Identification of Lipid Biomarkers To Discriminate between the

Subscriber access provided by READING UNIV

Article

Identification of Lipid Biomarkers to Discriminate the Different Production Systems of Asiago PDO Cheese Severino Segato, Gianni Galaverna, Barbara Contiero, Paolo Berzaghi, Augusta Caligiani, Angela Marseglia, and Giulio Cozzi J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.7b03629 • Publication Date (Web): 24 Oct 2017 Downloaded from http://pubs.acs.org on October 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

Journal of Agricultural and Food Chemistry

1

Identification of Lipid Biomarkers to Discriminate the Different Production Systems of

2

Asiago PDO Cheese

3 4

Severino Segato*,†, Gianni Galaverna‡, Barbara Contiero†, Paolo Berzaghi†, Augusta

5

Caligiani‡, Angela Marseglia‡, Giulio Cozzi†

6 7

†

8

Legnaro (PD), Italy

9

‡

Department of Animal Medicine, Production and Health, University of Padova, 35020

Department of Food and Drug, University of Parma, 43121 Parma, Italy

10 11

*Corresponding author:

12

Phone: ++39-049-827-2662; e-mail: [email protected]

13

ORCID:

14

Severino Segato: 0000-0003-0036-8502

1 Environment ACS Paragon Plus


15

ABSTRACT

16

Lipid fraction of Asiago Protected Designation of Origin (PDO) cheese was analysed to

17

identify specific biomarkers of its main production systems through a canonical discriminant

18

analysis. The three main production systems of the cheese were considered. Two were located

19

in the upland (UL): pasture based (P-UL) vs. hay based total mixed rations (H-UL). The third

20

was located in the lowland (LL) and processed milk from cows fed maize silage based rations

21

(maize silage lowland: MS-LL). The discriminant analysis selected 9 fatty acids and vitamin

22

A as lipid biomarkers useful to separate the three production systems. High contents of

23

conjugated linoleic acids, anteiso-C15:0 and vitamin A were discriminant factors for P-UL

24

cheese. The separation between H-UL and MS-LL cheese was less marked with the former

25

having the higher content of conjugated linoleic acids and some polyunsaturated n-6 fatty

26

acids, while the latter being identified by cyclopropane fatty acid and C9:0.

27 28

KEYWORDS: Asiago cheese, production system, lipid fraction, canonical discriminant

29

analysis

30 31

INTRODUCTION

32

Nowadays many types of cheese produced in the European Union are officially labelled

33

according to specific production methods (i.e, organic or traditional specialities guaranteed)

34

and/or to the geographical location of the milk production system (i.e., mountain products). In

35

order to avoid frauds, both the consumers and producers would benefit from a set of

36

identification biomarkers capable of discriminating and distinguish cheeses and dairy

37

products from each other. At this regard, the lipid fraction of dairy products has a relevant

38

role in the characterization of their nutritional1 and organoleptic quality.2,3 The outcomes of a

39

recent meta-analysis on the existing scientific literature comparing conventional and organic


Page 2 of 22

Page 3 of 22


40

bovine milk indicated that the supply of nutritionally desirable long-chain polyunsaturated n-3

41

fatty acids (PUFA) was improved in the lipid composition of organic milk.4 The presence of

42

conjugated linoleic acids (CLA), odd- and branched-chain fatty acids (OBCFA) and other

43

secondary fatty acids have also shown to improve cheese nutritional properties, since these

44

bioactive substances are positively associated with human health.5 Fat soluble vitamins are

45

further bioactive constituents of the cheese lipid fraction due to their anti-oxidant activity that

46

extends food shelf-life and prevents rancidity phenomena.6 Recent scientific evidences

47

showed that milk fatty acids (FA) profile and fat-soluble vitamins can be used to discriminate

48

different feeding systems of dairy cows.3,7,8 Therefore, the identification of qualified

49

biomarkers from the analysis of the lipid fraction could represent a promising approach to

50

trace the different production system of a given cheese.

51

Asiago is a semi-hard cheese produced from cow’s milk in a restricted area of north-eastern

52

Italy. Despite being processed and ripened according to its Protected Designation of Origin

53

(PDO),9 Asiago cheese has a wide variability in quality due to its different milk production

54

systems and cheese-making conditions.10 This study aimed at analysing lipid composition of

55

Asiago PDO cheese in order to identify specific biomarkers capable to discriminate its main

56

production systems.

