Journal of Dairy Science
Volume 89, Issue 9 , Pages 3296-3305, September 2006

Effect of Protein Composition on the Cheese-Making Properties of Milk from Individual Dairy Cows

  • A. Wedholm

      Affiliations

    • Department of Food Science, Swedish University of Agricultural Sciences, SE-750 07 Uppsala, Sweden
    • Corresponding Author InformationCorresponding author.
    • ,
  • L.B. Larsen

      Affiliations

    • Department of Food Science, Danish Institute of Agricultural Sciences, Research Centre Foulum, DK-8830 Tjele, Denmark
    • ,
  • H. Lindmark-Månsson

      Affiliations

    • Swedish Dairy Association, SE-223 63 Lund, Sweden
    • ,
  • A.H. Karlsson

      Affiliations

    • Department of Food Science, The Royal Veterinary and Agricultural University (KVL), DK-1958 Fredriksberg C, Denmark
    • ,
  • A. Andrén

      Affiliations

    • Department of Food Science, Swedish University of Agricultural Sciences, SE-750 07 Uppsala, Sweden

Received 13 January 2006; accepted 12 March 2006.

Article Outline

Abstract 

The objective of this study was to evaluate the effect of variations in milk protein composition on milk clotting properties and cheese yield. Milk was collected from 134 dairy cows of Swedish Red and White, Swedish Holstein, and Danish Holstein-Friesian breed at 3 sampling occasions. Concentrations of αS1-, β-, and κ-casein (CN), α-lactalbumin, and β-lactoglobulin (LG) A and B were determined by reversed phase liquid chromatography. Cows of Swedish breeds were genotyped for genetic variants of β- and κ-CN. Model cheeses were produced from individual skimmed milk samples and the milk clotting properties were evaluated. More than 30% of the samples were poorly coagulating or noncoagulating, resulting in weak or no coagulum, respectively. Poorly and noncoagulating samples were associated with a low concentration of κ-CN and a low proportion of κ-CN in relation to total CN analyzed. Furthermore, the κ-CN concentration was higher in milk from cows with the AB genotype than the AA genotype of κ-CN. The concentrations of αS1-, β-, and κ-CN and of β-LG B were found to be significant for the cheese yield, expressed as grams of cheese per one hundred grams of milk. The ratio of CN to total protein analyzed and the β-LG B concentration positively affected cheese yield, expressed as grams of dry cheese solids per one hundred grams of milk protein, whereas β-LG A had a negative effect. Cheese-making properties could be improved by selecting milk with high concentrations of αS1-, β-, and κ-CN, with high κ-CN in relation to total CN and milk that contains β-LG B.

Key words: cheese yield, milk protein composition, milk clotting properties, poorly coagulating milk

 

Back to Article Outline

Introduction 

In many milk-producing countries, a large fraction of the milk produced is used for cheese making. In the Scandinavian countries, about one-third of the total volume is used for this purpose. The quality and amount of cheese obtained, not only per volume of milk but also per gram of protein in cheese milk, is important for the economic outcome of the dairy industry. The milk clotting properties are important both with regard to quality and yield of cheese. It has been suggested that a firmer curd at cutting is positively correlated to yield of cheese (Aleandri et al., 1989;Martin et al., 1997). Johnson et al. (2001) showed that a firmer curd at cutting resulted in reduced-fat Cheddar cheeses with higher moisture content. However, the milk clotting properties are variable, and factors influencing these properties include the concentrations of total CN and calcium (Storry et al., 1983), pH (Najera et al., 2003), genetic polymorphism of milk proteins (Schaar et al., 1985;Mayer et al., 1997;Ikonen et al., 1999a), stage of lactation (Okigbo et al., 1985c;Ostersen et al., 1997), season (O’Brien et al., 1999), and feeding (Verdier-Metz et al., 1998). Differences in milk clotting properties between different breeds have been previously studied in the United Kingdom (Verdier-Metz et al., 1998), France (Macheboeuf et al., 1993), Italy (Chiofalo et al., 2000), Ireland (Auldist et al., 2002), and New Zealand (Auldist et al., 2004).

Genetic polymorphism of milk proteins has been associated with composition, production traits, and technological properties of milk. Concentration of β-LG is higher in milk with the AA genotype than with AB or BB (McLean et al., 1984;Ng-Kwai-Hang et al., 1987;Graml et al., 1989), which results in a lower CN number in AA milk (Lunden et al., 1997;Schaar, 1984;van den Berg et al., 1992). Several workers have reported significantly higher concentrations of κ-CN in milk with the B allele (McLean et al., 1984;van den Berg et al., 1992;Bobe et al., 1999). The B allele of κ-CN has also been associated with improved milk clotting properties (Schaar, 1984) and a higher cheese yield (Schaar et al., 1985), whereas the E allele has been related to unfavorable milk clotting properties (Ikonen et al., 1999a;Caroli et al., 2000). Many studies have been carried out on the effects of genetic polymorphism on milk clotting properties and cheese yield (Ng-Kwai-Hang, 1998), whereas studies evaluating the actual milk protein composition in relation to cheese-making properties of individual milk samples are less frequent (Auldist et al., 2004). The CN to total protein ratio has decreased in Swedish bulk milk during the last decades, making it less suitable for cheese making (Lindmark-Månsson et al., 2003). Deteriorating trends like this may also occur in other countries, and therefore, it is important to identify suitable quality markers for cheese milk. In Sweden, milk is currently graded according to concentrations of total protein and milk fat. It could be an economical advantage to the dairy industry if a specific marker could be used to identify milk suitable for cheese making; that is, good milk clotting properties and high cheese yield. The object of this study was thus to evaluate the effect of protein composition on milk clotting properties and cheese yield from milk of individual cows.

