J. Dairy Sci.

96:7490–7499
http://dx.doi.org/10.3168/jds.2013-7119
© American Dairy Science Association®, 2013. Open access under CC BY-NC-ND license.

Yield, changes in proteolysis, and sensory quality of Prato cheese

produced with different coagulants
L. S. Alves,* C. Merheb-Dini,*1 E. Gomes,† R. da Silva,† and M. L. Gigante*
*Faculty of Food Engineering, University of Campinas – UNICAMP, PO Box 6121, CEP 13083-970, Campinas, SP, Brazil
†Laboratory of Biochemistry and Applied Microbiology – Instituto de Biociências, Letras e Ciências Exatas (IBILCE) – Universidade Estadual
Paulista Júlio de Mesquita Filho (UNESP), Rua Cristóvão Colombo 2265, CEP 15054-000, São José do Rio Preto, SP, Brazil

ABSTRACT through limited proteolysis by selected proteinases, fol-

lowing micelle aggregation in the presence of calcium.
The objective of this research was to compare the Rennet, made up mostly of chymosin (EC 3.4.23.4) is
effect of 2 fungal proteases, one that is already com- the most-used protease for cheese manufacture (Kloos-
mercially established as a milk-clotting agent and terman, 1991). During enzymatic coagulation of milk,
another produced at the laboratory scale, on Prato κ-CN is the only protein hydrolyzed by chymosin, spe-
cheese composition, protein and fat recovery, yield, cifically at the bond Phe105-Met106, destroying micelle
and sensory characteristics. Cheeses were produced ac- stability (Fox et al., 2000). Afterward, the N-terminal
cording to the traditional protocol, using protease from part of the molecule, κ-CN fragment 1 to 105 (f1–105),
the fungus Thermomucor indicae-seudaticae N31 and known as para-κ-CN, remains attached to the micelle,
commercial coagulant from Rhizomucor spp. as clotting whereas the C-terminal part, known as glycomacrope-
agents. A 2 × 6 factorial design with 3 replications was ptide (f106–169), is lost to the aqueous medium (Fox
performed: 2 levels of coagulants and 6 levels of storage et al., 2000). The physicochemical modifications of the
time. After 5, 12, 19, 33, 43, and 53 d of refrigerated micelles continue after the first hydrolysis and on the
storage (12°C), cheeses were monitored for proteolysis, secondary phase of coagulation, the para-κ-CN micelles,
firmness, and casein degradation by capillary electro- in the presence of Ca2+ and at temperatures above
phoresis. Sensory acceptance was evaluated after 29 d 20°C, aggregate, forming a protein matrix referred to
of manufacturing. The different coagulants did not sta- as coagulum or gel (Fox et al., 2000). This coagulum is
tistically affect Prato cheese composition, protein and then treated, and the different treatments lead to the
fat recovery, and yield. Both cheeses presented good manufacture of the various types of cheeses. Chymosin
sensory acceptance. Proteolysis increased and firmness substitutes must reproduce its specific properties; that
decreased for both cheeses during the storage time, as is, exhibit high milk-clotting activity, which consists of
expected for Prato cheese. Caseins were well separated having specificity for κ-CN and low proteolytic activity
by capillary electrophoresis and the results showed, at pH and temperature values normally used in cheese
with good resolution, that the cheeses exhibited similar making (Fox et al., 2000). Currently, fermentation-
protein hydrolysis profile. Both cheeses presented good produced chymosin, highly purified chymosin obtained
sensory acceptance. The gathered data showed that using a host organism, and coagulants obtained from
the protease from T. indicae-seudaticae N31 presented microorganisms such as Rhizomucor miehei, Rhizomu-
similar action compared with the commercial enzyme, cor pusillus, Endothia parasitica, Aspergillus oryzae, and
indicating its efficiency as clotting agent for Prato Irpex lacteus, are extensively used for cheese manufac-
cheese manufacture. ture and among the microbial coagulants commercially
Key words: fungal enzyme, yield, sensory acceptance, available, the ones from fungal sources are the most
capillary electrophoresis used (Jacob et al., 2011).
The global enzyme market is expected to generate
INTRODUCTION sales of $8 billion in 2015, with $1.3 billion representing
the food and beverages segment and the highest sales
Enzymatic coagulation of milk for cheese manufac- occurring in the dairy market (Li et al., 2012). Boom-
ture involves specific modifications of the casein micelle ing of the market and its growing potential certainly
stimulates research for new microbial proteases to be
used as rennet substitutes, including the studies carried
Received June 11, 2013.
Accepted September 2, 2013. out with the following microorganisms: Nocardiopsis
1
Corresponding author: [email protected] spp. (Cavalcanti et al., 2005), Rhizopus oryzae (Kumar

