PaPer 416 Ital. J. Food Sci., vol. 27 - 2015 - Keywords: rennet paste, chemical and sensory characteristics - EFFECT OF ARTISANAL RENNET PASTE ON THE CHEMICAL, SENSORY AND MICROBIOLOGICAL CHARACTERISTICS OF TRADITIONAL GOAT’S CHEESE C. TRIPALDI*, G. PALOCCI, A. GARAVALDI1, T. BOGDANOVA2 and S. BILEI2 Consiglio per la Ricerca e la Sperimentazione in Agricoltura, Centro di Ricerca per la Produzione delle Carni e il Miglioramento Genetico, Via Salaria 31, 00016 Monterotondo, Roma, Italy 1Fondazione CRPA Studi Ricerche Onlus, Corso Garibaldi 42, 42100 Reggio Emilia, Italy 2Istituto Zooprofilattico Sperimentale delle Regioni Lazio e Toscana, Via Appia Nuova 1411, 00178 Roma, Italy *Corresponding author: carmela.tripaldi@entecra.it AbstrAct In a study using three replicates, Marzolina goat cheese made with artisanal rennet paste from goat kid was compared with cheese made with commercial liquid rennet from calf. samples of fresh cheese were subjected to chemical and microbiological analyses. samples of ripened cheese collected after 50 days of ripening were submitted to chemical and sensory analysis. results of this study show that cheese made with artisanal rennet pastes did not contain pathogenic micro- organisms and that this kind of rennet provided the enzymatic content necessary to achieve the typical characteristics of traditional cheeses. mailto:carmela.tripaldi%40entecra.it?subject= Ital. J. Food Sci., vol. 27 - 2015 417 INtrODUctION In recent years, there has been an increased interest in the safety and promotion of chees- es prepared according to local traditional pro- cesses. the renewed attention is given to arti- sanal rennet, which is considered one of the most important factors affecting the character - istics of some typical Mediterranean cheeses. ArPs usually originate from lamb or kid abo- masa, and farmers directly prepare them for use during cheese making. According to cur- rent legislation, the use of ArPs is allowed af- ter a special derogation from regulation (Ec) n. 852/2004 for foods with traditional charac- teristics (Ec reg. n. 2074/2005). to obtain this derogation, it has been necessary to study their safety and hygienic characteristics. results of a number of studies indicate that traditional cheese-making does not compromise health and hygiene (cOssEDDU and PIsANU, 1980; DEIANA et al., 1980; PIsANU and cOssEDDU, 1982; bAr- zAghI et al., 1997; cAlANDrEllI et al., 1997; IrIgOyEN et al., 2001; MOAtsOU et al., 2004; MOschOPOUlOU et al., 2007; trIPAlDI et al., 2012). Determining the role and influence of ArPs on the sensory characteristics of chees- es is also necessary. Marzolina is a traditional Italian cheese made from the milk of local goat breeds of the latium region, in the centre of Italy and from artisanal kid rennet pastes. Marzolina is characterised by a weigth of about 150 g and high salt content that preserves the cheese for long periods even in a natural room. ArP is traditionally used in the process for MArzOlINA cheese and its particu- lar flavor has been attributed to the use of ren- net pastes (Addis et al., 2005). however factors such as the time and the effort required to pre- pare the rennet at the farm, as well as the low- er demand for strongly flavoured cheeses, have contributed to the rapid replacement of ArP with commercial liquid rennet (clr). the aim of this trial was to study the hygien- ic and health characteristics of cheese from ArP so contributing to the approval of the derogation for using artisanal rennet pastes. Moreover this trial was finalised to evaluate the effect of ren- net pastes, as compared to liquid rennet, on the chemical, and sensory characteristics of Marzo- lina goat cheese. MAtErIAls AND MEthODs Marzolina cheese made with ArP from kid was compared with that made with clr from calf. both cheeses were produced by farmers. two kinds of rennet were used and three repli- cates were performed. the trial was carried out according to the process usually employed by farmers in a small cheese farm. thirty litres of milk for each kind of rennet in each replicate were processed. Abomasa were removed from suckling kids slaughtered at the age of 30–45 days and they were submitted to a drying phase preceded by a salting phase. ten abomasa were ground, merged and then utilised in all replicates of the trials. the commercial liquid calf rennet (Naturen®), used in the trial was produced by chr. hansen’s (Denmark). Data specifying proteolytic activity of clr were