Assessment of enzymatic methods in the δ18O value determination of the l -tyrosine p -hydroxy group for proof of illegal meat and bone meal feeding to cattle

Article abstract of DOI:10.1021/jf201217r

The δ¹⁸O value of the p-hydroxy group of l-tyrosine depends on the biosynthesis by plants or animals, respectively. In animal proteins it reflects the diet and is therefore an absolute indicator for illegal feeding with meat and bone meal. The aim of this investigation was to perform the positional ¹⁸O determination on l-tyrosine via a one-step enzymatic degradation. Proteins from plants, herbivores, omnivores, and carnivores were characterized by their δ¹³C, δ¹⁵N, and δ¹⁸O values, the latter for normalizing the positional δ¹⁸O values. Their l-tyrosine was degraded by tyrosine phenol lyase to phenol, analyzed as (2,4,6)-tribromophenol. Degradation by tyrosine decarboxylase yielded tyramine. The δ¹⁸O values of both analytes corresponded to the trophic levels of their sources but were not identical, probably due to an isotope effect on the tyrosine phenol lyase reaction. Availability of the enzyme, easy control of the reaction, and isolation of the analyte are in favor of tyrosine decarboxylase degradation as a routine method.

Full text of DOI:10.1021/jf201217r

ARTICLE
pubs.acs.org/JAFC
Assessment of Enzymatic Methods in the δ18O Value Determination of
the L-Tyrosine p-Hydroxy Group for Proof of Illegal Meat and Bone
Meal Feeding to Cattle
Nicole Tanz,†,‡ Roland A. Werner,# Wolfgang Eisenreich,^ and Hanns-Ludwig Schmidt*,†,§
†isolab GmbH, Woelkestrasse 9/I, D-85301 Schweitenkirchen, Germany
‡Lehrstuhl f€ur Lebensmittelchemie, Technische Universit€at M€unchen, Lise-Meitner-Strasse 34, D-85350 Freising, Germany
#Institut f€ur Agrarwissenschaften, ETH Z€urich, LFW C48.1, Universit€atsstrasse 2, CH-8092 Z€urich, Switzerland
^Lehrstuhl f€ur Biochemie, Technische Universit€at M€unchen, Lichtenbergstrasse 4, D-85747 Garching, Germany
§Lehrstuhl f€ur Biologische Chemie, Technische Universit€at M€unchen, Emil-Erlenmeyer-Forum 5, D-85350 Freising, Germany
ABSTRACT: The δ18O value of the p-hydroxy group of L-tyrosine depends on the biosynthesis by plants or animals, respectively. In
animal proteins it reflects the diet and is therefore an absolute indicator for illegal feeding with meat and bone meal. The aim of this
investigation was to perform the positional 18O determination on L-tyrosine via a one-step enzymatic degradation. Proteins from
plants, herbivores, omnivores, and carnivores were characterized by their δ13C, δ15N, and δ18O values, the latter for normalizing the
positional δ18O values. Their L-tyrosine was degraded by tyrosine phenol lyase to phenol, analyzed as (2,4,6)-tribromophenol.
Degradation by tyrosine decarboxylase yielded tyramine. The δ18O values of both analytes corresponded to the trophic levels of
their sources but were not identical, probably due to an isotope effect on the tyrosine phenol lyase reaction. Availability of the
enzyme, easy control of the reaction, and isolation of the analyte are in favor of tyrosine decarboxylase degradation as a routine
method.
KEYWORDS: L-tyrosine, positional δ18O value, plant origin, animal origin, tyrosine phenol lyase reaction, tyrosine decarboxylase
reaction, meat and bone meal, bovine spongiform encephalopathy
’ INTRODUCTION
can also be due to feeding with “organic” food, it cannot be an
absolute indication for feeding with meat and bone meal. Bahar
et al.6 demonstrated that the isotope characteristics of beef are
not absolutely indicative for the diet of the animals and do even
not permit a discrimination between products from conventional
and organic farming, respectively. Therefore, the δ15N value
cannot provide an absolute indication for feeding with meat and
bone meal, especially when the very first trophic level of the food
chain, the plant material, is unknown.
In 1994, the suspicion arose that bovine spongiform encepha-
lopathy (“mad cow disease”, related to scrapie in sheep and
CreutzfeldtꢀJakob disease in humans) could be transferred
among ruminants by meat and bone meal as a protein source
in their diet. As a consequence, in December 2000 the European
Commission banned the use of this diet additive from the feed of
livestock, especially cattle.1 Since then, the majority of meat and
bone meal is burned in power stations or cement factories, yet it
is still available as organic fertilizer. Meanwhile, a European-wide
control system for the identification of prions as the inducing
agent of the disease in slaughtered cattle has been established and
is routinely practiced.2 However, so far no method exists to prove
the illegal feeding of meat and bone meal to ruminant livestock.
Isotope ratio abundance determinations have been considered
as a suitable basis. From many investigations on food chains it is
known that the δ15N value of proteins is influenced by the diet
and is raised by ∼+2% per trophic level. This offers, for example,
the opportunity to reconstruct prehistoric nutrition or to distin-
guish between omnivores and vegetarians.3 In an experiment
with chicken, a correlation of the δ15N values of the animals’
protein to the relative amount of meat and bone meal in their
feed was shown.4 In a systematic study with cattle, a Spanish
group5 found that the δ15N value of hair and milk proteins from
animals grown under conditions of intensive farming was raised
by >+6.4% relative to those of a control group and postulated
that this was a basis for a method to prove meat and bone meal
feeding. However, as a relative enhancement of the δ15N value
In contrast, we have been able to show that the δ18O value of
the phenolic hydroxyl group of L-tyrosine is an absolute indicator
for the origin of the amino acid from plant or animal bio-
synthesis.7 In plants, L-tyrosine is synthesized via the shikimic
acid pathway. Here the oxygen atom in the para-position of the
aromatic ring of the amino acid originates from water via
erythrose and has, due to an equilibrium isotope effect, a δ18O
value of ∼+27% above that of leaf water.8 On the other hand,
only animals and some microorganisms are capable of synthe-
sizing L-tyrosine from L-phenylalanine, introducing the oxygen
atom in question by a monooxygenase reaction from O2; a
kinetic isotope effect leads to a δ18O value of ∼+7%.8 Therefore,
the δ18O value of the p-OH group of L-tyrosine from any protein
must be an absolute indicator for its (partial) origin from animal
production.
Received: February 26, 2011
Revised:
July 7, 2011
Accepted: July 8, 2011
Published: July 08, 2011
r
2011 American Chemical Society
9475
dx.doi.org/10.1021/jf201217r J. Agric. Food Chem. 2011, 59, 9475–9483
|

Journal of Agricultural and Food Chemistry
ARTICLE
Any determination of this δ18O value demands the conversion
of the amino acid into a derivative, in which the oxygen atoms of
the carboxyl group are eliminated. This positional isotope
analysis has originally been performed by means of a five-step
chemical degradation with a total yield of 60%, starting from 3 g
of the amino acid.7 This method is quite laborious for a routine
assay and demands large amounts of starting material. Therefore,
the aim of the present investigation was to develop a (one-step)
enzymaticprocessand to checktheestablishmentof theprocedure
as a routine assay. Potential candidates were seen in the tyrosine
phenol lyase and the tyrosine decarboxylase reactions. Problems
anticipated with both methods were the isolation of a few
milligrams of the corresponding analyte from a large volume of
incubation medium. In addition, for the first reaction, a possible
oxygen exchange of enzyme intrinsic water with a postulated keto-
quinoid intermediate of the reaction9,10 had to be excluded. With
the second reaction, the oxygen isotope ratio analysis of the
N-containing analyte tyramine could be especially problematic,
yet this problem has quite recently been overcome.11
sourced from Aldrich, Taufkirchen, Germany, and tin and silver capsules
for the isotope ratio analysis were bought from IVA Analysentechnik,
Meerbusch, Germany. L-Tyrosine from chicken feathers and from
human hair was kindly provided by G. Fronza from the Politecnico di
Milano, Milan, Italy. Any other ordinary chemicals were purchased
locally in the highest purity available.
