Determination of 2,3-dihydroxypropionamide, an oxidative metabolite of acrylamide, in human urine by gas chromatography coupled with mass spectrometry.

Abstract

The general population is exposed to acrylamide (AA) mainly through food and tobacco smoke. AA is classified as probably carcinogenic to humans. Glycidamide (GA), as the primary oxidative metabolite, was identified to be the ultimate genotoxic agent. This warrants full investigation of the oxidative pathway in AA metabolism and the share of the oxidative compared to the reductive pathway. 2,3-Dihydroxy-propionamide (OH-PA) as the direct hydrolysis product of GA has been shown to be a major urinary oxidative metabolite in human AA metabolism. We developed an analytical method to reliably quantify OH-PA in urine by GC-MS after a multistep procedure including "stripping" on a solid phase material, lyophilization, silylation and re-extraction. With a detection limit of 1 μg/L, our method is sensitive enough to quantify OH-PA in all urine samples of the general population. Within and between series precisions were between 1.9% and 8.2% and mean recoveries between 97% and 101%. We applied this method to 30 urine samples from the general population. In all the samples, OH-PA was present in concentrations between 6.8 and 109.4 μg/L (median, 49.7 μg/L) with no difference between smokers and non-smokers. OH-PA concentrations were approximately ten times higher than expected from the metabolism of AA via GA. Currently, we cannot confirm OH-PA to be a specific biomarker of the oxidative pathway of AA metabolism. Other sources than AA respectively GA might need to be considered for the formation of OH-PA.

Full text

Page 1

ORIGINAL PAPER
Determination of 2,3-dihydroxypropionamide, an oxidative
metabolite of acrylamide, in human urine by gas
chromatography coupled with mass spectrometry
Julia M. Latzin & Birgit K. Schindler & Tobias Weiss &
Jürgen Angerer & Holger M. Koch
Received: 17 November 2011 /Revised: 20 December 2011 /Accepted: 21 December 2011 /Published online: 26 January 2012
#Springer-Verlag 2012
Abstract The general population is exposed to acrylamide
(AA) mainly through food and tobacco smoke. AA is clas-
sified as probably carcinogenic to humans. Glycidamide
(GA), as the primary oxidative metabolite, was identified
to be the ultimate genotoxic agent. This warrants full inves-
tigation of the oxidative pathway in AA metabolism and the
share of the oxidative compared to the reductive pathway.
2,3-Dihydroxy-propionamide (OH-PA) as the direct hydro-
lysis product of GA has been shown to be a major urinary
oxidative metabolite in human AA metabolism. We devel-
oped an analytical method to reliably quantify OH-PA in
urine by GC-MS after a multistep procedure including
“stripping” on a solid phase material, lyophilization, silyla-
tion and re-extraction. With a detection limit of 1 μg/L, our
method is sensitive enough to quantify OH-PA in all urine
samples of the general population. Within and between
series precisions were between 1.9% and 8.2% and mean
recoveries between 97% and 101%. We applied this method
to 30 urine samples from the general population. In all the
samples, OH-PA was present in concentrations between 6.8
and 109.4 μg/L (median, 49.7 μg/L) with no difference
between smokers and non-smokers. OH-PA concentrations
were approximately ten times higher than expected from the
metabolism of AA via GA. Currently, we cannot confirm
OH-PA to be a specific biomarker of the oxidative pathway
of AA metabolism. Other sources than AA respectively GA
might need to be considered for the formation of OH-PA.
Keywords Acrylamide(AA).Glycidamide(GA).
Biologicalmonitoring.2,3-Dihydroxy-propionamide
(OH-PA).GC/MS.Urine
Introduction
Acrylamide (AA) is classified as probably carcinogenic to
humans by the International Agency for Research on Cancer
[1], and similar classifications have been adopted by the
German Research Foundation and other national and inter-
national bodies [1–5].
The general population is ubiquitously exposed to AAvia
the diet [6–9]. AA is formed in carbohydrate-rich food like
French fries, potato chips and pastry by the Maillard reac-
tion at temperatures above 120°C [7, 9]. The Joint FAO/
WHO Expert Committee on Food Additives estimated food
related AA daily intakes between 1 μg/kg body weight (bw)
respectively 4 μg/kg bw for high consumers [10]. Another
major source of AA exposure of the general population is
first- and second-hand tobacco smoke [11, 12].
AA is rapidly distributed throughout the body after expo-
sure [13, 14]. AA is partly oxidized to glycidamide (GA) by
CYP 2E1 (Fig. 1) [15, 16]. GA, a highly reactive epoxide, is
regarded to be the ultimate genotoxic agent in the metabolism
of AA. In rodents and humans as well, AA- and GA-derived
mercapturic acids and hemoglobin adducts have already been
determinedinbloodandurine[17–19].TheformationofDNA
adductshasbeendescribedinrodents[20,21].2,3-Dihydroxy-
propionamide (OH-PA) is the direct hydrolysis product of GA
and is excreted as such together with the mercapturic acids in
urine [22, 23]. OH-PA therefore has to be regarded as an
important metabolite in the oxidative pathway of AA metabo-
lism that is closer to the carcinogenic potential (via GA) than
AA metabolites of the reductive pathway.
J. M. Latzin:B. K. Schindler:T. Weiss:J. Angerer:
H. M. Koch (*)
Institute for Prevention and Occupational Medicine
of the German Social Accident Insurance,
Institute of the Ruhr Universität Bochum (IPA),
Bürkle-de-la-Camp-Platz 1,
44789 Bochum, Germany
e-mail: koch@ipa-dguv.de
Anal Bioanal Chem (2012) 402:2431–2438
DOI 10.1007/s00216-011-5692-x

