Some Properties of Fresh and Ripened Traditional Akcakatik Cheese

In total, 15 Akcakatik cheese samples produced with cow’s milk (approximately 600 g per sample) were analysed in this study. During the period June-September, seven fresh (not ripened) samples (named K1, K2, K3, K4, K5, K6, and K7) and eight ripened samples matured in cloth bags (named D1, D2, D3, D4, D5, D6, D7, and D8) for 2-4 months were collected from local markets in Burdur and Isparta provinces located in the Western Mediterranean region of Turkey. Cheese samples were transported and stored in the refrigerator at +4℃ until they were analysed.

The percentage of titratable acidity, total dry matter (TS), fat and NaCl were determined according to Anonymous (1989). Titratable acidity of cheese samples was expressed as lactic acid percentage (LA%). Total nitrogen (TN) and water soluble nitrogen (WSN) contents were determined by the Kjeldahl method (Grippon et al., 1975). The ripening index (RI) was calculated from the ratio of soluble nitrogen to TN (Uraz and Simsek, 1998). The pH values of the cheeses were measured with a pH meter (WTW, Weilheim, Germany). The ratio of black cumin and clove was determined by modifying the Sagdic et al. (2003) method. The chemical analyses applied to the cheeses were repeated three times.

Sample preparation: 10 g Akcakatik cheese samples were treated with 25 mL of perchloric acid (Sigma-Aldrich 99%; 10 g/100 mL) in the stainless steel container of a Waring blender (8011 ES HGB2WTS3, Torrington, CT) for preparation of samples for biogenic amine analysis. Most of the fat was crystallised by cooling to 3℃ and then the samples were centrifuged at 10,000 rpm for 10 min at 3℃ (NF-1200R-Nuve-Turkey). The creamed layer was removed and the supernatant filtered. 5 mL of the filtrate was filtered with a sterile 0.45 µm syringe filter (Sartorius, Göttingen, Germany). Finally, this filtrate was diluted with distilled water to 50 mL and then 10 mL of this was injected into the column. The column was YMC-Pack ODS-AM C18 5 mm (250mm-4.6mm I.D.). The method given by Oner et al. (2006) was applied exactly without any modification.

Apparatus and chromatographic conditions: Shimadzu (Kyoto, Japan) system HPLC consisting of an LC 10 ADVP pump, DAD detector, autosampler [(SIL 10 AD VP (70 Vial Model Rack 7)] and oven (CTO 10 ACVP) was used. The mobile phase consisted of 30 mL acetonitrile/100 mL water at pH 8. The flow-rate was 1.1 mL/min.

Standards: Tyramine, 99% (Aldrich Chemical Co., T90344) Putrescine dihydrochloride, ≥98% (Sigma Chemical Co., P7505) Cadaverine dihydrochloride, >98% (Sigma Chemical Co., C8561) Histamine, ~97% (Sigma Chemical Co., H7125) ß-phenylethylamine, 99% (Aldrich Chemical Co., 128945)

Standards Preparation: biogenic amine stock standard solutions at 1000 mg/L each by dissolving 100 mg of tyramine, 182.7 mg of putrescine dihydrochloride, 171.4 mg of cadaverine dihydrochloride, 96 mg of histamine, and 100 mg of phenylethylamine in separate 100-mL volumetric flasks. Bring each to volume with deionised water. Store stock solutions at 4℃ and protect from light. Prepare working standard solutions for generating the calibration curve with an appropriate dilution of the stock solutions in 0.4 M perchloric acid. These solutions should be prepared fresh weekly and stored at 4℃ when not in use.

The calibration curves for the biogenic amines are linear having r²=0.9997, phenethylamine r²=0.9998, putrescine r²=0.9999, cadaverine r²=0.9999, histamine r²=0.9998, and tyramine r²

Volatile flavor compounds analysis was carried out by gas chromatography (Perkin Elmer Auto System XL) using an automatic head-space sampler (Turbo Matrix 16, Perkin Elmer). Two gram cheese samples from each treatment group were transferred into headspace vials (6 mL) and heated at 85℃ for 5 min. Headspace vials were automatically injected on a CP WAX column (30m.x 0.32 mm I.D.x 1.2 mm film thickness; Perkin Elmer). The temperature was maintained at 35℃ for 5 min. The flame-ionisation detector was set at 240℃ and the injector at 180℃. Helium was used as the carrier gas (25 psi). Operating conditions were: the oven temperature was held at 5℃ for 2 min and the temperature was increased in increments up to 240℃ with a total cycle time of 20 min. Components were identified depending on their retention times by comparing with the retention times of corresponding standards. All samples were analysed in triplicate. Temperatures of transfer line, needle and vial oven of the HS method were adjusted to 120, 90 and 85℃, respectively. Injection, pressurise, withdrawal and thermostat times of the HS method were 0.08, 0.08, 0.5 and 5 min, respectively (Ulberth, 1991).

