TABLE OF CONTENTS

CHAPTER I Introduction 1

PART I Microbial ecology of traditional dairy equipment, raw milk 18 and cheeses

CHAPTER II Identification, typing, and investigation of the dairy characteristics 19 of lactic acid bacteria isolated from “Vastedda della valle del Belìce” cheeses

CHAPTER III Composition and characterisation of the lactic acid bacterial 44 biofilms associated with the wooden vats used to produce two traditional stretched cheeses

CHAPTER VI The influence of the wooden equipment employed for cheese 70 manufacture on the characteristics of a traditional stretched cheese during ripening

PART II Selection of lactic acid bacteria to improve cheese quality 96

CHAPTER V Selected lactic acid bacteria as a hurdle to the microbial spoilage 97 of cheese: Application on a traditional raw ewes’ milk cheese

CHAPTER VI In vivo application and dynamics of lactic acid bacteria for the 118 four-season production of Vastedda-like cheese

CHAPTER VII Activation of wooden vats with selected lactic acid bacteria for 144 the year-round production of traditional Vastedda-like cheese

CHAPTER VIII Large scale applications of selected lactic acid bacteria to improve 149 PDO Pecorino Siciliano cheese.

CHAPTER I

Introduction

1. INTRODUCTION 1 1.1. Cheese 1 1.2. Cheese making technology 2 1.3. Microorganisms involved in dairy productions 4 1.4. Lactic acid bacteria 5 1.4.1. General characteristics of lactic acid bacteria 5 1.4.2. Main genera and species of lactic acid bacteria of food interest 6 1.4.2.1. Genus Lactobacillus 6 1.4.2.2. Genus Lactococcus 8 1.4.2.3. Genus Leuconostoc 9 1.4.2.4. Genus Streptococcus 9 1.4.2.5. Genus Enterococcus 10 1.4.2.6. Genus Pediococcus 11 1.5. Characteristics of selection of lactic acid bacteria for dairy industry 11 1.6. Traditional Sicilian cheeses 13 REFERENCES 15

CHAPTER I

1. INTRODUCTION

1.1. Cheese

Cheese is a food with remote origins and it represents one of the first biotechnological transformations implemented by the human being. Cheese represents a relevant economic and food factor for many population, even for Southeastern Asian countries, which are not traditionally cheese consumers. Cheese “history” began with the acidified milks, produced as a means to store the excess milk. Cheese in the solid form originates from the habit of using the abomasum as a container to stock milk, which, in presence of rennet (a mixture of proteases naturally present in the glandular stomach of ruminants) coagulated. This product could be stored for long time (Salvadori Del Prato, 1998). At the same time, cheese production was also a way to preserve the two main constituents of milk, casein and fat. Moreover, cheese improves the nutritional value of milk. In the 18th century, the dairy processing technology acquired the rigorous scientific characters related to the knowledge in microbiology and, at the beginning of 20th century, the microorganisms that characterize milk and curd have been isolated and studied for their useful properties as well as for their pathogenic aspects (Vizzardi and Maffeis, 1990). According to the Italian law (D.R.L. n° 2033 del 1925) “cheese is the product derived from whole, partly or totally skimmed milk, or from the cream after acid or rennet coagulation, through the use of enzymes and sodium chloride”. According to this law, ricotta, yoghurt, and products derived from evaporated milk are not cheeses. Since there are many technologies for the production of cheese, there are many varieties of cheese products, that can be classified following different principles, including fat content, consistency, ripening and technology (Tab. 1), type of milk (Tab. 2) and, recently, the microbial starter involved in production (Tab. 3).

Table 1. Cheese classification (Fox et al., 2004).

Fat Consistency Ripening Technology

Fat Soft Fresh Uncooked curd

(fat >40% s.s.) (H2O >55% on weight) (15 days) (curd not subjected to cooking) Semi-fat Semi-soft Short-ripened Semi-cooked curd

(fat 20-40% s.s.) (H2O 45-54% on weight) (40 days-6 months) (curd cooked till at 48°C) Light Hard Ripened Cooked curd

(fat <20% s.s.) (H2O 38-44% on weight) (6 months-1 year) (curd cooked further on 48°C) Very hard Long ripened Stretched curd

(H2O 30-37% on weight) (> 1 year) (subjected to “stretching”)

1

CHAPTER I

Table 2. Cheese classification based on typology milk.

Typology of milk Cheese

Cow Caciocavallo, PDO Asiago, PDO Parmigiano Reggiano. Ewes’ PDO Pecorino Siciliano, PDO Pecorino Romano, PDO Vastedda valle del Belìce Buffalo Mozzarella of Bufala Goat Cheeses of Sicilian goat

Table 3. Cheese classification based on milk treatment and microbiology.

Milk Starter Strain

Pasteurized milk Selected starter strains Pasteurized milk Natural starter strains High Temperature/Short time (HTST) treated milk Natural starter strains Raw milk Selected starter strains

Raw milk Natural starter strains

Raw milk No starter strain addiction

1.2. Cheese making technology

Unlike other food fermentations, cheese production consists of two distinct microbial phenomena: cheese fermentation and cheese ripening (Fox et al., 2004). Each major phase includes several events (Fig. 1) and the microorganisms involved are different. During cheese production, the transformation of lactose into lactic acid is strictly linked to the presence and composition of the microbial flora, but for uncooked cheese, also to the presence of enzymes of rennet and original milk, especially if milk does not undergo a thermal treatment (Sciancalepore, 1998). During curdling, lactic acid bacteria (LAB) are responsible for the fermentation of lactose into lactic acid. These bacteria are designated as starter LAB (SLAB). Subsequently, in phase of ripening, other LAB develop the desired effects on the final cheese:, mainly flavor compound generation (Atamer et al., 2009; Fox and McSweeney, 1996). The last group is indicated as non-starter LAB (NSLAB). During the ripening of the cheese, the hydrolysis of the fat components plays an important role in the formation of the aroma of the product; interesting are the changes due to the activity of the lipase (mainly originating from microorganisms, molds and micrococci) that break triglycerides into glycerol and fatty acids. However, the most important phenomenon during cheese aging is the proteolysis of proteins which has a significant influence on the texture and organoleptic and nutritional characteristics of cheese. In fact, after rennet addition,

2

CHAPTER I

primary hydrolysis is carried out by chimosin, a proteinase of the rennet, which is the major responsible for the hydrolysis of large caseins into lower molecular weight products, able to be coagulated to form the curd (Sciancalepore, 1998). Important reactions due to the action of different enzymes regard the deamination of amino acids with production of ammonia and ketonic acid, decarboxylation, with liberation of carbon dioxide and the degradation of amino acids sulfur (cysteine and methionine) with production of sulphides (Sciancalepore, 1998).

Milk Standardization Pastourization

Seeding

Acidification

Coagulation

Curd Manufacture

Optional operations

Pressing

Salting

Cheese curd

Physical and chemical-physical modification of cheese

Loss of water

Crust formation

Lactose transformation Ripening

Lipolysis

Proteolysis

Secondary reactions

Mature cheese Fig. 1. General flow sheet of cheese production.

3

CHAPTER I

1.3. Microorganisms involved in dairy productions

Milk is a complex nutritional matrix in which a large variety of microorganisms can easily grow (Franciosi et al., 2011). The growth capacity of such microorganisms is also highly affected by environmental factors, like the temperature of the storage tanks and the time between milk collection and processing (Heeschen 1996; Slaghuis 1996; Murphy and Boor, 2000). Milk quality is affected by the group of microorganisms that are able to grow in this time and those that are entrapped in the continuous gel of casein micelles. The microorganisms that characterize cheese are those able to grow in complex foods (Jay et al., 2009); they can be pathogens, spoilage or pro-technological. The last group comprises the microorganisms responsible for curd acidification and transformation into mature cheese, as well as those able to improve te safety and stability of the final products. Besides LAB, other microorganisms may be important in cheese production: staphylococci, micrococci, coryneforms, propionibacteria, yeasts, and molds, provide well- defined functions, such as gas production, surface coloration, softening, and, through their diverse enzymatic systems, development of characteristic aromas (Parente and Cogan, 2004). On the basis of the microflora involved during production, another scheme of cheese classification has been proposed by Fox et al. (2004) (Fig. 2).

Rennet coagulation

Internal bacterially ripened Mould -ripened Surface-ripened

Surface mould Internal mould

Extra-hard Cheeses with eyes High-salt varietis Pasta filata varietis

Hard Swiss-type Ducht-type Semi-hard

Fig. 2. Classification of the cheeses based on the microbiology of fermentation. Source: Fox et al., 2004.

4

CHAPTER I

1.4. Lactic acid bacteria

LAB are used worldwide for the manufacture of several fermented foods and especially in the dairy industry to produce fermented milk, yogurt, sour cream and cheeses (Boucher et al., 2001; Deveau et al., 2006). LAB are ubiquitous in nature and their employment in the dairy industry ensures the productions of dairy foods characterized by high standard of quality and safety. For thousands of years, the human beings used unaware LAB to obtain dairy products, thanks to their unknown presence. Subsequently, LAB, in form of cultures obtained empirically from spontaneous or induced acidification of milk, were added to milk during cheese production. However, the use of selected LAB in cheese-making has changed the way of production (Moineau et al., 2002).

1.4.1. General characteristics of lactic acid bacteria

LAB constitute a microbial group particularly heterogeneous both from the morphological and the physiological point of view; LAB are prokaryotes, heterotrophic, Gram-positive cocci or rods. These non-sporulating, non-motile, anaerobic aerotolerant or microaerophilic bacteria that are tolerant to very small amounts of oxygen, do not possess catalase and nitrate reductase. Because of the lack of the respiratory chain, they show an energetic metabolism exclusively fermentative and obtain the necessary energy for the anabolic functions by the phosphorylation at the level of substrate (Kandler and Weiss, 1986). They are nutritionally fastidious and use fermentable carbohydrates as main energy source, for this reason they grow only in complex media containing all the growth factors they need. An ideal environment for LAB usually contains fermentable carbohydrates, protein hydrolysates (peptone and tryptone), complex extracts (meat extract and yeast extract), sources of fatty acids (the most common is the TWEEN 80, an ester of 'oleic acid), vitamins and minerals (Kandler and Weiss, 1986). LAB include different species distinguished for glucose fermentation (Fig. 3), growth at different temperatures and fermentation of carbohydrates other than glucose (e.g. carbohydrates pentose). The homofermentative species produce mainly lactic acid from the fermentation of hexose carbohydrates through the Embden-Meyerhof (glycolysis) pathway, and cannot ferment pentose carbohydrates; the heterofermentative species produce, besides lactic acid, acetic acid and/or ethanol and carbon dioxide through the 6- phosphogluconate/phosphoketolase (6PG/PK) pathway from the hexoses, while CO2 is not produced with pentose carbohydrates. However, a third group of LAB (facultative 5

CHAPTER I

heterofermentative) is able to perform both metabolic pathways, glycolysis in presence of hexoses and 6PG/PK in presence of pentoses.

A B GLUCOSE

homofermentative heterofermentative

Glucose-6-P Glucose-6-P Fructose-6-P 6-P-Gluconate Fructose-1,6-P Ribulose-5-P CO2 Xylulose-5-P Fructose-1,6-DP aldolase P-ketolase Dihidroxy acetone-P Glyceraldehyde-3-P Glyceraldehyde-3-P Acetyl-P Pyruvate Pyruvate Acetaldehyde H2O x 2

LACTATE LACTATE ETHANOL

(ACETIC ACID) Fig. 3. Lactic acid fermentation. A, homolactic fermentation; B, heterolactic fermentation.

1.4.2. Main genera and species of lactic acid bacteria of food interest

According to Gonzalez et al. (2000) the group of LAB includes the following genera: Aerococcus, Alloicoccus, Carnobacterium, Dolosigranulum, Enterococcus, Globicatella, Lactobacillus, Lactococcus, Lactosphaera, Leuconostoc, Oenococcus, Pediococcus, Streptococcus, Tetragenococcus, Vagococcus and Weissella. The main dairy LAB species belong to the genera Lactobacillus, Lactococcus, Leuconostoc, Streptococcus, Pediococcus and Enterococcus.

1.4.2.1. Genus Lactobacillus

The microorganisms belonging to this genus are widespread in nature, thanks to fact that they are able to grow in different environmental conditions. They play important roles in the preparation of various foods and are considered probiotics for humans and animals and are often used in industrial fermentation processes. This LAB genus comprises the largest number of species. To date, 213 Lactobacillus species (www.dsmz.de) have been counted, including also the only mobile species of the LAB group: Lactobacillus agilis, Lactobacillus ghanensis and Lactobacillus capillatus. Cell morphology is rod variable in length depending on the species; the cells may appear as long, thin sticks, sometimes curved, or short, or very short

6

CHAPTER I

and stubby (Fig. 4). The latter are commonly referred to by the term “coccobacilli” (König et al., 2009). In addition, the cell length also varies depending on the age, the substrate composition and the oxygen concentration. However, it may happen that some species (Lactobacillus fermentum and Lactobacillus brevis) might be characterized by pleomorphic morphology, including both long and short rod-shaped cells (Kandler and Weiss, 1986). Often cells are not individual, but in pairs or in short Fig. 4. Cell morphology. or long chains. The colony morphology is also variable: the form they take may be circular or irregular (jagged) when they grow on the surface of agar plates or lenticular when they are included in the volume of the agar layer; the margin can be smooth or corrugated; the thickness is often convex; dimensions are in the range 2-5 mm; the color can be white or transparent. From a metabolic perspective, the Lactobacillus genus is the most complex. It, in fact, comprises all three fermentative classes (obligated homolactic, obligated heterolactic and facultative heterolactic) described above. The species of this genus, which are used in the dairy sector, are given in table 4.

Table 4. Main Lactobacillus species involved in dairy productions.

Homofermentative species Facultative heterofermentative Heterofermentative species

Lactobacillus farciminis Lactobacillus alimentarius Lactobacillus fermentum Lactobacillus helveticus Lactobacillkus casei Lactobacillus buchneri Lactobacillus delbrueckii Lactobacillus pentosus Lactobacillus parabuchneri Lactobacillus plantarum Lactobacillus brevis Lactobacillus rhamnosus Lactobacillus curvatus

The thermophilic lactobacilli are unable to develop at low temperatures, in fact, bacterial growth is not observed already at 15°C. The thermophilic species have a fairly narrow habitat, as they ferment a few sugars (some strains of Lactobacillus helveticus ferment only two sugars). They rarely are lisogenic, often lack plasmids and have a high resistance to phage. The most thermophilic species is Lactobacillus delbrueckii subsp. lactis able to grow at 52°C. The facultative heterofermentative lactobacilli are mainly responsible for cheese ripening. From the physiological point of view, this group is very heterogeneous and includes, for the most part, mesophilic species with optimal temperatures for growth between 30 and 37°C. The mesophilic species are widespread in cheeses, fermented meat, fermented vegetables and

7

CHAPTER I

in the intestine of humans and animals, due to their very broad fermentative framework, high lysogeny, enzyme system, sensitivity to phage and plasmid equipment. Lactobacillus plantarum is one of the most widespread species of the genus Lactobacillus, and it is important for cheese ripening. It grows at 15°C but not at 45°C and produces isomers D and L of lactic acid and it is able to ferment many carbohydrates but not rhamnose. Lactobacillus casei and Lactobacillus paracasei are two important species in cheese ripening. Thanks to their peptidase activity and, for the ability to metabolize amino acids with the production of aromatic substances, they are used as additional cultures for cheese ripening. Finally, the obliged heterofermentative lactobacilli are a very heterogeneous group, both in terms of fermentation and for the level of similarity of DNA. The only two closest species are Lactobacillus buchneri and Lactobacillus kefir, that together with Lb. brevis are among the species of great interest in the dairy environment.

1.4.2.2. Genus Lactococcus

The genus Lactococcus includes species of cocci shape that occur singly, in pairs, or in chains (Fig. 5). The lactococci are mesophilic and are able to grow at or below 10°C, but they don't grow at 45°C and in present of NaCl 6,5% (Schleifer in Bergey’s Manual of Systematic Bacteriology, 1986). The metabolism of the species attributable at this genera is obligate homofermentative (Teuber et al., 1991). Currently are recognized six species of Lactococcus: Fig. 5. Cell morphology. Lactococcus garvie, Lactococcus piscium, Lactococcus plantarum, Lactococcus raffinolactis, Lactococcus chungangensis e Lactococcus lactis. The last species consist of three subspecies Lc. lactis subsp. lactis, Lc. lactis subsp. cremoris, e Lc. lactis subsp. hordniae (Casalta and Montel, 2007). Lc. lactis is used for the production of many cheeses, as starter cultures in single or mixed combination with other species. The strains of this species are technologically relevant in cheese making based on their acidification activity, generation of aromatic compounds and production of antimicrobial substances.

8

CHAPTER I

1.4.2.3. Genus Leuconostoc

The members of the Leuconostoc genes (leucus, clear, light; nostoc, algal generic name; leuconostoc, colorless nostoc) are facultatively anaerobic, catalase-negative, gram- positive cocci arranged in pairs and chains (Garvie, 1986) (Fig. 6). This genus includes microorganisms which have the Fig. 6. Cell morphology. technological function of produce acetoin and diacetyl, important for the formation of the aroma. The bacteria belonging to this genera are coccal- shaped, but also elongated, arranged in pairs or chain. All species included in this genera have an obligate heterofermentative metabolism. The optimal temperatures for growth are between 20° and 30°C, although some species are able to grow at refrigeration temperatures. The optimal habitat for leuconostocs is less acidic than that required by lactobacilli and they commonly prefer growth substrates with a pH close to neutrality. Furthermore, the development of these bacteria is highly dependent on the presence of certain amino acids such as thiamine, biotin and pantothenic acid (Dellaglio et al., 1995). The most interesting dairy species are Leuconostoc mesenteroides subsp. cremoris and Leuconostoc lactis. They both are used as starter in the production of butter and cheese, but L. mesenteroides is also found during ripening.

1.4.2.4. Genus Streptococcus

At this genus belong microorganisms of cocci shape reunited in long chains (Fig. 7). Based on morphological, physiological and biochemical characteristics the streptococci were divided into four groups: (Gobbetti and Corsetti, 2000): - pyogenic or hemolytic streptococci; - oral streptococci; - lactic streptococci; - faecal streptococci. Fig. 7. Cell morphology. The only streptococci of food interest within the group of lactic streptococci, while all others are often pathogens. Streptococcus thermophilus is the most important species of the genus Streptococcus of industrial use (Zirnstein and Hutkins, 2000).

9

CHAPTER I

This microrganism is a thermophilic Gram-positive bacterium with an optimal growth rate at 45°C but is able to to withstand pasteurization treatments. S. thermophilus is found in fermented milk products, and is generally used as starter coltures in the production of stretched cheeses and yogurt (alongside Lb. delbrueckii subsp. bulgaricus). Recently, it has been identified another species of this genus, Streptococcus macedonicus, isolated from Greek cheeses (Tsakalidou et al., 1998) and found associated with the ripening of many cheeses (Maragkoudakis et al., 2009).

1.4.2.5. Genus Enterococcus

Enterococci are cocci gathered, mainly in short chains (Fig. 8). They require different growth factors, in particular B-group vitamins and some amino acids (Devriese and Pot, 1995). Many species are halo-tolerant i.e. able to grow at high NaCl concentrations (6.5%) and are able to withstand a maximum pH of 9.6. Enterococci may survive intense pasteurization and most of the species grow at a temperature in the range 10 - 45°C Fig. 8. Cell morphology. (Moellering, 1992). Enterococcus faecium and Enterococcus faecalis are the species most frequently isolated from the digestive tract of humans, and together with Enterococcus durans are the species of enterococci mostly isolated from dairy products (Franz et al., 1999; Suzzi et al., 2000; Andrighetto et al., 2001; Gelsomino et al., 2001). Studies on the microflora of cheeses matured in the Mediterranean basin have shown that enterococci play an important role in cheese ripening. In fact, they show lipolytic and proteolytic activities higher than other bacteria, and like bacteria belonging to the Leuconostoc genus, enterococci are capable of producing high amounts of diacetyl, acetoin and acetaldehyde through the metabolism of citrate (Franz et al., 1999). The enterococci presence in milk and cheese has long been considered the result of unhygienic conditions adopted during the making processes. Fecal contamination may be direct or indirect and associated with various sources, such as water, skin of animals, equipment and vats used in the dairy factories. Despite some authors have argued that high levels of enterococci may cause a deterioration of the sensory properties of certain cheeses

10

CHAPTER I

(Thompson and Marth, 1986), several studies reported on their positive role in the typicality of products (Foulquié Moreno et al., 2006).

1.4.2.6. Genus Pediococcus

This genera includes species of spherical shape that can be divided on two floors to form tetrads; cells rarely occur in pairs and never alone (Fig. 9). The optimum growth temperature is between 25 and 30°C depending on the species. Some species can produce a pseudocatalase capable of hydrolyzing the hydrogen Fig. 9. Cell morphology peroxide. These bacteria have exclusively an obligate homofermentative metabolism and produce huge amounts of lactic acid. The species of greatest interest are Pediococcus acidilactici and Pediococcus pentosaceus, both able to ferment the monosaccharides and pentoses.

1.5. Characteristics of selection of lactic acid bacteria for dairy industry

LAB used to produce cheeses are SLAB and NSLAB. Basically, the SLAB must fulfill one purpose which is to acidify the curd in a short time and their role runs out fairly quickly, while the NSLAB play different roles during ripening. In general, NSLAB are already present as contaminants of milk and only recently they are object of selection. The role of the starter in cheese production is very important, since, thanks to their acidifying power, they favour the whey removal, create an acidic environment suitable for a long preservation, contrast the action of anti-dairy microorganisms and, in the case of stretched cheese, they give an appropriate plasticity degree to the curd, useful during the operation of stretching (Salvadori Del Prato, 1998). Starter cultures are divided into "mixed strain starters" (MSS), which include several species known in undefined proportions, and "defined strain starters" (DSS), which include several species known in definite proportions. The MSS are prepared by the cheesemaker or by specialized laboratories and they are distinct in MSS thermophilic, essentially consisting of S. thermphilus, Lb. delbrueckii and Lb. helveticus, reproduced at high temperatures in whey (natural whey starter culture “sieroinnesto”), and mesophilic, consisting of Lc. lactis and Leuconostoc spp., reproduced at 25-30°C in milk (natural milk starter culture, “lattoinnesto”). The DSS, distinct in

11

CHAPTER I

thermophilic and mesophilic as well, are exclusively prepared by specialized laboratories (Parente and Cogan, 2004). Unlike SLAB, NSLAB are generally selected for their ability to positively influence the organoleptic characteristics of the final products. During the ripening process, these bacteria, thanks to their hydrolytic enzymes, degrade the casein and peptides present in cheese and liberate the amino acids. Some amino acids possess aromatic features (Fox and Wallace, 1997), but for the majority of them the aroma results after various catabolic reactions, producing several compounds (ammonia, α-keto acids, amines, aldehydes, acids and alcohols) (Hemme et al., 1981). The action of NSLAB is defining in this process. Poveda et al. (2004) reported that the addition of L. plantarum to the mixture of starter strains determined an increase of free fatty acids compared to the control cheese obtained with the typical MSS. Lactobacillus reuteri has been found to affect the fraction of volatile compounds in Edam cheese (Tungjaroenchai et al., 2004). Furthermore, a strain of Lb. casei subsp. rhamnosus has increased the level of proteolysis during ripening of Greek Kefalograviera cheese (Michaelidou et al., 2003). However, the effect of the NSLAB can vary in function of the mixture of the strains (Skeie et al., 2008). In addition, the changing of the fat and proteic component during the ripening of the cheese has also a high impact on the structure and texture of the final cheese (Fox et al., 2004). These microorganisms also play a significant preservative role, as they are often able to inhibit the undesired microorganisms through the production of bacteriocins, compounds of protein nature endowed with antibacterial activity directed, generally, towards strains taxonomically closely related to the producing species. This ability is fundamental to the selective screening, since this characteristic determines an elongation of cheese shelf-life. Franciosi et al. (2009) reported that E. faecalis strains producers of bacteriocins persist in Puzzone of Moena cheese longer than non-producing strains, showing in this way that this character gives a competitive advantage to the producing strain and guarantees its action over time. However, we must keep in mind that the success of this organic preservation strategy depends heavily on the effectiveness of antimicrobial in situ (Settanni and Corsetti, 2008). Recently, the process of NSLAB selection also refers to other properties that affect their beneficial effects on human health. According to the “Consensus Document” of Concerted Action of the European Commission on Functional Food Science in Europe (FU.FO.S.E.), a food can be defined functional when it influences unequivocally one or more biological functions of the consumer, improving the health and well-being and reducing the risk of onset 12

CHAPTER I

of disease (Diplock et al., 1999). Foods containing probiotic bacteria, namely live microorganisms capable of giving the guest a healthy effect (Araya et al., 2002) belong to this category. Many NSLAB can be considered probiotics (Ouwehand et al., 2003; Saxelin et al., 2005) and the cheese is an optimal matrix to transfer living bacteria to the host for its protection during the gastrointestinal transit (Ross et al., 2002), so that they can express their positive capacity in the intestine. Other important features of NSLAB are related to the bioactive peptide production, specific protein fragments able to control hypertension (Kitts and Weiler, 2003), the ability to synthesize γ-aminobutyric acid (G.A.B.A.) from L-glutamate which has different physiological functions, including blood pressure regulation (Guin Ting Wong et al., 2003) and the antigenotoxic expression, that is the ability to inhibit the genotoxins that attack DNA, reducing intestine pathologies and cancer colon incidence (Massi et al., 2004). Therefore, the study of NSLAB has been for many years a strong point of the research activities in the food microbiology field. The current and future researches are aimed at selecting strains with many positive effects, but this will probably mean an adaptation or conversion of the existing cheese production processes. The current and future researches the dairy field are aimed at selecting strains with many positive effects, but this will probably mean an adaptation or conversion of the existing cheese production processes.

1.6. Traditional Sicilian cheeses

The Sicily region boasts a large number of traditional foods. In the dairy sector, the range of products is very large. Generally, Sicilian cheeses, often characteristic of very restricted areas, are niche products that bind strongly their hystory to the production area. Recently, the style of the "green consumerism", which is becoming increasingly popular, has resulted in a strong rediscovery of traditional and typical. In this context, thanks also to the food policies that are being taken to protect small productions, the consumer is becoming more sensitive to processed food in the traditional manner, as they are often characterized by a strong image of naturalness (Fig. 10). The best known Sicilian cheeses are: Caciocavallo Palermitano, Ragusano, Provola dei Nebrodi, Tuma persa and Fiore Sicano among those in cow's milk and Maiorchino, Pecorino Siciliano, Piacentinu Ennese and Vastedda della valle del Belìce among those in ewes’ milk.

13

CHAPTER I

These products are requested not only in the region but in the entire Italian mainland, since the various food and wine circuits of local products promotion have shown their appreciation.

Fig. 10. Traditional Sicilian cheeses.

The Sicilian cheeses with a protected designation of origin (PDO) are PDO Ragusano, PDO Pecorino Siciliano, PDO Vastedda della valle del Belìce and PDO Piacentinu Ennese. PDO products are characterized by a strong link with the production areas and the strict respect of the Production Guidelines approved by MIPAAF (Ministry of Agriculture and Forestry) and published in the Official Journal of the European Union. The strong link between the cheese and the native environment regards also raw materials, mainly milk, processing technology, production equipments and all those factors that link the product to its territory. The microflora characteristic, of a typical product from raw milk, influences many of the organoleptic characteristics of the cheese which is often the mirror of the environment and production systems. In fact, the binding of a cheese to the territory is not only found in the traditions handed down over time, but also and especially for the presence of species and strains of microorganisms that colonize the environments and processing equipments, contributing decisively to the typicality of the final product.

14

CHAPTER I

REFERENCES Andrighetto, C., Knijff, E., Lombardi, A., Torriani, S., Vancanneyt, M., Kersters, K., Swings, J., Dellaglio, F. (2001) Phenotypic and genetic diversity of enterococci isolated from Italian cheese. Journal of Dairy Research 68, 303–316. Araya, M., Morelli, L., Reid, G., Sanders, M.E., Stanton, C., Pineiro, M., Ben Embarek, P. (2002) Guidelines for the Evaluation of Probiotics in Food. Joint FAO/WHO Working Group Report on Drafting Guidelines for the Evaluation of Probiotics in Food, London (ON, Canada) April 30 and May 1. ftp://ftp.fao.org/es/esn/food/wgreport2.pdf Atamer, Z., Dietrich, J., Muller-Merbach, M., Neve, H., Heller, K.J., Hinrichs, J. (2009) Screening for and characterization of Lactococcus lactis bacteriophages with high thermal resistance. International Dairy Journal 19, 228–235 Boucher, I., Emond, E., Parrot, M., Moineau, S. (2001) DNA sequence analysis of three Lactococcus lactis plamids encoding phage resistance mechanisms. Journal of Dairy Science 84, 1610–1620 Casalta, E., Montel, M.C. (2007) Safety assessment of dairy microorganism. The Lactococcus genus. International Journal of Food Microbiology 126, 271–273 Dellaglio, F., Dicks, L.M.T., Torriani, S. (1995) The genus Leuconostoc. In Wood BJB, Holzapfel, W.H., (Eds.). The genera of Lactic Acid Bacteria, Black Academic & Professional, London, UK Deveau, H., Labrie, S.J., Chopin, M.C., Moineau, S. (2006) Biodiversity and classification of lactococcal phages. Applied and Environmental Microbiology 72, 4338–4346 Devriese, L.A., Pot, B. (1995) The genus Enterococcus. In Wood BJB, Holzapfel WH (Eds.). The genera of Lactic Acid Bacteria, Black Academic & Professional, London, UK Foulquié Moreno, M.R., Sarantinopoulos, P., Tsakalidou, E., De Vuyst, L. (2006) The role and application of enterococci in food and health. International Journal of Food Microbiology 106, 1–24. Fox, P.F., McSweeney, P.L.H. (1996) Proteolysis in cheese during ripening. Food Review International 12, 457– 509 Fox, P.F., Wallace, J.M. (1997) Formation of flavor compounds. Advances in Applied Microbiology 45, 17–85 Fox, P.F., Cogan, T.M., Guinee, T.P. (2004) Cheese: Chemistry, Physics and Microbiology. Elsevier Academic Press, London, UK Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Presence of enterococci in raw cow’s milk and “Puzzone di Moena” cheese. Journal of Food Processing and Preservation 33, 204–217 Franciosi, E., Settanni, L., Cologna, N., Cavazza, A., Poznanski, E. (2011) Microbial analysis of raw cows' milk used for cheese-making: influence of storage treatments on microbial composition and other technological traits. World Journal of Microbiology and Biotechnology, 27, 171–180 Franz, C.M.A.P., Worobo, R.W., Quadri, L.E.N., Schillinger, U., Holzapfel, W.H., Vederas, J.C., Stiles, M.E. (1999) Atypical genetic locus associated with constitutive production of enterocin B by Enterococcus faecium BFE900. Applied and Environmental Microbiology 65, 2170–2178 Garvie, E. I. (1986) Genus Leuconostoc. In P. H. A. Sneath, N. S. Mair, M. E. Sharp, and J. G. Holt (ed.), Bergey’s manual of systematic bacteriology, vol. 2. The Williams and Wilkins Co., Baltimore Gelsomino, R., Vancanneyt, M., Condon, S., Swings, J., Cogan. (2001) Enterococcal diversity in the environment of an Irish cheddar-type cheese making factory. International Journal of Food Microbiology 71, 177–188 Gobbetti, M., Corsetti, A. (2000) Streptococcus. In Robinson RK, Batt CA, Patel PD (Eds.). Encyclopedia of Food Microbiology, Academic Press, London, UK Gonzalez, C. J., Encinas, J.P., Garcia-Lopez, M.L., Otero, A. (2000). Characterization and identification of Lactic Acid Bacteria from freshwater fishes. Food Microbiology 17, 383–391 Guin Ting Wong, G., Bottiglieri, T., Carter Snead III, O. (2003) GABA, gammahydroxybutyric acid, and neurological disease. Annals of Neurology 6, 3–12

15

CHAPTER I

Heeschen, W.H. (1996) Bacteriological quality of raw milk. Legal requirements and payment systems. In Proceedings of IDF bacteriological quality of raw milk symposium, Wolfpassing, Austria. IDF, Brussels, Belgium Jay, J.M., Loessner, M.J., Golden, D.A. (2009) Microbiologia degli alimenti. Springer-Verlag, Italy Kandler, O., Weiss, N. (1986) “Genus Lactobacillus”. In Sneath, P.H.A., Mair, N.S., Sharpe, M.E., Holt, J.G., (Eds.). Bergey’s Manual of Systematic Bacteriology, Vol.2, 9th ed. Williams and Wilkins Baltimore Kitts, D.D., Weiler, K. (2003) Bioactive proteins and peptides from food sources. Applications of bioprocesses used in isolation and recovery. Current Pharmaceutical Design 9, 1309–1323 König, H., Fröhlich, J. (2009) Biology of Microorganisms on Grapes, in Must and in Wine. Institute of Microbiology and Wine Research, Johannes Gutenberg-University, 55099 Mainz, Germany Maragkoudakis, P.A., Papadelli, M., Georgalaki, M., Panayotopoulou, E.G., Martinez-Gonzalez, B., Mentis, A.F., Petraki, K., Sgouras, D.N., Tsakalidou, E. (2009) In vitro and in vivo safety evaluation of the bacteriocin producer Streptococcus macedonicus ACA-DC 198. International Journal of Food Microbiology 133, 141–147 Massi, M., Vitali, B., Federici, F., Matteuzzi, D., Brigidi, P. (2004) Identification method based on PCR combined with automated ribotyping for tracking probiotic Lactobacillus strains colonizing the human gut and vagina. Journal of Applied Microbiology 96, 777–786 Michaelidou, A., Katsiari, M.C., Voutsinas, L.P., Kondyli, E., Alichanidis, E. (2003) Effect of commercial adjunct cultures on proteolysis in low-fat Kefalograviera type cheese. International Dairy Journal 13, 743– 753 Moineau, S., Tremblay, D., Labrie, S. (2002) Phages of Lactic Acid Bacteria: from genomics to Industrial Applications. ASM News 68, 388–93 Murphy, SC., Boor, KJ. (2000) Trouble-shooting sources and causes of high bacteria counts in raw milk. Dairy, Food and Environmental Sanitation 20, 606–611 Ouwehand, A.C., Salvadori, B.B., Fondén, R., Mogensen, G., Salminen, S., Sellars, R. (2003) Health effects of probiotics and culture-containing dairy products in humans. Bulletin of International dairy Federation 380, pp. 4–19 Parente, E., Cogan, T.M. (2004) In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.). Cheese: Chemistry, Physics and Microbiology. Chapman and Hall, London Poveda, J.M., Cabezas, L., McSweeney, P.H.L. (2004) Free amino acid content of Manchego cheese manufactured with different starter cultures and changes throughout ripening. Food Chemistry 84, 213–218 Ross, R.P., Fitzgerald, G., Collins, K., Stanton, C. (2002) Cheese delivering biocultures e prebiotic cheese. Australian Journal of Dairy Technology 57, 71–78 Salvadori del Prato, O. (1998) Tecnologia Lattiero-Casearia. Edagricole, Bologna, Italy Saxelin, M., Tynkkynen, S., Mattila-Sandholm, T., de Vos, W. (2005) Probiotic and other functional microbes. From markets to mechanisms. Current Opinion in Biotechnology 16, 1–8 Schleifer, K.H. (1986) Gram-positive cocci. Sneath, P.H.A., Mir, N.S., Sharpe, M.E., Holt, J.G., Bergey’s Manual of Systematic Bacteriology. The Williams and Wilkins Co. Baltimore Sciancalepore, V. (1998) Industrie agrarie: olearia, enologica, lattiero-casearia. UTET, Milano, Italy Settanni, L., Corsetti, A. (2008) Application of bacteriocins in vegetable food biopreservation. International Journal of Food Microbiology 121, 123–138 Settanni, L., Moschetti, G., (2010) Non-starter Lactic Acid Bacteria used to improve cheese quality and provide health benefits. International Journal of Food Microbiology 27, 691–697 Skeie, S., Kieronczyk, A., Eidet, S., Reitan, M., Olsen, K., Østlie, H. (2008) Interaction between starter bacteria and aadjunct Lactobacillus plantarum INF15D on the degradation of citrate, asparagine and aspartate in a washed-curd cheese. International Dairy Journal 18, 169–177 Slaghuis B., 1996. Sources and significance of contaminants on different levels of raw milk production. In Proceedings of IDF symposium bacteriological quality of raw milk. Wolfpassing, Austria. IDF, Brussels, Belgium Suzzi, G., Caruso, M., Gardini, F., Lombardi, A., Vannini, L., Guerzoni, M., Andrighetto, C., Lanorte, M.T. (2000) A survey of the entrerococci isolated from an artisanal italian goat’s cheese (semicotto caprino). Journal of Applied Microbiology 89, 267–284

16

CHAPTER I

Teuber, M., Geis, A., Neve, M. (1991) The genus Lactococcus. In The Prokaryotes, Vol. 2, 2nd edn (eds Balows, A., Trüper, H.G., Dworkin, M., Harder, W. and Schleifer, K.H.) Springer-Verlag, New York, USA Thompson, T. L., Marth, E. H. (1986) Changes in Parmesan cheese during ripening. Microflora-coliforms, enterococci, anaerobs, propionibacteria and staphylococci. Milchwissenschaft 41, 201–204 Tsakalidou, E., Manolopoulou, E., Kabaraki, E., Zoidou, E., Pot, B., Kersters, K., Kalantzopoulos, G. (1994) The combined use of whole-cell protein extracts for the identification (SDS-PAGE) and enzyme activity screening of lactic acid bacteria isolated from traditional Greek dairy products. Systematic and Applied Microbiology 17, 444–458 Tungjaroenchai, W., White, C.H., Holmes, W.E, Drake, M.A. (2004) Influence of adjunct cultures on volatile free fatty acids in reduced-fat Edam cheeses. Journal of Dairy Science 87, 3224–3234 Vizzardi, M., Maffeis, P. (1990) Sviluppo agricolo nel Rinascimento italiano. Formaggi italiani - Storia e tecniche di preparazione. Edagricole, Bologna, Italy Zirnstein, G., Hutkins, R. (2000) Streptococcus thermophilus. In Robinson RK, Batt CA, Patel PD (Eds.). Encyclopedia of Food Microbiology, Academic Press, London, UK

17

PART I

Microbial ecology of traditional dairy equipment, raw milk and cheeses

18

CHAPTER II

Identification, typing, and investigation of the dairy characteristics of lactic acid bacteria isolated from

“Vastedda della valle del Belìce” cheeses

The present chapter has been published in

Dairy Science & Technology

94, 157-180

2014

19

CHAPTER II

ABSTRACT

Traditional cheeses made without starter cultures can be characterised by the attribute of instability. The addition of autochthonous starter cultures can ensure stability without compromising the characteristics of the final product. This study aimed to characterise the autochthonous lactic acid bacteria (LAB) population in Vastedda della valle del Belìce cheeses, which have a protected designation of origin (PDO) status, in order to develop an ad hoc starter culture to be used in its future production. Winter and spring productions were analysed to ensure isolation of specific LAB that had adapted to perform fermentation at low temperatures. Plate counts revealed total microbial numbers nearing 109 CFU/g. All of the cheese samples were dominated by coccus-shaped LAB. When enterobacteria were present, their concentrations were at similar levels (3.3-5.6 Log CFU/g) in both seasons. All of the colonies that differed in morphological appearance were isolated and differentiated on the basis of phenotypic characteristics and genetic polymorphisms, as analysed by random amplification of polymorphic DNA-PCR. A total of 74 strains were identified and further genotyped by sequencing the 16S rRNA gene, resulting in the identification of 16 LAB species belonging to five genera (Enterococcus, Lactobacillus, Lactococcus, Leuconostoc and Streptococcus). The species most frequently present were Streptococcus gallolyticus subsp. macedonicus, Streptococcus thermophilus, Lactococcus lactis and Leuconostoc mesenteroides. The 74 strains were also investigated in vitro for general dairy parameters such as acidification capacity, diacetyl generation and antibacterial activity. Several strains of the most frequently represented species displayed traits relevant to the production of PDO “Vastedda della valle del Belìce”.

Key words: Acidifying capacity; Bacteriocins; Diacetyl; Lactic acid bacteria; Raw ewes’ milk cheese.

20

CHAPTER II

1. INTRODUCTION

Stretched pasta filata cheeses owe their characteristics to the methods used in their production, which consists of two distinct steps. The first step involves an acidification of the curd, which results in it assuming a plastic consistency. The curd is then heated to a scalding temperature (80 – 90°C), allowing it be moulded into the final shape. Following these 2 steps, the cheese is left to ripen (Salvadori del Prato 1998). Within this group, only a few cheeses are produced from ewes’ milk. “Vastedda della valle del Belìce” (Vastedda) is a pasta filata cheese from the homonymous valley in Sicily, Italy; it is made from raw ewes’ milk without the addition of starter cultures, and has been named a protected designation of origin (PDO) cheese by the European Union. Soon after its production, the cheese is sealed under vacuum, refrigerated, and sold fresh in local markets (Mucchetti et al., 2008). From a hygienic perspective, raw milk cheeses deserve greater attention than cheeses made from thermally treated milk, because the final products can become contaminated with pathogenic microorganisms from the raw milk that subsequently survive the cheese making process (Donnelly 2004). However, the stretching phase, typical of the pasta filata preparation method, strongly contributes to the safety of the resulting cheeses because the high temperatures applied during processing inactivate most microorganisms. Cheese cannot be made without the activity of certain species of lactic acid bacteria (LAB) (Parente and Cogan, 2004). As such, cheese produced from raw milk without the addition of starter cultures relies on the presence of indigenous LAB in the milk and/or those transferred from the processing equipment or the transformation environment. Under these conditions, differences in microbial evolution may produce variability in the final characteristics of the cheese, which cannot easily be controlled by the cheese maker (Franciosi et al., 2008). Species and strain composition may be responsible for not only the inter-factory differences (Antonson et al., 2001; De Angelis et al., 2001) but also for the differences between cheeses produced at the same factory on different days or in different vats on the same day (Fitzsimons et al., 1999; Williams et al., 2002). The modern systematic approach to minimising microbial variability and obtaining a cheese with the desired characteristics, consistently over time, is based on the use of starter cultures. However, this technology can compromise the individual characteristics or typicality of the final product. In this context, the only innovation that can be used to attain the typicality of a given cheese is the addition of autochthonous microorganisms. These microorganisms can be indigenous to the milk, or they may derive from the environment or

21

CHAPTER II

equipment. In some cases, they are highly adapted to the production area and are responsible for the desirable sensory attributes of the cheese (Micari et al., 2007). Traditionally, Vastedda cheese was produced only during the summer, but due to the increasing demand for this cheese, it is now produced year-round. The particular flavour and typical organoleptic properties of raw milk cheeses are strongly associated with specific attributes of the raw milk (Beresford et al., 2001), and the generation of the aroma profile relies on metabolites produced by the indigenous microbial populations (Settanni and Moschetti 2010), which may vary across the different seasons. In order to identify the autochthonous LAB composition of the starter culture required to standardise the year-round production of Vastedda cheese, seasonal samples of this cheese were collected from different dairy factories. In order to select the strains best suited for use in Vastedda cheese production, the LAB populations in the cheese samples were enumerated, isolated and characterised with respect to their acidification capacity, diacetyl production and ability to inhibit the growth of undesirable bacteria.

22

CHAPTER II

2. MATERIALS AND METHODS

2.1. Sample collections

Samples of PDO “Vastedda della valle del Belìce” were collected from 7 markets supplied by 7 dairy factories located in western Sicily (Agrigento, Palermo and Trapani provinces), during the winter (2011-2012) and spring (2012) production periods (Table 1). The Vastedda cheeses were manufactured (Fig. 1) according to regulated production practices, that exclude the addition of starter LAB (GUE no. C 42/16 19.2.2010), vacuum packed and refrigerated. The market samples were collected 11 - 14 days after production, placed into a portable refrigerator and transferred to the Laboratory of Agricultural Microbiology (Department of Agricultural and Forestry Science - University of Palermo).

Fig. 1. Flow diagram of Vastedda della valle del Belìce cheese production.

2.2. Microbiological analysis

Cheese samples were homogenised with a stomacher (BagMixer® 400, Interscience, Saint Nom, France) for 2 min at the highest speed in a sodium citrate (20 g/L) cheese:diluent (1:9) 23

CHAPTER II

solution. Further serial decimal dilutions were prepared using Ringer’s solution (Oxoid, Milan, Italy). Cell suspensions were plated and incubated as follows: total mesophilic count (TMC) bacteria were plated on plate-count agar (PCA) supplemented with 1 g/L skimmed milk (SkM) and incubated aerobically for 72 h at 30°C; Enterobacteriaceae on double-layered violet-red-bile-glucose agar (VRBGA) and incubated aerobically for 24 h at 37°C; mesophilic and thermophilic rod-shaped LAB on de Man-Rogosa-Sharpe (MRS) agar, acidified to pH 5.4 with lactic acid (5 mol/L) and incubated anaerobically for 48 h at 30 and 44°C, respectively; and mesophilic and thermophilic coccus-shaped LAB on M17 agar, and incubated anaerobically for 48 h at 30 and 44°C, respectively. Anaerobiosis occurred in hermetically sealed jars added with the AnaeroGen AN25 system (Oxoid). All media were purchased from Oxoid. Microbiological counts were performed in triplicate. Statistical analyses were conducted using STATISTICA software (StatSoft Inc., Tulsa, OK, USA). Due to the winter-produced batch from factory G being unavailable, only the data from cheese factories A-F were analysed. Microbial data were analysed using a generalised linear model (GLM) that included the effects of the farms (Fa = A to F), the season (Se = W and S) and their interaction (Fa × Se) with one another. Student t-tests were used to compare the means and pairwise comparisons were evaluated with a post-hoc Tukey test. A P-value <0.05 was deemed significant.