Page 2

The ratio between the oxidative and reductive pathway of
AA metabolism deserves special attention with respect to
the risk assessment of AA. Rats and mice show different
ratios of oxidative versus reductive metabolites (GA vs.
AA). GA versus AA is 0.39 in rats, 1.35 in mice. For
humans a ratio of 0.21 has been determined [13, 24, 25].
In the case of mice and rats, these ratios clearly reflect the
interspecies differences in the potency of AA to induce
cancer. The similarity of GA/AA ratios in humans and rats
may indicate similar cancer risks which should be lower
than in mice [26].
Therefore, all metabolites indicative of the oxidative
metabolism pathway are of special interest for human bio-
monitoring studies. In a human metabolism study, OH-PA in
urine represented a share of 5.4% (on a molar basis) of the
oral AA dose [27]. Another study found a comparable share
of 3.3% after oral application, albeit at a 200-fold higher AA
dose [28]. The mercapturic acids GAMA and iso-GAMA,
other known urinary metabolites of the oxidative pathway,
represented 4.6% and 0.8% of the dose [27]. Hence, OH-PA
seems to be one of or the main urinary oxidative metabolite
in the AA metabolism, and it is essential to include this
oxidative metabolite when focusing on the share of the
oxidative pathway.
Up to now, no internal exposure data are available for this
important oxidative metabolite for the general population.
We therefore modified our initial method developed for the
investigation of the human AA metabolism with isotope-
labeled material to suit the needs for human biomonitoring
studies, resulting in a highly sensitive and selective method
for the determination of OH-PA in urine of the general
population and occupational AA exposed subjects.
Experimental
Materials
OH-PA and the labeled 2,3-dihydroxy-d3-propionamide
(d3-OH-PA) were custom-synthesized by Dr. Belov (Max
Planck Institute for Biophysical Chemistry, Göttingen) with a
purity of >95% (>98% isotope purity). Acetonitrile Secco-
Solv® and SupraSolv®, hexane SupraSolv®, toluene Supra-
Solv® and methanol p.a. were purchased from VWR
International GmbH (Darmstadt, Germany). N-(tert-butyldi-
methylsilyl)-N-methyl-trifluoroacetamide (MTBSTFA) and
anhydrous pyridine (99.8%) were purchased from Sigma-
Aldrich (Steinheim, Germany). The water was deionised (e.g.
Millipore technique or equivalent purity). Isolute ENV+
(100 mg/3 mL) SPE columns were purchased from Separtis
(Grenzach-Wyhlen, Germany). A capillary column (DB,
35 ms, 60 m in length, 0.25 mm inner diameter, 0.25 μm in
film thickness) was purchased from joint analytical systems
GmbH (Moers, Germany).
Preparation of stock solutions
A stock solution of OH-PA was prepared by dissolving
10 mg of OH-PA in 10 mL water (1 g/L). The stock solution
was diluted with water to a concentration of 8 mg/L (work-
ing solution A) and 0.8 mg/L (working solution B). Both
working solutions were stored at –20°C for the preparation
of urinary standards.
The stock solution of the internal standard was prepared
by dissolving 14 mg of d3-OH-PA in 10 mL of acetonitrile
(1.4 g/L). This solution was diluted with acetonitrile to a
Oxidation(CYP 2E1)
acrylamide
(AA)
glycidamide
(GA)
mercapturic
acids
2,3-dihydroxy-
propionamide
OH-PA
DNA-adducts
(proven in rodents)
reductive pathway
glutathione
oxidative pathway
hemoglobinadducts
hemoglobin
hemoglobinadducts
mercapturic
acids
hemoglobinglutathione
Fig. 1 Metabolism of
acrylamide
2432J.M. Latzin et al.

