casein concentrate and milk protein concentrate treated with transglutaminase in imitation cheese products—Unmelted texture

The amount of intact casein provided by dairy ingredients is a critical parameter in dairy-based imitation mozzarella cheese (IMC) formulation because it has a significant effect on unmelted textural parameters such as hardness. From a functionality perspective, rennet casein (RCN) is the preferred ingredient. Milk protein concentrate (MPC) and micellar casein concentrate (MCC) cannot provide the required functionality due to the higher steric stability of casein micelle. However, the use of transglutaminase (TGase) has the potential to modify the surface properties of MPC and MCC and may improve their functionality in IMC. The objective of this study was to determine the effect of TGase-treated MPC and MCC powders on the unmelted textural properties of IMC and compare them with IMC made using commercially available RCN. Additionally, we studied the degree of crosslinking by TGase in MPC and MCC retentates using capillary gel electrophoresis. Three lots of MCC and MPC retentate were produced from pasteurized skim milk via microfiltration and ultrafiltration, respectively, and randomly assigned to 1 of 3 treatments: no TGase (control); low TGase: 0.3 units/g of protein; and high TGase: 3.0 units/g of protein, followed by inactivation of enzyme (72°C for 10 min), and spray drying. Each MCC, MPC, and RCN was then used to formulate IMC that was standardized to 21% fat, 1% salt, 48% moisture, and 20% protein. The IMC were manufactured by blending, mixing, and heating ingredients (4.0 kg) in a twin-screw cooker. The capillary gel electrophoresis analysis showed extensive inter-and intramolecular crosslinking. The IMC formulation using the highest TGase level in MCC or MPC did not form an emulsion because of extensive crosslinking. In MPC with a high level of TGase, whey protein and casein crosslinking were observed. In contrast, crosslinking and hydrolysis of proteins were observed in MCC. The IMC made from MCC powder had significantly higher texture profile analysis hardness compared with the corresponding MPC powder. Further, many-to-one (multiple) comparisons using the Dunnett test showed no significant differences between IMC made using RCN and treatment powders in hardness. Our results demonstrated that TGase treatment causes crosslinking hydrolysis of MCC and MPC at higher TGase levels, and MPC and MCC have the potential to be used as ingredients in IMC applications.


INTRODUCTION
Intact casein in cheese is defined as casein that has not undergone hydrolysis during aging (Kapoor and Metzger, 2008), and it contributes to the structural network of processed cheese.Intact casein content is critical in dairy-based imitation mozzarella cheese (IMC) and process cheese (PC) formulations to achieve the desired unmelted textural properties.Ingredients contributing intact casein in IMC include rennet casein (RCN), acid casein, milk protein concentrates (MPC), micellar casein concentrate (MCC), and NDM.However, the amount of intact casein provided by these ingredients varies significantly and has a substantial effect on the functional quality of IMC.Rennet casein is the ingredient of choice because it is considered to be 100% intact and provides the highest viscosity per gram of protein (Kapoor and Metzger, 2008), resulting in desirable textural properties.However, RCN is not cost effective for the US dairy industry, and alternative ingredients are needed to substitute RCN in IMC applications.
Manufacturing of MPC involves ultrafiltration (UF) and diafiltration (DF) of skim milk to produce partly or completely delactosed high-protein dairy ingredients.In the United States, MPC is manufactured with a total protein level (on a DM basis) ranging from 56 to 85% (Marella et al., 2013).The protein in MPC is a mixture of caseins and whey proteins (WP) in the same ratio as in milk (Carter et al., 2021;Salunke et al., 2021).The use of MPC in PC and IMC formulations can lead to defects such as a weak body and soft texture, low cooked viscosity, and restricted melting properties (Kapoor and Metzger, 2008;Salunke and Metzger, 2022a,b).In contrast, MCC is obtained from microfiltration (MF) of skim milk.The MCC produced in an MF-based process has a higher level of intact casein and a reduced level of WP compared with MPC (Carter et al., 2021;Salunke et al., 2021).In earlier work, we observed that PC produced using MCC had a desirable texture and viscosity compared with PC produced using MPC (Metzger, 2007;Salunke andMetzger, 2008, 2022a).However, the firmness and cooked viscosity of PC produced with MCC was still substantially lower than that of PC produced with RCN.The MPC and MCC did not provide comparable functionality to RCN in PC and IMC applications because of the presence of κ-CN in the native casein micelle that might contribute to the defects in IMC when higher levels of MPC or MCC are used in formulations.This could be due to the higher steric stability of casein micelles in the presence of κ-CN and lead to a soft body and reduced viscosity of IMC.The steric stability of κ-CN can be modified to improve the functionality of casein to match its functionality to that of RCN.
Various physical, chemical, and enzymatic methods are available to modify milk protein functionality (Fox and Mulvihill, 1982;Gerrard, 2002).Crosslinking the proteins with transglutaminase (TGase, EC 2.3.2.13) is one method to modify the functionality and structural properties of proteins such as caseins and WP.Transglutaminase catalyzes the acyl transfer reaction between the protein-bound glutaminyl residues and the primary amines (Folk and Finlayson, 1977;Chen and Hsieh, 2016;Raak et al., 2018).Transglutaminase has the potential to modify the physical properties of caseins by covalently crosslinking the caseins and thereby reducing the steric stability of κ-CN.We hypothesize that it is possible to create a TGase-mediated crosslinked protein network in MPC and MCC that has similar functionality to the RCN in PC or IMC formulations.In the United States, TGase has GRAS (generally regarded as safe) status, and the USDA has authorized its use in some dairy products, including cheese and yogurt.Our objective was to develop novel MPC and MCC ingredients treated with TGase.Further, the new ingredients were characterized in terms of degree of crosslinking using capillary gel electrophoresis (CGE).The newly developed ingredients were used in IMC formulations to understand the unmelted textural properties in comparison with IMC manufactured using RCN.

