Detection of advanced glycation end-products (AGEs) during dry-state storage of -lactoglobulin/lactose

Article abstract of DOI:10.1071/CH12442

Glycation of whey proteins is primarily affected by temperature, water activity, and pH, and leads to changes of the functional and nutritional properties of the proteins. In the case of prolonged storage of mixtures of lactose and -lactoglobulin as a model for dried dairy products under mild heat treatment (60-70°C) in a restricted water environment (a_w 0.64) at pH 7, N^ε-(carboxymethyl)lysine (CML) and furosine were formed concomitant with glycation of β-lactoglobulin. Indirect ELISA using polyclonal antibodies against advanced glycation end products (AGEs) was shown to correlate with analyses of CML using HPLC and may be used for quality control of dried dairy products. The glycation changed the solubility properties of the protein by forming polymeric carbohydrate products of β-lactoglobulin and AGEs as characterised by SDS-PAGE.

Full text of DOI:10.1071/CH12442

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2012, 65, 1620–1624
Full Paper
http://dx.doi.org/10.1071/CH12442
Detection of Advanced Glycation End-Products (AGEs)
During Dry-State Storage of b-Lactoglobulin/Lactose
A
A B
,
A
Sisse Jongberg, Michael Rasmussen,
Leif H. Skibsted,
A C
,
and Karsten Olsen
A
Department of Food Science, Faculty of Science, University of Copenhagen,
Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark.
B
Current address: Department of International Health, Immunology and Microbiology,
Panum Instituttet, University of Copenhagen, Blegdamsvej 3, DK-2200 København N,
Denmark.
C
Corresponding author. Email: ko@life.ku.dk
Glycation of whey proteins is primarily affected by temperature, water activity, and pH, and leads to changes of the
functional and nutritional properties of the proteins. In the case of prolonged storage of mixtures of lactose and
b-lactoglobulin as a model for dried dairy products under mild heat treatment (60–708C) in a restricted water environment
(aw 0.64) at pH 7, Ne-(carboxymethyl)lysine (CML) and furosine were formed concomitant with glycation of
b-lactoglobulin. Indirect ELISA using polyclonal antibodies against advanced glycation end products (AGEs) was shown
to correlate with analyses of CML using HPLC and may be used for quality control of dried dairy products. The glycation
changed the solubility properties of the protein by forming polymeric carbohydrate products of b-lactoglobulin and AGEs
as characterised by SDS–PAGE.
Manuscript received: 29 June 2012.
Manuscript accepted: 20 October 2012.
Published online: 16 November 2012.
risk to human health as discussed by Ames[8] and Sebekova and
Somoza.[9]
Introduction
Whey proteins are known for their high nutritional value and
functional properties and are used as functional ingredients in
various food products.[1] Over recent years efforts have been put
into improving the functional properties of whey proteins. One
approach has been to utilise the Maillard reaction, i.e. the non-
enzymatic reaction between a free amino group and a reducing
carbohydrate. While introduction of sugar moieties in proteins
by the Maillard reaction decreases the nutritional value by loss
of the essential amino acid lysine, glycation through early
Maillard reactions has been shown to improve the thermal sta-
bility, foaming, and emulsifying properties of the major whey
protein b-lactoglobulin (b-Lg).[2,3]
The extent of modification and the effect on protein structure
and properties during Maillard type glycation depend on the
experimental conditions such as pH, temperature, and water
activity.[4] The kinetics of lactosylation of b-Lg is greatly
accelerated when incubated with lactose in a dry matrix system
compared with solution conditions.[5,6] Furthermore, dry-state
glycation does not significantly alter the native structure of the
b-Lg, whereas glycation in solution induces important structural
changes leading to dissociation of the dimer and formation of
high molecular weight species.[5] Although glycation through
early Maillard reactions can improve the functional properties of
