Evolution and antimicrobial resistance of enterococci isolated from Pecorino and goat cheese manufactured on-farm in an area facing constraints as per EU Regulation 1305/2013 in Umbria, Italy.

The latest EU regulation on geographical indications (EU Regulation No. 1151/2012) has introduced a set of new tools for the protection and enhancement of food products in rural areas, under the group name of optional quality term (OQT). The Commission Delegated EU Regulation, No. 665/2014, regulated the conditions for the use of the optional quality term mountain product (MP), to support the implementation of a mountain value chain. This new tool is aimed at promoting local development, maintaining the economic activities in mountain areas, and redistributing wealth, whilst, at the same time, promoting the territory. Pecorino and goat cheeses are typical Italian cheeses made usually with whole raw ewe's or raw goat's milk, without starter culture addition. In an attempt to characterize these productions, the aim of this study was to investigate the evolution of enterococci during the production and ripening of Pecorino cheese made in three different farms, located in Umbria, Italy in areas facing natural or other specific constraints as stipulated by Regulation 1305/2013 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD). Enterococci are enteric organisms which are commonly isolated from ewe and goat's milk production in Umbria, Italy. Counts of enterococci in raw milk ranged from 1.75 for ovine milk to 3.62 for ewe milk and a marked reduction was observed after thermization especially in ovine milk. Out of 100 isolates, 69 were E. faecium, 23 E. durans, 8 E. faecalis and 2 E. casseliflavus and the distribution of species between farms and between samples showed a prevalence of E. faecium in ovine farms and E. durans in ewes farms, with an equal dis-tribution between samples. High percentages of susceptible isolates were found for amoxicil-lin/clavulanic acid, ampicillin, chloramphenicol, sulphamethoxazole, sulphamethoxazole/ trimethoprim, ticarcillin, vancomycin. A high prevalence of resistant strains (>30%) was ob-served for amikacin, ciprofloxacin, ceftriaxone, kanamycin, tetracycline. A comparison of this re-sults with those of previous works on similar dairy products revealed high levels of resistance to antimicrobials which needs to be addressed. ©Copyright: the Author(s).

Introduction

One of the roles the EU covers is that of creating an equilibrium amongst the dimensions of governance, embedding, and marketing, in order to define policies and strategies to manage productive activities. One such activity is to define strategic choices for the promotion of European foodstuffs. In this context, the latest EU regulation on geographical indications (EU Regulation No. 1151/2012) has introduced a set of new tools for the protection and enhancement of food products in rural areas, under the group name of optional quality term (OQT). The Commission Delegated EU Regulation, No. 665/2014, regulated the conditions for the use of the optional quality term «mountain product» (MP), to support the implementation of a mountain value chain. This new tool aimed at promoting local development, maintaining the economic activities in mountain areas and redistributing wealth, whilst, at the same time, promoting the territory (Bonadonna et al. 2017). Umbria, Italy is a region with several areas facing natural or other specific constraints as stipulated by Regulation 1305/2013 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD).

Pecorino and goat cheeses are typical cheeses produced in these areas, made usually with whole raw ewe’s or raw goat’s milk, without starter culture addition. Thus, only bacteria from milk contribute to ripening changes in the cheese (Clementi et al., 1998b). Enterococci represent the predominant microbiota of these productions (Cenci-Goga et al., 1995; Clementi et al., 1998b; Serio et al., 2007; Serio et al., 2010; Litopoulou-Tzanetaki and Tzanetakis, 2011; Russo et al., 2018).

Enterococci are gram-positive bacteria and may fit within the general definition of lactic acid bacteria. With regard to safety and according to the Qualified Presumption of Safety (QPS) list from the European Food Safety Authority (EFSA) (https://www.efsa.europa.eu/en/topics/topic/qps), Enterococcus species are neither rec ommended for the QPS list (Hazards, 2017) nor have GRAS (Generally Regarded As Safe) status (Huys et al., 2013), in spite of recent scientific knowledge allowing differentiation of commensal from pathogenic strains (Hanchi et al., 2018; Graham et al., 2020; Dapkevicius et al., 2021). Modern classification techniques resulted, back in 1980s, in the transfer of some members of the genus Streptococcus, notably some of the Lancefield’s group D streptococci, to the new genus Enterococcus (Schleifer and Kilpper-Bälz, 1984). Enterococci can be used as indicators of faecal contamination and have been implicated in outbreaks of foodborne illness. On the other hand, they have been ascribed a beneficial or detrimental role in foods (Hanchi et al., 2018; Graham et al., 2020; Dapkevicius et al., 2021). In processed meats, enterococci may survive heat processing and cause spoilage, though in certain cheeses the growth of enterococci contributes to ripening and development of product flavour (Clementi et al., 1998b; Dapkevicius et al., 2021). Some enterococci of food origin produce bacteriocins that exert anti-Listeria activity (Franz et al., 2007). Enterococci are used as probiotics to improve the microbial balance of the intestine, or as a treatment for gastroenteritis in humans and animals (Cenci- Goga et al., 2015; Hanchi et al., 2018; Graham et al., 2020; Grispoldi et al., 2020). On the other hand, enterococci have become recognized as serious nosocomial pathogens causing bacteraemia, endocarditis, urinary tract and other infections. This is in part explained by the resistance of some of these bacteria to most antibiotics that are currently in use (Dapkevicius et al., 2021). Resistance is acquired by gene transfer systems, such as conjugative or nonconjugative plasmids or transposons. It appears that foods could be a source of vancomycinresistant enterococci (Cenci-Goga et al., 2004; Werner et al., 2013; Graham et al., 2020).