Sample Pretreatment, Protein Hydrolysis, and Isolation of
L-Tyrosine. The demand of oxygen for the isotope ratio determined
the amount of sample required. Usually, 10ꢀ15 g of dry protein was
needed. One hundred grams of meat was cut into pieces of ∼2 cm3 and
freeze-dried for 24 h. Then it was ground in an ordinary laboratory mill
and defatted by Soxhlet extraction with petroleum ether (bp 40ꢀ60 ꢀC)
for at least 6 h. Milk (0.5ꢀ1 L) was skimmed by centrifugation, and the
casein was precipitated at pH 4.3 with 1 N HCl. The precipitate was
centrifuged and washed three times with deionized water and then dried
by lyophilization. Aliquots of all proteins were preserved for isotope ratio
measurements.
Ten grams of dry protein was suspended in 150 mL of 6 N HCl and
4 mL of thioglycolic acid in a 250 mL two-necked flask with a reflux
condenser and gas inlet. The mixture was heated to reflux temperature
for 24 h under occasional rinsing with N2. The content was transferred to
a 500 mL round-bottom flask and evaporated to dryness in a rotatory
evaporator. The residue was resuspended three times in 250 mL of water
and dried using a rotatory evaporator. Then it was dissolved in 250 mL of
water and adjusted to pH ∼2.5 with 5 N NaOH. After the addition of 5 g
of charcoal, the suspension was heated to 80 ꢀC for 15 min under
stirring.12 Then it was filtered through filter paper and the charcoal
rinsed with 60 mL of hot water.
The combined filtrates were adjusted to pH 5 with 5 N NaOH and
concentrated until the onset of crystallization, which was completed
overnight in the refrigerator at 4 ꢀC. The precipitate was collected by
filtration, rinsed with cold water, and once recrystallized from hot water.
The purity of the product was checked by TLC [cellulose on plastic foils,
n-butanol/acetic acid/water (v/v/v 50:20:30), 0.3 mg ninhydrin in
100 mL n-butanol] and by elemental analysis in tandem with the isotope
ratio determination. Most products had a purity of >90% and were used
without further purification for incubation. A possible isotope fractiona-
tion of the process was controlled by determining the δ13C and δ15N
values of a sample of commercial L-tyrosine before and after recrystalli-
zation; they agreed within 0.1%.
’ MATERIALS AND METHODS
Protein Samples, Enzymes, and Chemicals. Despite many
attempts to obtain protein samples from meat and bone meal fed animals
or cattle infected with bovine spongiform encephalopathy, such samples
could not be obtained. It was also not possible to provide a “100% animal
originating L-tyrosine”, which would have to be synthesized by animal
L-phenylalanine hydroxylase. Hence, we had to verify the principle of our
method and the reliability of the methodology on the basis of proteins,
the origin and history of which were typical for relative amounts of plant-
and/or animal-originating L-tyrosine, respectively. For this reason, the
samples to be analyzed were selected according to the following criteria:
(a) plant proteins as starting material in the food chain, representing the
most dominant cattle feed and belonging to the main photosynthetic
plant groups; (b) samples from herbivores, mainly ruminants, with
different controlled diets; and (c) samples from omnivores and carni-
vores representing different trophic levels with respect to the L-tyrosine
source.
Wheat gluten, corn protein (zein), and soy protein were sourced from
local agricultural commerce and beef and turkey meat from a local
supermarket. Dog and cat meat samples were provided from the
Veterinarian Department and meat from a crocodile and its feed by
M. Starck, Department of Biology II, both of the Ludwig Maximilians
Universit€at, M€unchen, Germany. Milk from cattle (mixtures from many
cows) with controlled diet (i, meadowgrazing only; ii, 50%grass, 20%hay,
30% concentrate, mainly from wheat; iii, 40% concentrate of wheat, corn,
soy, rape, and molasses, 35% corn silage, 25% grasssilage) wasprovided by
D. Weiss, Freising, Germany. Goatmilk and meatfrom a suckling kid goat,
exclusively nourished by this milk, as well as the fodder (meadow, hay,
some concentrate) and the drinking water of the goat were obtained from
a local farm. A hare muscle sample was provided by a local hunter, and a
pike was provided by J. Lamina from the Animal Science Department of
the Technische Universit€at M€unchen, Freising, Germany.
Enzymatic Fission of L-Tyrosine by Tyrosine Phenol Lyase,
Isolation, and Derivatization of Phenol. As the activity of
tyrosine phenol lyase is inhibited by Na+ and Li+,10 any buffers were
prepared from KH2PO4 and KOH. L-Tyrosine equivalent to 30 mg of
phenol (0.33 mmol = 60 mg of L-tyrosine) was incubated, and the
volume of the incubation medium was adapted to the low solubility of
the substrate. The incubation procedure13,14 was modified, in that
formate dehydrogenase, lactate dehydrogenase, and formate were used
1
in excess relative to tyrosine phenol lyase; NAD+ was added at /10
equivalent. To 140 mL of 0.55 mM postassium phosphate buffer
(pH 8.3) in a 250 mL flask was added 7 mg of thioglycolic acid, and
the pH value was adjusted to 8.0 (solution A). To 15 mL of this solution
were added 2 mg of pyridoxal phosphate and 2 U of tyrosine phenol lyase
(25μL of a solutionof40 U/0.5 mLinphosphate buffer), and the solution
was preincubated at 27 ꢀC for 30 min (solution B). To the remaining
125 mLofsolution A were added 60ꢀ90 mgofL-tyrosine, 36 μL offormic
acid, and 23 mg of NAD+-dihydrate, and the solution was adjusted to pH
8.0. Then 2 U each of formate dehydrogenase and lactate dehydrogenase
were added, and the solution was brought to 27 ꢀC. It was combined with
solution B, and after 5 min of gentle stirring, the L-tyrosine was almost
completely dissolved. The incubation in a closed flask was 48 h at 27 ꢀC,
the enzymes being supplemented, units as above, after 24 h.
