Effect of brine concentration on physico-chemical characteristics, texture, rheological properties and proteolysis level of cheeses produced by an optimized wild cardoon rennet

Abstract

The main objective of this study was to test the efficiency of a wild cardoon (Cynara cardunculus L.) rennet, previously optimized by response surface methodology, in cheese making process; then to select the best brine concentration, leading to excellent cheese quality. Results showed that the optimized C. cardunculus rennet and chymosin produced curds with similar properties (yield, colour, texture, viscoelasticity), suggesting that this coagulant could replace successfully calf rennet. After brining at different salt concentrations (5, 7, 10 and 15%), we concluded that the use of 15% of salt in brine was an efficient way to reduce considerably the proteolysis level in C. cardunculus cheeses, stored for 28 d at 4 °C. At this salt level, the highest hardness, gumminess, viscoelasticity and yield of soft cheeses were also recorded. In conclusion, the satisfactory findings could open new opportunities to produce industrially the optimized C. cardunculus rennet and its cheeses in the Mediterranean area.

Introduction

Different types of plant coagulants were employed in milk gelation and cheese-making process, such as aqueous extracts from wild cardoon flowers, which have been used for ages in the production of goat’s and ewe’s cheeses, especially in Portugal and Spain. Cheeses made with cardoon extracts were normally produced on an artisanal scale. However, they have an important socio-economical contribution to dairy and agricultural sectors, at regional areas of Mediterranean countries (Ordiales et al. 2012).

Milk-clotting activities (MCA) of cardoon extracts were attributed to aspartic proteases, named cardosins A and B, resembling to chymosin and pepsin; respectively, in terms of specificity and activity (Ordiales et al. 2012). The main drawback of cardosins is that their industrial use, as calf rennet substitutes, is practically non-existent because of lower yields, texture defects (poor and creamy textures), and high bitterness of the produced cheeses. This is mainly due to their excessive proteolytic activity (Ben Amira et al. 2017a).

However, some improvements in the understanding of their action during milk gelation and the controlling of all parameters, influencing cheese making process, suggest a change. In fact, the use of a suitable wild cardoon extract with a highest specific MCA, the application of a minimal coagulant dose for milk gelation, and the control of proteolysis level during cheese storage, could promote excellent results in terms of rheological and sensory properties, as compared to animal rennet (Ben Amira et al. 2017a).

Besides, one of the interesting ways to improve quality characteristics of C. cardunculus cheeses, is the use of an appropriate salt concentration during cheese brining. In this step, salt is absorbed by cheeses with a concomitant decrease of salt concentration in the brine. This net movement of Na⁺ and Cl⁻ ions occurs because of osmotic pressure differences between cheese and brine (Guinee and Fox 2004). In fact, a sufficient salt content in cheese contributes directly to its flavor and improves its texture by limiting proteolytic activity (Kaya et al. 1999). For this reason, it seems to be important in this study, to select the best brining conditions by evaluating yields, textures and rheological properties of C. cardunculus cheeses. To our knowledge, there is no published study dealing with the salt effect on the quality of cheeses produced by wild cardoon rennet.

The first objective of the present study is to evaluate the quality characteristics of curds produced by commercial chymosin and wild cardoon rennet, previously optimized by response surface methodology (RSM) (Ben Amira et al. 2017b). This seems to be very important to test the efficiency of the cardoon rennet as an appropriate substitute of calf chymosin in cheese making process. The second purpose is to compare physico-chemical, textural, rheological properties, and RP-HPLC peptide profiles of C. cardunculus soft cheeses, brined at different salt concentrations. This could allow the selection of the best brining conditions leading to a better quality of the dairy products; thus, offering new opportunities to produce industrially the C. cardunculus cheeses in the Mediterranean area.

Materials and methods

Plant material

Flowers of Cynara cardunculus var. sylvestris were harvested from plants grown in the region of Bizerte in Tunisia, at the middle of the flowering season (at the end of June). Collected flowers were fully opened with violet color. The upper parts (petals and pistils) were cut and separated from the rest of the flowers. Then, they were carefully picked out to remove waste. All samples were then stored immediately at − 20 °C. After that, flowers were lyophilized (Heto Dry Winner) and ground. The obtained powders were finally stored in a freezer at − 20 °C, until the beginning of the extraction procedure.