7490
 PRATO CHEESE PRODUCED WITH DIFFERENT COAGULANTS 7491

et al., 2005), Mucor bacilliformis (Machalinski et al., MATERIALS AND METHODS

2006), Bacillus subtilis natto (Shieh et al., 2009), and
Aspergillus oryzae MTCC 5341 (Vishwanatha et al., Enzyme Production
2010).
With the increased search for new coagulants, a novel The fungus T. indicae-seudaticae N31 was obtained
source is reported from time to time. Recently, Mer- from the collection of the Laboratory of Biochemistry
heb-Dini et al. (2010) presented a coagulant protease and Applied Microbiology, Instituto de Biociências,
obtained from the thermophilic fungus Thermomucor Letras e Ciências Exatas, Universidade Estadual Pau-
indicae-seudaticae N31 isolated in Brazil from indus- lista Júlio de Mesquita Filho (IBILCE-UNESP, São
trial waste piles (Martin et al., 2010). The enzymatic José do Rio Preto, SP, Brazil). The steps of fungal
extract was produced by solid-state fermentation using growth, inoculum preparation and protease production
only wheat bran as substrate. In 24 h of fermentation, followed the methodology described by Merheb-Dini
the microorganism secreted the extracellular enzyme, et al. (2010), with modifications to achieve maximum
which exhibited maximum milk-clotting activity at pH enzyme yield and concentration under laboratory con-
5.7 and at 70°C and stability in the pH range from ditions. Culture medium containing 20 g of wheat bran
3.5 to 4.5 and from 5.0 to 6.0 and temperature range was prepared and sterilized (120°C/20 min) in 500-mL
from 35 to 40–45°C. In addition, it was shown that Erlenmeyer flasks, inoculated with 27 mL mycelial sus-
the peptide profile formed by the enzymes from T. pension to give a 60% initial moisture, and incubated at
indicae-seudaticae N31 and from R. miehei (Hannilase; 45°C for 24 h under stationary conditions. For enzyme
Chr. Hansen Brazil, Valinhos, SP, Brazil), a widely extraction, 160 mL of distilled water was added to the
used milk-clotting agent for cheese manufacture, was medium, the flasks were shaken at 100 rpm for 30 min,
very similar. These results encouraged laboratory-scale and the mixture was filtered and centrifuged at 30,996
research using protease from T. indicae-seudaticae N31 × g for 20 min at 5°C. The resulting solution, denomi-
for the manufacture of Prato cheese compared with a nated crude enzyme extract, was filtered on Whatman
fermentation-produced chymosin, HA-LA, produced grade no. 1 filter paper, and concentrated by UF (Quix-
by Chr. Hansen Brazil (Merheb-Dini et al., 2012). The Stand System; GE Healthcare Life Sciences, São Paulo,
results suggested the possibility of obtaining a product SP, Brazil).
with good technological quality.
In view of the previous results regarding the discov- Cheese Manufacture
ery of an enzyme produced rapidly by using a low-cost
substrate, with promising biochemical properties for The cheeses were produced in vats with heating-
use in cheese manufacture, the objective of this study cooling jackets, stirrers, and speed control by the tra-
was to produce pilot-scale Prato cheese with the new ditional manufacturing method as described by Mazal
enzyme and compare it with a well-established com- et al. (2007). For each process, 100 L of heat-treated
mercial coagulant of fungal source, commonly used whole milk (68°C for 2 min) were divided into 2 equal
for Prato cheese manufacture. For the most complete parts to be used for Prato cheese manufacture. The
characterization, cheese and whey composition were following treatments were carried out: (1) control Prato
evaluated, protein and fat recovery and adjusted cheese cheese using commercial microbial coagulant produced
yield were determined, sensory analysis was carried out, by Rhizomucor spp. (Bela Vista Produtos Enzimáticos,
and proteolysis during the storage time was investi- Alto Bela Vista, SP, Brazil) and (2) Thermomucor Prato
gated by capillary electrophoresis, a modern technique cheese produced with the enzyme from T. indicae-seu-
more efficient for casein separation than conventional daticae N31. For both treatments, 50% calcium chloride
gel electrophoresis and HPLC. Prato is a semihard solution (250 mg/kg) was added to the milk at 35°C,
(moisture content 36–45.9%), low-scald (≈42°C) cheese which was followed by the addition of 1% mesophilic
manufactured by enzymatic coagulation of milk and culture (Lactococcus lactis ssp. lactis and Lactococcus
ripened for at least 25 d (Ministério da Agricultura lactis ssp. cremoris; R704; Chr. Hansen Brazil), annatto
do Brasil, 1997), with soft texture and smooth taste, dye (80 mg/kg), and coagulant. Control Prato cheese
and therefore a good tool to evaluate enzyme action as was produced using 35 min of clotting time, according
both milk-clotting agent and proteolytic agent during to the manufacturer’s instructions. Thermomucor Prato
ripening. This research should be of great contribution cheese was produced using 55 min of clotting time after
to the fields of biotechnology and dairy technology, as standardizing the firmness at cutting of the curd to
it deals with the application of a new microbial enzyme that of the commercial coagulant from Rhizomucor spp.
for cheese making. (Lucey and Kelly, 1994). The curd firmness was evalu-