Regardless these considerations Enterococci are commonly found in milk and cheese (Tsanasidou et al., 2021) and there are two divergent opinions about the presence of enterococci in cheese. One is that enterococci should be considered more suitable than other groups commonly used as indicators of unhygienic procedures in food processing and handling (e.g. Enterobacteriaceae). This is related to their high heat resistance and salt tolerance (Sojeong et al., 2019). The other opinion is that enterococci have a possible contribution to the ripening of cheese due to their lipolytic, proteolytic and caseinolytic activities (Marilley and Casey, 2004). Moreover it has been suggested that enterococci contribute to flavour due to the attitude of produce acetoin, diacetyl and acetaldehyde (Marilley and Casey, 2004).

In an attempt to characterize cheese production from a mountain area in Umbria, Italy, and define safety aspects, the aim of this work was to investigate the evolution of enterococci during the production and ripening of Pecorino cheese made in an area facing natural and specific constraints as per EU Regulation 1305/2013, with two different cheesemaking processes from sheep and ewe milk and characterize Enterococcus spp. isolates for antibiotic susceptibility all along the cheesemaking and ripening process.

Materials and Methods

Study area

The three farms are located in Umbria, Italy in an area that extends from latitude 42.977715°N to 43.227086°N and from longitude 12.141079°E to 12.310128°E (Figure 1). The area, partly hilly and mountainous and partly flat and fertile owing to the valley of the Tiber River, is separated from the central Apennines by the Tiber Valley (“Val Tiberina”). The farms are located in areas facing natural or other specific constraints as stipulated by Regulation 1305/2013 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD). For the three farms, according to the cited Commission Delegated EU Regulation, No. 665/2014, most of the conditions for the use of the optional quality term «mountain product» (MP), to support the implementation of a mountain value chain, are present and might be the base for a regional development for rural areas. These conditions are stipulated by Regulation 1305/2013, article 32: mountain areas or areas, other than mountain areas, facing significant natural constraints. These are areas characterized by a considerable limitation of the possibilities for using the land and by an appreciable increase in production costs due to the existence, because of altitude, of very difficult climatic conditions, the effect of which is to substantially shorten the growing season. At a lower altitude, the presence over the greater part of the area in question of slopes too steep for the use of machinery or requiring the use of very expensive special equip ment, or a combination of these two factors, where the constraints resulting from each taken separately are less acute but the combination of the two gives rise to an equivalent constraint.

Figure 1.

Geographical position of the three farms located in Umbria, Italy.

General

The experiment was carried out on three farms in Umbria, farm A, has 500 Sardinian ewes, farm B has 200 Sardinian ewes, farm C has 80 Saanen goats. In farm A and B milk is collected with a milking machine and automatically filtered before the cheesemaking. In farm C goats are hand milked and the milk is filtered through a linen cloth.

Milk from each farm was collected, analysed, and subsequently used to produce Pecorino or goat cheese. On each farm two different cheesemaking processes were carried out from raw milk and from heat-treated milk to which the autochthonous cultures were added as a starter. Bacterial strains in the formulation used for all cheesemaking in this work were: Lactococcus lactis ssp. lactis, reference strain n. 340; L. lactis ssp. lactis, strain n. 16; Lactobacillus casei ssp. casei, strain n. 208. This formulation was used for all. The morphological, biochemical and physiological characterization, the growth curves at several temperatures, including refrigeration conditions, the acidifying activity and their ability to improve palatability of cheeses have been reported by the authors in previous papers (Cenci- Goga et al., 2012; Sechi et al., 2014b, 2014a; Cenci-Goga et al., 2015; Cenci- Goga et al., 2016; Cenci-Goga et al., 2018).

Each cheesemaking was replicated three times in three consecutive days on each farm. According to this experimental procedure there were 18 manufacturing processes: 9 from raw milk and 9 from heattreated milk.

Cheese manufacturing

Cheese was made from ewes and goats milk obtained from both evening and morning milking. The main steps of cheesemaking process were: milk coagulation in a tinned copper vat occurred at 37°C within about 30’ by adding liquid calf rennet (Lima, Perugia, Italy, titre 1:10.000). The curd was cut into nut-sized granules (10 to 20 mm), then stirring for 5 minutes, heated at 42-43°. After a pause of 5 to 10 minutes the curd was put into frames (20 cm diameter by 8 cm high), pressed by hand for a few seconds, drained for 18 to 20 hours and salted in brine (20% NaCl w/v, at 12-15°C for 30 h). The cheeses were ripened in non-conditioned storage rooms at 12-15°C and 83-87% RH for 60 days (40 days for goat cheeses).