supplied by chr. hansen’s. Enzimatic character- istics of ArP were determined according to the following methods. total milk clotting activities of the artisanal rennet pastes were determined according to IsO 23058 IDF 199: 2006 method known as rEM- cAt method. to determine the chymosin and pepsin content, test samples were prepared by dissolving 25 g of rennet paste in 100 g of buffer solution (ch3cOOh/ch3cOONa) at ph 5.5; the samples were centrifuged at 3000 rpm (2189 g, refrigerated centrifuge Alc 4237r, Alc, Milano, Italy) for 30 min at 4°c. the supernatant was analysed as described in the International IDF standard 110b: 1997. chymosin and pepsin enzymes were expressed as a percentage of the sample’s total milk clotting activity. lipase activ- ity was analysed as described in the Food chem- ical codex (1981). characteristics of both types of rennet are shown in table 1. During the study, raw goat milk was utilised and no starter cultures were added. After coagulation at a temperature be- low 35°c, the curd was cut, reduced to small granules and then packed in a small cylindri- cal mould. After about 12 hours from the start of cheese-making, the cheeses were subjected to dry salting and then air-dried in a natural room for one week. Finally, they were packaged under vacuum and stored at 4°c. During each experiment, samples of fresh cheese were collected before dry salting. sam- table 1 - Enzymatic activity of artisanal rennet pastes and commercial liquid rennet. Rennet g or mL of rennet total IMCU × g-1 or mL-1 total IMCU on 100 l of milk chymosin (% RU) chymosin IMCU ILU × g-1 on 100 l of milk of rennet on 100 l of milk of rennet ARP 19 129 2493 95,34 2377 35,76 CLR 20 162 3229 80,00 2583 ARP: Artisanal Rennet paste; CLR: Calf Liquid Rennet. IMCU: International Milk Clotting Unit; RU: Rennet Unit;ILU: International Lipase Units. 418 Ital. J. Food Sci., vol. 27 - 2015 ples of fresh cheese were subjected to chemical analyses, which were conducted in duplicate. Analyses consisted of measurements of mois- ture (IDF, 1986), total nitrogen (tN) (FIl-IDF, 1993), soluble nitrogen (sN) (FIl-IDF, 1991), fat (FIl-IDF, 2001), salt (IDF, 1988), ash (AOAc, 2000), and free fatty acids (FFAs). FFAs (mmol/ kg) were analysed by capillary gas chromatog- raphy (De Jong and badings, 1990) and ex- pressed as mmol/kg to assess each individual FFA independent of its molecular weight. sam- ples of fresh cheese were also subjected to mi- crobiological analysis. All samples were subject- ed to qualitative tests for Salmonella, Listeria monocytogenes and Escherichia coli O157, and to quantitative analyses for L. monocytogenes, sulphite-reducing clostridia, total microbial count at 30°c, coagulase-positive staphylococ- ci, and β-glucuronidase-positive E. coli (tripal- di et al., 2012). samples of cheese ripened for 50 days were also collected and then subject- ed to chemical and sensory analyses. Descriptive sensory analysis was performed to determine the differences in sensory charac- teristics of the two kinds of cheeses. ten pan- elists (5 males and 5 females; between 20 and 50 years of age) were selected and trained in accordance with IsO 8586-1:1993 e IsO 8586- 2:1994 standards. For the laboratory conditions, UNI IsO 8589 standard was followed. the test was carried out according to UNI 10957:2003 standard based on triplicate analysis of each sample. twenty-two descriptors were identified, as fol- lows: 3 visual (color intensity, color omogenei- ty, rind color), 5 olfactory (smell intensity, stable straw smell, lactic smell, vegetable smell, other smell), 5 basic taste and trigeminal sensations (salty, sweet, sour, bitter, piquant), 6 retro-olfac- tory (flavour intensity, stable straw flavour, lac- tic flavour, vegetable flavour, other flavour, per- sistence), and 3 tactile descriptors (adhesive- ness, moisture, firmness). Attributes were rat- ed on a continuous scale of values from 0 to 10 (0 = absence of the intensity, maximum inten- sity = 10). sensory descriptors and their rating scale were defined according to UNI 10957:2003 and European guide for the sensory evaluation of hard and semi-hard cheese standard (bérO- DIEr e al.1997; lAvANchy et al.1994). the glM procedure of sAs software (sAs In- stitute Inc., 2007) was used for statistical anal- ysis of chemical parameters by using the model Y ijl = μ + A i + B j + A i ×B j + E ijl where Y ijl = qualitative characteristics of the cheese A i = fixed effect of the kind of rennet (i = 1 for