Journal of Dairy Science Vol. 96 No. 12, 2013

7492 ALVES ET AL.

ated by inserting a sanitized spatula into the coagulum Cheese Characterization During Storage
at a 45° angle, gently lifting the spatula straight up,
and observing the curd as it split open (Mazal et al., The cheeses were monitored for pH, moisture con-
2007). After coagulation, the curd was cut into 1.0-cm tent, proteolysis, electrophoretic profile, and firmness
cubes and submitted to slow continuous mixing for 15 after 5, 12, 19, 33, 43, and 53 d of refrigerated storage
min, which was followed by removal of part (30%) of (12°C). The pH was measured as described above, and
the whey and further heating of the curd to 42°C by moisture content was determined according to AOAC
adding hot water (80°C) to increase the temperature International (2006; methods 33.7.03 and 926.08). For
by 1°C every 3 min. After cooking, all the whey was proteolysis measurements, pH 4.6 SN and 12% TCA
drained off, and the curd was placed in rectangular SN were determined by the macro-Kjeldahl method ac-
plastic molds (0.5 kg) and pressed at room temperature cording to Bynum and Barbano (1985), and expressed
with successive turns (0.1 MPa/15 min; 0.1 MPa/15 as a percentage of the TN of the cheese.
min; 0.24 MPa/30 min; and 0.31 MPa/90 min). Cheeses The electrophoretic profile was determined by capil-
were fermented for 5 h at room temperature and salted lary electrophoresis in a fused-silica capillary column of
in brine (20%) for 10 h at 5°C. Then, the cheeses were 57 cm (50-cm effective length to detector) × 75 μm, us-
dried at 12°C for 48 h and vacuum-packed into heat- ing a P/ACE MDQ system (Beckman Coulter, Santana
shrinkable plastic bags and stored at 12°C for 53 d. The de Parnaiba, SP, Brazil) and 32 Karat software (Beck-
experiment was repeated 3 times on different days and man Coulter), according to the conditions described by
Thermomucor and control Prato cheeses were produced Ortega et al. (2003) and Otte et al. (1997) with modifi-
in side-by-side vats. cations. To dissociate the casein, 20 mg of cheese were
dissolved into 1 mL of 10 mM sodium phosphate buffer
Sampling and Analysis containing 8 M urea and 10 mM dithiothreitol (DTT;
pH 8.0), and left for at least 1 h at room temperature.
Heat-treated milk was analyzed for the activity of The samples were filtered (0.45 μm) and injected for 5
alkaline phosphatase (EC 3.1.3.1) to check the effi- s at 3.45 ×10−3 Pa. Separation was carried out at 18.5
ciency of the heat treatment as described by Marshall kV and 23°C, and detection was performed at 214 nm
(1992). The milk was evaluated for pH, fat by the for 60 min. Between runs, the capillary was conditioned
Gerber method (British Standards Institution, 1989), by washing first with 0.5 M NaOH for 5 min, then with
TS (AOAC International, 2006; methods 33.2.44 and pure water for 5 min, and finally with running buffer
990.20), ash content (AOAC International, 2006; meth- [10 mM sodium phosphate containing 6 M urea and
ods 33.2.10 and 945.46), total nitrogen (TN; AOAC 0.05% hydroxypropyl methylcellulose (HPMC; pH 3.0)]
International, 2006; methods 33.2.11 and 991.21), pH for 5 min. Sigma-Aldrich Co., St. Louis, MO standards
4.6-soluble nitrogen (pH 4.6 SN; AOAC International, for α-CN, β-CN, κ-CN (10 mg/mL) were used for peak
2006; methods 33.2.64 and 998.05), and 12% TCA- identification.
soluble nitrogen (12% TCA SN; AOAC International, Firmness was measured using the texture profile
2006; methods 33.2.12 and 991.21). Total protein was analysis defined by Bourne (2002), using a TA XT2
calculated by multiplying TN by 6.38. Texture Analyzer (Stable Micro Systems Ltd., Godalm-
For each vat, all the whey was collected, mixed, and ing, Surrey, UK) with a 35-mm-diameter aluminum
sampled. The whey was analyzed for fat content by the probe. Sampling and sample treatment were according
Mojonnier method (AOAC International, 1995; meth- to Mazal et al. (2007) as follows: 8 cylindrical samples
ods 33.2.26 and 989.05), TS, ash, TN, pH 4.6 SN, and of 2-cm diameter and 2.4-cm height were sampled from
12% TCA SN, using the same methodologies previously the entire length of the cheeses with the aid of an alu-
mentioned. minum probe. The cylinders were wrapped in plastic
Five days after manufacture, the pH of the cheeses was film, packed in plastic bags, and kept in an ice-water
determined by introducing the electrode directly into bath (10°C) for at least 4 h for temperature stabili-
the triturated samples. The samples were analyzed for zation. The test was performed at speed of 100 mm/
moisture content (AOAC International, 2006; methods min and a 40% compression of the initial height, and
33.7.03 and 926.08), fat by the Gerber method (British repeated within 5 s.
Standards Institution, 1989), ash (AOAC International,
2006; methods 33.7.07 and 935.42), salt by the Volhard
method (Richardson, 1985), TN (AOAC International, Recovery of Milk Components and Cheese Yield
2006; methods 33.7.12A and 2001.14), pH 4.6 SN, and
12% TCA SN by the macro-Kjeldahl method according The recovery of milk components and the adjusted
to Bynum and Barbano (1985). yield were calculated as described by Mazal et al.
Journal of Dairy Science Vol. 96 No. 12, 2013
 PRATO CHEESE PRODUCED WITH DIFFERENT COAGULANTS 7493