Heat treatment (where applied) included a heating step before the rennet addition. Raw milk was heated to 65-66°C for 3-4 min in a double wall stainless steel vat. The thermic cycle 55-65-55°C (during heating and cooling) took about 14 min. After further cooling to 42°C, the starter was added at 42°C at a final concentration of 2% (as a full-coagulated 24h culture in sterile milk).

Sampling

The following samples were taken on each farm: raw milk, heat-treated milk (where applied), curd, 7 days cheese, 30 days cheese, full-ripened cheese. Triplicate samples were collected, transported to the laboratory in chilled containers and analysed on the same day. Sampling was made according to ISO 5538:2004 (ISO CH 2004).

Bacterial counts

The following groups were evaluated according to the methodology described previously by Cenci-Goga et al. (2018), briefly, total viable count: pour plates of Plate Count Agar (Difco, Detroit, Mi, USA), were incubated at 30°C for 72 h; all colonies were counted; enterobacteriaceae: pour plates of Violet Red Bile Agar (Difco) were incubated at 30°C for 48h; all pink to red colonies, irrespective of diameter or presence/absence of zone of precipitation were counted as presumed enterobacteriaceae; enterococci: surface-inoculated plates of Barnes Agar (Biolife, Milano, Italy) were incubated at 44°C for 72 h; all pink, red or maroon colonies, irrespective of diameter, were counted as presumed enterococci.

Isolation and identification of enterococci

Two to five colonies from each sample were sub-cultured from Barnes medium into brain heart infusion broth (BHI, Difco) at 37°C for 24h and then tested for the following characteristics: cell morphology after Gram staining, presence of catalase, growth in bile-esculin-azide agar (Coccosel agar, BioMérieux, Marcy-l’Etoile, France) at 37°C, growth in 6.5% NaCl BHI agar (Difco) at 37°C, growth in the presence of 4% bile salts (Coccosel agar, Biolife, with added 4% bile salts) at 37°C, haemolysis type on tryptic soy agar (Biolife) to which 5% of ram blood was added, at 37°C.

Complementary biochemical tests were performed on colonies grown on blood agar using the API 20 STREP (BioMérieux). Computer program Apiweb (BioMérieux) was used for the results.

Antibiotics susceptibility

Enterococcus spp. isolates were tested for antimicrobial susceptibility against a panel of 12 antimicrobials by the disk diffusion method (Kirby Bauer Test) as described by the Clinical and Laboratory Standards Institute (CLSI 2018, 2021). The following antimicrobials were tested: amikacin 30 mg, amoxicillin/clavulanic acid 30 (20 + 10) mg, ampicillin 10 μg, ceftriaxone 30 μg, chloramphenicol 30 μg, ciprofloxacin 5 μg, kanamycin 30 μg, sulphamethoxazole 25 μg, sulphamethoxazole/ trimethoprim 25 μg, tetracycline 30 μg, ticarcillin 75 μg, vancomycin 30 μg. This antimicrobial panel was selected to test the major groups of antimicrobials. Briefly, frozen isolates were thawed and cultured in BHI broth (Bio-Rad) at 35 to 37°C for 24 h. A portion of the culture broth was inoculated into 6 mL of 0.9% sterile physiological saline solution until a turbidity of 0.5 McFarland was reached (1.0 for vancomycin (Wongthong et al. 2015)). Using a sterile swab, the solution was spread on Muller-Hinton agar plates (Oxoid). Antimicrobial disks (Oxoid) were placed on Muller-Hinton agar plates which were incubated at 37°C for 18 to 24 h. At the end of incubation, the diameters of the growth inhibitory zones were measured, and these were interpreted using specific CLSI tables whereby the bacterium is classified as susceptible, intermediately susceptible or resistant (CLSI 2021).

pH and water activity

To measure the pH, a S40 Seven-Multi digital pH-meter (Mettler-Toledo Italia, Novate Milanese, Italy) was used after mixing 10 g of milk, curd or cheese with 90 ml of distilled water. A dew-point hygrometer HygroLab 3 (Rotronic, Huntington, NY, USA) was used to measure water activity (aw).

Results and discussion

Results of the determination made on Pecorino and goat cheese during manufacture and ripening are given in Tables 1-6.

Changes in total viable counts and Enterobacteriaceae

The differences observed on the three farms (Table 1) comprised a range of variability which is common to on-farm cheesemaking in the study area (Cenci-Goga et al., 1995; Clementi et al., 1998b; Avellini et al., 1999a) and elsewhere (Marilley and Casey, 2004; Litopoulou-Tzanetaki and Tzanetakis, 2011; Bonadonna et al., 2017; Russo et al., 2018; Dapkevicius et al., 2021).

Table 1.

Evolution of microbiota during production and ripening of farm manufactured cheeses (log cfu ml-1 or g-1) and aw for full ripened cheeses.