Experimental Design
Three replicates of MCC and MPC were produced using MF and UF, respectively, from 3 different lots of skim milk.Each lot of MCC or MPC was divided into 3 equal portions and subsequently subjected to 3 levels of TGase treatment, as per the experimental design.The TGase treatments were (1) control (C; no TGase), (2) low TGase (L; 0.3 U/g of protein), and (3) high TGase (H; 3.0 U/g of protein).After a 10-min heating at 72°C to inactivate the enzyme, each retentate was spray dried, and the obtained powders were subsequently used as an ingredient in the manufacture of IMC.A 2 × 3 factorial design experiment was set up with 2 product types (MPC or MCC) and 3 TGase enzyme levels: C, L, or H. Furthermore, rennet caseinbased IMC was manufactured using different lots of commercially available rennet casein (RCN treatment).

Manufacture of MPC and MCC and TGase Treatment
Pasteurized (63°C/30 min) skim milk (454 L) obtained from the Davis dairy plant (South Dakota State University, Brookings) was divided into 2 equal portions.One portion was processed into MPC using the UF process and the other into MCC using the MF process using the methods described by Salunke et al. (2021).

MPC Manufacture
Skim milk (227 L) was ultrafiltered to a final retentate volume of 45.4 L, achieving a volume reduction ratio of approximately 5 (on a feed volume basis).The UF process was carried out at 23.3°C using a 10-kDa polyether sulfone spiral-wound membrane (3838 element format with 1.1-mm feed spacer and 5.7-m 2 area; Parker Hannifin Corp.) at a transmembrane pressure of 276 kPa.Diafiltration water (40% of the feed volume) was added at 4 predetermined intervals (water added at concentration factors of ~1.6, 2.0, 2.5, and 3.0 in equal parts to maintain flux rate).

MCC Manufacture
Skim milk (227 L) was microfiltered to a final retentate volume of 45.4 L, achieving a volume reduction ratio of approximately 5 (on a feed volume basis).The MF process was carried out at 23.3°C using a 0.5-µm polyvinylidene fluoride spiral-wound membrane (3830 element format with 1.1-mm feed spacer and 4.3-m 2 area; Parker Hannifin Corp.) operating at a transmem- brane pressure of 86 kPa.Diafiltration water was added at 6 intervals totaling 100% (on a feed volume basis).Immediately after production, the retentates were frozen to −18°C until further use.

TGase Treatment and Powder Manufacture
Each lot of MCC and MPC retentate was divided into 3 equal portions.One portion served as control, and the other 2 were treated with 2 levels of TGase (Activa TI, Ajinomoto Food Ingredients LLC; declared activity of 100 U/g): low (0.3 U of TGase/g protein) and high (3.0U of TGase/g protein).For each TGase treatment, the required quantity of TGase was weighed and mixed with 100 mL of distilled water.In the control treatment, 100 mL of distilled water was added without enzyme.After enzyme addition and thorough mixing, each treatment was incubated at 50°C for 25 min.Subsequently, the retentates were heated to 72°C for 10 min to inactivate the enzyme and then cooled to 4°C.All the retentates were then spray dried (ASO 412E, Niro Inc.).During drying, the inlet air temperature was maintained at 205°C and the outlet temperature was maintained at 90°C.The powders were collected in plastic bags (Associated Bag Co.) and stored at room temperature until further analysis.