Rennet extracts preparation

Fresh extracts from wild cardoon flowers, were prepared according to the procedure of Ben Amira et al. 2017b. All extraction parameters were optimized by response surface methodology and the resulting extract was named “optimized rennet” (Ben Amira et al. 2017b). The powdered sample (1.5 g) was mixed with 10 mL of 0.1 mol/L phosphate-citrate buffer (pH 3), during 50 min. The homogenates were centrifuged, in a Beckman Coulter (Aventi J-E) centrifuge, at 8000 g for 10 min at 4 °C. Supernatants were then filtered through cheesecloth. The fresh extracts were finally stored at 4 °C and used the same day, or they were frozen at − 20 °C until further use.

Cheese making

Raw cow’s milk was collected from the farm of “Gembloux” in Belgium, and used the same day for cheese making process. Cheese curds were prepared with pasteurized cow’s milk (63 °C/30 min) and the optimized wild cardoon rennet or diluted calf chymosin (10% dilution) (minimal chymosin concentration 530 mg/L ± 4%; pH 5.6 ± 0.2; Berthelot 530, Belgium). Briefly, 200 mL of pasteurized milk was placed in a wide-mouth beaker and mixed with 70 μL of 10 g/100 mL CaCl₂ and the coagulant amount necessary to clot the milk within 60 min (1 mL of the optimized extract or 600 μL of the diluted chymosin). The selection of the extract dose was based on a visual control of curd, which showed an acceptable hardness for further analysis. The appropriate chymosin dose was applied, according to Blecker et al. (2012), who used the same rennet for milk gelation.

Milk coagulation was performed at 35 °C for all samples. After clotting the milk, the coagulum was cut manually, then drained by manual pressure within cheesecloth and decantation into perforated molds at room temperature for few hours. Finally, the obtained curd was molded into small containers and stored overnight under refrigeration (4 °C). For each coagulant, three miniature cheese curds (4.5 cm diameter–13 mm height) were prepared for each further test.

The properties of curds produced by wild cardoon rennet and chymosin were compared. Then, C. cardunculus curds were brined at different salt concentrations, in order to obtain soft brined cheeses, to be stored for 28 d, for further analyses.

Brine preparation

The preparation of brines was carried out, using different salt concentrations (5, 7, 10 and 15 g/100 mL (w/v) of NaCl). Salt was dissolved in hot water (95 °C/30 min) and homogenized for 15 min at room temperature. Then, solutions were refrigerated at 4 °C, before cheese immersion. Molded curds were brined at a ratio of (1:5) (w/v) for 16 h. Once retired from the salted solutions, cheeses were drained at room temperature on a perforated wooden plate. Finally, they were stored at 4 °C, from 5 to 28 d, for further analyses.

Physico-chemical analysis

Moisture, ash and pH

Moisture was determined after dessication of 2 g of cheese sample at 105 °C ± 2 °C, over night. The pH of cheeses was measured directly, using an electrode connected to a pH meter for semi-solid food products. Ash content was determined after incineration of 10 g of cheese sample in a muffle furnace at 550 °C for 16 h, and the result was expressed as a percentage of dry matter.

Cheese yield

Fresh curd or cheese yields were determined on the basis of cheese weights and volumes of milk. The percentage of cheese yield was calculated according to the following Eq. (1):

$$\text{Cheese yield}\left(\frac{\text{g}}{100\text{g}}\text{Fw}\right)=\frac{\text{Weight of cheese}\left(\text{g}\right)\times 100}{\text{Volume of milk}\left(\text{mL}\right)}$$ 1

Fw: Fresh weight.

Total nitrogen content

Total nitrogen or protein content in brined cheeses was determined using a Dumas Elementar Rapid N cube 161 15,054 (Heraeus company, Donaustrasse, Germany). Two hundred mg of cheese, previously crushed with a mortar and pestle, were coated in paper containing 300 mg of sand, and pressed in pellet form. The Samples were transferred to the combustion tube and the nitrogen quantification was based on the quantitative digestion of the sample at approximately 900 °C, in presence of excess oxygen (Saint-Denis and Goupy 2004). Total nitrogen content was evaluated for all cheese samples, after 5, 10 and 28 days of storage.

Colour evaluation

The CIE L* a* b* coordinates were determined using a spectrophotocolorimeter (ColorFlex EZ Hunter Lab, Murnau, Germany). The CIE L* a* b* system employs L* (lightness), a* (redness), and b* (yellowness) values.