food proteins, the formed Amadori product might react further
to form advanced glycation end products (AGEs). AGEs have
been associated with pathophysiologies such as diabetes as
reviewed by Singh, Barden, Mori, and Beilin.[7] The metabolic
fate of dietary AGEs is still not fully revealed and may pose a
Due to the complexity and widespread distribution of AGEs
in biological systems, and the fact that there currently are no
universally accepted method to detect AGEs, makes it not only
difficult to compare data from different laboratories, but also
difficult to characterise the progression of AGE formation in a
given biological system.[7] Further, the detection of AGEs is
complicated by the decreased solubility of the resulting polymer
carbohydrate products.[2]
The most common methods used for detection are HPLC
or immunochemical methods. The use of a polyclonal antibody
in the immunochemical methods, such as the enzyme-linked
immunosorbent assay (ELISA), compromises the specificity
of the analysis, whereas the chromatographic method HPLC
is highly specific and the hyphenated technique liquid
chromatography-mass spectrometry (LC-MS) may provide
specific identification of the actual reaction products. However,
a side effect of using chromatographic methods for detection
of Maillard or AGEs is the acid hydrolysis in the sample
preparation. Acid hydrolysis has been shown to alter the
proteins, and may induce protein modifications which are not
found in the original sample matrix.[10] Ne-(carboxymethyl)
lysine (CML) and e-N-(2-furoylmethyl)-L-lysine (furosine) are
markers of Amadori products in the Maillard reaction, and are
formed during the acid hydrolysis in the sample preparation
depending on the molar concentration of acid.[11] Scheme 1
presents the reaction between the amino acid residue lysine and
lactose to yield CML and furosine.
Journal compilation Ó CSIRO 2012
www.publish.csiro.au/journals/ajc

Dry-State Glycation of Whey Proteins and AGE Formation
1621
H
O
O
C
CHOH
CHOH
CH
O
C
O
O
H
N
H
C
H
C
ϩ
H2N (CH2)4
Gal
OH
NH
H
CHOH
CH2OH
NH2
Carboxymethyllysine
Lactose
Protein-bound lysine
Hydrolysis
(6 M HCl)
O
O
H
C
H
C
HN (CH2)4
H2C
N
(CH2)4
NH
HC
NH
O
C
O
H
N
C
O
HC OH
CHOH
CH
O
H
C
OH
CHOH
O CH
Hydrolysis
(8 M HCl)
NH2
Furosine ((2-furoylmethyl)-lysine)
Gal
Gal
O
CHOH
CHOH
CH2OH
CH2OH
Schiff base adduct
Amadori product
Scheme 1. Initial stage of Maillard reaction with the formation of furosine and CML.
(a)
(b)
Time [h]
Time [h]
MW
0
1
2
4
8
12 24 36 48
MW
0
1
2
4
8
12 24 36 48
200
200
116.3
97.4
66.3
55.4
116.3
97.4
66.3
55.4
36.5
31.0
36.5
31.0
21.5
14.4
21.5
14.4
6.0
6.0
Unresolved
insulin
Unresolved
insulin
Fig. 1. Glycation and Maillard cross-linking of a b-Lg/lactose model system incubated at 608C for up to 48 h (aw 0.64, pH 7) as
determined by (a) non-reducing or (b) reducing SDS-PAGE.
In contrast to furosine, the formation of CML is not unique,
and it may be formed by other pathways than acid hydrolysis,
such as by oxidative cleavage of the Amadori product,[12] and
may hence provide additional information of the degree of
protein modifications than furosine.[10] However, to avoid
overestimation of CML during the acid hydrolysis of the
samples it is essential to perform a chemical reduction of the
protein samples before hydrolysis as proposed by Hartkopf
et al.[13] CML is identified as a major AGE antigen recognised
in tissue proteins by polyclonal anti-AGE antibodies, emphasis-
ing the usefulness of CML as a biomarker of the Maillard
reaction.[14] This suggests that detection of furosine and CML
provides information of the progression of early stages as well as
advanced stages of the Maillard reaction and the potential for
AGE formation in a given product or biological system.
identify and further develop screening methods for analysing
AGEs formed in b-Lg/lactose as a model for dry dairy products
using indirect ELISA with a polyclonal anti-AGE antibody. The
detection of CML by ELISA was correlated with sodium
dodecyl sulfate polyacrylamide gel electrophoresis (SDS-
PAGE), and reversed phase HPLC (RP-HPLC) detection of
CML and furosine, and may be used in control of stability of dry
or intermediate moisture food products.