ArP; i = 2 for clr) B j = fixed effect of days of ripening (j = 1: 1d; j = 2: 50d) E ijl = residual of error Data processing of sensory evaluations was carried out according to UNI 10957:2003 stand- ard. rEsUlts AND DIscUssION Results of proteolytic and lipolytic activity of ARPs and CLR are summarised in Table 1. The clotting characteristics were different: 129.38 to- tal International Milk Clotting Units (IMCU) × g−1 and 95.34% of chymosin for ARP and 162.00 total IMCU × ml−1 and 80.00% of chymosin for CLR. The latter was subjected to a preliminary test and then added to milk in quantity of 19.93 mL/100 l of milk. This led to coagulation in approximate- ly 60 minutes, the clotting time used by farmers. During the experiment, the addition of rennet to 100 l of milk resulted in 2493 and 3229 total IMCU for ARP and CLR, respectively. Chymosin IMCU values, 2377 and 2583 per 100 l of milk for ARP and CLR, respectively, were more similar than that of total IMCU. . The lower activity of to- tal clotting of ARP could be due to dilution of en- zymes caused by the presence of milk in the abo- masa used to produce the rennet (PIRIsI et al., 2007). The high chymosin percentage (95.34 and 80.00) in the rennet paste could be attributed to the completely milk-based diet and/or to complete filling of the stomach with milk when the kid was killed (BUsTAMAnTe et al., 2000; ADDIs et al., 2005). We observed that in our study the kind of diet given to the kid and the status of the aboma- sum before the slaughtering agreed with the con- ditions found by other Authors (BUsTAMAnTe et al., 2000; ADDIs et al., 2005). The international lipase units (ILU) of the rennet paste, 35.76 ILU × g−1, was similar to the average value (36.18 ILU × g−1) obtained from other samples of rennet paste from the same region (TRIPALDI et al., 2012). The chemical characteristics of fresh and rip- ened goat cheese made using ARP and CLR are reported in Table 2. The moisture values were not largely affected by the kind of rennet in both fresh and ripened cheese (69.65 and 42.30% in ARP cheese and 69.81 and 40.87% in CLR cheese). Generally, the cheese moisture depends on the temperature and relative humidity conditions of cheese-making and the ripening conditions (IRI- Goyen et al., 2002); thus, it is difficult to find differences between the two cheeses in which the only change is the type of rennet. We can observe slight differences in protein con- tent between the two kinds of cheeses (10.60 vs 11.13 and 20.98 vs 21.30 in ARP- and CLR-treat- ed fresh cheese and in ARP- and CLR-treated rip- ened cheese, respectively). The fat content of the two kinds of cheese also differed slightly (13.29 vs 14.03% in ARP and CLR of fresh cheese and 28.19 vs 30.31% in ARP- and CLR-treated ripened cheese, respectively). on the contrary, sAnToRo and FACCIA (1998) observed a significant differ- ence in fat content in CanestratoPugliese cheese Ital. J. Food Sci., vol. 27 - 2015 419 made with rennet having different characteristics. They attributed this result to the different aggre- gation states of the casein micelles in the curd. The salt in moisture (s/M) content of samples of fresh cheese was similar (0.79% in ARP-treated cheese and 0.82% in CLR-treated cheese), while s/M content was higher in ARP-treated than in CLR-treated ripened cheese (7.40 vs 6.01%; P < 0.05). The difference in salt content in the two kinds of ripened cheese is probably due to man- ual dry salting, a practice that is subject to large variations. our results show that ripened Marzo- lina cheese has higher salt content in compari- son with the majority of ripened cheeses (2.67 in ARP-treated cheese and –3.13% in CLR-treated cheese), corresponding to about 5% of dry matter. soluble protein as a percentage of total pro- tein was higher in ARP-treated fresh cheese than in CLR-treated fresh cheese (4.96 vs 3.98%). on the contrary, higher values of this proteolysis index were found in ripened CLR-treated cheese in com- parison with those obtained in ripened ARP-treat- ed cheese (10.11 vs 9.12%). During cheese ripen- ing, the higher salt content of ARP relative to the one of CLR cheese may influence negatively the proteolytic process, as observed in Romano type cheese (GUInee and Fox, 1984). similar values of soluble n/total n at pH 4.6 were found in Protected Designation of origin (PDo) sheep cheese Canestrato Pugliese (CoRBo et al., 2001) at 1 and 35 days of ripening (5.59-6.78% and 8.65-11.50%, respectively). Values