	Aerobic viable count		Coliforms		Enterococcus spp.
	Mean	Sd	Mean	Sd	Mean	Sd
Farm A – raw milk
Raw milk	5,9	0,34	3,93	0,42	2,74	2,81
Curd	7,42	0,23	4,41	0,34	3,47	3,08
1 week	10,04	0,79	5,76	0,21	5,26	2,74
1 month	8,7	0,14	1,69	2,93	4,42	1,95
2 months (aw 0.925 – sd 0.005)	7,1	0,64	0.00	0.00	3,66	1,1
Farm A – heat treated milk
Raw milk	5,9	0,34	3,93	0,42	2,74	2,81
Heat treated milk	3,06	0,25	0.00	0.00	0,33	0,58
Curd	7,58	0,89	3,33	0,64	4,6	1,8
1 week	8,56	2,03	3,88	3,4	5,54	2,92
1 month	9,42	0,4	1,37	2,37	6,27	1,5
2 months (aw 0.936 – sd 0.006)	9,13	0,57	0.00	0.00	4,29	1,12
Farm B – raw milk
Raw milk	6,01	0,31	4,7	0,36	1,75	3,03
Curd	7,19	0,66	5,34	1,49	4,72	0,59
1 week	9,13	0,38	5,3	0,45	5,74	1,1
1 month	8,19	0,44	1,78	0,88	5,33	1,01
2 months (aw 0.924 – sd 0.001)	7,44	0,85	0.00	0.00	3,66	1,1
Farm B – heat treated milk
Raw milk	6,01	0,31	4,7	0,36	1,75	3,03
Heat treated milk	4,56	0,91	0,48	0,83	1,79	1,72
Curd	8,15	0,46	4,61	1,35	5,92	0,86
1 week	10	0,44	5,86	0,49	6,27	0,33
1 month	8,01	0,42	3,15	1,32	4,67	0,75
2 months (aw 0.934 – sd 0.005)	8,34	0,22	0.00	0.00	5,14	0,36
Farm C – raw milk
Raw milk	7,54	0,36	3,63	0,85	3,62	1,11
Curd	8,74	0,31	5,15	0,11	5,37	1,03
1 week	9,46	0,00	4,91	0,27	5,77	0,79
1 month	7,81	0,27	1,81	1,57	5,35	1,24
2 months (aw 0.919 – sd 0.003)	7,97	0,12	1,63	1,45	4,41	0,46
Farm C – heat treated milk
Raw milk	7,54	0,36	3,63	0,85	3,62	1,11
Heat treated milk	4,91	0,09	0,72	0,72	2,47	0,04
Curd	8,42	0,05	3,81	0,53	6,22	1,18
1 week	9,72	0,01	5,92	0,24	6,7	0,59
1 month	8,12	0,27	2,75	1,55	4,36	0,82
2 months (aw 0.928 – sd 0.015)	7,99	0,61	0,83	1,43	4,92	0,41

The total mesophilic aerobes in raw milk ranged from log 5.9 log cfu ml-1 in farm A to log 7.54 log cfu ml-1 in farm C. At the early stage of production of cheeses made with raw milk, counts increased up to 1 week and then decreased. In the cheesemaking processes from heat-treated milk added with autochthonous starter cultures bacterial population in milk, after the treatment, was reduced to log 3.06 log cfu ml-1 in farm A, log 4.56 log cfu ml-1 in farm B and log 4.91 log cfu ml-1 in farm C. After the treatment counts showed the same evolution described for cheeses made with raw milk.

Enterobacteriaceae in raw milk ranged from 3.9 log cfu ml-1 in farm A to 3.9 log cfu ml-1 in farm B. In cheeses made with raw milk Enterobacteriaceae increased up to curd production and then decreased being not detectable or at low concentration at the end of the ripening. On farm B and C bacterial reduction was observed on 1-week cheese whilst on farm A an additional growth occurred in cheese during the first week of ripening, after which there was a decrease in population. Enterobacteriaceae were not detectable in full-ripened cheese made with raw ewe’s milk whilst they were detected from full-ripened cheese made with raw goat’s milk. In cheeses made with heat-treated milk with added autochthonous starter cultures, Enterobacteriaceae underwent great reduction in heat-treated milk being not detectable in farm A, whilst they were only partially reduced in farm B and C (4.22 log of reduction in farm B and 2.91 in farm C). Although heavy curd recontamination always occurred, no Enterobacteriaceae were detectable in fullripened cheese made with heat-treated ewe’s milk, whilst they were still detected in cheeses made with heat-treated goat’s milk.

Samelis et al. (2009) studied the changes in microbial composition of raw milk induced by thermization and demonstrated that the treatment significantly reduced bacterial populations of raw milk. Although the 67°C treatment was much more effective against all bacterial groups than was the 60°C treatment, the latter had major inactivation effects against gramnegative bacteria but moderate effects against gram-positive bacteria. Reductions in populations of thermophilic LAB, including enterococci, in samples thermized at 60°C were not significant.