2.3. Lactic acid bacteria isolation and phenotypic grouping

Gram-positive (Gregersen KOH method) and catalase-negative bacterial cultures, presumptively considered LAB, were obtained by randomly picking a minimum of 4 morphologically similar colonies for each of the various colony shapes found on the MRS and M17 agar plated and transferring them to corresponding broth media. Catalase-negative bacteria were identified by their inability to catalyse H2O2 (5%) to water. The isolates were purified by successive sub-culturing and stored in broth media containing 20% glycerol (v/v) at ‒ 80°C until further analysis. The isolates were phenotypically characterised in order to obtain the initial groupings by observing the cell morphology of the LAB isolates using an optical microscope. Subsequently, the presumptive LAB isolates were subjected to further phenotypic assays. Rod- and coccus-shaped LAB isolates were grouped on the basis of their growth characteristics: growth at 15 and 45°C; resistance at 60°C for 30 min (Sherman, 1937) evaluated as described by Harrigan and McCance (1976); NH3 production from arginine

24

CHAPTER II

tested as described by Abd el Malek and Gibson (1948); aesculine hydrolysis determined by the method described by Qadri et al. (1980); acid production from arabinose, ribose, xylose, fructose, galactose, lactose, sucrose and glycerol; and CO2 production from glucose. The test for CO2 production was carried out in Durham’s tubes with the same growth medium used for isolation, except for the addition of citrate, which can alter gas formation when fermented by certain LAB. The M17 medium had glucose substituted for lactose. Positive results for this test were indicative of hetero-fermentative metabolism. The strains negative for this assay were inoculated into test tubes containing their optimal growth media prepared with a mixture of pentose carbohydrates (xylose, arabinose and ribose, 8 g/L each) in place of glucose as described by Settanni at al. (2012). The strains capable of growth in this media have a facultative hetero-fermentative metabolism, while the strains unable to grow in the media have an obligate homo-fermentative metabolism. The coccus-shaped isolates were further grouped by their ability to grow at pH 9.2 and in the presence of NaCl (6.5 g/L) to separate enterococci (that can grow in both conditions) from other dairy cocci.

2.4. Genotypic differentiation and identification of lactic acid bacteria

Cells were harvested from cheese isolate cultures grown overnight in broth media at the optimal temperatures and genomic DNAs were extracted using the Instagene Matrix kit (Bio- Rad, Hercules, CA) as described by the manufacturer. Crude cell extracts were used as templates for the PCRs. Strains were differentiated from one another by random amplification of polymorphic DNA (RAPD)-PCR analysis in a 25-µL volume using the single primers M13, AB111, and AB106 as described by Settanni et al. (2012). The PCR products were separated by electrophoresis on 2% (w/v) agarose gels (Gibco BRL, Cergy Pontoise, France) and visualised by UV trans-illumination after staining with the SYBR® safe DNA gel stain (Molecular probes, Eugene, OR, USA). The GeneRuler 100 base pair (bp) Plus DNA ladder (M·Medical Srl, Milan, Italy) was used as a molecular size marker and the RAPD patterns were analysed using Gelcompare II software version 6.5 (Applied-Maths, Sint- Marten- Latem, Belgium). The LAB with different RAPD-PCR profiles were identified genotypically by sequencing 16S rRNA. PCR reactions were performed as described by Weisburg et al. (1991) using the primers rD1 (5’-AAGGAGGTGATCCAGCC-3’) and fD1 (5’-AGAGTTTGATCCTGGCTC

25

CHAPTER II

AG-3’). The PCR products were visualised, and the amplicons corresponding in size to the molecular weight of the 16S rRNA genes, were excised and purified using the QIAquick purification kit (Quiagen S.p.a., Milan, Italy). The resulting DNA was sequenced by Consorzio Regionale di Ricerca BioEvoluzione Sicilia (Santa Margherita del Belìce, AG, Italy), using the same primers employed for the PCR amplifications. The sequences were identified by a BLAST search of the GenBank/EMBL/DDBJ database and were deposited under the accession numbers KC545881-KC545901 and KF147882-KF147890.

2.5. Technological screening of the lactic acid bacteria

The different LAB strains were tested for their acidification capacity, diacetyl formation and production of antimicrobial compounds. The acidifying capacity was assayed at the optimal growth temperature and at 7°C in 10 mL of full fat, ultra-high temperature (UHT) ewe’s milk inoculated with a 1% (v/v) cell suspension obtained by growing the cultures overnight in their optimal medium, centrifuging at 5000 × g for 5 min, washing, and re-suspending in Ringer’s solution. To standardise bacterial inocula, the cells were re-suspended in Ringer’s solution to an optical density at 600 nm of 1.00, corresponding to approximately 109 CFU/mL, as measured with a 6400 Spectrophotometer (Jenway Ltd., Felsted, Dunmow, UK) and confirmed by plate count with the optimal media. The incubations were at 30 or 44°C for mesophilic and thermophilic strains, respectively and at 7°C for both groups. The pH was measured at 2-h intervals for the first 8 h, and then at 24, 48 and 72 h after inoculation and incubation at 30 and 44°C. The pH was monitored for 7 d following inoculation and incubation at 7°C. Diacetyl production was determined as described by King (1948). Briefly, UHT milk was inoculated with LAB and incubated for 24 h at 30°C. A 1-mL aliquot of the culture was added to a tube containing 0.5 mL of α-naphthol (1%, w/v) and KOH (16%, w/v) and incubated at 30°C for 10 min. Diacetyl generation was indicated by the formation of a red ring at the top of the tube. The antimicrobial activity of each LAB strain was first detected by the agar-spot deferred method, and the strains displaying positive results were subsequently tested with the well diffusion assay (WDA) (Schillinger and Lücke, 1989). Both assays were performed as modified by Corsetti et al. (2008) using Lactobacillus sakei LMG2313, Listeria innocua

4202, and Listeria monocytogenes ATCC 19114 as indicator strains. All tests were carried out in triplicate. The sensitivity of the active supernatants to proteolytic enzymes was tested by

26

CHAPTER II

digesting them with proteinase K (12.5 U/mg), protease B (45 U/mg), and trypsin (10.6 U/mg) at a final concentration of 1 mg/mL in phosphate buffer (pH 7.0). After incubating for 2 h at 37°C, the remaining activity was measured by a second WDA (Settanni et al., 2005). All enzymes were purchased from Sigma-Aldrich (St. Louis, MO).

3. RESULTS

3.1. Microbiological analyses

The viable cell counts for the Vastedda cheeses are displayed in Table 1.

Table 1. Samples of PDO Vastedda della valle del Belìce cheese analysed.

Samplesa Days from Season Bacterial count production M17 MRS PCA VRBGA 30°C 44°C 30°C 44°C 30°C 37°C VC1 14 winter 7.9  0.4 8.1  0.3 6.6  0.6 7.1  0.5 7.5  0.3 <1 VC2 14 winter 8.9  0.9 8.9  0.4 8.4  0.2 8.4  0.4 8.6  0.3 3.7  0.2 VC3 14 winter 9.2  0.7 9.2  0.4 5.7  0.7 8.2  0.1 8.9  0.2 3.4  0.5 VC4 13 winter 8.1  0.4 8.6  0.6 8.1  0.7 8.0  0.4 7.9  0.3 <1 VC5 11 winter 7.4  0.8 7.9  0.5 6.7  0.4 6.3  0.3 7.6  0.6 5.1  0.1 VC6 14 winter 9.0  0.3 9.3  0.4 8.0  0.2 8.1  0.3 8.6  0.6 5.6  0.4 VC1 14 spring 8.6  0.7 9.0  0.8 8.4  0.3 8.4  0.6 8.1  0.6 <1 VC2 14 spring 8.8  0.4 8.8  0.6 5.7  0.5 7.0  0.4 8.5  0.3 4.1  0.2 VC3 14 spring 8.5  0.2 8.4  0.2 7.0  0.7 7.2  0.6 8.2  0.4 3.3  0.8 VC4 12 spring 8.5  0.1 8.6  0.5 8.3  0.9 8.3  0.8 8.0  0.6 <1 VC5 14 spring 8.6  0.4 8.5  0.3 8.2  0.5 8.2  0.4 8.2  0.5 5.3  0.5 VC6 13 spring 8.6  0.3 8.7  0.1 8.4  0.3 8.4  0.3 8.7  0.2 5.6  0.6 VC7 13 spring 8.4  0.4 8.9  0.0 8.4  0.2 8.4  0.1 8.7  0.2 <1 Statistical significanceb: Factory (F) *** NS *** ** *** *** Season (S) NS *** *** * * *** F*S *** *** *** *** NS *** a origin of samples: VC1, factory A, Salemi (TP); VC2, factory B, Partanna (TP); VC3, factory C, Sambuca di Sicilia (AG); VC4, factory D, S. Margherita Belìce (AG); VC5, factory E, Poggioreale (TP); VC6, factory F, Menfi (AG); VC7, factory G, Contessa Entellina (PA). b Evaluated considering data of Vastedda cheese of the factories A-F. P value: ***, P<0.001; **, P< 0.01; *, P<0.05; NS, not significant. Abbreviations: MRS, de Man-Rogosa-Sharpe agar for mesophilic rod LAB; M17, medium 17 agar for mesophilic coccus LAB; PCA, Plate Count agar for total mesophilic count; VRBGA, Violet Red Bile Glucose agar for Enterobacteriaceae; VC, Vastedda cheese. Results indicate mean values±S.D. of three plate counts.

Both the origin of the cheese factory and the season of production significantly affected growth of the mesophilic and thermophilic rod-shaped LAB and Enterobacteriaceae. There were no significant differences between the seasons for the mesophilic coccus-shaped LAB or between the different factories for the thermophilic coccus-shaped LAB. With the exception of the TMC, the interaction between the factories and seasons were significant (P<0.001) for all microbial groups. All of the cheese samples were dominated by coccus-shaped LAB, and there were no statistically significant differences between the population levels of thermophilic and

27

CHAPTER II

mesophilic bacteria. For the rod-shaped LAB, significant differences in the number of thermophilic and mesophilic bacteria present in the VC3 sample produced in the winter and the VC2 sample produced in the spring were observed. In general, the cell counts for the thermophilic groups were higher than those for the mesophilic LAB. Enterobacteria were present in the VC2, VC3, VC5 and VC6 samples from the winter and spring productions, with the bacterial concentrations in the VC5 and VC6 samples being the highest in both seasons.

3.2. Isolation and grouping of lactic acid bacteria

Based on their appearance, about four colonies with similar attributes (colour, edge, surface and elevation) were isolated from each thermophilic and mesophilic LAB culture for each of the morphologies identified on the agar plates. Colonies were picked from the plates inoculated with the most diluted sample. A total of 1044 colonies were collected from 39 cheese samples. All of the cultures were inspected microscopically and classified as cocci (928) or rods (116). Gram-positive and catalase-negative cultures were considered presumptive LAB, and produced 894 of the cocci and all 116 of the rod cultures requiring further examination. Based on several phenotypic features of the cultures and combinations of these features, the 1010 LAB cultures were separated into 29 groups (Table 2), 3 for rods and 26 for cocci. The largest groups, composed of more than 100 isolates, were groups 22 and 27. In contrast, groups 13, 14 and 18 had only 2 or 3 isolates. Groups 1 and 2 were determined to have a homo-fermentative metabolism and were further characterised by their growth or lack thereof in the presence of ribose. Group 1 was found to have an obligate homo-fermentative metabolism that was characterised by the ability to grow at 45°C, but not at 15°C, resulting in the classification of the group as thermophilic lactobacilli. Group 2 was determined to be composed of facultative homo-fermentative mesophilic lactobacilli that grew at 15°C, but not at 45°C.

28

Table 2. Phenotypic grouping of LAB isolated from PDO Vastedda della valle del Belìce cheeses.

Characters Clusters

(n=10)

1 (n=72) 2 (n=10) 3 (n=34) 4 (n=14) 5 (n=20) 6 (n=18) 7 (n=67) 8 (n=47) 9 (n=55) 10 (n=17) 11 (n=51) 12 (n=11) 13 (n=2) 14 (n=3) 15 (n=68) 16 17 (n=21) 18 (n=2) 19 (n=14) 20 (n=72) 21 (n=14) 22 (n=127) 23 (n=16) 24 (n=57) 25 (n=17) 26 (n=28) 27 (n=115) 28 (n=11) 29 (n=17)

Morphologya R R R C C C C C C C C C C C C C C C C C C C C C C C C C C

Cell dispositionb sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc lc lc lc lc lc lc lc lc lc lc lc

Growth:

15°C - + - + + + + + + + + + + + + + + + ------

45°C + - + ------+ + + - - + + + + + + + + + + + + +

pH 9.2 nd nd nd ------+ + + + ------

6.5% NaCl nd nd nd + + + + + + + + + + + - - + + ------

Resistance to 60°C ------+ + + + - - + + ------+ + + + +

Hydrolysis of: II CHAPTER

arginine - - + - - + - + ------+ + + + ------

aesculin - - - + + - - + + - - - + - + + + + + + + - - - - + - - -

Acid production from:

arabinose - - + - + + + + + - + + - - - - + + ------

ribose - + + + + + + + + + + + + + + + + + + + - + - - + + + - +

xylose - - + - + + + + + - + + - + - - + + ------

fructose + + + + + + + + + + + + + + + + + + - + + + + + + + + + +

galactose + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

lactose + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

sucrose + - + - + + + + + + + + + + + + + - + + + + + + - + + + +

glycerol + + + + + + + + + + + + + + + + + + + + + + - + - + + + -

CO2 from glucose - - + + + + + + + + + + + + ------a R, rod; C, coccus. b sc, short chain; lc, long chain. nd, not determined.

29

CHAPTER II

3.3. Differentiation and identification of lactic acid bacteria

Following the methodology described by De Angelis et al. (2001), approximately 30% of the isolates from each phenotypic group, for a total of 336 isolates, were selected from the different VC samples and subjected to RAPD analysis. The reproducibility of the RAPD-PCR fingerprints was assessed by comparing the PCR products obtained with the primers M13, AB106 and AB111 using DNA extracted from 3 separate cultures of 2 strains from each phenotypic group. The lowest similarity found was 92%, indicating the results were highly reproducible using this technique under the applied conditions (data not shown). The genotypic differentiation by RAPD-PCR analysis distinguished 74 strains (Fig. 2). Although the strains belonging to the phenotypic groups 15 and 27 had very similar phenotypic profiles, they clustered quite far apart from one another after the RAPD-PCR analysis. The 74 strains were identified by sequencing of the 16S rRNA gene (Tables 3 and 4). The sequence lengths ranged from 1,454–1,509 bp and covered most of the 16S rRNA genes occurring in LAB, which in some species is longer than 1600 bp. The BLAST searches produced percentages of identity with sequences available in the NCBI database of at least 97%, which is the minimum level of similarity required between 16S rRNA genes from two strains to be considered as belonging to the same species (Stackebrandt and Goebel 1994), although Stackebrandt and Ebers (2006) have proposed that “a 16S rRNA gene sequence similarity range above 98.7-99% should be mandatory for testing the genomic uniqueness of a novel isolate”. The unequivocal identification of the strains belonging to the Enterococcus genus was obtained by applying the multiplex PCR assay on the sodA gene as described by Jackson et al. (2004). All of the strains (31 mesophilic and 43 thermophilic) were confirmed to belong to 16 species included in the LAB group. The species with the highest number of strains were Leuconostoc mesenteroides (n = 17), Streptococcus thermophilus (n = 15), Lactococcus lactis (n = 7) and Streptococcus gallolyticus subsp. macedonicus (n = 13). One strain (PON203) could not be classified as any of the described species because it shared only 96% identity with Lactococcus lactis subsp. lactis S49-2 (Acc. No. HM058666).

30

CHAPTER II

Strains Phenotypicgroup 35 40 45 50 55 60 65 70 75 80 85 90 95 100

80.0 PON59 23 53.3 PON413 19 PON449 29 PON54 27 88.9 22 51.3 PON70 79.3 PON1 27 65.3 PON53 22 PON58 28 59.4 PON207 24 39.6 75.0 PON208 22

66.1 PON61 23

85.7 PON189 24 PON517 26 80.0 PON401 26 PON405 1 62.5 PON94 18 PON458 20 80.0 PON120 22 56.8 20 74.1 PON307 92.3 PON69 22 80.1 PON459 27

54.5 PON305 22 85.7 PON242 22 PON244 27 85.7 PON6 20 53.8 71.1 PON232 22 85.7 PON3 27 PON246 27 83.3 PON259 7 73.3 PON261 27 PON511 7 88.9 PON12 17 78.0 PON587 11 52.0 73.8 PON134 15 83.3 PON65 25 37.1 PON89 10 64.3 90.9 PON173 20 75.6 PON267 5 55.5 92.3 PON125 9 74.0 55.5 PON137 8

90.9 PON245 11 49.1 PON252 7 54.0 72.7 PON85 17 PON139 9 71.4 PON42 8 PON175 20 83.3 PON9 3 46.5 72.7 PON14 3 62.2 PON45 3 PON79 1 80.0 PON57 17 63.6 PON411 12 PON82 17 57.0 43.8 90.9 PON316 12 32.6 71.8 PON335 13 67.2 PON111 17 PON423 12 75.0 PON238 9 58.3 PON559 7 PON256 1 51.7 88.9 PON113 24 69.4 41.2 PON271 6 59.3 PON169 8 85.7 PON221 11 PON560 4 88.9 PON46 15 73.3 PON239 15 65.7 PON153 15 90.9 PON105 21 57.3 PON203 15 85.7 PON36 15 73.3 PON37 16 PON101 15 PON461 2

Fig. 2. Dendogram obtained from combined RAPD-PCR patterns of the LAB strains isolated from Vastedda della valle del Belìce cheeses. Scale bar indicate the percentage of similarity.

31

Table 3. Identification and inhibitory activity of mesophilic LAB isolated from PDO Vastedda della valle del Belìce cheeses.

Phenotypic Species Strain Cheese Season of first Persistencea Acc. No. BLAST Sequence Diacetyl Bacteriocin-like inhibitory activityb group sample isolation homology length (bp) production Indicator strainsc (%) 19114 4202 2313 2 Lb. curvatus PON461 VC4 spring n.e. KC545936 99 1,508 - - - - 4 Ln. mesenteroides PON560 VC5 spring n.e. KC545940 99 1,488 - - - - 5 Ln. mesenteroides PON267 VC1 winter X KC545923 99 1,486 - - - - 6 Ln. mesenteroides PON271 VC1 winter n.f. KC545924 99 1,483 - - - - 7 Ln. mesenteroides PON252 VC1 winter n.f. KC545920 99 1,485 + - - - 7 Ln. mesenteroides PON259 VC1 winter n.f. KC545921 100 1,498 + - - 1.10 ± 0.12 7 Ln. mesenteroides PON511 VC6 spring n.e. KC545937 99 1,487 - - - - 7 Ln. mesenteroides PON559 VC5 spring n.e. KC545939 99 1,492 - - - - 8 Ln. mesenteroides PON42 VC2 winter n.f. KC545888 99 1,483 - - - - 8 Ln. mesenteroides PON137 VC5 winter n.f. KC545907 99 1,478 - - - - 8 Ln. mesenteroides PON169 VC6 winter n.f. KC545910 99 1,487 + 1.10 ± 0.12 - 1.40 ± 0.10 9 Ln. mesenteroides PON139 VC5 winter n.f. KC545908 99 1,484 - - - - 9 Ln. mesenteroides PON125 VC5 winter n.f. KC545905 99 1,479 + - - -

CHAPTER II CHAPTER 9 Ln. mesenteroides PON238 VC2 winter n.f. KC545943 99 1,492 - - - - 10 Ln. mesenteroides PON89 VC5 winter X KC545899 99 1,498 + 1.60 ± 0.17 1.17 ± 0.06 - 11 Ln. mesenteroides PON221 VC2 winter n.f. KC545914 99 1,488 + - - 1.20 ± 0.00 11 Ln. mesenteroides PON245 VC1 winter n.f. KC545881 99 1,488 + - - - 11 Ln. mesenteroides PON587 VC5 spring n.e. KC545941 99 1,494 + - - - 12 Ln. lactis PON316 VC3 spring n.e. KC545927 100 1,474 - 1.60 ± 0.17 1.30 ± 0.06 1.10 ± 0.06 12 Ln. lactis PON411 VC1 spring n.e. KC545931 99 1,503 - 1.40 ± 0.17 1.30 ± 0.15 1.10 ±0.06 12 Ln. lactis PON423 VC1 spring n.e. KC545933 99 1,499 - 1.10 ± 0.12 1.30 ± 0.12 - 13 Ln. lactis PON335 VC3 spring n.e. KC545928 98 1,487 - 1.00 ± 0.06 1.20 ± 1.30 ± 0.12 14 Ln. lactis PON57 VC3 winter n.f. KC545892 99 1,486 + 1.30 ± 0.12 1.10 ± 0.12 - 15 Lc. lactis PON101 VC6 winter X KC545901 100 1,471 - - - - 15 Lc. lactis PON134 VC5 winter n.f. KC545906 99 1,484 + - - - 15 Lc. lactis PON153 VC6 winter n.f. KC545909 99 1,480 + - - 1.30 ± 0.10 15 Lactococcus spp. PON203 VC3 winter X KC545912 96 1,501 - - - - 15 Lc. lactis PON239 VC2 winter n.f. KC545916 99 1,469 - - - - 15 Lc. lactis PON36 VC2 winter X KC545886 99 1,484 + 1.10 ± 0.15 - 1.20 ± 0.12 15 Lc. lactis PON46 VC2 winter n.f. KC545890 99 1,483 - - - - 16 Lc. lactis PON37 VC2 winter n.f. KC545887 99 1,487 + - - - a evaluated only for the strains isolated from winter cheese productions. b Width of the inhibition zone (mm). Results indicate mean ± S.D. of three independent experiments. c Bacterial species: Listeria monocytogenes ATCC 19114; Listeria innocua 4202; Lactobacillus sakei 2313. Abbreviations: Lb., Lactobacillus; Ln., Leuconostoc; Lc., Lactococcus; n.e., not evaluated; n.f., not found in spring cheese. Symbols: X, found also in spring cheeses; -, no inhibition.

32

Table 4. Identification and inhibitory activity of thermophilic LAB isolated from PDO Vastedda della valle del Belìce cheeses.

Phenotypic Species Strain Cheese Season of first Persistencea Acc. No. BLAST Sequence Diacetyl Bacteriocin-like inhibitory activityb group sample isolation homology length (bp) production Indicator strainsc (%) 19114 4202 2313 1 Lb. delbrueckii PON79 VC4 winter X KC545897 100 1,500 - - - - 1 Lb. delbrueckii PON256 VC1 winter n.f. KC545944 99 1,456 - - - 1.00 ± 0.12 1 Lb. delbrueckii PON405 VC1 spring n.e. KC545930 99 1,507 - - - - 3 Lb. fermentum PON9 VC1 winter n.f. KC545884 99 1,509 - 1.30 ± 0.06 1.20 ± 0.00 1.30 ± 0.15 3 Lb. fermentum PON14 VC1 winter n.f. KC545885 99 1,505 - 1.30 ± 0.12 1.30 ± 0.12 1.40 ± 0.10 3 Lb. fermentum PON45 VC2 winter n.f. KC545889 99 1,508 - - 1.10 ± 0.15 - 17 E. durans PON12 VC12 winter n.f. KF147885 99 1,500 - - - - 17 E. gallinarum PON82 VC5 winter n.f. KF147889 99 1,495 - - - - 17 E. faecalis PON85 VC5 winter n.f. KC545898 99 1,494 + - - - 17 E. faecium PON111 VC6 winter n.f. KF147883 99 1,498 + 1.10 ± 0.12 1.00 ± 0.06 1.20 ± 0.00 18 E. faecium PON94 VC5 winter n.f. KC545900 99 1,497 + 1.00 ± 0.12 1.00 ± 0.06 1.00 ± 0.12 19 S. termophilus PON413 VC1 spring n.e. KC545932 99 1,472 - - - -

20 S. termophilus PON6 VC1 winter n.f. KC545883 100 1,481 + - - 1.00 ± 0.06 II CHAPTER 20 S. termophilus PON307 VC3 spring n.e. KC545926 100 1,473 + - - - 20 S. termophilus PON458 VC4 spring n.e. KC545934 100 1,508 - - 1.10 ± 0.06 - 20 S. gallolyticus subsp. PON173 VC6 winter n.f. KF147882 99 1,481 - - - - macedonicus 20 S. lutetiensis PON175 VC5 winter n.f. KF147890 99 1,480 + - - - 21 S. lutetiensis PON105 VC6 winter n.f. KC545902 99 1,483 + - - -

22 S. gallolyticus subsp. PON53 VC3 winter n.f. KC545891 99 1,483 - - - - macedonicus 22 S. termophilus PON69 VC5 winter X KC545895 99 1,484 + - - - 22 S. gallolyticus subsp. PON70 VC4 winter n.f. KC545896 99 1,487 - - - - macedonicus 22 S. termophilus PON120 VC5 winter n.f. KC545904 99 1,490 - - - - 22 S. gallolyticus subsp. PON208 VC3 winter n.f. KC545913 99 1,479 - - - - macedonicus 22 S. bovis PON232 VC2 winter n.f. KC545915 98 1,486 - 1.40 ± 0.15 1.10 ± 0.06 1.10 ± 0.06 22 S. termophilus PON242 VC1 winter n.f. KC545917 99 1,480 + - - - 22 S. termophilus PON305 VC3 spring n.e. KC545925 99 1,478 - - - - 23 S. gallolyticus subsp. PON59 VC3 winter n.f. KC545894 99 1,484 + 1.50 ± 0.17 1.00 ± 0.12 1.00 ± 0.06 macedonicus 23 S. gallolyticus subsp. PON61 VC3 winter n.f. KC545942 99 1,488 - 1.10 ± 0.12 - - macedonicus 24 S. termophilus PON113 VC5 winter n.f. KC545903 99 1,480 + - - - 24 S. gallolyticus subsp. PON189 VC3 winter n.f. KC545911 99 1,466 - - - 1.20 ± 0.00 macedonicus

33

24 S. gallolyticus subsp. PON207 VC3 winter n.f. KF147886 99 1,482 - - - - macedonicus 25 S. gallolyticus subsp. PON65 VC3 winter n.f. KF147884 99 1,483 + - - - macedonicus 26 S. termophilus PON401 VC1 spring n.e. KC545929 99 1,480 - - - - 26 S. infantariusus PON517 VC6 spring n.e. KC545938 99 1,486 - - - - 27 S. gallolyticus subsp. PON1 VC1 winter n.f. KC545945 98 1,496 + - - - macedonicus 27 S. termophilus PON3 VC1 winter n.f. KC545882 100 1,494 - - - - 27 S. gallolyticus subsp. PON54 VC3 winter n.f. KF147887 99 1,489 - - - - macedonicus 27 S. termophilus PON244 VC1 winter n.f. KC545918 100 1,454 + 1.30 ± 0.12 1.10 ± 0.12 - 27 S. termophilus PON246 VC1 winter n.f. KC545919 99 1,482 + - - - 27 S. termophilus PON261 VC1 winter n.f. KC545922 99 1,482 + - - - 27 S. termophilus PON459 VC4 spring n.e. KC545935 100 1,479 - - - - 28 S. gallolyticus subsp. PON58 VC3 winter n.f. KC545893 99 1,480 + - - - macedonicus 29 S. gallolyticus subsp. PON449 VC4 spring n.f. KF147888 99 1,478 + - - - macedonicus II CHAPTER a evaluated only for the strains isolated from winter cheese productions. b Width of the inhibition zone (mm). Results indicate mean ± S.D. of three independent experiments. c Bacterial species: Listeria monocytogenes ATCC 19114; Listeria innocua 4202; Lactobacillus sakei 2313. Abbreviations: Lb., Lactobacillus; E., Enterococcus; S., Streptococcus; n.e., not evaluated; n.f., not found in spring cheese. Symbols: X, found also in spring cheeses; +, positive for diacetyl production; -, negative for diacetyl production or, in case of antibacterial tests, no inhibition found.

34

CHAPTER II

3.4. Evaluation of the general dairy characteristics of the lactic acid bacteria

The results of the acidification capacity assay done at the optimal growth temperature for each of the 74 LAB strains are reported in Table 5 and 6. Milk cultures inoculated with S. thermophilus PON244 and PON305 (Table 5), Lactococcus spp. PON203 and Ln. lactis PON335 (Table 6) displayed the greatest decrease in pH as these strains after 24 h.

Table 5. Values of pH registered for the mesophilic LAB isolated during different productions of Vastedda della valle del Belìce cheeses.

Species Strains Time (h) 0 2 4 6 8 24 48 72 Lb. curvatus PON 461 6.57 6.57 6.36 6.36 6.30 4.74 3.90 3.69 Ln. mesenteroides PON 560 6.65 6.65 6.56 6.31 6.12 4.32 4.24 4.21 Ln. mesenteroides PON 267 6.65 6.65 6.35 6.23 6.20 4.42 4.17 4.16 Ln. mesenteroides PON 271 6.65 6.65 6.62 6.54 6.28 4.09 3.74 3.71 Ln. mesenteroides PON 252 6.65 6.65 6.35 6.34 6.11 4.36 4.12 4.10 Ln. mesenteroides PON 259 6.67 6.67 6.36 6.21 5.45 4.25 4.15 4.15 Ln. mesenteroides PON 511 6.65 6.65 6.57 6.28 5.68 4.30 4.28 4.27 Ln. mesenteroides PON 559 6.64 6.64 6.54 6.21 5.54 4.32 4.28 4.26 Ln. mesenteroides PON 42 6.66 6.66 6.49 6.41 6.23 4.33 4.09 4.07 Ln. mesenteroides PON 137 6.67 6.67 6.47 6.27 6.16 4.65 4.16 4.12 Ln. mesenteroides PON 169 6.68 6.68 6.48 6.39 6.38 4.51 4.19 4.19 Ln. mesenteroides PON 139 6.66 6.66 6.32 6.07 5.87 5.12 4.45 4.33 Ln. mesenteroides PON 125 6.59 6.59 6.47 6.36 6.18 4.82 4.14 3.86 Ln. mesenteroides PON 238 6.65 6.65 6.59 6.37 5.88 4.38 4.38 4.38 Ln. mesenteroides PON 89 6.67 6.67 6.52 6.39 6.38 4.78 4.15 4.18 Ln. mesenteroides PON 221 6.67 6.67 6.60 6.41 6.28 4.47 4.31 4.30 Ln. mesenteroides PON 245 6.61 6.61 6.55 6.36 5.89 4.46 4.18 4.03 Ln. mesenteroides PON587 6.66 6.66 6.42 6.19 6.06 4.85 4.33 4.29 Ln. lactis PON316 6.65 6.65 6.29 6.15 6.01 4.54 4.31 4.27 Ln. lactis PON411 6.67 6.67 6.27 6.16 6.12 5.04 4.41 4.33 Ln. lactis PON423 6.64 6.64 6.33 6.15 6.09 5.11 4.42 4.32 Ln. lactis PON335 6.65 6.65 6.53 5.99 5.20 4.35 4.32 4.30 Ln. lactis PON57 6.67 6.67 6.43 6.30 6.15 4.99 4.27 4.24 Lc. lactis PON 101 6.57 6.57 6.21 6.21 6.06 4.51 4.14 3.93 Lc. lactis PON 134 6.62 6.62 6.43 6.24 5.95 4.93 3.96 3.76 Lc. lactis PON 153 6.66 6.66 6.20 5.89 5.52 4.26 4.12 4.12 Lactococcus spp. PON 203 6.66 6.66 6.11 5.32 4.66 4.20 4.20 4.18 Lc. lactis PON 239 6.66 6.66 6.59 6.25 5.45 4.35 4.35 4.33 Lc. lactis PON36 6.66 6.66 6.19 5.90 5.45 4.27 4.25 4.18 Lc. lactis PON 46 6.66 6.66 6.63 6.51 6.43 4.45 4.36 4.29 Lc. lactis PON 37 6.63 6.63 6.06 5.83 5.71 4.99 4.39 4.09

35

CHAPTER II

Table 6. Values of pH registered for the thermophilic LAB isolated during different productions of Vastedda della valle del Belìce cheeses.

Species Strains Teime(h) 0h 2h 4h 6h 8h 24h 48h 72h Lb. delbrueckii PON79 6.67 6.67 6.59 6.47 6.37 4.63 4.30 3.90 Lb. delbrueckii PON256 6.65 6.65 6.28 6.18 5.92 4.29 3.91 3.77 Lb. delbrueckii PON405 6.63 6.63 6.57 6.53 6.27 4.14 3.72 3.61 Lb. fermentum PON9 6.64 6.64 6.40 6.20 6.10 4.41 4.18 4.03 Lb. fermentum PON14 6.68 6.68 6.51 6.48 6.45 4.97 4.17 4.17 Lb. fermentum PON45 6.63 6.63 6.26 6.24 6.21 4.86 4.22 4.00 E. durans PON12 6.66 6.61 6.29 6.18 6.03 5.45 4.21 4.11 E. gallinarum PON82 6.67 6.66 6.44 6.42 6.36 5.66 4.24 4.19 E. faecalis PON85 6.60 6.60 6.42 6.41 6.29 4.43 4.15 4.10 E. faecium PON111 6.61 6.59 6.59 6.55 6.22 4.51 4.09 3.99 E. faecium PON94 6.59 6.59 6.50 6.38 6.21 4.61 4.12 3.93 S. termophilus PON413 6.62 6.62 6.13 5.95 5.88 5.04 4.46 4.35 S. termophilus PON6 6.66 6.66 6.41 6.03 6.02 4.63 4.16 3.90 S. termophilus PON307 6.65 6.65 6.10 5.81 5.36 4.29 4.18 4.18 S. termophilus PON458 6.59 6.59 6.56 6.54 6.22 4.43 4.09 3.82 S. gallolyticus subsp. macedonicus PON173 6.66 6.66 6.30 6.29 6.13 4.99 4.18 4.15 S. lutetiensis PON175 6.62 6.59 6.51 6.38 6.18 4.69 4.15 3.99 S. lutetiensis PON105 6.66 6.66 6.39 6.29 6.26 5.45 4.21 4.19 S. gallolyticus subsp. macedonicus PON53 6.65 6.65 6.25 6.08 5.90 5.00 3.94 3.71 S. termophilus PON69 6.64 6.64 6.30 6.18 6.02 4.64 4.11 3.92 S. gallolyticus subsp. macedonicus PON70 6.55 6.55 6.32 6.28 6.04 5.13 4.14 3.84 S. termophilus PON120 6.58 6.58 6.60 6.53 6.30 4.64 4.16 3.89 S. gallolyticus subsp. macedonicus PON208 6.65 6.65 6.30 6.29 6.15 4.82 4.18 3.90 S. bovis PON232 6.64 6.64 6.19 6.03 5.64 4.70 4.28 4.00 S. termophilus PON242 6.65 6.65 6.08 5.48 4.78 4.20 4.14 4.14 S. termophilus PON305 6.64 6.64 5.94 5.26 4.68 4.23 4.21 4.21 S. gallolyticus subsp. macedonicus PON59 6.69 6.69 6.51 6.39 6.29 5.15 4.17 4.15 S. gallolyticus subsp. macedonicus PON61 6.65 6.65 6.50 6.34 6.16 4.70 4.23 4.03 S. termophilus PON113 6.69 6.69 6.62 6.58 6.50 4.38 4.10 4.10 S. gallolyticus subsp. macedonicus PON189 6.66 6.66 6.39 6.31 6.16 4.81 4.05 3.81 S. gallolyticus subsp. macedonicus PON207 6.66 6.64 6.59 6.50 6.21 4.11 3.85 3.74 S. gallolyticus subsp. macedonicus PON65 6.69 6.69 6.59 6.44 6.28 5.98 4.22 4.11 S. termophilus PON401 6.65 6.65 6.35 6.14 5.86 4.35 3.85 3.71 S. infantariusus PON517 6.65 6.65 6.28 6.24 6.23 5.06 3.79 3.63 S. gallolyticus subsp. macedonicus PON1 6.69 6.69 6.36 6.24 6.16 4.45 4.20 4.20 S. termophilus PON3 6.69 6.69 6.55 6.50 6.48 6.03 4.62 4.21 S. gallolyticus subsp. macedonicus PON54 6.68 6.66 6.55 6.49 6.17 4.33 3.99 3.91 S. termophilus PON244 6.65 6.65 5.92 5.21 4.62 4.18 4.16 4.16 S. termophilus PON246 6.65 6.65 6.31 6.27 6.02 5.12 3.91 3.79 S. termophilus PON261 6.66 6.66 6.61 6.51 5.68 4.26 4.02 3.81 S. termophilus PON459 6.66 6.66 6.64 6.64 6.46 4.40 4.01 3.81 S. gallolyticus subsp. macedonicus PON58 6.69 6.69 6.52 6.41 6.29 4.91 4.15 4.12 S. gallolyticus subsp. macedonicus PON449 6.67 6.65 6.58 6.41 6.16 4.79 4.33 4.13

36

CHAPTER II

The results from acidifying the milk during refrigeration (results not shown) were as follows: streptococci and lactobacilli reduced the milk pH to 6.51- 6.68 after 7 d, with the exception of S. lutetiensis PON105, which decreased the pH to 5.18; enterococci slightly reduced the milk pH, reaching 6.17 for E. gallinarum PON82 by day 7; all of the leuconostocs and lactococci reduced the milk pH, but the majority did not produce decreases below 5.50. Interestingly, 3 Lc. lactis (PON101, PON134 and PON153) and 4 Ln. mesenteroides strains (PON125, PON137, PON139 and PON271) acidified the milk to a pH of 4.52-4.65 after 7 d. Diacetyl production (Tables 3 and 4) was found to be strain-dependent with 13 mesophilic isolates and 18 thermophilic isolates showing a positive result. The majority of the LAB strains with this characteristic were among the Ln. mesenteroides and S. thermophilus species. Out of the LAB isolates, 23 exhibited antibacterial activities (Tables 3 and 4), inhibiting at least 1 of the indicator strains. All Lb. fermentum and Ln. lactis strains were antibacterial compound producers and the highest activity, as measured by the inhibition area, was observed for Ln. lactis PON316. Treating the culture supernatants with proteolytic enzymes eliminated all inhibitory activity, confirming that the toxins were proteinaceous, which is a general characteristic of bacteriocins (Jack et al., 1995). Because we did not characterise the amino acid and nucleotide sequences for these substances in this study, we will be referring to them as bacteriocin-like inhibitory substances (BLIS) (Corsetti et al., 2008). With the exception of the BLIS produced by the PON6, PON153, PON189, PON221, PON256 and PON259 isolates, all of the other BLIS were active against Listeria spp..

4. DISCUSSION

Starter LAB culures are added in cheese production to determine a rapid increase in the lactic acid concentration during curd acidification or fermentation (Settanni and Moschetti 2010). One of the most promising strategies to control undesirable indigenous microbial populations is to employ well-characterised LAB. However, to avoid a loss of typicality, the addition of autochthonous microorganisms is strongly suggested (Franciosi et al., 2008). Vastedda della valle del Belìce PDO is a particular type of pasta-filata cheese produced from raw milk following the same process used for the long-term ripening of other pasta-filata cheeses, which includes acidification of the curd between 24-48 h (GUE no. C 42/16 19.2.2010), but is then consumed soon after production. Since Vastedda cheese is intended for fresh consumption and is kept under refrigeration, a real ripening of this cheese does not

37

CHAPTER II

occur. Thus, although starter LAB are generally selected only for their acidification capacity (Settanni and Moschetti 2010), for Vastedda cheese production, the LAB should also produce the aroma necessary to maintain the typicality of this cheese within a short period of time. With the aim of developing an ad hoc starter culture for the year-round production of Vastedda cheese that retains its typical characteristics, different PDO Vastedda della valle del Belìce cheeses produced in the winter and spring were used as sources of autochthonous LAB. Starter LAB cultures employed in pasta filata cheese production are commonly thermophilic. However, because we did not believe that LAB isolated from cheeses produced during the summer would be capable of meeting our goal of developing a starter culture that could acidify the curd at low winter temperatures (below 15°C), the summer productions of Vastedda cheese were not included in this study. All of the cheese samples were dominated by coccus-shaped LAB, with the highest levels approaching 9.3 Log CFU/g. The TMC and LAB concentrations detected in this study were, on average, 1 Log cycle higher than those reported by Mucchetti et al. (2008), who studied the influence of cheese making technologies on the microbial populations of Vastedda cheese. This may be due to the difference in time between the productions and analyses of the cheeses used in the two studies. Members of the Enterobacteriaceae family were not always found, but their level was estimated to be up to 105 CFU/g. Interestingly, when they were present, the concentration of Enterobacteriaceae did not greatly vary between the winter and spring productions, suggesting that their presence may have been due to insufficient hygienic conditions during milking and/or cleaning of the cheese factories. Colonies of LAB were isolated from the cheese samples and purified in order to be characterised, differentiated, and identified. The isolates were subjected to several tests generally employed to perform the phenotypic grouping of LAB based on their cell morphology, physiological traits, and biochemical characteristics (Settanni et al., 2012). With this approach, 29 LAB groups were identified. A representative percentage of the isolates was examined by RAPD-PCR, a technique commonly applied, alone or in combination with other methodologies, to discriminate LAB strains associated with food matrices. In this study, 74 different strains were found, demonstrating that high levels of LAB biodiversity exist in Vastedda cheeses. Seventy-three strains were identified at the species level and all of them belonged to the LAB group. One isolate, PON203, remained unidentified, but it is probably a new Lactococcus species, given its 96% similarity to Lc. lactis subsp. lactis. The species S. thermophilus has been found to dominate bacterial populations during the manufacturing of 38

CHAPTER II

other Italian stretched cheeses lacking commercial starters such as Provolone del Monaco (Aponte et al., 2008), Scamorza Altamurana (Baruzzi et al., 2002), and Caciocavallo Palermitano (Settanni et al., 2012). This species has been detected in the LAB employed as thermophilic starter cultures, and it is generally present in the natural whey starter culture used for cooked cheese and pasta filata cheese production (Parente et al., 1998; Settanni and Moschetti 2010). Thermophilic streptococci have also been found in high numbers during the fermentation of the Greek cheese Kasseri (Tsakalidou et al., 1994, 1998; Moatsou et al., 2001), but we have found no reports of the S. thermophilus species in raw ewes’ milk pasta filata cheeses in the literature. Among the species identified with a large number of strains, Lc. lactis and Ln. mesenteroides were dominant in the Vastedda cheeses. These species are also the main components of the mesophilic milk starter cultures used for dairy fermentations (Settanni and Moschetti 2010). This phenomenon is due to their thermal resistance during scalding of the acidified curds. Several strains within these two species were able to develop colonies after treatment at 60°C for 30 min. Exposure of the strains to high temperatures during curd stretching (80–90°C) occurs for a short period, which may explain the presence of other leuconostocs and lactococci after 14 days of ripening. Strains in another large group belonged to the species S. gallolyticus subsp. macedonicus, previously classified as Strepotococcus macedonicus, but reclassified by Schlegel et al. (2003). This species is being considered to develop adjunct cultures for use in cheese making (Settanni et al., 2011). Three strains were identified as Lb. delbrueckii, which is present in other thermophilic starter cultures (Parente and Cogan, 2004), while the other species with dairy relevance (E. faecium, E. faecalis, Lb. curvatus and Lb. fermentum) are members of the non-starter LAB community (Chamba and Irlinger, 2004). Other Streptococcus spp. were identified in the samples of Vastedda cheese analysed, in addition to S. thermophilus and S. gallolyticus subsp. macedonicus. S. bovis is associated with dairy environments (Jans et al., 2012) and is considered a pathogenic bacterium (Vaska and Faoagali, 2009). S. infantarius has been found in raw ewes’ milk cheese (Todaro et al., 2011), and S. lutetiensis was reclassified from Streptococcus infantarius subsp. coli (Poyart et al., 2002). All of these species belong to the group D streptococci and are undesirable in cheese. Some strains (5%) were found to persist during the two consecutive seasons of production (Table 3), highlighting their strong adaptation to the environment and to the transformation methods. This phenomenon is unsurprising in traditional cheese making processes, especially in those carried out using wooden equipment, which allows the formation of microbial 39

CHAPTER II

biofilms. Several wooden vats have been found to host LAB (Lortal et al., 2009; Didienne et al., 2012) and some strains have persisted over time (Settanni et al., 2012). As Vastedda cheese production is carried out using wooden vats, they may provide a reservoir of LAB to inoculate the milk, resulting in the LAB found in the final products. With a view to selecting autochthonous LAB species to standardise the production of Vastedda cheese, all of the LAB strains were evaluated based on characteristics that facilitate cheese production. The acidification capacity is of paramount importance, and an optimal starter LAB culture contains isolates that can highly acidify milk within a short period of time (Franciosi et al., 2009). The presence of diacetyl, which is a flavour compound generated as an end product of citrate metabolism by some LAB, was also evaluated, and found to be produced by several strains. Furthermore, all of the strains were tested for antibacterial compound production, and a consistent percentage of the LAB isolated from Vastedda della valle del Belìce PDO were BLIS producers, a positive attribute that ensures a competitive advantage for the starter strains. Since Vastedda cheese is refrigerated soon after production, the presence of LAB able to ferment at low temperatures should ensure the development of useful characteristics. Acidification at low temperatures was mediated by some strains of Lc. lactis and Ln. mesenteroides, which, with the exception of Ln. mesenteroides PON271, were all resistant to high temperatures.