Back to Article Outline

Materials and Methods 

Milk Collection and Experimental Design 

Evening whole milk samples (approximately 10L) from 16 Swedish Red and White (SRB), 15 Swedish Holstein (SLB), and 14 Danish Holstein-Friesian (SDM) cows were collected in September 2003. Milk from SRB and SLB were collected at the experimental dairy herd at Jälla (Swedish University of Agricultural Sciences) and milk from SDM was collected at the Research Centre Foulum (Danish Institute of Agricultural Sciences). A second sampling from 22 SRB, 8 SLB, and 15 SDM took place in January 2004, and the last sampling from an additional 20 SRB, 9 SLB, and 15 SDM was carried out in May 2004. A total of 134 cows was thus included in the study and no cow was sampled twice. Feeding and management was carried out according to standard practices in the respective countries and the cows were held indoors on the 3 sampling occasions. All cows included in the study were healthy and milked twice a day. The cows were grouped into 4 classes according to stage of lactation: early (wk 6 to 15; n = 27), mid (wk 16 to 30; n = 69), late (wk 31 to 45; n = 30), and very late (wk 46 or later; n = 8). All breeds were represented within each stage of lactation class. The cows were also grouped into 4 classes according to lactation number: first lactation (1; n = 50), second lactation (2; n = 42), third lactation (3; n = 22), and fourth or more lactation (4; n = 20). All breeds were represented within each lactation number class.

Milk Composition 

Samples of fresh milk were analyzed for concentration of milk fat, lactose, urea, and pH by a Milkoscan FT 6000 (A/S N., Foss Electric, Hillerød, Denmark) and for somatic cells using a Fossomatic 5000 (A/S N., Foss Electric) at Steins Laboratory (Holstebro, Denmark). Calcium content was determined using atom absorption spectrophotometry after drying to ash at 525°C for 6h, followed by dissolving in acid and dilution in lanthanchloride solution. Concentrations of major milk proteins were determined using the reversed phase liquid chromatography (RP-HPLC) method modified by Wedholm et al. (2006) according to Bordin et al. (2001). No measurement was made for αS2-CN. Concentration of total protein analyzed was defined as the sum of the concentrations of αS1-CN, β-CN, κ-CN, β-LG A, β-LG B, and α-LA. Concentration of total CN analyzed comprised the sum of the concentrations of αS1-CN, β-CN, and κ-CN, and concentration of total whey protein analyzed comprised the sum of the concentrations of β-LG A, β-LG B, and α-LA. Proteolysis in individual pasteurized skimmed milk samples were determined as the level of free amino terminals using the fluorescamine method modified for milk samples (Larsen et al., 2004). Level of free amino terminals was expressed as leucine equivalents (in mM) according to a standard curve for leucine.

Genotyping for κ- and β-CN Genetic Variants 

Blood samples were collected from SRB and SLB and the DNA was genotyped for genetic variants of β- and κ-CN using pyrosequencing, as described by E. Hallén, T. Allmere, J. Näslund, A. Andrén, and A. Lundén, Departments of Food Science and Animal Breeding and Genetics, Swedish Univ. Agric. Sci., Uppsala, Sweden; personal communication).

Cheese Making 

The model cheeses were produced from skimmed milk to reduce the number of variables influencing cheese yield. After 2 d of cold storage (4°C), individual milk samples were preheated to 40°C, defatted, and heated in a pilot plate heating apparatus (72°C for 15s), as described by Allmere et al. (1998). Four liters of skimmed milk was inoculated with a commercial starter culture (0.1g/L of Lactobacillus helveticus 174 and 0.1g/L of Probat 404, Danisco, Sweden), and incubated at 30°C for 30min. This was followed by addition of chymosin (1.25 mL/L of Chy-Max Plus, 190 International Milk Clotting Units/mL, Christian Hansen A/S, Denmark) and gentle stirring. After 30min at 30°C, the gel formed was cut into 2-cm cubes. To allow syneresis, the curd was incubated at 50°C for another 30min during gentle stirring. The whey was removed and the curd was pressed (0.04 kg/cm2) for 20h at room temperature. After 2 wk of storage at 10°C, the cheeses were weighed to obtain the yield. To obtain dry weight from individual cheeses, 2 to 3g was taken from the interior of the cheeses, grated, and mixed with a fixed amount of sand. Cheese samples were incubated at 105°C overnight and then placed in a desiccator for 1h before weighing. The cheese yield was expressed in 3 different ways: in relation to amount of cheese milk used (as g of cheese/100g of milk or as g of dry cheese solids/100g of milk) or as a transition number (as g of dry cheese solids/100g of milk protein), as recommended by Emmons (1993).