Results and Discussion
The formation of AGEs during storage of whey proteins in the
presence of lactose was investigated in a model system prepared
from b-Lg and lactose. As presented in Fig. 1, a shift in the
electrophoretic mobility of b-Lg was observed already after 1 h
of dry-state glycation at 608C (aw 0.64). Band broadening also
occurred due to the formation of advanced Maillard reaction
products or multiple lactosylated b-Lg. A similar phenomenon
was described by Thomsen et al. as evaluated by LC-MS.[4]
As glycation is significantly accelerated in dry-state systems
compared with solution conditions it seems reasonable to expect
that formation of advanced Maillard products or AGEs might
also be accelerated. The aim of this study was accordingly to

1622
S. Jongberg et al.
1.50
1.25
1.00
0.75
0.50
0.25
0
300
250
200
150
100
50
5
4
3
2
1
0
0
0
10
20
30
40
50
0
10
20
30
40
50
Time [h]
Time [h]
Fig. 2. Formation of AGE products expressed as mean ꢁ sd in a b-Lg/
lactose model system incubated at 608C for up to 48 h (aw 0.64, pH 7)
measured by indirect ELISA using a polyclonal anti-AGE antibody. The
background of pure b-Lg is subtracted.
Fig. 3. Formation of CML expressed as mean ꢁ sd in a b-Lg/lactose model
system incubated at 608C for up to 48 h (aw 0.64, pH 7) measured by
RP-HPLC after protein hydrolysis (6 M HCl) and derivatisation with
OPA/NAC.
3.0
2.5
2.0
The increase in molecular weight stabilised after 8 h incubation,
and subsequently the formation of non-reducible higher
molecular weight compounds was observed, indicating forma-
tion of cross-linked advanced Maillard products.[15] The fact
that the polymers were non-reducible by dithiothreitol (DTT)
indicates that the polymerisation is not caused by reducible
disulfide bonds. The bands observed at ,40 kDa and ,65 kDa
might be b-Lg dimers and b-Lg trimers, respectively, and fur-
ther polymerisation products are also noticeable at higher
molecular weights (Fig. 1a, b). The band intensity of the
monomer b-Lg decreases after 12 h incubation due to either
reduced solubility of the glycated protein or formation of protein
polymerisation products.
1.5
Slope ϭ 0.344
1.0
Intercept ϭ 0.0054
R2 ϭ 0.999
0.5
0
Indirect ELISA of dry-state glycated b-Lg using a polyclonal
anti-AGE antibody (a-AGE pAB), showed a linear signal
increase up to 12 h after which the formation of AGE epitopes
halts (Fig. 2). The drop in signal intensity after 24 h may most
likely be attributed to the decreased protein solubility or inac-
cessibility of the epitopes due to cross-linking induced by
Maillard and not disulfide bonds as observed in SDS-PAGE
(Fig. 1). Beyond 48 h of incubation, formation of insoluble
aggregates prevented reliable ELISA result from being
obtained, and data are not presented.
The formation of the specific AGE-product, CML, was
determined by RP-HPLC (Fig. 3). Samples were reduced before
acid hydrolysis since acid hydrolysis has been shown to generate
artificially high levels of CML without prior reduction.[13]
Treatment with borohydride reduces lactose to lactitol, which
is unable to participate in the glycation processes with b-Lg. No
CML was detected in control samples of pure b-Lg and
lyophilised non-incubated b-Lg/lactose, but the formation of
CML in the incubated dry-state b-Lg/lactose system (Fig. 3)
followed a kinetic profile similar to that observed in ELISA
(Fig. 2).