57 58

MATERIALS AND METHODS

59

Cheese Production Systems and Experimental Design. The study considered the three

60

main cheese production systems (CPS) of the Asiago PDO which differ for size and location

61

of the cheese factories, and for cows feeding and management system (Table 1). Two

62

production systems represent the historic center of Asiago cheese and they are located in the

63

upland (UL) of Asiago (around 1000 m above sea level) in the Veneto Region (Italy). The

64

first is a seasonal pasture based system (P-UL) in which cows graze during the summer on



65

natural pastures and receive a daily amount of concentrates [3.0-5.0 kg of dry matter (DM)

66

per lactating cow] according to their milk yield. This is a very small scale production system

67

that usually processes milk produced by a single or a few dairy herds. The other is a hay

68

based production system (H-UL) with cows fed total mixed rations (TMR) made of local

69

permanent grass, cereal grains and protein sources. It represents small/medium size alpine

70

cheese factories that process milk produced from local dairy farms to produce cheese labelled

71

as mountain product in compliance with the EU regulation n. 665/2014. A third production

72

chain (MS-LL) features medium/large size cheese factories located in the lowland (LL)

73

surrounding the highland of Asiago. These productive sites process milk collected from dairy

74

farms in which high producing dairy cows are kept indoor and fed maize silage (MS) based

75

high energy TMR.

76

Three separate bulks of raw milk were collected within each CPS during the summer period

77

(from July to September) and they were processed to produce Asiago PDO cheese in different

78

sites. All wheels had a similar ripening time of about 6 months that was carried out in storage

79

bay at 10±2 °C and 80-85% relative humidity. The sampling protocol considered 81 wheels of

80

cheese (28 of P-UL, 26 of H-UL cheese and 27 of MS-LL, respectively).

81

Chemical Analysis. Cheese samples were analyzed in duplicate for moisture,11 total

82

protein and water soluble nitrogen,12 and sodium chloride content.13 Total lipids content was

83

determined using an accelerated solvent extraction method by ASE 200 (Dionex Corporation,

84

Sunnyvale, CA) with hexane/isopropanol (3:2, v/v) solution that was removed by means of a

85

rotary evaporator. According to the method proposed by Christie et al.,14 an aliquote of 40 mg

86

of anhydrous fat was used to methylate the fatty acids (FAs). Sodium methoxide (CH3ONa)

87

was used as catalyst and methyl 12-tridecenoate as an internal standard (Nu Chek Prep Inc.,

88

Elysian, MN). The analyses of FAs methyl esters were performed by two dimensional GC on

89

a 7890A instrument (Agilent Technologies, Santa Clara, CA) connected with a


Page 4 of 22

Page 5 of 22


90

chromatography data system (Agilent ChemStation) software (Agilent Technologies,

91

Wilmington, DE) and fitted with two capillary columns: the first used was a 75 m x 180 µm

92

i.d., 0.14 µm film thickness (Supelco, Bellefonte, PA); the second used was a 3.8 m x 250 µm

93

i.d., 0.25 µm film thickness (J&W Scientific, Folsom, CA). The carrier gas was hydrogen and

94

the injector temperature was 270 °C meanwhile the flame ionization detector operated at 250

95

°C. The oven temperature of the GC varied from 50-150 °C with an increase of 50 °C/min,

96

and then ramped up to 240 °C at 2 °C/min. The cheese FAs were identified by comparing

97

their retention time and peak position to those of a mixture of 52 pure FAs [reference mixture

98

(Nu Chek Prep Inc., Elysian, MN); reference mixture and bacterial acid methyl esters

99

(Supelco); PUFA No.3 from menhaden oil (Sigma-Aldrich, St. Louis, MO); individual CLA

100

isomers (Nu Chek Prep Inc.; Matreya LLC, State College, PA)]. Standard (dihydrosterculic

101

acid methyl ester, purity >98%) of cyclopropane FA (CPFA) was commercially available

102

(Abcam, Cambridge, UK) while ϖ-cyclohexyl tridecanoic acid (ϖ-C19) was synthetized as

103

reported previously.15 The two-dimensional chromatograms were analysed with the

104

comprensive GC x GC image software (Zoex Corp., Houston, TX). Based on peak areas,

105

single FAs were expressed as g/100 g of total FAs according to the correction factors of

106

AOAC 963.22 method.16 The analysis of vitamin A (retinol) and vitamin E (α-tocopherol)

107

was performed by high performance liquid chromatography according to the methods

108

proposed by Zahar and Smith17 and Indyk,18 respectively. A model 510 HPLC (Waters Corp.

109

Milford, MA) was equipped with a 150 mm x 4.6 mm i.d., 5 µm, RP-C18 Discovery column

110

(Supelco, St. Louis, MO) at 25 °C, mobile phase methanol/water (98:2, v/v) and model 2996

111

UV detector (Waters Corp.). The two vitamins were determined at 324 nm for retinol and 293

112

nm for α-tocopherol according to the retention times of reference standard (Sigma-Aldrich,

113

St. Louis, MO).