of the sol- uble n/total n at pH 4.6 (7.87%) of the tradition- al Italian cheese Piacentinu ennese at 2 months of ripening (FALLICo et al., 2006) are lower than those of ripened Marzolina cheese. Differences in all parameters between fresh and ripened samples of both kinds of cheese were ob- served. With ARP and CLR, the moisture of rip- ened cheese significantly decreased compared with that of fresh cheese. The other cheese com- ponents protein, fat, s/M and ash increased sig- nificantly during cheese ripening as a result of the decrease in moisture. The soluble protein/total protein ratio was significantly higher in ripened cheese as result of increased proteolysis (UPAD- HyAy et al., 2004). Table 3 shows the individual and total FFA (TFFA) content of cheese made with ARP com- pared with cheese made with CLR. The TFFA content was higher in ARP-treated cheese than in CLR-treated cheese: 8.94 versus 4.09 mmol× kg-1 and 39.51 versus 36.56 mmol× kg-1 in fresh and ripened cheese, respectively. The difference between fresh and ripened cheese of ARP- and CLR-treated cheeses was significant. short chain free fatty acids (sCFFAs) are the most abundant FFAs in ARP and CLR. similar to TFFAs, sCFFA was present at higher levels in ARP-treated cheese (5.58 vs 2.06 mmol× kg-1 and 21.46 vs 15.75 mmol× kg-1, in fresh and ripened cheese, respectively). The difference was signifi- cant between ripened cheeses made with ARP and with CLR and between fresh and ripened cheese of both ARP- and CLR-treated cheeses. Butyric and capric acids were the most abun- dant FFAs. Levels of butyric acid were the high- est in both kinds of fresh cheese. Levels of capric acid were the highest in both kinds of ripened cheese.. Butyric acid levels were significantly high- er in fresh or ripened ARP cheese than in CLR cheese (2.27 vs 0.81 mmol × kg-1and 4.89 vs 3.85 mmol × kg-1, P ≤ 0.05). Caproic acid only in rip- ened cheese was significantly higher in ARP-treated cheese than in CLR-treated cheese (4.85 vs 3.31 mmol × kg-1). Also Capric acid was higher in ARP ripened cheese than in CLR ripened cheese (8.39 vs 6.05 mmol × kg-1, P ≤ 0.06) The content of all individual fatty acids of both groups of ripened cheeses were higher than that of both groups of the fresh ones (P ≤ 0.01). Levels of medium-chain FFAs (MCFFAs) and long-chain FFAs (LCFFAs) were higher in fresh cheese made with ARP (1.91 vs 1.02 mmol × kg- 1and 1.44 vs 1.00 mmol ×kg-1 for MCFFAs and LCFFAs, respectively) and lower in ripened cheese made with ARP (10.58 vs 12.04 mmol × kg-1 and 7.48 vs 8.77 mmol × kg-1, respectively). The dif- table 2 - composition of fresh and ripened Marzolina goat cheese made using artisanal rennet paste (ArP) and commercial liquid rennet (clr) at 1 day and 1 month of ripening. Fresh cheese Ripened cheese P ARP + ++ CLR + ++ ARP + ++ CLR + ++ SE rennet ripening moisture (%) 69,65 ns a 69,81 ns a 42,30 ns b 40,87 ns b 1,15 NS * protein (%) 10,60 ns b 11,13 ns b 20,98 ns a 21,30 ns a 0,60 NS * soluble protein (% of total protein) 4,96 ns b 3,98 ns b 9,12 ns a 10,11 ns a 0,13 NS * fat (%) 13,28 ns b 14,03 ns b 28,19 ns a 30,31 ns a 0,73 NS * NaCl (% of moisture) 0,79 ns b 0,82 ns b 7,40 a a 6,01 b a 0,44 * * ash (%) 1,66 ns b 1,71 ns b 5,09 ns a 4,36 ns a 0,26 NS * ARP: Artisanal Rennet paste; CLR: Calf Liquid Rennet. + kind of rennet; ++ ripening time. a, b, *: p<0.05. 420 Ital. J. Food Sci., vol. 27 - 2015 table 3 - Free fatty acids (mmol x kg-1) in fresh and ripened Marzolina goat cheese made using artisanal rennet paste (ArP) and commercial liquid rennet (clr) at 1 day and 1 month of ripening. Fresh cheese Ripened cheese P ARP + ++ CLR + ++ ARP + ++ CLR + ++ SE rennet ripening C4:0 2,27 a B 0,81 b B 4,89 a A 3,85 b A 0,33 * ** C6:0 1,28 ns B 0,37 ns B 4,85 a A 3,31 b A 0,31 * ** C8:0 0,60 ns B 0,24 ns B 3,29 ns A 2,50 ns A 0,40 NS ** C10:0 1,42 ns B 0,62 ns B 8,39 ns A 6,05 ns A 0,80 NS ** SCFFAs 5,58 ns B 2,06 ns B 21,46 a A 15,75 b A 1,45 * ** C11:0 0,01 ns B 0,00 ns B 0,04 a A 0,03 b A 0,00 * ** C12:0 0,44 ns B 0,20 ns B 2,42 ns A 1,81 ns A 0,22 NS ** C14:0 0,47 ns b 0,22 ns B 2,75 ns a 3,13 ns A 0,43 NS * C15:0 0,03 ns b 0,02 ns B 0,18 ns a 0,23 ns A 0,03 NS * C16:0 0,94 ns b 0,56 ns B 5,05 ns a 6,69 ns A 0,84 NS * C16:1 0,02 ns b 0,01 ns b 0,13 ns a 0,15 ns a 0,03 NS * MCFFAs 1,91 ns B 1,02 ns B 10,58 ns A 12,04 ns A 1,55 NS ** C17:0 0,02 ns b 0,01 ns B 0,10 ns a 0,13 ns A 0,02 NS * C18:0 0,45 ns b 0,33 ns B 2,15 ns a 2,73 ns A 0,34 NS * C18:1 0,88 ns b 