Rheological properties

Viscoelastic properties of curds and brined cheeses were evaluated at room temperature, using an MCR 302 Rheometer (Anton Paar company, Austria). The measuring geometry was a “plan-plan”, fitted with a plate and a probe “PP50”. The cylindrical cheese sample (4.5 cm diameter) was placed on the rheometer tray (the same diameter as the cheese) and the probe was fixed in direct contact with the top of cheese. The distance between (tray-probe) was about 13 mm (thickness of cheese).

In order to determine the viscoelastic region of the sample, a first test was carried out by measuring the G′ and G″ parameters as a function of the strain, which varied from 0.01 to 100%, at a constant frequency (1 Hz). The fixed strain (γ), corresponding to the linearity zone was then used to evaluate the storage and loss moduli, as a function of the oscillation frequency (0.1–100 Hz) (constant strain).

Texture profile analysis

Texture profile analysis (TPA) of curds and brined cheeses was accomplished using a Texture analyzer (Texturometer SMS TAXT plus). The obtained cylindrical samples were cut using a plastic box of 13 mm high and ultra thin fishing line in order to get cheeses with dimensions of 4.5 cm diameter and 13 mm height. Curds or brined cheeses were held at room temperature for 10 min before evaluation. The probe used was a cylinder of 25 × 20 mm (diameter × height). Samples were analyzed at 20 °C and compressed by 50% in two compression cycles at a constant crosshead velocity of 1 mm/s and a waiting time of 10 s. Hardness, cohesiveness and springiness were recorded from two successive compression cycles.

RP-HPLC of the WSN fraction

The extraction of WSN fraction was performed according to the method of Kuchroo and Fox (1982), with modifications. 26.66 g of each brined cheese, was crushed using a mortar and pestle; then, mixed with 40 mL of water and homogenized using an Ultra-Turrax T 25 homogenizer for 2 min. The homogenate was held at 40 °C for 1 h, pH was adjusted to 4.4–4.6, with HCl (7 N), and the suspension was centrifuged for 30 min at 4 °C at 4800 rpm. Supernatant was filtered through CA Syringe Filters (0.22 μm–25 mm) to obtain the water soluble fraction. Then, 0.2% of sodium azide (0.1%) was added to this fraction in order to avoid contamination.

The samples (100 µL) were injected and analyzed on an Agilent 1100 series HPLC instrument using an Eclipse plus C18 (Zorbax) reverse phase chromatography peptide column (5 μm, 3 mm × 150 mm). Peptides were eluted using a gradient of 5 to 100% B, 95 to 0% A over 55 min (where A is 0.1 g/100 mL trifluoroacetic acid (TFA) in water and B is 0.086 g/100 mL TFA in acetonitrile).

Statistical analysis

Results were assessed using SPSS. 21, i.e. SPSS^® Statistics 21 (©Copyright IBM Corporation, USA). T-student tests were employed for comparison of curds properties produced by chymosin or C. cardunculus rennet. Duncan’s-tests were applied for comparison of C. cardunculus cheeses properties, brined at different salt concentrations. Two-way analyses of variance (ANOVA) were used to test the interaction between brine concentration and time. Data were considered significant if the P value was below 0.05.

Results and discussion

Curd properties

Yield

To evaluate any enzyme as a rennet substitute, yields of cheese curds produced by the C. cardunculus extract and calf chymosin were compared in Table 1. Results showed that there was no significance (P < 0.05) difference in the curd yields derived from the cardoon extract and calf chymosin (18.16% and 18.62%, respectively). This highlights the importance of both our coagulant extract, previously optimized by RSM (Ben Amira et al. 2017b), and the use of the appropriate dose for milk coagulation.

Table 1.

Evaluation of yields, texture, viscoelastic properties and colour of curds produced by C. cardunculus rennet and chymosin, stored for 2 days at 4 °C

Curds properties	Coagulants
	Cynara cardunculus rennet	Chymosin
Curd yield (g/100 g Fw)	18.16 ± 0.33a2	18.62 ± 0.53a
Texture
Hardness (N)	14.36 ± 0.34a	14.60 ± 0.65a
Springiness	0.85 ± 0.01a	0.86 ± 0.02a
Gumminess (N)	10.04 ± 0.39a	10.58 ± 0.22a
Cohesiveness (N)	0.69 ± 0.02a	0.72 ± 0.02a
Viscoelastic parameters
G′ (Pa) F11 = 158 rad/s	14,833 ± 378.59a	13,833 ± 1724a
G″ (Pa) F1 = 158 rad/s	5150 ± 135.27a	3980 ± 600.99b
G′ (Pa) F2 = 39.6 rad/s	12,400 ± 305.50a	11,030 ± 1669a
G″ (Pa) F2 = 39.6 rad/s	3673 ± 92.91a	2913 ± 486.44a
Colour parameters (Cie L*a*b*)
L*	89.29 ± 0.32a	88.75 ± 0.56a
a*	1.45 ± 0.12a	1.56 ± 0.23a
b*	16.24 ± 0.35a	17.10 ± 0.81a