Tyrosine phenol lyase (EC 4.1.99.2, from Citrobacter freundii, 2 ꢁ 40
U/0.5 mL phosphate buffer) was a kind gift of R. S. Phillips, Department
of Chemistry, University of Georgia, Athens, GA. Tyrosine decarbox-
ylase (EC 4.1.1.25, apoenzyme from Streptococcus faecalis NCTC 6783,
250 mg of 0.30 U/mg and 250 mg of 0.56 U/mg) was purchased from
Worthington Biochemical Corp., Freehold, NJ. Lactate dehydrogenase
(EC 1.1.1.27, 508 U/mg), formate dehydrogenase (EC 1.2.1.2, 0.47 U/
mg), pyridoxal phosphate, NAD+-dihydrate, and (2,4,6)-tribromophe-
nol from Fluka, Steinheim, Germany, and TLC cellulose plastic sheets
and L-tyrosine from E. Merck, Darmstadt, Germany, were purchased
from a local representative. 18O-enriched water (10 atom % 18O) was
This reaction time had been elaborated by the turnover control of a
representative sample. One milliliter of incubation medium samples was
9476
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
taken for HPLC control of substrate and product concentrations at
times 0, 24, and 48 h. Thirty-four microliters of 5 M trichloroacetic acid
was added to the samples. The precipitate formed was removed by
centrifugation. Eighty-five microliters of the supernatant was diluted with
water to 1 mL. This solution was injected into a Dionex HPLC system
BioLC, consisting of an AminoPac PA-10 analytical column, 2 ꢁ 250 mm,
with an AminoPac PA 10 Guardpre column, 2 mm (10ꢀ32), equipped
with an ED 50 electrochemical detector, a GS 50 gradient pump, and an
AS 50 autosampler (Dionex GmbH, Idstein, Germany). One milliliter of
each of a 20 μM solution of L-tyrosine (substrate) and a 420 μM solution
of phenol (product) was used as standard. After 48 h, L-tyrosine had
disappeared, and this time was used for any other incubations.
After the incubation, the pH value of the medium was adjusted to 3.5
by acetic acid. The precipitate was eliminated through centrifugation.
For the isolation of the phenol the supernatant was filtered for solid
phase extraction through a 12 mL Giga Tube with 1 g Strata X 33 μm
polymeric sorbent (Phenomenex Inc., Torrance, CA),15 conditioned
with 5 mL of methanol and 5 mL of deionized water. The column was
dried with N2, and the bound phenol was eluted with 20 mL of
dichloromethane. The phenol was extracted with 20 mL of 0.5 N
NaOH, and the solution was acidified with 6 N HCl. A solution of
0.5 mL of Br2 in 10 mL of acetic acid was added dropwise until the
supernatant remained slightly yellow. The precipitate of (2,4,6)-tribro-
mophenol was isolated by filtration, washed with ice-cold water, dried by
lyophilization, and weighed into silver capsules for the oxygen isotope
ratio analysis. The yields of this product relative to the applied L-tyrosine
did not exceed 70%.
To check the isolation procedure for isotope fractionations, we
determined the δ18O value of a sample of phenol and then prepared a
solution of ∼120 mg of this phenol in 200 mL of H2O and divided it into
two equal parts. In one of them, the derivatization to (2,4,6)-tribromo-
phenol was directly performed; the other part was submitted to the solid
phase and solvent extraction procedures prior to precipitation. The
δ18O values of the two (2,4,6)-tribromophenol samples were identical
within 0.1%, but 0.5% more negative than that of the original phenol.
This is above the SD of the isotope ratio measurement ((0.2%) but
within the overall reproducibility of the complex procedure (∼(1.0%).
For the control of a possible oxygen exchange of the keto-quinoid
intermediate with water, in some experiments 0.625 mL of the incuba-
tion medium was replaced by 18O-enriched water (10 atom % 18O). Five
milliliters of the labeled medium was preserved for the O-isotope ratio
analysis. After the isolation of the phenol, the solution was adjusted to
pH 9.3 and the water was recovered by distillation. The distillate was
cleared with 5 g of charcoal and, after the addition of another 0.045 mL
of 10% 18O-labeled water, reused for the preparation of another labeled
incubation medium.
Enzymatic L-Tyrosine Decarboxylation and Isolation of
Tyramine. With reference to Ziadeh et al.16 100 mg of L-tyrosine was
added to 100 mL of 0.1 M of acetate buffer (pH 5.5). Twenty milligrams
(6.0 or 11.2 U) of the tyrosine decarboxylase preparation (a raw dry
powder of a bacterial homogenate) and 0.5 mg of pyridoxal phosphate
were preincubated in another 5 mL of the same buffer at 37 ꢀC for 15 min.
Afterward, the two solutions were combined in a closed incubation
vessel, the gas phase of which was connected via an S-shaped glass tube
to a water-filled calibrated cylinder to control the turnover of the
substrate by volumetric measurement of CO2 production. The medium
was stirred at 37 ꢀC . After 20 h another 9 U of enzyme were added and at
maximum 36 h the gas production indicated the quantitative conversion
of the substrate (calculated as ∼14.0 mL for 100 mg of L-tyrosine). The
medium was centrifuged to eliminate the solid parts of the enzyme
preparation. The supernatant was transferred to a 300 mm ꢁ 10 mm i.d.
column with 10 g of Dowex 1X8 anion exchange resin, 200ꢀ400 mesh,
in Clꢀ form, retaining the acetate and pyridoxal phosphate ions. The
product was eluted with 100 mL of water (identified by ninhydrin for
tyramine and AgNO3 for Clꢀ). The tyramine chloride containing
fractions were combined and lyophilized. The remainder contained
unknown impurities from the enzyme preparation and had to be
submitted to preparative HPLC for further purification by a Knauer
HPLC preparative system with a Nucleosil 100 C18 column, 250 mm ꢁ
20 mm i.d. (Macherey-Nagel, D€uren, Germany), a K-1800 WellChrom
preparative pump, and a K2501 UV detector (Knauer, Berlin, Germany).
As tyramine is only partially soluble in water, the procedure had to be
repeated in up to 18 portions, dissolved in a few milliliters of H2O under
the addition of some drops of 6 N HCl. Tyramine was eluted with
deionized water (40 mL/min at 3 MPa). The fractions containing
tyramine were identified by their UV absorption at 254 nm. In total, the
product had to be isolated from several hundred milliliters of eluate by
lyophilization. The residue, a mixture of tyramine, tyramine hydrochlor-
ide, and NaCl, was directly used for isotope ratio analysis.
Isotope Ratio Analysis of Proteins, L-Tyrosine, Tyramine,
(2,4,6)-Tribromophenol, and Water. The indications of the
isotope ratios are given, relative to international standards from the
International Atomic Energy Agency (IAEA) in Vienna, Austria (VPDB
for carbon, AIR for nitrogen, VSMOW and IAEA 600, 601, and 602 for
oxygen), in δ value notations, based on isotope ratios R = [heavy isotope]/
[main isotope], for example, for carbon:
δ13C ¼ ðRSample=RVPDB ꢀ 1Þ ꢁ 1000 ð%Þ
ð1Þ
The determinations of the δ13C and δ15N values of the proteins and
the L-tyrosine samples were performed in a multielement isotope ratio
analyzer.17 The device simultaneously provided the elemental concen-
trations of C and N. Three milligrams of the samples in tin capsules were
measured against a casein laboratory standard (δ13CVPDB = ꢀ23.4%
15
and δ
N
= 6.2%); the analytical precision limits were for δ13C
AIR
(0.1% and for δ15N (0.2%. The δ18O values of the proteins and most
of the tyramine samples were determined at ETH Z€urich, and some
tyramine samples were analyzed by H.-J. Kupka, Elementar Analysen-
systeme GmbH, Hanau, Germany.11 Laboratory references calibrated
against VSMOW and other international standards (IAEA 600, IAEA
601, IAEA 602) were used for control, and the SD was (0.2%.