(2007). Protein and fat recovery (%R) were calculated ated the cheeses regarding the parameters appearance,
according to Equation 1: aroma, flavor, texture, and overall impression using a
9-point hedonic scale (1 = extremely disliked, 7 = liked
(m j × cij ) × 100 moderately, and 9 = liked extremely; Stone and Sidel,
%Rij = , [1] 1993). Results for overall impression were expressed as
m milk × cimilk
a percentage that represented the relationship between
the number of responses regarding the acceptance (6–
where i is the milk component (fat or protein), j is the 9), indifference (5), or rejection scores (1–4) to the total
sample (cheese or whey), m is the mass of the sample number of assessors. Purchase intent was evaluated on
(g), cij is the percentage of i in sample j, mmilk is the a 5-point hedonic scale (1 = certainly would not buy
mass of milk (g), and cimilk is the percentage of i in the this product; 5 = certainly would buy this product) and
milk. the results were expressed as a percentage between the
The adjusted yield (Yadj), which considers the de- number of responses regarding the acceptance (5 and
sirable salt and moisture contents of the cheeses, was 4), indifference (3), or rejection scores (1 and 2) to the
calculated according to Equation 2: total number of assessors. The results were submitted
to ANOVA, considering samples and assessors as varia-
Yadj =
tion causes at a 5% significance level.
Yact × [100 − (% real moisture content + % real salt content)]× 100
0
,
[100 − (% desirable moisture content + % desirable salt content)]
[2] RESULTS AND DISCUSSION

Characterization of the Raw Material

where Yact is the actual cheese yield [(g of cheese ×
100)/(g of milk)]; desirable salt content for Prato cheese The milk used in cheese manufacture met the stan-
was considered 1.6% and desirable moisture content for dards required by the Brazilian legislation (Ministério
Prato cheese was considered 42%. da Agricultura do Brasil, 2011) regarding the physico-
chemical parameters: minimum of 3% fat, acidity be-
Experimental Design and Statistical Analysis tween 14 and 18°D, relative density (15°C) from 1.028
to 1.034 g/mL, and a minimum of 2.9% total protein.
To assess the effect of the coagulant on cheese com- The heat treatment applied to milk was adequate, re-
position, the experiments were conducted in a complete sulting in negative alkaline phosphatase activity.
randomized block design with a 2 × 6 factorial arrange-
ment. The factor coagulant had 2 levels (commercial Effect of Coagulants on the
coagulant and coagulant produced at a laboratory scale Manufacture of Prato Cheese
by T. indicae-seudaticae N31), and the factor storage
time had 6 levels (5, 12, 19, 33, 43, and 53 d after Figure 1 shows the protein profile of milk used in the
manufacture). The blocks were the manufacture repli- cheese manufacture, which exhibits the casein fractions
cations (processes 1, 2, and 3 for each cheese). Analysis αs2-CN, αs1-CN 8P, αs1-CN 9P, κ-CN, and β-CN with
of variance was performed to assess the effect of treat- its genetic variants A1 and A2 (Ortega et al. 2003).
ments on the composition of cheese and whey, protein These protein fractions were also observed in the elec-
and fat recovery, and cheese yield, regarding only the trophoretic profile of the control and Thermomucor
effect of the coagulant. To evaluate cheese character- Prato cheese after manufacture, with the exception of
istics during storage, the independent effects of the κ-CN, which was hydrolyzed, resulting in the formation
coagulant and storage time were assessed, as well as of para-κ-CN, due to the action of the coagulants dur-
the interaction between these factors. The Tukey test ing the manufacturing process. After 5 d of manufac-
was used for comparison between means, considering ture (Table 1), both cheeses met the standards required
the 5% significance level using the software Statistica by the Brazilian legislation, concerning fat in DM
7.0 (StatSoft Inc., Tulsa, OK). (45–59.9%) and moisture content (36–45.9%; Ministé-
rio da Agricultura do Brasil, 1997), and presented a
Sensory Analysis composition similar to the results reported by Mazal et
al. (2007). The coagulants did not significantly affect
To abide by the Brazilian legislation that specifies cheese composition, except for total protein content,
the ripening period for Prato cheese to be at least 25 which was higher in Thermomucor Prato cheese (P <
d, the sensory acceptance test was performed 29 d after 0.05). However, when expressed on a dry basis, the
manufacturing with 100 untrained assessors, who evalu- protein content did not differ between treatments, indi-
Journal of Dairy Science Vol. 96 No. 12, 2013
7494 ALVES ET AL.

Table 1. Effect of treatments1 on the composition, firmness, protein and fat recovery, and adjusted yield of
control and Thermomucor Prato cheese after 5 d of manufacture (means ± SD; n = 3)

Control Prato Thermomucor

Item cheese Prato cheese P-value
pH 5.13 ± 0.11 5.08 ± 0.09 0.60
Moisture (%) 44.0 ± 1.54 41.36 ± 1.47 0.098
Fat (%) 27.83 ± 2.60 29.33 ± 2.43 0.50
FDB2 (%) 49.67 ± 3.80 49.98 ± 2.87 0.92
Total protein (total N × 6.38) (%) 22.97 ± 0.32 24.62 ± 0.41 0.0051
PDB3 (%) 41.03 ± 0.70 42 ± 1.02 0.24
Ash (%) 4.03 ± 0.38 4.08 ± 0.12 0.83
Salt (%) 1.93 ± 0.24 1.90 ± 0.04 0.86
S/M4 (%) 4.38 ± 0.38 4.63 ± 0.29 0.43
Firmness (g) 1,928 ± 390.10 2,496 ± 333.52 0.13
Protein recovery (%) 76.21 ± 5.66 77.02 ± 5.53 0.87
Fat recovery (%) 81.09 ± 6.63 80.62 ± 6.02 0.93
Adjusted yield (%) 9.57 ± 0.34 9.48 ± 0.40 0.52
1
Treatments were Prato cheese made with coagulant from Thermomucor indicae-seudaticae N31 and with com-
mercial coagulant.
2
Fat on dry basis.
3
Protein on dry basis.
4
Salt in moisture.