Rheological Measurements 

Immediately after rennet addition, a 12-mL sample was transferred to the C25 measuring cup of a Bohlin VOR Rheometer (Malvern Instruments, Nordic AB, Uppsala, Sweden). An oscillating technique was used with a frequency of 1Hz and the torsion bar 2 g·cm. Measurement temperature was 25°C. The elastic (storage) modulus, G′, was determined at a constant strain of 0.0412 and plotted against time to obtain a gelation profile of the curd. Measurements were carried out for 15min. The milk clotting time (MCT) was recorded as the time in minutes, from chymosin addition to the beginning of curd formation, to compare the clotting rate among the milk samples, and G′ (Pa) at 15min (G′15) was recorded to compare curd firmness.

Statistical Analysis 

Multivariate Regression 

Partial least square regression analysis (PLS1) was carried out using the software Unscrambler (version 9.0, CAMO ASA, Oslo, Norway). The x-variables consisted of the descriptive variables of breed (SRB, SLB, and SDM), stage of lactation (early, mid, late, and very late), lactation number (1, 2, 3, and 4,) and sample occasion (1, 2, and 3), and the continuous regressors of milk fat, total CN analyzed, κ-CN, αS1-CN, β-CN, β-LG A, β-LG B, total β-LG, α-LA, total whey protein analyzed, total protein analyzed (all g/100g of milk), total CN per total protein analyzed, κ-CN per total CN analyzed, αS1-CN per total CN analyzed, β-CN per total CN analyzed, β-LG A per total whey protein analyzed, β-LG B per total whey protein analyzed, α-LA per total whey protein analyzed, lactose (g/100g), urea (mM), SCC (log/mL), free amino terminals (mM of leucine), and calcium concentration (g/100g). The response variables (y-variables) analyzed were cheese yield expressed as grams of cheese per one hundred grams of milk, as grams of dry cheese solids per one hundred grams of milk, and as grams of dry cheese solids per one hundred grams of milk protein. Discriminant PLS1 was carried out to compare the milk composition of poorly coagulating and noncoagulating milk with normally coagulating milk (see definitions in first section of Results). Standardized (centered: μ = 0, and normalized: 1/SD) variables and full cross validation was used.

ANOVA 

The GLM procedure from SAS (Version 8e, SAS Institute Inc., Cary, NC) was used to calculate least squares means and standard errors and for pairwise testing of significant differences in concentration of κ-CN among the genotype classes κ-CN-AA, AB, and AE. The model contained the fixed effects of κ- and β-CN genetic variants, breed, sampling, stage of lactation and lactation number. Interactions between main effects were found to be insignificant and therefore were excluded from the statistical model. Concentration of total CN analyzed was included as covariate in the statistical model to adjust the main effects before ANOVA was accomplished. Significant differences between least squares means were evaluated based on F-values, using the option PDIFF. The GLM procedure was also used to test the overall differences in cheese yield between poorly coagulating and normally coagulating milk. The statistical model used included the fixed effect of coagulation class (0 or 1).

Back to Article Outline

Results 

Milk Clotting Properties and Cheese Yield 

The analyzed milk samples were grouped into 4 classes according to their clotting properties; that is, well-coagulating milk, milk with low clotting rate, poorly coagulating milk, and noncoagulating milk (Table 1). Milk samples that started coagulation within 15min and had firm curd at cutting (30min) were referred to as well-coagulating milk. Milk samples were assigned to the low clotting rate class if they did not start coagulation within 15min, but formed a firm curd at cutting. Poorly coagulating milk had a pronounced weaker curd at cutting and a G′15 value of less than 15Pa. Milk that did not start coagulation within 30min was referred to as noncoagulating milk. Table 2 shows means and standard errors of cheese yield obtained from poorly coagulating and well-coagulating milk. Average cheese yields were higher for cheeses prepared from normally coagulating than from poorly coagulating milk, irrespective of the cheese yield formula used, but the difference between coagulation groups was not statistically significant (Table 2). Coagulation groups tended to differ significantly, however, when yield was expressed as grams of cheese per one hundred grams of milk (P = 0.055).

Table 1. Classification of the milk samples according to their clotting properties expressed as the milk clotting time (MCT), elastic modulus at 15min (G′15), and curd firmness at 30min
MCTG′15 (Pa)Curd firmness at 30min1No. of samplesPercentage of all samples
Well-coagulating milk<15≥ 15Firm8161
Milk with low clotting rate15–300Firm118
Poorly coagulating milk<30<15Weak3828
Noncoagulating milk>300No curd43
Total134100

1Curd firmness at 30min was evaluated subjectively: firm = firm enough to form cubes; weak = not firm enough to form cubes.