For the δ18O value determination of (2,4,6)-tribromophenol at the
Max-Planck-Institute for Biogeochemistry, Jena, Germany, samples
corresponding to 0.3ꢀ1.0 mg of O in silver capsules were converted
by high-temperature conversion at 1400 ꢀC. The CO formed was
analyzed in a mass spectrometer against laboratory standards, calibrated
to VSMOW and other international standards. Error limits are given
with the results. The δ18O value measurements of the incubation water
were performed by Hydroisotop GmbH, Schweitenkirchen, Germany.
Five milliliters of the medium was equilibrated with CO2, the δ18O value
of which was determined relative to a laboratory standard, calibrated to
VSMOW and other international standards. The error limit was 0.2%.
’ RESULTS AND DISCUSSION
Origins and Isotopic Characteristics of the Proteins, δ13C
and δ15N Values. The δ13C values of the plant proteins
(Table 1) are typical for C3- and C4-plants. As expected, the
δ13C values of the meat and casein from cattle and goat, which
had been fed exclusively with hay and grass, and of the meat of the
hare, living on local C3-plants, are nearly identical or quite close
to those of these plants. Any other animal proteins show a
distinct 13C enrichment, probably due to a more or less greater
direct or indirect presence of corn (products) in the diets. The
diet influence must be overlapped by a trophic shift (usually
1ꢀ2.5% per level).3 The unexpected 13C depletion of the
crocodile meat relative to the beef feed can be explained only
9477
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
Table 1. Origin, History, and Isotope Characteristics of Protein Samples Used for the Proof of a Correlation between the Trophic
State of L-Tyrosine and the δ18O Value of Its p-OH Group
protein, name, origina
history, dietb
δ
C
VPDB
(%)
δ
N
(%)
δ
O
(%)
Δδ
O
c (%)
VSMOW
13
15
18
18
AIR
VSMOW
wheat gluten, comm
zein, comm
cattle feed, tec qual
cattle feed, tec qual
cattle feed, tec qual
unknown
ꢀ28.0
ꢀ14.8
ꢀ25.2
ꢀ22.1
ꢀ23.2
ꢀ27.1
ꢀ27.3
ꢀ21.3
ꢀ26.7
ꢀ25.7
ꢀ28.2
ꢀ23.5
ꢀ21.8
ꢀ22.0
ꢀ30.1
+1.3
+2.0
+0.6
+4.6
+5.4
+5.4
+5.3
+5.6
+5.2
+6.5
+7.9
+1.3
+7.5
+7.3
+25.1
+22.5
+18.9
+10.4
+9.7
soy protein, comm
chopped beef, comm
cow's milk casein, comm
cow's milk casein 1
cow's milk casein 2
cow's milk casein 3
goat's milk casein
kid goat (muscle)
hare (muscle)
(0
unknown
ꢀ0.7
ꢀ0.2
ꢀ1.3
+0.4
ꢀ1.9
ꢀ1.7
+0.4
ꢀ0.3
+1.1
ꢀ0.5
ꢀ1.7
+2.2
exclusively grass
grass, hay, conc
corn + grass silage, conc
grass, hay, conc
exclusively goat milk
local plants
+10.2
+9.1
+10.8
+8.5 (ꢀ10.3)d
+8.7
+10.8
turkey (muscle), comm
dog (muscle)
unknown
+10.1
unknown
+11.5
cat (muscle)
unknown
+9.9
pike (muscle), local river
crocodile (muscle), terra
unknown
+16.5
+5.4 [+4.7]e
+8.7 (ꢀ11.1)d
+12.6
beef
ꢀ23.4 [ꢀ21.5]e
a comm, commercial; terra, terrarium. b tec qual, technical quality; conc, concentrate. c δ value shift of the sample relative to the average δ18O value of the
animal proteins (+10.4%). d Data in parentheses are of the corresponding drinking water. e Data in square brackets are for the diet beef. The SE is
e0.3% for all isotope measurements.
by the assumption that the sample investigated was not repre-
sentative for the ordinary diet of the animal.
from +20 to +26% can be expected for the plant proteins
investigated in this work. The experimental δ18O values in Table 1
agree excellently with this expectation. An estimation of the δ18O
values of carbohydrates and hence of the p-OH group of plant
L-tyrosine (shift +27 ( 2% above water) suggests δ18O values
between +25 and +29%.
With regard to the nitrogen data, the low δ15N value of the soy
protein is typical for a legume; any other δ15N values are
probably dominated by the actual fertilization. A trophic shift
can be seen in the δ15N values from plants to cattle products and
carnivores (usually 2ꢀ4% per trophic level)3 and even between
the goat's milk casein and the kid goat meat. The unexpected high
15N enrichment of the hare muscle protein can tentatively be
attributed to autocoprophagy, recorded for these animals, or
simply to the possibility that a main part of the animal’s diet was
plants from a field fertilized with organic manure. This under-
scores our reservation against the suitability of the δ15N value as
an indicator for proof of feeding meat and bone meal.
A similar estimation of the δ18O value of animal proteins is by
far more complex. The source of the major oxygen component,
the peptide oxygen, is the “cell water”, a product of the local
drinking water, the diet, and oxidation water. No data exist about
correlations between the δ18O values of animal tissue water, as
isolated, for example, by meat squeezing, and body proteins.
Aside from the drinking water, mainly the diet has a large
influence on the δ18O value of tissue water,21 also manifested
by differences between tissue water of different species from the
same area.22 These observations are confirmed by the finding that
the δ18O value of cattle hair from animals grazing in grassland in
our area was +11 ( 1%, but +7 ( 1% in winter, when the
animals’ diet was hay and grass silage.23 Collagen should be slightly
“lighter” due to its content of Hypro with the OH group
introduced by a monooxygenase reaction. Our data in Table 1
are, for most herbivores (mean value +10.4 ( 1%), in line with
these results. Some differences may be inferred from the water
source. That of the crocodile might have been water from the
terrarium, 18O enriched by evaporation. As to the relative deple-
tion of the pike protein (δ18O value of river water ꢀ11.1%), one
has to take into account that the δ18O value of dissolved oxygen in
an aquatic system, hence the oxidation water, depends on the
photosynthetic activity in and the oxygen saturation of the water
and can be quite different from that of atmospheric O2.21,24 The
δ18O value of the goat products cannot be explained.
Origins and Isotopic Characteristics of the Proteins, δ18O
Values. The δ18O values of the proteins are most interesting in
the present context. Proteins contain approximately 25% O. The
majority of the oxygen atoms are bound in the carbonamide
function, a smaller part being bound in carboxyl (Glu, Asp) and
hydroxyl groups (Ser, Tyr, Hypro) of side chains. Free carboxyl
groups show, at equilibrium, an 18O enrichment of ∼+19%
relative to the surrounding water.8 The oxygen atoms of peptide
groups are expected to be more enriched, probably up to
∼+24%, due to an isotope effect on the amino acid activation
for the protein biosynthesis. The δ18O values of the OH groups
of Ser and plant-originating Tyr should be identical to those of
carbohydrates, namely ∼+27% above water. Taking these facts
into account, we estimate the average 18O enrichment of plant
proteins relative to the mean δ18O value of leaf water to be
∼+23%. Yet this value is plant-typical and underlies (local)
climatic, seasonal, diurnal, and plant physiological parameters.18,19
As an example, the mean value of the leaf water of a Rhus typhina
tree from this area was ꢀ2.2%.20 Average values for grain plants in
our surroundings are between ꢀ3 and +2% (A. Rossmann,
private communication), and for soy plants with lower transpira-
tion a value of ∼ꢀ5% can be assumed. From this, δ18O values
An estimation of the δ18O values of animal proteins on the
basis of the drinking water alone (δ18O value ꢀ10.5 ( 0.5% in
our area) or of known data on tissue water (ꢀ4 ( 2%)22,25
would lead to unsatisfactory results. Therefore, our data must
definitely in part be caused by organic material of the diet of the
9478
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
animals. This provides an opportunity to correct the uncertain-
ties on the experimental δ18O values of L-tyrosine contributed by
the natural δ18O value range of the original plant diet ((2% on
27% for carbohydrates and the phenolic group of L-tyrosine).