Chemical Analysis of Liquid and Dried Samples
Proximate analysis of retentate, TGase-treated retentate, powders, and IMC samples were analyzed using their respective standard methods (Hooi et al., 2004).Lactose was analyzed using the HPLC (Beckman-Coulter Inc.) method described by Amamcharla and Metzger (2011).Calcium content was measured by atomic absorption spectroscopy (AAnalyst 200, Perki-nElmer Instruments LLC) using the method described by Metzger et al. (2000).In addition, CGE was carried out on TGase-treated retentates to determine the extent of crosslinking.

CGE
The control and TGase-treated samples were diluted to a protein level of 10 mg/mL using HPLC-grade water.The CGE was carried out using a Beckman P⁄ACE MDQ capillary electrophoresis system (Beckman-Coulter Inc.) equipped with a UV detector set at 214 nm using the method as described by Salunke and Metzger (2022b).Separation was obtained using a 50-µm bare fused-silica capillary.All reagents and available pure standards (used for identifying proteins) were of analytical grade from Sigma or obtained as a part of the ProteomeLab SDS-molecular weight (MW) Analysis Kit (Beckman-Coulter Inc.) designed for the separation of protein-SDS complexes using a replaceable gel matrix.The SDS-MW size standard (recombinant proteins 10-225 kDa) was used to estimate the MW of the proteins in the sample.Separation was performed at a constant voltage of 15 kV (25°C and 2,000 kPa) with reverse polarity in SDS-MW gel buffer.The area of each peak was calculated from the electropherogram.The area of each identified casein fraction, serum protein (SP) fraction, and NPN fraction (all positive peaks below 10 kDa) was calculated.Based on the MW standards migration time the peak areas were divided into 4 major groups representing major protein fractions: <10 kDa (NPN and low MW peptides), 10-20 kDa (major SP and medium MW peptides), 20-50 kDa (CN fractions), and >50 kDa (high MW fractions).The degree of crosslinking was interpreted based on the changes in relative peak area in each group.The relative peak area for each group was obtained by adding the area of each positive peak within the group and dividing by total area (Salunke and Metzger, 2022b).Additionally, the appearance and disappearance of peaks and the qualitative analysis of peak shape and size are reported.

IMC Manufacture and Analysis
IMC Formulation.The IMC formulations were developed and balanced using Techwizard, a spreadsheet software provided by Owl Software.The IMC was formulated (Table 1) to target moisture, fat, salt, and protein contents of 48.0, 21.0, 1.0, and 20.0%, respectively.In each formulation, the MPC or MCC was used as a sole protein source.For the RCN treatment, 3 lots of commercially available rennet casein were used.Vegetable shortening (Crisco, all-vegetable shortening, The J. M. Smucker Co.) was used as the fat source.Sodium aluminum phosphate basic powder (Kasal, Innophos) was used as the emulsifying salt.The other ingredients in the formulation included deproteinized whey powder (Davisco Foods International Inc.), citric acid (KIC Chemical Inc.), lactic acid 85% (wt/wt; Fisher Scientific), and iodized salt (Great Value, Wal-Mart Stores Inc.).
IMC Manufacture.The IMC was manufactured in a twin-screw cooker (Blentech Corp.).A preblend of all the ingredients (except lactic acid) was prepared (4.0 kg) in the twin-screw cooker by mixing at 50 rpm for 20 min with no heating.The temperature of the preblend was then increased to 74°C over 5 min and held for an additional 4 min.The auger speed was increased to 120 rpm during the heating and holding periods.After the holding time, lactic acid was added and mixed for an additional 30 s. Immediately after manufacture, the IMC in molten condition was filled in plastic contain-ers with lids (Pro-kal-1602 Squat Polypropylene deli containers with lids, Fabri-Kal), and transferred immediately to a cold room (4°C).The IMC was stored at 4°C until further analysis.
Chemical Analysis of IMC.After 1 wk of storage at 4°C, the pH of the samples was measured in duplicate using a pH meter (Corning pH meter 340, Corning Inc.) equipped with an Accumet gel-filled glass electrode with a spear tip (Fisher Scientific).The proximate analysis of the IMC samples (moisture, fat, and total protein) was carried out using standard methods described by Hooi et al. (2004).