Table 2. Cheese yield (expressed in 3 ways) obtained from poorly coagulating and well-coagulating milk samples
g of cheese/100g of milkg of dry cheese solids/100g of milkg of dry cheese solids/100g of milk protein
Poorly coagulating milk (n = 38)7.69±0.193.03±0.0792.44±1.61
Well-coagulating milk (n = 81)8.14±0.133.17±0.0593.20±1.14
Degrees of freedom111
P-value0.0550.1390.682

Effects of Analyzed Milk Variables on Milk Clotting Properties and Cheese Yield 

Means and standard deviations of analyzed milk proteins, milk fat (before skimming), calcium, free amino terminals, pH, SCC, G′15, MCT, and cheese yield are presented in Table 3 and 4. Results from the PLS1 analyses are presented in Figures 1 and 2. Concentrations of lactose, urea, free amino terminals, pH, and SCC did not significantly affect milk clotting properties or cheese yield and did not improve the multivariate regression model when included. These variables were therefore excluded from the final model. The cheese yield, expressed as grams of cheese per one hundred grams of milk (Figure 1a), was improved at increasing concentrations of milk fat (P<0.01), κ-CN (P<0.001), αS1-CN (P<0.001), β-CN (P<0.001), total CN analyzed (P<0.001), β-LG B (P<0.01), total β-LG (P<0.001), total protein analyzed (P<0.001), and total whey protein analyzed (P<0.001). The same list of variables was also important for the cheese yield when expressed as grams of dry cheese solids per one hundred grams of milk (results not shown). The significant effect of calcium concentration on cheese yield (Figure 1a,b) was very low. Cheese yield, in relation to exploitation of milk protein for cheese making (expressed as g of dry cheese solids/100g of milk protein), was improved at increasing concentrations of β-LG B (P<0.001), total CN per total protein analyzed (P<0.01), β-LG B per total whey protein analyzed (P<0.001), and α-LA per total whey protein analyzed (P<0.01). The β-LG A concentration and the β-LG A to total whey protein ratio analyzed were negatively associated with grams of dry cheese solids per one hundred grams of milk protein (P<0.01 and P<0.001, respectively). Cheese yield, expressed as grams of cheese per one hundred grams of milk (Figure 1a), was further positively associated with late lactation (P<0.01) and lactation number 1 (P<0.05), but negatively associated with early lactation (P<0.01) and lactation number 4 (P<0.01). Cheese yield was also positively associated with sampling occasion number 1 (in September; Figure 1a). No significant differences in cheese yield or milk clotting properties between the investigated breeds of SRB, SLB, and SDM were observed. Poorly coagulating and noncoagulating milk was associated with lower concentration of κ-CN (P<0.001) and κ-CN per total CN analyzed (P<0.001; Figure 2). Poorly coagulating and noncoagulating milk were most frequent within lactation number 1 (P<0.05).

Table 3. Means and SD of analyzed milk proteins,1 milk fat, lactose, urea, calcium, pH, free amino terminals, and SCC
MeanSDdfMinimumMaximum
Total protein analyzed23.390.341332.404.44
Total CN analyzed32.730.391331.833.74
CN/total protein analyzed40.810.031330.730.89
αS1-CN1.050.171330.661.42
β-CN1.250.211330.771.73
κ-CN0.440.101330.200.71
κ-CN/total CN analyzed0.160.031330.090.23
αS1-CN/total CN analyzed0.380.031330.260.44
β-CN/total CN analyzed0.460.031330.370.55
Total whey proteins analyzed50.620.141330.360.97
β-LG A0.290.231330.0060.82
β-LG B0.250.141330.0070.66
Total β-LG0.540.131330.280.89
α-LA0.080.021330.050.15
β-LG B/total whey proteins analyzed0.440.291330.0060.91
β-LG A/total whey proteins analyzed0.420.311330.0070.92
α-LA/total whey proteins analyzed0.140.041330.080.26
Milk fat (g/100g of milk)4.920.931333.088.30
Lactose (g/100g of milk)4.780.2312784.165.31
Urea (mM)4.961.451271.679.12
Calcium (g/100g of milk)0.120.0113180.090.16
pH6.700.061336.556.84
Free amino terminals (mM of leucine)1.010.261330.571.79
SCC (log/mL)4.990.3712684.306.00

1Concentration (g/100g of milk), in relation to total CN analyzed or total whey proteins analyzed.

2Concentration (g/100g of milk) of αS1-CN, β-CN, κ-CN, β-LG, and α-LA determined by reverse phase (RP)-HPLC.

3Concentration (g/100g of milk) of αS1-CN, β-CN, and κ-CN determined by RP-HPLC.

4The CN to total protein ratio determined from the RP-HPLC analysis.

5Concentration (g/100g milk) of β-LG and α-LA determined by RP-HPLC.

6Milk samples that lack the A-variant of β-LG.

7Milk samples that lack the B-variant of β-LG.

8Number of observations reduced due to missing data.

Table 4. Means and SD of milk clotting time (MCT) and elastic modulus at 15min (G′15), and cheese yield
MeanSDdfMinimumMaximum
Milk clotting properties
MCT (min)4.812.2310710.810.8
G′15 (Pa)42.0441.0511211225
Cheese yield
g of cheese/100g of milk8.001.1312425.2010.80
g of dry cheese solid/100g of milk3.120.4311422.104.20
g of dry cheese solid/100g of milk protein92.819.26114258.06129.68

1Number of observations reduced due to milk samples that did not start coagulation within 15min and to missing data.