0,63 ns b 4,57 ns a 5,05 ns a 0,89 NS * C18:2 0,06 ns b 0,03 ns B 0,41 ns a 0,56 ns A 0,10 NS * C18:3 0,03 ns b 0,01 ns B 0,23 ns a 0,30 ns A 0,05 NS * LCFFAs 1,44 ns b 1,00 ns B 7,48 ns a 8,77 ns A 1,40 NS * TFFAs 8,94 ns B 4,09 ns B 39,51 ns A 36,56 ns A 4,24 NS ** ARP: Artisanal Rennet paste; CLR: Calf Liquid Rennet. + kind of rennet; ++ ripening time. SCFFAs = short chain free fatty acids MCFFAs = medium chain free fatty acids LCFFAs = long chain free fatty acids. TFFAs = total free fatty acids. a, b, *: p<0.05; A, B, **: p<0.01 ference was significant for both groups of FFAs only between fresh and aged cheeses. The higher butyric acid content of cheese made with ARPs compared with cheese made with CLR is confirmed by the high specificity of the enzy- matic activity of pregastric lipase for sCFAs, es- pecially butyric acid, esterified to the sn-3 posi- tion of triglycerides (Pitas and Jensen, 1970; Kim and Lindsay, 1993). FonTeCHA et al. (2006) found higher butyric acid content in spanish goat cheese made with rennet paste compared with cheese made with CLR. The higher concentration of capric acid com- pared with butyric acid in ripened cheeses was ob- served in other goat cheeses (BUFFA et al., 2001; PoVeDA and CABezAs, 2006; ATAsoy and TURKo- GLU, 2009). According to BUFFA et al. (2001), the capric acid content of cheese from goat milk in- creased during ripening while butyric acid varied slightly from the start to the end of ripening. The small increase in butyric acid content was prob- ably due to its metabolic conversion to aromatic compounds (BUFFA et al., 2001). sensory attributes of each kind of cheese are shown in Fig. 1. Mean values of the following sensory descriptors were significantly different Fig. 1 - sensory profile of rip- ened Marzolina goat cheese made using artisanal rennet 14 pastes (ArP) and commer- cial liquid rennet (clr). Ital. J. Food Sci., vol. 27 - 2015 421 (P<0.05) in the two kinds of cheese (ARP vs CLR, respec- tively): color omogeneity (6.6 vs 7.1), rind color (3.7 vs 4.0), stable straw smell (4.4 vs 4.0), lactic smell (4.2 vs 4.6), other smell (3.3 vs 2.8), salty (4.2 vs 3.6), sweet (2.9 vs 3.2), sour (3.6 vs 3.3), bitter (2.9 vs 2.6), piquant (3.9 vs 3.3), flavour intensity (7.1 vs 6.8), stable straw flavour (4.7 vs 4.5), lactic flavour (4.4 vs 4.8), other flavour (3.8 vs 3.4), persistence (7.2 vs 6.7), adhesiveness (4.3 vs 5.0), moisture (4.1 vs 4.5) and firmness (5.7 vs 4.9). There was no significant interaction between assessor and replicate, suggesting high repeatability of panellist as- sessment in the three replicates. sample-replicate and sam- ple-panellist interactions were not significant, showing ei- ther homogeneity of samples in the three replicates or good agreement among panellist assessments during sensory evaluation. Analysis of the eyes and slits in cheese showed a higher number of samples made with ARPs having these defects compared with those in cheese samples made with CLR (82 vs 57). The basic tastes salty, acid, bitter and piquant were more pronounced in cheeses made with ARP than in cheeses made with CLR. In addition, the smell and flavour of cow- shed were more marked in cheese made with ARP than in cheese made with CLR. Its texture was firmer, less tacky and less moist than the latter kind of cheese. The odour and flavour given by lactic acid to cheeses made with CLR are more dominant than in cheeses made with ARP. Gener- ally, the sweet, salty and sour attributes of taste were less pronounced. The texture was tackier, more moist and less firm than that of cheese made with ARP. It is noteworthy that basic tastes salty and piquant, which are more pronounced in cheese made with ARP than in cheese made with CLR, agree with the chemical results. As reported by Addis et al. (2005), cheese made with ARPs has a major amount of butyric acid, which may be responsible for piquancy in cheese (Rennet paste has been associated with piquancy or pungency and with char- acteristic flavours of certain cheeses from the Mediterra- nean basin (AnIFAnTAKIs, 1976; neLson et al., 1977; Woo and LInDsAy, 1984; BATTIsToTTI and CoRRADInI, 1993; BARzAGHI et al., 1997; CALAnDReLLI et al., 1997). In a study on Idiazabal cheese (eTAyo et al., 2006), cheese made with lamb rennet pastes showed higher butyric acid content and received higher scores compared to cheese made with commercial liquid lamb rennet. . A larger number of eyes and slits in cheese made using ARPs was also observed by