Free Maillard Reaction Products in Milk Reflect Nutritional Intake of

Subscriber access provided by UNIV TORONTO

Article

Free Maillard reaction products in milk reflect nutritional intake of glycated proteins and can be used to distinguish “organic” and “conventionally” produced milk Uwe Schwarzenbolz, Thomas Hofmann, Nina Sparmann, and Thomas Henle J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.6b01375 • Publication Date (Web): 23 May 2016 Downloaded from http://pubs.acs.org on June 5, 2016

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 33

Journal of Agricultural and Food Chemistry

Free Maillard reaction products in milk reflect nutritional intake of glycated proteins and can be used to distinguish “organic” and “conventionally” produced milk

Uwe Schwarzenbolz, Thomas Hofmann, Nina Sparmann, Thomas Henle

Institute of Food Chemistry, Technische Universität Dresden, D-01062 Dresden, Germany

Corresponding author: T. Henle Tel.: +49-351-463-34647 Fax: +49-351-463-34138 Email: [email protected]

1 ACS Paragon Plus Environment


1

Page 2 of 33

Abstract

2 3

Using LC-MS/MS and isotopically labelled standard substances, quantitation of free Maillard

4

reaction products (MRPs), namely Nε-(carboxymethyl)lysine (CML), 5-(hydroxymethyl)-1H-

5

pyrrole-2-carbaldehyde (pyrraline, PYR), Nδ-(5-hydro-5-methyl-4-imidazolon-2-yl)-ornithine

6

(MG-H) and Nε-fructosyllysine (FL) in bovine milk was achieved. Considerable variations in

7

the amounts of the individual MRPs were found, most likely as a consequence of the

8

nutritional uptake of glycated proteins. When comparing commercial milk samples labelled as

9

originating from “organic” or “conventional” farming, respectively, significant differences in

10

the content of free PYR (organic milk: 20-300 pmol/mL; conventional milk: 400-1000

11

pmol/mL) were observed. An analysis of feed samples indicated that rapeseed and sugar beet

12

are the main sources for MRPs in conventional farming. Furthermore, milk of different dairy

13

animals (cow, buffalo, donkey, goat, ewe, mare, camel) as well as for the first time human

14

milk was analyzed for free MRPs. The distribution of their concentrations, with FL and PYR

15

as the most abundant in human milk and with a high individual variability, also points to a

16

nutritional influence As the components of concentrated feed do not belong to the natural

17

food sources of ruminants and equidae, free MRPs in milk might serve as indicators for an

18

adequate animal feeding in near-natural farming and can be suitable parameters to distinguish

19

between an “organic” and “conventional” production method of milk.

20 21

Keywords

22

Milk, organic and conventional farming, Maillard reaction, pyrraline, animal feeding

23


Page 3 of 33


24

Introduction

25

Since mankind started to use fire for cooking, Maillard reaction products are part of the

26

human diet.1 MRPs can be divided into compounds from the “early” stage, namely Amadori

27

products, and “advanced glycation end products” (AGE) such as the lysine derivatives N-ε-

28

carboxymethyllysine (CML) or pyrraline (PYR), respectively, and the arginine derivative

29

MG-H1. 2,3 The daily amount of MRPs taken up with a regular Western diet was estimated to

30

1000 mg for Amadori products, calculated for N-ε-fructosyllysine (FL), and 25 to 75 mg for

31

the AGEs CML and PYR.4 For CML and FL a release of protein-bound derivatives into

32

peptides, which are small enough to be absorbed (< 1000 Da), is described.5 Depending on the

33

individual chemical structure there are reports about the metabolic transit of MRPs. Amadori

34

products were found to pass through the intestinal mucosa only by diffusion to absorption

35

rates of 1-3%,6 while MRPs like CML, MG-H1 and PYR can be transported actively into

36

mucosa cells by the PEPT-1 transporter when bound in dipeptides.7 Within the epithelial

37

cells, after peptidolysis, MRPs with hydrophobic side chains, namely PYR, can pass through

38

the basolateral membrane, while MRPs with charged or polar side chains, like CML and MG-

39

H1 remain trapped inside the epithelial cells. Nevertheless, in infants the plasma and urine

40

levels of CML can be influenced through the diet8, most likely because of the still developing

41

barrier function of the epithelium. Consequently, from dietary studies it could be concluded

42

that the excretion of PYR, measured in urine

43

this particular MRP.

44

Due to the fact that dietary MRPs like pyrraline can be found in the urine originating from

45

blood clearance, it can be speculated, that mammals also excrete AGEs by lactation.