cating that the difference found in the protein contents Prato cheeses have technological implications being
was due to the nonsignificant difference of 3% moisture very important from the standpoint of cheese yield and
between the control Prato cheese (44.0 ± 1.54%) and quality, as discussed further on.
Thermomucor Prato cheese (41.36 ± 1.47%). Despite Cheese manufacture may be considered as a dehydra-
not being statistically significant, the 3% difference in tion process in which the casein and fat of milk are con-
moisture content between Thermomucor and control centrated (Lucey and Kelly, 1994) and, therefore, pro-

Figure 1. Capillary electrophoresis curves of proteins present in milk and in Prato cheese after manufacture. Labeling of peaks is according
to Ortega et al. (2003) and Otte et al. (1997). AU = arbitrary units.

Journal of Dairy Science Vol. 96 No. 12, 2013

 PRATO CHEESE PRODUCED WITH DIFFERENT COAGULANTS 7495

tein and fat recoveries and cheese yield are important the pH behavior during cheese manufacture influences
tools to evaluate the potential use of a new coagulating the characteristics of the final product once it favors
enzyme, as the one used in this study. The coagulants protein contraction and, consequently, syneresis (Fox
did not significantly affect either milk protein and fat et al., 2000). Therefore, the greater pH decrease during
recoveries in the cheese or the adjusted cheese yield Thermomucor Prato cheese production may have con-
(Table 1). Control and Thermomucor Prato cheeses ex- tributed to the lower moisture content of this cheese,
hibited 76.21 ± 5.66 and 77.02 ± 5.53% protein recov- which in turn possibly contributed to its proteolysis
ery, respectively, and 81.09 ± 6.63 and 80.62 ± 6.02% being lower and firmness being higher during the stor-
fat recovery, respectively. These values are slightly age time (Table 3).
higher than the ones for Prato cheese made with chy- No significant interaction was observed between
mosin, with 73% protein recovery and 78% fat recovery, treatments and storage time for any of the evaluated
as reported by Mazal et al. (2007). The adjusted yields parameters (Table 3) and both cheeses presented the
for control and Thermomucor Prato cheeses were 9.57 same behavior during storage, with increased protein
± 0.34 and 9.48 ± 0.40%, respectively, and were very fractions [pH 4.6 SN and 12% TCA SN (%TN)] and
similar to the 9.3% adjusted cheese yield of the Prato reduced firmness (Table 3 and Figure 2). Although in
cheese studied by Mazal et al. (2007). The different the model system with sodium caseinate the pattern
coagulants did not influence the protein and fat loss of peptide fragments formed by the action of protease
to the whey (Table 2), indicating a similar action dur- from T. indicae-seudaticae N31 and from R. miehei
ing Prato cheese manufacture. Values for protein and (Hannilase) showed similarities (Merheb-Dini et al.,
fat recoveries in whey (Table 2) complement the ones 2010), the control Prato cheese produced with protease
obtained in cheese (Table 1). These results are of great from Rhizomucor spp. in the current study presented
economic importance for the protease from T. indicae- higher values of pH 4.6 SN (%TN; 9.75 ± 3.81%) than
seudaticae N31. the Thermomucor Prato cheese (7.81 ± 3.34%), suggest-
ing higher proteolytic activity during the storage time.
Effect of Coagulant on Ripening of Prato Cheese Proteolysis assessed by pH 4.6 SN (%TN) is primarily
related to natural proteinases and milk-clotting agents,
Table 3 shows the effects of treatment, storage time, which degrade protein into higher-molecular-weight
as well as the interaction of these factors on pH, mois- peptides (Fox, 1989). The higher proteolytic activity
ture content, proteolysis, and firmness of the cheeses. of the commercial coagulant, besides being related to
The treatments, storage time, and the interaction of the higher moisture content, may also be due to higher
these factors did not significantly affect the pH of the amount of coagulant retained in the curd. Protease
samples, which was, on average, 5.09 ± 0.03 and 5.10 from R. miehei is characterized by its relatively high
± 0.02 for control and Thermomucor Prato cheeses, heat stability (up to 60°C; Walsh and Li, 2000), which
respectively. The moisture content was affected by the may have contributed to its higher residual activity af-
treatments, being lower for Thermomucor Prato cheese ter the heat treatment of the curd (42°C), leading to an
(41.07 ± 0.47%) than for the control Prato cheese (43.26 increased cleavage of caseins during storage. Protease
± 0.55%). This difference in moisture content is prob- from T. indicae-seudaticae N31 exhibits lower heat sta-
ably due to the pH decrease during cheese manufacture, bility, up to 40 to 45°C, (Merheb-Dini et al., 2010) and
which was higher for Thermomucor Prato cheese (0.59) probably underwent higher denaturation during the
than for control Prato cheese (0.26). It is known that heat treatment of the curd. The observed differences in

Table 2. Effect of treatments1 on the composition and protein and fat recovery of control and Thermomucor
Prato cheese whey (means ± SD; n = 3)