FeRRAnDInI et al. (2012). This could be attributed to different textural properties (FeRRAnDInI et al., 2011) of cheese made with the two kinds of rennet. In fact, results of microbiological analy- ses carried out on ARP used in this study (TRIPALDI et al., 2012) exclude the presence of microorganisms mark- ers of hygiene characteristics including germs causing mi- crobiological spoilage in cheese. Table 4 displays the microbiological characteristics of the cheese. Salmonella and L. monocytogenes, pathogens con- sidered as health markers (Reg. Ce 2073/2005), were un- detected by qualitative analyses. β-Glucuronidase-positive E. coli and coagulase-positive staphylococci, which include Staphylococcus aureus, are considered as hygiene mark- ers (Reg Ce 2073/2005). The maximum count tolerated for coagulase-positive staphylococci in cheese from raw milk is 105cfu × g-1. samples with more elevated counts must be ta b le 4 - M ic ro b io lo g ic a l ch a ra ct er is ti cs o f fr es h M a rz o li n a g o a t ch ee se m a d e u si n g a rt is a n a l re n n et p a st e (A r P ) a n d c o m m er ci a l li q u id r en n et ( c l r ). Sa m pl e Re pl ic at e Sa lm on el la s pp Li st er ia m on oc yt og en es E. c ol i E. c ol i C oa gu la se p os iti ve En te ro ba ct er ia ce ae Su lp hi te re du ci ng To ta l m es op hi lic (in 2 5 g) (in 2 5 g) (c fu × g -1 ) O :1 57 st ap hy lo co cc i (c fu × g -1 ) cl os tr id ia co un t (in 2 5 g) (c fu × g -1 ) (c fu × g -1 ) (c fu × g -1 ) A R P 1 ne ga tiv e ne ga tiv e <1 0 ne ga tiv e <1 0 <1 0 <2 2. 6x 10 7 C LR 1 ne ga tiv e ne ga tiv e <1 0 ne ga tiv e <1 0 <1 0 <2 3. 0x 10 8 AR P 2 ne ga tiv e ne ga tiv e <1 0 ne ga tiv e <1 0 3. 5x 10 3 <2 9. 3x 10 5 C LR 2 ne ga tiv e ne ga tiv e <1 0 ne ga tiv e <1 0 4. 8x 10 3 <2 1.3 x1 06 AR P 3 ne ga tiv e ne ga tiv e 10 0 ne ga tiv e <1 0 6. 2x 10 3 <2 4. 9x 10 6 C LR 3 ne ga tiv e ne ga tiv e <1 0 ne ga tiv e <1 0 <1 0 <2 1.5 x1 05 422 Ital. J. Food Sci., vol. 27 - 2015 analysed for staphylococci enterotoxins. Counts of coagulase-positive staphylococci in all samples were lower than the detection limit of the meth- od (10 cfu × g-1). The count of β-glucuronidase- positive E. coli was 100 cfu × g-1 in only one sam- ple, but lower than the detection limit in other samples (10 cfu× g-1). Qualitative analysis for E. coli o157 in our samples gave negative results. sulphite-reducing clostridia is another group of microbial pathogens considered as hygiene mark- ers, but legislation has not established permissible levels for these in food. Their levels in all samples were lower than the detection limit of the meth- od (2 cfu × g-1). The mean total mesophilic count in our sam- ples was 5.6 × 107cfu× g-1. Microbiological anal- yses of the cheese samples confirmed the re- sults obtained during the monitoring of some ARPs collected in the same region of Marzolina production (TRIPALDI et al., 2012). similar re- sults were obtained in three PDo raw ewe milk cheeses from spain, as Manchego, Idiazabal and zamorano cheese (eTAyo et al., 2006), where the hygienic quality of cheeses made with lamb rennet paste is comparable to that of cheeses manufactured with non-paste commercial ren- net. Another study on Idiazabal did not detect E. coli, Clostridium, Salmonella or L. monocy- togenes, and levels for other microorganisms were below the limits of the european legisla- tive standards for cheese manufactured with raw milk (GIL et al., 2007). the results of our study show that treat- ment with ArPs did not favour the growth of microbial pathogens and that ArPs provided the enzymatic content necessary to achieve the typical characteristics of traditional cheeses. cONclUsIONs butyric acid was the main marker of chees- es made with ArPs because of the high speci- ficity of enzymatic activity of pre-gastric lipase for butyric acid. results of sensory evaluation show that the piquant flavours as well as the odour and flavour of cowshed were more pro- nounced in cheeses made with ArP, confirm- ing the results for other Mediterranean chees- es made with rennet paste. therefore, the use of ArPs provided the enzymatic content neces- sary to achieve the typical characteristics of tra- ditional cheeses. At the same time, this kind of rennet did not favour the growth of microbial pathogens in cheese. AcKNOWlEDgEMENts this work was funded by the Agenzia regionale per lo svi- luppo e l’Innovazione dell’Agricultura del lazio, Italy (Ar- sIAl) and to the ruggieri farm from Itri (lt) for providing the Marzolina cheese samples. rEFErENcEs Addis M., Piredda g., Pes M., Di salvo r., scintu M.F. and Pirisi A. 2005. Effect of the use of three different lamb paste rennets on lipolysis of the PDO Pecorino romano cheese. Int. Dairy J. 15: 563. Addis M., Piredda g. and Pirisi A. 2008. the use of lamb paste rennet in the traditional sheep cheese production small rum. res. 79: 2. Anifantakis E. 1976. Influence d’unepresured’agneausur la qualité du fromageKefalotyri. lait, 56: 76. Association of Official Analytical chemists international. 