¹F Frequency; Fw Fresh weight

²All values given are means of three repetitions (x̅ ± SD). SD Standard deviation. Values with different superscript letters within the same line are different (P < 0.05)

Based on information from previous studies, most aspartic proteases from plants cannot successfully replace calf rennet, because of the great loss of protein during cheese making (Mazorra-Manzano et al. 2013). For example, curds made by kiwi (17.8%), melon (15.1%), and ginger (15.4%) extracts presented lower values than chymosin curd (20.2%). This was in agreement with the results of Bruno et al. (2010) who found lower Bromelia hieronymi cheese yield (14.52%), compared to that of chymosin (16.05%).

Colour

Results in Table 1 showed that the two types of curds (C. cardunculus and chymosin) were characterized by similar values (P > 0.05) of all Cie L* a* b* parameters. These findings were in agreement with those observed by the naked eye, confirming that C. cardunculus coagulant and chymosin produced curds with the same colour properties.

This efficiency of C. cardunculus rennet on curd colour arises probably from the use of a minimal dose for milk gelation, allowing reduced amounts of colored pigments in gel, then in cheese matrix. In fact, the extract used in this study was previously optimized by RSM, in order to maximize its specific activity (MCA) (Ben Amira et al. 2017b). So, even if a minimal dose was used, the obtained curd would reach excellent properties due to the high MCA of the coagulant extract. Furthermore, the latter was prepared at low pH (pH 3), involving the extraction of the highest enzyme fraction as a percentage of total proteins. Therefore, more purified coagulant enzymes with less contaminant proteins and colored pigment were recovered, which was an advantage to improve colour.

Rheological properties

Concerning rheological characteristics, both storage and loss moduli of each curd (cardoon or chymosin) were dependent on oscillatory frequency. They increased with the rise of angular frequency. Also, they demonstrated similar shapes and trends. The G′ was greater than G′′ at any frequency for all samples, which indicated a dominant contribution of the elastic component to the visco-elasticity (Kahyaoglu and Kaya 2003; Madadlou et al. 2005). As a result, this reflects the typical behavior for a viscoelastic solid product (Ustunol et al. 1995). Similar results were found by Li et al. (2013).

In fact, in order to detect differences between curd samples, the dynamic moduli obtained at two selected frequencies (158 and 39.6 rad/s) were compared in Table 1. The G′ values showed no significant differences between curds derived from chymosin and cardoon extract (P > 0.05). Thus, the two coagulants contributed to the same viscoelastic properties and rigidity of curds matrix, which could be translated by similar internal cohesive forces. These results arose probably from the same viscoelasticity and firmness of milk gels, previously made by the two coagulants (Ben Amira et al. 2017b).

Texture

As shown in Table 1, texture measurements revealed similar values (P > 0.05) of hardness (14.36 and 14.60 N), springiness (0.85 and 0.86), gumminess (10.04 and 10.58 N), and cohesiveness (0.69 and 0.72 N) of curds produced by the two coagulants, respectively. The same texture was observed for curds made by kiwi extract and chymosin, which exhibited similar values of the parameters describing elasticity (7.10 and 6.38 mm), and cohesion (0.33 and 0.35 N), respectively (Mazorra-Manzano et al. 2013). They presented also the highest hardness values of curds (6.19 and 5.52 N). In fact, the latter were considerably lower than those obtained in our study (Table 1), due to the differences in cheese making and texture measurements methods, the plant extracts used, and moisture contents in the produced curds.

On the other side, texture results in Table 1 correlated well with those of Brutti et al. (2012), who found that the semi-hard cheese manufactured with Onopordosine, showed similar texture characteristics to other commercial cheeses of the same type. However, it seemed that our findings were better than those of most cheeses produced by plant rennets, which revealed an unsatisfactory creamy texture (Hashim et al. 2011).