They not only directly influence the plant-originating L-tyrosine
p-OH group but also indirectly influence the protein δ18O values.
Hence, in reverse, the shifts of the latter from the average value
+10.4% (Δδ values) (Table 1) should be a basis for the
“normalization”of the experimental δ18O values of the p-OH
group of L-tyrosine, obtained by hydrolysis of the proteins.
Assessment of the Tyrosine Phenol Lyase Reaction for the
Positional δ18O Value Determination. Tyrosine phenol lyase
(β-tyrosinase, EC 4.1.99.2) is a bacterial pyridoxal phosphate
dependent enzyme, for example, from Escherichia intermedia
A-2113 or from Citrobacter freundii.9,10 The catalyzed reaction,
degrading the amino acid into phenol, pyruvate, and ammonia, is
reversible, mainly due to the relatively low solubility of L-tyrosine
in water at pH 7.0. Hence, the enzyme has been used to
synthesize isotopically labeled (substituted) L-tyrosine from
(substituted) phenol, pyruvate, and ammonia.26 To force the
reaction to a quantitative fission of L-tyrosine, we coupled it to an
irreversible enzymatic system, eliminating pyruvate by its reduc-
tion to lactate with NADH, regenerated by the formate dehy-
drogenase (FDH) reaction (TPL = tyrosine phenol lyase, LDH =
lactate dehydrogenase):
Table 2. Dependence of the δ18O Value of the p-OH Group
of L-Tyrosine from Degradation by Tyrosine Phenol Lyase on
the Incubation Medium
18
18
δ
O
(%)
δ
O
(%)p-OH
VSMOW
VSMOW
L-tyrosineb
protein and origina
casein, comm
incubation water
ꢀ10.0
+238.5
+20.4 ( 0.9
+22.6 ( 0.2
wheat gluten
beef comm
ꢀ10.0
+27.3 ( 0.1
+24.0 ( 0.2
+156.2
ꢀ10.0
+132.1
+23.9 ( 0.1
+22.4 ( 0.7
a comm, commercial. b The SD relates to the isotope ratio measurement.
quite probable alternative, as the water molecules are situated
closer to the side chain than to the para-position of the aromatic
ring of the substrate.9 (iii) The half-life of the intermediate is
far below that of the exchange reaction (>CO + HOH T
>C(OH)2). As a matter of fact, the minimum half-life of oxygen
exchange between carbonyl or quinoid groups and water is on the
order of minutes or seconds,27,28 whereas that of the keto-
quinoid intermediate in question must be on the order of
milliseconds. Furthermore, the “concentration” of the two
reactants is certainly not in favor of a fast exchange reaction. In
conclusion, our results are in agreement with the structure of the
active site of the enzyme and the kinetic probability of the
exchange process. Therefore, the tyrosine phenol lyase reaction
is suitable for our purpose, but this had to be experimentally
proved.
HOꢀC6H4ꢀCH2ꢀCHðNH2ÞꢀCOOH þ H2O
s
s
TPL
HOꢀC H þ CH ꢀCOꢀCOOH þ NH
F
R
6
5
3
3
CH3ꢀCOꢀCOOH þ NADH þ Hþ
s
s
LDH
CH ꢀCHOHꢀCOOH þ NADþ
F
R
3
Phenol as the Analyte for the Positional Oxygen Isotope
Analysis of L-Tyrosine. First, the selected L-tyrosine samples
were characterized by their bulk δ13C and δ15N values (Table 3)
and compared with the original proteins (Table 1). With two
exceptions, all amino acids are depleted in 13C by a few %
relative to the corresponding proteins. This is in line with general
experiences for products of the shikimic acid pathway.29 Whereas
the δ15N values of most plant and herbivorous L-tyrosine samples
are quite similar to those of the corresponding proteins, samples
from the omnivorous and carnivorous groups show positive and
negative 15N shifts relative to their protein, respectively. This can
be attributed to their different origins and the complex metabo-
lism of L-tyrosine.
Most informative in the present context are the differences
between the positional δ18O values of L-tyrosine (Table 3) and
the bulk δ18O values of their proteins (Table 1). As outlined
above, the positional δ18O values of the plant L-tyrosine samples,
correlated to leaf water, are expected to be +27 ( 2%, confirmed
by the experimentally found value of 27.5%. Leaf water is also
the main source of the protein oxygen; however, the expected
shift for plant proteins is +23 ( 2.5%, also experimentally
confirmed (Table 1). In contrast, most positional δ18O values
for L-tyrosine from animal-originating samples differ from those
of their proteins by 5ꢀ14%. This is in line with the expected
origin of the oxygen in the animal proteins from “cell water” and
diet and of the phenolic OH group of L-tyrosine in the proteins
from plant leaf water, animal diet, and O2. Among the L-tyrosine
samples of casein from milk of cows with controlled diet, the
δ18O value of the product from animals with corn and grass silage
NADþ þ HCOOH
s
NADH þ Hþ þ CO2
f
FDH
L-Tyrosine samples from different origins were incubated at pH
8.0 with the enzyme under these conditions for 48 h at 27 ꢀC.
The phenol formed was isolated and analyzed as (2,4,6)-
tribromophenol.
Reaction Mechanism and Possible Oxygen Exchange of
the Product with Water. The mechanism of the tyrosine phenol
lyase reaction9,10 includes the formation of a substrate keto-
quinoid intermediate in the enzymeꢀsubstrate complex. As the
latter contains also two molecules of water in the active site, it
would be of importance to study whether the intermediate can
exchange oxygen atoms with this water. Therefore, the reaction
was first studied in the presence of oxygen-18-labeled water
(Table 2). The data demonstrate unequivocally that the 18O
content of the medium water does not affect the δ18O value of
the phenolic OH group of L-tyrosine. The differences between
the samples obtained in normal and 18O-labeled water suggest
that they originate from experimental errors due to the complex
analyte preparation.