2000. 17th Ed. Official Methods of Analysis, gaithers- burg, MD, UsA. Atasoy A.F. and türkoğlu h. 2009. lipolysis in Urfa cheese Produced from raw and Pasteurized goats’ and cows’ Milk with Mesophilic or thermophilic cultures During ripening. Foodchemistry 115: 71. barzaghi s., Davoli E., rampilli M. and contarini g. 1997. la lipolisi nel formaggio Provolone: ruolo del caglio in pas- ta. sci. tecn. latt.-cas. 48: 146. battistotti b. and corradini c. 1993. Italian cheese. In: Fox P.F. (Ed.), cheese: chemistry, physics and microbiological, vol 2, 2nd edition, chapman and hall, london, UK, 221. bérodier F., lavanchy P., zannoni M., casals J., herrero l. and Adamo c. 1996. guide d’évaluation Olfacto-gusta- tive des Fromages à Pâte Dure et semidure. Poligny: AIr 2039 gEcOtEFt. IsbN 295-10884-0-X.buffa M., guamis b., Pavia M. and trujillo A.J. 2001. lipolysis in cheese made from raw, pasteurized or high-pressure-treated goats’milk. Int. Dairy J. 11: 175. bustamante M., chavarri F., santisteban A., ceballos g., hernandez I., Miguelez M.J., Aranburu I., barron l.J.r., virto M. and De renobales M. 2000. coagulating and lip- olytic activities of artisanal lamb rennet pastes. J. Dairy res., 67, 393. calandrelli M., rubino r., Masoero g., clementi F., Morone g. and Pizzillo M. 1997. Effetto della tecnica di preparaz- ione del caglio di capretto sulle sue caratteristiche mi- crobiologiche e sulla composizione chimica del formaggio caprino semicotto. sci. tecn. latt.-cas. 46: 343. commission regulation (Ec) No 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs. corbo M.r., Albenzio M., De Angelis M., sevi A. and gobbet- ti M. 2001. Microbiological and biochemical properties of canestratoPugliese hard cheese supplemented with bifi- dobacteria. J. Dairy sci. 84, 551. cosseddu A.M. and Pisanu s. 1980. sulla presenza di sta- filococchi patogeni nel latte ovino industriale e nei cagli di agnello e capretto prodotti in sardegna. Il latte 5: 203. Deiana P., Farris g.A., Fatichenti F., carini s., lodi r. and todesco r. 1980. Impiego di caglio di agnello e di capret- to nella fabbricazione di formaggio Fiore sardo: aspetti microbiologici e tecnologici. Il latte 5: 191. De Jong c. and badings h.t. 1990. Determination of free fatty acids in milk and cheese. J. high resol. chromat. 13: 94. Etayo I., Pérez-Elortondo J., gil P.F., Albisu M., virto M., conde s., rodriguez barron l.J., Najera A.I., gomez- hidalgo M.E., Delgado c., guerra A. and de renobales M. 2006. hygienic quality, lipolysis and sensory prop- erties of spanish Protected Designation of Origin ewe’s milk cheeses manufactured with lamb rennet pastes. lait 86: 415. Fallico v., tuminello l., Pediliggieri c. horne J., carpino s. and licitra g. 2006. Proteolysis and microstructure of Pi- acentinuEnnese cheese made using different farm tech- nologies. J.Dairy sci., 89: 37. Ferrandini E., lopez M.b., castillo M. and laencina J. 2011. Influence of an artisanal lamb rennet paste on proteoly- sis and textural properties of Murcia al vino cheese. Food chem. 124, 583. Ferrandini E., castillo M., de renobales M., virto M.D., gar- rido M.D., rovira s. and lopez M.b. 2012. Influence of an artisanal lamb rennet paste on the lipolytic profile of Murcia al vino cheese. J. Dairy sci. 95: 583. Ital. J. Food Sci., vol. 27 - 2015 423 FIl-IDF. 1991. chemical methods for evaluating proteolysis in cheese maturation, No. 261. brussels, belgium: Inter- national Dairy Federation. FIl-IDF. 1993. Milk. Determination of nitrogen content. Kjel- dahl method; standard 20b. Annex I: Modified proce- dure for milk products. brussels; belgium: Internation- al Dairy Federation. FIl-IDF 172: 2001 / IsO 14156:2001- Milk and Milk prod- ucts-Extraction methods for lipids and liposoluble com- pounds. Foltman b. 1993. general and molecular aspects of rennets. In: cheese: chemistry, Physics and Microbiology. vol. 1, pp. 37-88 (Ed. PF Fox). london: chapman and hall. Fontecha J., castillo I., blasco l., Alonso l. and Juarez M. 2006. Effect of artisanal kid rennet paste on liplysis in semi-hard goat cheese. Food chem. 98: 253. gil P.F., conde s., Albisu M., Pérez-Elortondo J., Etayo I., virto M. and de renobales M. 2007. hygienic quality of ewes’ milk cheeses manufactured with artisan-produced lamb rennet pastes. J. Dairy res. 74: 329. guinee t.P. and Fox P.F. 1984. studies on romano type cheese: general proteolysis. Irish J. Food sci. technol. 