All these findings confirmed that the Tunisian C. cardunculus extract, previously optimized by RSM, could replace successfully the commercial chymosin in cheese making process, regarding to colour, yield, texture and viscoelasticity. However, the better understanding of cheese properties evolution during storage, and monitoring particularly proteolysis level and mechanic behaviors of brined cheeses, are also very relevant.

Physico-chemical properties of C. cardunculus cheeses brined at different salt concentrations

In this study, physico-chemical properties of C. cardunculus cheeses, brined at different salt concentrations (5, 7, 10 and 15%), were evaluated. All parameters (moisture, ash, pH, cheese yield and total nitrogen content (% N)) were determined, after 28 days of storage at 4 °C (Fig. 1a–e).

Fig. 1.

Physico-chemical properties of C. cardunculus cheeses, brined at different salt concentrations, and stored for 28 d at 4 °C

Results in Fig. 1a showed a significant decrease of moisture values from 59.57% at 5% salt, to 56.68% at 15% salt. This finding could be explained by the fact that the high salt content promotes hydrophobic interactions, and can cause a faster expulsion of serum from brined cheese, leading to a decrease in cheese moisture content (McMahon et al. 2009).

According to Kaya (2002), moisture loss and salt uptake are mutual diffusion processes leading to an osmotic dehydration of the product. The more salt-in-moisture content increases, the more moisture content decreases. For cheeses brined at lower salt percentages, the elevated moisture content could be related to caseins proteolysis, which is possibly due to the act of residual coagulant enzymes (cardosins) or adventitious microflora.

The ash contents of brined cheeses, stored for 28 d at 4 °C, were presented in Fig. 1b. As expected, this parameter raised significantly with the increase of salt values, from 2.16% DM (5% salt) to 3.68% DM (15% salt) (P < 0.05). It can be assumed that the increased values of ash, at high salt concentration, were the result of reduced moisture and increased protein interactions. This would increase protein-bound minerals, such as colloidal calcium phosphate.

Results in Fig. 1c showed that pH of all samples was in the range of (6.49—6.59). These values were too close to the starting milk pH (6.6), revealing that pH decreased slightly after cheese manufacturing and storage, for 28 d at 4 °C.

Results in Fig. 1e showed that maximal values of 15.56% and 15.95% (P > 0.05) were obtained for samples brined at 10 and 15% of salt, respectively. These data demonstrated that the elevated salt level in cheese have led to higher protein interactions within cheese matrix, with a minimum of loss during storage (Fig. 2).

Fig. 2.

Evolution of cheese yield of C. cardunculus cheeses brined at different salt concentrations (5, 7, 10 and 15%), as a function of storage days (4 °C). Salt concentrations: ( ) 5%, ( ) 7%, ( ), 10%, ( ) 15%

Viscoelastic properties and texture of C. cardunculus cheeses brined at different salt concentrations

Frequency sweep tests were employed to evaluate whether the salt level affects the visco-elastic characteristics of C. cardunculus cheeses during storage. Results in Fig. 3 Ia, b showed that brined cheeses exhibited the same general shape of dynamic moduli (G′, G″) evolution, as described previously for curds. Also, increasing the salt level in brine allowed a remarkable rise of both storage and loss modulus evolution, as a function of angular frequency. This suggested a clear improvement of cheese viscoelasticity, due to high salt content.

Fig. 3.

I & II. I a–d. Evolution of dynamic moduli (G′, G″) of C. cardunculus soft cheeses brined at different salt concentrations (5, 7, 10 and 15%) and stored for 28 d at 4 °C, as a function of angular frequency (rad/s). Salt concentrations: ( ) 15%, ( ) 10%, ( ) 7%, ( ) 5%. II. Evolution of storage modulus (G′) of C. cardunculus cheeses, brined at different salt concentrations (5, 7, 10 and 15%) and stored for 5, 10 and 28 d at 4 °C, as a function of angular frequency. ( ) 5 d ( ) 10 d ( ) 28 d

In order to detect differences between cheese samples, elastic moduli obtained at two selected frequencies (158 and 39.6 rad/s) were compared in Table 2a–c. For the two frequencies, the difference between G′ values of cheeses, brined at various salt concentrations, began to be more pronounced from the tenth day of storage (Table 2b). A clear significant difference (P < 0.05) was observed between cheeses brined at 5% and those brined at 15% of salt (24,760 Pa and 36,275 Pa, respectively (158 rad/s), suggesting that the elastic behavior of cheeses was greatly improved at high salt level. The same result was obtained for values found at a lower frequency (39.6 rad/s) (Table 2b), and a similar shape was also observed after 28 d.

