polysaccharide complexes (Knight et al., 1988). In brief, solutions were made of 2% (w/v) of the complexes in 20 mM NaCl solution at a pH of 7.0 and 4.5. One drop of the sample solution was dropped from 1.5 meter onto a pre-cooled -80 ºC slide and annealed at -8 ºC using a cryo-stage HCS 302 (Instec Instruments, Boulder, CO) for 30 min. Three pictures of each drop were taken using polarized light microscopy (Leica, DM2700 M, Wetzlar, Germany) with a built-in digital camera (Leica, DMC 4500, Wetzlar, Germany). This served as one replicate and the measurement was done in duplicate. PEG of the same concentration and pH was used as a negative control. The Feret diameter was determined by using a combination of Cellpose and Fiji (Saad et al., 2023). The average Feret diameter was calculated from 2 replicates with each replicate having three images and measurements. Analysis of particle size of complexes using the Zetasizer To validate complexation of the proteins and polysaccharide, particle size analysis was performed. Select complexes were chosen based on their IRI activity and their performance in 36 the dairy system in the preliminary trials. Particle size distribution measurements were performed in triplicates using Malvern Zetasizer, model 3000, (Malvern Instruments, Worcestershire, England). Samples were prepared at 1 mg/mL in 20 mM NaCl at both pH 7.0 and 4.5 and placed in a cuvette. Particle size distribution was reported as a percentage of the total volume of particles in a particular size range. Confocal microscopic observation for complexation validation Fast green and rhodamine B stain solutions were prepared at 1 mg/mL in DI water and Nile red was prepared at the same concentration in methanol to stain the protein, polysaccharides, and lipids, respectively. Complexes were made at 5% (w/v) in DI water with a total volume of 1 mL. Then, 20 L of each dye was added to the sample dispersion, vortexed, and incubated overnight at 4 ºC. A drop of the sample was placed on a microscope slide and a cover slip was placed on top. The samples were visualized under an inverted confocal laser scanning microscope (CLSM) (Leica Microsystems, Baden-Wurttemberg, Germany) at 63x magnification with oil immersion. The Nile red dye was excited at 488 nm and the emission was collected between 520-590 nm. The Fast green was excited at 633 nm and the emission was collected between 660-750 nm. The rhodamine B was excited at 540 nm and the emission was collected between 553-623 nm. Amphiphilicity of complexes qualitatively evaluated by emulsion stability testing Oil in water emulsions were created with a composition of 94% DI water, 5% soybean oil, and 1% (w/v) of the complexes or controls. Nile red dye was added to the soybean oil for better visualization during testing. Tween 80 (1%) was used as a positive control and the proteins and polysaccharides were also evaluated individually as controls. The emulsions were homogenized using a Fisherbrand™ 850 homogenizer (Fisher Scientific (Ontario, Canada) at 37 10,000 rpm for 3 minutes. Images of the emulsions were taken at 0, 10, and 30 minutes and were evaluated qualitatively. Evaluation of antifreeze activity of complexes in dairy products Sour cream sample preparation and freezing conditions Half and half cream (1200 mL) was heated to 30 ºC while stirring and then cooled to 25 ºC before adding 0.83 g of culture. The mixture was then fermented at 25 ºC for ~16 hours. The final pH was ~ 4.35. To create a smooth finished product the sour cream was mixed in a Kitchen Aid mixer on the lowest stir setting for 5 minutes. Sour cream was weighed out to 120 g and placed in plastic screw top containers. The lyophilized complexes were then added at 3% weight (as-is), which is equivalent to ~ 18 % dry-weight basis. The complexes were mixed by hand until fully incorporated. The lyophilized hydrolyzed WPI and LBG complex resulted in a hard mass after complexation that needed to be rehydrated before incorporation into the sour cream. The hydrolyzed WPI and LBG complex was hydrated in 3 mL of DI water overnight before being mixed into the sour cream. The protein and polysaccharides were added individually as solids at their respective ratio (5:1) as the controls and mixed by hand until fully incorporated. A “Fresh/Frozen + Water” treatment was added as a control to account for the additional water added in order to hydrate the hydrolyzed WPI and LBG complex. All samples were hydrated overnight at 4 ºC before being frozen at -22 ºC for 4 days. The samples were then thawed overnight at 4 ºC and equilibrated at room temperature for ~2-3 hours before testing. All treatments were performed in duplicate. Yogurt preparation and freezing conditions To increase the total solids in the milk, whole milk powder was added at 3.33% (w/v) to the whole milk (1200 mL) and hydrated overnight while stirring. The milk was heated to 90 ºC 38 while constantly stirring and held at this temperature for 10 minutes to denature the whey proteins. The milk was then cooled to ~43 ºC before adding 0.21 g of culture. The mixture was fermented at 49 ºC for 1 hour, then the temperature was reduced to 30 ºC for another 4 hours, and lastly the temperature was