5. CONCLUSION

Vastedda della valle del Belìce PDO cheeses produced in the winter and spring were characterised by high levels of thermophilic and mesophilic LAB. All of the major groups of strains belonged to species commonly employed as starter cultures in different cheese productions. Some strains belonging to the thermophilic species Lb. delbrueckii and S. thermophiles and the mesophilic species Lc. lactis and Ln. mesenteroides displayed characteristics favourable to dairy production, suggesting that they may be useful in the Vastedda cheese making process. Thus, the strains Lb. delbrueckii PON79, PON256, and PON405; Lc. lactis PON36, PON153, and PON203; Ln. mesenteroides PON169, PON259, and PON559; and S. thermophilus PON244, PON120 and PON242 will be evaluated in situ for their potential to act as starter cultures for the continuous four-season production of Vastedda cheese.

40

CHAPTER II

REFERENCES

Abd-el-Malek, Y., Gibson, T. (1948) Studies in the bacteriology of milk. I. The streptococci of milk. Journal of Dairy Research 15, 233–240 Aponte, M., Fusco, V., Andolfi, R., Coppola, S. (2008) Lactic acid bacteria occurring during manufacture and ripening of Provolone del Monaco cheese: detection by different analytical approaches. International Dairy Journal 18, 403–413 Antonsson, M., Molin, G., Ardo, Y. (2003) Lactobacillus strains isolated from Danbo cheese as adjunct cultures in a cheese model system. International Journal of Food Microbiology 85, 159–169 Baruzzi, F., Maturante, A., Morea, M., Cocconcelli, P.S. (2002) Microbial community dynamics during the Scamorza Altamurana cheese natural fermentation. Journal of Dairy Science 85, 1390–1397 Beresford. T.P., Fitzsimons, N.A., Brennan, N.L., Cogan, T.M. (2001) Recent advances in cheese microbiology. International Dairy Journal 11, 259–274 Chamba, J.F., Irlinger, F. (2004) Secondary and adjunct cultures. In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.). Cheese: Chemistry, Physics and Microbiology. Elsevier, London Corsetti, A., Settanni, L., Braga, T.M., De Fatima Silva Lopes, M., Suzzi, G. (2008) An investigation on the bacteriocinogenic potential of lactic acid bacteria associated with wheat (Triticum durum) kernels and non- conventional flours. LWT-Food Science and Technology 41, 1173–1182 De Angelis, M., Corsetti, A., Tosti, N., Rossi, J., Corbo, M.R., Gobbetti, M. (2001) Characterization of non- starter lactic acid bacteria from Italian ewe cheeses based on phenotypic, genotypic, and cell wall protein analyses. Applied and Environmental Microbiology 67, 2011–2020 Didienne, R., Defargues, C., Callon, C., Meylheuc, T., Hulin, S., Montel, M.C. (2012) Characteristics of microbial biofilm on wooden vats (‘gerles’) in PDO Salers cheese. International Journal of Food Microbiology 156, 91–101 Donnelly, C.W. (2004) In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.). Cheese: Chemistry, Physics and Microbiology. Chapman and Hall, London Fitzsimons, N.A., Cogan, T.M., Condon, S., Bereford, T. (1999). Phenotypic and genotypic characterization of nonstarter lactic acid bacteria in mature cheddar cheese. Applied and Environmental Microbiology 65, 3418–3426 Franciosi, E., Settanni, L., Carlin, S., Cavazza, A., Poznanski, E. (2008) A factory-scale application of secondary adjunct cultures selected from lactic acid bacteria during “Puzzone di Moena” cheese ripening. Journal of Dairy Science 91, 2981–2991 Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Biodiversity and technological potential of wild lactic acid bacteria from raw cows’ milk. International Dairy Journal 19, 3–11 Harrigan, W.F., McCance, M.E. (1976) Laboratory Methods in Food and Dairy Microbiology. Academic Press, New York. Jack, R.W., Tagg, J.R., Ray, B. (1995) Bacteriocins of Gram-positive bacteria. Microbiological Reviews 59, 171–200 Jackson, C.R., Fedorka-Cray, P.J., Barrett, J.B. (2004) Use of a genus- and species-specific multiplex PCR for identification of enterococci. Journal of Clinical Microbiology 42, 3558–3565 Jans, C., Lacroix, C., Meile, L. (2012) A novel multiplex PCR/RFLP assay for the identification of Streptococcus bovis/Streptococcus equinus complex members from dairy microbial communities based on the 16S rRNA gene. FEMS Microbiology Letters 326, 144–150 King, N. (1948) Modification of Voges–Proskauer test for rapid colorimetric determination of acetyl methyl carbimol plus diacetyl in butter. Dairy Industries 13, 860–866 Lortal, S., Di Blasi, A., Madec, M.N., Pediliggieri, C., Tuminello, L., Tanguy, G., Fauquant, J., Lecuona, Y., Campo, P., Carpino, S., Licitra, G. (2009) Tina wooden vat biofilm. A safe and highly efficient lactic acid bacteria delivering system in PDO Ragusano cheese making. International Journal of Food Microbiology 132, 1–8 Micari, P., Sarullo, V., Sidari, R., Caridi, A. (2007) Physico-chemical and hygienic characteristics of the Calabrian raw milk cheese, Caprino d’Aspromonte. Turkish Journal of Veterinary and Animal Sciences 31, 55–60

41

CHAPTER II

Moatsou, G., Kandarakis, I., Moschopoulou, E., Anifantakis, E., Alichanidis, E. (2001) Effect of technological parameters on the characteristics of kasseri cheese made from raw or pasteurized ewes’ milk. International Journal of Dairy Technology 54, 69–77 Mucchetti, G., Bonvini, B., Remagni, M.C., Ghiglietti, R., Locci, F., Barzaghi, S., Francolino, S., Perrone, A., Rubiloni, A., Campo, P., Gatti, M., Carminati, D. (2008) Influence of cheese-making technology on composition and microbiological characteristics of Vastedda cheese. Food Control 19, 119–125 Parente, E., Cogan, T.M. (2004) In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.). Cheese: Chemistry, Physics and Microbiology. Chapman and Hall, London Parente, E., Moschetti, G., Coppola, S., (1998) Starter cultures for Mozzarella cheese. Annals of Microbiology 48, 89–109 Poyart, C., Quesne, G., Trieu-Cuot, P. (2002) Taxonomic dissection of the Streptococcus bovis group by analysis of manganese-dependent superoxide dismutase gene (sodA) sequences: reclassification of 'Streptococcus infantarius subsp. coli' as Streptococcus lutetiensis sp. nov. and of Streptococcus bovis biotype 11.2 as Streptococcus pasteurianus sp. nov. International Journal of Systematic and Evolutionary Microbiology 52, 1247–1255 Qadri, S.M.H., Desilva, M.I., Zubairi, S. (1980) Rapid test for determination of esculin hydrolysis. Journal of Clinical Microbiology 12, 472–474 Salvadori del Prato, O. (1998) Tecnologia Lattiero-Casearia. Edagricole, Bologna, Italy Schillinger, U., Lücke, F.K. (1989) Antibacterial activity of Lactobacillus sake isolated from meat. Applied and Environmental Microbiology 55, 1901–1906 Schlegel, L., Grimont, F., Ageron, E., Grimont, P.A.D., Bouvet, A. (2003) Reappraisal of the taxonomy of the Streptococcus bovis/Streptococcus equinus complex and related species: description of Streptococcus gallolyticus subsp. gallolyticus subsp. nov., S. gallolyticus subsp. macedonicus subsp. nov. and S. gallolyticus subsp. pasteurianus subsp. nov. International Journal of Systematic and Evolutionary Microbiology 53, 631–645 Settanni, L., Di Grigoli, A., Tornambé, G., Bellina, V., Francesca, N., Moschetti, G., Bonanno, A. (2012) Persistence of wild Streptococcus thermophilus strains on wooden vat and during the manufacture of a Caciocavallo type cheese. International Journal of Food Microbiology 155, 73–81 Settanni, L., Franciosi, E., Cavazza, A., Cocconcelli, P.S., Poznanski, E. (2011) Extension of Tosèla cheese shelf-life using non-starter lactic acid bacteria. Food Microbiology 28, 883–90 Settanni, L., Massitti, O., Van Sinderen, D., Corsetti, A. (2005) In situ activity of a bacteriocin-producing Lactococcus lactis strain. Influence on the interactions between lactic acid bacteria during sourdough fermentation. Journal of Applied Microbiology 99, 670–681 Settanni, L., Moschetti, G. (2010) Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiology 27, 691–697 Sherman, J.M. (1937) The streptococci. Bacteriological Reviews 1, 3–97 Stackebrandt, E., Goebel, B.M. (1994) Taxonomic Note: A Place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present species definition in bacteriology. International Journal of Systematic Bacteriology 44, 846–849 Stackebrandt, E., Ebers, J. (2006) Taxonomic parameters revisited: tarnished gold standards. Microbiology Today 33, 152–155 Todaro, M., Francesca, N., Reale, S., Moschetti, G., Vitale, F., Settanni, L. (2011) Effect of different salting technologies on the chemical and microbiological characteristics of PDO Pecorino Siciliano cheese. European Food Research and Technology 233, 931–940 Tsakalidou, E., Manolopoulou, E., Kabaraki, E., Zoidou, E., Pot, B., Kersters, K., Kalantzopoulos, G. (1994) The combined use of whole-cell protein extracts for the identification (SDS-PAGE) and enzyme activity screening of lactic acid bacteria isolated from traditional Greek dairy products. Systematic and Applied Microbiology 17, 444–458 Tsakalidou E, Zoidou E, Pot B, Wassill L, Ludwig W, Devriese LA, Kalantzopoulos G, Schleifer KH, Kersters K (1998) Identification of streptococci from Greek Kasseri cheese and description of Streptococcus macedonicus sp. nov. International Journal of Systematic Bacteriology 48, 519–527 Vaska, V.L., Faoagali, J.L. (2009) Streptococcus bovis bacteraemia: identification within organism complex and association with endocarditis and colonic malignancy. Pathology 41, 183–186 42

CHAPTER II

Weisburg, W. Barns, S.M., Pelletier, D.A., Lane, D.J. (1991) 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology 173, 697–703 Williams AG, Choi, SC, Banks JM (2002) Variability of the species and strain phenotype composition of the non-starter lactic acid bacterial population of Cheddar cheese manufactured in a commercial creamery. Food Reviews International 35, 483–493

43

CHAPTER III

Composition and characterisation of the lactic acid bacterial biofilms associated with the wooden vats used to produce two

traditional stretched cheeses

The present chapter has been submitted for publication in

International Journal of Food Microbiology

44

CHAPTER III

ABSTRACT

The biofilms of 12 wooden vats used for the production of the traditional stretched cheeses Caciocavallo Palermitano and PDO Vastedda della valle del Belìce were investigated. Salmonella spp. and L. monocytogenes were never detected, total coliforms were at low numbers with E. coli found only in three vats. Coagulase-positive staphylococci (CPS) were below the detection limit, whereas lactic acid bacteria (LAB) dominated the surfaces of all vats. In general, the dominance was showed by coccus LAB. Enterococci were estimated at high numbers, but almost at 2 Log cycles lower than other LAB. LAB populations were investigated at species and strain level and for their technological properties relevant in cheese production. Eighty-five strains were analyzed by a polyphasic genetic approach and allotted into 15 species within the genera Enterococcus, Lactobacillus, Lactococcus, Leuconostoc, Pediococcus and Streptococcus. Enterococcus faecium was found in all wooden vats and the species most frequently isolated were Enterococcus faecalis, Lactococcus lactis, Leuconostoc mesenteroides, Pediococcus acidilactici and Streptococcus thermophilus. The study of the quantitative data on acidification rate, autolysis kinetics and diacetyl production by cluster and principal component analysis led to the identification of some strains with promising dairy characteristics. The majority of LAB strains were proteolytic. Thirty-one LAB inhibited sensitive strains by bacteriocin-like inhibitory substances (BLIS). This work confirmed that the wooden vat is a safe system for cheese production.

Key words: Enterococci; Lactic acid bacteria; Raw milk; Technological screening; Traditional cheese; Wooden vat.

45

CHAPTER III

1. INTRODUCTION

Traditional Italian cheeses are often manufactured with raw milk without the addition of commercial or natural starter cultures using wooden equipment. In order to transform milk into cheese, the presence of lactic acid bacteria (LAB) is needed (Settanni and Moschetti, 2010). During cheese production, LAB are distinct in two groups: starter LAB (SLAB), responsible for the acidification of curd; non starter LAB (NSLAB), implicated in the maturation. When selected cultures are not inoculated in milk, the main sources of microbial contamination that might provide desirable LAB are generally the milk, the equipment used during processing and the dairy environment (Beresford et al., 2001; Franciosi et al., 2008). Lortal et al. (2009) assessed that lactic acid is produced both by the natural milk microbiota and that from the biofilms of the wooden vat surfaces. “The structure of the wood as porous, would absorb and trap bacteria that may contaminate food products” is what the US Food and Drug Administration declared during the launch of the warning about Italian and French cheeses ripened on wooden planks (Cutini, 2014). Lortal et al. (2009) stated that, although wood is a traditional and natural material used in cheese production (vat and shelves), the European rule discussions highlight regularly the question of food safety regarding this material. Regarding cheese ripening, it has been demonstrated that the aging on wooden planks reduces the presence of bacteria dangerous for humans, such as L. monocytogenes, showing instead a potential of wood for bio-protection against food pathogens (Mariani et al., 2011). Since several traditional cheeses are made in wooden vats, the microbial characterization of this recipients has been object of different research groups, mainly Italian and French, in the last few years (Didienne et al., 2012; Licitra et al., 2007; Lortal et al., 2009; Settanni et al., 2012). They found LAB at dominant levels, while the undesired microorganisms, including coliforms and coagulase-positive staphylococci (CPS), were at very low densities and, overall, the presence of the pathogenic bacteria L. monocytogenes and Salmonella spp. was never revealed. The wooden vats used to produce different raw cow’s milk pasta-filata cheeses in Sicily (southern Italy) have been found to host high levels of LAB, mainly belonging to the species Streptococcus thermophilus (Licitra et al., 2007; Settanni et al., 2012). Pasta-filata technology consists of a first acidification step and a subsequent scalding of the acidified curd to be moulded into the final shape (Salvadori del Prato, 1998). The dominant role of wooden vat SLAB over the indigenous SLAB hosted in the bulk milk has been demonstrated for cheeses manufactured with this technology (Lortal et al., 2009; Settanni et al., 2012).

46

CHAPTER III

The analysis of the composition of the biofilms associated with the wooden vats used to produce different French cheeses revealed the presence of several dairy LAB, including the SLAB species Lactobacillus helveticus, Lactococcus lactis and Leuconostoc mesenteroides and different NSLAB such as Lactobacillus plantarum and Lactobacillus casei (Didienne et al., 2012). Settanni et al. (2012) also evidenced the presence of other NSLAB in the wooden vat biofilms, in particular enterococci, that are considered important for the typicality of traditional cheeses (Foulquié Moreno et al., 2006), even though the safety status of some species of the Enterococcus genus must be considered. Recently, enterococci of wooden vat origin were detected in ripened cheeses, demonstrating the ability to follow the different stages of cheese making and to persist during ripening, thus influencing the characteristics of the final products (Di Grigoli et al., 2015). Although the dominance and the role of the SLAB of wooden vat origin has been verified for pasta-filata cheeses made with raw cow’s milk (Settanni et al., 2012), the contribution of the wooden vat SLAB during the manufacture of raw ewe’s milk cheeses has not been investigated yet. Furthermore, very little is known on the potential of the wooden vat NSLAB during cheese ripening. Since the wooden vat biofilms are living systems whose activities are defining during cheese production, an investigation of the characteristics of SLAB and NSLAB of these ecosystems deserves attention. For this reason, the objectives of the present work were: to quantify the transfer of microbial flora from wooden vats to milk; to enumerate LAB population of the wooden vats used to produce two different pasta-filata cheeses; to identify LAB at species level and differentiate the strains; and to evaluate their technological dairy traits in vitro.

47

CHAPTER III

2. MATERIALS AND METHODS

2.1. Biofilm collection

Twelve wooden vats (Table 1) used in 11 dairy factories located in western Sicily (Italy) producing two different pasta-filata (stretched) cheese typologies, Caciocavallo Palermitano cheese and PDO Vastedda della valle del Belìce cheese, obtained with raw cows’ and raw ewes’ milk, respectively, without the addition of starter cultures, were microbiologically investigated.

Table 1. Characteristics of the wooden vats.

Wooden City of dairy factory Age Type of Cheese Milk Milk Type of washing vat (province)a (years) wood processed volume (L) 1 Godrano (PA) 28 chestnut Caciocavallo Palermitano Bovine 160 HDW (in winter); CW (in summer) 2 Godrano (PA) 10 chestnut Caciocavallo Palermitano Bovine 400 HDW 3 S. Margherita Belìce (AG) 5 douglas Vastedda della valle del Belìce Ovine 200 HDW 4 Menfi (AG) 5 douglas Vastedda della valle del Belìce Ovine 250 CW 5 Terrasini (PA) 10 chestnut Caciocavallo Palermitano Bovine 170 HDW (in winter); CW (in summer)

6 Cinisi (PA) 10 chestnut Caciocavallo Palermitano Bovine 300 HDW 7 Godrano (PA) 10 chestnut Caciocavallo Palermitano Bovine 250 HDW 8 Terrasini (PA) 10 chestnut Caciocavallo Palermitano Bovine 300 HDW (in winter); CW (in summer) 9 Salemi (TP) 5 douglas Vastedda della valle del Belìce Ovine 190 CW 10 Salemi (TP) 7 douglas Vastedda della valle del Belìce Ovine 190 CW 11 Partanna (TP) 5 douglas Vastedda della valle del Belìce Ovine 150 HDW 12 Godrano (PA) 20 chestnut Caciocavallo Palermitano Bovine 220 HDW (in winter); CW (in summer) a AG, Agrigento; PA, Palermo; TP, Trapani. Abbreviations: HDW, Hot deproteinized whey (whey resulting after separation of whey proteins coagulated by thermal treatment); CW, cold water.

Vat surfaces (400 cm2) were sampled, just before cheese production took place, as reported by Didienne et al. (2012) using UV-treated paper squares positioned halfway up the side and on the bottom. Milk samples were also collected in this study to evaluate the effect of the vat biofilms on the microbial counts of milks. To this purpose, milks were sampled before transfer in the wooden vats (milk before contact, MBC) and after 5 min of contact with the vat surfaces (milk after contact, MAC) as reported by Didienne et al. (2012). MBC and MAC bulks were subjected to agitation before sampling. The samples were collected in duplicate at two week intervals in the spring season 2012. All samples were transported to the laboratory, under refrigeration with a portable fridge, where they were immediately analysed.

48

CHAPTER III

2.2. Microbiological analyses

Wooden vat surface and milk samples were microbiologically investigated for total mesophilic counts (TMC), enterococci, mesophilic and thermophilic rod-shaped LAB, mesophilic and thermophilic coccus-shaped LAB as reported by Settanni et al. (2012). Total coliforms were analysed following the ISO 4832 (2006), Escherichia coli the ISO 16649-2 (2001), CPS the ISO 6888-2 (1999) and Amend 1 (2003), Salmonella spp. with the screening method AFNOR BIO 12/22-05/07 (2007), enzyme immunoassay Enzyme Linked Fluorescent Assay (ELFA) performed with the automated system VIDAS (bioMérieux), and Listeria monocytogenes by the method ELFA AFNOR BIO 12/11-03/04 (2004). All media were purchased from Oxoid (Milan, Italy). Microbiological counts were performed in triplicate. Statistical analyses were conducted using STATISTICA software (StatSoft Inc., Tulsa, OK, USA). Microbial data were analysed using a generalised linear model (GLM) that included the effects of the wooden vat. Student t-tests were used to compare the means and pairwise comparisons were evaluated with a post-hoc Tukey test. A P-value <0.05 was deemed significant.

2.3. Isolation and phenotypic grouping of lactic acid bacteria

Four colonies of presumptive LAB (Gram-positive, determined by KOH method, and catalase negative, determined by transferring fresh colonies from a Petri dish to a glass slide and adding 5%, w/v, H2O2) with similar appearance (colour, edge, surface and elevation) were isolated for all morphologies detected from the highest plated dilutions on MRS, M17 and KAA to represent their diversity and transferred into the corresponding broth media, except for the isolates from KAA which were inoculated in M17 broth. The isolates were then purified with successive sub-culturing and stored in broth media containing 20% glycerol (v/v) at −80°C until further analysis. Phenotypic characterisation was carried out to obtain an initial grouping of isolates as reported by Gaglio et al. (2014) based on cell morphology, growth at 15 and 45°C, resistance at 60°C for 30 min, NH3 production from arginine, aesculin hydrolysis, acid production from the carbohydrates reported by Di Grigoli et al. (2015), and

CO2 production from glucose following the method described by Settanni et al. (2012). The coccus-shaped isolates were further grouped by their ability to grow at pH 9.2 and in the presence of NaCl (6.5 g/L) to separate enterococci which are able to grow in both conditions from other LAB. 49

CHAPTER III

2.4. Strain differentiation and identification of lactic acid bacteria

Genomic DNA for PCR assays was prepared from LAB isolates after their overnight growth in broth media at 30°C. Cells were harvested, and DNA was extracted using an InstaGene Matrix kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s protocol. Crude cell extracts were used as templates for PCR. Strain differentiation was performed with random amplification of polymorphic DNA (RAPD)-PCR analysis as previously described by Settanni et al. (2012). Genotypic identification of the different LAB strains was carried out by 16S rRNA gene sequencing as reported by Di Grigoli et al. (2015). The resulting DNA was sequenced by Istituto Zooprofilattico Sperimentale della Sicilia “A. Mirri” (IZS – Palermo, Italy) using the same primers employed for the PCR amplifications. The sequences were compared with those available in the GenBank/EMBL/DDBJ (http://www.ncbi.nlm.nih.gov) (Altschul et al., 1997) and EzTaxon-e (http://eztaxon- e.ezbiocloud.net/) (Chun et al., 2007) databases. The last database compares a given sequence to those of type strains only. The isolates were considered to represent the species in question if 97% or higher similarity was detected. The multiplex PCR assay based on the sodA gene reported by Jackson et al. (2004) was applied to better classify Enterococcus species.

2.5. Characterization of the technological properties of lactic acid bacteria

The different LAB strains were tested for some of the technological traits useful during cheese production: acidification capacity; diacetyl formation; autolytic properties; proteolytic activities; and production of antimicrobial compounds. The acidifying capacity was assayed at the optimal growth temperature for each strain in 100 mL of full fat, ultra-high temperature (UHT) milk inoculated with a 1% (v/v) cell suspension obtained by growing the cultures overnight in their optimal medium, centrifuging at 5,000 × g for 5 min, washing and re-suspending in Ringer’s solution. To standardise bacterial inocula, the cells were re-suspended in Ringer’s solution to an optical density at 600 nm of 1.00, corresponding to approximately 109 CFU/mL, as measured with a 6400 Spectrophotometer (Jenway Ltd., Felsted, Dunmow, UK) and confirmed by plate count with the optimal media. The incubations were at 30 or 44°C for mesophilic and thermophilic strains, respectively. The pH was measured on aliquots of 4 mL, aseptically collected from each test flask, at 2-h intervals for the first 8 h, and then at 24, 48 and 72 h after inoculation and incubation at 30 and 44°C.

50

CHAPTER III

Diacetyl production was determined as described by King (1948). LAB were inoculated in UHT milk and after the incubation at 30°C for 24 h, aliquots of 1 mL were added to 0.5 mL of α-naphthol (1%, w/v) and KOH (16%, w/v) and maintained at 30°C for 10 min. Diacetyl generation was indicated by the formation of a red ring at the top of the tube. The autolysis of whole cells was determined applying the method described by Mora et al.

(2003). OD600 was measured at 2-h intervals for the first 8 h and then 24, 48, and 72 h after inoculation. The extracellular protease activity of LAB was determined on agar plates, with the method described by Vermelho et al. (1996). Bovine serum albumin (BSA) and gelatine (Sigma- Aldrich, Milan, Italy) were used as protease substrates; they were incorporated into each medium at 1% (w/v). The antimicrobial activity of LAB was first detected by the agar-spot deferred method, and the strains displaying positive results were subsequently tested with the well diffusion assay (WDA) (Schillinger and Lücke 1989). Both assays (in triplicate) were performed as modified by Corsetti et al. (2008). Listeria innocua 4202, Listeria monocytogenes ATCC 19114 and Lactobacillus sakei 2313 were used as indicator strains. The supernatants showing inhibitory properties were subjected to the action of proteolytic enzymes: proteinase K (12.5 U/mg); protease B (45 U/mg); and trypsin (10.6 U/mg) at a final concentration of 1 mg/mL in phosphate buffer (pH 7.0). The tests for residual activities were performed after 2 h at 37°C. All enzymes were purchased from Sigma-Aldrich.

2.6. Multivariate analysis

Acidification rate, autolysis kinetics and production of diacetyl of LAB were subjected to multivariate analysis to evaluate their dairy potential. To this purpose, data were first elaborated through a hierarchical cluster analysis based on euclidean distances for grouping the strains into homogeneous groups according to their activities and, then, by principal component analysis (PCA). The analysis was performed with the average quantitative activity data of the 85 LAB strains assayed in this work. For the selection of the number of Principal Components or Factors, an arbitrary criterion was followed and only factors with eigen-values higher than 0.90 were retained. Cases introduced in the analysis were the 85 strains, while explanatory variables were the three technological traits considered. Statistical data elaboration was achieved by STATISTICA software.

51

CHAPTER III

3. RESULTS AND DISCUSSION

3.1. Microbiological analyses

Table 2 shows the viable counts of the microbial groups harboured on the wooden vats before milk was added and those present in milk before and after contact with the container surfaces. The results of Salmonella spp. and L. monocytogenes are not reported in table, because no surface and milk sample was scored positive for their presence. Lortal et al. (2009) assessed that the inability of pathogens to adhere or to survive in wooden vat biofilms is mainly due to the acidic conditions, determined by LAB that ferment lactose from the residual whey (Settanni et al., 2012), the competition for nutrients and the high temperatures applied for cheese cooking. TMC of wooden vats was in the range 4.32 – 6.40 Log CFU/cm2. In general, the surfaces of the vats hosted low numbers of total coliforms, except WV4 for which this bacterial group was registered at the same level of TMC. WV4 was also characterized for the highest cell density of E. coli (3.18 Log CFU/cm2). This bacterium was also found in the inner surfaces of the vats WV5 and WV12, but, in general, low levels of E. coli were registered in this study. A similar observation was reported by Didienne et al. (2012). CPS were undetectable in any vat. A low frequency of this microbial group was reported for the wooden vat biofilms analysed for pasta-filata cheeses other than those investigated in the present study (Didienne et al., 2012; Lortal et al., 2009). All the wooden vat surfaces harboured high numbers of LAB. The majority of biofilms showed the presence of coccus LAB at dominant levels, while the sample WV9 showed mesophilic and thermophilic rod LAB at almost 1 Log cycle higher than cocci. Sample WV7 displayed the highest cell density of LAB, since mesophilic cocci were detected at 6.85 Log CFU/cm2. High numbers of LAB were already reported in previous wooden vat biofilm inspections (Didienne et al., 2012; Licitra et al., 2007; Lortal et al., 2009; Settanni et al., 2012) and the dominance of cocci over rods is common to other studies. E.g. Didienne et al. (2012) detected lactococci at higher levels than lactobacilli for the majority of the vats analysed. Furthermore, within the coccus LAB community, Lortal et al. (2009) found thermophilic LAB at the highest levels, while Settanni et al. (2012) reported that mesophilic LAB were more represented than thermophilic LAB onto the surface of the vat used for Caciocavallo Palermitano cheese. These discordances might be explained by the diverse milk typologies processed and the different washing methodologies. Enterococci, detected in all wooden vat samples, were at almost 2 Log cycles lower than the LAB counted on M17 and MRS and the highest levels were found for WV4, which was filled in with ewe’s milk. No big differences, on average, were

52

CHAPTER III

registered for the levels of enterococci detected in the wooden vats used to transform different types of milk, even though the lowest concentrations were found for raw cow’s milk. The levels of enterococci found in this study were in the same order of magnitude of only one of the wooden vats used for making Ragusano cheese (Lortal et al., 2009). Milk samples before contact with the wooden vat surfaces were all dominated by LAB. In fact, the levels of TMC and LAB were almost comparable. Interestingly, six samples showed enterococci at the same level of LAB. Total coliforms, E. coli and CPS were investigated in milk only after resting in the wooden vats. After contact, total coliforms were in the range 1.53 – 5.86 Log CFU/mL and the highest levels were reached for sample MAC4, for which also the density of E. coli was the highest. E. coli was undetectable in samples MAC1, MAC2, MAC6, MAC9 and MAC10. CPS, undetectable in all wooden vat biofilms were, instead, present in all milks with the highest level (4.15 Log CFU/mL) registered for the sample MAC4. The comparison with TMC indicated the dominance of LAB over the other microbial groups investigated. A general trend was observed after contact with the wooden vat surfaces: the levels of LAB increased for all those samples of milk characterised by levels lower than 6 Log CFU/mL at delivery. The best example was provided by the sample MAC9 whose levels of mesophilic and thermophilic LAB were barely 3.60 and 3.85, respectively, before contact, but increased strongly (at 5.85 and 5.59 Log CFU/mL, respectively) after resting in the wooden vat (WV9) that hosted 6.04 Log CFU/cm2 of mesophilic rod LAB and 6.05 Log CFU/cm2 of thermophilic rod LAB. On the contrary, MAC10, which originated from the same bulk milk of MBC9, did not show a high increase in mesophilic and thermophilic rod LAB number, after contact with WV10 taht hosted lower levels of these bacterial groups than WV9. This phenomenon could be explained with the observation of Lortal et al. (2009) who registered until a level of 106 CFU/mL of thermophilic LAB released into milk, after contact with wooden vats. Thus, the effect of LAB inoculation by the vat surfaces is relevant only for milks with LAB loads lower than the maximum level reachable with the vat release, because, once in milk, vat LAB are at higher levels than indigenous milk LAB.

53

CHAPTER III

Table 2. LAB concentrationsa in biofilms of wooden vats and milk before and after contact with the vat surfaces.

Sample Bacterial counts TMC Total E. coli CPS Enterococci Rod LAB Rod LAB Coccus LAB Coccus LAB coliforms MRS-30°C MRS-44°C M17-30°C M17-44°C Wooden vat: WV1 4.94 ± 0.24 0.30 ± 0.04 <1 <1 2.30 ± 0.21 4.72 ± 0.23 4.91 ± 0.20 5.32 ± 0.30 5.86 ± 0.13 WV2 5.72 ± 0.24 2.30 ± 0.13 <1 <1 3.24 ± 0.07 3.91 ± 0.20 5.43 ± 0.21 5.42 ± 0.16 5.65 ± 0.21 WV3 4.48 ± 0.25 0.85 ± 0.07 <1 <1 3.17 ± 0.17 3.58 ± 0.13 3.04 ± 0.10 4.94 ± 0.13 5.23 ± 0.14 WV4 6.14 ± 0.18 6.04 ±0.06 3.18 ±0.20 <1 4.42 ±0.13 5.76 ± 0.21 5.85 ± 0.11 6.19 ± 0.24 6.33 ± 0.13 WV5 6.08 ± 0.11 3.18 ± 0.14 2.46 ± 0.17 <1 2.85 ± 0.11 4.75 ± 0.20 4.72 ± 0.17 6.18 ± 0.16 5.81 ± 0.10 WV6 5.02 ± 0.08 0.70 ±0.04 <1 <1 4.15 ± 0.21 4.83 ± 0.16 4.85 ± 0.11 5.77 ± 0.18 5.61 ± 0.11 WV7 6.40 ± 0.11 0.70 ± 0.10 <1 <1 4.03 ± 0.04 4.10 ± 0.14 5.75 ± 0.21 6.85 ± 0.08 6.51 ± 0.16 WV8 5.43 ± 0.06 2.00 ± 0.17 <1 <1 2.15 ± 0.21 5.07 ± 0.10 5.10 ± 0.14 5.72 ± 0.08 5.48 ± 0.11 WV9 5.72 ± 0.13 1.79 ± 0.13 <1 <1 3.50 ± 0.13 6.04 ± 0.06 6.05 ± 0.10 4.55 ± 0.21 5.28 ± 0.11 WV10 4.32 ± 0.10 0.60 ± 0.07 <1 <1 3.35 ± 0.08 4.61 ± 0.08 4.94 ± 0.07 4.51 ± 0.16 3.62 ± 0.11 WV11 5.19 ± 0.13 1.38 ± 0.07 <1 <1 2.54 ± 0.06 3.80 ± 0.14 3.96 ±0.08 4.86 ± 0.13 5.11 ± 0.16 WV12 5.36 ± 0.08 2.81 ± 0.16 2.00 ± 0.07 <1 3.30 ± 0.11 5.18 ± 0.14 4.99 ± 0.13 5.43 ± 0.18 5.49 ± 0.17 Statistical significance *** *** *** *** *** *** *** *** Milk before contact: MBC1 5.41 ± 0.16 nd nd nd 4.39 ± 0.13 4.92 ± 0.05 4.81 ± 0.14 5.00 ± 0.15 5.12 ± 0.10 MBC2 6.53 ± 0.23 nd nd nd 6.83 ± 0.25 5.94 ± 0.21 6.23 ± 0.17 6.45 ± 0.25 6.48 ± 0.08 MBC3 7.11 ± 0.11 nd nd nd 4.10 ± 0.19 7.21 ± 0.08 7.14 ± 0.23 7.01 ± 0.11 7.11 ± 0.14 MBC4 6.17 ± 0.12 nd nd nd 5.13 ± 0.07 5.52 ± 0.11 5.38 ±0.15 6.07 ± 0.30 6.06 ± 0.21 MBC5 5.41 ± 0.05 nd nd nd 3.70 ± 0.30 3.71 ± 0.18 4.05 ± 0.11 5.63 ± 0.13 4.77 ± 0.23 MBC6 4.81 ± 0.11 nd nd nd 4.70 ± 0.11 4.18 ± 0.14 4.32 ± 0.07 4.53 ± 0.21 4.40 ± 0.06 MBC7 5.84 ± 0.15 nd nd nd 5.62 ± 0.21 6.14 ± 0.18 6.09 ± 0.13 6.15 ± 0.14 6.03 ± 0.25 MBC8 7.06 ± 0.17 nd nd nd 4.40 ± 0.08 6.94 ± 0.15 6.91 ± 0.17 6.76 ±0.06 6.83 ± 0.07 MBC9b 3.95 ± 0.23 nd nd nd 3.44 ± 0.24 3.60 ± 0.25 3.85 ± 0.18 3.74 ± 0.10 3.88 ± 0.23 MBC10b 3.95 ± 0.23 nd nd nd 3.44 ± 0.24 3.60 ± 0.25 3.85 ± 0.18 3.74 ± 0.10 3.88 ± 0.23 MBC11 5.02 ± 0.10 nd nd nd 3.53 ± 0.10 3.72 ± 0.23 3.79 ± 0.18 4.92 ± 0.13 5.04 ± 0.16 MBC12 5.56 ± 0.22 nd nd nd 3.13 ± 0.07 5.36 ± 0.13 5.61 ± 0.21 5.61 ± 0.10 5.44 ± 0.11 Statistical significance *** *** *** *** *** *** Milk after contact: MAC1 5.53 ± 0.23 1.53 ± 0.14 <1 3.20 ± 0.16 4.96 ± 0.20 5.57 ± 0.20 5.59 ± 0.16 6.13 ± 0.25 6.24 ± 0.23 MAC2 6.15 ± 0.07 3.42 ± 0.13 <1 2.30 ± 0.10 6.53 ± 0.14 5.73 ± 0.14 6.16 ± 0.25 6.47 ± 0.17 6.45 ± 0.10 MAC3 7.39 ± 0.13 3.08 ± 0.10 2.73 ± 0.17 3.74 ± 0.13 5.15 ± 0.30 6.19 ± 0.23 6.81 ± 0.07 7.53 ± 0.24 7.54 ± 0.18 MAC4 7.13 ± 0.30 5.86 ± 0.07 4.11 ± 0.24 4.15 ± 0.18 5.34 ± 0.25 7.18 ± 0.18 6.07 ± 0.08 7.25 ± 0.20 7.19 ± 0.25 MAC5 5.72 ± 0.07 2.67 ± 0.20 1.23 ± 0.10 2.54 ± 0.25 4.74 ± 0.17 4.48 ± 0.25 4.30 ± 0.24 5.38 ± 0.13 4.77 ± 0.20 MAC6 5.80 ± 0.17 1.73 ± 0.13 <1 3.83 ± 0.20 5.04 ± 0.25 4.12 ± 0.10 3.70 ± 0.10 5.43 ± 0.16 5.18 ± 0.23 MAC7 6.78 ± 0.24 3.18 ± 0.24 1.23 ± 0.16 2.48 ± 0.18 6.15 ± 0.16 6.39 ± 0.13 6.33 ± 0.20 6.28 ± 0.27 6.35 ± 0.10 MAC8 7.03 ± 0.17 3.71 ± 0.10 2.18 ± 0.27 3.79 ± 0.23 5.16 ± 0.10 6.70 ± 0.25 6.55 ± 0.23 6.68 ± 0.18 6.73 ± 0.17 MAC9 4.48 ± 0.20 2.00 ± 0.18 <1 2.52 ± 0.10 4.07 ± 0.17 5.85 ± 0.14 5.59 ± 0.21 4.65 ± 0.25 5.14 ± 0.14 MAC10 4.85 ± 0.10 1.60 ± 0.10 <1 2.62 ± 0.08 4.30 ± 0.07 4.62 ± 0.27 4.49 ± 0.10 5.30 ± 0.30 5.69 ± 0.07 MAC11 5.51 ± 0.18 3.62 ± 0.16 2.49 ± 0.18 3.71 ± 0.20 4.51 ± 0.20 4.78 ± 0.08 4.57 ± 0.17 5.54 ± 0.20 6.60 ± 0.16 MAC12 6.29 ± 0.23 4.11 ± 0.27 3.54 ± 0.25 1.34 ± 0.16 4.08 ± 0.27 5.91 ± 0.16 5.86 ± 0.07 5.86 ± 0.24 6.00 ± 0.14 Statistical significance *** *** *** *** *** *** *** *** *** a Log CFU/cm2 for surfaces; Log CFU/mL for milk samples. b Originating from the same bulk milk before contact with two different wooden vats. Results indicate mean values ± S.D. of six plate counts (carried out in triplicate for two collection times). Abbreviations: MBC, milk before contact; MAC, milk after contact; TMC, total mesophilic counts; CPS, coagulase-positive staphylococci P value: ***, P≤0.001.

54

CHAPTER III

3.2. Isolation and grouping of lactic acid bacteria

A total of 713 colonies were collected from the 12 wooden vats. All of the cultures were inspected microscopically and classified as 542 cocci and 171 rods. After Gram determination and catalase test, 492 coccus-shaped and 165 rod-shaped Gram-positive and catalase-negative cultures were further examined. Based on the combination of the phenotypic features evaluated, the 657 LAB cultures were separated into 24 groups (Table 3). The highest number of groups was observed for cocci. Almost 46% of the total isolates was in group XVI. More than the half of the groups comprised a few isolates. Among rods, the groups showing a homo-fermentative metabolism were further examined for their growth or lack thereof in the presence of pentose carbohydrates. Only group I was found to have an obligate homo-fermentative metabolism. The ability to grow at 45°C but not at 15°C resulted in the classification of the groups I, II and VII as thermophilic LAB.

3.3. Genetic differentiation and identification of lactic acid bacteria

The isolates representative of each phenotypic group, for all wooden vats, were subjected to RAPD analysis. The reproducibility of this technique was verified by comparing the PCR products obtained with the three primers using DNA extracted from three separate cultures of two strains per morphology. RAPD profiles were analysed separately for each cell morphology resulting in four dendrograms (Fig. 1) for 85 dominant strains. As expected, the most numerous dendrogram included LAB cocci in short chains. The 85 strains were subjected to the 16S rRNA gene sequencing. The sequences were compared with those available in two distinct databases; all strains were clearly identified as members of the LAB community, since sequence similarity was higher than 97% in both databases (Table 4) with species within the genera Enterococcus, Lactobacillus, Lactococcus, Leuconostoc, Pediococcus and Streptococcus. Five strains, allotted into the Enterococcus genus, could not be identified at species level. Due to the different results of BLAST and EzTaxon search, these strains were further analysed by a species-specific multiplex PCR strategy which identified one Enterococcus faecium and four Enterococcus durans. The highest number of strains (n=39) belonged to E. faecium. All LAB included in the dendrograms of Fig. 1A,C,D, belonged to a single genus: Lactobacillus for LAB rods; Streptococcus for LAB cocci in long chains; Pediococcus for LAB cocci in tetrads. All strains belonging to a given species clustered closely, specifically 55

CHAPTER III

all Lactobacillus brevis, Lactobacillus reuteri (Fig. 1A), Streptococcus thermophilus (Fig. 1C), Pediococcus acidilactici and Pediococcus lolii (Fig. 1D). Regarding LAB cocci in short chains, basically the majority of strains clustered per species; one major cluster was showed by Lactococcus lactis, Leconostoc mesenteroides, E. durans and E. faecalis, while basically five main clusters were obtained for E. faecium (Fig. 1B). However, some strains clustered with different species, e.g. E. durans strains which resulted mixed within the E. faecalis group, Lc. lactis, Ln. mesenteroides and Ln. pseudomesenteroides that clustered together with E. faecium. All the species identified are commonly associated with raw milk and cheeses (Franciosi et al., 2011; Settanni and Moschetti, 2010; Wouters et al., 2002), including stretched cheeses (Gaglio et al., 2014; Morea et al., 2007; Piraino et al., 2008; Settanni et al., 2012) and several of them were also found associated wooden vats used for cheese making in Italy and France (Didienne et al., 2012; Licitra et al., 2007; Settanni et al., 2012).

56

Table 3. Phenotypic grouping of the LAB forming biofilms on the wooden vat surfaces.

Characters Clusters

I (n=8) I (n=6)II (n=16) III (n=8) IV (n=9) V (n=8) VI (n=16) VII (n=35) VIII (n=23) IX (n=8) X (n=6) XI (n=10) XII (n=30) XIII (n=61) XIV (n=7) XV (n=301) XVI (n=24) XVII (n=17) XVIII (n=7) XIX (n=10) XX (n=6) XXI (n=5) XXII (n=20) XXIII (n=16) XXIV Morphologya R R R R R R R C C C C C C C C C C C C C C C C C Cell dispositionb sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc sc t t t t t lc lc lc

Growth:

15°C - - + + + + - + + + + + + + + + + + + + + - - - 45°C + + - - + - + - - - + + + + + + + + + + + + + + pH 9.2 n.d. n.d. n.d. n.d. n.d. n.d. n.d. + - - - - + + + + + + + + + - - - 6.5% NaCl n.d. n.d. n.d. n.d. n.d. n.d. n.d. - + + + + + + + + + + + + + - - - Resistance to 60°C + + - - + - + + + - + - + + + + + + + + - + + +

Hydrolysis of: I CHAPTER

arginine - + - - + - + + + - + - + + - + - + - + - + + + aesculin + + - + + + - + + - + + + + + + - + + + - + - + c Acid production from: arabinose - + + + + + - - + + + + + + + + + + + + + + + + ribose - + + + + + + + + + + + + + + + + + + + + + + +

I xylose - + + + + + - - + + + + + + + + + + + + + + + + I

glycerol + + + + + + + + + + + + + + + + - + - + - + + + CO2 from glucose - + + + - - + - + + + + ------a R, rod; C, coccus. b sc, short chain; t, tetrads; lc long chain. c fructose, galactose, lactose and sucrose were fermented by all isolates. n.d., not determined.