Protein degradation levels of the seven fresh and eight ripened Akcakatik cheeses were evaluated by Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) using a mini vertical gel system (E-C 120 mini vertical gel system E-C 250-90, EC Apparatus Corporation, USA) previously described by Laemmli (1970). Protein standards (Catalog number SDS7) were obtained from Sigma Chemical (St. Louis. MO) to identify degraded or breakdown products of protein molecules. Each sample was diluted with sample buffer to give a final concentration of 1 mg/ml protein. An acrylamide to bisacrylamide ratio of 30% w/v (29:1) was used and the acrylamide concentrations in the running and stacking gel were 12 and 5%, respectively. Samples were heated in boiling water for 3 min and 20 μl was applied to each well. Electrophoresis was performed at a constant 60 mA. Gels were stained with Coomassie Brilliant Blue R250 solution (Merck, Darmstadt, Germany). After the staining, gel was destained and preserved in 10% acetic acid solution. The developed gel pictures were used to quantify degradative products of casein fractions by an advanced computerised optical densitometer (UV Transilluminator 2000. Bio Rad. Italy) located at the Laboratory of Research Centre in Süleyman Demirel University, Isparta-Turkey. The final numerical values of each breakdown product of protein molecules were quantitatively analysed.

10 grams of each cheese sample were weighed and homogenised with 90 mL sterile saline solution (0.85% NaCl, w/v). Then 10-fold dilutions were made with the same solution. Total aerobic mesophilic bacteria (TAMB) were enumerated by spread on Plate Count Agar (Merck) incubated at 30℃ for 72 h. LAB count was determined by spread on de Man, Rogosa and Sharpe Agar (Merck) containing 0.14% sorbic acid (w/v) incubated at 37℃ for 48 h. Yeast-moulds were enumerated by spread on Dichloran Rose Bengal Chloramphenicol Agar (DRBC, Merck) incubated at 25℃ for five days. The coliform group bacteria count was determined by the pouring plate method. For the pour plate, a 1 mL sample was mixed with 15 mL melted Violet Red Bile Agar (VRB, Merck). VRB plates were incubated at 37℃ for 48 h. Escherichia coli were enumerated by the double-layer plate method described by Cakir (1999). For double-layer plates, 1 mL sample was mixed with 15 mL melted Tryptic Soy Agar (TSA, Merck) and poured into sterile empty Petri plates. Then the plates were incubated at 35℃ for 2 h. After the incubation, TSA plates were overlaid with 5 mL of VRB agar. Then the plates were incubated at 45.5℃ for 24 h before counting.

For identification of LAB, isolates were randomly selected from MRS agar (Merck), M17 agar (Merck) and Neutral Red Chalk Lactose agar (NRCLA) (Harrigan and McCance, 1966). The isolates were tested by Gram staining, catalase and cultural tests growth in MRS broth (10℃, 15℃, 45℃ and pH 9.6) and tolerance of different salt levels (4%, 6.5%, 8% or 10% NaCl) (Huggins, 1984; Holt et al., 1994). Gram positive, catalase-negative cocci and rods were selected as presumptive LAB and they were identified phenotypically based on the production of acids from saccharides and related compounds. The production of acids from saccharides and related compounds was investigated using API 50 CH (BioMérieux, Marcy I’Étoile, France) according to the manufacturer’s instructions.

The mean value and standard deviation were calculated from the data obtained from the three separate experiments. T-test was used to evaluate the significance levels of the data with SPSS/PC software (Version 15.0) for Windows (SPSS Inc., Chicago II, USA). The level of confidence required for significance was set at p<0.05 (Duzgunes et al., 1987).