46

Cattle are ruminants and their digestive system is specialized on feed with high amounts of

47

roughage. The milk yield of dairy cows in conventional farming increased significantly over

48

the past years.10 To cover the energy consumption for the production of up to 40-50 l milk per

9

accounts for 50% of the ingested amount of



49

day, a cow needs feed with a higher energy density and more nitrogen compared to its natural

50

food sources. Thus the addition of molasses, coarse colza and soy are indispensable in

51

conventional farming with the consequence of elevated MRP levels in cow’s nutrition11.

52

It is known that MRPs, administered with feed ingredients, are able to influence the rumen

53

flora12, but in contrast to pet food13, to the best of our knowledge, there are no reports on the

54

amount of MRPs in feed used for cattle.

55

Caused by an increasing consumer’s demand, the production of organic food grew in recent

56

years and was covering about 7-8% of the total agricultural production in European countries

57

like Danmark or Germany in 2012. 14

58

Milk and milk products are of particular importance, as they have the highest sale revenue

59

among all the different categories. Currently, the demand is still exceeding the supply, so it is

60

essential, to find ways to distinguish between “organic” and “conventional” products. For

61

milk, the ratios of carbon and nitrogen isotopes were suggested to be valid parameters.15 As

62

the studies showed that the δ13C value as single indicator is not sufficient, additional

63

parameters like the α−linoleic acid content16 or the δ15N nitrogen isotope ratio17 were

64

included for an unequivocal differentiation between organic and conventional milk. Recently,

65

with a sophisticated analysis of the fatty acid and antioxidant profile of milk fat, combined

66

with multivariate statistics, Kusche et al.

67

between samples of organic and conventional farming. Although the stable isotope ratio as

68

well as the spectrum of fatty acids in milk vary with the season and may possess regional

69

variations, in combination these parameters offer effective tools to distinguish between

70

organic and conventional milk and within limitations their significance can be transferred on

71

processed milk products as well.19

72

Additionally, for milk products like cheese, phytanic and pristanic acid, microbiological

73

degradation products of chlorophyll in the rumen were suggested as indicators of organic

74

feeding.20 A target value of 200 mg/100g lipid of phytanic acid was proposed to be a 4

18

demonstrated that it is possible to distinguish

ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33


75

threshold, but some conventional products also reached the lowest levels of products from

76

organic farming. To increase the accuracy of the discrimination between organic and

77

conventional, the authors additionally suggest to validate the ratio of the diastereomers of

78

phytanic acid and to include the δ13C values in the decision.21

79

As feed is able to influence the composition of milk, the hypothesis for the work presented

80

here was that higher levels of free MRPs should be detectable in milk of mammals which

81

have consumed heated food/feed. Especially, if cows are fed concentrated feed and are able to

82

absorb MRPs in their intestine, these substances could be helpful to indicate feed ingredients,

83

which are not part of the natural food sources. For this reason we adapted an LC-MS/MS

84

method22 and examined commercially available milk samples from organic and conventional

85

production for free MRPs and compared the values with the free MRP content of milk from

86

other species as well as with the MRP content in cows feed.

87 88

Materials and Methods

89

Chemicals. Nonafluoropentanoic acid (NFPA) was from Sigma (Seelze, Germany), CML and

90

[2H2]CML were obtained from PolyPeptide (Strasbourg, France). [13C6,15N2] lysine and [13C6]

91

arginine were purchased from Euriso-top (Saarbrücken, Germany). HPLC grade methanol and

92

acetonitrile were from VWR Prolabo (Darmstadt, Germany).

93

The synthesis of the following MRPs was performed according to the literature stated: FL,23

94

[13C6,15N2]FL,24 PYR,25 and MG-H1.7 The synthesis of [13C6,15N2] PYR and [13C6] MG-H1

95

were performed in the same manner but using [13C6,15N2] lysine (PYR) and [13C6] arginine

96

(MG-H1) instead of the unlabeled amino acids.

97

All other chemicals were from Merck (Darmstadt, Germany).