57

WV_tine

A WV_tine B

Strain Species Phenotypic group Strain Species Phenotypic group

20

30

40

50

60

70

80

90

100

10 20 30 40 50 60 70 80 90 100 WVS430 Lb. brevisLactobacillus brevisXI 11 WVS418 Enterococcus faecium 11 93.3 WVS430 85.7 WVS418 E. faecium XI WVS58 Enterococcus faecium 3 XI 64.9 WVS58 E. faecium 63.0 WVS432WVS432 Lb. brevisLactobacillus brevis 11 III WVS292WVS292 E. faeciumEnterococcus faecium 8 IX 85.7 VIII 48.5 WVS416WVS416 Lb. brevisLactobacillus brevis 9 49.3 WVS388WVS388 E. faeciumEnterococcus faeciumX 10 WVS223 Lb. reuteriLactobacillus reuteriVII 7 WVS426 Enterococcus faecalis 11 76.2 WVS223 41.3 80.0 WVS426 E. faecalis 36.8 XI WVS243WVS243 Lb. reuteriLactobacillus reuteriVII 7 WVS442WVS442 E. faecalisEnterococcus faecalisXII 12 WVS87WVS87 E. duransEnterococcus durans 4 26.6 WVS279WVS279 Lb. plantarumLactobacillus plantarumVIII 8 IV WVS315 Enterococcus faecalis 2 87.5 WVS315 E. faecalis II WVS270WVS270 Lb. paracaseiLactobacillus paracaseiVIII 8 18.4 40.0 55.0 WVS53 E. faecalisEnterococcus faecalisII 2 33.3 WVS53 WVS303 Lactobacillus delbrueckii 1 WVS296 Enterococcus faecalis 1 WVS303 Lb. delbrueckii I 50.2 WVS296 E. faecalis I WVS356 E. faecalisEnterococcus faecalis 9 WVS354WVS354 Lb. fermentumLactobacillus fermentumXI 11 62.5 WVS356 IX WVS43WVS43 E. faecalisEnterococcus faecalisII 2 WVS335 Enterococcus faecium 9 88.9 WVS335 E. faecium IX 43.4 84.5 WVS343WVS343 E. faeciumEnterococcus faeciumX 10 WVS395WVS395 E. faeciumEnterococcus faeciumX 10 73.5 WVS393 Enterococcus faecium 10 90.9 WVS393 E. faecium X 60.6 87.1 WVS398WVS398 E. faeciumEnterococcus faeciumX 10 WVS422WVS422 E. faeciumEnterococcus faeciumXI 11 WVS340WVS340 E. faeciumEnterococcus faeciumIX 9 WVS359 Enterococcus faecium 9 94.1 WVS359 E. faecium IX 28.2 91.5 WVS360WVS360 E. faeciumEnterococcus faeciumIX 9 WV_tine 72.7 WVS439WVS439 E. faeciumEnterococcus faeciumXI 11 67.7 WVS231WVS231 E. faeciumEnterococcus faeciumVII 7 61.6 WVS142WVS142 E. faeciumEnterococcus faeciumV 5 40.5 WVS233WVS233 E. faeciumEnterococcus faeciumVII 7 C 38.6 WVS238WVS238 E. faeciumEnterococcus faeciumVII 7 WVS147 Leuconostoc mesenteroides 5 50.0 WVS147 Ln. mesenteroides V Strain Species Phenotypic group WVS181WVS181 Ln. pseudomesenteroidesLeuconostoc pseudomesenteroidesVI 6

CHAPTER I CHAPTER

10

20

30

40

50

60

70

80

90 100 WVS1 Enterococcus faecium 1 83.3 WVS1 E. faecium I WVS50 Streptococcus termophilus 2 WVS4 E. faeciumEnterococcus faecium 1 76.9 WVS50 S. thermophilus II 72.1 WVS4 I WVS31 Enterococcus faecium 2 33.3 WVS307 S. thermophilusStreptococcusI thermophilus 1 92.3 WVS31 E. faecium II 53.2 36.2 WVS7WVS7 E. faeciumEnterococcus faeciumI 1 20.1 WVS271 S. thermophilusStreptococcusVIII thermophilus 8 WVS443WVS443 E. faeciumEnterococcus faeciumXII 12 6.3 WVS18 S. thermophilusStreptococcus thermophilus 1 80.0 I 46.9 WVS466WVS466 E. faeciumEnterococcus faeciumXII 12 WVS105 S. lutetiensisStreptococcus lutetiensis 2 WVS106 Enterococcus faecium 4 WVS105 II 85.7 WVS106 E. faecium IV 80.4 WVS108WVS108 E. faeciumEnterococcus faeciumIV 4 WVS45WVS45 E. faeciumEnterococcus faeciumII 2 25.0 WVS195 Enterococcus faecium 6 41.4 72.7 WVS195 E. faecium VI

60.8 WVS74WVS74 E. faeciumEnterococcus faeciumIII 3 I 51.0 WVS333WVS333 E. faeciumEnterococcus faeciumVIII 8 I

WVS346WVS346 E. faeciumEnterococcus faeciumX 10 WVS143 E. faeciumEnterococcus faeciumV 5 76.9 WVS143 48.1 WVS42 Enterococcus faecium 2 69.6 WVS42 E. faecium II WVS221 Ln. pseudomesenteroidesLeuconostoc pseudomesenteroides 7 80.0 WVS221 VII

57.6 WVS264WVS264 E. faeciumEnterococcus faeciumVIII 8 WV_tine WVS37 Enterococcus faecium 2 76.9 WVS37 E. faecium II 69.7 WVS39WVS39 Lc. lactisLactococcus lactis II 2 18.4 WVS66WVS66 E. faeciumEnterococcus faeciumIII 3 WVS137 Enterococcus faecium 5 58.8 WVS137 E. faecium V WVS153WVS153 E. faeciumEnterococcus faeciumV 5 D 43.4 WVS141 Leuconostoc mesenteroides 5 94.1 WVS141 Lc. mesenteroides V

84.9 39.2 WVS159WVS159 Lc. mesenteroidesLeuconostoc mesenteroidesV 5 Strain Species Phenotypic group WVS161 Lc. mesenteroidesLeuconostoc mesenteroides 5

30 40 50 60 70 80 90 100 WVS161 V 15.0 WVS187 Lactococcus lactis 6 WVS434 P. acidilacticiPediococcus acidilacticiXI 11 33.2 WVS187 Lc. lactis VI 52.2 WVS434 WVS119 Lactococcus lactis 3 60.0 WVS119 Lc. lactis III 30.0 WVS451WVS451 P. acidilacticiPediococcus acidilacticiXII 12 53.6 WVS320WVS320 Lc. lactisLactococcus lactis II 2 WVS383WVS383 P. acidilacticiPediococcus acidilacticiX 10 WVS49WVS49 Lc. lactisLactococcus lactis II 2 9.2 26.4 WVS183 P. acidilacticiPediococcus acidilacticiVII 7 WVS464WVS464 E. faecalisEnterococcus faecalisXII 12 90.0 WVS183 WVS215WVS215 E. faeciumEnterococcus faeciumVII 7 64.6 WVS458WVS458 P. acidilacticiPediococcus acidilacticiII 2 22.2 WVS467WVS467 E. faecalisEnterococcus faecalisXII 12 WVS280 Pediococcus acidilactici 8 37.4 WVS280 P. acidilactici VIII WVS113 E. duransEnterococcus durans 4 85.7 WVS113 IV WVS222 Pediococcus lolii 6 83.3 WVS222 P. lolii VI 39.3 WVS275WVS275 E. duransEnterococcus durans VIII 8 WVS325WVS325 P. lolii Pediococcus loliiXII 12 WVS23WVS23 E. duransEnterococcus durans I 1

Fig. 1. Dendrogram obtained from combined RAPD-PCR patterns of LAB strains from wooden vats generated with three primers. A, LAB rods; B, LAB cocci in short chains; C, LAB cocci in long chains; D, LAB cocci in tetrads. Upper line indicate the percentage of similarity. Abbreviations: E., Enterococcus; Lb., Lactobacillus; Lc., Lactococcus; Ln., Leuconostoc; P., Pediococcus; S., Streptococcus.

58

CHAPTER III

Table 4. Identification of LAB strains isolated from wooden vats.

Strain Species Phenotypic % similarity (accession no. of closest relative) by: Multiplex Acc. No. Sequence group PCR length BLAST EzTaxon (sodA gene) (bp) WVS87 E. durans/E. hirae XIII 98 (FJ917736.1) 98.07 (AJ420801) E. durans KM257654 1302 WVS113 E. durans/E. hirae XIII 99 (KF149306.1) 99.78 (CP003504) E. durans KM257658 1376 WVS275 E. durans/E. hirae XIII 99 (HQ677826.1) 99.71 (CP003504) E. durans KM257683 1358 WVS23 E. durans/E. thailandicus XIII 99 (KF149142.1) 99.67 (EF197994) E. durans KM257641 1205 WVS296 E. faecalis XIV 98 (KJ395380.1) 98.76 (ASDA01000001) n.n. KM257687 1381 WVS43 E. faecalis XIV 99 (KJ725229.1) 98.76 (ASDA01000001) n.n. KM257646 1268 WVS53 E. faecalis XIV 99 (KJ726743.1) 100 (ASDA01000001) n.n. KM257650 1350 WVS315 E. faecalis XIV 99 (KF826013.1) 99.92 (ASDA01000001) n.n. KM257690 1332 WVS356 E. faecalis XIV 99 (HQ721273.1) 98.85 (ASDA01000001) n.n. KM257696 1482 WVS426 E. faecalis XIV 99 (KJ702541.1) 99.48 (ASDA01000001) n.n. KM257710 1351 WVS442 E. faecalis XIV 99 (KJ725229.1) 99.92 (ASDA01000001) n.n. KM257715 1241 WVS464 E. faecalis XV 99 (HQ721272.1) 98.46 (ASDA01000001) n.n. KM257719 1550 WVS467 E. faecalis XIV 98 (KF826013.1) 98.04 (ASDA01000001) n.n. KM257721 1483 WVS1 E. faecium XVI 98 (KF149258.1) 97.85 (DQ411813) n.n. KM257637 1351 WVS4 E. faecium XVI 99 (KJ675503.1) 99.11 (DQ411813) n.n. KM257638 1239 WVS7 E. faecium XVI 98 (KF149320.1) 97.97 (DQ411813) n.n. KM257639 1381 WVS31 E. faecium XVI 99 (JQ735957.1) 98.36 (DQ411813) n.n. KM257642 1280 WVS37 E. faecium XVI 99 (KF245564.1) 99.48 (DQ411813) n.n. KM257643 1340 WVS42 E. faecium XVI 99 (KJ156957.1) 98.64 (DQ411813) n.n. KM257645 1323 WVS45 E. faecium XVI 98 (KC351906.1) 97.40 (DQ411813) n.n. KM257647 1425 WVS58 E. faecium XVI 99 (AB932549.1) 99.17 (DQ411813) n.n. KM257651 1325 WVS66 E. faecium XVI 99 (KC222512.1) 99.33 (DQ411813) n.n. KM257652 1345 WVS74 E. faecium XVI 99 (JQ735957.1) 98.16 (DQ411813) n.n. KM257653 1308 WVS106 E. faecium XVI 97 (JX556411.1) 97.23 (DQ411813) n.n. KM257656 1510 WVS108 E. faecium XVI 99 (JN560912.1) 97.87 (DQ411813) n.n. KM257657 1409 WVS137 E. faecium XVI 99 (KF245564.1) 99.92 (DQ411813) n.n. KM257660 1299 WVS142 E. faecium XVI 99 (KF318400.1) 99.62 (DQ411813) n.n. KM257662 1309 WVS143 E. faecium XVI 98 (KF149258.1) 97.77 (DQ411813) n.n. KM257663 1348 WVS153 E. faecium XVI 99 (KF245564.1) 100.00 (DQ411813) n.n. KM257665 1354 WVS195 E. faecium XVI 99 (KF245564.1) 99.47 (DQ411813) n.n. KM257671 1337 WVS215 E. faecium XVI 98 (KF826025.1) 97.36 (DQ411813) n.n. KM257672 1376 WVS231 E. faecium XVI 99 (AB932549.1) 99.72 (DQ411813) n.n. KM257676 1411 WVS233 E. faecium XVI 99 (KC478513.1) 99.64 (DQ411813) n.n. KM257677 1374 WVS238 E. faecium XVI 98 (FJ481129.1) 97.77 (DQ411813) n.n. KM257678 1260 WVS264 E. faecium XVI 98 (KF149258.1) 97.70 (DQ411813) n.n. KM257680 1303 WVS292 E. faecium XVI 99 (KF245564.1) 99.23 (DQ411813) n.n. KM257686 1316 WVS333 E. faecium XVI 99 (KF245564.1) 99.50 (DQ411813) n.n. KM257693 1396 WVS335 E. faecium XVI 99 (KF245564.1) 99.92 (DQ411813) n.n. KM257694 1303 WVS340 E. faecium XVI 99 (KJ803878.1) 99.49 (DQ411813) n.n. KM257697 1381 WVS359 E. faecium XVI 100 (KF245564.1) 100.00 (DQ411813) n.n. KM257698 1359 WVS360 E. faecium XVI 99 (KC478513.1) 99.93 (DQ411813) n.n. KM257699 1420 WVS343 E. faecium XVI 99 (KF245564.1) 99.85 (DQ411813) n.n. KM257700 1362 WVS346 E. faecium XVI 99 (KF245564.1) 99.93 (DQ411813) n.n. KM257701 1360 WVS388 E. faecium XVI 99 (KF149258.1) 98.64 (DQ411813) n.n. KM257703 1246 WVS393 E. faecium XVI 100 (KF245564.1) 100.00 (DQ411813) n.n. KM257704 1262 WVS398 E. faecium XVI 99 (KF245564.1) 99.93 (DQ411813) n.n. KM257706 1441 WVS418 E. faecium XVI 99 (KJ726575.1) 100.00 (DQ411813) n.n. KM257708 1212 WVS422 E. faecium XVI 99 (KF245564.1) 100.00 (DQ411813) n.n. KM257709 1260 WVS439 E. faecium XVI 99 (KC478513.1) 99.86 (DQ411813) n.n. KM257714 1423 WVS443 E. faecium XVI 98 (KF149258.1) 98.02 (DQ411813) n.n. KM257716 1311 WVS466 E. faecium XVI 99 (JX290550.1) 98.28 (DQ411813) n.n. KM257720 1222 WVS395 E. faecium/E. lactis XVI 99 (KF245564.1) 100.00 (GU983697) E. faecium KM257705 1362 WVS416 Lb. brevis IV 99 (KC336482.1) 99.55 ( KI271266) n.n. KM257707 1327 WVS430 Lb. brevis III 99 (KJ702494.1) 100 (EF120367) n.n. KM257711 1297 WVS432 Lb. brevis III 99 (KF923752.1) 99.51 (CP000416 ) n.n. KM257712 1238 WVS303 Lb. delbrueckii I 99 (KF029498.1) 99.49 (JQ801728) n.n. KM257688 1475 WVS354 Lb. fermentum II 99 (HG798472.1) 97.33 (AJ575812) n.n. KM257695 1202 WVS270 Lb. paracasei V 100 (NR_121787.1) 100.00 (D16550) n.n. KM257681 1342 WVS279 Lb. plantarum VI 99 (EU626009.1) 99.17 (ACGZ01000098) n.n. KM257684 1325 WVS223 Lb. reuteri VII 99 (EU547308.1) 98.98 (AP007281) n.n. KM257675 1384 WVS243 Lb. reuteri VII 99 (HM218331.1) 99.41 (AP007281) n.n. KM257679 1359 WVS39 Lc. lactis VIII 98 (JQ411245.1) 96.88 (AB100803) n.n. KM257644 1480 WVS49 Lc. lactis VIII 99 (KF149646.1) 99.34 (AB100803) n.n. KM257648 1358 WVS320 Lc. lactis VIII 100 (KJ186940.1) 100.00 (AB100803) n.n. KM257691 1300 WVS119 Lc. lactis VIII 99 (KJ702498.1) 99.34 (AB100803) n.n. KM257659 1373 WVS187 Lc. lactis VIII 99 (KJ690910.1) 99.92 ( EU770697) n.n. KM257670 1208 WVS141 Ln. mesenteroides X 99 (KF149523.1) 100.00 (AB023246) n.n. KM257661 1372 WVS147 Ln. mesenteroides IX 99 (KC865284.1) 99.64 (CP000414) n.n. KM257664 1408 WVS159 Ln. mesenteroides IX 99 (NR_074957.1) 99.16 (HM443957) n.n. KM257666 1473 WVS161 Ln. mesenteroides IX 98 (HF562947.1) 99.62 (AB023246) n.n. KM257667 1326 59

CHAPTER III

WVS181 Ln. pseudomesenteroides XII 99 (KF263162.1) 99.92 (AEOQ01000906) n.n. KM257668 1256 WVS221 Ln. pseudomesenteroides XI 99 (KJ026683.1) 99.55 (AEOQ01000906) n.n. KM257673 1342 WVS183 P. acidilactici XIX 98 (KF057958.1) 99.27 (GL397069) n.n. KM257669 1381 WVS280 P. acidilactici XVII 99 (KF057958.1) 99.79 (GL397069) n.n. KM257685 1416 WVS383 P. acidilactici XVII 99 (KF057958.1) 100.00 (GL397069) n.n. KM257702 1433 WVS434 P. acidilactici XVIII 98 (JX885673.1) 99.03 (GL397069) n.n. KM257713 1349 WVS451 P. acidilactici XVII 99 (KF057958.1) 99.92 (GL397069) n.n. KM257717 1212 WVS458 P. acidilactici XVIII 99 (KF057958.1) 100.00 (GL397069) n.n. KM257718 1410 WVS325 P. lolii XX 99 (JX311435.1) 99.86 (BANK01000051) n.n. KM257692 1435 WVS222 P. lolii XXI 99 (JX311435.1) 99.86 (BANK01000051) n.n. KM257674 1430 WVS105 S. lutetiensis XXII 99 (EU163444.1) 99.76 (DQ232532) n.n. KM257655 1400 WVS18 S. thermophilus XXIV 99 (KJ026586.1) 100.00 (AY188354) n.n. KM257640 1399 WVS307 S. thermophilus XXIV 99 (KJ833590.1) 98.85 (AY188354) n.n. KM257689 1312 WVS50 S. thermophilus XXIII 99 (HM059000.1) 99.86 (AY188354) n.n. KM257649 1454 WVS271 S. thermophilus XXIII 99 (KF286609.1) 98.84 (AY188354) n.n. KM257682 1473 Abbreviations: E., Enterococcus; Lb., Lactobacillus; Lc., Lactococcus; Ln., Leuconostoc; P., Pediococcus; S., Streptococcus; Acc. No., accession number; n.n., not necessary.

3.4. Species distribution

A total of 16 species of LAB was identified at dominating levels in the biofilms associated with the wooden vats used for the production of Vastedda della valle del Belìce and Caciocavallo Palermitano cheeses (Table 5). The species isolated from all wooden vat surfaces was E. faecium found at cell densities among 103 – 106 CFU/cm2. Lb. brevis, Lb. delbrueckii, Ln. mesenteroides, Lb. reuteri, and Streptococcus lutetiensis were found associated only to a single wooden vat, with the last 2 species detected at 106 CFU/cm2. Although Lb. reuteri was found at the same level of the other species (E. faecium, Ln. pseudomesenteroides and P. lolii) isolated from the highest dilution plates of the gauzes used to collect the biofilms from WV7, S. lutetiensis was 2 Log cycles higher than E. durans and E. faecium present in WV4. S. thermophilus was isolated only from the vats used for Caciocavallo Palermitano cheese making, always at 105 CFU/cm2. P. acidilactici and E. faecalis were largely distributed among the LAB biofilms analysed, both in the range 103 – 105 CFU/cm2. The highest biodiversity in terms of enterococci was determined for WV1 which was the oldest vat used in cheese manufacture (28 years). In general, a very low complexity of the lactobacilli community was found in the biofilms, at least at the highest densities of detection. In particular, the vats made with douglas wood, except WV9, were characterized by the absence of Lactobacillus isolates. S. thermophilus were only detected in the vats made with chestnut wood, especially with those used since long time. All bovine milk cheese productions were performed in chestnut wooden vats. The vats (WV1, WV2, WV6, WV7, WV8 and WV11) for which a co-dominance of several species (4 – 6) was found were all subjected to the washing with hot deproteinized whey after cheese making, even though the vats WV1, WV8 and WV12 were washed with cold water in the period June – September.

60

CHAPTER III

Surprisingly, despite the large number of strains detected, the number of Enterococcus species displayed by the surfaces of the vats (WV3, WV4, WV9, WV10 and WV11) used to transform raw ewe’s milk was quite limited. The high number of strains identified as members of the genus Enterococcus, especially those belonging to the species E. faecium, was not surprising for the vats filled in daily with ewe’s milk. In fact, the species E. faecium, E. faecalis and E. durans comprise the enterococci most prevalent in artisanal European raw ewe’s cheeses (Prodromou et al., 2001; Todaro et al., 2011). Even though enterococci were not found at consistent levels in French wooden vats (Didienne et al., 2012), their presence was already reported for those used in Sicily (Licitra et al., 2007; Settanni et al., 2012). A recent work by Di Grigoli et al. (2015) showed that three enterococci belonging to three species (E. faecalis, E. casseliflavus and E. gallinarum), isolated from the wooden vat surface, were able to persist at different times of ripening of Caciocavallo Palermitano cheese, demonstrating the influence of the equipment during the production of traditional cheeses and highlighting the importance of this group of LAB to confer typicality. In particular, the strain found longer during ripening was an E. faecalis. The species E. faecalis, that together with E. faecium, E. durans, E. mundtii and E. casseliflavus, is commonly found in many raw materials and foods (Corsetti et al., 2007; Franciosi et al., 2009; Settanni et al., 2014).

61

Table 5. Distributiona of LAB species among wooden vats.

LAB species Wooden vats

WV1 WV2 WV3 WV4 WV5 WV6 WV7 WV8 WV9 WV10 WV11 WV12

E. durans ■ (4) ■ (4) ■ (4) E. faecium ■ (4) ■ (3) ■ (3) ■ (4) ■ (4) ■ (3) ■ (6) ■ (5) ■ (4) ■ (4) ■ (3) ■ (5) E. faecalis ■ (4) ■ (3) ■ (5) ■ (5) ■ (5) Lb. brevis ■ (4) Lb. delbrueckii ■ (4) Lb. fermentum ■ (6) Lb. paracasei ■ (5) Lb. plantarum ■ (5) Lb. reuteri ■ (6)

CHAPTER I CHAPTER Lc. lactis ■ (5) ■ (5) ■ (5) Ln. mesenteroides ■ (5) Ln. pseudomesenteroides ■ (4) ■ (6)

P. acidilactici ■ (3) ■ (5) ■ (4) ■ (3) ■ (5)

P. lolii ■ (5) ■ (6) I I S. lutetiensis ■ (6)

S. thermophilus ■ (5) ■ (5) ■ (5) a The number reported between brackets refers to the highest concentration (Log cycle) of detection. Abbreviations: E., Enterococcus; Lb., Lactobacillus; Lc., Lactococcus; Ln., Leuconostoc; P., Pediococcus; S., Streptococcus.

62

CHAPTER III

3.5. Technological traits of lactic acid bacteria

The results of the technological characterization based on acidification, autolysis and production of diacetyl were considered simultaneously. The grouping of LAB according to these three activities was achieved by hierarchical clustering analysis (Fig. 2).

120

100 Single Linkage Euclidean distances

80

60

(Dlink/Dmax)*100 40

20

0

WVS4 WVS7 WVS1

WVS58 WVS43 WVS49 WVS37 WVS42 WVS18

WVS53 WVS74 WVS39 WVS87 WVS66 WVS50 WVS31 WVS45 WVS23

WVS187 WVS279 WVS271 WVS333 WVS320 WVS356 WVS426 WVS292 WVS418 WVS275 WVS106 WVS119 WVS398 WVS354 WVS280 WVS221 WVS222 WVS430 WVS307 WVS181 WVS159 WVS432 WVS416 WVS141 WVS161 WVS143 WVS439 WVS270 WVS383 WVS451 WVS243 WVS223 WVS215 WVS231 WVS233 WVS434 WVS153 WVS335 WVS238 WVS195 WVS147 WVS442 WVS296 WVS443 WVS315 WVS393 WVS346 WVS458 WVS466 WVS264 WVS142 WVS108 WVS343 WVS395 WVS464 WVS137 WVS325 WVS388 WVS113 WVS183 WVS340 WVS360 WVS422 WVS359 WVS105 WVS303

WVS467 Strains Fig. 2. Hierarchical clustering analysis grouping the assayed strains according to their technological traits (acidification rate, autolysis kinetics and diacetyl production).

Two strains (WVS187 and WVS279) resulted different from the others, which clustered in two main groups. A small group of strains (WVS271, WVS333, WVS320, WVS58, WVS43, WVS467, WVS 356, WVS426, WVS292, WVS418, WVS49, WVS275) clustered with 48% of the relative linkage distance. The remaining strains were clustered together in a major group according to a relative linkage distance below 33%. This analysis is indicative that most of the strains tested had fairly similar activities and only some of them showed relevant characteristics.

63

CHAPTER III

PCA individuated two Factors with eigen-values higher than 0.90, indicating that the three variables analyzed could be grouped into two Factors which explained 78.55% of the total variance. The relationship between these factors and the original variables can be deduced from the projection of the latter onto the plane formed by the two factors (Fig. 3). In this way, Factor 1, representing 46.46% of the total variance, was directly related (positive correlation in Fig. 3) to autolysis and diacetyl production. On the contrary, Factor 1 was inversely related to acidification. Conversely, Factor 2 (which represented 32.09% of the total variance) was directly related to diacetyl production and barely to acidification, while it was inversely related to autolysis activity (Fig. 3). Projection of the variables on the factor-plane ( 1 x 2)

1,0 Diacetyl production

0,5

Acidification 0,0

Factor 2 : 32,09% : 2 Factor

Autolysis -0,5

-1,0

-1,0 -0,5 0,0 0,5 1,0 Factor 1 : 46,46% Fig. 3. Projection of the variables (acidification, autolysis and diacetyl production) onto the plane formed by Factor 1 and Factor 2.

The projection of the cases (LAB strains) onto the planes formed by the two Factors led to the segregation of several isolates clearly differentiated from the rest of LAB (Fig. 4).

64

CHAPTER III

Projection of the cases on the factor-plane ( 1 x 2) Cases with sum of cosine square >= 0,00 4,0

3,5

3,0 WVS333

WVS279 WVS271 2,5 WVS467 WVS320 WVS275WVS43 WVS58 WVS356 2,0 WVS426WVS292 WVS418WVS49 1,5

1,0 WVS354

WVS223WVS153 0,5 WVS434WVS233 WVS222WVS243 WVS221WVS181WVS307WVS430WVS4 WVS231 WVS416 WVS147 WVS1 WVS50 Factor 2: 32,09% 2: Factor 0,0 WVS432WVS159 WVS451WVS142WVS270WVS18 WVS42 WVS137WVS183 WVS395 WVS141 WVS335 WVS215WVS53WVS87 WVS398 WVS280WVS143WVS439WVS238WVS195 WVS7WVS393WVS466WVS343WVS39WVS464 WVS303WVS113WVS37 WVS443WVS346WVS458WVS264 WVS45WVS23 -0,5 WVS296WVS315 WVS325WVS422WVS359 WVS161WVS442WVS383WVS66WVS340WVS105 WVS187 WVS108WVS74 WVS31WVS360 WVS388 WVS106WVS119 -1,0 -1,5 -2,0

-2,5 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 Factor 1: 46,46% Fig. 4. Projection of the cases (LAB strains) onto the plane formed by Factor 1 and Factor 2.

The majority of isolates were grouped close to the intersection between both factors, slightly in the negative axis of Factor 2, while several strains (WVS43, WVS49, WVS58, WVS271, WVS275, WVS292, WVS320, WVS356, WVS418, WVS426, WVS467) were differently clustered into the upper-right quarter, positively related with both Factors. Only three cases (WVS279, WVS333, WVS354) were clearly located outside the two groups, negatively correlated with Factors 1 and positively with Factor 2. Therefore, globally, results found by PCA were in agreement with those obtained by clustering. In general, optimal SLAB are characterised by a fast and appropriate acidification and a rapid autolysis, whereas optimal NSLAB show opposite performances (Franciosi et al. 2009; Settanni et al., 2013). The group of the fastest acidifiers included some strains of Lc. lactis and S. thermophiles which were also positive for diacetyl production, thus proving that the wooden vats might act as sources of cultures useful in cheese making. Wooden vat LAB strains were particularly active in hydrolysing gelatine and BSA, since after incubation, a clear halo was detected for 74 strains in presence of both substrates (results not shown). On the contrary, E. faecalis WVS464, Lb. delbrueckii WVS303, Lb. reuteri WVS243 and P. acidilactici WVS393 did not hydrolyse nor gelatine neither BSA. Three E. faecium and two P. acidilactici strains were active only on BSA, while barely one strain per each of these two species hydrolysed only gelatine. Lortal et al. (2009) supposed that an additional hypothesis to explain the absence of foodborne pathogens in the wooden vat biofilms is due to the presence of bacteriocin

65

CHAPTER III

producers. For this reason, all LAB were also investigated for this character. The antimicrobial activity was registered for 31 strains (Table 6).

Table 6. Inhibitory characteristics of LAB isolated from wooden vatsa.

LAB culture Bacteriocin-like inhibitory activityb Indicator strainsc 19114 4202 2313 E.durans WVS23 1.2 ± 0.10 1.3 ± 0.12 - E. faecium WVS1 1.5 ± 0.15 2.0 ± 0.00 - E. faecium WVS4 1.4 ± 0.06 2.2 ± 0.06 - E. faecium WVS7 1.4 ± 0.06 1.5 ± 0.12 - E. faecium WVS31 1.7 ± 0.15 - - E. faecium WVS42 1.3 ± 0.17 1.5 ± 0.15 - E. faecium WVS45 2.6 ± 0.12 1.8 ± 0.17 - E. faecium WVS66 1.6 ± 0.10 1.4 ± 0.00 1.3 ± 0.06 E. faecium WVS74 - - 1.7 ± 0.06 E. faecium WVS106 1.6 ± 0.17 - - E. faecium WVS143 - 1.8 ± 0.06 - E. faecium WVS231 1.4 ± 0.15 1.4 ± 0.17 - E. faecium WVS264 1.6 ± 0.12 2.0 ± 0.15 - E. faecium WVS292 1.6 ± 0.17 2.4 ± 0.06 - E. faecium WVS388 1.4 ± 0.10 1.2 ± 0.12 - E. faecalis WVS464 - - 1.4 ± 0.15 Lb. brevis WVS430 - 1.8 ± 0.06 2.0 ± 0.06 Lb. brevis WVS432 - 1.4 ± 0.06 2.0 ± 0.00 Lb. delbrueckii WVS303 - 1.8 ± 0.12 - Lb. plantarum WVS279 1.7 ± 0.06 1.6 ± 0.10 2.3 ± 0.06 Lb. reuteri WVS223 1.4 ± 0.06 1.6 ± 0.00 - Lb. reuteri WVS243 1.3 ± 0.12 1.6 ± 0.06 - Lc. lactis WVS39 1.5 ± 0.00 1.4 ± 0.10 - Lc. lactis WVS49 - - 1.3 ± 0.10 Lc. lactis WVS119 - - 1.5 ± 0.12 Ln. mesenteroides WVS141 1.2± 0.12 1.1 ± 0.12 - Ln. pseudomesenteroides WVS181 - - 1.4 ± 0.00 P. acidilactici WVS280 - - 1.5 ± 0.10 P. acidilactici WVS383 1.5 ± 0.12 1.2 ± 0.15 1.6 ± 0.06 P. lolii WVS222 1.5 ± 0.00 1.4 ± 0.00 - S. lutetiensis WVS105 1.2 ± 0.15 1.8 ± 0.00 - Abbreviations: E., Enterococcus; Lb., Lactobacillus; Lc., Lactococcus; Ln., Leuconostoc; P., Pediococcus; S., Streptococcus. a The results are shown only for the strains showing inhibitory activity towards at least one indicator bacterium. b Width of the inhibition zone (millimeters). Results indicate mean ± SD of three independent experiments. c Bacterial species: Listeria monocytogenes ATCC 19114; Listeria innocua 4202; Lactobacillus sakei 2313.

The highest activity in terms of number of indicators inhibited was displayed by E. faecium WVS66, Lb. plantarum WVS279 and P. acidilactici WVS383 which inhibited all three sensitive organisms, but the highest inhibitory power in terms of width of the clear area in plate was registered for E. faecium WVS45 and WVS292 and Lb. plantarum WVS279. These attributes confer competitive advantages to the producing strains. Furthermore, the anti- Listeria effect found in some LAB, might contribute to the safety of the microbial biofilms during cheese production. All antibacterial compounds were proteins and for this reason indicated as bacteriocin-like inhibitory substances (BLIS) (Corsetti et al., 2008). This study, together with previous studies performed on other food chains (Francesca et al., 2013)

66

CHAPTER III

demonstrated that high percentages of LAB present also onto the surfaces of the equipment used for food production are found to be BLIS producers.

4. CONCLUSIONS

Species and strain composition of LAB associated with the wooden vats used to produce Caciocavallo Palermitano and Vastedda della valle del Belìce cheeses are not particularly influenced by the origin of milk. However, since both cheeses are obtained applying the technology of stretched cheeses, future investigations will include the wooden vats used for the production of other traditional cheeses processed with different technologies. A high percentage of the strains isolated from the biofilms analysed were enterococci and, as being a member of this group their use in food applications needs to be validated by the absence of risks for consumer (Settanni et al., 2014). Further studies are being prepared to better investigate on the safety of these enterococci in terms of antibiotic resistance and virulence as well as cellular toxicity. The technological characterization of the LAB found at high numbers onto the surfaces of the wooden vats showed interesting dairy properties. Multivariate analysis proved to be useful tools in the management of the large amount of data generated in this study in relation to the diverse LAB biochemical activities. Several strains showed the capacity of inhibiting undesired bacteria. This observation strengthens the safety of use of wooden vats for traditional cheese production. However, further studies are necessary to determine the influence of these microorganisms during cheese production.

67

CHAPTER III

REFERENCES

AFNOR. (2004) Certificate n. BIO 12/11-03/04. VIDAS Listeria monocytogenes II (VIDAS LMO2) – Enrichment stage at 37°C. Agence Française de Normalisation, Saint-Denis La Plaine, France AFNOR. (2007) Certificate n. BIO 12/22-05/07. VIDAS Immuno-concentration Salmonella II (VIDAS ICS2)+SLM. Agence Française de Normalisation, Saint-Denis La Plaine, France Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W. (1997) Grapped BLAST and PSI- BLAST: a new generation of protein database search programs. Nucleic Acids Research 25, 3389–3402. Beresford, T.P., Fitzsimons, N.A., Brennan, N.L., Cogan, T.M. (2001) Recent advances in cheese microbiology. International Dairy Journal 11, 259–274 Chun, J., Lee, J-H., Jung, Y., Kim, M., Kim, S., Kim, B.K., Lim, Y.W. (2007) EzTaxon: a web-based tool for the identification of prokaryotes based on 16S ribosomal RNA gene sequences. International Journal of Systematic and Evolutionary Microbiology 57, 2259–2261 Corsetti, A., Settanni, L., Chaves López, C., Felis, G. E., Mastrangelo, M., Suzzi, G. (2007) A taxonomic survey of lactic acid bacteria isolated from wheat (Triticum durum) kernels and non-conventional flours. Systematic and Applied Microbiology 30, 561–571 Corsetti, A., Settanni, L., Braga, T.M., De Fatima Silva Lopes, M., Suzzi, G. (2008) An investigation on the bacteriocinogenic potential of lactic acid bacteria associated with wheat (Triticum durum) kernels and non- conventional flours. LWT-Food Science and Technology 41, 1173–1182 Cutini, F. (2014) Food: Divina (LN), the FDA declares war on our chesees. AgenParl (Agenzia Parlamentare per l’informazione politica ed economica). Available from http://www.agenparl.com/?author=5&paged=26 Accessed 7.31.14 Didienne, R., Defargues, C., Callon, C., Meylheuc, T., Hulin, S., Montel, M.C. (2012) Characteristics of microbial biofilm on wooden vats (‘gerles’) in PDO Salers cheese. International Journal of Food Microbiology 156, 91–101 Di Grigoli, A., Francesca, N., Gaglio, R., Guarrasi, V., Moschetti, M., Scatassa, M.L., Settanni, L., Bonanno, A. (2015) The influence of the wooden equipment employed for cheese manufacture on the characteristics of a traditional stretched cheese during ripening. Food Microbiology 46, 81–91 Francesca, N., Sannino, C., Moschetti, G., Settanni, L. (2013) Microbial characterisation of fermented meat productions from the Sicilian breed “Suino Nero dei Nebrodi”. Annals of Microbiology 63, 53–62 Franciosi, E., Settanni, L., Carlin, S., Cavazza, A., Poznanski, E. (2008) A factory-scale application of secondary adjunct cultures selected from lactic acid bacteria during “Puzzone di Moena” cheese ripening. Journal of Dairy Science 91, 2981–2991 Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Biodiversity and technological potential of wild lactic acid bacteria from raw cows’ milk. International Dairy Journal 19, 3–11 Franciosi, E., Settanni, L., Cologna, N., Cavazza, A., Poznanski, E. (2011) Microbial analysis of raw cows' milk used for cheese-making: influence of storage treatments on microbial composition and other technological traits. World Journal of Microbiology and Biotechnology 27, 171–180 Foulquié Moreno, M.R., Sarantinopoulos, P., Tsakalidou, E., De Vuyst, L. (2006) The role and application of enterococci in food and health. International Journal of Food Microbiology 106, 1–24 Gaglio, R., Francesca, N., Di Gerlando, R., Cruciata, M., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. (2014) Identification, typing, and investigation of the dairy characteristics of lactic acid bacteria isolated from 'Vastedda della valle del Belìce' cheese. Dairy Science & Technology 94, 157–180 ISO. (1999) ISO 6888-2:1999/Amd 1:2003. Microbiology of food and animal feeding stuffs - Horizontal method for the enumeration of coagulase-positive staphylococci (Staphylococcus aureus and other species) - Part 2: Technique using rabbit plasma fibrinogen agar medium. International Organization Standardization, Geneva, Switzerland ISO. (2001) ISO 16649-2. Microbiology of food and animal feeding stuffs - Horizontal method for the enumeration of beta-glucuronidase-positive Escherichia coli - Part 2: Colony-count technique at 44 degrees C using 5-bromo-4-chloro-3-indolyl beta-D-glucuronide. International Organization Standardization, Geneva, Switzerland ISO. (2006) ISO 4832. Microbiology of food and animal feeding stuffs - Horizontal method for the enumeration of coliforms - Colony-count technique. International Organization Standardization, Geneva, Switzerland 68

CHAPTER III

Jackson, C.R., Fedorka-Cray, P.J., Barrett, J.B. (2004) Use of a genus- and species-specific multiplex PCR for identification of enterococci. Journal of Clinical Microbiology 42, 3558–3565 King, N. (1948) Modification of Voges–Proskauer test for rapid colorimetric determination of acetyl methyl carbimol plus diacetyl in butter. Dairy Industries 13, 860–866 Licitra, G., Ogier, J.C., Parayre, S., Pediliggieri, C., Carnemolla, T.M., Falentin, H., Madec, M.N., Carpino, S., Lortal, S. (2007) Variability of the bacterial biofilms of the “tina” wood vat used in the Ragusano cheese making process. Applied and Environmental Microbiology 73, 6980–6987 Lortal, S., Di Blasi, A., Madec, M.N., Pediliggieri, C., Tuminello, L., Tanguy, G., Fauquant, J., Lecuona, Y., Campo, P., Carpino, S., Licitra, G. (2009) Tina wooden vat biofilm. A safe and highly efficient lactic acid bacteria delivering system in PDO Ragusano cheese making. International Journal of Food Microbiology 132, 1–8 Mariani, C., Oulahal, N., Chamba, J.F., Dubois-Brissonnet, F., Notz, E., Briandet, R. (2011) Inhibition of Listeria monocytogenes by resident biofilms present on wooden shelves used for cheese ripening. Food Control 22, 1357–1362 Mora, D., Musacchio, F., Fortina, M. G., Senini, L., Manachini, P. L., (2003) Autolytic activity and pediocin-induced lysis in Pediococcus acidilactici and Pediococcus pentosaceus strains. Journal of Applied Microbiology 94, 561–570 Morea, M., Matarante, A., Di Cagno, R., Baruzzi, F., Minervini, F. (2007) Contribution of autochthonous non- starter lactobacilli to proteolysis in Caciocavallo Pugliese cheese. International Dairy Journal 17, 525–534 Piraino, P., Zotta, T., Ricciardi, A., McSweeney, P.L.H., Parente, E. (2008) Acid production, proteolysis, autolytic and inhibitory properties of lactic acid bacteria isolated from pasta filata cheeses: a multivariate screening study. International Dairy Journal 18, 81–92 Prodromou, K., Thasitou, P., Haritonidou, E., Tzanetakis, N., Litopoulou-Tzanetaki, E. (2001). Microbiology of “Orinotyri”, a ewe’s milk cheese from the Greek mountains. Food Microbiology 18, 319–328 Salvadori del Prato, O. (1998). Trattato di Tecnologia Casearia. Edagricole , Bologna, Italy Schillinger, U., Lücke, F. K. (1989) Antibacterial activity of Lactobacillus sake isolated from meat. Applied and Environmental Microbiology 55, 1901–1906 Settanni, L., Moschetti, G. (2010) Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiology 27, 691–697 Settanni, L., Di Grigoli, A., Tornambé, G., Bellina, V., Francesca, N., Moschetti, G., Bonanno, A. (2012) Persistence of wild Streptococcus thermophilus strains on wooden vat and during the manufacture of a Caciocavallo type cheese. International Journal of Food Microbiology 155, 73–81 Settanni, L., Gaglio, R., Guarcello, R., Francesca, N., Carpino, S., Sannino, C., Todaro, M. (2013) Selected lactic acid bacteria as a hurdle to the microbial spoilage of cheese: application on a traditional raw ewes’ milk cheese. International Dairy Journal 32, 126–132 Settanni, L., Guarcello, R., Gaglio, R., Francesca, N., Aleo, A., Felis, G.E., Moschetti, G. (2014) Production, stability, gene sequencing and in situ anti-Listeria activity of mundticin KS expressed by three Enterococcus mundtii strains. Food Control 35, 311–322 Todaro, M., Francesca, N., Reale, S., Moschetti, G., Vitale, F., Settanni, L. (2011) Effect of different salting technologies on the chemical and microbiological characteristics of PDO Pecorino Siciliano cheese. European Food Research and Technology 233, 931–940 Vermelho, A.B., Meirelles, M.N.L., Lopes, A., Petinate, S.D.G., Chaia, A.A., Branquinha, M.H. (1996) Detection of extracellular proteases from microorganisms on agar plates. Memorias do Istituto Oswaldo Cruz 91, 755–760 Wouters, J.T.M., Ayad, E.H., Hugenholtz, J., Smit, G. (2002) Microbes from raw milk for fermented dairy products. International Dairy Journal 12, 91–109

69

CHAPTER IV

The influence of the wooden equipment employed for cheese manufacture on the characteristics of a traditional stretched

cheese during ripening

The present chapter has been published in

Food Microbiology

46, 81-91

2015

70

CHAPTER IV

ABSTRACT

The influence of the wooden equipment used for the traditional cheese manufacturing from raw milk was evaluated on the variations of chemico-physical characteristics and microbial populations during the ripening of Caciocavallo Palermitano cheese. Milk from two farms (A, extensive; B, intensive) was processed in traditional and standard conditions. Chemical and physical traits of cheeses were affected by the farming system and the cheese making technology, and changed during ripening. Content in NaCl and N soluble was lower, and paste consistency higher in cheese from the extensive farm and traditional technology, whereas ripening increased the N soluble and the paste yellow and consistency. The ripening time decreased the number of all lactic acid bacteria (LAB) groups, except enterococci detected at approximately constant levels (104 and 105 CFU/g for standard and traditional cheeses, respectively), till 120 d of ripening. In all productions, at each ripening time, the levels detected for enterococci were lower than those for the other LAB groups. The canonical discriminant analysis of chemical, physical and microbiological data was able to separate cheeses from different productions and ripening time. The dominant LAB were isolated, phenotypically characterised and grouped, genetically differentiated at strain level and identified. Ten species of LAB were found and the strains detected at the highest levels were Pediococcus acidilactici and Lactobacillus casei. Ten strains, mainly belonging to Lactobacillus rhamnosus and Lactobacillus fermentum showed an antibacterial activity. The comparison of the polymorphic profiles of the LAB strains isolated from the wooden vat with those of the strains collected during maturation, showed the persistence of three enterococci in traditional cheeses, with E. faecalis found at dominant levels over the Enterococcus population till 120 d; the absence of these strains in the standard productions evidenced the contribution of vat LAB during Caciocavallo Palermitano cheese ripening.

Key words: Enterococcus, lactic acid bacteria biodiversity, “pasta filata” cheese, ripening, wooden dairy plant equipment.

71

CHAPTER IV

1. INTRODUCTION

Many traditional cheeses are manufactured in small size farms with raw milk from animals of indigenous breeds that are fed mainly on natural pasture. This is the case of Caciocavallo Palermitano cheese, a “pasta-filata” product, manufactured within the Palermo province (Sicily, Italy) mainly with milk from the autochthonous breed cows (Cinisara and Modicana) processed raw. The traditional cheese making is carried out employing the wooden dairy equipment without the addition of lactic acid bacteria (LAB) (Bonanno et al., 2004). Recently, some variations in the traditional production system of Caciocavallo Palermitano cheese have been registered for some dairy factories, especially those characterised by high volumes of milk. Since cheese cannot be made without the action of certain species of LAB (Parente and Cogan, 2004), any innovation based on the thermal treatment of milk may compromise the characteristic features that contribute to the definition of cheese typicality. In general, cheese production comprises two different microbiological steps in which different LAB are involved: starter LAB (SLAB) during manufacturing, and non starter LAB (NSLAB) during ripening (Settanni and Moschetti, 2010). The microbiota of a typical cheese is defined for the final characteristics of the resulting product and it often reflects the environment and the system of production (Micari et al., 2007). The typical flavour of a given cheese depends also on the microbial activity during ripening, especially due to the enzymatic degradation of milk lactose, fat and protein, producing volatile organic compounds with aromatic properties (Urbach, 1997). Several cheeses are niche products that are linked to the production area not only for the traditions that are handed down over time, but also and, above all, for the presence of microbial species and strains colonizing the environment of transformation and the equipment employed during processing (Settanni and Moschetti, 2014). This phenomenon is particularly evident when the equipments used for transformation are made with material (e.g. wood) that can help the formation of microbial biofilms (Lortal et al., 2009; Didienne et al., 2012) strongly contributing to the typicality definition. Currently, some Caciocavallo Palermitano producers follow a standard scheme: milk is pasteurised, the equipment is in stainless steel and commercial SLAB are added into the milk before coagulation. This actual trend involves the simultaneous presence of “traditional” and “standard” products, both designed as “Caciocavallo Palermitano” on the market, that indeed differ substantially (Settanni et al., 2010). A study conducted on the microbiological characterization of both traditional and standard technologies applied to obtain this cheese

72

CHAPTER IV

revealed that, following the traditional protocol, a clear dominance of the Streptococcus thermophilus strains of wooden vat origin emerged during the entire cheese manufacture till stretched curd moulding, highlighting the influence of the traditional equipment during the first stages of the production process (Settanni et al., 2012). However, other species of vat origin were identified as members of the NSLAB population, some of which are reported to be linked to the cheese typicality. With this in mind, the traditional cheeses have been followed from the manufacturing stage to ripening, then sampled at different times. The specific objectives of this work were to: enumerate and isolate LAB from traditional and standard productions after 30, 60 and 120 days of ripening; characterize, differentiate and identify all dominant LAB; compare the polymorphic profiles of the strains isolated during ripening with those previously isolated from the wooden vat before milk was processed into cheese; evaluate the chemical and physical changes of the cheeses during ripening.