For the interpretation of the result with respect to the structure
of the enzyme and the mechanism of the catalyzed reaction,9,10
the following three alternatives have to be discussed: (i) The two
water molecules in the active site do not equilibrate with the
water of the medium. A definite decision about this possibility
cannot be drawn from our result. (ii) The water molecules are
fixed at their position and do not come into contact with the
oxygen atom of the carbonyl group of the intermediate. This is a
9479
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
Table 3. δ13C and δ15N Values of L-Tyrosine Samples and δ18O Values of Its p-OH Group, Analyzed on (2,4,6)-Tribromophenol
from Phenol by the Tyrosine Phenol Lyase Degradation of L-Tyrosine
protein
L-tyrosine
L-tyrosine (p-OH group)
18
ꢀΔδ
O
VSMOW
13
15
18
18
name
and origina
history,
dietb
δ
C
δ
N
δ
O
(%),
values
δ
O
VSMOW
f ꢁ 100 (%),
for f see eq 2
VPDB
AIR
VSMOW
exptlc
(%)
(%)
(%) from Table 1
(%), norme
wheat gluten, comm
zein, comm
cattle feed, tec qual
cattle feed, tec qual
unknown
ꢀ27.8
ꢀ13.2
ꢀ26.4
ꢀ25.8
ꢀ30.0
ꢀ30.3
ꢀ25.1
ꢀ29.8
ꢀ30.5
ꢀ26.7
ꢀ21.6
+5.1
+2.6
+2.3
+6.8
+5.6
+5.5
+5.5
+5.4
+5.5
+3.0
ꢀ2.0
+27.3 ( 0.1
+27.6 ( 0.1
+23.9 ( 0.1
+23.5
+27.3
+27.6
+23.9
+24.2
+19.5
+20.6
+21.8
+20.3
+19.7
+21.5
+24.9
chopped beef, comm
cow's milk casein, comm
cow's milk casein 1
cow's milk casein 2
cow's milk-casein 3
goat's milk casein
(0
15.5
14.0
37.5
32.0
26.0
33.5
36.5
27.5
10.5
unknown
+0.7
+0.2
+1.3
ꢀ0.4
+1.9
+1.7
+0.3
?
exclusively grass
diet grass, hay, conc
corn + grass silage, conc
grass, hay, conc
exclusively goat milk
unknown
+19.3 ( 0.8
+19.3 ( 0.8
+22.2 ( 0.8
+18.4 ( 1.1
+18.0 ( 1.8
+21.2 ( 0.2
+24.9 ( 0.5
(+18.9 ( 0.4)d
+19.6 ( 0.4
(+17.8 ( 0.2)d
+18.3 ( 0.6
+17.7
kid goat (muscle)
turkey (muscle), comm
chicken (feathers), China unknown
human (hair), China
unknown
ꢀ28.3
+1.4
?
+19.6
37.0
dog (muscle)
unknown
unknown
beef
ꢀ26.2
ꢀ26.5
ꢀ27.0
+5.9
+5.5
+6.9
ꢀ1.1
+0.5
ꢀ2.2
+17.2
+18.2
+15.5
49.0
44.4
57.5
cat (muscle)
crocodile (muscle), terra
+17.7 ( 1.1
a comm, commercial; terra, terrarium. b conc, concentrate; tec qual, technical quality. c exptl, experimental. d Values by chemical degradation (see ref 7).
e norm, normalized.
diet is the most positive one, indicating the highest percentage of
“plant L-tyrosine”. This result excludes a potential contribution of
“animal L-tyrosine” from the silage feed but in contrast suggests
that the silage process and/or the rumen fermentation may even
contribute to the “plant L-tyrosine”. In addition, the result again
confirms theinfluenceof the dietonthe δ18O value of thesamples’
OH group and justifies their correction by “normalization”.
The individual fractions fof “animal L-tyrosine”(δ18Ovalue+7%)
in the proteins can be calculated from the normalized δ18O values
(δ18Onorm) of the L-tyrosine of the samples. With an average δ18O
value of “plant L-tyrosine” (fraction 1 ꢀ f) of +27%, δ18Onorm is
enzyme from S. faecalis used in this work decarboxylates
L-tyrosine and L-3,4-dihydroxyphenylalanine, but is inactive versus
L-phenylalanine and L-tryptophan. The turnover rate of the
catalyzed reaction (TDC = tyrosine decarboxylase)
HOꢀC6H4ꢀCH2ꢀCHðNH2ÞꢀCOOH þ H2O
s
f
HOꢀC H ꢀCH ꢀCH ꢀNH þ CO
6
4
2
2
2
2
TDC
can, at the incubation pH of 5.5, easily be controlled by
measuring the volume of the produced CO2. Ziadeh et al.16 used
enzymes from S. faecalis for the selective conversion of
L-phenylalanine and L-tyrosine into the corresponding amines
for a selective δ15N value determination. Although the authors
found no difference between the δ15N values of the amino acids
and the extracted amines, we do not believe that the applied
ethanol extraction of tyramine at pH 5.5 is quantitative, because
the compound is amphoteric. Therefore, we tested anion ex-
change chromatography, as we assumed that this method would
be suited to isolate the analyte tyramine in one step. However,
impurities from the raw enzyme preparation forced us to add a
second purification step, preparative HPLC, which in its turn
demanded, because of the low solubility of tyramine in water, the
manipulation of large solvent amounts. The undefined composi-
tion of the resulting analyte, a mixture of tyramine, tyramine
hydrochloride, and NaCl, demands the provision of a purer
enzyme preparation for the application of the method as a routine
assay with ion exchange chromatography as the sole purification
step. A general serious problem has been so far the oxygen isotope
ratio determination of N-containing organic analytes, which had
only been overcome by high-temperature conversion. The isotope
ratio determinations of the tyramine samples have been performed
with this method at ETH Z€urich. However, on the basis of a recent
δ18Onorm ð%Þ ¼ ½ð1 ꢀ fÞ ꢁ 27 þ f ꢁ 7ꢂð%Þ
f ¼ ð27 ꢀ δ18OnormÞ=20
ð2Þ
The expression of f in % ( f ꢁ 100) (Table 3) shows that the
proteins from herbivores and some omnivores contain between 15
and 40% “animal L-tyrosine”, among them the goat kid at the upper
limit, whereas the carnivorous proteins attain values >40%. This result
is convincing proof for the correctness of the principle of the method.
The L-tyrosine samples from human hair and chicken feathers are from
China, and their preparation is unknown. The difference between the
data found by biochemical and chemical analyte preparation is
tentatively attributed to an isotope effect on one of them.
Assessment of the Tyrosine Decarboxylase Reaction for the
Positional δ18O Value Determination, Tyramine as Analyte.
Tyrosine decarboxylase (EC 4.1.1.25) is a pyridoxal phosphate
dependent enzyme occurring in plants, insects, and mammals;
enzymes from different origins have different substrate specifi-
cities. The enzyme from animals, involved in the metabolism of
neurotransmitters, degrades only aromatic amino acids.30 The
9480
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
Table 4. δ13C and δ15N Values of L-Tyrosine Samples and δ18O Values of Its p-OH Group, Analyzed on Tyramine from the
Tyrosine Decarboxylase Degradation of L-Tyrosine
protein
L-tyrosine
L-tyrosine (p-OH group)
f ꢁ 100 (%),c for f see eq 2
13
15
18
18
18
name
and origina
history/
dietb
δ
C
δ
N
δ
O
ꢀΔδ
O
δ O
VSMOW
VPDB
AIR
VSMOW
VSMOW
(%)
(%)
(%), exptl (%) values from Table 1 (%), norme
TDC
TPL
zein, comm
cattle feed, tec qual
cattle feed, comm
exclusively grass
ꢀ13.2
ꢀ28.0
ꢀ30.0
+2.6
+0.8
+5.6
+5.5
+5.5
+1.4
+5.5
+6.9
+30.4 ( 0.3
+28.5 ( 0.5
+24.9 ( 0.7
+24.2 ( 0.2
+20.0 ( 0.4
+23.1 ( 0.5
+19.3 ( 0.8
+21.2 ( ?