Table 2.

Texture profile analysis and storage modulus (G′) values recorded at two angular frequencies of C. cardunculus cheeses, brined at different salt concentrations and stored for 28 days

Salt concentration (%)	Hardness (N)	Springiness (cm2)	Gumminess (N)	Cohesiveness	G′ (Pa)1	G′ (Pa)
					F1 = 158 rad/s	F2 = 39.6 rad/s
(a) 5 Days
5	18.93 ± 1.16a,A2	0.92 ± 0.01 a,A	14.67 ± 0.79 a,A	0.77 ± 0.01a, A	27,260 ± 5417a,A	21,780 ± 4378a,A
7	22.45 ± 0.70b,A	0.91 ± 0.02 a, A	17.46 ± 0.42 b, A	0.77 ± 0.01a,A	25,520 ± 1465a,A	20,700 ± 1111a,A
10	25.74 ± 3.26b,A	0.84 ± 0.02 a, A	18.32 ± 1.63 b, A	0.71 ± 0.03a,A	29,580 ± 6790a,A	23,880 ± 5264a,A
15	23.59 ± 0.16b,A	0.63 ± 0.08b,A	11.68 ± 1.99c,A	0.49 ± 0.08b,A	35,060 ± 15263a,A	28,340 ± 11827a,A

Salt concentration (%)	Hardness (N)	Springiness (cm2)	Gumminess (N)	Cohesiveness	G′ (Pa)1	G′ (Pa)
					F1 = 158 rad/s	F2 = 39.6 rad/s
(b) 10 Days
5	22.01 ± 1.84a, B	0.83 ± 0.02a,B	15.73 ± 1.32a,A	0.71 ± 0.03a,B	24,760 ± 3158a,A	20,400 ± 2472a,A
7	23.86 ± 1.51a,b,A	0.81 ± 0.06a,B	16.09 ± 1.84a,A	0.67 ± 0.04a,B	26,160 ± 3932a,A	21,760 ± 3241a,A
10	26.07 ± 0.69b,A	0.69 ± 0.05b,B	13.67 ± 2.48a,B	0.52 ± 0.03b,B	30,420 ± 3218a,b,A	24,680 ± 2281a,b,A
15	24.62 ± 2.98b,A	0.81 ± 0.04a,B	15.38 ± 0.93a,B	0.63 ± 0.04b,a,B	36,275 ± 7454b,A	28,700 ± 5856b,A

Salt concentration (%)	Hardness (N)	Springiness (cm2)	Gumminess (N)	Cohesiveness	G′ (Pa)1	G′ (Pa)
					F1 = 158 rad/s	F2 = 39.6 rad/s
(c) 28 Days
5	15.82 ± 1.10a,C	0.87 ± 0.02 a,C	11.61 ± 0.89a,B	0.73 ± 0.03a,A,B	23,125 ± 2804a,A	18,550 ± 2066a,A
7	23.32 ± 0.21b,A	0.80 ± 0.03a,b,B	15.62 ± 0.96b,A	0.67 ± 0.03a,b,B	25,320 ± 3546a,b,A	21,160 ± 2687a,b,A
10	24.31 ± 0.35b,A	0.74 ± 0.06b,A,B	15.28 ± 1.06b,A,B	0.63 ± 0.05b,A,B	30,240 ± 2818a,b,A	24,820 ± 2160b,A
15	48.34 ± 1.17c,B	0.49 ± 0.05c,C	17.55 ± 2.28b,B	0.36 ± 0.05c,C	32,340 ± 9721b,A	26,860 ± 7352b,A

¹G′ values were measured at two angular frequencies (158 rad/sec and 39.6 rad/sec)

²All values given are means of three repetitions (x̅ ± SD); SD Standard deviation. Values with different superscript letters within the same column are different (P < 0.05)

Capital superscript letters (A–C) indicate significant differences between samples, brined at the same salt concentration, with different storage time. Lower-cased superscript letters (a–c) indicate significant differences according to the brine concentration, at the same storage time

Regarding viscoelastic stability, monitoring G′ values during storage was reported in Fig. 3 II and Table 2a–c. It revealed that for each salt concentration, the G′ evolution as a function of angular frequency was stable, and there was no significant difference recorded between 5, 10 and 28 d of storage. Consequently, the viscoelasticity of all brined cheeses was maintained during 28 d of storage. The slight reduction observed at 5% of salt (Fig. 3 II) was probably related to a higher proteolytic activity. This was confirmed previously by results of cheese yields, total nitrogen, and moisture contents.