73

CHAPTER IV

2. MATERIALS AND METHODS

2.1. Cheese production and sample collection

The experimental cheeses object of this study were previously produced (Settanni et al., 2012; Bonanno et al., 2013) using the bulk milk from two farms (A and B) located within the Palermo province (Sicily, Italy). In the farm A, cows of autochthonous breed (Cinisara) were fed mainly at natural pasture, whereas in the farm B specialized dairy cows of Brown breed received a diet based on hay and concentrate. Four productions, two traditional (TA and TB) carried out following the local cheese making protocol with the wooden equipment, and two standard (SA and SB) carried out in a stainless steel vat added with a commercial SLAB culture (LYOBAC-D T, Alce International s.r.l., Quistello, Italy), were performed in a dairy factory close to the farms. The main wooden equipment for Caciocavallo Palermitano cheese production (Tornambé et al., 2009) consisted of a vat (tina) for milk coagulation, a stick (rotula) for curd breaking, a bowl (cisca) for curd pressing, a cane plan (cannara) for residual whey loss by pressing, a horizontal stick (appizzatuma) for curd acidification, a truncated conical vat (piddiaturi) for curd stretching through a stick (maciliatuma) and a form (tavuleri) for moulding. Each cheese production was performed in triplicate in three consecutive weeks for a total of three cheese making trials. Cheeses from all productions were sampled at 30, 60 and 120 days during ripening from the same form by covering the cut surface with paraffin. On the whole, 36 samples [three replicates for each of the four productions (TA, TB, SA and SB) at the three aging times (30, 60 and 120 d)] were collected. For microbiological analysis, cheese samples were transferred into sterile Stomacher bags, kept into a portable cooler during transport and, once in laboratory, immediately analysed. Successively, the samples were stored frozen (-20°C) until other analysis.

2.2. Chemico-physical analyses

Each cheese sample was analysed for dry matter (DM), fat, protein (N × 6.38), and ash content according to International Dairy Federation (IDF) standards [4A:1982 (IDF,1982), 5B:1986 (IDF, 1986), 25:1964 (IDF, 1964a), and 27:1964 (IDF, 1964b), respectively]. Soluble nitrogen (N) was determined on an aqueous filtrate using the Kjeldahl method (MAF, 1986), and NaCl according to the IDF procedure (17A:1972; IDF, 1972).

74

CHAPTER IV

Color was measured by Minolta Chroma Meter (CR-300; Minolta, Osaka, Japan) using illuminant C, and expressed as lightness (L*), redness (a*), and yellowness (b*), according to the (CIE) L*a*b* system. The maximum resistance to compression (compressive stress, N mm-2) was measured with an Instron 5564 tester (Instron, Trezzano sul Naviglio, Milano, Italy).

2.3. Microbiological analyses

Twenty-five grams of each cheese sample were suspended in 225 mL sodium citrate (2% w/v) solution and homogenised for 2 min at high speed with a stomacher (BagMixer® 400, Interscience, Saint Nom, France). Further serial decimal dilutions were performed in Ringer’s solution (Sigma-Aldrich, Milan, Italy). Total mesophilic counts (TMC), total psychrotrophic counts (TPC), coliforms, enterococci, pseudomonads, mesophilic and thermophilic rod LAB, mesophilic and thermophilic cocci LAB, and yeasts were cultivated and incubated as reported by Settanni et al. (2012). Microbiological counts were performed in duplicate.

2.4. Isolation of lactic acid bacteria and phenotypic grouping

After growth, at least four colonies for each different morphology of presumptive LAB, including enterococci, were picked up from count plates and transferred to the corresponding broth media. The isolates from kanamycin aesculin azide (KAA) were cultivated in M17 broth, while the cultures from whey-based agar medium (WBAM) were inoculated into de Man-Rogosa-Sharpe (MRS) broth medium. The isolates were purified by successive sub- culturing, checked microscopically for purity and cell morphology and those Gram-positive (Gregersen KOH method) and catalase negative [determined by transferring fresh colonies from a Petri dish to a glass slide and adding 5% (w/v) H2O2] were stored in glycerol at −80°C. Phenotypic characterization was carried out as reported by Gaglio et al. (2014) based on growth at 15 and 45°C, resistance at 60°C for 30 min, NH3 production from arginine, aesculine hydrolysis, acid production from arabinose, ribose, xylose, fructose, galactose, lactose, sucrose and glycerol, and CO2 production from glucose. For coccus isolates, the sub- grouping also included the evaluation of growth at pH 9.2 and in presence of NaCl 6.5% (w/v) since, unlike other dairy cocci, enterococci can grow in both conditions.

75

CHAPTER IV

2.5. Genotypic differentiation and identification of lactic acid bacteria

DNA from LAB cultures was extracted by cell lysis using the Instagene Matrix kit (Bio- Rad, Hercules, CA) as described by the manufacturer. Crude cell extracts were then used as templates for PCR. Strain differentiation was performed by random amplification of polymorphic DNA-PCR (RAPD-PCR) following the scheme reported by Settanni et al. (2012) by means of T1 Thermocycler (Biometra, Göttingen, Germany) to generate amplicons and the pattern analysis software package Gelcompare II Version 6.5 (Applied Maths, Sin-Martens-Latem, Belgium) to analyze their profiles. Genotypic identification of the LAB characterised by different RAPD-PCR patterns was carried out by 16S rRNA gene sequencing (Weisburg et al., 1991). DNA fragments of about 1600 bp were purified by the QIAquick purification kit (Quiagen S.p.a., Milan, Italy) and sequenced at PRIMM (Milan, Italy). The sequences were compared by a BLAST search in GenBank/EMBL/DDBJ database. Furthermore, the multiplex PCR assay based on sodA gene reported by Jackson et al. (2004) was applied to confirm species identity of enterococci.

2.6. Antibacterial substances produced by lactic acid bacteria

The antibacterial activity of each LAB was evaluated against three strains (Lactobacillus sakei LMG2313, Listeria innocua 4202, and Listeria monocytogenes ATCC 19114) highly sensitive to bacteriocins (Hartnett et al., 2002; Corsetti et al., 2008). The inhibitory activities were first tested through the agar-spot deferred method, and the strains displaying antimicrobial properties were further subjected to the well diffusion assay (WDA) as reported by Corsetti et al. (2008). All tests were carried out in triplicate. The proteinaceous nature of the active compounds was tested against proteolytic enzymes as described by Settanni et al. (2005). All enzymes were purchased from Sigma-Aldrich (St. Louis, MO).

2.7. Statistical analysis

The GLM and CANDISC procedures of the SAS software package version 9.2 (SAS, 2010) were used for the statistical analysis. Chemico-physical and microbiological data were analysed by GLM procedure including the effects of farm (F = A, B), cheese technology (TC = T, traditional; S, standard), ripening time (R = 30, 60, 120 d), and their interaction F*TC*R. The Student “t” test was used for means comparisons at P≤0.05 significance level. A

76

CHAPTER IV

multivariate statistical approach were performed by a canonical discriminant analysis according to the CANDISC procedure, in order to ascertain the ability of chemical and physical parameters and microbiological counts in discriminating cheeses from different productions and during ripening.

3. RESULTS

3.1. Chemico-physical analyses

Yield, chemical composition, and color parameters (L*, a* and b*) of cheeses were affected by the farm (Table 1). On the whole, cheeses produced in the extensive farm A showed higher yield and protein percentage, lower fat, NaCl and soluble N contents, and a less intense yellow color, as indicated by the lower b* values, than farm B. Cheese making technology significantly influenced all chemico-physical parameters. Compared with cheeses from standard productions, those produced with the traditional technology had higher DM and, consequently, lower cheese yield. Moreover, traditional cheeses showed higher fat content and values of L*, a* and b* color indexes, and were more resistant to compression, indicating a more compact cheese paste than S cheeses; in addition, their content of NaCl and soluble N was lower. A significant trend due to ripening time was observed for most of the cheese parameters. A marked increase in soluble N and compressive stress test was detected during ripening, whereas L* and b* decreased and increased, respectively, only in traditional cheese, explaining the significant F*TC*R interactions. However, between 30 and 60 d of aging, a reduction in both DM and NaCl content was registered.

3.2. Microbial evolutions during ripening

The viable counts of the 10 microbial groups investigated in this study are reported in Table 2. Coagulase positive staphylococci and clostridia were not investigated since they were not detected in the stretched curd processed into cheese (Settanni et al., 2012). The effects of farm, cheese making conditions and ripening time affected significantly the development of total psychrotrophic microorganisms and, consequently, pseudomonads. In general, except for enterococci, the most evident effect on the growth of the several microbial groups analysed was showed by the ripening time (P<0.001). The interactions of the three effects considered resulted significant only for psychrotrophic microorganisms and coliforms.

77

CHAPTER IV

A general decreasing trend was observed for TMC, coliforms, yeasts and all LAB groups, during ripening. Enterococci within each production did not statistically (P>0.05) vary at the different times of analysis, but their levels estimated in the cheeses from traditional productions were, on average, about 1 Log cycle higher than those from the corresponding standard productions. However, the group of enterococci was counted at levels lower than those detected for the other LAB groups in all productions for each collection time. The highest levels were observed for mesophilic rod LAB at 30 d of ripening, while the lowest levels were registered for thermophilic coccus LAB at 120 d.

78

Table 1. Chemico-physical characteristics of cheese samples collected through Caciocavallo Palermitano cheese ripening (means±SD).

Production TA TB SA SB Statistical significance f

30 d 60 d 120 d 30 d 60 d 120 d 30 d 60 d 120 d 30 d 60 d 120 d Farm Technology Ripening F*TC*R (F) (TC) (R)

Cheese yield, % 8.79±0.55 8.63±0.51 8.38±0.50 7.78±0.37 7.52±0.33 7.26±0.35 9.33±0.65 9.15±0.61 8.85±0.58 8.07±0.14 7.85±0.12 7.56±0.13 *** * * NS

DM, % 61.72±1.00 60.58±1.10 62.68±0.58 65.15±0.68 64.64±1.60 68.91±1.01 59.37±2.22 56.72±1.98 60.22±2.10 61.39±2.76 57.94±0.75 62.21±1.96 *** *** *** NS

Protein, % DM 48.62±0.39 49.00±0.33 49.34±0.41 46.40±1.87 46.77±2.71 47.25±1.60 48.76±1.82 49.47±2.12 48.79±1.85 48.53±1.63 49.00±1.33 49.02±1.37 * + NS NS

Fat, % DM 39.14±0.75 39.94±0.51 39.90±0.51 41.11±2.65 41.77±2.63 41.29±1.94 37.39±2.18 38.17±1.92 37.68±2.07 39.04±1.40 38.60±1.36 38.82±1.65 * *** NS NS

Ash, % DM 8.32±0.21 7.24±0.33 7.72±0.65 9.20±0.66 7.93±0.49 8.62±1.00 10.03±1.35 8.48±1.00 10.08±1.96 9.78±0.48 8.61±0.51 9.21±0.24 NS *** ** NS

NaCl, g/100 g 2.39±0.14 1.62±0.21 2.01±0.35 3.48±0.20 2.55±0.05 3.28±0.61 3.58±0.75 2.40±0.77 3.41±1.11 3.64±0.55 2.69±0.34 3.33±0.25 ** ** *** NS

Soluble N, % DM 0.58±0.24 0.87±0.15 1.01±0.03 0.67±0.34 0.97±0.13 1.08±0.12 1.05±0.36 1.18±0.23 1.62±0.15 1.21±0.32 1.70±0.21 1.74±0.41 * *** *** NS

a

SN/TN 7.67±3.22 11.37±1.91 13.11±0.47 9.14±4.25 13.17±1.00 14.60±1.07 13.81±4.78 15.31±3.64 21.20±2.62 15.99±4.63 22.24±3.10 22.76±5.93 * *** *** NS CHAPTER

L* b 83.60±0.66abc 83.10±0.66abcd 80.28±0.86def 85.57±1.7a 84.67±0.7ab 80.90±1.12cdef 81.76±3.19bcde 81.39±2.56bcde 78.07±2.44f 79.02±1.73ef 77.94±1.91f 72.28±3.30g + *** *** * a* c -4.10±0.10 -4.30±0.27 -5.04±0.09 -4.27±0.23 -4.06±0.03 -4.13±0.12 4.54±0.51 -4.47±0.05 -5.32±0.23 -5.29±0.39 -5.02±0.32 -6.04±0.49 ** *** *** + b* d 26.79±0.55ef 28.09±1.30de 29.04±1.11cd 29.81±0.59bc 30.81±1.12ab 32.16±0.86a 25.27±0.92f 25.32±1.24f 25.19±1.16f 29.53±0.17bcd 29.50±0.59bcd 28.25±1.46cde *** *** NS *

IV CS e, N/mm2 0.25±0.03 0.24±0.05 0.41±0.13 0.27±0.12 0.31±0.05 0.36±0.19 0.15±0.05 0.16±0.02 0.23±0.02 0.13±0.03 0.11±0.01 0.23±0.18 NS *** * NS

a SN/TN = soluble N/total N; b L* = lightness; c a* = redness; d b* = yellowness. e CS= compressive stress. f P value: ***, P≤0.001; **, P≤0.01; *, P≤0.05; +, P≤0.10; NS, not significant. Means within a row with different superscripts (a, b, c, d, e, f, g) differ (P≤0.05).

79

Table 2. Microbial load (Log CFU/g) of cheese samples collected through Caciocavallo Palermitano cheese ripening (means±SD).

Production TA TB SA SB Statistical significance b

30 d 60 d 120 d 30 d 60 d 120 d 30 d 60 d 120 d 30 d 60 d 120 d Farm Technology Ripening F*TC*R (F) (TC) (R)

PCA-SkM 7°C 3.20±0.26ab 2.67±0.15c 2.13±0.11d 3.47±0.30a 3.20±0.10ab 2.97±0.21bc 2.20±0.40d 1.67±0.35e 1.77±0.30de 3.07±0.40abc 3.10±0.20abc 3.17±0.15ab *** *** *** **

PCA-SkM 30°C 7.73±0.40 7.13±0.30 6.40±0.56 7.77±0.25 7.43±0.21 6.57±0.40 7.63±0.40 7.53±0.50 6.67±0.68 8.20±0.36 7.17±0.40 7.20±0.40 NS NS *** NS

VRBA 4.10±0.40ab 3.80±0.26abc 2.87±0.35de 3.93±0.51ab 3.13±0.50cd 2.67±0.42de 3.20±0.26cd 2.67±0.25de 2.40±0.40e 4.33±0.61a 3.60±0.10bc 2.60±0.60de NS NS *** *

KAA 5.40±0.36 5.63±0.25 5.57±0.38 5.63±0.23 5.73±0.50 5.90±0.20 4.43±0.45 4.83±0.35 4.33±0.35 4.47±0.50 4.93±0.45 4.83±0.06 NS *** NS NS

PAB 2.63±0.15 2.30±0.26 2.33±0.21 3.57±0.21 3.53±0.31 2.90±0.20 2.27±0.30 2.30±0.30 1.90±0.10 3.00±0.50 2.83±0.35 2.87±0.15 *** *** * NS

MRS 8.07±0.30 7.43±0.40 6.80±0.35 8.20±0.36 7.63±0.32 6.47±0.55 8.10±0.66 7.00±0.10 6.07±0.29 7.93±0.31 7.23±0.25 6.97±0.45 NS NS *** NS

M17 30°C 7.17±0.21 6.93±0.11 6.83±0.11 7.13±0.23 6.53±0.32 5.87±0.42 7.67±0.23 6.97±0.65 6.30±0.50 7.77±0.70 7.23±0.75 6.77±0.45 NS * *** NS

WBAM 6.97±0.06 6.53±0.45 5.80±0.30 7.13±0.65 6.83±0.50 6.10±0.72 7.23±0.23 6.80±0.30 6.13±0.25 7.53±0.55 6.60±0.40 6.27±0.25 NS NS *** NS

CHAPTER IV CHAPTER

M17 44°C 3.20±0.66 2.67±0.70 2.13±0.35 3.47±0.65 3.20±0.17 2.97±0.23 2.20±0.51 1.67±0.22 1.77±0.06 3.07±0.46 3.10±0.46 3.17±0.30 NS * *** NS

DRBC 7.73±0.50 7.13±0.25 6.40±0.36 7.77±0.31 7.43±0.15 6.57±0.58 7.63±0.20 7.53±0.06 6.67±0.50 8.20±0.55 7.17±0.45 7.20±0.20 * NS *** NS

Results indicate mean values ± S.D. of six plate counts (carried out in duplicate for three independent productions). a Abbreviations: PCA-SkM 7°C, plate count agar added with skimmed milk incubated at 7°C for total psychrotrophic counts; PCA-SkM 30°C, plate count agar added with skimmed milk incubated at 30°C for total mesophilic counts; VRBA, violet red bile agar for coliforms; KAA, kanamycin aesculin azide agar for enterococci; PAB, Pseudomonas agar base for pseudomonads; MRS, de Man-Rogosa-Sharpe agar for mesophilic rod LAB; M17 30°C, medium 17 agar incubated at 30°C for mesophilic coccus LAB; M17 44°C, medium 17 agar incubated at 44°C for thermophilic coccus LAB; WBAM, whey-based agar medium for thermophilic rod LAB; DRBC, dichloran rose bengal chloramphenicol agar for yeasts. b P value: ***, P≤0.001; **, P≤0.01; *, P≤0.05; NS, not significant.

Means within a row with different superscripts (a, b, c, d, e) differ (P≤0.05).

80

CHAPTER IV

3.3. Canonical discriminant analysis

The canonical discriminant analysis, performed simultaneously on chemical, physical and microbiological data, was able to distinguish clearly the Caciocavallo Palermitano cheeses manufactured according the different productions and during ripening. The plot generated by the canonical discriminant analysis (Fig. 1) showed a wider separation, due to the canonical variable 1 (y-axis), among cheeses produced with different technologies (traditional and standard). In addition, a discriminant effect of the farm, also due mainly to the canonical variable 1, emerged within both cheese technologies. Whereas the separation effect of ripening, linked to the canonical variable 2 (x-axis), was quite evident, even though weaker for traditional cheeses of the farm A.

30

25 TB120TB120 TB60

TB60 20 TB120

TB60 15 TA60 TB30 TB30 TA60 TA30 TA60 TB30 10 TA30 TA120 TA120 TA30 TA120 5

0 SB30 SB60 SB60 SB30 SB60

-5 SB30 can 1 (54.0 %) (54.0 1 can SB120 -10 SB120 SB120 -15

SA60 SA30 -20 SA60 SA30 SA30

SA60 -25

SA120 SA120 -30 SA120

-35 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30

can 2 (26.1 %) Fig. 1. Plot from canonical discriminant analysis in which Caciocavallo Palermitano cheeses from different productions are distributed in function of canonical variables 1 and 2 based on chemico-physical parameters and microbiological counts.

Table 3 shows the correlation coefficients for the parameters considered with the canonical variables. The variables 1, which contributed to separate the cheeses on the basis of technology and, to a lesser extent, of farm, explained the 47% of variance and was mainly 81

CHAPTER IV

correlated to chemical and physical parameters, especially the yellow index b* (0.76), even though the highest coefficient was recorded with the enterococci (0.81). The canonical variable 2, responsible for the separation among cheeses due to the ripening time, explained the 26% of the variance, and was mainly and positively correlated to all microbiological groups, with the exception of enterococci, with which it showed a lower and negative correlation.

Table 3. Canonical discriminant analysis: correlation coefficients for chemico-physical parameters and microbiological counts with the canonical variables 1 and 2 in the canonical discriminant analysis of Caciocavallo Palermitano cheese from different productions during ripening.

Canonical variable 1 Canonical variable 2 Variance % 46.78 26.08 Cheese yield -0.5270 -0.1802 DM 0.6769 -0.1227 Protein -0.3586 -0.1064 Fat 0.6276 -0.0212 Ash -0.5092 0.2755 NaCl -0.2954 0.3592 Soluble N -0.5313 0.0402 SN/TN a -0.5074 0.0461 L* b 0.4781 0.0126 a* c -0.3785 0.0224 b* d 0.7579 0.2528 CS e, N mm-2 0.4720 -0.4366 PCA-SkM 7°C 0.6086 0.4207 PCA-SkM 30°C -0.0610 0.5863 VRBA 0.3176 0.5753 KAA 0.8116 -0.2498 PAB 0.5656 0.4600 MRS 0.2357 0.5483 M17 30°C -0.2118 0.5854 WBAM -0.0485 0.6064 M17 44°C 0.3927 0.4565 DRBC -0.0355 0.6468 a SN/TN = soluble N/total N; b L* = lightness; c a* = redness; d b* = yellowness; e CS= compressive stress. Abbreviations: PCA-SkM 7°C, plate count agar added with skimmed milk incubated at 7°C for total psychrotrophic counts; PCA-SkM 30°C, plate count agar added with skimmed milk incubated at 30°C for total mesophilic counts; VRBA, violet red bile agar for coliforms; KAA, kanamycin aesculin azide agar for enterococci; PAB, Pseudomonas agar base for pseudomonads; MRS, de Man-Rogosa-Sharpe agar for mesophilic rod LAB; M17 30°C, medium 17 agar incubated at 30°C for mesophilic coccus LAB; M17 44°C, medium 17 agar incubated at 44°C for thermophilic coccus LAB; WBAM, whey-based agar medium for thermophilic rod LAB; DRBC, dichloran rose bengal chloramphenicol agar for yeasts.

82

CHAPTER IV

3.4. Isolation and grouping of lactic acid bacteria

On the basis of appearance, about four colonies showing similar morphological characteristics were isolated from each medium used for LAB counts, at the highest dilutions of samples, in order to detect the dominant strains. A total of 882 colonies were collected from 36 cheese samples. All cultures were subjected to microscopic inspection and separated in 683 cocci and 199 rods. After Gram characterisation and catalase test, 612 cocci and 191 rods were still considered presumptive LAB cultures, as being Gram-positive and catalase- negative. Based on several phenotypic features of the cultures and combinations of these features, the 803 LAB cultures were separated into 13 groups (Table 4), 7 for cocci and 6 for rods. The most numerous groups were group I and III, including 195 and 169 isolates, respectively. However, the unequivocal determination of the fermentative metabolism of LAB included between the groups IX to XIII needed the evaluation of their growth in presence of pentose sugars, that evidenced an obligate homofermentative metabolism for the isolates of group IX, and showed a facultative heterofermentative metabolism for the isolates of the groups X to XIII.

83

Table 4. Phenotypic grouping of LAB isolates collected through Caciocavallo Palermitano cheese ripening.

Characters Clusters

I (n=195) II (n=40) III (n=169) IV (n=81) V (n=56) VI (n=41) VII (n=30) VIII (n=48) IX (n=71) X (n=28) XI (n=12) XII (n=9) XIII (n=23)

Morphologya C C C C C C C R R R R R R

Cell dispositionb sc sc tr tr tr tr tr sc sc sc sc sc sc

Growth:

15°C + + + + + + + - - + + + +

45°C + + + + + + + + + - - - +

pH 9.2 + + + + + + + n.d. n.d. n.d. n.d. n.d. n.d.

6.5% NaCl + + + + + + + n.d. n.d. n.d. n.d. n.d. n.d.

Resistance to 60°C + + + - + + + + - + + - +

Hydrolysis of:

arginine + - - - - + + - - - + - -

aesculin + - - - + - + - - - + - + I CHAPTER

Acid production from:

arabinose + + + + + + + + - + + + +

ribose + + + + + + + + - + + + +

xylose + + + + + + + + - + + + +

fructose + + + + + + + + + + + - +

galactose + + + + + + + + + + + + + V

lactose + + + + + + + + + + + + +

sucrose + + + + + + + + + + + + +

glycerol + + + + + + + + + + + + +

CO2 from glucose ------+ - - - - -

Growth in presence of n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. - + + + + pentose carbohydrates a R, rod; C, coccus. b sc, short chain; tr, tetrads. n.d., not determined.

84

CHAPTER IV

3.5. Differentiation and identification of lactic acid bacteria

Two hundred and forty-one isolates (about 30% of the total cultures collected) were selected from each phenotypic group (PG) as being representative of the different productions and ripening times and subjected to the RAPD analysis. The genotypic differentiation distinguished 30 strains as shown by the resulting dendrogram (Fig. 2).

RAPD_FMAC_II_parte Phenotypic group

45 50 55 60 65 70 75 80 85 90 95 100 FMAC278. IV 94.1 91.5 FMAC13. VI

86.8 FMAC8. IX FMAC67. III 93.3 FMAC163. I

83.4 FMAC62. XIII 94.1

90.8 FMAC134B. I 85.8 FMAC63. VII

84.0 82.9 FMAC152. II FMAC219. I FMAC22. III 94.1 FMAC7. V 79.5 89.5 FMAC31. III 95.2 FMAC4. IV FMAC104. I 94.7 77.3 FMAC98. I FMAC43. X 94.1

91.5 FMAC16. X

86.6 FMAC61. III 71.7 FMAC17. X 93.3 FMAC37. XI FMAC9. IX 69.4 95.2 87.9 FMAC2. IX FMAC225. IX FMAC240. XIII 65.6 72.7 FMAC1. VIII 90.9 FMAC283. VIII 43.7 FMAC19. X 88.9 FMAC21. XII FMAC55. XI Fig. 2. Dendrogram obtained from RAPD-PCR patterns of LAB strains from traditional and standard Caciocavallo Palermitano cheese productions.

85

CHAPTER IV

All 30 strains were identified by 16S rRNA gene sequencing. The BLAST search evidenced a percentage of identity with sequences available in the NCBI database of at least 97%, which is considered the minimum level of similarity for 16S rRNA genes of two strains belonging to the same species (Stackebrandt and Goebel, 1994). This method allowed the identification of all strains at species level (Table 5) and all of them were confirmed to belong to the group of LAB. The species with the highest number of strains were Pediococcus acidilactici and Lactobacillus casei.

Table 5. Identification of LAB strains through Caciocavallo Palermitano cheese production.

Strains Phenotypic group Cheese samples Ripening time Species Ac. No. Bacteriocin-like inhibitory activitya Indicator strainsb 19114 4202 2313 FMAC98 I TB 30 d E. casseliflavus KF060255 - - - FMAC104 I TB 60 d E. gallinarum KF060264 - - - FMAC134B I SB 30 d E. durans KF029506 - - - FMAC163 I TA 120 d E. casseliflavus KF060266 - - - FMAC219 I SB 120 d E. faecalis KF060261 - - - FMAC152 II TA 30 d E. casseliflavus KF060267 - - - FMAC22 III TB 30 d P. acidilactici KF060269 - - - FMAC31 III TB 120 d P. acidilactici KF060262 - - - FMAC61 III SB 60 d P. pentosaceus KF060257 - - - FMAC67 III SB 120 d P. pentosaceus KF029505 - - - FMAC4 IV TA 30 d P. acidilactici KF060271 - - - FMAC278 IV SB 30 d P. acidilactici KF060253 - - - FMAC7 V TA 60 d P. acidilactici KF060254 - - - FMAC13 VI TA 120 d P. acidilactici KF060270 - - - FMAC63 VII SB 60 d P. pentosaceus KF060272 - - - FMAC1 VIII TA 30 d L. fermentum KF060268 1.8 ± 0.06 1.9 ± 0.10 2.0 ± 0.00 FMAC283 VIII SB 60 d L. fermentum KF060259 - 1.4 ± 0.17 1.2 ± 0.10 FMAC2 IX TA 30 d L. delbrueckii KF029498 1.3 ± 0.15 1.5 ± 0.10 1.8 ± 0.06 FMAC8 IX TA 60 d L. delbrueckii KF060252 - - - FMAC9 IX TA 60 d L. delbrueckii KF060256 - 1.4 ± 0.10 - FMAC225 IX TA 30 d L. delbrueckii KF060263 - - - FMAC16 X TA 120 d L. casei KF029499 - - - FMAC17 X TA 120 d L. casei KF029500 - - - FMAC19 X TB 30 d L. casei KF029501 - 1.6 ± 0.17 - FMAC43 X SA 60 d L. casei KF029503 - - - FMAC37 XI SA 30 d L. casei KF029502 1.2 ± 0.06 1.1 ± 0.12 1.2 ± 0.10 FMAC55 XI SB 30 d L. casei KF060258 1.4 ± 0.00 1.7 ± 0.06 1.8 ± 0.00 FMAC21 XII TB 30 d L. paracasei KF060265 1.4 ± 0.12 1.5 ± 0.10 1.7 ± 0.17 FMAC62 XIII SB 60 d L. rhamnosus KF029504 1.9 ± 0.00 2.1 ± 0.00 2.1 ± 0.06 FMAC240 XIII TA 120 d L. rhamnosus KF060260 1.3 ± 0.06 1.5 ± 0.12 1.2 ± 0.15 a Width of the inhibition zone (mm). Results indicate mean ± S.D. of three independent experiments. b Bacterial species: Listeria monocytogenes ATCC 19114; Listeria innocua 4202; Lactobacillus sakei 2313. Abbreviations: E., Enterococcus; P., Pediococcus; L., Lactobacillus. Symbols: plus sign positive for diacetyl production; minus sign negative for diacetyl production or, in case of antibacterial tests, no inhibition found.

86

CHAPTER IV

3.6. Comparison of RAPD profiles of vat and cheese lactic acid bacteria

In order to evaluate the persistence of the LAB of wooden vat origin during the ripening of Caciocavallo Palermitano cheese, the RAPD profiles of the strains, isolated from the vat before milk was processed into cheese, were compared to those of the LAB strains collected during cheese maturation. The direct comparison of the RAPD profiles (Fig. 3) was able to evidence the persistence of three enterococci identified as Enterococcus faecalis FMA721, Enterococcus gallinarum FMA288 and Enterococcus casseliflavus FMA108 at the time of isolation from the wooden vat and identified, in this work, from ripened cheeses as E. faecalis FMAC219, E. gallinarum FMAC104 and E. casseliflavus FMAC98, respectively. All three strains were not found, at least at dominant levels, in the standard productions and were isolated from both traditional (A and B) productions. In particular, E. casseliflavus FMAC98 was isolated from TA and TB no longer than 30 d, E. gallinarum FMAC104 was isolated at 30 d from TA and TB, but at 60 d only from TB, while E. faecalis FMAC219 persisted till 120 d in both productions.

Fig. 3. Persistence of LAB [carried out by RAPD (with primer M13) profile comparison] of wooden vat origin during the ripening of traditional Caciocavallo Palermitano cheeses. Lines M, 1-kb DNA molecular size markers (Invitrogen). Lines: 1, E. faecalis FMA721; 2, E. gallinarum FMA288; 3, E. casseliflavus FMA108.

3.7. Inhibitory activity of lactic acid bacteria

In order to evaluate the competitive advantages of the strains isolated during ripening of Caciocavallo Palermitano cheeses, the strains were tested for antibacterial compound production against three indicator strains with high sensitivity to bacteriocins. Ten strains, all lactobacilli, showed an antibacterial activity at least against one of the indicator strains, with L. rhamnosus FMAC62 and L. fermentum FMAC1 showing the highest inhibition both in terms of number of indicator strains and width of the inhibition areas (Table 5). All active compounds were inactivated by proteolytic enzymes (data not shown), proving their protein

87

CHAPTER IV

nature. Since the active substances were not characterised for amino acid and/or gene sequences, they shall be referred to as bacteriocin-like inhibitory substances (BLIS) (Corsetti et al., 2008).

4. DISCUSSION

In the last years, although the innovation in food technologies allowed the production of several safe foods with extended shelf-life, a re-discovery of traditional products is being registered (Settanni and Moschetti, 2014). Regarding the traditional food processes, especially those applied in cheese manufacture, they are not immutable in principle (Cavazza et al., 2011). Several factors may change over time: hygiene of breeding, milk production, transformation environment conditions, manufacturer’s origin and experience and government regulations. As a matter of fact, all these changes may affect the concept of food typicality which is expression of the characteristics of a territory, its history and tradition (Iannarilli, 2002). This work was carried out within a research project aimed to determine the influence of the traditional equipment on the quality of Caciocavallo Palermitano cheese during ripening. Chemical and physical traits of cheeses resulted highly influenced by the farming system and the cheese making technology, as also emerged by the canonical discriminant analysis, confirming the contribution of the environment of milk production and cheese manufacture in characterizing cheese quality. Under extensive farming system (farm A), the higher cheese yield was linked to the higher casein content that characterized the low milk produced by the rustic autochthonous Sicilian cows, whereas the lower cheese fat was probably connected to the milking system of the autochthonous cows, according to which the last and most fat rich- milk is destined for the calf (Alabiso et al., 2000). Moreover, in cheeses from farm A, the lower NaCl may depend on the higher moisture which diluted the salt, the lower N soluble content may reflect the lower milk urea (Martin et al., 1997), and the lower yellow index (b*) was probably due to a lower level of carotenoids in comparison with the cheese from intensive farm B where the cows consumed a maize-based concentrate. The strong pressure action exerted on the cheese paste to eliminate the residual whey during the traditional cheese making technology (Tornambè et al., 2009) may be responsible for the higher DM, the reduced cheese yield, and the more compact cheese paste in comparison with the standard technology. The lower NaCl content of traditional cheeses could be imputable to either lower moisture or harder paste consistency, both contributing to reduce salt absorption. The

88

CHAPTER IV

indigenous lactic microorganisms active in the traditional cheese making (Settanni et al., 2012) could be responsible for the higher color indexes, indicating a more intense cheese color (Buffa et al., 2001), and also for the less pronounced proteolytic activity, resulting in the lower levels of N soluble, in comparison with the selected LAB used in the standard productions. During ripening, as expected, both traditional and standard cheeses showed an increasing trend for soluble N, derived from microbial proteolysis, and paste consistency, whereas the changes in the color indexes (L* increase and a* decrease) interested only the traditional cheeses, due to the higher ability of the native microbiota to confer a more intense yellow color during ripening than the microorganisms of the commercial starter (Buffa et al., 2001). The reduction in both DM and NaCl content between 30 and 60 d of aging, observed in all cheeses, was opposite to that expected; this particular trend could be explained by the sampling method using the paraffin covering that, reducing the exposure of the central slice of cheese to the air, may have slowed down dehydration and NaCl penetration into the 60-d aged cheese samples (Bonanno et al., 2013). The evaluation of the microbiological characteristics of traditional and standard productions, conducted in a previous study (Settanni et al., 2012), revealed that, following the traditional protocol, a clear dominance of the Streptococcus thermophilus strains, typical thermophilic SLAB, of wooden vat origin was registered during the entire cheese manufacture till stretched curd moulding. In that study, other species found in the wooden vat were identified and recognized as common members of the NSLAB population. Thus, the main hypothesis of the present experimentation was the persistence of some LAB forming biofilms on the wooden vat surface during the ripening of cheese manufactured traditionally. This because the persistence of vat LAB in cheese, over time, undoubtedly shows a clear active role of these bacteria during ripening. The traditional cheeses followed at the manufacturing stage, together with those carried out in standard conditions, were then subjected to ripening and the data retrieved from the samples collected at 30, 60 and 120 days of ripening are showed in this paper. In general, the ripening time significantly affected the development of all LAB groups, except enterococci, as clearly confirmed by the canonical discriminant analysis where all the microbiological groups contributed to separate cheeses at different ripening time, except enterococci. In fact, this last group was detected at levels lower than those detected for the other LAB groups in all productions at each collection time and, interestingly, their levels registered in the traditional cheeses were almost one order of magnitude higher than those observed for the corresponding standard cheeses. The other LAB groups showed a decreasing 89

CHAPTER IV

trend in concentration over time. The highest levels were observed for mesophilic rod LAB at 30 d of ripening (in the range 107 – 108 CFU/g) while the lowest levels were registered for thermophilic coccus LAB at 120 d (in the range 105 – 106 CFU/g). These levels are in the same range of those reported for other Caciocavallo cheeses produced in southern Italy such as Caciocavallo Pugliese analyzed at 60 d that showed ca. 108 CFU/g for mesophilic rod LAB and ca. 106 CFU/g for mesophilic coccus LAB, while thermophilic LAB were not higher than 106 CFU/g (Gobbetti et al., 2002) and different Caciocavallo produced in Calabria, Campania and Basilicata regions, collected from retail markets, for which the maximum levels were in the range 8.8 – 8.9 Log CFU/g on MRS (Piraino et al., 2005). The different dominating LAB colonies were isolated from the various plate counts of the cheese samples and 803 isolates were first subjected to several phenotypic tests from which thirteen groups were obtained. Two-hundred and forty-one isolates representative of the different ripening times of the traditional and standard cheeses were differentiated by RAPD- PCR in 30 strains, evidencing a limited LAB biodiversity in the experimental cheeses. All 30 strains were identified by 16S rRNA gene sequencing as P. acidilactici, P. pentosaceus, E. casseliflavus, E. gallinarum, E. faecalis, L. rhamnosus, L. casei, L. delbrueckii, L. fermentum and L. paracasei. Except E. casseliflavus, E. gallinarum, and L. delbrueckii, most of the species identified are commonly reported to be part of the NSLAB population in several cheeses (Settanni and Moschetti, 2010). Furthermore, the species L. fermentum, L. paracasei, L. rhamnosus, L. casei, L. delbrueckii, and E. faecalis were found in other Caciocavallo type cheeses produced in South Italy (Gobbetti et al., 2002; Piraino et al., 2005; Morea et al., 2007). In our work, enterococci were detected at approximately constant levels (105 CFU/g for traditional and 104 CFU/g for standard cheeses), till 120 d of ripening. Similar levels of enterococci have been reported for other Italian raw cows’ milk cheeses (Franciosi et al., 2009). In order to better investigate the low biodiversity of LAB isolated during ripening, all LAB strains were tested for antibacterial compound production, since this character may confer advantages from competitiveness with other strains. Ten lactobacilli showed the capacity of producing BLIS and this allows to improve the safety, control the fermentation microbiota, speed maturation and increase the shelf life of the final cheeses (Deegan et al., 2006; Garde et al., 2007). In order to evaluate the influence of the LAB forming biofilms on the wooden vat during the ripening of Caciocavallo Palermitano cheese, the strains isolated from the vat before milk addition were compared (by RAPD profiles) to those collected during cheese maturation. 90

CHAPTER IV

Three strains belonging to the species E. faecalis, E. casseliflavus and E. gallinarum were found in the vat and also during ripening, even though only E. faecalis was detected till the end of the ripening period at levels of 105 CFU/g that were dominant within the enterococcal population. These strains were not found in the standard productions, evidencing the defining role of the wooden vat in enriching the milk LAB diversity at the time of cheese making. The presence of the enterococci in cheese is usually attributed to faecal contamination; but some authors assessed that the presence of these bacteria in the food matrices is not always due to the direct contact with contaminated material (Mundt, 1986; Birollo et al., 2001). Although the enterococci do not represent the major humans pathogenic, they are recognized as responsible of numerous nosocomial infections (Coque et al., 1996). However, several authors suggest that the presence of some strains of Enterococcus, established their harmlessness, is desirable, especially in long ripened traditional cheeses, as their contribution in developing the aroma is believed to be fundamental. Moreover, numerous strains of this group, originating from raw milk, are tightly linked to typicality of the final cheese (Foulquié Moreno et al., 2006). The presence of enterococci at dominant levels during ripening has been reported for cheeses produced in the Mediterranean basin, as well as for other Sicilian cheeses (Randazzo et al., 2008). In this work, within the enterococci isolated, only E. faecalis is generally associated with cheese (Settanni and Moschetti, 2010). The influence of the enterococci bacteria on the sensory properties of cheese seems to be due to specific biochemical traits such as proteolytic and lipolytic activities, citrate utilization, and production of several aromatic volatile compounds (Oliszewski et al., 2013). Due to their positive contribution to cheese flavour and their role in acceleration of the ripening process (Gardiner et al., 1999), enterococci are being proposed as adjunct cultures (De Vuyst et al., 2011; Oliszewski et al., 2013). The other NSLAB species had not been previously isolated from the wooden vat. This finding cannot exclude their presence in the wooden vat at subdominant (undetectable) levels or in a dormant/viable but not cultivable (VBNC) state. However, since standard cheese making was carried out in a stainless steel vat, the strain comparison between LAB collected from both traditional and standard productions, clarified the doubt on the origin of the strains. From this comparison, only for Lb. delbrueckii the milk or rennet origin has to be excluded, since it was only found in the traditional production (both A and B), highlighting the higher LAB biodiversity of the traditional cheese productions compared to the standard ones.

91

CHAPTER IV

5. CONCLUSIONS

In this work, the Caciocavallo Palermitano cheeses manufactured with traditional or standard technologies, performed using bulk milk derived from two farming systems, were analysed at different times in order to investigate the changes in chemical, physical and microbiological characteristics induced by the production system and during ripening. With regards to the objectives, several conclusions may be drafted. The contribution of the farming systems and cheese technology in changing chemical and physical traits was mainly evidenced by the lower contents in NaCl and soluble N, and the higher paste consistency recorded in cheeses from extensive farm and traditional technology, whereas during ripening the soluble N content and the paste yellow and consistency increased. LAB levels of the experimental cheeses were in the same range of those reported for other Caciocavallo cheeses produced in southern Italy. The ripening time affected the development of several LAB groups except enterococci, whose concentration was constant during ripening. The majority of the 10 LAB species identified are commonly reported to be part of the NSLAB population in several cheeses, including Italian Caciocavallo cheeses. The persistence of LAB from the wooden vat during ripening was evaluated by direct comparison of the polymorphic profiles and three strains belonging to the species E. faecalis, E. casseliflavus and E. gallinarum were found to be present during cheese maturation only in the traditional productions. In particular, E. faecalis FMA288 was found to dominate the enterococcal population at the end of the ripening period, evidencing the defining role of the wooden vat in the modification of LAB composition during Caciocavallo Palermitano cheese ripening. Other studies are being prepared in order to evaluate the individual contribution of each persistent strain to the aromatic and safety aspects of the final cheeses.