Behaviour of enterococci

Enterococci counts in milk were always lower than 4 log cfu ml-1. The changes observed were similar to those observed previously (Clementi et al., 1998b). At the early stage of production of cheeses made with raw milk, counts increased up to 1 week and then slightly decreased. In the cheesemaking processes from heat-treated milk with added autochthonous starter cultures no evident reduction of enterococci population was observed after the treatment. After the heat treatment counts showed the same behaviour described for cheeses made with raw milk.

Other authors have studied Enterococcus spp. populations in pecorino and other traditional cheeses. For instance Serio et al. (2007) found that the presence of enterococci in Pecorino Abruzzese ranged from log 4 cfu g-1 to log 6 cfu g-1 at the beginning of ripening and reached log 9 cfu g-1 in ripened cheeses. Similar results were described by Litopoulou-Tzanetaki and Tzanetakis (2011) which described enterococci levels in traditional Greek cheeses as ranging from log 6 cfu g-1 to log 10 cfu g-1 for the different kind of products examined.

Identification

One-hundred strains of enterococci were identified. Forty-six from farm A, thirty- two from farm B and twenty-two from farm C. Sixty-nine were identified as E. faecium, twenty-two as E. durans, seven as Enterococcus faecalis and two as E. casseliflavus (Tables 2 and 3). Enterococcus faecium was found more frequently on farm A (80.4%) and on farm B (68.8%), whilst on farm C more than 95% of the isolates consisted of Enterococcus faecium and E. durans altogether. No strains of Enterococcus faecalis were found on Farm A, whilst E. casseliflavus was found only on farm B. Many enterococci such as E. faecium and E. durans (isolated during this study), are prominent probiotic candidates because they possess many desirable characteristics such as cholesterol-lowering ability, production of antimicrobial compounds against pathogens, adhesion to intestinal cells and immunomodulation (Bagci et al., 2019). Enterococci are gastrointestinal commensal bacteria in humans and other animals predominantly considered non-pathogenic. Two species in particular, E. faecalis and E. faecium, despite used in probiotics formulation (Grispoldi et al., 2020) have also been recognised as nosocomial opportunistic pathogens in healthcare settings (Gao et al., 2018).

Table 2.

Distribution of Enterococcus spp. per sample (cheesemaking processes from heat-treated milk).

Species	N.	Milk	Curd	7-day cheese	30-day cheese	Full ripened cheese
E. faecium	69	8(5)	4(8)	10(6)	4(5)	10(9)
E. durans	22	2 (-)	1(4)	3(4)	2 (-)	5(1)
E. faecalis	8	- (-)	1 (-)	1 (-)	- (-)	5 (-)
E. casseliflavus	2	- (-)	- (-)	- (-)	- (-)	- (2)
Total	100	10(5)	6(12)	14(10)	6(5)	20(12)

Recently, Russo et al. (2018) described the prevalence distribution of enterococci in Ragusano and Pecorino Siciliano: 35% of the isolate were ascribed to E. durans, 35% to E. faecalis, 28% to E. faecium. Overall data from previous studies are quite similar and the prevalence of E. faecium/E. durans, followed by E. faecalis is reported quite constantly (Andrighetto et al., 2001; Templer and Baumgartner, 2007; Jamet et al., 2012).

Bacterial cells under microscopic observation after Gram staining appeared Gram positive, in pairs or in short chains, ovoid elongated in direction of the chain. All strains were catalase negative. This result together with cells morphology confirmed the identification of the strains as belonging to the genus Streptococcus. Enterococci were identified as such on the basis of their growth in the presence of bile-esculin, 6.5% of NaCl solution, and 4% bile salts, and on the basis of the hydrolysis of the arginine. TTC reduction on Barnes medium differentiated E. faecalis from E. faecium. In fact, culture grown in the presence of triphenyl tetrazolium chloride (TTC) on Barnes medium were white or with a red centre and white border. The former was presumptively identified as E. faecium and the latter as E. faecalis prior to biochemical tests with API 20 Strep (BioMérieux). No strains were ß-haemolytic.

The principal physiological and biochemical characteristics of the isolates are given on Table 4.

Table 4.

Principal physiological and biochemical characteristics of enterococci isolated from pecorino and goat cheese at various stages of ripening (figures are numbers of strains).

	E. faecium	E. durans	E. faecalis	E. casseliflavus
β-haemolysis	0	0	0	0
β-haemolysis	0	0	0	0
β-haemolysis	69	22	7	2
Growth in bile-esculin	69	22	7	2
Growth in the presence of 6.5% nacl	69	22	7	2
Growth in the presence of 4% bile salts	69	22	7	2
Growth in the presence of TTC	69	22	7	2
Hydrolysis of arginine	69	22	7	2
Hydrolysis of hippurate	34	16	7	2
Acetoin production	69	22	7	2
Fermentation of:
Mannitol	69	0	7	2
Sorbitol	0	0	7	0
Raffinose	0	0	0	2
Inulin	0	0	0	0
L-arabinose	68	0	0	2
Ribose	69	22	7	2
Lactose	69	22	7	2
Trehalose	69	22	7	2
Starch	68	16	7	2
Glycogen	0	0	0	0
Total number of strains	69	22	7	2