+30.4
+28.5
+25.1
+23.8
+21.7
+23.1
+19.8
+19.0
soy protein, comm
cow's milk casein 1
cow's milk casein 3
kid goat (muscle)
human (hair), China
cat (muscle)
+0.2
ꢀ0.4
+1.7
?
9.5
16.0
26.5
19.5
36.0
40.0
37.5
26.0
36.5
37.0
44.4
57.5
corn + grass silage, conc ꢀ25.1
exclusively goat milk
unknown
ꢀ30.5
ꢀ28.3
ꢀ26.5
ꢀ27.0
unknown
+0.5
ꢀ2.2
crocodile (muscle), terra beef
a comm, commercial; terra, terrarium. b tec qual, technical quality; comm, commercial; conc, concentrate. exptl; experimental. d TDC, by tyrosine
c
decarboxylase; TPL, by tyrosine phenol lyase, Table 3. e norm, normalized.
methodological development,11 a simpler and more reliable perfor-
mance will be available in the future.
We incubated preferably samples that had already been
analyzed by the tyrosine phenol lyase method. Again, the general
agreement of the results (Table 4) with the theory confirms the
correctness of the concept and the suitability of the method. Only
the positional δ18O value of L-tyrosine from zein exceeds the
expected limit for the amino acid from plants (+27 ( 2%). As
also the relatively high δ18O value of the casein from milk of cows
fed corn silage (cow's milk casein 3) is in line with this result
(Table 1), these results may be due to the special water manage-
ment of C4-plants, again demonstrating the influence of the
organic part of the diet on the δ18O values of the animal
products. These findings will be taken into account in the future
for the normalization of the experimental δ18O values. Also, any
other data in Table 4 appear to be reasonable.
However, the absolute δ18O values by the tyrosine decarboxylase
process are generally at least by 2% more positive and the fvalues more
than 10% below those obtained by the tyrosine phenol lyase method
(Table 3) for the same samples. The suspicion arises that a systematic
error in one of the methods or between them might be the reason. As
the L-tyrosine conversions had been quantitative and the isolation
methods of the analytes do not imply, as far as could be controlled,
isotope fractionations, the only explanation that we can offer for this
discrepancy is an oxygen isotope effect on the tyrosine phenol lyase
reaction itself, probably on the conversion of the of the COH bond into
a CO bond at the formation of the keto-quinoid intermediate.
Proposal for Practical Application. The practical check of
the content of “animal L-tyrosine” in unknown samples should be
performed by means of a graph showing the relative fraction “f ꢁ
100” in percent as a function of the normalized δ18O value of the
sample (Figure 1). A straight line between the extreme value for a
100% “animal L-tyrosine”, +7% for any oxygen atom introduced by
a monooxygenase reaction as represented by that of vanillin,8 and
the mean theoretical and experimental value for plant originating
L-tyrosine, +27%, will permit the normalized experimental data to
be assigned to the “degree of carnivorous feed”. Independent from
proof of the absolute correctness of the data, the most complete
results actually exist from the tyrosine phenol lyase method, and
they will be used here for a demonstration of the practical check of
meat and bone meal feeding on unknown samples.
Figure 1. Relative fraction f ꢁ 100 (%) of animal originating L-tyrosine
from individual protein samples as a function of the normalized δ18O
values of the p-OH group of the amino acid: data from L-tyrosine
degradation by tyrosine phenol lyase (A) and by tyrosine decarboxylase
(B), respectively. The theoretical δ18O values for the extreme values are
symbolized by solid squares. That of “animal L-tyrosine” (+7 ( 1%) is
represented by vanillin, and that of “plant L-tyrosine” (+27 ( 2%) is
identical with several experimental data. Comm, commercial. For space
reasons the names of some samples are abbreviated.
As the actual (f ꢁ 100) values for herbivores, mainly cattle, and
some omnivores are below 40% and those for carnivores between
9481
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
40 and 55%, values for cattle above 40% would be suspicious. A
δ18O value shift of +2% corresponds to an increase of 10% in
animal protein in the feed source. However, at present this is also
near the experimental error limit, but we are convinced that an
optimization will soon be possible.
Muscle Foods Analysis; Nollet, L. M. L., Toldrꢀa, F., Eds.; CRC Press,
Taylor and Francis Group: Boca Raton, FL, 2009; pp 767ꢀ787.
(4) Carrijo, A. S.; Pezzato, A. C.; Ducatti, C.; Sartori, J. R.; Trinca, L.;
Silva, E. T. Traceability of bovine meat and bone meal in poultry by
stable isotope analysis. Braz. J. Poult. Sci. 2006, 8, 63–68.
(5) Delgado, A.; Garcia, N. δ15N and δ13C analysis to identify cattle
fed on feed containing animal proteins. A safety/quality index in meat,
milk and cheese. Contribution to the 6th International Symposium on Food
Authenticity and Safety, Eurofins, Nantes, 2001, 28ꢀ30.
Summary. The present results confirm the correctness of the
principle of using the δ18O value of the p-OH group of L-tyrosine
from animal proteins, normalized by means of their δ18O values, as an
absolute indication of the protein source in the animals’ diet. There-
fore, they provide proof of the illegal use of meat and bone meal in
feed. They also demonstrate that a one-step enzymatic method is
realistic for the positional oxygen isotope ratio analysis on the amino
acid. Of the two methods tested, the tyrosine decarboxylase process
offers the best prospect to be implemented as a routine method. The
enzyme is commercially available and, provided the provision of a
purer product, the tyramine isolation can be simplified to a one-step
procedure. A special advantage is also the easy volumetric turnover
control of the reaction. Finally, a recent advance in the oxygen isotope
ratio determination of N-containing analytes11 eliminates any pro-
blem in the oxygen isotope ratio analysis of tyramine.
(6) Bahar, B.; Schmidt, O.; Moloney, A. P.; Scrimgeour, C. M.;
Begley, I. S.; Monahan, F. J. Seasonal variation in the C, N, and S stable
isotope composition of retail organic and conventional Irish beef. Food
Chem. 2008, 106, 1299–1305.
(7) Fronza, G.; Fuganti, C.; Schmidt, H.-L.; Werner, R. A. The δ18O-
value of the p-OH group of L-tyrosine permits the assignment of its origin
to plant or animal sources. Eur. Food Res. Technol. 2002, 215, 55–58.
(8) Schmidt, H.-L.; Werner, R. A.; Rossmann, A. 18O pattern and
biosynthesis of natural plant products. Phytochemistry 2001, 58, 9–32.
(9) Sundararaju, B.; Antson, A. A.; Phillips, R. S.; Demidkina, T. V.;
Barbolina, M. V.; Gollnick, P.; Dodson, G. G.; Wilson, K. S. The crystal
structure of Citrobacter freundii tyrosine phenol-lyase complexed with
3-(40-hydroxyphenyl)propionic acid, together with site-directed muta-
genesis and kinetic analysis, demonstrates that arginine 381 is required
for substrate specificity. Biochemistry 1997, 36, 6502–6510.
(10) Sundararaju, B.; Chen, H.; Shilcutt, S.; Phillips, R. S. The role of
glutamic acid-69 in the activation of Citrobacter freundii tyrosine phenol-
lyase by monovalent cations. Biochemistry 2000, 39, 8546–8555.