raised to 35 ºC for the last two hours to set the yogurt. The final pH was 4.6. The yogurt was then mixed in a Kitchen Aid mixer on the lowest stir setting for 1 minute to produce a smooth finished product. Yogurt was weighed out to 120 g and placed in plastic screw top containers. The lyophilized complexes were then added at 3% weight (as-is), which is ~ 19.5% dry-weight basis, and mixed until fully incorporated. The same hydration procedure for the hydrolyzed WPI and LBG complex as used for the sour cream was used in the yogurt. The protein and polysaccharides were added individually as solids at their respective ratio (5:1) as the controls and mixed by hand until fully incorporated. All samples were hydrated overnight at -4 ºC before being frozen at -40 ºC for 2 days and then at -22 ºC for 2 days. The different freezing conditions used for the yogurt was chosen to see if faster freezing and temperature change showed any difference in IRI activity or textural quality of the complexes. The samples were then thawed overnight at 4 ºC and equilibrated at room temperature for ~2-3 hours before testing. All treatments were performed in duplicate. Texture analysis to evaluate freeze induced damage A T.A.TXT2 texture analyzer (Texture Technologies Corp., South Hamilton, MA) was used to perform a single penetration test with a 2.5 cm diameter Perspex cylindrical probe and this was repeated in triplicate on a single sample. The test speed was 1.5 mm/s, the distance of penetration was 15 mm, and the trigger force was 5 g. The average firmness, cohesiveness, and consistency were recorded as the peak positive force, peak negative force, and positive area under the curve, respectively. 39 Lacunarity image analysis to quantify graininess due freeze induced damage This method was developed to quantify the visual graininess that has been observed in sour cream and yogurt after freezing. In brief, 0.3 g of each sample was gently mixed with 1 mL of water. Different sample dilutions and light exposures were investigated to optimize the sample prep conditions. It was determined that a 30% dilution with full light exposure produced the most consistent and accurate results. A drop was then placed onto a microscope and two images were taken at 10x objective using light microscopy (Leica, DM2700 M, Wetzlar, Germany) with a built-in digital camera (Leica, DMC 4500, Wetzlar, Germany). This was repeated in duplicate for each sample. Using Fraclac with Fiji, the lacunarity of the image was measured using a sliding box method (Karperien, 1999-2013). The analysis was run using the gray scale differential setting, the minimum pixel size was 5, the max size of the box was set to 45% of the image, the sliding distance was set to 5 pixels for both the X and Y direction, and a block series scaling method was used. Qualitative analysis of freeze induced damage by confocal laser scanning microscopy (CLSM) Fast green was used to stain protein and Nile red to stain fat in sour cream and yogurt samples to evaluate freeze induced matrix damage using confocal microscopy. Dye solutions were made at 1 mg/mL in methanol for Nile red and DI water for Fast green. Samples of 1.0 g were weighed and diluted in 9 mL of DI water. Each dye was added at 100 L and the samples inverted by hand to mix. The samples were left to stain overnight before being visualized under the confocal microscope. A single drop was placed on a microscope slide and a cover slip was placed on top. The samples were visualized under an inverted CLSM (Leica Microsystems, Baden-Wurttemberg, Germany) at 10x objective. The Nile red dye was excited at 488 nm and 40 the emission was collected between 520-590 nm while the Fast green was excited at 633 nm and the emission was collected between 660-750 nm. Syneresis measured by water separation To evaluate syneresis of the samples, 15 mL centrifuge tubes were filled with 10 mL of yogurt or sour cream after freezing. The samples were then left to naturally settle overnight and the water that had separated from the sample was calculated as % of total volume. This was done in duplicate. Statistical Analysis Complexes from the same batch were applied to two different batches of sour cream and yogurt. Results were reported as mean ± standard deviation. A nested design was used since not all of the treatments appeared in both the unhydrolyzed and hydrolyzed groups. The texture analysis data was log transformed to meet the assumption of an ANOVA (i.e. the data is normally distributed). An ANOVA test was conducted using JMP v17.0 (SAS Institute, Cary, NC, USA). Statistical significance was determined as P<0.05. A Tukey HSD post-hoc test was used to compare the means. Results and Discussion Degree of hydrolysis of the polysaccharide and protein hydrolysates In order to create complexes that have the potential to have increased mobility in the food matrix to act at the ice-water interface, the proteins and polysaccharides may need to be broken into lower molecular weight (MW) molecules. The LBG and LC were hydrolyzed by cellulase for 48 hours and DH results are shown in Table 3. The DH of both polysaccharides was low, however the reducing sugar content increased 30 and 22 times for