IMC Functional Analysis (Unmelted Textural Properties).
The IMC unmelted functional properties were studied using texture profile analysis (TPA).For TPA, samples were prepared using a cheese borer (20 mm internal diameter), cut to a height of 20 mm, and wrapped in plastic film to prevent moisture loss.Then, TPA was performed using a TA.XT2 Texture Analyzer (Texture Technologies Corp./Stable Microsystems) as described by Drake et al. (1999).After samples (n = 4) were tempered for 15 min at 20°C, they were twice compressed to 80% of their original height with a 50mm cylindrical flat probe at a crosshead speed of 0.8 mm/s (uniaxial 2-bite compression test).

Statistical Analysis
Samples of retentate, TGase-treated retentate, powders, and IMC were analyzed for their composition and degree of crosslinking.The IMC were characterized in terms of unmelted textural properties.Statistical analysis was done using Proc GLM (SAS Institute Inc.).The IMC with a high TGase level did not form an emulsion, and were omitted from further statistical analysis.A 2 × 2 factorial analysis with a type I error rate (α) of 0.05 to test for significant differences among the treatments was used.The IMC manufactured using RCN was compared with all treatment IMC using the Dunnett test.

MCC and MPC Powders
Mean values for total protein (DM basis), moisture, lactose, fat, calcium, and ash in MCC and MPC powders are shown in Table 2.As expected, protein and calcium contents were higher in MCC powders, whereas lactose content was significantly lower in MCC compared with MPC powders (P < 0.05), reflecting the different manufacturing processes.There was no significant difference in the fat content of the powders (P = 0.46); however, there were small but significant differences in moisture (P < 0.05) content because of drying conditions.