2Number of observations was reduced due to noncoagulating milk (i.e., cheese yield was equal to zero) and due to missing data.

  • View full-size image.
  • Figure 1. 

    Results from the partial least squares regression analysis presented as the weighted regressions between x-variables of analyzed milk proteins (determined by HPLC), milk fat, calcium, breed, stage of lactation, lactation number, and sampling occasion, and the y-variables g of cheese/100g of milk (panel a) and g of dry cheese solids/100g of milk protein (panel b). Black bars represent significant regression coefficients. R2 is the correlation coefficient between measured and predicted y variables. All y-variables followed normal distributions. aMilk fat concentration before skimming; btotal concentration of αS1-CN, β-CN, and κ-CN; ctotal concentration of β-LG; dtotal concentration of αS1-CN, β-CN, κ-CN, β-LG, and α-LA; etotal concentration of β-LG and α-LA. *P<0.05; **P<0.01; ***P<0.001.

  • View full-size image.
  • Figure 2. 

    Results from the discriminant partial least squares regression analysis presented as weighted regressions between x-variables of analyzed milk proteins (determined by HPLC), milk fat, calcium, breed, stage of lactation, lactation number, and sampling occasion, and the y-variable of poorly coagulating and noncoagulating milk. R2 is the correlation coefficient between measured and predicted y-variables. The number of observations in the poorly coagulating and noncoagulating group was 42, whereas in the normally coagulating group, it was 81. aMilk fat concentration before skimming; btotal concentration of αS1-CN, β-CN, and κ-CN; ctotal concentration of β-LG; dtotal concentration of αS1-CN, β-CN, κ-CN, β-LG, and α-LA; etotal concentration of β-LG and α-LA. *P<0.05; **P<0.01; ***P<0.001.

The PLS1 model explained less than 50% of the total variation in milk clotting properties (Figure 2) and in cheese yield expressed as grams of dry cheese solids per one hundred grams of milk protein (Figure 1b). This was indicated by the correlation coefficient (R2) between measured and predicted y-variables, which were less than 0.50 (Figure 1b and 2). A larger amount of the total variation could be explained by the model when describing variations in cheese yield expressed as grams of cheese per one hundred grams of milk with R2 = 0.65 (Figure 1a).

Influence of κ-CN Genotype on κ-CN Concentration and Milk Clotting Properties 

Table 5 shows least squares means and standard errors of concentration of κ-CN in milk from SRB and SLB cows with the κ-CN AA, AB, and AE genotypes. Concentration of κ-CN was significantly higher in milk with the AB genotype compared with milk with the AA genotype (P<0.05). No difference in κ-CN concentration was found between the AB and AE genotypes or between the AA and AE genotypes. Furthermore, no effect of β-CN genotype on κ-CN concentration was found. Comparing poorly and noncoagulating milk with well-coagulating milk, the AE genotype was more frequent within the poorly and noncoagulating group (Table 5).

Table 5. Influence of κ-CN genotype (AA, AB, and AE) on concentration of κ-CN in milk samples from Swedish Red and White (SRB) and Swedish Holstein (SLB) cows (least squares means and SE) and frequency of genotypes within the poorly coagulating and noncoagulating group and within the well-coagulating group
κ-CN genotype1β-CN genotype2P-valueBreed (SRB or SLB) P-valueCovariateStage of lactation, lactation number, and sampling P-valuedf
AA (25 SRB, 24 SLB)AB (18 SRB, 4 SLB)AE (13 SRB, 3 SLB)Total CN analyzed3P-value
κ-CN (g/100g of milk)0.42±0.02a0.48±0.03b0.46±0.03ab0.35440.40550.0001>0.0515
Frequency of genotype within group4
Poorly and noncoagulating462133
Well-coagulating602911

a,bMeans with different superscripts differ (P<0.05).

1The κ-CN genotypes detected within the Swedish herd were AA, AB, AE, BB, and BE. Only one cow was genotyped BE and two cows genotyped BB. These were therefore excluded from the statistical analysis.

2β-CN genotypes detected within the Swedish herd were A1A1 (10 SRB and 3 SLB), A1A2 (29 SRB and 16 SLB), A2A2 (15 SRB and 10 SLB), and A2B (2 SRB and 2 SLB).

3Concentration of αS1-CN, β-CN, and κ-CN determined by reverse phase HPLC.

4Percentage of all cows genotyped AA, AB or AE.

Back to Article Outline

Discussion 

The results from this investigation showed the effect of variations in the milk protein composition, from individual cows, on milk clotting properties and yield of model cheeses. Studies on variation in cheese yielding capacity in relation to milk composition from a substantial number of individual milk samples are rare (Ng-Kwai-Hang et al., 1989), whereas studies on cheeses produced from pooled milk samples are more frequent (Mayer et al., 1997;Ikonen et al., 1999b;Auldist et al., 2004). There are also few reports on milk clotting properties recorded on a rheometer (Sbodio et al., 1997) because it has been more common to use a formagraph (Aleandri et al., 1989;Ikonen et al., 1999b;Auldist et al., 2004).