92

CHAPTER IV

REFERENCES

Alabiso, M., Di Grigoli, A., Bonanno, A., Alicata, M. L., Bongarra, M., Calagna, G., Console, A. (2000) Effetto del diverso comportamento al rilascio del latte sulla produzione quanti-qualitativa in bovine Modicane (Effect of the different behaviour to milk release on milk yield and quality in Modicana cows). Proceedings of the 54th Congress of Società Italiana delle Scienze Veterinarie (S.I.S.Vet.). Riva del Garda (TN), Italy. 457–458 Birollo, G.A., Reinheimer, J.A., Vinderola, C.G. (2001) Enterococci vs. nonlactic acid microflora as hygiene indicators for sweetened yoghurt. Food Microbiology 18, 597–604 Bonanno, A., Di Grigoli, A., Tornambè, G., Formoso, B., Alicata, M.L., Procida, G., Manzi, P., Marconi, S., Pizzoferrato, L. (2004) Effects of feeding regime on nutritional and aromatic characteristics of Caciocavallo Palermitano cheese. Proceedings of the 6th International Meeting on Mountain Cheeses, “Dairy food biodiversity: flavour and health properties”. Ragusa (RG), Italy. 43–50 Bonanno, A., Tornambè, G., Bellina, V., De Pasquale, C., Mazza, F., Maniaci, G., Di Grigoli, A. (2013) Effect of farming system and cheesemaking technology on the physicochemical characteristics, fatty acid profile, and sensory properties of Caciocavallo Palermitano cheese. Journal of Dairy Science 96, 710–724 Buffa, M.N., Trujillo, A.J., Pavia, M., Guamis, B. (2001) Changes in textural, microstructural and colour characteristics during ripening of cheese made from raw, pasteurized or high-pressure-treated goats’ milk. International Dairy Journal, 11, 927–934 Cavazza, A., Franciosi, E., Settanni, L., Monfredini, L., Poznanski, E. (2011) Mantenere la tipicità in un mondo che cambia. Terra Trentina 5, 43 (in Italian) Coque, T.M., Tomayko, J.F., Ricke, S.C., Okhyusen, P.C., Murray, B.E. (1996) Vancomycin-resistant enterococci from nosocomial, community, and animal sources in the United States. Antimicrobial Agents and Chemotherapics, 40, 2605–2609 Corsetti, A., Settanni, L., Braga, T.M., De Fatima Silva Lopes, M., Suzzi, G. (2008) An investigation on the bacteriocinogenic potential of lactic acid bacteria associated with wheat (Triticum durum) kernels and non- conventional flours. LWT-Food Science and Technology 41, 1173–1182 De Vuyst, L., Vaningelgem, F., Ghijsels, V., Tsakalidou, E., Leroy, F. (2011) New insights into the citrate metabolism of Enterococcus faecium FAIR-E 198 and its possible impact on the production of fermented dairy products. International Dairy Journal 21, 580–581 Deegan, L.H., Cotter, P.D., Hill, C., Ross, P. 2006 Bacteriocins: biological tools for bio-preservation and shelf- life extension. International Dairy Journal 16, 1058–1071 Didienne, R., Defargues, C., Callon, C., Meylheuc, T., Hulin, S., Montel, M.C. 2012 Characteristics of microbial biofilm on wooden vats (“gerles”) in PDO Salers cheese. International Journal of Food Microbiology 156, 91–101 Foulquié Moreno, M.R., Sarantinopoulos, P., Tsakalidou, E., De Vuyst, L. (2006) The role and application of enterococci in food and health. International Journal of Food Microbiology 106, 1–24 Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Presence of enterococci in raw cow’s milk and “Puzzone di Moena” cheese. Journal Food Processing and Preservation 33, 204–217 Gaglio, R., Francesca, N., Di Gerlando, R., Cruciata, M., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. (2014) Identification, typing, and investigation of the dairy characteristics of lactic acid bacteria isolated from 'Vastedda della valle del Belìce' cheese. Dairy Science & Technology 94, 157–180 Garde, S., Avila, M., Fernandez-Garcıa, E., Medina, M., Nuñez, M. (2007) Volatile compounds and aroma of Hispanico cheese manufactured using lacticin 481- producing Lactococcus lactis subsp. lactis INIA 639 as an adjunct culture. International Dairy Journal 17, 717–726 Gardiner, G.E., Ross, R.P., Wallace, J.M., Scanlan, F.P., Jägers, P.P., Fitzgerald, G.F., Collins, J.K., Stanton, C. (1999) Influence of a probiotic adjunct culture of Enterococcus faecium on the quality of cheddar cheese. Journal of Agricultural and Food Chemistry 47, 4907–4916 Gobbetti, M., Morea, M., Baruzzi, F., Corbo, M.R., Matarante, A., Considine, T., Di Cagno, R., Guinee, T., Fox, P.F. (2002) Microbiological, compositional, biochemical, and textural characterization of Caciocavallo Pugliese cheese during ripening. International Dairy Journal 12, 511–523

93

CHAPTER IV

Hartnett, D.J., Vaughan, A., Van Sinderen, D. (2002) Antimicrobial producing lactic acid bacteria isolated from raw barley and sorghum. Journal of the Institute of Brewing 108, 169–177 Iannarilli, A. (2002) The Italian food guide: the ultimate guide to the regional foods of Italy. Touring Editore: Milan, Italy IDF (International Dairy Federation), (1964a) Determination of the protein content of processed cheese products. Standard FIL-IDF 25:1964. International Dairy Federation: Brussels, Belgium IDF (International Dairy Federation), (1964b) Determination of the ash content of processed cheese products. Standard FIL-IDF 27:1964. International Dairy Federation: Brussels, Belgium IDF (International Dairy Federation), (1972) Cheese-determination of chloride content. Standard FIL-IDF 17A:1972. International Dairy Federation: Brussels, Belgium IDF (International Dairy Federation), (1982) Cheese and processed cheese product. Determination of the total solids content. Standard FIL-IDF 4A: (1982) International Dairy Federation: Brussels, Belgium IDF (International Dairy Federation), (1986) Cheese and processed cheese product. Determination of fat content- gravimetric method (Reference method). Standard FIL-IDF 5B:1986. International Dairy Federation: Brussels, Belgium Jackson, C.R., Fedorka-Cray, P.J., Barrett, J.B. (2004) Use of a genus- and species-specific multiplex PCR for identification of enterococci. Journal of Clinical Microbiology 42, 3558–3565 Lortal, S., Di Blasi, A., Madec, M.-N., Pediliggieri, C., Tuminello, L., Tanguy, G., Fauquant, J., Lecuona, Y., Campo, P., Carpino, S., Licitra, G. (2009) Tina wooden vat biofilm. A safe and highly efficient lactic acid bacteria delivering system in PDO Ragusano cheese making. International Journal of Food Microbiology 132, 1–8 Martin, B., Coulon, J.B., Chamba, J.F., Bugaud, C. (1997) Effect of milk urea content on characteristics of matured Reblochon cheeses. Lait 77, 505–514 Micari, P., Sarullo, V., Sidari, R., Caridi, A. (2007) Physico-chemical and hygienic characteristics of the Calabrian raw milk cheese, Caprino d’Aspromonte. Turkish Journal of Veterinary & Animal Sciences 31, 55–60 Morea, M., Matarante, A., Di Cagno, R., Baruzzi, F., Minervini, F. (2007) Contribution of autochthonous non- starter lactobacilli to proteolysis in Caciocavallo Pugliese cheese. International Dairy Journal 17, 525–534 Oliszewski, R., Wolf, I.V., Bergamini, C.V., Candioti, M., Perotti, M.C. (2013) Influence of autochthonous adjunct cultures on ripening parameters of Argentinean goat’s milk cheeses. Journal of the Science of Food and Agriculture 93, 2730–2742 Parente, E., Cogan, T.M. (2004) Starter cultures: General aspects. In Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.), Cheese: Chemistry, Physics and Microbiology. Chapman and Hall: London, 123–148 Piraino, P., Zotta, T., Ricciardi, A., Parente, E. (2005) Discrimination of commercial Caciocavallo cheeses on the basis of the diversity of lactic microflora and primary proteolysis. International Dairy Journal 15, 1138–1149 Randazzo, C.L., Pitino, I., De Luca, S., Scifò, G.O., Caggia, C. (2008) Effect of wild strains used as starter cultures and adjunct cultures on the volatile compounds of the Pecorino Siciliano cheese. International Journal of Food Microbiology 122, 269–278 Settanni, L., Di Grigoli, A., Tornambé, G., Gambino, A., Bonanno, A. (2010) Indagine microbiologica della lavorazione tradizionale del Caciocavallo Palermitano. Agrisicilia 11, 51–55 (in Italian) Settanni, L., Di Grigoli, A., Tornambé, G., Bellina, V., Francesca, N., Moschetti, G., Bonanno, A. (2012) Persistence of wild Streptococcus thermophilus strains on wooden vat and during the manufacture of a Caciocavallo type cheese. International Journal of Food Microbiology 155, 73–81 Settanni, L., Massitti, O., Van Sinderen, D., Corsetti, A. (2005) In situ activity of a bacteriocin-producing Lactococcus lactis strain. Influence on the interactions between lactic acid bacteria during sourdough fermentation. Journal of Applied Microbiology 99, 670–681 Settanni, L., Moschetti, G. (2010) Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiology 27, 691–697 Settanni, L., Moschetti, G. (2014) New trends in technology and identity of traditional dairy and fermented meat production processes. Trends in Food Science and Technology 37, 51–58

94

CHAPTER IV

Stackebrandt, E., Goebel, B.M. (1994) Taxonomic note. A place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present species definition in bacteriology. International Journal Systematic Bacteriology 44, 846–849 Tornambé, G., Di Grigoli, A., Alicata, M.-L., De Pasquale, C., Bonanno, A. (2009) Comparing quality characteristics of “Caciocavallo Palermitano” cheese from traditional and intensive production systems. Proceeding of the 15th Meeting of the FAO-CIHEAM Mountain Pastures Network. Les Diablerets, Switzerland. 153–156 Urbach, G. (1997) The flavour of milk and dairy products: II. Cheese: contribution of volatile compounds. International Journal of Dairy Technology 50, 79–89 Weisburg, W., Barns, S.M., Pelletier, D.A., Lane, D.J. (1991) 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology 173, 697–703

95

PART II

Selection of lactic acid bacteria to improve cheese quality

96

CHAPTER V

Selected lactic acid bacteria as a hurdle to the microbial spoilage

of cheese: Application on a traditional raw ewes’ milk cheese

The present chapter has been published in

International Dairy Journal

32, 126–132

2013

97

CHAPTER V

ABSTRACT

To evaluate the efficacy of lactic acid bacteria (LAB) to improve the hygienic safety of a traditional raw milk cheese, the PDO Pecorino Siciliano cheese made from raw ewes’ milk was used as a model system. Different PDO Pecorino Siciliano curds and cheeses were used as sources of autochthonous LAB subsequently used as starter (SLAB) and non-starter (NSLAB). They were screened for their acidification capacity and autolysis. SLAB showing the best performances were genotypically differentiated and identified and two strains of Lactococcus lactis subsp. lactis were selected. Among the NSLAB, Enterococcus faecalis, Lactococcus garviae and Streptococcus macedonicus strains were selected. The five cultures were used in different individual or dual inocula to produce experimental cheeses in a dairy factory whose productions were characterised by high numbers of undesirable bacteria. At 5- month of ripening, the experimental cheeses produced with LAB were characterised by undetectable levels of enterobacteria and pseudomonads and the typical sensory attributes.

Key words: Hygienic safety; Non starter lactic acid bacteria; Starter lactic acid bacteria; PDO Pecorino Siciliano cheese; Typicality; Undesired microorganisms.

98

CHAPTER V

1. INTRODUCTION

Cheese classification based on raw materials and microbial inocula includes six categories (Mucchetti and Neviani, 2006): pasteurised milk and selected starters; pasteurised milk and natural starters; thermal treated milk and natural starters; raw milk and selected starters; raw milk and natural starters; raw milk without starters. From a hygienic perspective, the latter cheese category is the one that deserves major attention, since final cheeses can become contaminated by pathogenic microorganisms as a result of their presence in raw milk and their subsequent survival during the cheese making process (Donnelly, 2004). Regarding pathogenic bacteria, the factors that mainly contribute to the safety of cheese are milk quality, starter cultures or native lactic acid bacteria (LAB), pH, salt, control of ripening conditions and chemical changes that occur in cheese during ripening (Johnson et al., 1990). Cheese cannot be made without the action of certain species of LAB (Parente and Cogan, 2004). Thus, a cheese production performed with raw milk without starter addition relies on the presence of indigenous LAB in milk and/or those transferred by the equipment used for the processing and from the environment. However, this may also determine a great variability of the final characteristics of the cheese that cannot be easily controlled by the cheese maker (Franciosi et al., 2008). Considering that the microbiology of the cheeses produced with raw milk without starters can be unpredictable, the addition of selected LAB may drive the fermentation process in an appropriate direction (Caplice and Fitzgerald, 1999). Raw milk is generally used to produce extra-hard cheeses that are ripened for a long period. However, some hygienic issues with respect to the presence of some pathogenic bacteria have been found during the long ripening of traditional cheeses, such as PDO Pecorino Siciliano, an extra-hard Italian cheese produced with raw ewes’ milk (Todaro et al., 2011). For the reasons mentioned above, the present work was aimed to evaluate the efficacy of autochthonous LAB to improve the hygienic safety of a typical cheese obtained with raw milk. The autochthonous LAB were expected to be adapted to the technology, as well as to the cheese typology. The PDO Pecorino Siciliano cheese was used as a model cheese to convert the production process from a production performed with raw milk without starters to a production carried out with raw milk and natural starters. The specific objectives for this study were: to isolate and select starter LAB (SLAB) from acidified PDO Pecorino Siciliano curds; to select non starter LAB (NSLAB); to produce experimental cheeses with different inocula of SLAB alone or in combination with NSLAB autochthonous for PDO Pecorino

99

CHAPTER V

Siciliano cheese; to evaluate the improvement of the hygienic conditions of the final products and the preservation of their typical properties by sensory analysis.

100

CHAPTER V

2. MATERIALS AND METHODS

2.1. Isolation and grouping of starter lactic acid bacteria from curd samples

Curd samples were provided by three dairy factories producing PDO Pecorino Siciliano cheese following the traditional production protocol that excludes the addition of starter LAB (Official Gazzette of the Italian Republic G.U.R.I. n. 295 of 12-22-1955). The dairy factories were located within Trapani and Agrigento provinces/districts (Sicily, Italy). Five curd samples were collected from each factory in two consecutive weeks 24 h after adding rennet. The acidified curds were transferred into sterile plastic bags and transported for approximately 90 min in a portable fridge at 8°C. Once in laboratory, 10 g of each sample were homogenised into 90 mL of sodium citrate (2% w/v) solution by means of a Stomacher (BagMixer® 400, Interscience, Saint Nom, France) and serially diluted in Ringer’s solution (Sigma-Aldrich, Milan, Italy). Presumptive rod LAB were grown on the Man-Rogosa-Sharpe (MRS) agar (Oxoid, Milan, Italy), acidified to pH 5.4 with lactic acid (5 mol/L), while presumptive coccus LAB were grown on M17 agar (Oxoid). Both agars were incubated anaerobically at 30°C for 48 h. After growth, the presumptive LAB colonies were picked up, purified and phenotypically characterised as reported by Settanni et al. (2012).

2.2. Acidification and autolysis of starter and non starter lactic acid bacteria

SLAB identified in this study and NSLAB previously isolated from PDO Pecorino Siciliano cheese (Todaro et al., 2011) were evaluated for their ability to acidify milk and to undergo autolysis. LAB cultures were grown overnight in M17 or MRS medium and centrifuged at 5,000 x g for 5 min. The cells were suspended in and washed with Ringer's solution. The acidifying capacity was assayed in 10 mL full fat ultra-high temperature treated (UHT) milk inoculated with 1% (v/v) of cell suspension, to reach a final concentration of about 107 CFU/mL and incubated at 30°C. Measurements of pH were carried out at 2 h intervals for the first 8 h and then 24, 48 and 72 h after inoculation. Autolysis of whole cells was determined in buffer solution (potassium phosphate, 50 mmol/L, pH 6.5) following the method of Mora et al. (2003) using a 6400 Spectrophotometer (Jenway Ltd., Felsted Dunmow, UK) at 600 nm. Optical density (OD) was measured at 2 h intervals for the first 8 h and then 24, 48 and 72 h after inoculation.

101

CHAPTER V

2.3. Genotypic differentiation and identification of starter lactic acid bacteria

Before genetic identification was carried out, the presumptive SLAB isolates showing the best acidifying and autolytic performances were differentiated at strain level by random amplification of polymorphic DNA-PCR (RAPD-PCR) analysis as reported by Settanni et al. (2012). Cell lysis for DNA extraction was performed on overnight cultures by the Instagene Matrix kit (Bio-Rad, Hercules, CA, USA) as described by the manufacturer. Genotypic identification was carried out by 16S rRNA gene sequencing following the scheme applied by Settanni et al. (2012).

2.4. Experimental cheese productions

The strains within SLAB and NSLAB groups showing the best acidifying and autolytic performances, alone or in combination, were selected to be employed in cheese production. Cells were centrifuged and washed as reported above and re-suspended in Ringer’s solution till reaching an OD of ca. 1.00 which approximately corresponds to a concentration of 109 CFU/mL as evaluated by plate count. Cheese trials were carried out at a dairy factory located in Menfi (Italy), which was one of the providers of both curd and ripened PDO Pecorino Siciliano cheeses used to isolate SLAB and NSLAB, respectively. This dairy factory produces PDO Pecorino Siciliano cheeses daily employing the same wooden vat since seven years. The bulk milk (250 L) used for the experimental cheese making was first put in contact with that traditional wooden vat, under manual agitation, for 15 min, which represents the time that commonly occurs before rennet addition. After that, the milk was transferred into seven plastic vats (37 L each); the milk in six of the vats was inoculated with LAB (described under section 3.4.) to obtain the experimental cheeses (EC1-EC6), while the milk in one vat was supplemented with the same volume of Ringer’s solution without bacteria and represented the control vat to obtain the control cheese (CC). SLAB and NSLAB were inoculated at a final concentration of approximately 107 and 103 CFU/mL, respectively. The cheese productions followed then the traditional protocol (Fig. 1) and the cheeses were ripened for five months. The cheese trials were carried out in duplicate in two consecutive weeks.

102

CHAPTER V

Fig. 1. Flow diagram of PDO Pecorino Siciliano cheese production. RH, relative humidity.

2.5. Analyses of the experimental cheeses

Temperature and pH of milk and curd samples were measured by a portable pH meter (waterproof pHTestr 30, Eutech Instruments, Nijkerk, The Nederlands). The different samples (milks and curds) to be microbiologically investigated were collected during cheese

103

CHAPTER V

production in sterile containers, immediately lowered in temperature and transported for 90 min under refrigeration with a portable fridge to the laboratory of Agricultural Microbiology (University of Palermo). Other curd samples were collected to be followed for pH decrease and LAB counts during the first hours (2, 4, 6, 8, 24 and 48) after production and were kept at ambient temperature during transport and for 48 h. After a 5 month ripening period, the 14 cheeses were sampled and subjected to the same analyses as performed on the refrigerated curd samples mentioned above. The decimal dilutions of milk (10 mL) samples were prepared in Ringer's solution. The first dilution of curd (10 g) and cheese (25 g) samples was performed in sodium citrate solution as reported above (paragraph 2.1.), while further serial dilutions were carried out in Ringer's solution. The total mesophilic count (TMC), total psychrotrophic counts (TPC), number of Enterobacteriaceae, enterococci, pseudomonads, positive coagulase staphylococci (PCS), rod and coccus LAB, yeasts and clostridia were estimated as reported by Settanni et al. (2012). The microbiological counts were carried out in duplicate. Detection of Listeria monocytogenes was carried out on 25 g of cheese sample, after pre- enrichment, as described by Mucchetti et al. (2008). The concentration of salt (NaCl) in the final cheeses was determined by the Volhard method (AOAC, 1975). Microbial data were statistically analysed by the STATISTICA software (StatSoft Inc., Tulsa, OK, USA) using a generalised linear model (GLM) including the effects of sample; the Student “t” test was used for mean comparison. The post-hoc Tukey method was applied for pairwise comparison. Significance level was P<0.05.

2.6. Sensory analysis

In order to evaluate the influence of the several bacterial inocula in the definition of the final characteristics of cheese, the different cheese productions, at 5 month ripening, were subjected to the sensory evaluation. To define the sensory profile of the experimental cheeses, a descriptive panel of nine graders (five females and four males, 30 – 50 years old) performed the organoleptic evaluation of the cheeses. All panellists were trained at CoRFiLac (Ragusa, Italy), which is a consortium whose research activities are focused on the local dairy products (www.corfilac.it), and participate at their sensory profiling and other types of sensory analysis during the whole year. All the graders were familiar with the descriptive sensory analysis of

104

CHAPTER V

the Pecorino cheese variety and were specifically trained for the PDO evaluation of cheese products during the previous years of PDO Pecorino Siciliano certification. The score chart used for the expert panellists of CoRFiLaC in this study was the same as used for the certification of the PDO Pecorino Siciliano. The scores of the nine graders were reported in the score chart sequentially for appearance (colour, oil, eyes after cutting of the cheese, uniformity attributes), smell (odour, pasture, unpleasent attributes), taste (taste, salt, spicy, bitter) and consistency (soft/hard, saliva evoking, dispersion attributes). The evaluations were acquired by the software Compusense five v4.6 (Compusense, Guelph, Canada). The sensory tests were carried out following the ISO 13299 (2003) indications. The graders were not informed about the experimental design and had no specific information about the individual cheese samples tested. The graders operated in individual chambers (ISO 8589:2007). The seven cheeses made in each of the two consecutive weeks were tested in a randomised order of presentation. The samples (pieces of about 3 × 3 × 2 cm in size) were left at ambient temperature (ca. 20°C) for 60 min before administration and they were presented in coded white plastic plates. Two evaluation sessions were performed. Sensory evaluations were statistically analysed using the GLM procedure in SAS 2004, version 9.1.2 (Statistical Analysis System Institute Inc., Cary, NC, USA). The discrimination efficiency of the attributes for each assessor was tested by a 2-factor analysis of variance (ANOVA), with graders (i=1..9) and experimental cheeses (j=1..7) as fixed factors. Least square means (LSM) were compared using T test (P < 0.05).

3. RESULTS AND DISCUSSION

3.1. Isolation and phenotypic grouping of starter lactic acid bacteria from curd samples

The samples of PDO Pecorino Siciliano curd, acidified for approximately 24 h at ambient temperature, as performed at the dairy factory in the routine cheese production, contained between 7.5 and 8.9 Log CFU/g of presumptive coccus LAB, while presumptive rod LAB were in the range 6.2 – 8.2 Log CFU/g. On the basis of appearance (colour, morphology, edge, surface and elevation) at least 3 – 5 identical colonies per curd sample were randomly picked up from M17 plates, forming a total of 129 cultures. They were considered presumptive LAB, as being Gram-positive and catalase-negative. Phenotypic characterisation allowed the separation of all coccus LAB isolates collected from the acidified curds into three groups: group I (102 isolates) included LAB forming short chains which were able to grow at 15°C and at pH 9.2; group II (21 isolates) included LAB 105

CHAPTER V

forming short chains which were able to grow at 15 and 45°C, at pH 9.2 and in presence of 6.5% NaCl; group III (6 isolates) included LAB forming long chains which were able to grow at 45°C, but not in the other conditions tested. All isolates were characterised by a homofermentative metabolism of lactose which is a basic characteristic for application in cheeses for which the presence of eyes is undesired.

3.2. Evaluation of acidification and autolysis of starter and non starter lactic acid bacteria

About 30% of the SLAB isolates of each phenotypic group, or at least one isolate per curd for the less numerous groups, forming a total 40 isolates, were randomly chosen and, together with 22 NSLAB previously isolated and identified from ripened PDO Pecorino Siciliano cheese (Todaro et al., 2011), subjected to the evaluation of their aptitudes in cheese making. The acidification capacity and the autolysis were tested, so that the optimal SLAB were characterised by a fast and appropriate acidification and a rapid autolysis. Optimal NSLAB, however, showed opposite performances (Franciosi et al., 2009). The results of the acidification and autolysis of the 62 LAB (results not shown) indicated that eight SLAB cultures (CAG4, CAG5, CAG12, CAG23, CAG25, CAG37, CAG60 and CAG70) showed a rapid decrease of milk pH (Fig. 2A), but only two of them (CAG4 and CAG37) were characterised also by a rapid autolysis (Fig. 2B). Regarding NSLAB, identified at species level by Todaro et al. (2011), the combination of both parameters indicated the strains PSL67, PSL71 and PSL72 as the weakest acidifiers with pH 6.14, 6.16 and 6.21, respectively, at 24 h after inoculation with the slowest autolytic activity with 0.088, 0.085, 0.096 OD decrease, respectively, after 72 h. These three strains were identified as: Lactococcus (L.) garvieae, Enterococcus (E.) faecalis and Streptococcus (S.) macedonicus, respectively. The three strains were chosen not only for their technological potential: E. faecalis has been reported to be linked to the typicality of final products (Foulquié Moreno et al., 2006); L. garvieae and S. macedonicus are found in raw milk (Franciosi et al., 2009) and they are commonly isolated from several Italian cheeses, including PDO Pecorino Siciliano cheeses (Todaro et al., 2011), and possess properties useful to provide cheese typicality during ripening (Fortina et al., 2007; Settanni et al., 2011).

106

CHAPTER V

A 7.07

6.56,5

6.06

55,5.5 pH 5.05

4.54,5

4.04

3.53,5

3.03 0 10 20 30 40 50 60 70 80 Time (h)

B 1.401,40

1.201,20

1.001,00 600

OD 0.800,80

0.600,60

0.400,40

0.200,20 0 10 20 30 40 50 60 70 80 Time (h) Fig. 2. Acidification (A) and autolysis (B) of SLAB isolated from acidified PDO Pecorino Siciliano curds. Symbols: ♦, isolate CAG4; ■, isolate CAG5; ▲, isolate CGA12; ◊, isolate CAG23; ∆, isolate CAG25; ●, isolate CAG37; □, isolate CAG60; ○, isolate CAG70; ×, isolate CAG76 (slow acidifier, added as control in A). Bars represent standard deviation of the mean. Vertical bars not visible are smaller than symbol size.

107

CHAPTER V

3.3. Recognition of starter lactic acid bacteria at strain and species level

The eight isolates of the SLAB group characterised by a strong and fast decrease of milk pH were analysed by RAPD-PCR and recognised as different strains (Fig. 3). All eight strains listed above belonged to the phenotypic group I. Strains CAG4 and CAG37 which showed the fastest autolysis, isolated from two distinct curds, were subjected to the analysis of 16S rRNA gene sequencing. Both strains were identified as Lactococcus lactis subsp. lactis (Acc. No. KC351901, KC351902) which is commonly found during the acidification of several cheeses and used as mesophilic starter (Settanni and Moschetti, 2010).

1 2 3 4 5 6 7 8 9 10

Fig. 3. RAPD-PCR profiles of rapid acidifying SLAB isolated from PDO Pecorino Siciliano cheese obtained with primer M13. Lanes: 1, GeneRuler 100bp Plus DNA ladder; 2, strain CAG4; 3, strain CAG5; 4, strain CAG12; 5, strain CAG23; 6, strain CAG25; 7, strain CAG37; 8, strain CAG60; 9, strain CAG70; 10, negative control.

3.4. Evolution of chemical parameters and microbial populations during experimental cheese making and ripening

L. lactis subsp. lactis CAG4 and CAG37 were selected as starter cultures, while L. garvieae PSL67, E. faecalis PSL71 and S. macedonicus PSL72 were chosen as secondary adjunct cultures in cheese making. The experimental PDO Pecorino Siciliano cheese productions were carried out in a dairy factory whose ripened cheese were characterised by

108

CHAPTER V

high numbers of undesired (pathogenic/spoilage) bacteria (Todaro et al., 2011). The raw ewes’ milk used was characterised by a concentration of TMC of 6.2 Log CFU/mL which is higher than the limit for the “good microbiological quality” in Europe for raw ewes’ milk (<500,000 CFU/mL) to be processed into cheese with a manufacturing process which do not involve any heat treatment [Reg. (CE) 853/2004 (Alleg. II, Sez. IX, Cap. I, Parag. III)]. Cheese production were performed with different inocula of SLAB (alone or in combination) or SLAB and the three species of NSLAB as listed below: CC, control cheese not inoculated; EC1, with L. lactis CAG4; EC2, with L. lactis CAG37; EC3, with L. lactis CAG4/L. lactis CAG37; EC4, with L. lactis CAG4 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC5, with L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC6, with L. lactis CAG4/L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72. The pH drop followed during 48 h after curd production is shown in Fig. 4. All inoculated curds (EC) showed a faster decrease of pH than the control curd (CC). No statistical significant differences in acidification were found between the curds inoculated with L. lactis subsp. lactis CAG4 and CAG37 alone (EC1 and EC2) or in combination with NSLAB (EC4 and EC5). However, when L. lactis subsp. lactis CAG4 and CAG37 where inoculated together (EC3 and EC6) the fastest pH drop was observed.

6.56,5

6.06

55,5.5

pH 5.05

4.54,5

4.04

3.53,5 0 10 20 30 40 50 Time (h) Fig. 4. pH during the acidification of experimental PDO Pecorino Siciliano cheese curds. Abbreviations: CC, control curd; EC, experimental curd. Symbols: ♦, CC; □, EC1; ▲, EC2; ■, EC3; ●, EC4; ○, EC5; ∆, EC6. Bars represent standard deviation of the mean. Vertical bars not visible are smaller than symbol size.

109

CHAPTER V

The bulk milk, after resting in the wooden vat, hosted 6.5 Log CFU/mL of TMC and it was dominated by coccus LAB (Table 1). After inoculation, all experimental vats showed a concentration of coccus LAB 1 Log cycle higher than the control vat (results not shown). Except rod LAB, whose concentration was affected by the addition of the selected SLAB, all other microbial populations did not show differences in the levels detected for the different vats. After coagulation, a 10-fold increase in concentration was registered for the majority of the microbial groups (Table 1). On the contrary, the almost complete disappearance of clostridia was evidenced by the MPN technique. At seven days from inoculation, curds evolved almost similarly, but CC was characterised by higher levels of Enterobacteriaceae and pseudomonads than ECs. Furthermore, all curds were dominated by LAB and, regarding this population, the differences between CC and ECs were less evident, showing that the indigenous LAB were able to develop during the seven days after curd production. After five months of ripening, several microbiological data were almost superimposable among the different vats (Table 1), including CC, but the concentration of Enterobacteriaceae and pseudomonads, undetectable for all ECs, were 3.7 Log CFU/g in CC. A significant reduction of Enterobacteriaceae concentration caused by L. lactis subsp. lactis has been reported for Serra de Estrela cheese made from raw ewes’ milk (Macedo et al., 2004), whereas no previous study has evaluated the inhibitory effect of the addition of lactococci on the growth of pseudomonads in raw ewes’ milk cheeses. Regarding pseudomonads, known agents of food spoilage, the higher the pH the higher their concentrations (Hayes, 1995); thus, the rapid decrease of pH due to the activity of SLAB reduced the Pseudomonas spp. numbers. Although the stressing conditions of cheese during ripening should determine the reduction of Enterobacteriaceae population, the presence of these bacteria in raw ewes’ milk cheeses at consistent levels is not a rare finding (Tavaria and Malcata, 2000; Prodromou et al., 2001). L. monocytogenes was not detected in any cheese. The concentration of salt was registered at 6.59, 6.45 and 6.64% (w/dw) when the strain L. lactis subsp. lactis CAG37 was employed as starter culture in EC2, EC3 and EC5, respectively. Lower levels of salt were determined in CC (5.01%, w/dw) and in cheese with L. lactis subsp. lactis CAG4 as starter, 5.13 and 5.87% (w/dw) for EC1 and EC4, respectively. The faster and stronger acidification caused by L. lactis subsp. lactis CAG37 determined a higher syneresis of the curd and the subsequent higher salt concentration (Salvadori Del Prato, 1998) in EC2, EC3 and EC5 than CC.

110

Table 1. Microbial loadsa of samples collected through experimental PDO Pecorino Siciliano cheese productions.

Samples pH Media PCA-SkM PCA-SkM VRBGA KAA PAB BP MRS M17 YGC RCMb 7°C 30°C Bulk milk 6.34 ± 0.02 4.9 ± 0.1 6.2 ± 0.2 3.3 ± 0.5 3.5 ± 0.6 3.6 ± 0.2 2.9 ± 0.3 4.9 ± 0.1 6.0 ± 0.2 <1 1.6 Wooden vat surface n.d. 4.1 ± 0.2 6.2 ± 0.9 1.5 ± 0.3 3.3 ± 0.8 0.5 ± 0.3 <1 3.7 ± 0.1 5.7 ± 0.7 1.8 ± 0.4 1.6 Bulk milk in wooden vat 6.34 ± 0.05 5.0 ± 0.5 6.5 ± 0.5 3.9 ± 0.4 3.9 ± 0.2 3.3 ± 0.2 2.8 ± 0.1 5.0 ± 0.1 6.1 ± 0.3 <1 1.6

Curds at T0: CC 6.30 ± 0.03 6.0 ± 0.6 6.8 ± 0.3 4.6 ± 0.2 4.6 ± 0.2 4.3 ± 0.3 3.6 ± 0.1 5.8 ± 0.5 6.9 ± 0.2 <2 1.6 EC1 6.30 ± 0.08 5.8 ± 0.1 8.4 ± 0.6 4.8 ± 0.2 4.7 ± 0.5 4.0 ± 0.4 3.5 ± 0.2 7.4 ± 0.6 8.3 ± 0.6 <2 1.6 EC2 6.18 ± 0.03 5.5 ± 0.2 8.5 ± 0.4 4.8 ± 0.0 4.4 ± 0.3 3.9 ± 0.4 3.7 ± 0.3 7.2 ± 0.3 8.5 ± 0.4 <2 1.6

EC3 6.16 ± 0.06 6.2 ± 0.2 8.4 ± 0.4 4.6 ± 0.2 4.4 ± 0.2 4.1 ± 0.6 3.6 ± 0.3 7.6 ± 0.2 8.4 ± 0.5 <2 1.6 CHAPTER EC4 6.15 ± 0.07 6.0 ± 0.7 8.1 ± 0.5 4.6 ± 0.2 4.7 ± 0.1 3.8 ± 0.2 3.8 ± 0.6 7.3 ± 0.2 8.5 ± 0.4 <2 1.6 EC5 6.05 ± 0.02 5.5 ± 0.4 8.5 ± 0.2 4.6 ± 0.4 4.6 ± 0.2 4.2 ± 0.6 3.9 ± 0.4 7.5 ± 0.2 8.0 ± 0.2 <2 1.6 EC6 5.96 ± 0.06 5.6 ± 0.5 8.5 ± 0.6 4.7 ± 0.2 4.7 ± 0.3 4.2 ± 0.2 4.0 ± 0.3 7.1 ± 0.4 8.0 ± 0.7 <2 1.6

V

Curds after 7 d: CC 4.23 ± 0.09 5.3 ± 0.5 6.9 ± 0.3 5.8 ± 0.7 4.8 ± 0.4 4.9 ± 0.2 2.7 ± 0.3 7.8 ± 0.4 7.8 ± 0.7 <2 0 EC1 4.18 ± 0.07 6.5 ± 0.3 7.6 ± 0.3 3.4 ± 0.2 4.5 ± 0.2 4.0 ± 0.5 2.6 ± 0.3 8.0 ± 0.3 8.1 ± 0.3 <2 0 EC2 4.20 ± 0.06 6.8 ± 0.8 7.7 ± 0.4 3.8 ± 0.3 4.4 ± 0.3 3.9 ± 0.2 2.3 ± 0.3 8.2 ± 0.3 8.2 ± 0.4 <2 0 EC3 4.16 ± 0.05 6.9 ± 0.4 7.9 ± 0.3 3.7 ± 0.2 4.3 ± 0.4 3.9 ± 0.4 2.3 ± 0.3 8.3 ± 0.4 8.4 ± 0.4 <2 1.6 EC4 4.18 ± 0.09 6.7 ± 0.6 7.6 ± 0.5 3.8 ± 0.4 4.9 ± 0.4 4.2 ± 0.4 2.6 ± 0.5 8.3 ± 0.4 8.5 ± 0.5 <2 0 EC5 4.18 ± 0.06 6.5 ± 0.6 7.5 ± 0.3 3.4 ± 0.5 4.7 ± 0.5 4.1 ± 0.2 2.4 ± 0.2 8.4 ± 0.2 8.4 ± 0.3 <2 0 EC6 4.18 ± 0.06 7.2 ± 0.6 7.8 ± 0.6 3.5 ± 0.5 4.5 ± 0.5 4.1 ± 0.4 2.3 ± 0.1 8.7 ± 0.6 8.8 ± 0.3 <2 0 Ripened cheeses: CC 5.57 ± 0.06 3.2 ± 0.1 7.7 ± 0.1 3.7 ± 0.1 5.1 ± 0.3 3.7±0.4 <2 7.6 ± 0.4 7.3 ± 0.2 <2 0 EC1 5.47 ± 0.03 <2 7.4 ± 0.3 <1 5.4 ± 0.5 <2 <2 7.6 ± 0.2 7.3 ± 0.1 <2 0

111

EC2 5.57 ± 0.02 <2 7.4 ± 0.5 <1 5.5 ± 0.3 <2 <2 7.4 ± 0.1 7.2 ± 0.5 <2 0 EC3 5.62 ± 0.04 <2 7.2 ± 0.6 <1 5.7 ± 0.6 <2 <2 7.5 ± 0.4 7.4 ± 0.4 <2 0 EC4 5.51 ± 0.02 <2 7.4 ± 0.3 <1 5.6 ± 0.2 <2 <2 7.6 ± 0.3 7.4 ± 0.4 <2 0 EC5 5.51 ± 0.06 <2 7.5 ± 0.3 <1 5.5 ± 0.6 <2 <2 7.4 ± 0.3 7.2 ± 0.3 <2 0 EC6 5.55 ± 0.06 <2 7.8 ± 0.1 <1 5.5 ± 0.6 <2 <2 7.2 ± 0.3 7.4 ± 0.1 <2 0 Statistical significancec P<0.001 P<0.001 P<0.05 P<0.001 P<0.001 P<0.001 - P<0.001 P<0.001 - - a Log CFU/mL for milk samples, Log CFU/g for curds and cheeses, Log CFU/cm-2 for wooden vat surface. b As estimated by MPN. c Statistical significance is referred to ripened cheeses. Abbreviations: PCA-SkM 7°C, plate count agar added with skimmed milk incubated at 7°C for total psychrotrophic counts; PCA-SkM 30°C, plate count agar added with skimmed milk incubated at 30°C for total mesophilic counts; VRBGA, violet red bile glucose agar for Enterobacteriaceae; KAA, kanamycin aesculin azide agar for enterococci; PAB, Pseudomonas agar base for pseudomonads; BP, Baird Parker for positive coagulase staphylococci; MRS, de Man-Rogosa-Sharpe agar for mesophilic rod LAB; M17 agar for mesophilic coccus LAB; YGC, yeast glucose dichloran rose bengal chloramphenicol agar for yeasts; RCM, reinforced clostridial medium for clostridia; n.d., not determined.

CHAPTER CHAPTER Sample designation: CC, control cheese; EC1, with L. lactis CAG4; EC2, with L. lactis CAG37; EC3, with L. lactis CAG4/L. lactis CAG37; EC4, with L. lactis CAG4 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC5, with L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC6, with L. lactis CAG4/L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72. Results indicate mean values ± S.D. of four plate counts (carried out in duplicate for two independent productions).

V

112

CHAPTER V

3.5. Sensory evaluation

The sensory evaluation carried out by the expert graders recognised both control cheeses produced in this study as typical PDO Pecorino Siciliano cheeses. The sensory profiles of the experimental cheeses compared to the control cheese (Table 2) showed that only five attributes (colour, eyes, taste, salt and saliva evoking) were significantly different among cheeses. The most notable differences were evidenced by saliva evoking and eyes. The highest number and diameter of eyes were displayed by the control cheese. The presence of less eyes in the experimental cheeses may be the effect of a rapid inhibition of coliforms in these cheeses. The inoculation with L. lactis subsp. lactis CAG4 (EC1) showed less differences from the control cheese, especially for colour, oil, odour intensity, unpleasant, salt and dispersion. Thus, the addition of L. lactis subsp. lactis CAG4 did not alter the typicality of the final cheese. The finding that inocula of L. lactis subsp. lactis at 107 CFU/mL did not modify the aroma of cheese is not surprising; Centeno et al. (2002) reported that not all strains of L. lactis were able to enhance the flavour intensity of raw ewes’ milk cheeses, even though their levels of inoculation were high. In our study, the bulk milk hosted a concentration level of TMC even higher (more than 1 Log cycle) than that of the milk processed by Centeno et al. (2002). The addition of NSLAB was not effective in the modification of the sensory characteristics of the final cheeses. This could be due to the low level of inocula (approximately 103 CFU/mL of milk) chosen to avoid the interference in the acidification process or negative influences in the mature cheeses by NSLAB (Franciosi et al., 2008).

113

Table 2. Sensory characteristics of experimental PDO Pecorino Siciliano cheeses (LSM) ripened for five months.

Attributes Cheese samples SEM Significancea

CC EC1 EC2 EC3 EC4 EC5 EC6 Graders Cheese colour 6.48 A 6.38 A 6.40 A 5.95 AB 6.56 A 5.66 B 5.96 AB 0.19 * * oil 3.10 3.00 2.56 2.90 2.94 2.81 2.92 0.13 *** ns eyes 2.93 A 2.53 A 2.09 B 2.63 A 2.25 AB 2.29 AB 2.60 A 0.17 *** ** uniformity 11.89 12.19 12.38 11.86 12.36 11.82 11.80 0.20 *** ns odour intensity 7.87 7.98 8.07 8.44 8.02 8.43 8.43 0.17 *** ns pasture 5.02 5.25 5.08 5.13 5.05 5.30 5.12 0.14 *** ns unpleasant 1.63 1.67 1.82 1.87 1.73 1.64 1.86 0.16 *** ns CHAPTER taste intensity 7.97 ab 8.12 ab 7.79 a 8.14 b 8.18 b 8.42 b 8.15 b 0.13 *** * salt 4.62 a 4.77 a 5.35 b 5.01 ab 4.57 a 5.06 ab 5.02 ab 0.22 ns *

V

bitter 6.05 6.35 6.31 6.03 5.68 6.40 6.02 0.24 ** ns spicy 1.64 1.86 1.70 1.77 1.65 1.91 1.75 0.13 *** ns soft/hard 6.29 6.05 6.80 6.39 6.28 6.48 6.38 0.18 ** ns saliva evoking 11.17 ac 11.74 b 10.79 c 11.30 ab 11.45 ab 11.38 ab 11.31 ab 0.18 *** ** dispersion 5.02 4.96 5.39 4.51 4.62 5.26 4.50 0.22 *** ns a P value: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ns = not significant. Lowercase (a, b, c) and uppercase (A, B, C) letters indicate different statistical significances at P values of ≤0.05) and P≤0.01) Abbreviations: LSM, least square means; SEM, standard error of means. Sample designation: CC, control cheese; EC1, with L. lactis CAG4; EC2, with L. lactis CAG37; EC3, with L. lactis CAG4/L. lactis CAG37; EC4, with L. lactis CAG4 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC5, with L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72; EC6, with L. lactis CAG4/L. lactis CAG37 and L. garvieae PSL67/E. faecalis PSL71/S. macedonicus PSL72.

114

CHAPTER V

4. CONCLUSIONS

The addition of selected autochthonous LAB in the raw ewes’ milk showing a low hygienic quality determined the production of PDO Pecorino Siciliano cheese characterised by acceptable hygienic conditions. In particular, the individual addition of L. lactis subsp. lactis CAG4 determined also the preservation of the typical sensory profile of this traditional cheese. Works are being prepared in order to test the resistance of this strain to the most common dairy viruses and to evaluate its performances in the several dairy factories producing PDO Pecorino Siciliano cheese, which are gathered into a consortium for the protection of this traditional cheese production.

115

CHAPTER V

REFERENCES

Association of Official Analytical Chemists (1975) Official methods of analysis. 12th ed. Assoc. Offic. Anal. Chem., Washington, DC Caplice, E., Fitzgerald, G. F. (1999) Food fermentations: role of microorganisms in food production and preservation. International Journal of Food Microbiology 50, 131–149 CE Regulation 853 (2004) Available from http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2004:1 39:0055:0205:IT:PDF. Accessed 1.15.13 Centeno, J. A., Tomillo, F. J., Fernández-García, E., Gaya, P., Nuñez, M. (2002) Effect of Wild Strains of Lactococcus lactis on the Volatile Profile and the Sensory Characteristics of Ewes’ Raw Milk Cheese. Journal of Dairy Science 85, 3164–3172 Donnelly, C. W. (2004) Growth and survival of microbial pathogens in cheese. In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds) Cheese: Chemistry, Physics and Microbiology. Chapman and Hall: London 541–560 Fortina, M. G., Ricci, G., Foschino, R., Picozzi, C., Dolci, P., Zeppa, G., Cocolin, L., Manachini, P.L. (2007) Phenotypic typing, technological properties and safety aspects of Lactococcus garvieae strains from dairy environments. International Dairy Journal 103, 445–453 Foulquié Moreno, M. R., Sarantinopoulos, P., Tsakalidou, E., De Vuyst, L. (2006) The role and application of enterococci in food and health. International Journal of Food Microbiology 106, 1–24 Franciosi, E., Settanni, L., Carlin, S., Cavazza, A., Poznanski, E. (2008) A factory-scale application of secondary adjunct cultures selected from lactic acid bacteria during “Puzzone di Moena” cheese ripening. Journal of Dairy Science 91, 2981–2991 Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Biodiversity and technological potential of wild lactic acid bacteria from raw cows’ milk. International Dairy Journal 19, 3–11 G.U.R.I. 295 (1955) Riconoscimento delle denominazioni circa i metodi di lavorazione, caratteristiche merceologiche e zone di produzione dei formaggi. Available from http://www.normattiva.it/uri- res/N2Ls?urn:nir:presidente.repubblica:decreto:1955;1269. Accessed 1.4.13 Hayes, P.R. (1995) Food microbiology and hygiene. Chapman and Hall: London ISO 13299 (2003) Sensory analysis — Methodology — General guidance for establishing a sensory profile ISO 8589 (2007) Sensory analysis – General guidance for the design of test rooms Johnson, E. A., Nelson, J. H., Johnson, M. (1990) Microbiological safety of cheese made from heat treated milk. Part I: executive summary, introduction and history. Journal of Food Protection 53, 441–452 Macedo, A. C., Tavares, T. G., Malcata, F. X. (2004) Influence of native lactic acid bacteria on the microbiological, biochemical and sensory profiles of Serra da Estrela cheese. Food Microbiology 21, 233– 240 Mora, D., Musacchio, F., Fortina, M. G., Senini, L., Manachini, P. L. (2003) Autolytic activity and pediocin- induced lysis in Pediococcus acidilactici and Pediococcus pentosaceus strains. Journal of Applied Microbiology 94, 561–570 Mucchetti, G., Bonvini, B., Remagni, M.C., Ghiglietti, R., Locci, F., Barzaghi, S., Francolino, S., Perrone, A., Rubiloni, A., Campo, P., Gatti, M., Carminati, D. (2008) Influence of cheese-making technology on composition and microbiological characteristics of Vastedda cheese. Food Control 19, 119–125 Mucchetti, G., Neviani, E. (2006) Microbiologia e tecnologia lattiero-casearia. Qualità e sicurezza. Milano: Tecniche Nuove Parente, E., Cogan, T.M. (2004) In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.) Cheese: Chemistry, Physics and Microbiology. Chapman and Hall, London 123–148 Prodromou, K., Thasitou, P., Haritonidou, E., Tzanetakis, N., Litopoulou-Tzanetaki, E. (2001) Microbiology of ‘‘Orinotyri’’, a ewe’s milk cheese from the Greek mountains. Food Microbiology 18, 319–328 Salvadori del Prato, O. (1998) Tecnologia Lattiero-Casearia. Edagricole, Bologna, Italy Settanni, L., Moschetti, G. (2010). Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiology 27, 691–697 Settanni, L., Franciosi, E., Cavazza, A., Cocconcelli, P. S., Poznanski, E. (2011) Extension of Tosèla cheese shelf-life using non-starter lactic acid bacteria. Food Microbiology 28, 883–890

116

CHAPTER V

Settanni, L., Di Grigoli, A., Tornambé, G., Bellina, V., Francesca, N., Moschetti, G., Bonanno, A. (2012) Persistence of wild Streptococcus thermophilus strains on wooden vat and during the manufacture of a Caciocavallo type cheese. International Journal of Food Microbiology 155, 73–81 Tavaria, F. K., Malcata, F. X. (2000) On the microbiology of Serra da Estrela cheese: geographical and chronological considerations. Food Microbiology 17, 293–304 Todaro, M., Francesca, N., Reale, S., Moschetti, G., Vitale, F., Settanni, L. (2011) Effect of different salting technologies on the chemical and microbiological characteristics of PDO Pecorino Siciliano cheese. European Food Research and Technology 233, 931–940

117

CHAPTER VI

In vivo application and dynamics of lactic acid bacteria for the

four-season production of Vastedda-like cheese

The present chapter has been published in

International Journal of Food Microbiology

177, 37–48

2014

118

CHAPTER VI

ABSTRACT

Twelve lactic acid bacteria (LAB), previously selected in vitro (Gaglio et al., 2014), were evaluated in situ for their potential to act as starter cultures for the continuous four-season production of Vastedda-like cheese, made with raw ewes’ milk. The strains belonged to Lactobacillus delbrueckii, Lactococcus lactis subsp. cremoris, Leuconostoc mesenteroides subsp. mesenteroides and Streptococcus thermophilus. LAB were first inoculated in multiple- strain combinations on the basis of their optimal growth temperatures in three process conditions which differed for milk treatment and medium for strain development: process 1, growth of strains in the optimal synthetic media and pasteurised milk; process 2, growth of strains in whey based medium (WBM) and pasteurised milk; process 3, growth of strains in WBM and raw milk. The strains that acidified the curds in short time, as shown by pH drop, were all mesophilic and were then tested in single inocula through the process 3. Randomly amplified polymorphic DNA (RAPD)-PCR analysis applied to the colonies isolated from the highest dilutions of samples confirmed the dominance of the added strains after curd acidification, stretching and storage. After 15 d of refrigerated storage, the decrease in pH values showed an activity of the mesophilic strains at low temperatures, but only Lc. lactis subsp. cremoris PON153, Ln. mesenteroides subsp. mesenteroides PON259 and PON559 increased their number during the 15 days at 7°C. A sensory evaluation indicated that the cheeses obtained applying the protocol 3 and inoculated with lactococci as the most similar to the protected denomination of origin (PDO) cheese and received the best scores by the judges. Thus, the experimental cheeses obtained with raw milk and inoculated with single and multiple combinations of lactococci were subjected to the analysis of the volatile organic compounds (VOCs) carried out by headspace solid phase micro extraction (SPME) technique coupled with gas chromatography with mass spectrometric detection (GC/MS). The dominance of lactococci over thermophilic LAB of raw milk was verified during summer production and, based on the combination of VOC profiles and sensory evaluation of the final cheeses, the multi-strain Lactococcus culture resulted the most suitable starter preparation for the full-year production of Vastedda-like cheese.