Texture profile results of cheeses demonstrated that the analyzed samples stored for 28 d, were highly influenced by salt content in cheese. Results in Table 2c showed that hardness increased considerably with the rise of brine concentration. In fact, this salt effect on these parameters was more remarkable at elevated storage times (Table 2a–c). This could be explained by a progressive whey release and a reduction of cheese moisture, as a function of time, under low storage temperature (4 °C) and increased salt diffusion.

The present finding was also clearly demonstrated when examining results in Table 2c and those in Fig. 1a, b, d. The latter revealed that hardness correlated well with cheese physico-chemical properties and composition. From these data, it can be concluded that increasing salt level leads to elevated cheese hardness values, arising from reduced moisture, low total nitrogen content, and high cheese yield (low protein loss) after 28 d of storage (Fig. 1a, b, d).

The obtained hardness results correlated well with those of Kaya (2002) who found that cheese samples kept in 20 and 25% of salt were very hard, compared to those brined at lower concentrations. In addition, the authors concluded that a percentage of salt in brine less than 15% causes a weak cheese structure, which is in agreement with our findings in Table 2c.

In contrast with hardness and gumminess, springiness values obtained after 28 d, decreased significantly (P < 0.05) with elevated salt levels (Table 2c). This result was in agreement with cohesiveness values, which showed a similar evolution, as a function of salt concentration and storage time. The present finding was the outcome of low water contents in cheeses, brined at 15%, leading to the lack of “water-proteins” bindings, which would be implied in the establishment of cohesiveness and internal springiness.

Regarding texture stability, it has been shown that hardness and gumminess of cheeses brined at 15% of salt, increased during storage, as opposed to those brined at 5%. Concerning the latter, texture evolution showed a significant decrease of all these parameters (between 10 and 28 d). These reductions were mainly attributed to the softening of cheese protein matrix, brought by hydrolysis of αs₁-casein (Fontecha et al. 1996).

RP-HPLC peptide profiles of WSN fractions from C. cardunculus cheeses, brined at different salt concentrations

Peptide profiles of the WSN fractions from C. cardunculus cheeses were investigated by RP-HPLC, in Fig. 4a–d. Each peak corresponds at least to one peptide; and the peak area is normally proportional to the peptide concentration. In all cases, it was apparent that most peptides were eluted during the first 30 min of analysis.

Fig. 4.

a–d. RP-HPLC peptide profiles of water soluble nitrogen (WSN) fractions from C. cardunculus semi-hard cheeses, brined at different salt concentrations a 5%, b 7%, c 10% and d 15%, and stored for 28 d at 4 °C

When comparing chromatograms generating from cheeses brined at different salt concentrations, a similar general shape was observed (Fig. 4). This was completely evident, due to the use of the same coagulant extract in milk gelation, and the presence of residual cardosins in each cheese sample, which cleave the same peptide bonds. However, the differences in salt content led to fluctuations of peak numbers and areas. In fact, increasing salt concentration from 5 to 15% reduced greatly the intensity and the area of peaks. After peaks quantification, it can be concluded that the four major peaks areas (I, II, III & IV), were reduced at rates of 33.51, 56.50, 66.77 and 59.86%, respectively, from 5 to 15% of salt concentration.

These results reflect clearly the main contribution of high salt content to the decrease of both proteolysis level and released peptides concentration. This could be very advantageous to reduce undesired bitterness in C. cardunculus cheeses, which is an important criterion for subsequent consumer acceptance of the final products.

Conclusion

According to this study, the wild cardoon (C. cardunculus) rennet previously optimized by RSM, could be used as an efficient substitute of calf chymosin in cheese making process. Furthermore, our results allowed us to select the best salt concentration for C. cardunculus soft cheese brining (15%), leading to excellent results in terms of physico-chemical properties, yield, texture and viscoelasticity, after 28 d of storage (4 °C). These findings were in close agreement with RP-HPLC peptide profiles of brined cheeses, which revealed a remarkable reduction of proteolysis level, at 15% of salt. In this study, the satisfactory findings could open new opportunities to produce industrially the C. cardunculus rennet and its cheeses in the Mediterranean area.