It has been shown that milk with impaired clotting properties was not improved by mixing it with an equal amount of well-coagulating milk (Okigbo et al., 1985b). The same authors showed that the milk clotting properties of poorly coagulating milk were improved by addition of calcium chloride and by reduction of milk pH. However, the clotting properties were not improved to the same level as in well-coagulating milk. The association we found between poorly coagulating milk and lactation number 1 (Figure 2) is in line with the results of Schaar (1984). He observed that milk samples from cows within lactation number 1 and 2 formed weaker curds than milk from cows in later lactations and suggested that this effect could be due to differences in composition and distribution of milk salts.

We found that the κ-CN concentration and its proportion in relation to αS1-CN and β-CN were significantly lower in poorly coagulating and noncoagulating milk. This finding is in line with the results of St-Gelais and Haché (2005), who showed poorer milk clotting properties of milk enriched with β-CN powder. They also found that cheeses produced from β-CN enriched milk were harder. The higher κ-CN concentration found in milk from cows with the κ-CN AB genotype compared with the AA genotype (McLean et al., 1984;van den Berg et al., 1992;Bobe et al., 1999; Table 5) and the positive association between κ-CN concentration and well-coagulating milk indicate that the AB genotype can influence milk clotting properties, indirectly, via the κ-CN concentration. However, the higher frequency of κ-CN AE in poorly and noncoagulating milk (Table 5) suggests that functionalities of κ-CN, other than solely the concentration, can contribute to the milk clotting properties. The high frequency of κ-CN AE observed in poorly coagulating milk (Table 5) also supports the significant correlation between the κ-CN E allele and poor milk clotting ability reported by Ikonen et al. (1999a).

Noncoagulating milk has been observed in several milk clotting studies (Okigbo et al., 1985a;Ikonen et al., 1999a;Nsofor, 2000). It has been suggested that noncoagulating milk could be due to genetic parameters because high heritability estimates for milk clotting properties have been observed (Ikonen et al., 1999a). We found that the average calcium concentration among noncoagulating milk samples was 0.10 g/100g of milk (results not shown); that is, below the average of all analyzed milk samples (0.12 g/100g; Table 3). It is thus possible that the lower calcium concentration in combination with a low concentration of κ-CN (Figure 2) was responsible for the noncoagulating milk samples in this study. However, only 4 noncoagulating milk samples were observed (Table 1), which is not enough observations for a statistical evaluation. Hence, it would be interesting to evaluate milk composition in a substantial number of noncoagulating milk samples.

According to the literature, associations between milk clotting properties and cheese yield vary. Aleandri et al. (1989) and Martin et al. (1997) suggested that a firm curd at cutting was associated with high cheese yield, whereas Ikonen et al. (1999b) did not find a significant difference in yield between cheeses made from poorly coagulating and well-coagulating milk. Riddell-Lawrence and Hicks (1989) reported that curd-healing time affected cheese yield more than coagulum strength at cutting. However, it has also been suggested that curd firmness at cutting is important for the sensory attributes of cheeses. Johnson et al. (2001) reported that an increase in cheese moisture content, due to firmer curd at cutting, resulted in softer and smoother texture of reduced-fat Cheddar cheeses. In the present study, the difference (although not significant) in cheese yield between the poorly coagulating and the well-coagulating milk was more pronounced when expressing yield as grams of cheese per one hundred grams of milk than as grams of dry cheese solids per one hundred grams of milk (Table 2). This suggests that an increase in cheese yield due to a firmer curd at cutting could be associated with an increase in the water-holding capacity of cheeses prepared from well-coagulating milk, as reported by Johnson et al. (2001).

The significantly positive correlation between β-LG B and cheese yield (Figure 1) was expected because it has been suggested that the B allele correlates with higher CN number (Schaar, 1984;van den Berg et al., 1992;Lunden et al., 1997) because of the lower concentration of β-LG in this milk (McLean et al., 1984;Ng-Kwai-Hang et al., 1987;Graml et al., 1989). In our study, the CN to total protein ratio analyzed did not affect the cheese yield expressed as grams of cheese per one hundred grams of milk (Figure 1a), whereas the yield expressed as grams of dry cheese solids per one hundred grams of milk protein was affected (Figure 1b). The latter expression would thus be the best to explain the effect of CN to total protein ratio analyzed on cheese making properties of milk. The milk fat was important for the cheese yield even though the milk was defatted before cheese making. This was probably indirectly due to the strong significant correlation (P<0.001; results not shown) between concentration of total protein and milk fat.

The small amount (less than 50%) of the total variation in milk clotting properties (Figure 2) and grams of dry cheese solids per one hundred grams of milk protein (Figure 1b) explained by milk protein composition implies that other variables not measured in our study contributed significantly to these traits. Auldist et al. (2004) evaluated the effect of milk composition on milk clotting properties from Friesian and Jersey dairy cows. They also found that milk protein composition only partly explained the total variation in milk clotting properties. However, the process from milk to cheese is very complex with several significant variables, which makes it difficult to find only one or a few markers for the cheese-making properties.