The antimicrobial susceptibility test data are shown in Tables 5 and 6. High percentages (>80%) of susceptible strains were found for amoxicillin/clavulanic acid, ampicillin, chloramphenicol, sulphamethoxazole, sulphamethoxazole/ trimethoprim, ticarcillin, vancomycin. A high prevalence of resistance strains (>30%) was observed for amikacin, ciprofloxacin, ceftriaxone, kanamycin, tetracycline. It is important to note that 5% of the strains (one strain of E. durans and 4 strains of E. faecium all isolated from farm B) were resistant to vancomycin. The mechanisms of antibiotic resistance development and diffusion of the genes that encode for antibiotic resistance are very complex and still not completely understand. In the veterinary and agricultural fields, the responsibility has often been placed on the excessive use of them in production animals. Consequently, it is possible to hypothesize that the antibiotic resistance patterns observed in our study were influenced by the pharmacological treatments in the farms necessary to treat pathological conditions. At the same time, it is well known that wild bacterial populations can play the role of reservoirs for antibiotic resistance genes and could become part of the strains isolated during the study (European Centre for Disease Prevention Control, 2019; Dapkevicius et al., 2021). For instance, the last two decades, Enterococci have become major nosocomial pathogens. An increasing number of these infections are due to enterococci that are resistant to vancomycin. Accurate detection of vancomycin-resistant enterococci (VRE) is important so that appropriate therapy and infection control measures may be applied, including veterinary surveillance (European Centre for Disease Prevention Control, 2019). Russo et al. (2018) examined the susceptibility of Enterococcus spp. isolated from Sicilian cheeses to the most clinical antibiotics by microdilution method and found that 97% of the isolates, out of 110, exhibited a multidrug- resistant phenotype (resistance to at least three antimicrobials) with 14 strains, 13 E. faecalis and one E. faecium, resistant to 7 out of 9 antimicrobials tested. Furthermore, 19 strains (15 E. durans, 2 E. faecalis and 2 E. faecium) showed resistance to 5 antimicrobials. In particular, out of 15 E. durans, all were resistant to erythromycin, 14 to rifampicin and ampicillin and 7 to penicillin. Only three (2.7%) strains (E. durans, E. faecalis, and E. faeci um) were susceptible to all tested antibiotics. These results highlighted the highest occurrence of resistance for rifampicin, erythromycin and ampicillin and the lowest for gentamicin. More recently Výrostková et al. (2021) reported that isolates of Enterococcus spp. from sheep and goat cheeses showed high antibiotic resistance to vancomycin (84.62%; 44/52 isolates), teicoplanin (84.62%; 44/52 isolates), erythromycin (76.92%; 40/52 isolates), and rifampicin (76.92%; 40/52 isolates). Lower antibiotic resistance was detected against nitrofurantoin (46.15%; 24//52 isolates) and minocycline (38.46%; 20/52 isolates). Serio et al. (2007) analysed enterococci isolates from Pecorino Abruzzese and found a high incidence of antibiotic resistance, with a prevalence of erythromycin resistance especially for E. faecium (75.7%), followed by E. faecalis (48.3%), and E. durans (37.5%). Results for vancomycin by disk diffusion test have been criticized and in 2011 the CLSI M100-S19 document has recommended the disuse of vancomycin disks for staphylococci and informed that studies on the action of teicoplanin in disk-diffusion testing should be performed (Camargo et al. 2011), however, for screening purposes, such as in this study, the disk diffusion test is still a viable choice, especially for enterococci (Hegstad et al. 2014; Wongthong et al. 2015).

Table 5.

Antimicrobial susceptibility data per Enterococcus spp.

	Total n=100			E. durans n=22			E. faecium n=69			E. faecalis n=7			E. casseliflavus n=2
	S	I	R	S	I	R	S	I	R	S	I	R	S	I	R
AK
total	31	21	48	7	5	10	21	15	33	3	1	3	0	0	2
%	31	21	48	31.8	22.7	45.5	30.4	21.7	47.8	42.9	14.3	42.9	0	0	100
AMC
total	99	1	0	22	0	-	68	1	0	7	0	0	2	0	0
%	99	1	0	100	0	-	98.6	1.5	0	100	0	0	100	0	0
AMP
total	98	0	2	22	0	0	67	0	2	7	0	0	2	0	0
%	98	0	2	100	0	0	97.1	0	2.9	100	0	0	100	0	0
C
total	94	5	1	22	0	0	63	5	1	7	0	0	2	0	0
%	94	5	1	100	0	0	91.3	7.3	1.5	100	0	0	100	0	0
CIP
total	23	47	30	8	9	5	15	29	25	0	7	0	0	2	0
%	23	47	30	36.4	40.1	22.7	21.7	42.0	36.2	0	100	0	0	100	0
CRO
total	8	50	42	0	15	7	5	32	32	2	2	3	1	1	0
%	8	50	42	0	68.2	31.8	7.3	46.4	46.4	28.6	28.6	42.9	50	50	0
K
total	0	26	74	0	5	17	0	21	48	0	0	7	0	0	2
%	0	26	74	0	22.7	77.3	0	30.4	69.6	0	0	100	0	0	100
RL
total	100	0	0	22	0	0	69	0	0	7	0	0	2	0	0
%	100	0	0	100	0	0	100	0	0	100	0	0	100	0	0
SXT
total	89	4	7	19	0	3	62	4	3	6	0	1	2	0	0
%	89	4	7	86.4	0	13.6	89.9	5.8	4.38	85.7	0	14.3	100	0	0
TE
total	45	16	39	13	3	6	27	11	31	4	2	1	1	0	1
%	45	16	39	59.1	13.6	27.3	39.1	15.9	44.9	57.1	28.6	14.3	50	0	50
TIC
total	91	7	2	19	3	0	65	2	2	6	1	0	1	1	0
%	91	7	2	86.4	13.6	0	94.2	2.9	2.9	85.7	14.3	0	50	50	0
VA
total	82	13	5	19	2	1	57	8	4	6	1	0	0	2	0
%	82	13	5	86.4	9.1	4.5	82.6	11.6	5.8	85.7	14.3	0	0	100	0