98

Milk samples. A total number of 32 commercially available whole milk samples (protein: 3.4

99

+/ 0.1%; fat: 3.6 +/- 0.1%) from conventional or organic production (according to 5 ACS Paragon Plus Environment


Page 6 of 33

100

manufacturer´s data) were purchased from local stores and included 12 pasteurized (2

101

conventional, 10 organic), 13 extended shelf life (ESL) (7 conventional, 6 organic), 4 ultra-

102

high temperature treated (2 conventional, 2 organic) and 3 raw (1 conventional, 2 organic)

103

milk samples. Two conventionally produced raw cow´s milk samples were obtained from a

104

local

105

Großerkmannsdorf, Germany). These samples were taken from individual cows directly from

106

the udder and all single feed ingredients (coarsely grinded soy, rapeseed and corn, sugar beet

107

pellets, molasses, cereals and palm oil) and readily mixed cow’s feed, which was given to the

108

cows on the day of milking, was recorded. Furthermore, one milk sample each of buffalo

109

(Landgut Chrusdorf, Penig, Germany), ewe (Schuberts Milchschafhof, Saultitz, Germany),

110

goat (Ziegenhof Winter, St. Egidien, Germany), camel (Camel Milk Vitality, Eggerding,

111

Austria), horse (NaturOase Sachsen Oederan, Germany) and donkey (Asinerie d’Embazac,

112

L’Isle Jourdain, France) were analyzed. 17 samples of human milk were kindly donated by 5

113

volunteers.

114

All milk samples were freeze dried after purchase or sampling and stored at -18°C. Before

115

analysis, the milk was reconstituted with double distilled water.

116

Feed samples. All ingredients, namely grinded corn, soy and rapeseed, sugar beet pellets,

117

cereals, palm oil and molasses, as well as the complete feed mixture, which additionally

118

contained silage, were obtained from a local farmer (Landwirtschaftliches Unternehmen "An

119

der Dresdner Heide", Großerkmannsdorf, Germany).

120

Incubation of human milk

121

Each 3 mL of human milk were incubated at 37° C after addition of one small crystal of

122

thymol for 2, 4, 6, 20 and 24 h in sealed reaction tubes. Immediately after incubation the

123

samples were frozen and kept at -18°C until analysis.

farmer

(Landwirtschaftliches

Unternehmen

"An

der

Dresdner

Heide",


Page 7 of 33


124

Sample preparation

125

Free MRPs in milk. 1 mL of reconstituted milk was mixed with 10 mL of an ice cold

126

mixture of acetonitrile with methanol (50/50, v/v) in a test tube and 10 µL of internal standard

127

mixture (see table S1 in the supplement) were added. After cooling at -25 °C for 1h, the

128

samples were centrifuged at 3000 g for 5 min. The supernatant was decanted into a poly

129

propylene tube and evaporated on a Zymark TurboVab LV Evaporator (Hopkinton, MA,

130

USA). The precipitate was dissolved in 1.4 mL water and 1.4 mL hexane. For defatting, the

131

organic phase was discarded after centrifugation (1000 g; 2 min) and to the remaining water

132

phase, 100 µL of NFPA were added. This mixture was applied to a solid phase extraction

133

(SPE) cartridge (Waters Oasis HLB 6 cc, Milford, MA, USA), previously prepared by

134

consecutively flushing with 2 mL methanol, 2 mL methanol/10 mM NFPA (50/50,v/v) and 4

135

mL 10 mM NFPA. Following sample application, the SPE cartridge was washed with 4 mL

136

of a mixture of methanol and 10 mM NFPA (5/95,v/v) and the analytes were eluted with 5

137

mL of a mixture of methanol and 10 mM NFPA (90/10,v/v) into test tubes. The eluate was

138

evaporated, redisolved in 40 µL of water and applied to HPLC-MS. All samples were

139

analyzed at least in triplicate.

140

Quantitation of free and total MRPs in cow’s feed. Single feed ingredients and readily

141

mixed cow’s feed were grinded as fine as possible in a laboratory mill (Retsch, Haan,

142

Germany) and homogenized. For analysis of free MRPs, each sample was weight in according

143

to the expected concentrations (between 1.5 and 3 g) and extracted with defined volumes

144

(between 6 and 16 mL) of 0.1 M HCl using an ultra-turrax disperser and an ultrasonic bath.

145

These mixtures were centrifuged and 1 mL of the supernatant was treated in the same way as

146

described for milk samples, starting with the addition of 10 mL of an ice cold mixture of

147

acetonitrile with methanol (50/50, v/v).