Control Prato Thermomucor Prato

Item cheese whey cheese whey P-value
Fat (%) 0.54 ± 0.06 0.63 ± 0.11 0.29
Total protein (total N × 6.38) (%) 0.78 ± 0.03 0.85 ± 0.03 0.056
pH 4.6-soluble N (%) 0.13 ± 0.01 0.14 ± 0.01 0.48
12% TCA-soluble N (%) 0.04 ± 0.003 0.05 ± 0.01 0.057
TS (%) 7.04 ± 0.09 7.16 ± 0.2 0.20
Ash (%) 0.52 ± 0.01 0.53 ± 0.03 0.67
Protein recovery (%) 22.42 ± 1.68 25.18 ± 2.81 0.22
Fat recovery (%) 13.51 ± 1.47 16.15 ± 2.48 0.19
1
Treatments were Prato cheese made with coagulant from Thermomucor indicae-seudaticae N31 and with com-
mercial coagulant.

Journal of Dairy Science Vol. 96 No. 12, 2013

7496 ALVES ET AL.

Table 3. Summary of the ANOVA: effect of treatment, storage time, and treatment × storage time interaction
on pH, moisture, pH 4.6-soluble nitrogen [pH 4.6 SN; % total nitrogen (TN)], 12% TCA-soluble nitrogen (12%
TCA SN; %TN), and firmness of Prato cheeses

P-value

pH 4.6 SN 12% TCA SN Firmness

Factor df pH Moisture (%TN) (%TN) (g)
Treatment (T)1 2 0.68 <0.0001 <0.0001 0.38 <0.0001
Storage time (ST)2 5 0.55 0.27 <0.0001 <0.0001 <0.0001
T × ST 5 0.86 0.73 0.59 0.84 0.98
1
Treatments were Prato cheese made with coagulant from Thermomucor indicae-seudaticae N31 and with com-
mercial coagulant.
2
Storage time: 5, 12, 19, 33, 43, and 53 d after manufacture.

proteolysis and moisture content between the cheeses The hydrolysis of αs1-CN 8P formed the fraction αs1-I-
of the present study probably reflected on firmness CN 8P that was visible on the fifth day of manufacture
differences. The control Prato cheese presented higher for both cheeses. The hydrolysis of αs1-CN 9P formed
proteolysis [pH 4.6 SN (%TN) = 9.75 ± 3.81%], higher the fraction αs1-I-CN 9P, strongly visible after 33 d of
moisture content (43.26 ± 0.55%), and less firmness ripening for both cheeses. The β-CN genetic variants A1
(1,737.55 ± 153.19 g) than Thermomucor Prato cheese, and A2 were also hydrolyzed, possibly by plasmin, with
whose values for 4.6 pH SN (%TN), moisture con- small peaks of γ-CN arising over time, especially γ2-CN,
tent, and firmness were 7.81 ± 3.34%, 41.07 ± 0.47%, whose increase was visible from d 33 in both cheeses.
and 2,296.23 ± 177.56 g, respectively. It is clear that The hydrolysis profile of both cheeses produced with
cheeses with higher moisture content ripened faster and commercial microbial enzyme (obtained from Rhizomu-
presented lower firmness, given that the hydrolysis of cor spp.) and the enzyme from T. indicae-seudaticae
casein and its degradation products favors hydration N31 was equivalent to the classic and expected hydro-
and weakening of the protein matrix (Fox et al., 2000), lysis profile of cheeses produced with chymosin [i.e.,
resulting in greater softening of the cheese. hydrolysis of αS1-CN by the residual coagulant during
The hydrolysis profile of the different casein frac- the early stages of maturation in the bond Phe23-Phe24,
tions during cheese storage can be seen in the capillary resulting in the peptides αS1-CN f24–199 (αS1-I-CN)
electropherogram shown in Figure 3. We observed that and αS1-CN f1–23, and hydrolysis of β-CN by plas-
the hydrolysis of different protein fractions was similar min, resulting in γ-CN; Fox et al., 2000]. The set of
for both Thermomucor (Figure 3A) and control Prato parameters studied during ripening suggests that the
cheese (Figure 3B). The fractions αs1-CN 8P and αs1- enzyme from T. indicae-seudaticae N31 has potential
CN 9P were degraded in the first 19 d of ripening, and as a coagulating agent for the production of Prato
the last one disappeared from d 33 in the control Prato cheese, equivalent to the commercial enzyme used for
cheese and from d 43 in the Thermomucor Prato cheese. comparison. However, to reach more conclusive results
on proteolysis, as to whether the higher content of pH
4.6 SN (%TN) in the control cheeses were caused by
a difference in the activity, specificity, inactivation, or
retention of the coagulant or by the difference in mois-
ture content, or by combinations of these possibilities,
future work regarding the determination of residual
coagulant activity in the cheeses and producing cheeses
with comparable moisture content could be carried out.