Degree of Crosslinking: CGE
Figures 1 and 2 show typical CGE electropherograms of MCC and MPC treated with TGase, respectively.Changes in protein fractions can be measured in CGE by analyzing changes in peak shape and size and the emergence of new peaks or disappearance of existing peaks in the CGE electropherogram.In the electropherograms, the protein fractions <10 kDa represent NPN and low MW protein fractions (peptides), the 10-20 kDa represent SP and medium MW peptides, the 20-50 kDa represents CNs fractions, and the >50 kDa MW represents high MW protein fractions consisting of TGase crosslinked monomers of CN or whey protein.The quantification of change in MW caused by TGase was analyzed by measuring the peak area and calculating the ratio of protein at various MW ranges.The mean square values and the P-values of the protein fractions at <10 kDa, 10-20 kDa, 20-50 kDa, and >50 kDa MW obtained from the CGE electropherograms are shown in Table 3, and the mean values are shown in Table 4.There was a significant (P < 0.05) effect of product type on peak areas in the 10-20 and 20-50 kDa ranges, and there was a significant (P < 0.05) effect of enzyme level in the 20-50 and >50 kDa ranges.The significance (P < 0.05) of the product type and enzyme level indicates that there was a change in peak areas, whereas the significance (P < 0.05) of the interaction term (product type × enzyme level) in the <10 and 20-50 kDa regions indicates that these changes were not linear.
The control samples (MCC-C and MPC-C) had peak areas similar to the retentate peaks seen in a previous study (Salunke et al., 2021).The quantitative analysis also showed similar results, with MCC-C having less noncasein N and more CN than the MPC-C samples (Salunke et al., 2021).The 10-20 kDa region contained the noncasein N peaks, whereas the 20-50 kDa MW region contained the CN peaks (Figures 1a and 2a), with significant differences (P < 0.05) in MCC and MPC (Salunke et al., 2021).
In MCC and MPC, addition of TGase at low levels (MCC-L and MPC-L) caused a broadening of peaks, a change in peak size and shape, and the emergence of a new peak >35 kDa (Figures 1b and 2b).The action of TGase resulted in crosslinking (inter-and intramolecular) and consequently increased the MW of the proteins.Intramolecular crosslinking does not increase the MW of the protein; however, intermolecular crosslinking increases the MW based on the number and type of proteins involved in crosslinking.Broadening of peaks indicated intramolecular (within) covalent crosslinking, whereas the appearance of high MW peaks and the emergence of new peaks compared with control samples indicated intermolecular (between) crosslinking.The inter-and intra-crosslinking in MCC retentates treated with TGase have been reported previously by Salunke and Metzger (2022b).
In MPC-L (Figure 2b), the SP peaks were identified clearly, and broadening of peaks was seen in some low MW SP.Quantitatively, there was an increase in the 10-20 kDa area in MPC-L, indicating the formation of new peaks due to crosslinking of various medium MW peptides (Table 4).The MPC samples had higher noncasein N content, and as the TGase level increased, the peak area decreased in the 10-20 kDa region, with a simultaneous increase in the high MW (<50 kDa) region in MPC-L, indicating crosslinking between CN and whey proteins.A similar phenomenon was observed in MCC-L samples, wherein the CN peak (20-50 kDa) area decreased as TGase level increased, with a simultaneous increase in <50 kDa peak area (Table 4).
The reduction in peak height and broadening of the individual fraction is clearly evident in MCC-H and MPC-H samples (Figures 1c and 2c).New peaks appeared in both higher and lower MW areas, indicating extensive inter-and intramolecular crosslinking and rearrangements.At higher levels of TGase (MCC-H and MPC-H), the κ-CN peak disappeared and there was a reduction in the peak height and area of β-CN, a broadening and merging of peaks, and formation of new high MW peaks, indicating extensive inter-and intramolecular crosslinking.A change in the shape of the CGE electropherogram also indicated structural changes on CN (κ-CN) due to crosslinking.This indicated that κ-CN and β-CN were preferentially crosslinked because of their peripheral location and dynamic nature, respectively.Sharma et al. (2001) and Hsieh and Pan (2012) reported crosslinking of κ-CN and β-CN in skim milk due to TGase action.Very subtle changes were  observed in α S -CN.α S -Caseins are the backbone of the micelle structure, and Glu sites are buried inside α S -CN due to its structure and are therefore not easily accessible to TGase and not easily modified by it (Sharma et al., 2001).Additionally, in the MPC-H treatment, we observed disappearance of peaks, broadening of peaks, and development of new peaks in the 10-20 kDa region, indicating that the SP is modified or crosslinked at the high TGase level.The significant (P < 0.05) increase in peak area (<50 kDa) in MPC-H indicated intermolecular crosslinking between CN and SP (Table 4).
As shown in Table 4 and evident from Figures 1c and  2c, there was a significantly (P < 0.05) higher level of protein <10 kDa MW in MCC-H compared with MPC-H and a significantly (P < 0.05) lower level in MCC-C.In the MPC treatments, particles <10 kDa in MPC decreased as TGase level increased, whereas in the MCC, <10 kDa particles increased with TGase level.The high amount of medium MW protein fractions in MCC-H was surprising and indicated that protein was hydrolyzed.Milk contains many endogenous enzymes that are proteolytic in nature and can cause hydrolysis of proteins (Fox and McSweeney, 1998).However, the possibility of hydrolysis by endogenous enzymes is unlikely because all treatments were handled similarly, and protein hydrolysis was seen only in the MCC-H sample.These results indicate that TGase may hydrolyze some protein fractions (Table 4).However, we observed only a nominal increase in the NPN content of MCC analyzed by the Kjeldahl method (Salunke et al., 2021).
A possible mechanism for the presence of medium MW peptides in the MCC-H treatment is hydrolysis or deamidation of Glu residues by TGase.It is well established that TGase results in crosslinking of peptide-bound glutamine and lysine residues, forming ε-(γ-glutamyl) lysine isopeptide bonds and high MW polymers.However, in the absence of amine substrates, TGase can catalyze the deamidation of glutamine residues (Motoki et al., 1986;Griffin et al., 2002;Stamnaes et al., 2008).Stamnaes et al. (2008) reported that higher amounts of substrate (CN here) and deamidation reaction can cause partial hydrolysis of proteins.Transglutaminase can catalyze the replacement of some amide groups of protein-bound glutamine residues and catalyze their hydrolysis, as demonstrated by its action on substrate proteins and their chemically acetylated and deaminated modifications (Mycek and Waelsch, 1960).
At the high TGase level, we observed a simultaneous increase in peaks >50 kDa in MCC and MPC, with a higher increase in MPC.This indicates an increase  (Sharma et al., 2001;Hsieh and Pan, 2012;Chen and Hsieh, 2016) in milk and caseinate solutions.Similar to our results, Partschefeld et al. (2007), using gel permeation chromatography, reported that at low enzyme activity, mainly dimers and trimers were formed, whereas at higher activity, oligomers and polymers were formed.Salunke and Metzger (2022b) reported an increase in crosslinking and the presence of high MW proteins in MCC retentate treated with TGase.
Analysis of the degree of crosslinking shows that there was a change in MW indicating extensive inter-and intramolecular crosslinking by TGase enzyme, modifying the surface properties of casein micelles.