Key words: Fermentation; Lactic acid bacteria; Pilot plant; Raw milk; Starter cultures; Traditional cheese.

119

CHAPTER VI

1. INTRODUCTION

In the last years, the request for traditional dairy products increased and this phenomenon is still on the increase. Furthermore, the consumers demanding foods with no or reduced chemical preservatives for food conservation (Leite et al., 2006), determined a re-discovery of typical cheeses produced in restricted areas (Settanni et al., 2012a). In Italy, the majority of ewes’ milk cheeses are processed with raw milk and are considered to be traditional; some of them enjoy a “protected designation of origin” (PDO) status (Todaro et al., 2011). Within this group, “Vastedda della valle del Belìce” cheese is typical of the homonymous valley of Sicily (Italy); it is produced without the addition of starter cultures (GUE no. C 42/16 19.2.2010) applying the technology of stretched (“pasta filata”) cheeses consisting of an acidification followed by the scalding of the acidified curd (Salvadori del Prato, 1998). “Vastedda della valle del Belìce” cheese does not undergo a ripening process, it is sealed under vacuum and kept under refrigeration until consumption which can take place after a short time from production (Mucchetti et al., 2008). This is in contrast with the general trend to use raw milk for extra-hard cheeses that are ripened for a long period (Settanni et al., 2013). Raw milk cheeses deserve greater attention than cheeses made with thermal treated milk since the final products can become contaminated by pathogenic microorganisms as a result of their presence in raw milk, as well as their subsequent survival during the cheese making process (Donnelly, 2004). This is particularly true when the cheeses are consumed fresh, even though the stretching phase at high temperatures of pasta filata cheeses contributes to the safety of the resulting products. Since cheeses can be obtained only if lactic acid bacteria (LAB) are present in milk before coagulation (Parente and Cogan, 2004), they should be contaminants of the milk or the equipment used for cheese making, or they must be added (Settanni and Moschetti, 2010). The last strategy may compromise the typicality of the final product, but the addition of autochthonous strains that derive from the environment/equipment and are highly adapted to the production area may provide the typical characteristics to cheeses (Micari et al., 2007). An optimal starter culture for typical cheeses should drive the fermentation process in an appropriate direction by inhibiting the undesired microorganisms to warrant the hygienic aspects, but do not alter the typical sensory profile of the traditional cheese (Settanni et al., 2013).

120

CHAPTER VI

“Vastedda della valle del Belìce” cheese was traditionally produced only during the summer season, but it is currently requested throughout the year. This results in cheeses characterized by marked differences among the seasons, especially between summer and winter productions, and the final quality is unpredictable. Regarding the microbiology of cheeses, it might be greatly affected by the different temperatures registered in Sicily between summer, when the environmental temperatures may reach 30 – 35°C, and winter with temperatures below 15°C. Since the modern systematic approach to minimising microbial variability and obtaining cheeses with the desired characteristics, consistently over time, is based on the use of starter cultures, they should be constituted by autochthonous microorganisms which can ensure the maintenance of the typicality (Gaglio et al., 2014). At this respect, the strain selection for the four-season production of “Vastedda della valle del Belìce” cheese, has to consider the heat resistance during stretching and the capacity to carry out the acidification at high temperatures during summer as well as at low temperatures during winter. With the aim to convert the production process for PDO “Vastedda della valle del Belìce” cheese from a production performed with raw milk without starters to a production carried out with raw milk and natural starters, during all four seasons of the year, the objectives of the present work were: to evaluate the in vivo acidifying ability of twelve strains of LAB; to evaluate their dynamics and to monitor their levels during the refrigerated storage of the experimental cheeses; and to select the strains providing the best cheeses in terms of sensory characteristics and volatile organic compounds.

121

CHAPTER VI

2. MATERIALS AND METHODS

2.1. Strains and growth conditions

Twelve LAB strains (Lactobacillus delbrueckii PON79, PON256 and PON405, Lactococcus lactis PON36, PON153 and PON203, Leuconostoc mesenteroides PON169, PON259 and PON559, Streptococcus thermophilus PON244, PON120 and PON242), isolated from PDO Vastedda della Valle del Belìce cheese samples and selected as technologically relevant in cheese making based on their acidification activity, generation of aromatic compounds and production of antimicrobial substances (Gaglio et al., 2014), from a total of 74 strains, were used in this study. The strains belonging to the culture collection of the of the Agricultural Microbiology laboratory of the Department of Agricultural and Forestry Science - University of Palermo (Palermo, Italy), were grown overnight as follows: lactobacilli in MRS broth incubated at 42°C, streptococci in M17 broth incubated at 44°C, lactococci and leuconostocs in M17 broth at 30°C. All media were purchased from Oxoid (Milan, Italy). Lc. lactis strains were genetically identified at subspecies level applying the PCR amplification of the gene acmA as described by Garde et al. (1999). The RAPD analysis tool reported by Pérez et al. (2002) was used to discriminate among the different subspecies of Ln. mesenteroides.

2.2. Cheese production

The LAB strains were first mixed together in different multi-strain inocula: all strains belonging to each species in triple combination (Lb, lactobacilli; Lc, lactococci; Ln, leunostocs; St, streptococci); all thermophilic strains (Lb-St, lactobacilli and streptococci); all mesophilic strains (Lc-Ln, lactococci and leuconostocs). The bacterial mixtures were prepared after the individual overnight growth of each strain in different conditions: process 1, after growth in the optimal synthetic medium, re-suspended in Ringer’s solution [overnight cultures were centrifuged at 5,000 × g for 5 min, washed twice in Ringer’s solution and re- suspended in the same solution till reaching an optical density (OD) at 600 nm of ca. 1.00 which approximately corresponds to a concentration of 109 CFU/mL] and inoculated in pasteurised (72°C for 15 s) ewes’ milk; process 2, after growth in whey-based medium (WBM), prepared as reported by Settanni et al. (2012a), and inoculated in pasteurised ewes’ milk; process 3, after growth in WBM and inoculated in raw ewes’ milk. The preparation of the cells for the processes 2 and 3 was the same as that reported for the process 1, but when

122

CHAPTER VI

the OD600 was approximately 1.00, the cells were centrifuged once again and re-suspended in WBM in place of Ringer’s solution. The inocula were added to a final concentration of approximately 107 CFU/mL in milk. The multi-strain inocula included all strains at the same final concentration. Control cheeses were produced without addition of starters: CP1, control for cheese making with pasteurised ewes’ milk and Ringer’s solution (process 1); CP2, control for cheese making with pasteurised ewes’ milk and WBM (process 2); CP3, control for cheese making with raw ewes’ milk and WBM (process 3) (Fig. 1).

A B

INOCULA plastic vats plastic vats INOCULA LAB grown in synthetic media and LAB grown in whey based medium re-suspended in Ringer’s Solution CP1 CP1

lactobacilli lactobacilli PON79 - PON256 - PON405 Lb PON79 - PON256 - PON405 Lb

streptococci streptococci PON120 - PON242 - PON244 St PON120 - PON242 - PON244 St

Pasteurised Pasteurised lactococci ewes’ milk lactococci ewes’ milk PON36 - PON153 - PON203 Lc PON36 - PON153 - PON203 Lc

leuconostocs leuconostocs Ln Ln PON169 - PON259 - PON559 PON169 - PON259 - PON559

streptococci streptococci PON120 - PON242 - PON244 PON120 - PON242 - PON244 + + lactobacilli Lb-St lactobacilli Lb-St PON79 - PON256 - PON405 PON79 - PON256 - PON405

lactococci lactococci PON36 - PON153 - PON203 PON36 - PON153 - PON203 + Lc-Ln + Lc-Ln leuconostocs leuconostocs PON169 - PON259 - PON559 PON169 - PON259 - PON559

C

plastic vats plastic vats INOCULA INOCULA LAB grown in whey-based medium LAB grown in whey-based medium

CP1

lactobacilli Lc. lactis subsp. cremoris PON36 PON79 - PON256 - PON405 Lb 36

streptococci Lc. lactis subsp. cremoris PON153 PON120 - PON242 - PON244 St 153

lactococci Raw ewes’ milk Lc. lactis subsp. cremoris PON203 PON36 - PON153 - PON203 Lc 203

leuconostocs Ln 169 Ln. mesenteroides subsp. mesenteroides PON169 PON169 - PON259 - PON559

streptococci PON120 - PON242 - PON244 + Ln. mesenteroides subsp. mesenteroides PON259 lactobacilli Lb-St 259 PON79 - PON256 - PON405

lactococci PON36 - PON153 - PON203 + Ln. mesenteroides subsp. mesenteroides PON559 leuconostocs Lc-Ln 559 PON169 - PON259 - PON559

Fig. 1. Experimental design of Vastedda-like cheese productions performed at pilot plant scale. A, process 1; B, process 2; C, process 3.

The experimental cheese making trials were carried out in a dairy pilot plant (Istituto Zooprofilattico Sperimentale della Sicilia “Adelmo Mirri”, Palermo, Italy) using the

123

CHAPTER VI

POLYFOOD mod. SI-050 (INVENTAGRITM, Modena, Italy). Each trial was performed with 10 L of milk; according to the experimental plan, milk (heated at 38°C) was inoculated with the corresponding bacterial mixture and added with 3 g of rennet paste (Clerici Sacco

International, Cadorago, Italy). After curdling, 1 L of H2O kept at 60°C was added to each trial during the curd cutting until obtaining small rice-size grains. The curds were put into perforated containers and the pH was monitored electrometrically using the portable pH meter Russell RL060P (Thermo Fisher Scientific, Beverly, MA) at 1 h intervals for the first 6 h and, subsequently, at 12 h intervals until the value into the range 5.20 – 5.40 was reached. Room temperature of experimental dairy factory where the acidification step took place was monitored by a 175-T2 data logger (Testo, Settimo Milanese, Italy). After acidification, the curds were stretched under hot (85°C) water and moulded to circular shape. The cheeses were salted in a brine containing 20% NaCl (w/v) for 30 min, air dried for 24 h and then kept refrigerated (at 7°C) under vacuum. The strains whose combinations showed the best results in terms of kinetics of curd acidification were then tested individually (as reported above) in cheese making (Table 1) applying the conditions of the process 3, after growth in WBM and inoculated in raw ewes’ milk. Cheese trials were carried out in duplicate in two consecutive weeks during February 2013. pH measurements were carried out in duplicate for each trial at each time.

2.3. Microbiological analyses

The curds were collected soon after transfer into the perforated containers and, then, before stretching (acidified curds). The final cheeses were analysed soon after stretching and after 15 days of refrigerated storage. The first dilution of each sample (10 g) was performed in sodium citrate (2% w/v) solution by homogenisation in a stomacher (BagMixer® 400, Interscience, Saint Nom, France) for 2 min at the highest speed. Further serial dilutions were continued in Ringer’s solution. Microbial suspensions were plated and incubated as follows: total mesophilic count (TMC) on plate count agar (PCA) with 1 g/L added skimmed milk (SkM), incubated aerobically at 30°C for 72 h; total psychrotrophic counts (TPC) on PCA-SkM, incubated aerobically at 7°C for 7 d; mesophilic and thermophilic rod LAB on MRS agar, acidified at pH 5.4 with lactic acid (5 mol/L), incubated anaerobically for 48 h at 30 and 44°C, respectively; mesophilic and thermophilic coccus LAB on M17 agar, incubated

124

CHAPTER VI

anaerobically for 48 h at 30 and 44°C, respectively. TPC were monitored only during cheese storage. Microbiological counts were carried out in duplicate.

Table 1. Changes of pH during experimental Vastedda-like cheese making.

Production Inocula pH curd at T0 Time of curd pH in the pH cheese at T0 pH cheese at 15 d processa range 5.2 – 5.4 (value)

1 CP1 6.43±0.01 120 h (5.20±0.02) 5.57±0.02A 5.58±0.01A 1 Lb 6.23±0.02 72 h (5.36±0.01) 5.48±0.01A 5.49±0.01A 1 St 6.27±0.00 120 h (5.39±0.01) 5.40±0.00A 5.51±0.02A 1 Lc 6.45±0.02 5 h (5.23±0.03) 5.54±0.02A 5.51±0.01A 1 Ln 6.48±0.00 5 h (5.24±0.01) 5.52±0.01A 5.51±0.02A 1 Lb-St 6.46±0.01 72 h (5.39±0.01) 5.49±0.01A 5.47±0.00A 1 Lc-Ln 6.46±0.02 5 h (5.19±0.01) 5.60±0.02A 5.44±0.01B 2 CP2 6.42±0.01 120 h (5.21±0.00) 5.49±0.01A 5.47±0.01A 2 Lb 6.08±0.01 72 h (5.40±0.00) 5.49±0.00A 5.48±0.02A 2 St 6.24±0.00 120 h (5.31±0.00) 5.50±0.01A 5.48±0.00A 2 Lc 6.14±0.02 3 h (5.26±0.01) 5.50±0.00A 5.38±0.02B 2 Ln 6.35±0.02 5 h (5.32±0.02) 5.38±0.01A 5.29±0.01B 2 Lb-St 6.23±0.03 48 h (5.26±0.00) 5.51±0.01A 5.52±0.00A 2 Lc-Ln 6.32±0.02 5 h (5.21±0.02) 5.53±0.00A 5.24±0.01B 3 CP3 6.48±0.00 72 h (5.35±0.00) 5.60±0.01A 5.57±0.01A 3 Lb 6.55±0.01 48 h (5.23±0.02) 5.50±0.02A 5.49±0.00A 3 St 6.42±0.02 72 h (5.29±0.01) 5.45±0.01A 5.43±0.02A 3 Lc 6.11±0.00 6 h (5.32±0.02) 5.50±0.00A 5.10±0.01B 3 Ln 6.35±0.01 6 h (5.35±0.00) 5.45±0.01A 5.19±0.00B 3 Lb-St 6.39±0.02 72 h (5.25±0.01) 5.48±0.02A 5.44±0.01A 3 Lc-Ln 6.37±0.00 6 h (5.28±0.01) 5.51±0.01A 5.17±0.00B 3 Lc. lactis subsp. cremoris PON36 6.66±0.01 5 h (5.29±0.01) 5.49±0.00A 5.29±0.02B 3 Lc. lactis subsp. cremoris PON153 6.67±0.00 24 h (5.28±0.01) 5.45±0.02A 5.30±0.00B 3 Lc. lactis subsp. cremoris PON203 6.65±0.02 5 h (5.33±0.01) 5.52±0.00A 5.32±0.01B 3 Ln. mesenteroides subsp. 6.43±0.00 6 h (5.37±0.00) 5.48±0.01A 5.33±0.00B mesenteroides PON169 3 Ln. mesenteroides subsp. 6.47±0.02 24 h (5.39±0.00) 5.49±0.00A 5.28±0.02B mesenteroides PON259 3 Ln. mesenteroides subsp. 6.41±0.00 24 h (5.33±0.02) 5.54±0.02A 5.31±0.00B mesenteroides PON559 a 1, growth of bacteria in the optimal synthetic media, re-suspension in Ringer’s solution and inoculation in pasteurised ewes’ milk; 2, growth of bacteria in WBM and direct inoculation in pasteurised ewes’ milk; 3, growth of bacteria in WBM and direct inoculation in raw ewes’ milk. Abbreviations: CP1, control for the production protocol 1; CP2, control for the production protocol 2; CR, control for the production protocol 3; Lb, lactobacilli; Lc, lactococci; Ln, leunostocs; St, streptococci; Lb-St, lactobacilli and streptococci; Lc-Ln, lactococci and leuconostocs. Results indicate mean values ± SD of four measurements (carried out in duplicate for two independent productions). Uppercase letters indicate different statistical significances (overall P < 0.05, Tukey's correction). Means within a given column with the same letter are not statistically different from each other.

2.4. Monitoring of the bacterial inocula

The presence of the microorganisms added as starter cultures in the cheese trials with individual inocula (Lactococcus lactis PON36, PON153 and PON203 and Leuconostoc mesenteroides PON169, PON259 and PON559) was confirmed, after colony isolation from the highest dilution of sample suspensions, by microscopic inspection and randomly amplified

125

CHAPTER VI

polymorphic DNA (RAPD) analysis. DNA from broth cultures was extracted by the Instagene Matrix kit (Bio-Rad, Hercules, CA) as described by the manufacturer. Crude cell extracts were used as template for PCR. RAPD-PCR was performed by means of T1 Thermocycler (Biometra, Göttingen, Germany) as reported by Settanni et al. (2012b). Amplified DNAs from the isolates of a given trial, together with that of the pure culture(s) corresponding to the same trial, were loaded onto a gel in order to recognize the inoculated bacteria.

2.5. Sensory analysis

The effect of the different bacterial inocula on the final characteristics of the cheeses was evaluated by sensory analysis on samples kept refrigerated for 15 d. A PDO Vastedda della Valle del Belìce cheese production conserved in the same conditions as the experimental cheeses was used for sensory comparison. The descriptive panel consisted of thirteen judges (seven females and six males, 28 – 52 years old) familiar with the sensory analysis of cheeses, but not specifically trained in the evaluation of Vastedda della Valle del Belìce cheese. The judges were asked to score fourteen descriptors regarding the aspect (colour, oil, presence of eyes and uniformity of structure), the smell (strength of odours, pasture and pungent odour), the taste (taste intensity, salt, bitter and spicy) and the consistency (soft/hard, saliva evoking and dispersion). The sensory analysis was conducted following the ISO 13299 (2003) indications. The panellists performed the analysis in individual chambers and had no specific information about the experimental design. All cheeses were administered in pieces of about 3 × 3 × 2 cm in size left at ambient temperature (ca. 20°C) for 60 min and presented in coded white plastic plates in a randomised order.

2.6. Analysis of cheese volatile organic compounds

The cheeses which reached the best scores in terms of appearance, smell, taste and consistency were analysed for their volatile organic compound (VOC) composition after 15 d of refrigerated storage. VOCs were determined using the headspace solid phase micro extraction (SPME) technique coupled with gas chromatography with mass spectrometric detection (GC/MS). The cheeses, frozen at -20°C, were manually grated and 10 g of each cheese were transferred into a vial, added with 10 mL H2O and 30 μL of internal standard solution [4-methyl-2-pentanone (4.14 g/L) and isobutyric acid (20 g/L) in H2O]. The vials,

126

CHAPTER VI

kept under magnetic stirring, were heated at 60°C until melting (Carlin and Versini, 2005) and the headspace was collected by a DBV-Carboxen-PDMS fibers (Supelco, Bellefonte, PA) for 30 min. The SPME fibre was inserted directly into a Finnegan Trace MS for GC/MS (Agilent 6890 Series GC system, Agilent 5973 Net Work Mass Selective Detector, Milan, Italy) equipped with a DB-WAX capillary column (Agilent Technologies, 30 m, 0.250 mm i.d., film thickness 0.25 μm, part n° 122-7032). The GC temperature was 40°C for 2 min (during splitless injection), from 40 to 60°C, 4°C/min, 60°C for 2 min, from 60 to 190°C, 2°C/min, from 190 to 230, 5°C/min, 230°C for 15 min; injector 250°C, Fid 250°C, transfer line 230°C, carrier helium 1 mL/min.; EM. 70 eV. Mass spectra were recorded by electronic impact (EI) at 70 eV using ion source temperatures at 200°C. The scan mode was used to detect all the compounds in the range m/z 33 – 495 atomic mass unit (amu). The individual peaks were identified by comparison of their retention indices to those of authentic samples, as well as by comparing their mass spectra with the NIST/EPA/NIH Mass Spectral Library database (Version 2.0d, build 2005). The results were expressed in milligrams per kilogram as 4- methyl-2-pentanone. All solvents and reagents were purchased from WWR International (Milan, Italy). Chemical and physical determinations were performed in triplicate and the results expressed as means ± standard deviation.

2.7. Induction of the lytic cycle

The search of the lysogenic state of the LAB showing the best results after sensory evaluation and VOC analysis was performed by the chemical induction of the lytic cycle, adapting the method of Cochran and Paul (1998). Overnight grown broth cultures were treated by adding mitomycin C (1 mg/mL). Controls were left untreated. After incubation for 24 h at room temperature in the dark, the presence of lytic phages was detected by the plaque assay modified (PAm) as described by Franciosi et al. (2009).

2.8. Summer production of Vastedda-like cheese

The strains showing the best performances during winter production were evaluated during summer to evaluate their ability to compete with the indigenous thermophilic LAB of raw milk. Summer production was carried out in duplicate between June and July 2013. In order to test the selected strains in “working” rather than “experimental” conditions, cheese making was performed at industrial level in the dairy factory “Il Cacio Siciliano” located in Belmonte

127

CHAPTER VI

Mezzagno (Palermo, Italy), applying the same conditions as (process 3) the winter production. Acidification of curds, refrigerated storage of cheeses, strain recognition and VOC analysis were conducted as reported above.

2.9. Statistical analyses

Data of acidification and volatile organic compound concentration were statistically analysed using the ANOVA procedure. Differences between means were determined by Tukey's multiple-range test. Significance level was P <0.05. Sensory evaluations were analysed using the generalized linear model (GLM) procedure. The discrimination efficiency of the attributes for each assessor was tested by a 2-factor analysis of variance (ANOVA), with judges (i=1..11) and experimental cheeses (j=1..27) as fixed factors. Least square means (LSM) were compared using T test (P < 0.05). All statistical analyses were conducted using the software SAS 2004, version 9.1.2 (Statistical Analysis System Institute Inc., Cary, NC, USA).

3. RESULTS AND DISCUSSION

3.1. Effect of bacterial inocula on the acidification kinetics of curds

In this study, we evaluated the dynamics of mesophilic and thermophilic LAB strains during winter production of Vastedda-like cheese, assuming that mesophilic strains are able to perform curd acidification during summer whereas thermophilic strains may not be able to carry out this process during winter, when the temperatures are too low. The cheese making was conducted at pilot-scale level in standard conditions in order to keep all process variables, except temperature, under control and to avoid the casual effects of the factory-scale productions (Settanni et al., 2011). The entire experimentation was performed in February, which is one of the coldest months in Sicily. The temperature was not kept controlled during curd acidification to mimic the conditions of the artisanal cheese factories. One of the medium used for LAB growth was prepared with whey, since LAB starters for the production of pasta-filata cheeses are commonly provided in form of natural whey starter cultures (NWSC) (Parente et al., 1998). However, the strains tested in this study were isolated from cheese and, before the inocula could be provided in form of whey starter cultures, their optimal development in whey had to be checked.

128

CHAPTER VI

The acidification of curds determined by the addition of the selected LAB was followed until the value was comprised in the range 5.20 – 5.40 (Table 1) which corresponds to the level of acidity allowing the stretching of the curd (Niro, 2011). At those pH values most of dicalcium-para-caseinate gets converted into mono-calcium paracaseinate which provides the strings and sheen to the cheese (Kosikowski, 1958). After genetic identification at subspecies level, all lactococci resulted to belong to Lc. lactis subsp. cremoris, while all leuconostocs were Ln. mesenteroides subsp. mesenteroides. Room temperatures registered during the acidification step were in the range 10.9 – 12.9°C during night (6 p.m. – 6 a.m.) and 13.8 – 16.5°C during the day (6 a.m. – 6 p.m.). Both curds of the control cheeses produced with pasteurised milk (CP1 and CP2) could be stretched after 5 d, while the curds of control cheese made with raw milk (CP3) reached a pH suitable for stretching at the third day of acidification. The triple inocula of lactobacilli acidified both pasteurised and raw milk before the triple inocula of streptococci. Generally lactobacilli and streptococci needed more time than leuconostocs and lactococci to acidify the curds. Lactobacilli determined decrease in pH value in the range 5.20 – 5.40 after 48 h in raw milk, while at least 72 h were necessary in pasteurised milk. Hence, the thermophilic combinations were not able to acidify the in the time requested by the disciplinary of PDO “Vastedda della valle del Belìce” cheese (GUE no. C 42/16 19.2.2010). Regarding mesophilic strains, lactococci were stronger acidifier than leuconostocs and the best results were registered after the application of process 2, for which the pH reached 5.26 at 3 h from milk curdling. Although the application of the process 3 reduced the time for the acidification by thermophilic strains, it delayed the drop of pH by lactococci and leuconostocs. This observation might be explained with the fact that the LAB, presumably mesophilic, present in raw milk overcame the thermophilic Lb, St and Lb-St inocula, but have undertaken a competition with Lc, Ln and Lc-Ln inocula. However, both mesophilic groups reached the wanted pH within 6 h from coagulation. The multiple strain/species combination showed results similar to those of the triple strain combinations of the single species, except Lc-Ln for the process 2 which reached 5.21 after 5 h, while Lc alone at 3 h. Due to the rapid acidification of curds determined by the mesophilic strains in combination, all lactococci and leuconostocs were tested singly applying the process 3 to register their behaviour in raw milk. Lc. lactis subsp. cremoris PON 153, Ln. mesenteroides subsp. mesenteroides PON259 and PON559 were not able to reach at least 5.40 within 6 h, but Lc. lactis subsp. cremoris PON 36 and PON 203 determined a final pH of 5.29 and 5.33, respectively at the 5th hour from curdling. As a matter of fact, Lc. lactis subsp. cremoris PON153, Ln. mesenteroides subsp. 129

CHAPTER VI

mesenteroides PON259 and PON559 showed an acidification kinetics compatible with the 24 h laid down by the PDO disciplinary, while the other mesophilic strains determined a too rapid process. However, a rapid acidification of the curd is important to prevent the growth of undesirable (spoilage/pathogenic) microorganisms. Immediately after stretching, all Vastedda cheeses showed a pH value ranging between 5.38 and 5.60. The increase of pH after the stretching phase is a common phenomenon due to the loss of acids during treatment with hot water (Mucchetti and Neviani, 2006). After 15 d of refrigerated storage, the only pH that decreased consistently (P < 0.05) were those determined by Lc, Ln and Lc-Ln inocula, showing a certain activity of the mesophilic combinations at low temperatures. The final pH registered in presence of Lc. lactis subsp. cremoris and Ln. mesenteroides subsp. mesenteroides inoculated singly were higher than those displayed by the multiple combinations, but their values were significantly (P < 0.05) different from those registered soon after production.

3.2. Microbial dynamics

Changes in the levels of concentration of TMC, mesophilic and thermophilic rod and coccus LAB during the acidification of curds and, subsequently, the refrigerated storage are shown in Table 2 and 3. In table 3, the evolution of TPC is also reported. All the three curds obtained from the control cheese productions (CP1, CP2 and CP3) were characterized by the highest counts on M17 agar, indicating the dominance of mesophilic coccus LAB, which concentrations were comparable with those of TMC (Table 2). All curds obtained from the milks inoculated with the termophilic LAB, rods or cocci, mono-species or multi-species combinations, did not show increases of their numbers during the acidification of any of the process considered. On the contrary, all curds started with mesophilic LAB showed an increase of concentrations of this group of at least 1 Log; the highest levels were registered for the combination Lc-Ln of the process 1 and 2 which reached 10.1 and 10.0 Log CFU/g, respectively, in 5 hours. The single inocula of Lc. lactis subsp. cremoris and Ln. mesenteroides subsp. mesenteroides were all characterized by the increase of the mesophilic LAB concentration, but Ln. mesenteroides subsp. mesenteroides PON259 increased this level until barely 8.2 Log CFU/g, whereas all other five strains determined a final level above 9.00 Log CFU/g. After stretching (Table 3), the levels of TPC ranged between 4.7 and 8.0 Log CFU/g. The lowest concentration of TPC was displayed by CP2, while the highest by Lc-Ln of the process

130

CHAPTER VI

2. The levels of TMC of the cheeses were due to the LAB inoculated. At the 15th day of refrigerated storage, only the cheeses started with mesophilic LAB showed the final levels of this group at high concentrations. Also control cheeses were characterized by high levels of mesophilic LAB which were almost 2 Log cycles higher than thermophilic LAB. The cheeses obtained with single inocula of mesophilic LAB were not all characterized by high numbers of mesophilic LAB after 15 days of storage: except Lc. lactis subsp. cremoris PON153 and Ln. mesenteroides subsp. mesenteroides PON259, which reached the levels of 9.0 and 8.5 Log CFU/g, respectively, showing a consistent increase of their concentrations, the other 4 mesophilic strains did not seem to have increased during refrigeration. Although the growth at sub-optimal temperatures results in a dramatic slow down of metabolism of L. lactis, strains of this species can develop at 4°C (Panoff et al., 1994). The slowed metabolism of the mesophilic strains may play a relevant role to impact the sensory profile of cheeses that are kept refrigerated after production and are not subjected to ripening.

131

Table 2. Microbial evolution during curd acidification.

Production Inocula Sample a protocol Curd at T0 Curd before stretching PCA-SkM 30°C M17 30°C MRS 42°C M17 44°C PCA-SkM 30°C M17 30°C MRS 42°C M17 44°C

1 CP1 4.0±0.4 3.7±0.5 2.0±0.3 2.5±0.1 7.5±0.4 7.5±0.5 5.7±0.1 5.7±0.4 1 Lb 7.0±0.2 n.d. 7.9±0.2 n.d. 7.9±0.5 n.d. 7.6±0.5 n.d. 1 St 7.5±0.6 n.d. n.d. 7.4±0.4 7.7±0.3 n.d. n.d. 7.8±0.2 1 Lc 8.0±0.4 8.5±0.5 n.d. n.d. 8.7±0.2 9.7±0.1 n.d. n.d. 1 Ln 7.9±0.1 8.6±0.3 n.d. n.d. 8.9±0.5 9.9±0.2 n.d. n.d. 1 Lb-St 7.3±0.5 n.d. 7.5±0.4 7.2±0.1 7.4±0.4 n.d. 7.8± 7.6±0.1 1 Lc-Ln 8.5±0.1 9.1±0.3 n.d. n.d. 8.9±0.2 10.1±0.1 n.d. n.d. 2 CP2 4.4±0.4 4.1±0.3 2.5±0.2 2.5±0.3 7.5±0.3 7.4±0.2 5.3±0.2 6.0±0.3 2 Lb 6.9±0.6 n.d. 7.7±0.3 n.d. 7.6±0.3 n.d. 7.4±0.1 n.d. 2 St 6.3±0.3 n.d. n.d. 7.7± 7.7±0.2 n.d. n.d. 7.3±0.6

2 Lc 8.6±0.3 8.4±0.4 n.d. n.d. 9.8±0.4 9.7±0.2 n.d. n.d. CHAPTER 2 Ln 8.2±0.1 8.1±0.2 n.d. n.d. 9.5±0.3 9.7±0.4 n.d. n.d. 2 Lb-St 6.9±0.4 n.d. 7.2±0.4 7.0±0.5 7.5±0.2 n.d. 7.7±0.4 7.9±0.6 2 Lc-Ln 8.4±0.3 8.7±0.2 n.d. n.d. 8.9±0.3 10.0±0.1 n.d. n.d. 3 CP3 5.7±0.2 6.0±0.4 3.2±0.6 4.9±0.5 7.3±0.4 7.6±0.1 4.6±0.5 5.3±0.4

VI 3 Lb 7.1±0.3 n.d. 7.6±0.2 n.d. 7.5±0.1 n.d. 7.9±0.1 n.d.

3 St 7.3±0.5 n.d. n.d. 7.3±0.6 7.6±0.3 n.d. n.d. 7.9±0.1 3 Lc 7.5±0.6 7.9±0.2 n.d. n.d. 9.0±0.4 9.8±0.1 n.d. n.d. 3 Ln 7.6±0.4 8.1±0.1 n.d. n.d. 8.9±0.1 9.5±0.1 n.d. n.d. 3 Lb-St 7.7±0.1 n.d. 7.7±0.2 7.3±0.5 7.6±0.4 n.d. 7.5±0.4 7.7±0.3 3 Lc-Ln 7.7±0.3 8.3±0.1 n.d. n.d. 9.1±0.1 9.7±0.4 n.d. n.d. 3 Lc. lactis subsp. cremoris PON36 7.5±0.6 8.1±0.4 n.d. n.d. 9.0±0.1 9.5±0.2 n.d. n.d. 3 Lc. lactis subsp. cremoris PON153 7.8±0.4 7.9±0.2 n.d. n.d. 8.7±0.3 9.4±0.5 n.d. n.d. 3 Lc. lactis subsp. cremoris PON203 7.3±0.6 7.9±0.1 n.d. n.d. 8.6±0.5 9.3±0.4 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON169 7.5±0.1 7.8±0.3 n.d. n.d. 8.5±0.2 9.5±0.2 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON259 7.6±0.4 7.5±0.1 n.d. n.d. 7.8±0.6 8.2±0.1 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON559 7.5±0.2 7.2±0.4 n.d. n.d. 8.6±0.4 9.0±0.3 n.d. n.d. a 1, growth of bacteria in the optimal synthetic media, re-suspension in Ringer’s solution and inoculation in pasteurised ewes’ milk; 2, growth of bacteria in WBM and direct inoculation in pasteurised ewes’ milk; 3, growth of bacteria in WBM and direct inoculation in raw ewes’ milk. Abbreviations: CP1, control for the production protocol 1; CP2, control for the production protocol 2; CP3, control for the production protocol 3; Lb, lactobacilli; Lc, lactococci; Ln, leunostocs; St, streptococci; Lb-St, lactobacilli and streptococci; Lc-Ln, lactococci and leuconostocs. Results indicate mean values ± SD of four plate counts (carried out in duplicate for two independent productions). 132

Table 3. Microbial evolution during refrigerated Vastedda-like cheese storage.

Production Inocula Sample a protocol Cheese at T0 Cheese at 15 d PCA-SkM PCA-SkM M17 MRS M17 PCA M17 MRS M17 PCA 7°C 7°C 30°C 30°C 42°C 44°C 30°C 30°C 42°C 44°C 1 CP1 6.0±0.1 7.9±0.5 7.9±0.1 7.0±0.3 7.3±0.2 7.9±0.5 7.6±0.1 8.5±0.5 5.0±0.2 7.4±0.1 1 Lb 6.2±0.3 7.6±0.2 n.d. 6.9±0.3 n.d. 8.7±0.3 8.0±0.2 n.d. 7.0±0.3 n.d. 1 St 6.7±0.1 7.3±0.2 n.d. n.d. 7.5±0.5 7.9±0.4 8.0±0.4 n.d. n.d. 6.2±0.3 1 Lc 6.7±0.5 8.6±0.4 8.6±0.1 n.d. n.d. 7.5±0.1 8.7±0.2 8.6±0.4 n.d. n.d. 1 Ln 7.1±0.4 8.6±0.2 8.5±0.5 n.d. n.d. 7.2±0.3 8.7±0.5 8.7±0.3 n.d. n.d. 1 Lb-St 6.0±0.2 7.0±0.2 n.d. 7.6±0.4 7.7±0.3 6.5±0.1 9.0±0.1 n.d. 6.5±0.5 7.0±0.1 1 Lc-Ln 7.4±0.5 7.9±0.1 8.1±0.5 n.d. n.d. 8.0±0.3 9.0±0.2 9.4±0.1 n.d. n.d. 2 CP2 4.7±0.3 8.9±0.3 8.9±0.3 6.2±0.3 6.3±0.5 8.0±0.4 8.3±0.3 9.0±0.5 5.7±0.1 6.5±0.4 2 Lb 6.7±0.1 6.5±0.4 n.d. 6.7±0.3 n.d. 7.0±0.1 7.7±0.2 n.d. 7.8±0.3 n.d. 2 St 6.4±0.2 6.5±0.1 n.d. n.d. 6.6±0.5 6.7±0.3 7.0±0.1 n.d. n.d. 6.7±0.2

2 Lc 7.0±0.1 9.1±0.5 8.8±0.1 n.d. n.d. 7.7±0.1 9.4±0.3 9.4±0.4 n.d. n.d. CHAPTER 2 Ln 7.0±0.4 9.1±0.1 8.9±0.3 n.d. n.d. 7.8±0.4 9.0±0.2 9.0±0.5 n.d. n.d. 2 Lb-St 6.7±0.2 7.1±0.1 n.d. 6.9±0.3 6.7±0.1 7.8±0.2 7.9±0.3 n.d. 7.0±0.4 6.9±0.3 2 Lc-Ln 8.0±0.3 10.0±0.4 8.9±0.1 n.d. n.d. 8.3±0.3 8.4±0.5 8.3±0.2 n.d. n.d. 3 CP3 6.4±0.1 8.0±0.3 7.7±0.2 4.7±0.3 6.9±0.2 7.5±0.1 7.8±0.4 7.9±0.4 7.0±0.1 7.5±0.3

VI 3 Lb 6.1±0.4 7.3±0.3 n.d. 7.1±0.4 n.d. 7.7±0.5 8.2±0.4 n.d. 7.2±0.1 n.d.

3 St 6.2±0.5 7.7±0.4 n.d. n.d. 7.5±0.1 7.2±0.3 8.5±0.3 n.d. n.d. 6.3±0.4 3 Lc 5.7±0.3 8.3±0.2 8.4±0.3 n.d. n.d. 7.8±0.2 8.5±0.1 9.0±0.2 n.d. n.d. 3 Ln 6.3±0.2 8.1±0.1 8.2±0.2 n.d. n.d. 7.5±0.3 8.5±0.2 8.7±0.2 n.d. n.d. 3 Lb-St 6.6±0.4 7.4±0.5 n.d. 7.1±0.3 7.2±0.4 7.0±0.1 8.2±0.4 n.d. 6.4±0.3 6.7±0.1 3 Lc-Ln 6.9±0.5 8.1±0.2 8.0±0.3 n.d. n.d. 7.9±0.4 8.3±0.3 8.6±0.5 n.d. n.d. 3 Lc. lactis subsp. cremoris PON36 6.0±0.1 6.2±0.2 6.0±0.1 n.d. n.d. 6.5±0.4 6.0±0.1 5.7±0.3 n.d. n.d. 3 Lc. lactis subsp. cremoris PON153 6.0±0.2 8.2±0.4 8.0±0.2 n.d. n.d. 8.6±0.2 8.4±0.2 9.0±0.1 n.d. n.d. 3 Lc. lactis subsp. cremoris PON203 5.9±0.2 8.4±0.1 7.5±0.2 n.d. n.d. 7.4±0.3 7.8±0.4 7.6±0.2 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON169 6.0±0.4 6.3±0.1 5.7±0.4 n.d. n.d. 7.2±0.4 7.2±0.3 6.8±0.2 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON259 7.7±0.2 7.1±0.5 8.1±0.1 n.d. n.d. 8.6±0.3 8.4±0.1 8.5±0.4 n.d. n.d. 3 Ln. mesenteroides subsp. mesenteroides PON559 6.5±0.1 8.6±0.4 8.3±0.3 n.d. n.d. 8.2±0.4 8.5±0.3 8.6±0.1 n.d. n.d. a 1, growth of bacteria in the optimal synthetic media, re-suspension in Ringer’s solution and inoculation in pasteurised ewes’ milk; 2, growth of bacteria in WBM and direct inoculation in pasteurised ewes’ milk; 3, growth of bacteria in WBM and direct inoculation in raw ewes’ milk. Abbreviations: CP1, control for the production protocol 1; CP2, control for the production protocol 2; CP3, control for the production protocol 3; Lb, lactobacilli; Lc, lactococci; Ln, leunostocs; St, streptococci; Lb-St, lactobacilli and streptococci; Lc-Ln, lactococci and leuconostocs. Results indicate mean values ± SD of four plate counts (carried out in duplicate for two independent productions). 133

CHAPTER VI

3.3. Strain recognition

The isolates collected from the highest dilutions of samples, were analysed at the strain level by means of RAPD-PCR with primer M13 in order to monitor the dynamics of the added strains. The direct comparison of the RAPD patterns (Fig. 2) allowed the recognition of the cultures in all 24 trials inoculated with selected LAB alone or in combination.

Lb St Lc Ln Lb-St Lc-Ln

M

PON405 PON120 PON244 PON36 PON169 PON256 PON120 PON242 PON36 PON559 PON79 PON256 PON242 PON153 PON203 PON259 PON559 PON79 PON405 PON244 PON153 PON203 PON169 PON259

A

7.7 7.4 7.9 7.1 7.4 7.0 7.0 7.5 7.3 7.1 7.6 7.1 7.3 7.1 7.5 6.8 7.2 6.9 7.8 8.1 7.5 7.6 7.8 7.3 Curd T0 Acidified curd 7.1 6.7 7.6 7.2 7.8 7.7 8.9 9.7 9.1 8.8 9.9 9.1 7.3 7.2 7.8 7.0 7.6 7.1 9.5 10.1 9.8 9.0 9.3 9.1

. Lb St Lc Ln Lb-St Lc-Ln

.

M

PON79 PON242 PON153 PON559 PON405 PON153 PON259 PON256 PON405 PON120 PON244 PON36 PON203 PON169 PON259 PON79 PON256 PON120 PON242 PON244 PON36 PON203 PON169 PON559

B

7.3 7.5 7.7 7.3 7.7 7.4 7.1 7.4 7.3 7.1 7.4 7.3 6.8 6.8 7.2 6.8 7.0 6.7 7.1 7.7 7.2 7.0 6.9 6.9 Curd T0 Acidified curd 7.3 7.1 7.4 7.1 7.3 7.4 8.5 9.7 8.9 8.5 9.7 8.9 7.1 7.2 7.7 7.1 7.9 6.9 8.9 10.0 9.1 8.5 7.8 8.1

Lb St Lc Ln Lb-St Lc-Ln Lc Lc Lc Ln Ln Ln

M

PON405 PON120 PON244 PON36 PON169 PON256 PON120 PON242 PON36 PON559 PON79 PON256 PON242 PON153 PON203 PON259 PON559 PON79 PON405 PON244 PON153 PON203 PON169 PON259

PON36 PON153 PON169 PON259 PON559 PON203

C

7.0 7.1 7.6 7.0 7.3 7.0 6.7 6.9 6.5 7.0 7.1 6.8 6.8 7.0 7.7 6.9 7.3 6.9 6.8 7.3 7.1 6.8 7.0 7.0 7.1 Curd T0 7.9 7.9 7.8 7.5 7.2 Acidified curd 6.9 7.4 7.9 7.1 7.9 7.2 8.9 9.8 9.1 8.5 9.5 8.5 7.1 7.0 7.5 7.1 7.7 7.0 8.5 9.7 8.4 8.4 8.9 8.3 9.5 9.4 9.3 9.5 8.2 9.0 Fig. 2. Bacterial recognition during experimental Vastedda-like cheese productions by means of RAPD-PCR analysis with primer M13. A, process 1; B, process 2; C, process 3. Abbreviations: Lb, lactobacilli; St, streptococci; Lc, lactococci; Ln, leuconostocs; M, marker. The maximum level of detection (Log CFU/g) of the single inocula in curds is reported before and after acidification.