S: susceptible, I: intermediate, R: resistant; AK: amikacin 30 g, AMC: amoxicillin/clavulanic acid 30 (20 + 10) g, AMP: ampicillin 10 μg, C: chloramphenicol 30 μg, CIP: ciprofloxacin 5 μg, C RO: ceftriaxone 30 μg, K: kanamycin 30 μg, RL: sulphamethoxazole 25 μg, SXT: sulphamethoxazole/trimethoprim 25 μg, TE: tetracycline 30 μg, TIC: ticarcillin 75 μg, VA: vancomycin 30 μg.

Results for the evolution of pH throughout ripening are shown in Figure 2 while Table 1 lists aw values for full ripened cheeses. pH ranged from 6.7 for milk samples to values around 5 for 1-week cheeses with a slight increase toward the end of ripening when all samples were above 5.5. From 1-week all cheeses made with heat treated milk with added starter culture had consistently lower values, due to the acidification induced by lactic acid bacteria. aw for full ripened cheeses was in the range 0.92-0.93. These data are common for those productions (Clementi et al. 1998a; Avellini et al. 1999b; Cenci-Goga et al. 2021; Celano et al. 2022; Daghio et al. 2022; Di Trana et al. 2022; Gaglio et al. 2022).

Figure 2.

Evolution of pH during production and ripening of farm manufactured cheeses (n = 3).

Conclusions

The results clearly supported the technological and hygienic reasons that cheese processors have for applying mild thermization rather than typical pasteurization treatments in raw ewe’s and goat’s milk intended for use in traditional hard cheese processing. In principle, empirically applied thermization treatments should contribute to the reduction of the total microbial load of raw milk to levels assuring an optimal fermentation and increased safety of the resultant cheese while selecting for the beneficial part of the natural milk flora that maintains viability. Thermization of raw milk samples at 56°C for 30 s practically eliminated undesirable Enterobacteriaceae. In our previous study we demonstrated thermization of ewe’s and goat’s milk almost completely eliminated the coliforms and staphylococci in farm goat cheese, although these bacteria were present in high concentrations in raw milk. S. aureus strains (10% of total staphylococci) were isolated from raw milk cheese samples only (Clementi et al., 1998b; Avellini et al., 1999a).

Regarding the antimicrobial susceptibility, in Turkey a study by Sanlibaba and Senturk (2018) in 215 traditional cheese samples identified 99.1% enterococcal isolates that were highly resistant to nalidixic acid (100%), kanamycin (98.6%), and rifampicin (78.4%), and were resistant to ampicillin, ciprofloxacin, erythromycin, tetracycline, penicillin G, chloramphenicol, gentamycin, and streptomycin (Sanlibaba and Senturk, 2018). On the other hand, according to Hollenbeck and Rice (2012) the resistance profile of Enterococcus species according was as follows: erythromycin (49.2%); vancomycin (37.3%); and tetracycline (45.8%). Concurrently, the detected occurrence of antibiotic resistance genes in theses tested enterococci was: ermA 44.8%, vanA 63.6%, tetA 51.9%, tetM 55.6%, ermB 13.8%, and vanB 22.7%. This study may reveal that RTE food products may be reservoirs of detectable enterococci such as E. casseliflavus, E. durans, E. hirae, E. gallinarum.

Also Serio et al. (2007) found a high incidence of antibiotic resistance, with a prevalence of erythromycin resistance especially for E. faecium (75.7%), followed by E. faecalis (48.3%), and E. durans (37.5%). Food isolates resistant to this macrolide have been reported by other authors, such as Russo et al. (2018), but in lower percentages and with a higher incidence for E. faecalis than for E. faecium. The widespread erythromycin resistance could be related to the presence in enterococci of plasmids and transposons (Foulquié Moreno et al. 2006).