IMC Manufacture and Composition
The IMC formulations with high TGase treatment (MCC-H and MPC-H) did not form emulsions during manufacture.Consequently, no functional analysis was performed on these treatments.This may have been caused by the high level of crosslinking (Table 4; Figures 1c and 2c) that limits the emulsion capacity of casein.To form a proper emulsion, disaggregation of casein is essential.In PC manufacture, insoluble calcium phosphate para-caseinate or casein is converted into soluble or hydrated dispersible forms in the presence of emulsifying salt, heat, and shear action (Kapoor and Metzger, 2008;Guinee, 2011).Similar reactions occur in IMC manufacture even though the source ingredient may be different.Hence, at a higher degree of crosslinking (MCC-H and MPC-H), it was difficult to disaggregate casein and form an emulsion due to the strong covalent bond aggregates of proteins (Table 4; Figure 1c and 2c).The covalent crosslinking from TGase has been shown to be stable against many disruptive forces, including heat, high pressure, ethanol, urea, and emulsifying salts (O'Sullivan et al., 2002;Mounsey et al., 2005;Huppertz and de Kruif, 2007).Intramicellar crosslinking enhances the stability of casein micelles, and the net-like crosslinks formed within the external region of the micelles adopt the stabilizing role of colloidal calcium phosphate within the micelles, thus making the micelles strong against disrupting influences (Partschefeld et al., 2007).
The mean composition of the IMC is shown in Table 5.All replicates used in this study were balanced for moisture, fat, salt, protein, and lactose.Even though the RCN, MCC, and MPC powders were different compositionally, all IMC formulations were balanced for similar proximate composition.No significant differences (P > 0.05) were observed in the pH, fat, moisture, or protein content of the IMC.The proximate composition of the IMC (fat, protein, and moisture) analyzed (Table 5) was close to the formulated targets.The small differences in composition (Table 5) may be due to small variations in manufacturing and packing conditions.

Unmelted Functional Properties: TPA Parameters
The unmelted functional properties were measured using TPA.The mean square values and P-values of the unmelted functional properties are shown in Table 6, and the mean values are shown in Table 7.There was a significant (P < 0.05) replicate effect in TPA hardness, springiness, and resilience.This could be due to the variation in TGase-induced crosslinking of proteins, as there was no significant difference in the proximate composition of IMC (Table 5).Samples of retentate, TGase-treated retentate, powders, and IMC were analyzed for their composition and degree of crosslinking.A significant (P < 0.05) effect of product type on TPA hardness, fracturability, springiness, gumminess, chewiness, and resilience of the manufactured IMC was found.There was a significant (P < 0.05) effect of enzyme level on adhesiveness, springiness, chewiness, and resilience of the manufactured IMC.All of the TPA parameters except cohesiveness were higher for IMC made with MCC powders compared with IMC made from MPC powders.The hardness and gumminess values of MCC-C and MCC-L were similar but significantly (P < 0.05) higher than those of MPC-C and MPC-L (Table 7).Lower hardness values are typical of soft body defects reported with the use of MPC in IMC.Salunke and Metzger (2022a) reported soft body issues in PC products made using MPC.The fracturability and chewiness values were significantly (P < 0.05) higher for MCC-L and significantly (P < 0.05) lower for MPC-C compared with MCC-C and MPC-L.Fracturability and chewiness values were significantly (P < 0.05) higher for MCC-C than MPC-L.The adhesiveness values significantly (P < 0.05) decreased in both MCC and MPC samples when TGase was added.Springiness and resilience values of MCC-L were significantly (P < 0.05) higher and that of MPC-C significantly (P < 0.05) lower compared with MCC-C and MPC-L samples.The addition of TGase and formation of crosslinking bonds decreased adhesiveness and increased springiness, chewiness, and resilience in both products, indicating a substantial change in IMC texture.The higher amount of crosslinking observed in MCC-L and MPC-L (Table 4, Figures 1 and 2) compared with control samples changed the adhesiveness, springiness, chewiness, and resilience of the IMC.Breene (1975) defined TPA hardness as a measure of the unmelted texture of a cheese that describes firmness; TPA hardness is the main parameter defining the functionality of IMC.During PC manufacture, extensive protein-based interactions occur that lead to a strong protein network (Garimella Purna et al., 2006).The levels of intact casein (Berger et al., 1998), as well as its pH and calcium-to-casein ratio, influence the extent of casein hydration during the manufacture of PC, and intact casein, in turn, influences the degree of emulsification, the degree of casein aggregation, and the elasticity of the final PC (Guinee et al., 2004;Guinee, 2007).
In IMC applications, the ability of casein to form a gel network is exploited, and casein provides a desirable  (Fox and McSweeney, 1998).The denaturation of whey proteins can lead to precipitation and flocculation, resulting in undesirable product characteristics (gelation or high viscosity) and a soft body.The use of young cheese in PC products has been shown to increase hardness because of intact casein (Garimella Purna et al., 2006).Higher hardness in samples having MCC in PC formulations has been reported, and the differences were attributed to a higher intact CN and a lower WP content in the MCC used (Salunke and Metzger, 2022a).We expect similar results when MCC providing intact CN is used in IMC formulation.