Back to Article Outline

Conclusions 

A surprisingly large fraction of the individual milk samples analyzed had poor clotting properties (about 30%), with a tendency of impaired cheese yield. A low concentration of κ-CN and a low proportion of κ-CN in relation to total CN analyzed were associated with poorly coagulating and noncoagulating milk. However, the analyzed milk protein composition only partly explained the total variation in milk clotting properties. Nevertheless, our results suggest that milk superior for cheese making is high in concentrations of αS1-, β-, and κ-CN, has high κ-CN in relation to total CN, and contains β-LG B. Furthermore, a high concentration of CN to total protein analyzed was significant for the transfer of proteins from milk to cheese.

Back to Article Outline

Acknowledgments 

The authors wish to thank PhD student Jessica Näslund, Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences for genotyping the genetic variants of β- and κ-CN. At the same university, Toomas Allmere, Department of Food Sciences, is acknowledged for planning and supervision of this project, and Roger Andersson, Department of Food Sciences, for good advice regarding the statistical analyses. Gudrun Franzén and Lena Hagenvall are acknowledged for the milk sample collection. At the Danish Institute of Agricultural Sciences, Hanne Damgaard Poulsen, Department of Animal Health, Welfare and Nutrition is acknowledged for performing the calcium analyses and Stina Greis Handberg and Helle Louise Christensen, Department of Food Quality, for excellent technical assistance. The Chy-Max Plus was a kind gift from Marianne Harboe, Christian Hansen A/S, Denmark. The financial support from the Swedish Farmers’ Foundation for Agricultural Research, Danish Dairy Research Foundation, and the Innovation Law is gratefully acknowledged.

Back to Article Outline

Supplementary data 

Download file

download text

Interpretive summary.