0
1
2
3
4
5
6
7
8
9
Average # of lactosylated Lys
Fig. 4. Formation of furosine expressed as mean ꢁ sd as a function of the
average number of lactosylated lysine residues in a b-Lg/lactose dry matrix
model system incubated at 708C for up to 120 min (aw 0.64, pH 7) quantified
by RP-HPLC after acid hydrolysis (8 M HCl) of the protein.
formation of the Amadori products required to generate protein
bound AGEs (see Scheme 1). No lag-phase was observed for the
formation of AGEs as determined by the ELISA method
demonstrating the non-specificity of the polyclonal anti-AGE
antibody, which may recognise other protein modifications,
such as the Amadori compounds generated before CML forma-
tion. The HPLC method was found to have a larger dynamic
range compared with the ELISA method, which was ascribed
the increment in protein solubility during the sample preparation
by which the protein aggregates are hydrolysed. As indicated in
Fig. 3, the formation of CML stabilises after an average of five
lysine residues are converted to CML.
Furosine is another marker of Amadori compounds in the
Maillard reaction, and in contrast to CML furosine is only
formed during acid hydrolysis above 7.75 M HCl[16] (see
Scheme 1). Furosine was quantified by HPLC analysis after
acid hydrolysis (8 M HCl) of b-Lg/lactose subjected to dry-state
glycation at 708C (aw 0.64, pH 7). Figure 4 shows that the
formation of furosine correlates positively (R2 ¼ 0.999) with the
Interestingly, CML was detected after only 1 h of incubation
indicating a rapid formation of AGEs in the dry-state system.
A linear correlation (R2 ¼ 0.994) between formation of CML
and time of incubation was seen after the lag-phase correspond-
ing to a rate of reaction of 0.25 nmol CML mgꢀ1 protein minꢀ1
in the time range of 2–12 h. The lag-phase of CML formation
measured by HPLC may likely be attributed to the initial

Dry-State Glycation of Whey Proteins and AGE Formation
1623
average number of lactosylated lysine residues, as determined
by Thomsen et al.[4] The slope of the linear regression indicates
that 0.34 mol furosine is formed for each mol lactosylated
protein. This correlates with the fact that not all lactosylated
protein is hydrolysed to generate furosine, but other reaction
products such as CML may also be formed. The reaction rate for
the formation of furosine was found to be 1.6 nmol furosine
mgꢀ1 protein minꢀ1, however, the lag-phase of furosine was
shorter compared with CML suggesting that formation of
furosine correlates well with the progression of lactosylation
as described by Thomsen et al.[4]
were fixed overnight in fixation solution (50 % ethanol, 7 %
acetic acid, and 43 % Milli-Q water) at room temperature on a
rocking table. Following staining overnight by the fluorescence
SYPRO Ruby Protein Gel Stain (Invitrogen, Carlsbad, CA,
USA), the gels were photographed by a charge-coupled device
(CCD) camera (Raytest, Camilla II, Straubenhardt, Germany).
Indirect AGE-ELISA
The AGE content of incubated b-Lg was determined by indirect
ELISA using a polyclonal anti-AGE antibody. All incubation
steps were performed on a rocking platform. 96 well microplates
were coated overnight with 50 mL/well of 0.5 mg mLꢀ1 b-Lg/
lactose dissolved in water and diluted in coating buffer (50 mM
carbonate, pH 9.5). The plates were washed three times with
PBS/0.05 % TweenÒ 20 (PBST) and blocked with 300 mL/well
of 0.5 % gelatin/PBST (blocking buffer) for 2 h at room tem-
perature. The plates were washed three times with PBST and
50 mL/well of primary antibody (rabbit anti-AGE polyclonal
antibody, Biologo AGE101–0.2; Kronshagen, Germany)
diluted 1 : 10000 in blocking buffer was applied and the plates
incubated for 1 h at room temperature. The plates were
washed three times with PBST and 50 mL/well of secondary
antibody (HRP conjugated goat anti-rabbit IgG, ABIN237503;
Antibodies-online GmbH, Aachen, Germany) diluted 1 : 10000
in blocking buffer was applied. The plates were incubated for 1 h
at room temperature and washed 5 times with PBST. Aliquots of
50 mL/well of TMB ONE (Kem-En-Tec Diagnostics A/S,
Taastrup, Denmark) were applied and the plates were wrapped
in aluminium foil and incubated for 20 min at room temperature.