The results of the chemical analyses applied to the samples (pH, titratable acidity, total dry matter, fat, NaCl, total nitrogen, water soluble nitrogen and clove and black cumin amounts) are given in Table 1. In this study, pH values of Akcakatik cheeses were determined between 3.68 and 4.29. The average of titratable acidity values of fresh and ripened Akcakatik cheeses was determined as 2.30 ± 0.62 and 2.67 ± 0.31%, respectively. There is no significant difference in the total dry matter of the samples between the groups. The average of total dry matter of fresh and ripened Akcakatik cheeses was found to be 70.20 ± 4.97 and 73.38 ± 8.78%, respectively. Fat and NaCl contents of cheese samples showed changes between 16.50 - 40% and 4.32 - 16.54%, respectively. Total nitrogen contents of the fresh and ripened cheeses were determined on average to be 2.55 ± 0.54 and 2.77 ± 0.48%, respectively. WSN values of fresh and ripened cheeses were detected on average as 0.21 ± 0.13 and 0.51 ± 0.11%, respectively. WSN levels were found to be high in all ripened cheese and two fresh cheese samples (K6 and K7). Ripening index of fresh and ripened cheeses was detected as 7.85 ± 3.41 and 18.35 ± 1.75%, respectively. The clove amounts of the cheeses were determined to be 0.06 g/ 10 g as the lowest, 0.47g/10 g as the highest and 0.31 ± 0.27g/10 g on average. The average of the clove and black cumin amounts in fresh cheese samples was found to be higher than ripened cheese samples. The black cumin amount of all cheese samples was determined to be lower when compared to the clove amount except for ripened cheese sample D2. The difference was determined as statistically significant in terms of titratable acidity (LA%), fat%, NaCl%, clove amounts (g/10 g), water soluble nitrogen% and ripening index% (p<0.01). Total nitrogen, water soluble nitrogen, and ripened index of ripened cheese samples showed that protein degradation was higher in ripened Akcakatik cheese samples than in fresh samples.

pH value was in conformity with the findings of Simsek and Gun (2009). These researchers reported that lactic acid and dry substance contents of Akcakatik cheeses were approximately 0.866 and 73.96%, respectively (Simsek and Gun, 2009). In addition, Kirdar (2004) detected the lactic acid and dry substance contents of Akcakatik cheeses at between 0.9-2.16 and 24.5-59.63%, respectively. NaCl results of this study were found to be higher than Simsek and Gun (2009) and similar to Kırdar (2004). Labneh contains 6.42-10.70% fat, 14.12-16.47% milk solids (non-fat), 20.54-24.61% TS, 1.07-1.33% minerals (ash) and a pH range of 3.67-4.05 (Salji et al., 1983). Guler and Uraz (2003) reported that the total nitrogen content of Tulum cheese was found to average 3.34% and its minimum and maximum values were 2.81 and 4.02%, respectively. Dinkci et al. (2012) indicated that the WSN value of Tulum cheese was 0.56% and the ripening index was 16.86%. Oner et al. (2003) reported that the WSN and ripening index of the Tulum cheeses were 0.98 and 26.51%, respectively. Renner (1983) indicated that the ripening index of a cheese is between 10-60%, depending on the cheese type, when its maturation has not completed. Laban zeer produced in Egypt and considered as a nutritious fermented dairy product was found to be rich in essential amino acids and total protein averaged 17.14% and 8.77% soluble N, respectively (Ibrahim et al., 1999). Chemical analysis results of Akcakatik cheese were lower than the results of Tulum, Labneh and Laban cheeses.

Biogenic amine levels of fresh and ripened Akcakatik cheese samples are given in Table 2. Pheniletilamine, putrescine and cadaverine were not determined in any cheese samples. On the other hand, tryptamine was detected on average as 0.03 ± 0.01 and 0.02 ± 0.00 mg/kg, respectively. In general, histamine levels in ripened cheese samples were found to be higher than in fresh cheese samples. Maximum and minimum histamine levels in ripened cheese samples were detected as 5.80 and 44.6 mg/kg, respectively. In addition, tyramine was detected in only three of the ripened cheese samples (D4, D6 and D8) with an average of 0.013 ± 0.01 mg/kg.

Butikofer et al. (1990) carried out studies on Gorgonzola, Emmental, Camembert, Roquefort and Feta cheeses and reported that Gorgonzola cheese was found to have the highest biogenic amine content with 4 mg/kg pheniletilamine, 69 mg/kg putrescine, 120 mg/kg cadaverine, 123 mg/kg histamine and 822 mg/kg tyramine. Durlu-Ozkaya et al. (2000) indicated that pheniletilamine was detected at the highest level (average 33.2 mg/100 g) among the biogenic amine contents in Tulum cheese. It is known that 70-1,000 mg histamine taken orally leads to clinical intoxication (Edwards and Sandine, 1981; Chang et al., 1985). The results of this study showed that histamine levels in fresh and ripened Akcakatik cheese samples were below the toxic level.