8: 105. International Dairy Federation. 1986. cheese and processed cheese products. Determination of dry matter. IDF stand- ard No. 4. FIl-IDF, brussels, belgium. International Dairy Federation. 1988. Determination of dry salt content. IDF standard No. 12b. FIl-IDF, brussels, belgium. Irigoyen A., Izco J.M., Ibàñez F.c. and torre, P. 2001. In- fluence of rennet milk clotting activity on the proteolyt- ic, and sensory characteristics of an ovine cheese. Food chem. 72: 137. Irigoyen A., Izco J.M., Ibàñez F.c. and torre, P. 2002. In- fluence of calf or lamb rennet on the physicochemical, proteolytic, and sensory characteristics of an ewe’s-milk cheese. Int. dairy J., 12: 27. IsO 8586-1 – sensory analysis- general guidance for selec- tion, training and monitoring of assessors. Part 1: select- ed assessors (first edition 1993-03-15). IsO 8586-2 – sensory analysis- general guidance for selec- tion, training and monitoring of assessors. Part 2: Ex- perts (first edition 1994-04-15). Kim h.J. and lindsay r.c. 1993. release of the volatile branched-chain and other fatty acids from ruminant milk fats by various lipases. J. Dairy sci. 76: 677. lavanchy P., berodier F., zannoni M., Noel y., Adamo c., squella J. and herrero l. 1993. l’évaluation sensori elle de la texture des fromages a pâte dure ou semidure. Etude interlaboratoires. lebensmittel Wissenschafi und tech- nologie 26 59-68. Moatsou g., Moschopoulou E., georgala A., zoidou E., Kandarakis I., Kaminarides s. and Anifantakis E. 2004. Effect of artisanal liquid rennet from kids and lambs abo- masa on the characteristics of Feta cheese. Food chem. 72: 237. Moschopoulou E., Kandarakis I. and Anifantakis E. 2007. characteristics of lamb and kid artisanal liquid ren- net used for traditional Feta cheese manufacture. small rum. res. 72: 237. Nelson J.h., Jensen r.g. and Pitas r.E. 1977. Pregastric esterase and other oral lipases. J. Dairy sci, 60: 327. Pisanu s. and cosseddu A.M. 1982. sopravvivenza di “sal- monella abortusovis” e brucella melitensis” nel caglio di capretto. Il latte 7: 729. Pirisi A., Pinna g., Addis M., Piredda g., Mauriello r., De Pascale s., caira s., Mamone g., Ferranti P., Addeo F. and chianese l. 2007. relationship between the enzy- matic composition of lamb rennet paste and proteolyt- ic, lipolytic pattern and texture of PDO Fiore sardo ovine cheese. Int. Dairy J. 17: 143. Pitas r.E. and Jensen J.h. 1970. Action of pregastric ester- ase on synthetic tryglicerides containing butyric acid. J. Dairy sci. 53: 1083. Poveda J. M. and cabezas l. 2006. Free fatty acid compo- sition of regionally-produced spanish goat cheese and relationship with sensory characteristics”. Food chem- istry, 95: 307. reg. (Ec) n. 2074/2005 laying down implementing meas- ures for certain products under regulation (Ec) n. 853/2004, 854/2004, 882/2004, 852/2004, 853/2004 and 854/2004. santoro M. and Faccia M. 1998. Influence of mould size and rennet on proteolysis and composition of canestratoPug- liese cheese. Ital. J. Food sci. 10: 217. sAs Institute Inc. 2007. sAs/stAt release 9.12. cary, Nc, UsA. tripaldi c., Palocci g., bilei s., bogdanova t., scintu M.F. and Addis M. 2012. Physical, chemical, enzymatic and microbiological characteristics of artisanal rennet pastes from the centre of Italy. Ital. J. Food science, 24: 70. UNI IsO 8589 2001. Analisi sensoriale - criteri generali per la progettazione di locali destinati all’analisi. UNI 10957: 2003 Analisi sensoriale- Metodo per la defi- nizione del profilo sensoriale degli alimenti e delle bev- ande. Upadhyay v.K., Mcsweeney P.l.h., Magboul A.A.A. and Fox P.F. 2004. Proteolysis in cheese during ripening. In: “cheese: chemistry, physics and microbiology. vol. 1. general aspects. Fox P.F., Mcsweeney P.l.h., cogan t.M., guinee t.P. Ed. Applied science; london, 207. Woo A.h. and lindsay r.c., 1984. concentration of major free fatty acids and flavour development in Italian cheese varieties. J. Dairy sci. 67, 9. Paper Received May 29, 2014 Accepted December 1, 2014