Sensory Acceptance and Purchase Intent Test

The sensory evaluation was performed after approval

by the Ethics in Research Committee (University of
Campinas, Campinas, SP, Brazil; register no. 470/2011)
Figure 2. Effect of storage time on pH 4.6-soluble nitrogen (pH 4.6 and after confirming that the products met the stan-
SN; ), 12% TCA-soluble nitrogen (12% TCA SN; ), and firmness dards required for microbiological safety established by
( ) of Prato cheese during storage (n = 3). Values with the same let-
ters (a–d) are not significantly different (P < 0.05) for each analytical the Brazilian legislation for Prato cheese (Agência Na-
determination. cional de Vigilância Sanitária do Brasil, 2001), includ-
Journal of Dairy Science Vol. 96 No. 12, 2013
 PRATO CHEESE PRODUCED WITH DIFFERENT COAGULANTS 7497

Figure 3. Capillary electrophoresis curves presenting casein breakdown during storage of Thermomucor Prato cheese (A) and control Prato
cheese (B). Labeling of peaks is according to Otte et al. (1997) and Rehn et al. (2010). AU = arbitrary units.

Journal of Dairy Science Vol. 96 No. 12, 2013

7498 ALVES ET AL.

Table 4. Scores for the attributes of control and Thermomucor Prato cheeses obtained in the sensory evaluation
(means ± SD; n = 100)

Control Prato Thermomucor

Attribute cheese Prato cheese P-value
1
Appearance 7.75 ± 1.00 7.55 ± 1.01 0.16
Aroma1 7.31 ± 1.23 7.43 ± 1.20 0.48
Flavor1 6.98 ± 1.40 7.07 ± 1.43 0.65
Texture1 7.20 ± 1.55 6.99 ± 1.41 0.32
Overall impression1 7.13 ± 1.31 7.06 ± 1.33 0.71
Purchase intent2 3.81 ± 1.09 3.80 ± 1.02 0.95
1
Scale: 1 = extremely disliked, 7 = liked moderately, and 9 = liked extremely.
2
Scale: 1 = certainly would not buy this product, 5 = certainly would buy this product.

ing coliform at 45°C [most probable number (MPN)/g Brazil) and Technological and Foundation for Research
<3], coagulase-positive Staphylococcus (cfu/g <10), Support of the State of São Paulo (FAPESP), process
Salmonella spp. (absent in 25 g of sample), and Listeria 2011/51158-8 and 2011/50844-5, for financial support.
monocytogenes (absent in 25 g of sample). As can be
seen in Table 4, the control and Thermomucor Prato REFERENCES
cheeses showed no significant differences for the sensory Agência Nacional de Vigilância Sanitária do Brasil. 2001. Regulamen-
parameters studied (appearance, aroma, flavor, texture, to técnico sobre padrões microbiológicos para alimentos. Resolução
RDC No. 12-02/01/2001. Accessed Jan. 3, 2013. http://portal.anvisa.
and overall impression). Both cheeses presented high gov.br/wps/wcm/connect/a47bab8047458b909541d53fbc4c6735/
acceptability indexes for all parameters, with scores RDC_12_2001.pdf?MOD=AJPERES.
between 7 and 8, which are equivalent to the “like mod- AOAC International. 1995. Official Methods of Analysis. 16th ed.
AOAC International, Arlington, VA.
erately” and “like very much” categories of the sensory AOAC International. 2006. Official Methods of Analysis. 18th ed.
evaluation scale. With respect to the overall impression, AOAC International, Arlington, VA.
88% of the assessors scored in the acceptance range, Bourne, M. 2002. Food Texture and Viscosity: Concept and Measure-
ment. 2nd ed. Academic Press, New York, NY.
5% reported having no opinion, and 7% rejected the British Standards Institution. 1989. Determination of fat content of
cheeses. With respect to purchase intention, 66% of the milk and milk products (Gerber method)—Methods. British Stan-
assessors declared that they would buy Thermomucor dards Institution, London, UK.
Bynum, D. G., and D. M. Barbano. 1985. Whole milk reverse osmosis
and control Prato cheeses and only 13% declared that retentate for Cheddar cheese manufacture: Chemical changes dur-
they would not buy the cheeses. ing aging. J. Dairy Sci. 68:1–10.
Proteases from microbial sources generally present Cavalcanti, M. T. H., C. R. Martinez, V. C. Furtado, B. B. Neto, M. F.
Teixeira, J. L. Lima Filho, and A. L. F. Porto. 2005. Milk-clotting
high proteolytic activity, which may result in bitterness protease production by Nocardiopsis sp. in an inexpensive medium.
in aged cheese varieties, due to their action on highly World J. Microbiol. Biotechnol. 21:151–154.
hydrophobic sequences of casein (Sousa et al., 2001). Fox, P. F. 1989. Proteolysis during cheese manufacture and ripening.
J. Dairy Sci. 72:1379–1400.
The higher proteolysis of control Prato cheese did not Fox, P. F., T. P. Guinee, T. M. Cogan, and P. L. H. McSweeney. 2000.
reflect in the sensory assessment of cheese texture and Fundamentals of Cheese Science. Aspen Publishers Inc., Gaithers-
flavor and the good sensory acceptance of the Ther- burg, MD.
Jacob, M., D. Jaros, and H. Rohm. 2011. Recent advances in milk clot-
momucor and control Prato cheeses indicate that the ting enzymes. Int. J. Dairy Technol. 64:14–33.
enzyme from T. indicae-seudaticae N31 presented simi- Kloosterman, J. 1991. The role of biotechnology in the manufacturing
lar action to the enzyme from R. miehei, being able to of wholesome natural ripened cheese. Food Biotechnol. 5:207–215.
Kumar, S., N. S. Sharma, M. R. Saharan, and R. Singh. 2005. Ex-
impart typical sensory properties to the finished cheese. tracellular acid protease from Rhizopus oryzae: Purification and
characterization. Process Biochem. 40:1701–1705.
CONCLUSIONS Li, S., X. Yang, S. Yang, M. Zhu, and X. Wang. 2012. Technology
prospecting on enzymes: Application, marketing and engineering.
The use of protease from T. indicae-seudaticae N31 Comput. Struct. Biotechnol. J. 2:e201209017.
Lucey, J., and J. Kelly. 1994. Cheese yield. Int. J. Dairy Technol.
in Prato cheese manufacture presented satisfactory 47:1–14.
results regarding the main requirements when choos- Machalinski, C., M. L. Pirpignani, C. Marino, A. Mantegazza, and
ing a clotting agent: good cheese yield and proteolysis M. B. de Jiménez Bonino. 2006. Structural aspects of the Mucor
bacilliformis proteinase, a new member of the aspartyl-proteinase
during ripening that resulted in the development of soft family. J. Biotechnol. 123:443–452.
texture and good sensory acceptance. Marshall, R. T. 1992. Standard Methods for the Determination of
Dairy Products. 16th ed. American Public Health Association,
ACKNOWLEDGMENTS Washington, DC.
Martin, N., M. A. U. Guez, L. D. Sette, R. Da Silva, and E. Gomes.
2010. Pectinase production by a Brazilian thermophilic fungus
The authors thank the National Council for Scientific Thermomucor indicae-seudaticae N31 in solid-state and submerged
and Technological Development (CNPq, Brasília, DF, fermentation. Microbiology 79:306–313.