Our results, with an overall high level of resistance of Enterococcus spp. to erythromycin (48% of the strains), ciprofloxacin (30%), ceftriaxone (42%), kanamycin (74%) and tetracycline (39%), including a 5% of the strains resistant to vancomycin (and 13% intermediate) demonstrated that the resistance to antimicrobial of Enterococcus spp. is distributed over a wide variety of antibiotics groups. According to Werner et al. (2013) the genomic composition of enterococci, their robust nature, the frequent occurrence in many natural habitats and their flexibility to respond to varying environmental conditions make them a central hub (a «drug resistance gene trafficker») for resistance gene acquisition, conservation and dissemination, especially among related Gram-positive bacteria

In Pecorino and goat cheese the ripening processes are the result of the natural microbiological contamination of milk and of the characteristics cheesemaking technology. Therefore, not only lactic acid bacteria but also other bacterial groups, such as enterococci, must be considered. This study demonstrated that, regardless the cheesemaking technology, heavy curd recontamination occurs and that enterococci are always detected at high concentration at the end of ripening. A large proportion of isolates are also resistant to antibiotics and this pose the question whether microorganism belonging to Enterococcus spp. should be awarded in future the QPS or GRAS status.

The development of mountain farming and the promotion of mountain food production are a way to encourage sustainable development of mountain areas, which are generally considered places with specific geographic and climatic constraints. In order to promote strategies for mountain products, the European Union with the Regulation (EU) No 1151/2012 reserves the use of the term “Mountain Product” to food products produced and processed in mountain areas. This regulation was supplemented by the Delegated Act (EU) No 665/2014, which specifies the conditions of use of the optional quality term “Mountain Product”. It is of the utmost importance to assess the interest in the application of the new mountain label and the perception of the food mountain products by the Umbria Region. However, we believe that the creation of a label to protect and certify mountain food could be the basis to improve the promotion of mountain quality food products and the sustainability of these areas.

Table 3.

Distribution of Enterococcus spp. per farm.

Species	n.		Farm A		Farm B		Farm C
	n.	%	n.	%	n.	%	n.	%
E. faecium	69	69	37	80.4	22	68.8	10	45.4
E. durans	22	22	9	19.6	2	6.2	11	50.0
E. faecalis	7	7	-	-	6	18.8	1	4.6
E. casseliflavus	2	2	-	-	2	6.2	-	-
Total	100	100	46	100	32	100	22	100

Table 6.

Antimicrobial susceptibility data per farm.

	Total			Farm A n=46			Farm B n=32			Farm C n=22
	S	I	R	S	I	R	S	I	R	S	I	R
AK
total	31	21	48	17	9	20	9	3	20	5	9	8
%	31	21	48	37.0	19.6	43.5	28.1	9.4	62.5	22.7	40.9	36.4
AMC
total	99	1	0	46	0	0	31	1	0	22	0	0
%	99	1	0	100	0	0	96.9	3.1	0	100	0	0
AMP
total	98	0	2	46	0	0	30	0	2	22	0	0
%	98	0	2	100	0	0	93.8	0	6.3	100	0	0
C
total	94	5	1	45	0	1	27	5	0	22	0	0
%	94	5	1	97.8	0	2.2	84.4	15.6	0	100	0	0
CIP
total	23	47	30	17	15	14	2	18	12	4	14	4
%	23	47	30	37.1	32.6	30.4	6.3	56.3	37.5	18.2	63.6	18.2
CRO
total	8	50	42	2	15	29	6	19	7	0	16	6
%	8	50	42	4.3	32.6	63.0	18.8	59.4	21.9	0	72.7	27.3
K
total	0	26	74	0	16	30	0	7	25	0	3	19
%	0	26	74	0	34.8	65.2	0	21.9	78.1	0	13.6	86.4
RL
total	0	0	100	0	0	46	0	0	32	0	0	22
%	0	0	100	0	0	100	0	0	100	0	0	100
SXT
total	89	4	7	43	0	3	26	4	2	20	0	2
%	89	4	7	93.5	0	6.5	81.3	12.5	6.3	90.9	0	9.1
TE
total	45	16	39	21	6	19	13	7	12	11	3	8
%	45	16	39	45.7	13.0	41.3	40.6	21.9	37.5	50	13.6	36.4
TIC
total	91	7	2	46	0	0	27	3	2	18	4	0
%	91	7	2	100	0	0	84.4	9.4	6.3	81.8	18.2	0
VA
total	82	13	5	43	3	0	18	9	5	21	1	0
%	82	13	5	93.5	6.52	0	56.3	28.1	15.6	95.5	4.5	0

S: susceptible, I: intermediate, R: resistant; AK: amikacin 30 g, AMC: amoxicillin/clavulanic acid 30 (20 + 10) g, AMP: ampicillin 10 μg, C: chloramphenicol 30 μg, CIP: ciprofloxacin 5 μg, C RO: ceftriaxone 30 μg, K: kanamycin 30 μg, RL: sulphamethoxazole 25 μg, SXT: sulphamethoxazole/trimethoprim 25 μg, TE: tetracycline 30 μg, TIC: ticarcillin 75 μg, VA: vancomycin 30 μg.