The reaction was stopped by adding 50 mL/well of 0.3 M H2SO4
and the absorbance measured at 450 nm in a TECAN GENios
Plus plate reader (Salzburg, Austria).
Conclusion
Detection of AGEs in b-Lg/lactose by polyclonal anti-AGE
antibody was found to correlate with the formation of CML as
determined by HPLC analysis. Detection by the immuno-
chemical methods using polyclonal antibody may apply as a
screening method for the formation of AGEs in dry-state sys-
tems. However, at later stages the decrease in protein solubility
due to aggregation compromises the detection and other meth-
ods must be applied. The results indicate that the formation of
AGE-products is an inevitable consequence of glycation of
b-Lg, and the reactions are expected to occur also under milder
storage conditions, where glycation takes place. Consequently,
if storage conditions of functionally glycated food products are
not carefully controlled it could lead to formation of undesired
AGEs.
Experimental
Sample Preparation
b-Lactoglobulin (b-Lg) was purified according to Kristiansen,
Otte, Ipsen, and Qvist[17] except that a mixture of the genetic
variants A and B was used. Lyophilised b-Lg (5 g) and lactose
(10 g) (AnalRÒ from VWR, Poole, UK) was dissolved in 50 mM
phosphate buffer (pH 7.0) in a molar ratio of 1 : 100 and the
solution was frozen overnight at ꢀ408C. The frozen solution
was lyophilised at ꢀ508C in a BOC Edwards Modylyo lyophi-
lizer (Buch & Holm, Herlev, Denmark) to generate the dry
matrix. After lyophilisation the b-Lg/lactose dry matrix was
stored at ꢀ248C until it was transferred to plastic beakers
without lid in 500 mg aliquots and placed in desiccators con-
taining KI (aw ¼ 0.64) and incubated at 608C for SDS-PAGE,
ELISA, and HPLC analysis of CML or at 708C for the HPLC
analysis of furosine. Samples were collected at different times
between 1–48 h from the 608C batch and after 30, 60, 75, and
120 min from the 708C batch. All samples were stored at ꢀ248C
until analysis. The b-Lg/lactose dry matrix as well as pure b-Lg
were used as controls.
Determination of CML by RP-HPLC
Determination of CML was done according to the method of
Drusch, Faist, and Erbersdobler[18] with minor modifications.
An aliquot of 100 mg of each b-Lg/lactose dry matrix was dis-
solved in 8 mL 0.2 M borate buffer (pH 9.5) in a 32 mL screw
cap glass tube (Schott AG) and 2 mL sodium borohydride (1 M
in 0.1 M NaOH) was added for reduction. The samples were
incubated for 4 h at room temperature. After reduction, 12 M
HCl was added dropwise to a final concentration of 6 M HCl.
The solutions were purged with nitrogen and tubes were sealed
under nitrogen atmosphere. The samples were hydrolysed by
incubation for 20 h at 1108C. The hydrolysate was filtered
through a 0.20 mm filter and HCl was removed by rotary evap-
oration, resuspended in H2O : ethanol : triethylamine (2 : 2 : 1)
and evaporated to dryness to remove traces of HCl. The dry
samples were resuspended in eluent A (48 mM phosphate
buffer, 4 v/v % methanol, pH 6.5) and filtered (0.20 mm). To
avoid HPLC column clotting, samples were purified on solid
phase extraction C18 columns (500 mg, 6 mL from Mallinckrodt
Baker B.V. Deventer, The Netherlands). The columns were
equilibrated with 5 mL methanol (analytical grade from Merck,
Darmstadt, Germany) and 2 ꢂ 5 mL eluent A. Samples were
applied (500 mL) and eluted with 4 mL eluent A : methanol
(80 : 20 v/v %).