As a result of analysis of Akcakatik cheese samples with GC/HS, aroma components detected at their highest amounts are given in Table 3. Volatile flavor compound analysis showed that acetaldehyde, diacetyl, ethanol and ethyl acetate are major carbonyl compounds in Akcakatik cheese samples. Acetaldehyde, diacetyl, ethanol and ethyl acetate in ripened cheese were detected on average as 31.62 ± 3.23, 73.42 ± 57.24, 8.04 ± 7.60 and 60.21 ± 68.66 ppb, respectively. According to statistical analysis, there is no statistically significant difference between the carbonyl compounds of the groups. 2-butanon and acetone were not detected. Carbonyl compounds of the Akcakatik cheese samples in this study were detected at lower levels than the results of Atamer et al. (2004) who analysed carbonyl components in the yogurts.

The composition of α, β and γ-caseins formed from other peptides (%) are given in Table 4. These values were obtained by taking the average of the areas on the gel of two different gel bands. In general, the composition of the samples belonging to α-casein in ripened Akcakatik cheese was detected as 27.30 ± 2.25%. β-casein and γ-casein compositions were detected as 29.61 ± 0.75 and 43.43 ± 2.94%, respectively. When the findings were statistically reviewed, it was observed that the difference among the groups was significant (p<0.01). While no difference was detected between α and β casein ratios of ripened and fresh cheese samples, the difference between data formed from γ-caseins peptides of ripened samples was detected as significant. However, an increase was observed in peptide compositions while α and β casein compositions decreased in the samples during the ripening. The protein slabs obtained from SDS-PAGE showed that all of the Akcakatik cheeses were made from cow’s milk.

Hassan El-Deeb (1988) detected the lowest α-casein ratio as 40.1% in Roquefort cheese and the highest ratio as 55.3% in Kariesh cheese among nine commercial mature cheeses. Hernandez et al. (1988) reported that while the α-casein ratio of soft goat cheeses was detected as 21.1% on the second day of maturation, this ratio decreased to 13.1% at the end of the 28^th day. While α-casein ratio was detected as 61.4% in the külek cheeses on the first day, it was detected as 49.4% at the end of the 90^th day (Dervisoglu and Yazici, 2001). While β-casein ratio is 29.9% in soft cheeses at the beginning of maturation, it decreases to 21.7% at the end of the 28^th day (Hernandez et al., 1988). The decrease in casein compositions (α- and β-casein), which is evidenced by the maturation reported by other researchers, is similar to the casein composition of Akcakatik cheese. Degradation of casein is probably the most important biochemical event during the ripening of most cheese varieties, with a major impact on flavor and texture (Fox, 1989). In this study, γ-caseins and peptides showed an increase at the maturation of Akcakatik cheese. As γ-caseins are fragmentation products of β caseins they increase depending on fragmentation of the β caseins during maturation. It was shown that: β_a contains γ₁-casein [β-casein (29–209)] and the correlated peptide [β-casein (30–209)]; β_b contains γ₂-casein [β-casein (106–209)] and γ₃-casein [β-casein (108–209)] (Mayer, 1996; Giaschi et al., 2001),

Microbiological analysis results of Akcakatik cheese samples are given in Table 5. Microbiological analysis results were not found to have statistically significant differences (p>0.05) between fresh and ripened cheeses. The results of microbiological analysis showed that TAMB counts of fresh and ripened cheese samples ranged from 7.21 to 8.13 and from 7.12 to 8.52 log cfu/g, respectively. When compared with our results, lower TAMB counts were reported by Kırdar et al. (2017) for ripened Akcakatik cheese samples collected from Burdur province in Turkey. On the other hand, some researchers reported a high count of TAMB in traditional Turkish cheeses such as Urfa (Yetişmeyen and Yıldız, 2003), Cokelek (Onganer and Kirbag, 2009) and Aho (Temiz and Kılıç, 2016), as confirmed in this study. Yetişmeyen and Yıldız (2003) found that the average TAMB number of Urfa cheese samples is 1.0x10⁹ cfu/g. Onganer and Kirbag (2009) found that the average count of TAMB in Cokelek cheese samples was determined as 8.49 ± 0.79 log cfu/g. Similarly, Temiz and Kılıç (2016) reported that the TAMB count of Aho cheese, a traditional Turkish cheese made from cow’s milk, ranged from 6.20 log cfu/g to 7.44 log cfu/g, and the highest value was determined as 7.58 log cfu/g. Bacterial contamination of raw milk usually originates from the environment, udders and milking equipment (Rysha et al., 2014). The high TAMB number in both fresh and ripened Akcakatik cheeses is worrying for consumer health because pathogen bacteria can be found within the number assessed as TAMB.