134

CHAPTER VI

This approach allowed to assess the evolution of the added LAB. In the multi-strain combinations, it allowed to establish the dominance of Lc. lactis subsp. cremoris PON153 and Ln. mesenteroides subsp. mesenteroides PON259 over the other mesophilic strains and revealed that the thermophilic combinations were not able to develop at dominant levels during the acidification step, but found them still viable after 15 days at 7°C (results not shown). The RAPD profiles of the LAB isolated at the highest concentrations from the control curds and cheeses excluded the presence of any of the 12 LAB used as starter in raw and pasteurized milk (results not shown). RAPD analysis is currently applied to monitor starter cultures at strain level during food production.

3.4. Sensory evaluation

The results of the sensory evaluation carried out by the judges on the 27 experimental cheeses and a PDO Vastedda della Valle del Belìce cheese, in duplicate, are reported in Table 4. The addition of cheeses processed with the traditional PDO protocol is fundamental for the evaluation of the final characteristics imparted by the different strains (Settanni et al., 2013). Except the uniformity of structure, which was not significantly different among judges, all other sensory attributes were different both for judges and cheeses. The less notable differences were evidenced by pasture and pungent odour among judges and strength of odours among cheeses. Control experimental cheeses, except CP3 regarding the presence of eyes, showed scores different from that of the PDO cheese for all other attributes. On the contrary, PDO cheese showed the same scores as the cheeses obtained applying the protocol 3 with lactococci for oil, eyes, strength of odours, spicy, pungent odour, salt, saliva evoking and dispersion and the same scores as the cheeses obtained with the same process inoculated with leuconostocs for colour, pasture and pungent odour and soft/hard consistency. All other attributes of PDO cheese were almost comparable with the results registered for the experimental cheeses of the process 3 inoculated with the mesophilic strains, but the results showed by lactococci were most superimposable than those of leuconostocs, on average.

3.5. Volatile organic compound composition

Based on the above results, only the experimental cheeses inoculated with lactococci were compared with PDO cheese regarding the VOCs. The results from chromatographic analysis of the cheeses are reported in Table 5. In the headspace of the cheeses, 18 compounds were

135

CHAPTER VI

identified: 7 acids, 5 alcohols, 3 aldehydes and 3 esters. Control cheese CP3 was characterized by the lowest number of molecules registered, whereas only PDO cheese showed the presence of 18 compounds. The strain Lc. lactis subsp. cremoris PON153, which determined the highest number of volatile compounds among the experimental cheeses, generated a massive increase in the concentrations of all acids and shared a similar VOC profile with the PDO cheese. In particular, heptanoic acid, ethyl esanoate and ethyl octanoate were detected in PDO cheese and the cheese inoculated with Lc. lactis subsp. cremoris PON153. Ethyl decanoate, was detected also in cheese inoculated with Lc. lactis subsp. cremoris PON203 and with the triple Lc combination, but it was at higher concentration in PDO cheese and the cheese processed with Lc. lactis subsp. cremoris PON153. The last strain determined also the highest level of 2,3-butanediol, but, in general, high concentrations of alcohols were due to the Lc combination. The highest production of aldehydes was registered for the cheese inoculated with Lc. lactis subsp. cremoris PON203.

136

Table 4. Sensory characteristics of experimental Vastedda-like cheeses.

Production Cheese samples Attributes protocola Colour Oil Eyes Uniformity Strenght of Pasture Pungent Taste Salt Bitter Spicy Soft/hard Saliva Dispersion odours odour intensity evoking PDO Market cheese 3.46 1.08 1.62 1.46 2.31 1.15 1.23 1.77 1.69 1.31 1.15 3.92 1.31 3.08 1 CP1 3.62 2.54 3.00 3.77 2.08 1.00 1.92 2.69 1.92 3.77 1.31 3.46 1.85 2.85 1 Lb 4.23 1.54 2.69 3.77 2.23 1.00 1.62 2.00 1.85 3.00 1.00 3.38 1.23 2.77 1 St 3.92 1.77 2.08 1.92 1.54 1.00 2.08 3.08 2.54 2.54 1.00 4.00 2.08 2.77 1 Lc 3.15 1.00 1.31 1.31 1.46 1.15 1.46 1.46 1.62 1.62 1.00 3.77 1.77 2.38 1 Ln 3.15 1.00 1.80 1.80 1.69 1.00 1.54 2.15 1.15 2.00 1.15 3.85 1.85 3.00 1 Lb-St 4.23 1.46 2.23 2.85 2.15 1.00 1.62 1.92 2.00 1.85 1.00 3.00 1.38 2.15 1 Lc-Ln 3.15 1.08 1.92 1.90 1.46 1.00 1.15 1.92 1.23 2.15 1.00 3.85 1.62 3.00 2 CP2 4.08 1.38 2.08 2.85 2.23 1.00 1.38 2.08 1.62 1.92 1.77 3.15 1.85 2.77 2 Lb 3.77 1.77 2.08 4.69 1.92 1.00 1.92 2.77 1.62 4.00 1.00 4.08 1.54 2.92

2 St 3.46 1.54 1.46 1.92 1.85 1.00 1.69 3.00 2.85 2.08 1.38 3.92 2.15 2.85 CHAPTER 2 Lc 4.54 1.08 1.31 1.62 2.23 1.62 1.69 2.31 1.46 2.62 1.08 3.00 1.92 2.77 2 Ln 4.54 1.00 1.77 1.92 1.92 1.15 1.38 1.92 1.46 1.46 1.00 2.85 1.46 2.38 2 Lb-St 4.38 1.23 1.77 1.92 1.46 1.15 2.38 2.77 1.77 1.92 1.15 3.08 2.00 2.85 2 Lc-Ln 3.15 1.00 2.54 2.54 1.69 1.00 1.62 2.08 1.23 1.85 1.08 4.00 1.62 3.00

VI 3 CP3 3.47 1.00 1.62 1.62 2.62 1.08 1.85 1.31 1.62 1.54 1.08 4.15 1.85 3.00

3 Lb 3.77 1.15 2.54 3.77 1.92 1.00 1.15 1.38 1.62 1.62 1.15 3.00 1.38 3.00 3 St 3.77 1.00 1.92 1.92 2.00 1.00 1.38 1.38 1.38 1.08 1.08 4.15 1.08 3.00 3 Lc 3.31 1.38 1.31 1.31 2.15 1.00 1.00 1.62 2.00 1.54 1.15 3.15 1.54 3.62 3 Ln 3.77 1.23 1.86 1.31 1.92 1.00 1.08 1.77 1.38 1.85 1.62 3.38 1.46 2.85 3 Lb-St 3.31 1.00 1.00 1.00 2.15 1.00 1.77 1.62 1.23 1.31 1.15 3.92 1.77 3.15 3 Lc-Ln 3.31 1.00 3.62 2.08 2.08 1.08 1.62 2.23 1.08 1.77 1.00 3.77 1.38 3.00 3 Lc. lactis subsp. cremoris PON36 3.77 1.08 1.62 1.92 2.00 1.08 1.00 1.62 1.15 1.46 1.23 3.77 1.00 3.08 3 Lc. lactis subsp. cremoris PON153 3.77 1.00 1.92 1.92 2.31 1.00 1.23 1.54 1.69 1.00 1.23 3.31 1.46 2.54 3 Lc. lactis subsp. cremoris PON203 4.08 1.00 1.31 1.31 2.23 1.00 1.15 1.46 1.38 1.46 1.15 3.23 1.31 2.85 3 Ln. mesenteroides subsp. 3.46 1.00 1.81 2.54 1.77 1.15 1.00 1.08 1.00 1.46 1.00 3.54 1.23 2.85 mesenteroides PON169

137

3 Ln. mesenteroides subsp. 3.92 1.00 1.92 1.92 2.15 1.08 1.46 1.00 1.38 1.15 1.08 3.54 1.00 2.85 mesenteroides PON259 3 Ln. mesenteroides subsp. 3.92 1.00 1.91 1.62 2.46 1.08 1.23 1.08 1.08 1.38 1.00 3.92 1.15 2.62 mesenteroides PON559 SEM 0.23 0.16 0.38 0.43 0.24 0.08 0.25 0.28 0.22 0.28 0.14 0.22 0.18 0.16 Statistical significance: Judges *** *** ** NS *** * * *** *** *** *** *** *** ** Cheeses *** *** *** *** * *** ** *** *** *** ** *** *** *** a 1, growth of bacteria in the optimal synthetic media, re-suspension in Ringer’s solution and inoculation in pasteurised ewes’ milk; 2, growth of bacteria in WBM and direct inoculation in pasteurised ewes’ milk; 3, growth of bacteria in WBM and direct inoculation in raw ewes’ milk. Abbreviations: CP1, control for the production protocol 1; CP2, control for the production protocol 2; CP3, control for the production protocol 3; Lb, lactobacilli; Lc, lactococci; Ln, leunostocs; St, streptococci; Lb-St, lactobacilli and streptococci; Lc-Ln, lactococci and leuconostocs; LSM, least square means; SEM, standard error of means. Results indicate mean values. Values reported in bold were identical for PDO cheese and experimental cheeses processed with mono-culture inocula. P value: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ns = not significant.

CHAPTER CHAPTER

VI

138

Table 5. Analysis of volatile organic compounds emitted from 15-d refrigerate stored experimental Vastedda-like cheeses.

Chemical Cheeses compoundsa Winter production Summer production PDO CP3 PON36 PON153 PON203 Lc PDO CP3 PON36 PON153 PON203 Lc Isoamyl acohol 0.50 ± 0.05A 0.14 ± 0.02B 0.71 ± 0.04B 0.55 ± 0.04A 1.51 ± 0.07B 1.89 ± 0.07B 0.54 ± 0.08A 1.05 ± 0.11B 1.43 ± 0.19B 0.97 ± 0.09B 1.17 ± 0.10B 1.2 6± 0.21B Ethyl esanoate 0.47 ± 0.08A n.d. n.d. 0.67 ± 0.04A n.d. n.d. 0.52 ± 0.08A n.d. n.d. 0.29 ± 0.03B n.d. n.d. 1-Pentanol 0.05 ± 0.02A 0.04 ± 0.01A 0.08 ± 0.02A 0.03 ± 0.01A 0.07 ± 0.01A 0.06 ± 0.01A 0.08 ± 0.01A n.d. n.d. n.d. n.d. n.d. Acetoin 0.45 ± 0.02A 0.14 ± 0.02B 0.31 ± 0.02B 0.36 ± 0.02B 0.60 ± 0.04B 0.05 ± 0.01B 0.14 ± 0.02A n.d. n.d. 0.06 ± 0.01B n.d. n.d. 1-Hexanol 0.02 ± 0.01A n.d. 0.06 ± 0.01B n.d. 0.02 ± 0.01A 0.05 ± 0.01A 0.09 ± 0.01A n.d. n.d. 0.01 ± 0B n.d. n.d. 2-Nonanone n.d. n.d. n.d. n.d. n.d. n.d. 0.21 ± 0.02A 0.02 ± 0B n.d. 0.12 ± 0.01B 0.01 ± 0.01B 0.03 ± 0.01B Ethyl octanoate 0.39 ± 0.04A n.d. n.d. 0.51 ± 0.04B n.d. n.d. 0.46 ± 0.08A n.d. n.d. 0.06 ± 0.01B n.d. n.d. Acetic acid 4.44 ± 0.11A 0.65 ± 0.04B 0.49 ± 0.04B 4.26 ± 0.09A 0.09 ± 0.02B 1.90 ± 0.07B 1.38 ± 0.13A 0.17 ± 0.02B 0.16 ± 0.01B 0.73 ± 0.05B 0.17 ± 0.02B 0.33 ± 0.03B Benzaldehyde 0.07 ± 0.02A n.d. 0.10 ± 0.02A 0.03 ± 0.01B 0.05 ± 0.01A 0.06 ± 0.01B 0.06 ± 0.01A 0.01 ± 0B 0.02 ± 0B 0.01 ± 0B 0.01 ± 0B 0.02 ± 0.01B 2,3-Butanediol isomer 4.47 ± 0.05A 1.07 ± 0.05B n.d. 5.44 ± 0.11B 0.01 ± 0.01B 3.79 ± 0.10B 0.07 ± 0.02A 0.08 ± 0.02A 0.11 ± 0.02A 0.07 ± 0.03A 0.08 ± 0.01A 0.14 ± 0.03B Phenylacetaldehyde 0.07 ± 0.02A n.d. 0.06 ± 0.01A n.d. 0.05 ± 0.01A n.d. n.d. n.d. n.d. 0.01 ± 0 n.d. n.d. Butyric acid 4.51 ± 0.11A 0.58 ± 0.04B 1.49 ± 0.07B 5.21 ± 0.11B 0.30 ± 0.04B 1.32 ± 0.07B 4.40 ± 0.21A 0.29 ± 0.06B 0.59 ± 0.03B 2.35 ± 0.04B 0.31 ± 0.10B 2.25 ± 0.02B

CHAPTER CHAPTER Ethyl decanoate 0.60 ± 0.04A n.d. n.d. 0.57 ± 0.04A 0.01 ± 0.01B 0.05 ± 0.01B 0.54 ± 0.04A n.d. n.d. n.d. n.d. 0.01 ± 0B 17.57 ± 22.59 ± 11.75 ± 11.00 ± Hexanoic acid 0.28A 1.95 ± 0.07B 5.75 ± 0.11B 19.67 ± 0.19B 1.12 ± 0.06B 3.41 ± 0.09B 0.15A 1.87 ± 0.10B 4.66 ± 0.09B 0.04B 1.67 ± 0.13B 0.07B 2-Phenylethanol 0.14 ± 0.03A 0.02 ± 0.01B 0.05 ± 0.01B 0.10 ± 0.01A 0.16 ± 0.02A 0.28 ± 0.02B 0.07 ± 0.01A 0.09 ± 0.02A 0.07 ± 0.01A 0.05 ± 0.01A 0.11 ± 0.02A 0.11 ± 0.01A Heptanoic acid 0.27 ± 0.03A n.d. n.d. 0.31 ± 0.02A n.d. n.d. 0.26 ± 0.03A n.d. n.d. n.d. n.d. n.d.

15.18 ± 10.44 ± VI Octanoic acid 0.31A 0.53 ± 0.04B 1.78 ± 0.07B 23.19 ± 0.21B 0.29 ± 0.02B 0.84 ± 0.07B 0.98A 0.40 ± 0.12B 0.15 ± 0.01B 5.17 ± 0.06B 0.17 ± 0.02B 0.32 ± 0.08B

Nonanoic acid 0.12 ± 0.02A 0.03 ± 0.01B 0.06 ± 0.01B 0.22 ± 0.02B 0.01 ± 0.01B 0.05 ± 0.01B 0.09 ± 0.01A 0.02 ± 0B 0.01 ± 0B 0.01 ± 0B 0.01 ± 0B 0.02 ± 0.01B Decanoic acid 7.58 ± 0.18A 0.07 ± 0.01B 0.42 ± 0.04B 9.38 ± 0.12B 0.08 ± 0.01B 0.11 ± 0.02B 2.75 ± 0.01A 0.16 ± 0.04B 0.11 ± 0.02B 1.13 ± 0.03B 0.02 ± 0B 0.03 ± 0.01B a The chemicals are shown following their retention time. Abbreviations: PDO, production denomination of origin; CP3, control for the production protocol 3; Lc. lactis subsp. cremoris PON36; Lc. lactis subsp. cremoris PON153; Lc. lactis subsp. cremoris PON203. Results indicate mean values of three measurements and are expressed (in mg/kg) as 4-methyl-2-pentanone. n.d., not detected. Uppercase letters indicate different statistical significances (overall P < 0.05, Tukey's correction). Means between columns of a given season with the same letter are not statistically different.

139

CHAPTER VI

3.6. Induction of the lytic cycle

On the basis of the results obtained by the winter productions, lactococci were selected as starters for Vastedda-like cheese production. However, lysogenic strains cannot be proposed as starters during cheese making, since accidental and spontaneous induction of the lytic cycle leads to production failure (Franciosi et al., 2009). Thus, all three lactococci were subjected to a chemical induction of the lytic cycle. The results of PAm (not shown) excluded the presence of lysogenic bacteriophages in Lc. lactis subsp. cremoris PON36, PON153 and PON203.

3.7. Summer production of Vastedda cheese

On the basis of the results obtained by the winter productions, lactococci were selected for the four-season production of Vastedda-like cheese. For this reason, the Lactococcus strains needed to be tested during summer, when their dominance during process can be compromised by the indigenous thermophilic LAB of raw milk which may find favourable conditions for development. All lactococci in single and multiple combinations, after growth in whey, were used to produce Vastedda-like cheese with raw milk during the last week of June and the first week of July at industrial level to test the performances of the strains outside the experimental cheese factory. Room temperature was, on average 23.5°C during night and 27.1°C during the day. RAPD analysis (results not shown) allowed to recognise each Lc. lactis in the corresponding single inocula, but only Lc. lactis subsp. cremoris PON153 was found at dominant levels after acidification, stretching of curd and 15-d of refrigerated storage of the cheese obtained with the mixed strain starter. Furthermore, VOC analysis of the resulting cheeses (Table 5) showed that also for the summer production the cheese processed with Lc. lactis subsp. cremoris PON153 showed the VOC profile most similar to that of the PDO cheese.

4. CONCLUSIONS

The uncontrolled evolution of microorganisms may lead to variable results in cheese production (Settanni and Moschetti, 2010). The selection of strains with given technological characteristics is important not only to drive the fermentation process, but also to maintain a certain typicality of traditional cheeses (Settanni et al., 2013). In order to set up a pool of LAB strains to be used for the four-season production of Vastedda-like cheese, the dynamics of twelve strains isolated from PDO cheeses and belonging to Lb. delbrueckii, Lc. lactis subsp.

140

CHAPTER VI

cremoris, Ln. mesenteroides subsp. mesenteroides and S. thermophilus, which displayed a dairy potential in vitro (Gaglio et al., 2014), were evaluated at pilot plant scale under controlled conditions in different single or multiple inocula. On the basis of the results shown for the winter and summer productions, at pilot scale and industrial level, respectively, and combining VOC and sensory evaluation, the multi-strain combination of lactococci was selected to act as starter preparation for the four-season production of Vastedda-like cheese. Even though the strains Lc. lactis subsp. cremoris PON36, PON153 and PON203 did not contain lysogenic phages, studies are being prepared to test their resistance to the most common dairy phages and to evaluate their performances in the several dairy factories producing PDO Vastedda della valle del Belìce cheese, which are gathered into a consortium for the protection of this traditional cheese production.

141

CHAPTER VI

REFERENCES

Carlin, S., Versini G. (2005) La caratterizzazione dei formaggi trentini attraverso la frazione volatile. Caratterizzazione di formaggi tipici dell’arco alpino: Il contributo della ricerca. Gasperi, F., Versini, G., ed. Temi, San Michele all’Adige, Italy Cochran, P. K., Paul, J. H. (1998) Seasonal abundance of lysogenic bacteria in a subtropical estuary. Applied and Environmental Microbiology 64, 2308–2312 Donnelly, C.W., 2004. Growth and survival of microbial pathogens in cheese. In: Fox, P. F., McSweeney, P. L. H., Cogan T. M., Guinee T. P. (Eds.), Cheese: Chemistry, physics and microbiology. Chapman and Hall, London 541–560 Franciosi, E., Settanni, L., Cavazza, A., Poznanski, E. (2009) Biodiversity and technological potential of wild lactic acid bacteria from raw cows’ milk. International Dairy Journal 19, 3–11 Gaglio, R., Francesca, N., Di Gerlando, R., Cruciata, M., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. (2014) Identification, typing, and investigation of the dairy characteristics of lactic acid bacteria isolated from 'Vastedda della valle del Belìce' cheese. Dairy Science & Technology 94, 157–180 Garde, S., Babin, M., Gaya, P., Nuñez, M., Medina, M. (1999) PCR amplification of the gene acmA differentiates Lactococcus lactis subsp. lactis and L. lactis subsp. cremoris. Applied and Environmental Microbiology 65, 5151–5153 ISO, 2003. ISO 13299. Sensory analysis – Methodology - General guidance for establishing a sensory profile. International Standardisation Organisation, Geneva, Switzerland Kosikowski, F.V. (1958) Problems in the Italian soft cheese industy. Journal of Dairy Science 41, 455–458 Leite, S.E., Montenegro, S.T.L., de Oliveira, L.E. (2006) Sensitivity of spoiling and pathogen food-related bacteria to Origanum vulgare L. (Lamiaceae) essential oil. Brazilian Journal of Microbiology 37, 527–532 Micari, P., Sarullo, V., Sidari, R., Caridi, A. (2007) Physico-chemical and hygienic characteristics of the Calabrian raw milk cheese, Caprino d’Aspromonte. Turkish Journal of Veterinary and Animal Sciences 31, 55–60. Mucchetti, G., Bonvini, B., Remagni, M. C., Ghiglietti, R., Locci, F., Barzaghi, S., Francolino S., Perrone, A., Rubiloni, A., Campo, P., Gatti, M., Carminati, D. (2008) Influence of cheese-making technology on composition and microbiological characteristics of Vastedda cheese. Food Control 19, 119–125 Mucchetti, G., Neviani, E. (2006) Microbiologia e tecnologia lattiero-casearia. Qualità e sicurezza, ed. Tecniche Nuove, Milan, Italy Niro, S. (2011) Innovazione di processo e di prodotto in formaggi a pasta filata. PhD thesis, Università degli Studi del Molise, Italy Parente, E., Cogan, T.M. (2004) Starter cultures: general aspects. Ed. Elsevier, London, UK Parente, E., Moschetti, G., Coppola, S. (1998) Starter cultures for Mozzarella cheese. Annali di Microbiologia ed Enzimologia 48, 89–109 Panoff, J.M., Legrand, S., Thammavongs, B., Boutibonnes, P. (1994) The cold shock response in Lactococcus lactis subsp lactis. Current Microbiology 29, 213–216 Pérez, G., Cardell, E, Zárate, V. (2002) Ramdom amplified polymorphic DNA analysis for differentiation of Leuconostoc mesenteroides subspecies isolated from Tenerife cheese. Letters in Applied Microbiology 34, 82–85 Salvadori del Prato, O. (1998) Trattato di Tecnologia Casearia. Ed. Edagricole, Bologna, Italy Settanni, L., Di Grigoli, A., Tornambé, G., Bellina, V., Francesca, N., Moschetti, G., Bonanno, A. (2012a) Persistence of wild Streptococcus thermophilus strains on wooden vat and during the manufacture of a Caciocavallo type cheese. International Journal of Food Microbiology 155, 73–81 Settanni, L., Franciosi, E., Cavazza, A., Cocconcelli, P.S., Poznanski, E. (2011) Extension of Tosèla cheese shelf-life using non-starter lactic acid bacteria. Food Microbiology 28, 883–890 Settanni, L., Gaglio, R., Guarcello, R., Francesca, N., Carpino, S., Sannino, C., Todaro, M. (2013) Selected lactic acid bacteria as a hurdle to the microbial spoilage of cheese: application on a traditional raw ewes’ milk cheese. International Dairy Journal 32, 126–132 Settanni, L., Miceli, A., Francesca, N., Moschetti, G. (2012b) Investigation of the hygienic safety of aromatic plants cultivated in soil contaminated with Listeria monocytogenes. Food Control 26, 213–219

142

CHAPTER VI

Settanni, L., Moschetti, G. (2010) Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiology 27, 691–697 Todaro, M., Francesca, N., Reale, S., Moschetti, G., Vitale, F., Settanni, L. (2011) Effect of different salting technologies on the chemical and microbiological characteristics of PDO Pecorino Siciliano cheese. European Food Research and Technology 233, 931–940

143

CHAPTER VII

Activation of wooden vats with selected lactic acid bacteria for the

year-round production of traditional Vastedda-like cheese

The present chapter reports a work in progress

144

CHAPTER VII

1. INTRODUCTION

Many traditional cheeses are manufactured in small size farms with raw milk from animals of indigenous breeds that are fed mainly on natural pasture. Some of these cheeses enjoy a “protected designation of origin” (PDO) status. Within this group, “Vastedda della valle del Belìce” cheese is typical of the homonymous valley of Sicily (Italy); it is produced with raw ewes’ milk without the addition of starter cultures (GUE no. C 42/16 19.2.2010) applying the technology of stretched (“pasta filata”) cheeses consisting of an acidification followed by the scalding of the acidified curd. Traditionally, Vastedda cheese was produced only during the summer, but due to the increasing demand for this cheese, it is now produced year-round. The production of Vastedda cheese is carried out in wooden vats that are characterized by the presence of microbial biofilms hosting lactic acid bacteria (LAB) responsible for the acidification of the curds.

2. OBJECTIVES

The objectives of this research are:  Monitoring of microbial biofilms of new formation onto the surfaces of “virgin” wooden vats;  Isolation and identification of LAB isolated from the surfaces of control wooden vats;  Activation of wooden vats with a mixture of lactococci previously isolated from PDO vastedda cheeses (Gaglio et al., 2014a) and applied in vivo in standard conditions (Gaglio et al., 2014b).

3. MATERIALS AND METHODS

3.1. Wooden vats activation

Four chestnut wooden vats of 100 L volume were purchased by a local artisan producer. The vats were used in controlled conditions in a dairy pilot plant (Istituto Zooprofilattico Sperimentale della Sicilia “Adelmo Mirri”, Palermo, Italy) (TZ, Fig. 1A), where only experimental productions are performed, working the milk of a single farm selected for the high hygienic characteristics, and in a dairy (“Il Cacio Siciliano” located in Belmonte Mezzagno, Palermo - Italy) (TA, Fig. 1B) which transforms bulk milk of several farms and the cheese is produced concomitantly with the daily cheese making. Both trails included a control vat (TA1 and TZ1) activated with the whey of the previous day cheese production and 145

CHAPTER VII

the experimental vat (TA2 and TZ2) activated with a whey containing the Lactococcus lactis subsp. cremoris PON36, PON153 and PON203 inocula produced as reported by Gaglio et al. (2014b).

Fig. 1. Wooden vats used for Vastedda cheese production. A, vats used in controlled conditions; B, vats used in dairy trial.

Before activation, the 4 wooden vats were daily treated with hot water for 1 month to remove the tannic components. Vastedda cheese was produced following the PDO protocol every day for the first 5 d and every 5 d for 1 month. Several samples of vat surfaces, milk, whey, curd and cheese were collected during production (Table 1).

4. RESULTS

The results of microbial loads of total psychrotrophic, total mesophilic and LAB during Vastedda cheese production and reported in Table 1. TPC were significantly less concentrated that TMM in all vats. Vat surfaces of the dairy trial showed a clear dominance of mesophilic LAB both in control and experimental vats. Lower counts of LAB were detected onto the surfaces of control and experimental vats kept in the controlled dairy pilot, with the high counts revealed by M17 that is indicated for lactococci. After contact with vat surfaces, the counts of lactococci increased in both conditions for the experimental vats, but high counts of LAB were also estimated on MRS for the control vats. Regarding pathogenic bacteria, L. monocytogenes and Salmonella spp. were never detected, while E. coli was found onto the surfaces of the vats only during the first days of activation, and disappeared over time. E. coli was present in cheese productions of the first days of the wooden vat activation for the control

146

CHAPTER VII

vats, but is was not detected after stretching. No production carried out in experimental vats showed the presence of E. coli in the acidified curds.

Table 1. Microbial load of total psychrotrophic, total mesophilic and LAB during Vastedda cheese production.

Sample Media PCA-SkM PCA-SkM M17 30°C M17 44°C MRS 30°C MRS 42°C 7°C 30°C Production TA1 Vat surfaces 4.22±0.99 6.67±0.16 6.78±0.66 3.77±0.91 6.48±0.11 1.91±0.95 Bulk milk 7.57±0.41 7.40±0.25 7.64±0.76 1.52±0.65 6.23±0.44 1.05±0.82 Bulk milk after resting 7.35±0.55 7.53±0.61 7.17±0.25 5.01±0.36 6.08±0.41 1.83±0.98 Curd 7.86±0.46 8.02±0.63 7.97±0.25 6.26±0.35 7.05±0.19 2.33±0.91 Whey 7.57±0.84 7.85±0.49 8.20±0.22 4.37±0.99 7.89±0.18 3.80±0.93 Deproteinized whey 6.10±0.88 5.47±0.94 5.47±0.91 2.49±0.85 5.20±0.74 2.51±0.77 Curd before stretching 9.10±0.16 9.10±0.16 9.42±0.20 8.04±0.51 8.76±0.47 2.51±0.99

Cheese at T0 7.31±0.34 7.31±0.34 8.15±0.82 8.16±0.99 7.24±0.43 6.72±0.35 Cheese at 15 d 8.18±0.11 8.38±0.23 8.22±0.53 8.07±0.13 4.36±0.81 2.53±0.75 Production TA2 Vat surfaces 3.41±0.99 5.87±0.90 7.07±0.92 2.36±0.91 6.82±0.69 1.77±0.85 Bulk milk after resting 7.19±0.47 7.43±0.19 7.85±0.51 3.19±0.81 6.37±0.41 2.38±0.93 Milk with whey starter 7.45±0.85 7.72±0.48 7.75±0.26 4.84±0.02 7.54±0.08 4.11±0.59 cultures Curd 8.32±0.97 8.06±0.48 8.22±0.80 3.89±0.92 8.40±0.17 3.39±0.81 Whey 5.33±0.74 8.37±0.54 9.01±0.23 4.71±0.79 8.81±0.51 1.16±0.77 Deproteinized whey 2.13±0.72 4.76±0.88 5.62±0.89 1.53±0.69 5.72±0.81 1.22±0.69 Curd before stretching 6.17±0.89 9.06±0.32 9.69±0.04 8.08±0.92 9.67±0.13 4.78±0.76

Cheese at T0 0 7.06±0.97 7.43±0.45 4.41±0.97 7.01±0.70 3.83±0.74 Cheese at 15 d 5.25±0.92 6.99±0.68 7.99±0.60 6.77±0.84 5.45±0.95 2.06±0.85 Production TZ1 Vat surfaces 0 2.83±0.88 4.56±0.84 2.87±0.85 2.20±0.75 2.87±0.42 Bulk milk 4.80±0.34 5.39±0.73 4.83±0.35 2.60±0.92 3.79±0.56 1.75±0.18 Bulk milk after resting 4.56±0.09 5.08±0.31 4.48±0.40 3.02±0.65 3.84±028 2.07±0.81 Curd 5.76±0.98 6.06±0.38 5.16±0.67 3.06±0.70 4.51±0.98 1.59±0.88 Whey 5.63±0.94 5.37±0.95 6.58±0.87 2.22±0.81 6.09±0.99 0 Deproteinized whey 8.65±0.91 8.65±0.89 8.85±0.60 8.02±0.66 8.80±0.72 7.00±0.19 Curd before stretching 8.55±0.76 8.91±0.51 9.00±0.46 5.26±0.98 8.51±0.73 2.06±0.84

Cheese at T0 8.67±0.93 8.72±0.98 8.64±0.96 7.67±0.89 8.61±0.83 6.81±0.72 Cheese at 15 d 7.66±0.42 7.80±0.46 8.28±0.42 7.82±0.64 7.87±0.31 6.66±0.91 Production TZ2 Vat surfaces 3.26±0.89 5.02±0.98 3.71±0.99 0.80±0.87 2.51±0.97 1.30±0.81 Bulk milk after resting 1.55±0.93 5.50±0.82 4.86±0.78 2.48±0.89 3.45±0.97 1.09±0.94 Milk with whey starter 3.28±0.89 6.22±0.40 7.37±0.15 3.36±0.83 7.19±0.06 1.97±0.81 cultures Curd 2.46±0.83 7.06±0.85 7.53±0.34 4.70±0.60 7.62±0.54 2.87±0.85 Whey 3.93±0.84 7.37±0.85 8.86±0.35 0.80±0.91 8.72±0.39 0 Deproteinized whey 4.53±0.82 3.47±0.13 3.55±0.84 0 3.52±0.74 0 Curd before stretching 8.15±0.91 8.26±0.85 9.49±0.20 7.48±0.65 9.38±0.31 3.99±0.83

Cheese at T0 7.37±0.04 7.90±0.32 7.97±0.14 7.40±0.75 5.49±0.73 5.68±0.75 Cheese at 15 d 2.42±0.89 6.41±0.68 8.00±0.51 6.70±0.73 7.73±0.57 3.61±0.93

The colonies of LAB grown at the highest levels on MRS and M17 both at 30 and 44°C were isolated and preliminary characterized. At the present time, the genetic identification of the LAB of wooden vat origin showed the dominance of Leuconostoc lactis, Lc. lactis subsp. lactis and Streptococcus thermophilus in control vats, while in the experimental vats were

147

CHAPTER VII

also detected Lc. lactis subsp. cremoris, and several enterococci in addition to the species found in control vats. The comparison of the polymorphic profiles showed that the lactococci found in the experimental vats were those inoculated with the whey during the activation.

REFERENCES

Gaglio, R., Francesca, N., Di Gerlando, R., Cruciata, M., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. (2014) Identification, typing and investigation of the dairy characteristics of lactic acid bacteria isolated from “Vastedda della valle del Belìce” cheeses. Dairy Science & Technology 94,157–180 Gaglio, R., Scatassa, M. L., Cruciata, M., Miraglia, V., Corona, O., Di Gerlando, R., Portolano, B., Moschetti, G., Settanni, L. (2014) In vivo application and dynamics of lactic acid bacteria for the four-season production of Vastedda-like cheese. International Journal of Food Microbiology 177, 37–48

148

CHAPTER VIII

Large scale applications of selected lactic acid bacteria to improve

PDO Pecorino Siciliano cheese

The present chapter represents a work in progress

149

CHAPTER VIII

1. INTRODUCTION

Pecorino Siciliano cheese is, probably, the oldest European cheese (Betta and Cantarelli, 2002). The PDO disciplinary goes back to 1955 (GURI n. 295 of 12-22-1955) and provides the use of entire ewe’s raw milk produced in Sicily, the use of traditional wooden equipment, the application of dry salting and a ripening period of at least 4 months. PDO Pecorino Siciliano cheese is obtained without the addition of starters. Thus, the microflora acting during cheese making and ripening is autochthonous, deriving from milk or the transformation environment. Due to the high numbers of undesired microorganisms, especially those potentially pathogenic for consumers, found in several productions of this cheese, the revision of the production protocol has been suggested (Todaro et al., 2011). The addition of autochthonous lactic acid bacteria (LAB) may be defining to improve the hygienic conditions of PDO Pecorino Siciliano cheese.

2. OBJECTIVES

The objectives of this research are:  Experimental cheese production with inoculation of autochthonous SLAB (Lactococcus lactis CAG4 e Lactococcus lactis CAG37) (Settanni et al., 2013) and NSLAB (Lactococcus garviae PSL67, Enterococcus faecalis PSL71 e Streptococcus macedonicus PSL72) (Todaro et al., 2011) in mixed combinations, for the production of PDO Pecorino Siciliano, in different Sicilian dairy factories;  Estimation of concentration of the different microbial groups at each phase both for control and experimental cheeses productions;  Verify the presence of LAB and the absence of undesirable microorganisms on the wooden vat surface;  Comparison of the physical and microbiological characteristics between control and experimental cheeses productions.

3. MATERIALS AND METHODS

3.1. Cheese production

The bacterial mixtures were prepared after the individual overnight growth of each strain. The LAB strains were first mixed together after growth in the optimal synthetic medium, re-

150

CHAPTER VIII

suspended in Ringer’s solution [overnight cultures were centrifuged at 5,000 × g for 5 min, washed twice in Ringer’s solution and re-suspended in the same solution till reaching an optical density (OD) at 600 nm of ca. 1.00 which approximately corresponds to a concentration of 109 ]. The experimental cheese making trials were carried out in seven different dairy factories located in province of Agrigento, Catania, Palermo, Ragusa and Trapani.

Table 1. City of dairy factory

Dairy Type of City (province)a Age (years) factory wood

1 Aidone (EN) Rovere 8

2 Ramacca (CT) Rovere 1

3 Salemi (TP) Duglas 9

4 Santa Margherita del Bèlice (AG) Duglas 5

5 Castronovo di Sicilia (PA) Duglas 6

6 Santo Stefano di Quisquina (AG) Rovere 1

7 Ragusa (RG) Rovere 6 a AG, Agrigento; CT, Catania; PA, Palermo; RG, Ragusa; TP, Trapani.

The inocula were added to a final concentration of approximately 107 CFU/mL for SLAB and 103 CFU/mL for NSLAB in 50 L of bulk milk after resting (Fig. 1).

Bulk milk 100 L

10 min

Adding of 500 ml of Addition of 500 ml of inocula physiological solution (CAG4,CAG37,PSL67,PS L71,PSL72

Control milk 50 L Milk with inocula 50 L

Fig. 1. Experimental design of PDO Pecorino Siciliano cheese productions.

151

CHAPTER VIII

REFERENCES

Betta, P., Cantarelli, F. (2002) Dal Mito alla storia, il Pecorino Siciliano. In: CO.RE.R.A.S. (ed) Palermo, Italy Settanni, L., Gaglio, R., Guarcello, R., Francesca, N., Carpino, S., Sannino, C., Todaro, M. (2013) Selected lactic acid bacteria as a hurdle to the microbial spoilage of cheese: Application on a traditional raw ewes’ milk cheese. International Dairy Journal 32, 126–132 Todaro, M., Francesca, N., Reale, S., Moschetti, G., Vitale, F., Settanni, L. (2011) Effect of different salting technologies on the chemical and microbiological characteristics of PDO Pecorino Siciliano cheese. European Food Research and Technology 233, 931–940

152

ACKNOWLEDGMENTS

Desidero ringraziare innanzitutto il mio tutor, Prof. Luca Settanni, a lui va la mia più profonda e sincera gratitudine per tutto quello che mi ha insegnato in questi anni, per aver creduto in me ed avermi trasmesso l’entusiasmo con cui affrontare la ricerca. Ringrazio il Prof. Giancarlo Moschetti per i suoi suggerimenti e spunti critici durante le sperimentazioni. Ringrazio il Prof. Baldassare Portolano, responsabile scientifico del progetto

PON01_02249 “Application of molecular biotechnologies and pro-technological microorganisms for the characterisation and valorisation of dairy and bakery chains of typical products” of the Italian Ministry of Education, University and Research (CUP: B11C11000430005), con i cui fondi è stata finanzia la mia attività di ricerca. Ringrazio il gruppo di Zootecnia del Dipartimento SAF, in particolare nelle persone della Prof.ssa Adriana Bonanno, del Dott. Massimo Todaro e del Dott. Antonino Di Grigoli per il supporto tecnico-scientifico fornitomi in questi anni.

Ringrazio, inoltre, la Dott.ssa Maria Luisa Scatassa, la Dott.ssa Isabella Mancuso e Salvatore Marceca dell’Istituto Zooprofilattico Sperimentale della Sicilia “A. Mirri” per il loro prezioso aiuto datomi durante lo svogimento delle caseificazioni. Un ringraziamento speciale a tutti i miei colleghi, ma soprattutto amici, che mi hanno sostenuto ed aiutato durante questi anni: Dott. Nicola Francesca, Dott. Antonio Alfonzo, Dott.ssa Margherita Cruciata, Dott.ssa Rosa Guarcello e tutti gli altri collaboratori del laboratorio di Microbiologia Agraria del Dipartimento SAF. Ci tengo a far sapere a tutti loro che lavorare insieme, seppur tra mille difficoltà, è stato davvero bello. Intendo ringraziare, il Rag. Luigi Sempione e tutto il personale del caseificio “Il Cacio Siciliano” per la cordialità e il calore con cui sono stato accolto in azienda, e per l’atmosfera serena e piacevole che ha accompagnato la mia attività di ricerca industriale. Un ringraziamento dovuto, che non sarà mai sufficiente per quello che fa e ha fatto per me, va alla mia famiglia che mi ha sempre appoggiato e supportato in tutte le mie scelte. Ringrazio, infine, immensamente Patrizia che mi ha sempre sostenuto nonostante la lontananza fisica.

PUBLICATIONS

Settanni, L., Gaglio, R., Guarcello, R., Francesca, N., Carpino, S., Sannino, C., Todaro, M. (2013) Selected lactic acid bacteria as a hurdle to the microbial spoilage of cheese: Application on a traditional raw ewes’ milk cheese. International Dairy Journal 32, 126– 132 Alfonzo, A., Ventimiglia, G., Corona, O., Di Gerlando, R., Gaglio, R., Francesca, N., Moschetti, G., Settanni, L. (2013) Diversity and technological potential of lactic acid bacteria of wheat flours. Food Microbiology 36, 343–354 Settanni, L., Guarcello, R., Gaglio, R., Francesca, N., Aleo, A., Felis, G. E., Moschetti, G. (2014) Production, stability, gene sequencing and in situ anti-Listeria activity of mundticin KS expressed by three Enterococcus mundtii strains. Food Control 35, 311– 322 Gaglio, R., Francesca, N., Di Gerlando, R., Cruciata, M., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. (2014) Identification, typing and investigation of the dairy characteristics of lactic acid bacteria isolated from “Vastedda della valle del Belìce” cheeses. Dairy Science & Technology 94,157–180 Gaglio, R., Scatassa, M. L., Cruciata, M., Miraglia, V., Corona, O., Di Gerlando, R., Portolano, B., Moschetti, G., Settanni, L. (2014) In vivo application and dynamics of lactic acid bacteria for the four-season production of Vastedda-like cheese. International Journal of Food Microbiology 177, 37–48 Di Grigoli, A., Francesca, N., Gaglio, R., Guarrasi, V., Moschetti, M., Scatassa, M. L., Settanni, L., Bonanno, A. (2015) The influence of the wooden equipment employed for cheese manufacture on the characteristics of a traditional stretched cheese during ripening. Food Microbiology 46, 81–91 Scatassa, M. L., Gaglio, R., Macaluso, G., Francesca, N., Randazzo, W., Cardamone, C., Di Grigoli, A., Moschetti, G., Settanni, L. Composition and characterisation of the lactic acid bacterial biofilms associated with the wooden vats used to produce two traditional stretched cheeses. Submitted for publication in International Journal of Food Microbiology

CONFERENCES

Gaglio, R., Scatassa, M. L., Francesca, N., Cruciata, M., Di Gerlando, R., Miraglia, V., Guarcello, R., Portolano, B., Moschetti, G., Settanni, L. Development of an ad hoc starter culture preparation for the four-season production of Vastedda della valle del Belìce cheese. 2nd International Conference of Microbial Diversity. Torino (TO) 23–25 Ottobre 2013. 295–296. Poster presentation. Guarcello, R., Gaglio, R., Francesca, N., Aleo, A., Felis, G. E., Moschetti, G., Settanni, L. Microbial interactions in food model systems: In situ antilisterial activity of mundticin KS producing strains. 2nd International Conference of Microbial Diversity. Torino (TO) 23–25 Ottobre 2013. 336–337. Poster presentation. Scatassa, M.L., Gaglio, R., Cruciata, M., Mancuso, I., Di Gerlando, R., Moschetti, G., Portolano, B., Settanni, L. Activation of wooden vats with selected lactic acid bacteria for the year-round production of traditional Vastedda-like cheese. 1st International Meeting on Milk, Vector of Development. Rennes (France) 21–23 Maggio 2014. 218. Poster presentation. Galluzzo, P., Settanni, L., Gaglio, R., Cascone, G., Macaluso, G., Moschetti, G., Portolano, B., Caracappa, S. Preliminary data on hydrolytic activity of lactic acid bacteria on β- lactoglobulin in milk. LXVIII Convegno Nazionale SISVet. Pisa (PI) 16–18 Giugno 2014. 138–139. Poster presentation. Gaglio, R., Scatassa, M.L., Cruciata, M., Miraglia, V., Corona, O., Di Gerlando, R., Portolano, B., Moschetti, G., Caracappa, G., Settanni, L. Application of selected lactic acid bacteria for the year-round production of vastedda-like cheese. LXVIII Convegno Nazionale SISVet. Pisa (PI) 16–18 Giugno 2014. 153–154. Oral communication.

Faculdade de Medicina

Veterinária

ABROAD TRAINING

Professor Constança Pomba, Head of the Laboratory of Antimicrobial and Biocide Resistance, Faculty of Veterinary Medicine, University of Lisbon, Av Universidade Técnica de Lisboa 1300-477 LISBOA, Portugal

Lisboa, 1 December 2014

That Raimondo Gaglio, Ph.D. student in “Sistemi Agro-Ambientali – Indirizzo Agro- Ecosistemi Mediterranei” at University of Palermo (Italy) and working at the Department of Agricultural and Forest Science - Agricultural Microbiology Unit, worked in Department of Clinics, Faculty of Veterinary Medicine, University of Lisbon for three months between 1 september 2014 – 1 december 2014. During this stay, he worked in the Laboratory of Antimicrobial and Biocide Resistance, in the phenotypic and genotypic characterization of a collection of Enterococcus spp. isolated from traditional cheese productions carried out in Sicily (south Italy).

Lisboa, 1 December 2014

Signed by

Dr. Maria Constança Matias Ferreira Pomba Associate Professor of the Faculty of Veterinary Medicine Head of the Laboratory of Antimicrobial and Biocide Resistance

Av.Universidade Técnica 1300- 477 LISBOA  PORTUGAL  213652800/213652837   213652810/213652897  http://www.fmv.utl.pt [email protected]