Back to Article Outline

References 

  1. Aleandri R, Schneider JS, Buttazzoni LG. Evaluation of milk for cheese production based on milk characteristics and formagraph measures. J. Dairy Sci. 1989;72:1967–1989
  2. Allmere T, Andrén A, Lunden A, Björck L. Interactions in heated skim milk between genetic variants of β-lactoglobulin and κ-casein. J. Agric. Food Chem. 1998;46:3004–3008
  3. Auldist MJ, Johnston KA, White NJ, Fitzsimons WP, Boland MJ. A comparison of the composition, coagulation characteristics and cheese making capacity of milk from Friesian and Jersey dairy cows. J. Dairy Res. 2004;71:51–57
  4. Auldist M, Mullins C, O’Brien B, O’Kennedy BT, Guinee T. Effect of cow breed on milk coagulation properties. Milchwissenschaft. 2002;57:140–143
  5. Bobe G, Beitz DC, Freeman AE, Lindberg GL. Effect of milk protein genotypes on milk protein composition and its genetic parameter estimates. J. Dairy Sci. 1999;82:2797–2804
  6. Bordin G, Raposo FC, Calle B, Rodriguez AR. Identification and quantification of major bovine milk proteins by liquid chromatography. J. Chromatogr. A. 2001;928:63–76
  7. Caroli A, Bolla P, Budelli E, Barbieri G, Leone P. Effect of κ-casein E allele on clotting aptitude of Italian Friesian milk. Zootec. Nutr. Anim. 2000;26:127–130
  8. Chiofalo V, Maldonato R, Martin B, Dupont D, Coulon JB. Chemical composition and coagulation properties of Modicana and Holstein cows’ milk. Ann. Zootech. 2000;49:497–503
  9. Emmons DB. Definition and expression of cheese yield. In:  Emmons DB editors. Factors Affecting the Yield of Cheese. Brussels, Belgium: International Dairy Federation; 1993;p. 12–20
  10. Graml R, Weiss G, Buchberger J, Pirchner F. Different rates of synthesis of whey-protein and casein by alleles of the β-lactoglobulin and αS1-casein locus in cattle. Genet. Sel. Evol. 1989;21:547–554
  11. Ikonen T, Ahlfors K, Kempe R, Ojala M, Ruottinen O. Genetic parameters for the milk coagulation properties and prevalence of non-coagulating milk in Finnish dairy cows. J. Dairy Sci. 1999;82:205–214
  12. Ikonen T, Ruottinen O, Syvaoja EL, Saarinen K, Pahkala E, Ojala M. Effect of milk coagulation properties of herd bulk milks on yield and composition of Emmental cheese. Agric. Food Sci. Finland. 1999;8:411–422
  13. Johnson ME, Chen CMJ, Jaeggi JJ. Effect of rennet coagulation time on composition, yield and quality of reduced-fat Cheddar cheese. J. Dairy Sci. 2001;84:1027–1033
  14. Larsen LB, Rasmussen MD, Bjerring M, Nielsen JH. Proteases and protein degradation in milk from cows infected with Streptococcus uberis. Int. Dairy J. 2004;14:899–907
  15. Lindmark-Månsson H, Fonden R, Pettersson HE. Composition of Swedish dairy milk. Int. Dairy J. 2003;13:409–425
  16. Lunden A, Nilsson M, Janson L. Marked effect of β-lactoglobulin polymorphism on the ratio of casein to total protein in milk. J. Dairy Sci. 1997;80:2996–3005
  17. Macheboeuf D, Coulon JB, D’Hour P. Effect of breed, protein genetic variants and feeding on cows’ milk coagulation properties. J. Dairy Res. 1993;60:43–54
  18. Martin B, Chamba JF, Coulon JB, Perreard E. Effect of milk chemical composition and clotting characteristics on chemical and sensory properties of Reblochon de Savoie cheese. J. Dairy Res. 1997;64:157–162
  19. Mayer HK, Ortner M, Tschager E, Ginzinger W. Composite milk protein phenotypes in relation to composition and cheese making properties of milk. Int. Dairy J. 1997;7:305–310
  20. McLean DM, Bruce Graham ER, Ponzoni RW. Effects of milk protein genetic variants on milk yield and composition. J. Dairy Res. 1984;51:531–546
  21. Najera AI, de Renobales M, Barron LJR. Effects of pH, temperature, CaCl2 and enzyme concentrations on the rennet-clotting properties of milk: A multifactorial study. Food Chem. 2003;80:345–352
  22. Ng-Kwai-Hang KF. Genetic polymorphism of milk proteins: Relationships with production traits, milk composition and technological properties. Can. J. Anim. Sci. 1998;78:131–147
  23. Ng-Kwai-Hang KF, Hayes JF, Moxley JE. Variation in milk protein concentrations associated with genetic polymorphism and environmental factors. J. Dairy Sci. 1987;70:563–570
  24. Ng-Kwai-Hang KF, Politis I, Cue IR, Marziali AS. Correlations between coagulation properties of milk and cheese yielding capacity and cheese composition. Can. Inst. Food Sci. Technol. J. 1989;22:291–294
  25. Nsofor LM. Curd-forming casein concentration in poor/non-chymosin coagulating milk from individual cows. J. Food Sci. Technol. 2000;37:153–157
  26. O’Brien B, Mehra R, Connolly JF, Harrington D. Seasonal variation in the composition of Irish manufacturing and retail milks 1. Chemical composition and renneting properties. Irish J. Agric. Food Res. 1999;38:53–64
  27. Okigbo LM, Richardson GH, Brown RJ, Ernstrom CA. Casein composition of cow's milk of different chymosin coagulation properties. J. Dairy Sci. 1985;68:1887–1892
  28. Okigbo LM, Richardson GH, Brown RJ, Ernstrom CA. Effects of pH, calcium chloride and chymosin concentration on coagulation properties of abnormal and normal milk. J. Dairy Sci. 1985;68:2527–2533
  29. Okigbo LM, Richardson GH, Brown RJ, Ernstrom CA. Variation in coagulation properties of milk from individual cows. J. Dairy Sci. 1985;68:822–828
  30. Ostersen S, Foldager J, Hermansen JE. Effects of stage of lactation, milk protein genotype and body condition at calving on protein composition and renneting properties of bovine milk. J. Dairy Res. 1997;64:207–219
  31. Riddell-Lawrence S, Hicks CL. Effect of curd firmness on stirred curd cheese yield. J. Dairy Sci. 1989;72:313–321
  32. Sbodio OA, Tercero EJ, Coutaz VR, Luna JA, Martinez E. Simultaneous interaction of pH, CaCl2 addition, temperature and enzyme concentration on milk coagulation properties. Food Sci. Technol. Int. 1997;3:291–298
  33. Schaar J. Effects of κ-casein genetic variants and lactation number on the renneting properties of individual milks. J. Dairy Res. 1984;51:397–406
  34. Schaar J, Hansson B, Pettersson HE. Effects of genetic variants of κ-casein and β-lactoglobulin on cheesemaking. J. Dairy Res. 1985;52:429–437
  35. St-Gelais D, Haché S. Effect of β-casein concentration in cheese milk on rennet coagulation properties, cheese composition and cheese ripening. Food Res. Int. 2005;38:523–531
  36. Storry EJ, Grandison AS, Millard D, Owen AJ, Ford GD. Chemical composition and coagulating properties of renneted milks from different breeds and species of ruminant. J. Dairy Res. 1983;50:215–229
  37. van den Berg G, Escher JTM, de Koning PJ, Bovenhuis H. Genetic polymorphism of κ-casein and β-lactoglobulin in relation to milk composition and processing properties. Neth. Milk Dairy J. 1992;46:145–168
  38. Verdier-Metz I, Coulon JB, Pradel P, Viallon C, Berdague JL. Effect of forage conservation (hay or silage) and cow breed on the coagulation properties of milks and on the characteristics of ripened cheeses. J. Dairy Res. 1998;65:9–21
  39. Wedholm A, Hallén E, Larsen LB, Lindmark-Månsson H, Karlsson AH, Allmere T. Comparison of milk protein composition in a Swedish and a Danish dairy herd using reversed-phase HPLC. Acta Agric. Scand., A. Anim. Sci. 2006;(accepted)

PII: S0022-0302(06)72366-9

doi:10.3168/jds.S0022-0302(06)72366-9

Journal of Dairy Science
Volume 89, Issue 9 , Pages 3296-3305, September 2006

Article Tools

* Image Usage & Resolution