In fresh and ripened Akcakatik cheeses, the average numbers of LAB were detected as 7.28 ± 0.38 and 7.74 ± 0.61 log cfu/g, respectively. LAB play an important role in the production of a variety of dairy products probably through three major biochemical conservations of milk components glycolysis, proteolysis and lipolysis, so contributing to their typical taste and flavor (Leroy and De Vuyst, 2004; Steele et al., 2013; Chen et al., 2017).

Yeast-mould count of fresh and ripened cheese samples ranged from 5.55 to 7.71 and from 6.85 to 8.20 log cfu/g, respectively. These results showed that ripened cheese samples had higher yeast-mould counts than fresh cheese samples. Conversely to our results, yeast-mould count was determined on average to be 2.04 ± 0.19 log cfu/g by Kırdar et al. (2017) for ripened Akcakatik cheese samples. Our results showed that all fresh and ripened Akcakatik cheese samples had higher yeast-mould counts than allowed in the Turkish Food Codex. According to the Turkish Food Codex, a maximum count of 100 cfu/g of yeast-moulds is allowed in cheese and yogurt (Anonymous, 2010). However, high yeast-mould counts were detected in other types of traditional Turkish cheeses such as Urfa (Yetişmeyen and Yıldız, 2003), Carra (Aygun et al., 2005), Cokelek (Onganer and Kirbag, 2009), and Aho (Temiz and Kılıç, 2016), as confirmed in this study.

Coliform group bacteria were detected only in sample K5 (2.00 log cfu/g). In the other samples, coliform group bacteria were found below the level of 1.00 log cfu/g. Escherichia coli was not detected in all samples, as confirmed by Kırdar et al. (2017). The Turkish Food Codex suggests that dairy products should contain no more than 100 cfu/g coliform bacteria, and no E. coli (Anonymous, 2010). Both fresh and ripened Akcakatik cheese samples were found to be legal according to the Turkish Food Codex. Conversely to our results, E. coli was detected in some traditional Turkish cheeses such as Orgu (Turkoglu et al., 2003), Golot (Tuncturk and Ozdemir, 2005), Cokelek (Onganer and Kirbag, 2009) and Beyaz (Sener and Cakici, 2013).

A total of 60 presumptive LAB were isolated from Akcakatik cheese samples. The isolates were found to be Gram positive and catalase negative cocci (55%, 33 strains) or rods (45%, 27 strains). Coccoid-shaped 16 and 17 isolates were selected on M17 agar and NRCLA plates, respectively. All rod-shaped isolates were isolated on MRS agar plates. All isolates failed to grow in MRS broth at 10℃, 15℃, pH 9.6 and presence of 4%, 6.5%, 8% or 10% NaCl. On the other hand, they were grown in MRS broth at 45℃. The saccharide fermentation test results showed that dominant LAB species isolated from Akcakatik cheese samples are Streptococcus salivarius subsp. thermophilus (55%) and Lactobacillus delbrueckii subsp. bulgaricus (31.67%). These results are not surprising because Akcakatik cheese is produced from yogurt. Yogurt is a fermented dairy product produced by the protocooperative action of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus (Tamime and Robinson, 1985; Erkus et al., 2014). In addition, we also isolated five Lactobacillus acidophilus (8.33%), two Lactobacillus delbrueckii subsp. lactis (3.33%) and one Lactobacillus helveticus (1.67%) strains. These species were only isolated from ripened Akcakatik samples. Similar to our results, Roushan et al. (2014) identified Lactobacillus helveticus (14%), Lactobacillus acidophilus (6%) and Lactobacillus delbrueckii subsp. lactis (2%) from Iranian traditional yogurts by phenotypic methods.