114

Statistical Analysis. All the statistical analyses were performed using SAS 9.4 software

115

(SAS Institute Inc., Cary, NC). Cheese chemical data were submitted to ANOVA (PROC-

116

GLM), adopting a linear model that considered the fixed effects of production system and

117

milk bulk. The two degrees of freedom of CPS factor were used to perform the following set

118

of orthogonal contrasts: P-UL vs. (H-UL + MS-LL)/2 and H-UL vs. MS-LL. The dataset of

119

the cheese lipid fraction was submitted to a preliminary selection in order to avoid

120

redundancy among variables. The procedure deleted variables strongly correlated among each

121

other (r value > 0.95 for the Pearson coefficient). A stepwise discriminant analysis (PROC

122

STEPDISC) was then applied on the remaining FAs and fat-soluble vitamins to reduce the

123

number of variables according to CPS. A canonical discriminant analysis (PROC CANDISC)

124

was performed to achieve the most discriminant variables that maximize the distance among

125

CPS groups. Two canonical functions named CAN1 and CAN2 were obtained and the final

126

pool of FAs and vitamins highly correlated to them (r value > 0.60) was identified. The

127

degree of dissimilarity among production systems was measured by squared Mahalanobis

128

distances (D2-Mahalanobis) and the reliability of the canonical discriminant model was finally

129

assessed by a cross-validation.

130 131

RESULTS AND DISCUSSION

132

Cheese Chemical Composition, Fatty Acids Profile and Fat Soluble Vitamins Content.

133

Chemical composition of the cheese samples belonging to the three CPS are reported in Table

134

2. Cheese from MS-LL had the highest moisture content and the lowest fat content. Limited

135

difference in chemical composition were detected between P-UL and H-UL except for water

136

soluble nitrogen and salt content. The FA profiles of the cheese samples are reported in Table

137

3. Cheese from the pasture based system had the highest content of the short-chain FA C4:0,

138

whereas it had a lower concentration of medium-chain FAs (from C8:0 to C14:0) than H-UL


Page 6 of 22

Page 7 of 22


139

and MS-LL cheese. Pasture based cheese was also associated to a higher (P < 0.001)

140

concentration of MUFAs and PUFAs and a reduction of SFAs. Moreover, consistent with

141

previous studies on alpine pasture cheese and milk,19,20 P-UL cheese had the richest CLAs

142

content (P < 0.001). The concentration of these healthy constituents in milk and dairy foods is

143

increased by the intake of fresh grass that are rich in C18 precursors, such as linolenic and

144

linoleic acid as well as by the prevention of the reduction to stearic acid of the vaccenic acid

145

produced in the rumen.21,22 Under grazing conditions, the dietary C18:3 n-3 is mainly

146

biohydrogenated and the derivatives are trans-11 C18:1 and C18:0 that are strongly

147

desaturated to CLA in the mammary gland. Odd- and branched chain FAs (OBCFA) are

148

further healthy compounds of the lipid fraction in milk and dairy products and in this study,

149

their total content (P = 0.008) and in particular that of iso- and anteiso-C15:0, iso-C16:0,

150

anteiso-C17:0 and iso-C18:0 (P < 0.001) increased in P-UL cheese according to the higher

151

forage content of the diets fed in this CPS.23 Among cyclic FAs, ϖ-cyclohexyl tridecanoic (ϖ-

152

C19) was similar across CPS, whereas cyclopropane FA (CPFA), was detected only in MS-

153

LL cheese. The statistical contrast between H-UL and MS-LL samples was not significant for

154

most FAs (Table 3). However, H-UL cheese had a lower concentration of SFAs (P < 0.001)

155

and higher concentration of MUFAs (P = 0.020), PUFAs (P < 0.001) and cis-9,trans-11 CLA

156

(P < 0.001). Vitamin A (retinol) and vitamin E (α-tocopherol) concentrations were highest (P

157

< 0.001) in P-UL cheese, while there was no difference between H-UL and MS-LL (Table 3).

158

Fresh grass is rich of carotenoids and α-tocopherol while sun exposure and oxidation during

159

harvest, ensiling and storage cause their losses in preserved forages explaining the lower

160

content of their derivatives in food products from hay and/or silage based dairy systems.24

161

Canonical Discriminant Analysis. A canonical discriminant analysis (CDA) was carried out

162

by using the dataset relative to the lipid fraction of the 81 cheese samples to categorize the

163

three CPS. Two significant functions (CAN1 and CAN2) (Wilks’ Lambda = 0.0023, approx.



164

F value = 60.5, df1= 38, df2= 116, P < 1·10-4) accounted for 91.4% and 8.6% of the total

165

variability, respectively. The stepwise procedure used by the discrimination model selected 15

166

variables among all lipid compounds: 14 FAs and vitamin A (Table 4). However, only 9 FAs

167

and vitamin A had correlation coefficient values with the total canonical structure higher than

168

0.60 (Figure 1). A scattergram based on the standardized canonical coefficients was created to

169

plot the separation of the cheese samples according to the three CPS (Figure 2). The squared

170

Mahalanobis distances among them indicated that P-UL samples were extremely different

171

from both H-UL (D2-Mahalanobis = 272; P < 0.001) and MS-LL (D2-Mahalanobis = 260; P

Identification of Lipid Biomarkers To Discriminate between the

Recommend Documents