(11) Sieper, H.-P.; Kupka, H.-J.; Lange, L.; Rossmann, A.; Tanz, N.;
Schmidt, H.-L. Essential methodological improvements in the oxygen
isotope ratio analysis of N-containing organic substances. Rapid Com-
mun. Mass Spectrom. 2010, 24, 2849–2858.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: +49-871-44497. Fax: +49-871-44497. E-mail: hlschmidt@
web.de.
Funding Sources
This work was in part funded by the European Commission,
under the FP6 Food Quality and Safety Priority, within the
framework of the Integrated Project TRACE ꢀ 006942 “Tracing
Food Commodities in Europe”.
(12) Greenstein, J. P.; Winitz, M. Chemistry of Amino Acids; Wiley:
New York, 1961; p 2351.
(13) Kumagai, H.; Yamada, H.; Matsui, H.; Ohkishi, H.; Ogata, K.
Tyrosine phenol lyase. I. Purification, crystallisation, and properties. J.
Biol. Chem. 1970, 245, 1767–1772.
(14) Palcic, M. M.; Shen, S.-J.; Schleicher, B.; Kumagai, H.; Sawada,
S.; Yamada, H.; Floss, H. G. Stereochemistry and mechanism of
reactions catalysed by tyrosine phenol lyase from Escherichia intermedia.
Z. Naturforsch. C 1987, 42, 307–318.
(15) Countryman, S.; Shock, D.; Dixon, A.; Tauscher, J. High
recoveries of phenols from water with the polymeric sorbent Strata-
TMX; Application Sheet; Phenomenex: Torrance, CA, 2003.
(16) Ziadeh, B. I.; Michaud, A. L.; Saad, N. M. R.; Lewis, B. A.; Rafi,
M.; Pencharz, P. B.; Brenna, J. T. Enzymatic decarboxylation of tyrosine
and phenylalanine to enhance volatility for high-precision isotopic
analysis. Anal. Chem. 2002, 74, 479–483.
’ ACKNOWLEDGMENT
We thank R. S. Phillips, Department of Chemistry, University
of Georgia, Athens, GA, for the gift of the tyrosine phenol lyase,
and G. Fronza, Politecnico di Milano, Milan, Italy, for a gift of
L-tyrosine from chicken feathers and human hair. We are grateful
to W. A. Brand, Max-Planck-Institute for Biogeochemistry, Jena,
Germany, for the oxygen isotope ratio determinations of (2,4,6)-
tribromophenol. We thank P. Schieberle, Lehrstuhl f€ur Lebens-
mittelchemie, Technische Universit€at M€unchen, Germany, and
A. Rossmann from our laboratory for fruitful discussions,
A. Rossmann for help with the multielement isotope ratio determi-
nations, S. Hutter and C. Schwarz for experimental support, and
S. Harrison, School of Agriculture, Food Science and Veterinary
Medicine, Dublin, Ireland, for the revision of the manuscript.
(17) Sieper, H.-P.; Kupka, H.-J.; Williams, T.; Rossmann, A.; Rummel,
S.; Tanz, N.; Schmidt, H.-L. A measuring system for the fast simultaneous
isotope ratio and elemental analysis of carbon, hydrogen, nitrogen and sulfur
in food commodities and other biological material. Rapid Commun. Mass
Spectrom. 2006, 20, 2521–2527.
(18) Sternberg, L.; Pinzon, M. C.; Anderson, W. T.; Jahren, A. H.
Variation in oxygen isotope fractionation during cellulose synthesis: intramo-
lecular and biosynthetic effects. Plant, Cell Environ. 2006, 29, 1881–1889.
(19) Barbour, M. M. Stable oxygen isotope composition of plant
tissue: a review. Funct. Plant Biol. 2007, 34, 83–94.
’ ABBREVIATIONS USED
AIR, air (nitrogen); EC, Enzyme Commission (number); NCTC,
National Collection of Type Culture; VPDB, Vienna PeeDee
Belemnite; VSMOW, Vienna Standard Mean Ocean Water.
(20) Werner, R. A.; Rossmann, A.; Schwarz, C.; Bacher, A.; Schmidt,
H.-L.; Eisenreich, W. Biosynthesis of gallic acid in Rhus typhina:
discrimination between alternative pathways from natural oxygen abun-
dances. Phytochemistry 2004, 65, 2809–2813.
’ REFERENCES
(1) EG 2000/766. Entscheidung des Rates vom 04. Dezember 2000€uber
Schutzmassnahmen in Bezug auf die transmissiblen spongiformen Enzepha-
lopathien und die Verf€utterung von tierischem Protein. 2000/766/EG.
(2) European Union. Commision Regulation (EC) No. 1139/2003.
Off. J. Eur. Union 2003, L160, 22ꢀ32.
(21) Kohn, M. J. Predicting animal δ18O: accounting for diet and
physiological adaptation. Geochim. Cosmochim. Acta 1996, 60, 4811–4829.
(22) Thiem, I.; L€upke, M.; Seifert, H. Factors influencing the 18O/16O-
ratio in meat juices. Isot. Environ. Health Stud. 2004, 40, 1919–197.
(23) Schnyder, H.; Auerswald, K. Isotopes as natural recorders of
grassland ecosystem functioning and change. In Multifunctional Grasslands
(3) Schmidt, H.-L.; Rossmann, A.; Rummel, S.; Tanz, N. Stable
isotope analysis for meat authenticity and origin check. In Handbook of
9482
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483

Journal of Agricultural and Food Chemistry
ARTICLE
in a Changing World; Organization Committee of 2008 IGC/IRS Con-
ference, Eds.; Guangdong People’s Publishing House: Guangzhou, China,
2008; Vol. I, pp 46ꢀ51.
(24) Kaiser, A. S. Dissolved oxygen dynamics in the Dunnville Marsh
of the Grand River, Ontario, Canada. Thesis, University of Waterloo,
2009.
(25) Hegerding, L.; Seidler, D.; Danneel, H.-J.; Gessler, A.; Nowak,
B. Sauerstoffisotopen-Verh€altnis-Analyse zur Herkunftsbestimmung
von Rindfleisch. Fleischwirtschaft 2002, 82, 95–100.
(26) Augustyniak, W.; Kanski, R.; Kanska, M. Synthesis of tritium
labelled [20,6]-L-tyrosine. J. Labelled Compd. Radiopharm. 2004, 47,
977–981.
(27) Samuel, D.; Silver, B. L. Oxygen isotope exchange reactions of
organic compounds. Adv. Phys. Org. Chem. 1965, 3, 123–186.
(28) Byrn, M.; Calvin, M. Oxygen-18 exchange reactions of alde-
hydes and ketones. J. Am. Chem. Soc. 1966, 88, 1916–1922.
(29) Schmidt, H.-L. Fundamentals and systematics of the non-
statistical distributions of isotopes in natural compounds. Naturwis-
senschaften 2003, 90, 537ꢀ552; 2004, 91, 148.
(30) Facchini, P. J.; Huber-Allanach, K. L.; Tari, L. W. Plant aromatic
L-amino acid decarboxylases: evolution, biochemistry, regulation, and
metabolic engineering applications. Phytochemistry 2000, 54, 121–138.
9483
dx.doi.org/10.1021/jf201217r |J. Agric. Food Chem. 2011, 59, 9475–9483