SDS-PAGE
Polyacrylamide gel electrophoresis was performed using the
NuPAGEÒ NovexÒ Bis-Tris gel system from InvitrogenTM
(Carlsbad, CA). b-Lg/lactose samples were dissolved in water
and diluted in 1 : 1 in sample buffer (prediluted NuPAGEÒ LDS
Sample buffer (InvitrogenTM, Carlsbad, CA) with and without
50 mM DTT as reducing agent). Subsequently, a 10 mL sample
mixture was loaded onto a 12 % NuPAGEÒ acrylamide gel
(InvitrogenTM, Carlsbad, CA) together with aliquots of 2.5 mL of
Mark12TM molecular weight standard (InvitrogenTM, LC5677)
were applied to each gel. Electrophoresis was run in cassettes
containing ice-cold SDS TRIS-acetate Running buffer (Invi-
trogen, Carlsbad, CA, USA). After electrophoresis, the gels
Prior to RP-HPLC analysis hydrolysed samples were deri-
vatised with o-phthaldialdehyde/N-acetyl-L-cysteine (OPA/
NAC).[19] Samples were mixed 1 : 1 with OPA/NAC reagent
(25 mM OPA, 125 mM NAC in 0.2 M borate buffer (pH 9.5)
with 4 v/v % methanol) and after 1 h of derivatisation, the 20 mL

1624
S. Jongberg et al.
sample was injected on a RP C18 HPLC column (Zorbax
300SB-C18, Agilent Technologies). The samples were eluted
with an isocratic gradient over 6 min (100 % eluent A, 48 mM
phosphate buffer, 4 v/v % methanol, pH 6.5) followed by
a linear gradient (0–30 % eluent B, 1 mL minꢀ1) over 14 min.
The column was washed with 100 % eluent B after each run
(6 min, 1 mL minꢀ1) and re-equilibrated with eluent A over
7 min (1 mL minꢀ1). The concentration of CML in samples
was determined from a four-point linear calibration curve of
CML (Chemos GmbH, Regenstauf, Germany) ranging for
50–200 pmol CML injected. The standard was measured in
triplicates, while samples were measured in duplicates.
References
[1] S. Damodaran, in Food Proteins and their Applications (Eds
S. Damodaran, A. Paraf) 1997, pp. 57–110 (Marcel Dekker:
New York, NY).
[2] F. Chevalier, J. M. Chobert, Y. Popineau, M. G. Nicolas, T. Haertle,
Int. Dairy J. 2001, 11, 145. doi:10.1016/S0958-6946(01)00040-1
[3] A. Medrano, C. Abirached, L. Panizzolo, P. Moyna, M. C. Anon, Food
Chem. 2009, 113, 127. doi:10.1016/J.FOODCHEM.2008.07.036
[4] M. K. Thomsen, K. Olsen, J. Otte, K. Sjøstrøm, B. B. Werner,
L. H. Skibsted, Int. Dairy J. 2012, 23, 1. doi:10.1016/J.IDAIRYJ.
2011.10.008
[5] F. Morgan, D. Molle, G. Henry, A. Venien, J. Leonil, G. Peltre,
D. Levieux, J. L. Maubois, S. Bouhallab, Int. J. Food Sci. Technol.
1999, 34, 429. doi:10.1046/J.1365-2621.1999.00318.X
[6] S. J. French, W. J. Harper, N. M. Kleinholz, R. B. Jones, K. B. Green-
Church, J. Agric. Food Chem. 2002, 50, 820. doi:10.1021/JF010571Q
[7] R. Singh, A. Barden, T. Mori, L. Beilin, Diabetologia 2001, 44, 129.
doi:10.1007/S001250051591
[8] J. M. Ames, Mol. Nutr. Food Res. 2007, 51, 1085. doi:10.1002/MNFR.