Comparison with RCN
The IMC manufactured using RCN as an ingredient was compared with IMC manufactured using MCC and MPC powders using the Dunnett test.The mean values and the Dunnett test comparison with treatment powders for all functional properties are shown in Table 8.There were significant (P < 0.05) differences in IMC made using RCN and MCC and MPC powders in fracturability and cohesiveness.No significant (P < 0.05) differences were observed between RCN and MCC and MPC powders in TPA hardness or adhesiveness.The differences in the IMC were due to the differences in CN availability, SP content, and calcium level.
There is a difference between the intact casein found in rennet casein or cheese and native casein obtained through membrane filtration (UF, MF, or diafiltration).Almost all casein micelle structure models, including the recent dual-binding model of the casein micelle, state that κ-CN protrudes from the surface layer as a flexible hair, having hydrophilic properties (de Kruif et al., 2012).This κ-CN provides steric stability to casein micelles.The κ-CN on the casein micelles surface has a highly negative charge due to glycomacropeptide (GMP), which provides strong repulsive steric interactions that prevent casein aggregation.Due to rennet action during the manufacture of cheese and rennet casein, the GMP portion is cleaved off (Fox and McSweeney, 1998), changing the charge profile and steric stability of κ-CN and causing aggregation of paracasein, which is highly insoluble.The rennet casein has a high pH and high ash content, especially Ca and P found as colloidal calcium phosphate (Augustin et al., 2011).The intact casein obtained through MF and UF is soluble and has a reduced level of SP and soluble constituents (calcium and other minerals, and lactose, depending on the diafiltration used), and MF and UF concentrate the casein micelles, maintaining their integrity and soluble nature.The surface of the casein micelle is highly negatively charged, and this charged surface of the MCC may inhibit the dispersion of the native casein by emulsifying salts.The action of TGase brings about changes in surface properties (Table 4; Figure 1 and 2) and charge so that these products can function similarly to rennet casein.The action of TGase causes proteins to become insoluble (Salunke and Metzger, 2011), similar to rennet casein.When rennet casein is used, even after the addition of emulsifying salts, the small aggregated form of insoluble para casein remains, and after heat- ing, mixing, and cooling, it forms a strong IMC network that gives firmer texture and long strands during melting.Hence, the use of rennet casein provides the desirable strong IMC network.
The main IMC requirement of hardness improved in MCC-L after TGase action; IMC made from MCC-L had greater hardness than RCN, whereas IMC made from MPC-L had lower hardness compared with RCN.The IMC manufactured from MPC treatments could not provide better functionality partly because of the presence of SP.The SP content in the formulations provided by MPC, MCC, and RCN was 2.55, 1.38, and 0%, respectively.These results indicate that noncasein N had a detrimental effect on the functional properties of the IMC due to SP denaturation.The TGase treatment did modify some functional properties in MPC; however, they could not match those provided by MCC or RCN.However, the TGase treatment of MCC had a positive effect and improved certain functional properties.Salunke and Metzger (2022b) reported a significant increase in the hardness of PC products manufactured using TGase-treated MCC retentate and similar hardness in PC products compared with PC made using RCN.Other researchers have also reported a change in the texture of cheese when adding TGase to milk or curd.Nasr (2019) reported improvements in textural characteristics of mozzarella cheese by adding 0.04% TGase to milk.Li et al. (2019) reported that protein structure was enhanced by TGase in mozzarella cheese.Topcu et al. (2020) used microbial TGase (0.75 U/g of protein) with emulsifying salt in the cheese curd for the manufacture of processed Kashar cheese and found a different texture from control cheese.