148

Total MRPs were analyzed after enzymatic digestion according to the literature26. In brief,

149

each sample, representing 2–4 mg of protein, was incubated for 24 h at 37 °C in the presence 7 ACS Paragon Plus Environment


150

of 1 FIP-U of pepsin in 1.0 mL of 0.02 N HCl. Subsequently, 250 µL of 2 N TRIS buffer, pH

151

8.2 and 400 PU of pronase E were added and the solution was incubated for further 24 h at

152

37 C. Finally, 0.4 U of aminopeptidase M and 1 U of prolidase were added. After an

153

additional incubation for 24 h at 37 °C, the solutions were lyophilized, reconstituted in 1.4 mL

154

of water. After addition of 10 µL internal standard mixture and 1.4 mL of heptane the solution

155

was vortexed and centrifuged. The organic phase was discarded and to the remaining water

156

phase, 100 µL of NFPA were added. The subsequent sample preparation followed the

157

protocol described in the “free MRPs in milk” section by applying sample solution to the

158

SPE.

159

Acid hydrolysis. To verify the extent of the enzymatic hydrolysis, a comparative acid

160

hydrolysis of feed and milk samples was performed according to Henle et al. 1991 27 but with

161

lithium buffers on a Sykam Amino Acid Analyzer S 433 (chromatographic system and all

162

buffers were purchased from Sykam, Fürstenfeldbruck, Germany). For this purpose, a sample

163

amount equivalent to 10 mg protein was weight into a reaction tube and suspended in 6 N

164

HCl, the tube was flushed with nitrogen, then closed with a lid and heated at 110° C for 23 h.

165

Subsequent an aliquot of 1 mL hydrolysate was evaporated to dryness, solved in 500 µL

166

loading buffer and 50 µL applied to the amino acid analysis. These experiments were carried

167

out in triplicate.

168

HPLC-MS analysis of Maillard reaction products

169

The selected MRPs were analyzed on an Agilent 1200 series chromatographic system

170

connected to an Agilent 6410 mass spectrometer with electro spray ion source (ESI-MS, all

171

from Agilent, Waldbronn, Germany). The general working conditions were 350° C gas

172

temperature, gas flow 11 l/min, nebulizer pressure 35 psi and a capillary voltage of 4000 V.

173

For quantitative analysis, the MRM modus with ion transitions shown in table 1 and the

174

following chromatographic conditions were used. The elution buffers were 5 mM NFPA in 8 ACS Paragon Plus Environment

Page 8 of 33

Page 9 of 33


175

water (eluent A) and 5 mM NFPA in acetonitrile (eluent B). The gradient program started at

176

85% A which reduced to 15% within 21 minutes. Subsequent to holding 15% A for 9 minutes

177

the column was equilibrated (15 min) to the initial conditions. A combination of a ZORBAX

178

300SB18, 2.1x50 mm; 1.8 µm column (Agilent, Waldbronn, Germany) with a Hypercarb

179

2,1x150 mm, 5 µm (Thermo Fisher Scientific, Dreieich, Germany) and the flow rate was 0.2

180

mL/min.

181

Method validation and statistics

182

The limit of detection (LOD) and limit of quantitation (LOQ) were estimated from signal to

183

noise ratios (s/n) multiplied by 3 (LOD) or 9 (LOQ), respectively. Signal to noise ratios were

184

taken from chromatograms of human milk samples which showed the lowest signal for the

185

respective analyte.

186

To assay the recovery of the tested analytes, a sample of human milk was spiked with

187

standard (table 2) and analyzed according to the protocol above. The standard deviation of

188

repeatability was calculated from multiple analysis (n=8) of a single human milk sample.

189

Statistical parameters were calculated by using the program Sigmaplot version 12.0

190

(build12.0.0182, Systat software, Chicago, USA). First, the data were tested for normality and

191

depending on the individual results, mean or median with according confidence intervals

192

where calculated following ISO 16269 statistical interpretation of data28 and ISO 2602

193

statistical interpretation of test results29, respectively. The visiualization of the data in figure 2

194

was prepared based on Massart et al. (2005)30 with whiskers which were calculated as 1.5-

195

fold the interquartile range31.

196 197

Results and discussion

198

Method and validation 9 ACS Paragon Plus Environment


199

Selected MRPs, namely FL, CML, MG-H and PYR, were separated via RP-HPLC in the free

200

form after precipitation of proteins and solid phase extraction using two consecutive columns

201

with a mass spectrometric detection in MRM mode. As can be seen in Figure 1, MRPs eluted

202

between 5 and 14 minutes. Identification of each substance was achieved by comparing

203

retention time and mass spectra with isotopically labeled reference compounds [13C6,15N2]FL,

204

[2H2]CML, [13C6,15N2]PYR and [13C6]MG-H1. For MG-H1, the different isomers, which form

205

in the reaction between methylglyoxal and arginine32, could not be separated and, therefore,

206

contribute all to the area of a peak at retention time 10 min, which was named “MG-H”. For

207

quantitation, a standard of [13C6]MG-H1 was used under the assumption that all isotopes show

208

equivalent signal responds.