200600304
[9] K. Sebekova, V. Somoza, Mol. Nutr. Food Res. 2007, 51, 1079.
doi:10.1002/MNFR.200700035
[10] H. F. Erbersdobler, V. Somoza, Mol. Nutr. Food Res. 2007, 51, 423.
doi:10.1002/MNFR.200600154
[11] J. A. B. Baptista, R. C. B. Carvalho, Food Res. Int. 2004, 37, 739.
doi:10.1016/J.FOODRES.2004.02.006
[12] M. U. Ahmed, S. R. Thorpe, J. W. Baynes, J. Biol. Chem. 1986, 261,
Determination of Furosine by RP-HPLC
An aliquot of 50 mg b-Lg/lactose dry matrix was diluted in 3 mL
8 M HCl in a 10 mL Pyrex vial with screw cap and hydrolysed
according to the method described by Resmini, Pellegrino, and
Battelli[20] with minor modifications. Nitrogen was bubbled
through the sample for 20 s, before incubating the vial at 1108C
for 23 h. The hydrolysate was purified on solid phase extraction
C18 columns (500 mg, 6 mL from Mallinckrodt Baker B.V.,
Deventer, The Netherlands). The columns were pre-wetted with
5 mL methanol (analytical grade from Merck, Darmstadt,
Germany) and 2 ꢂ 5 mL distilled water. Aliquots of 500 mL
hydrolysate was added and eluted with 4.0 mL 3 M HCl; the
eluate was injected on an Ion-pair RP column (Synergi 4 mm
4889.
[13] J. Hartkopf, C. Pahlke, G. Ludemann, H. F. Erbersdobler, J. Chroma-
togr. A 1994, 672, 242. doi:10.1016/0021-9673(94)80613-6
[14] S. Reddy, J. Bichler, K. J. Wells-Knecht, S. R. Thorpe, Biochemistry
1995, 34, 10872. doi:10.1021/BI00034A021
[15] F. Chevalier, J. M. Chobert, D. Molle, T. Haertle, Lait 2001, 81, 651.
doi:10.1051/LAIT:2001155
[16] R. Krause, K. Knoll, T. Henle, Eur. Food Res. Technol. 2003, 216, 277.
[17] K. R. Kristiansen, J. Otte, R. Ipsen, K. B. Qvist, Int. Dairy J. 1998, 8,
113. doi:10.1016/S0958-6946(98)00028-4
[18] S. Drusch, V. Faist, H. F. Erbersdobler, Food Chem. 1999, 65, 547.
doi:10.1016/S0308-8146(98)00244-1
˚
Fusion-RP 80A column; Phenomenex, Allerød, Denmark). The
,
injection volume was 20 mL, the flow rate was 0.5 mL minꢀ1
the run-time was 30 min, and the temperature was 23 ꢁ 18C. The
samples were eluted with an isocratic mobile phase consisting of
0.5 w/v % ammonium hydrogen carbonate with 1.1 v/v % acetic
acid (pH 4.4). Furosine was detected by a Shimadzu Diode
Array Detector (SPD-M10A) at 280 nm. The concentration of
furosine in samples was determined from a seven point cali-
bration ranging from 2.33–13.3 mmol Lꢀ1 furosine.
[19] D. Aswad, Anal. Biochem. 1984, 137, 405. doi:10.1016/0003-2697
(84)90106-4
[20] P. Resmini, L. Pellegrino, G. Battelli, Italian J. Food Sci. 1990, 3, 173.
Acknowledgements
This work was supported by the Danish Dairy Research Foundation and the
Danish Council for Independent Research Technology and Production
Sciences (FTP-project ‘Milk Protein Modification by Reducing Sugars’).