CONCLUSIONS
The TGase treatment significantly affected the degree of crosslinking and functionality of the IMC.Treatment with TGase caused covalent crosslinking of proteins (caseins and WP) in MCC and MPC, as confirmed by CGE.The higher level of TGase in MCC retentate caused some hydrolysis of crosslinked protein.The IMC formulations using MCC or MPC treated with the highest TGase concentration did not form emulsions due to excessive crosslinking.By changing the enzyme level, incubation time, or substrate concentrations, new MPC and MCC products with enhanced functional properties can be manufactured.The IMC made from MCC (with or without TGase) had higher TPA hardness than the corresponding MPC samples.A comparison of IMC made using TGase-treated MCC and MPC to that made using RCN indicated no difference in TPA hardness.We concluded that TGase treatment causes crosslinking and hydrolysis of MCC and MPC at higher levels, and MPC and MCC have the potential to be used as ingredients in IMC applications and to replace RCN in IMC formulations.
Salunke et al.: MICELLAR CASEIN CONCENTRATE AND IMITATION CHEESE TEXTURE

Table 3 .
Salunke et al.: MICELLAR CASEIN CONCENTRATE AND IMITATION CHEESE TEXTURE Mean squares and P-values (in parentheses) of ratio of group peak area to total peak area obtained in capillary gel electropherogram of control and transglutaminase-treated micellar casein concentrate (MCC) and milk protein concentrate (MPC) powders

Table 4 .
Ratio of group peak area to the total peak area obtained in capillary gel electropherogram of control and transglutaminase-treated micellar casein concentrate (MCC) and milk protein concentrate (MPC) powders MCC-C = micellar casein concentrate powder, control (no transglutaminase, TGase); MCC-L = micellar casein concentrate powder, low TGase level; MCC-H = micellar casein concentrate powder, high TGase level; MPC-C = milk protein concentrate powder, control; MPC-L = milk protein concentrate powder, low TGase level; MPC-H = milk protein concentrate powder, high TGase level; RCN = rennet casein powder. in MW of MCC-H and MPC-H treatments.Various researchers have reported increases in CN crosslinking, degree of polymerization, and MW a-dMeans (±SD) in a row with common superscripts do not differ (P < 0.05; n = 3).1

Table 5 .
Salunke et al.: MICELLAR CASEIN CONCENTRATE AND IMITATION CHEESE TEXTURE Mean (±SD) values (n = 3) of the composition of the 5 imitation mozzarella cheese treatments manufactured in a twin-screw cooker

Table 6 .
Salunke et al.: MICELLAR CASEIN CONCENTRATE AND IMITATION CHEESE TEXTURE Mean squares and P-values (in parentheses) of texture profile analysis parameters of the 4 imitation mozzarella cheese treatments manufactured in a twin-screw cooker

Table 7 .
Mean (±SD) values (n = 3) of the functional properties of the 4 imitation mozzarella cheese treatments manufactured in a twin-screw cooker firm unmelted texture and stringy, elastic melted texture.Moreover, MCC provides a higher level of intact CN than MPC.The average contributions of SP in the IMC formulation were 1.38 and 2.55% for MCC and MPC, respectively.Milk protein concentrate, which has more SP, causes textural defects and a decrease in functionality, as SP can crosslink among themselves and with caseins at high temperatures

Table 8 .
Salunke et al.: MICELLAR CASEIN CONCENTRATE AND IMITATION CHEESE TEXTURE Mean values (n = 3) of the functional properties of the rennet casein imitation mozzarella cheese (IMC) and comparison with IMC treatments manufactured in a twin-screw cooker IMC made with rennet casein powder; MCC-C = micellar casein concentrate powder-control (no transglutaminase, TGase); MCC-L = micellar casein concentrate powder, low TGase level; MPC-C = milk protein concentrate powder, control; MPC-L = milk protein concentrate powder, low TGase level.