Journal of Dairy Science Vol. 96 No. 12, 2013

 PRATO CHEESE PRODUCED WITH DIFFERENT COAGULANTS 7499
Mazal, G., P. C. B. Vianna, M. V. Santos, and M. L. Gigante. 2007. of bovine caseins by capillary zone electrophoresis. Food Contr.
Effect of somatic cell count on Prato cheese composition. J. Dairy 14:307–315.
Sci. 90:630–636. Otte, J., M. Zakora, K. R. Kristiansen, and K. B. Qvist. 1997. Analy-
Merheb-Dini, C., G. A. C. Garcia, A. L. B. Penna, E. Gomes, and sis of bovine caseins and primary hydrolysis products in cheese by
R. da Silva. 2012. Use of a new milk-clotting protease from Ther- capillary zone electrophoresis. Lait 77:241–257.
momucor indicae-seudaticae N31 as coagulant and changes during Rehn, U., M. A. Petersen, K. Hallin Saedén, and Y. Ardö. 2010. Rip-
ripening of Prato cheese. Food Chem. 130:859–865. ening of extra hard cheese made with mesophilic DL-starter. Int.
Merheb-Dini, C., E. Gomes, M. Boscolo, and R. da Silva. 2010. Produc- Dairy J. 20:844–851.
tion and characterization of a milk-clotting protease in the crude Richardson, G. H. 1985. Standard Methods for Examination of Dairy
enzymatic extract from the newly isolated Thermomucor indicae- Products. 15th ed. American Public Health Association, Wash-
seudaticae N31: (Milk-clotting protease from the newly isolated ington, DC.
Thermomucor indicae-seudaticae N31). Food Chem. 120:87–93. Shieh, C.-J., L.-A. P. Thi, and I.-L. Shih. 2009. Milk-clotting enzymes
Ministério da Agricultura do Brasil. 1997. Regulamento técnico para produced by culture of Bacillus subtilis natto. Biochem. Eng. J.
fixação de identidade e qualidade do queijo Prato. Portaria 358- 43:85–91.
04/09/1997. Ministério da Agricultura, Pecuária e Abastecimen- Sousa, M. J., Y. Ardö, and P. L. H. McSweeney. 2001. Advances in
to, Brasília, Brazil. Accessed Jan. 3, 2013. http://sistemasweb. the study of proteolysis during cheese ripening. Int. Dairy J.
agricultura.gov.br/sislegis/action/detalhaAto.do?method=consult 11:327–345.
arLegislacaoFederal. Stone, H., and J. L. Sidel. 1993. Sensory Evaluation Practices. 2nd ed.
Ministério da Agricultura do Brasil. 2011. Regulamentos Técnicos Academic Press, New York, NY.
de Produção, Identidade e Qualidade do Leite tipo A, do Leite Vishwanatha, K. S., A. G. A. Rao, and S. A. Singh. 2010. Production
Cru Refrigerado, do Leite Pasteurizado e o Regulamento Técnico and characterization of a milk-clotting enzyme from Aspergillus
da Coleta de Leite Cru Refrigerado e seu Transporte a Granel. oryzae MTCC 5341. Appl. Microbiol. Biotechnol. 85:1849–1859.
Instrução Normativa No. 62-29/09/2011. Accessed Jan. 3, 2013. Walsh, M. K., and X. Li. 2000. Thermal stability of acid proteinases.
http://sistemasweb.agricultura.gov.br/sislegis/action/detalhaAto. J. Dairy Res. 67:637–640.
do?method=consultarLegislacaoFederal.
Ortega, N., S. M. Albillos, and M. D. Busto. 2003. Application of
factorial design and response surface methodology to the analysis

Journal of Dairy Science Vol. 96 No. 12, 2013