LBG and LC respectively. This is an indication of sufficient breakdown of the polysaccharide. The lower-than-expected DH 41 Table 3: Reducing sugar content after 48 hr cellulase hydrolysis of LBG and LC samples (of 10 mg/mL). Reducing Sugar Content (mg/mL) Unhydrolyzed Hydrolyzed DH (%) LBG 0.01 ± 0.00 0.30 ± 0.00 2.86 LC 0.01 ± 0.00 0.22 ± 0.01 2.08 Standard deviation is between two replicates. is most likely due to the relatively less -1,4 glycosidic bonds compared to cellulose. Cellulase is an enzyme that is specific to -1,4-glycosidic bonds (Yoon et al., 2014). Due to this specificity, it is typically used to hydrolyze cellulose which is made of glucose monomers (Chandel et al., 2012, Yoon et al., 2014). LBG is a galactomannan that has a -1,4-mannose backbone and D- galactopyranosyl side branches linked by -1,6- linkages (Barak and Mudgil, 2014). Carrageenans are sulfated galactans that are made of alternating -1,3 and -1,4 glycosidic bonds between D-galactopyranosyl units (Damodaran and Parkin, 2017). WPI and SPI were hydrolyzed by Alcalase which is a non-specific protease. Due to the random nature of this enzyme, various MW distributions of peptides can be formed (Tacias- Pascacio et al., 2020). The results of the hydrolysis are shown in Table 4, which shows high efficiency of hydrolysis. Multiple hydrolysis times were tested for each protein and since there were similar MW distributions among times, 10 minutes was chosen for use in this work. WPI had more than 60% less than 1 kDa and 20% between 1 and 5 kDa. The small size of these hydrolysates could potentially result in poor complexation if the polysaccharides are too large compared to the protein hydrolysates. 42 Table 4: Native protein and hydrolysate MW distribution determined by HPLC-SEC. Peptide Size Distribution (%) Protein Time (min) <1 kDa 1-5 kDa 5-10 kDa >10 kDa SPI 0 20.8 18.9 9.3 51.1 2 59.0 25.9 9.5 5.7 5 66.6 19.4 10.6 3.3 10 54.6 26.1 9.1 10.2 WPI 0 0.3 19.8 2.0 78.0 2 51.8 25.7 10.9 11.7 5 52.4 25.8 9.9 11.9 10 65.3 20.8 7.8 8.3 43 Ice recrystallization inhibition (IRI) activity of protein-polysaccharide and hydrolysate complexes The IRI activity of the protein-polysaccharide and hydrolysate complexes are shown in Table 5. Not all complexes that were tested during preliminary testing were chosen for final evaluation. This is because that during preliminary testing some samples were found to be not IRI active in the model system but performed better in dairy system as found by texture analysis and vice versa. Various complexes were chosen for final evaluation based on their performance in the model and dairy system. The average percent decrease relative to PEG (negative control) is reported in Table 5. Complex treatment and hydrolysis treatment showed a significant effect (P<0.05). At both pH, the unhydrolyzed complexes of SPI and LBG, WPI and LBG, and the hydrolyzed SPI and lecithin complexes had the highest IRI activity as indicated by the highest reduction of ice crystal size relative to PEG. The unhydrolyzed WPI and LC complex did not show IRI activity at either pH, this was most likely due to its limited dispersibility. Besides the unhydrolyzed WPI and LC and the hydrolyzed WPI and LBG complex, all the complexes had IRI activity even at pH 4.5 which was important to establish because this is similar to the acidic environment of the sour cream and yogurt. All of the controls also showed to be inactive indicating a synergistic effect of the proteins and polysaccharides when complexed together. However, the hydrolysate complexes did not seem to have increased IRI activity compared to the unhydrolyzed complexes as what was first hypothesized. From the data of the model system, all complexes should have IRI activity under similar pH conditions in sour cream and yogurt and justify their evaluation in these dairy systems. In a study done by Gaukel et al. (2014), a synergistic effect between fish antifreeze proteins and sodium alginate in a sucrose solution was seen. Recrystallization inhibition activity 44 Table 5: IRI activity of unhydrolyzed and hydrolyzed complexes tested at 2% in 20 mM NaCl at pH 4.5 and 7.0. Ice Crystal Size Reduction Relative to PEG (%) pH 4.5 pH 7.0 Controls Unhydrolyzed Hydrolyzed Unhydrolyzed Hydrolyzed SPI -17.94 ± 9.60c -15.70 ± 12.19c 10.50 ± 14.51cd 7.71 ± 7.91cde WPI -15.35 ± 0.55c -15.72 ± 7.74c 1.43 ± 2.80cde 3.55 ± 7.72cde LBG -27.16 ± 16.11c -28.89 ± 9.40c 7.24 ± 4.65cde -26.92 ± 7.81e LC -30.69 ± 8.50c -26.69 ± 4.31c -2.10 ± 3.66cde 16.23 ± 7.05cd Lecithin -39.11 ± 10.71c — -3.25 ± 18.36cde — Complexes SPI and LBG 37.80 ± 7.69a — 30.74 ± 13.36abc — WPI and LBG 33.51 ± 19.60ab -6.45 ± 9.29bc 63.91 ± 3.28a -1.15 ± 3.88cde WPI and LC -19.71 ± 6.80c 0.79 ± 9.50abc -8.36 ± 9.01cde 21.51 ± 5.82bcd SPI and Lecithin — 27.43 ± 4.25ab — 54.98 ± 2.51ab Empty cells indicated by “—” were complexes that were not chosen to be further investigated due to their poor performance in preliminary testing. Standard deviation is between the average of three images from two replicate analysis (N=2). Means were compared within each pH group for both unhydrolyzed and hydrolyzed; different letters indicate treatments were significantly different (P<0.05).