209

Overlay chromatograms for selected samples of raw milk from either conventional or organic

210

production, respectively, are shown in Figure 1. Additionally, the isotopically labelled

211

internal standards are shown. It can be seen, that all investigated MRPs are clearly separated

212

and can be detected in noticeable amounts. To enhance the selectivity of the mass

213

spectrometric detection, the elution time was divided into 3 different time segments 0-8 min,

214

8-11.5 min and 11.5-20 min with MRM transitions as indicated in table 1.

215

The recovery of the analytes was tested by analyzing a sample of human milk with and

216

without the addition of MRP standards. As shown in table 2, significant amounts of free

217

MRPs were present in the original sample and the addition of standard increased these figures

218

to expected levels. Recovery rates between 95 and 102% were calculated. The limits of

219

detection (LOD) were calculated from signal to noise ratios and are indicated in table 3.

220

To verify the influence of the enzymatic digestion which is needed for the analysis of protein-

221

bound MRPs in feed, a blank sample (water) and a standard mixture in water were treated

222

according to the enzymatic digestion procedure as described above. From this experiment it

223

could be seen, that lysine, which could additionally be recorded in the chromatogram (data

224

not shown), is liberated through self-digestion of the enzymes, but none of the desired target 10 ACS Paragon Plus Environment

Page 10 of 33

Page 11 of 33


225

molecules is affected by the sample preparation. The effectiveness of the enzymatic digestion

226

was determined to 91% by amino acid analysis in comparison to acid hydrolysis based on the

227

release of hydrophobic amino acids (valine, leucine, isoleucine; data not shown).

228

Finally, the potential influence of the freeze drying process, which was used to stabilize the

229

collected milk samples before analysis, was tested with cow’s milk by analyzing one sample

230

before and after drying (each in triplicate). The results showed that freeze drying caused no

231

significant decrease of most of analytes (table S2 in the supplement). The exceptions were

232

free FL (10%) and free lysine (20%). In milk there is an approximate 7000-fold molar excess

233

of lactose compared to lysine, therefore a reaction with this disaccharide, e.g. towards N-ε-

234

lactulosyllysine could explain this effect for lysine. For FL it is conceivable, that the free pair

235

of electrons at the nitrogen reacts with lactose or degradation towards lysine takes place33.

236

Free MRPs in commercially available milk

237

A total number of 32 commercially available milk samples and 2 samples taken directly on a

238

farm were analyzed for the content of free MRPs. Individual values for all samples can be

239

found as Table S3 and S4 in the supplementary material. As depicted in Figure 2, the early

240

Maillard product FL in commercially available milk can be found in concentrations between

241

112 pmol/mL and 960 pmol/mL. For CML, lower values between 47 pmol/mL and 119

242

pmol/mL were analyzed. The data for the free arginine derivative MG-H ranged between 8

243

pmol/mL and 67 pmol/mL and values between 14 pmol/mL and 1295 pmol/mL were assigned

244

to the PYR content. Especially the same order of magnitude for FL and PYR is unusual, when

245

compared to the values of their protein-bound analogues. According to literature data34,24, in

246

milk products and other food there is at least a 10 fold surplus in the content of protein-bound

247

FL compared to PYR. With the exception of FL, which we found in higher amounts, our

248

observations fit quite well to the results which were presented by Hegele et al. (2008)24. These

249

authors also showed that the technological processing has only limited influence on the

250

concentration of free MRPs. 11 ACS Paragon Plus Environment


Page 12 of 33

251

In order to compare the milk samples with respect to the production method, samples were

252

grouped according to their labeling into “organic” and “conventional” products.

253

As depicted in Figure 2, in average, slightly higher values for FL were present in the

254

conventional milk samples (median organic: 344, 95% confidence interval: 307 to 461

255

pmol/mL, mean conventional: 628 +/- 445 pmol/mL. If samples with FL concentrations

256

higher or lower than 1.5xIRQ31 (see Figure 2) are regarded as outliers, the values for organic

257

samples range from 112 to 959 pmol/ml, whereas the calculation for conventional milk

258

delivered an interval between 318 and 932 pmol/ml. Consequently, no significant difference

259

(P

Free Maillard Reaction Products in Milk Reflect Nutritional Intake of

Recommend Documents