Page 3

concentration of 3.5 mg/L (working solution internal
standard).
Sample preparation
The urine samples were stored at –20°C before analysis. They
were allowed to equilibrate to room temperature and were
shaken thoroughly for homogenization. One milliliter of urine
was spiked with 30 μL of the internal standard working
solution. Two millilitres of deionised water were added. After
homogenization, this solution was transferred onto an ENV+
cartridge that was previously washed and conditioned with 2×
2 mL methanol and 2×2 mL water. The sample was loaded on
the cartridge with a flow of ca. 1 mL/min which then was
suckeddrywiththeapplicationofavacuum.Inthis“stripping”
procedure,interferingmatrixisretainedontheENV+material
while the target analyte elutes without retention. The eluate
(approximately 3 ml) was frozen for at least 3 h at –20°C and
was then subjected to a lyophilization process overnight
at–46°Capplyingavacuumof0.01mbar(Zirbus,Laboratory
freeze dryer VaCo 2).
The lyophilizate was suspended in 1.5 mL acetonitrile by
vortex mixing (60 s) and ultrasonification (3 min). After
centrifugation (800×g, 5 min), 1,400 μL of the supernatant
was transferred carefully into a 1.8-mL glass screw-cap vial
without whirling up the sediment. The acetonitrile was
evaporated to dryness under a gentle stream of nitrogen.
After adding 50 μL of pyridine and 50 μL of MTBSTFA,
the vials were closed with the screw cap and the samples
were left at 65°C for 2 h for derivatization in an oven.
Then, 500 μL of deionised water was added to hydrolyze
the excess MTBSTFA. Five hundred microliters of hexane
was added to extract the derivatized OH-PA. The samples
were shaken for 10 min and were centrifuged (2,000×g,
10 min). Four hundred microlitres of the hexane layer was
transferred carefully into another 1.8-mL glass screw-cap
vial. Five hundred microlitres of hexane was added, and
extraction was performed as before (shaking 10 min; cen-
trifugation, 2,000×g; 10 min). Both extraction solutions
were combined and hexane was evaporated completely
under a gentle stream of nitrogen. The dry residue was re-
dissolved in 250 μL toluene. In some cases, sediments can
occur which can be separated by mild centrifugation. One
hundred microliters of the supernatant was transferred into a
microinsert, and 0.5 μL of this solution was injected into the
GC-MS. The creatinine content of each sample was deter-
mined according to Jaffe [29].
Calibration procedure and quality control
Pooled urine obtained from five non-smokers from the
general population was spiked with standard solutions
resulting in six different calibration standards in a
concentration range between 17.2 and 300 μg/L. The pooled
urine with a native, non-avoidable background level of OH-
PA was used as an unspiked blank. Calibration standards in
water were prepared the same way. These calibration stand-
ards were divided into 1-mL aliquots and frozen at –20°C
until use. The calibration standards were treated like the
samples during sample preparation. The quotient of the peak
area of OH-PA and the peak area of the labeled d3-OH-PA as
a function of the spiked concentration was plotted for quan-
tification. As quality control material was not commercially
available, two different pooled urine samples were used.
Both of these samples contained native OH-PA. One sample
was not spiked (Qlow), having a native OH-PA concentration
of 35.5 μg/L, the other sample was spiked with 75.25 μg/L
(Qhigh) resulting in a final OH-PA concentration of
176.8 mg/L. One-millilitre aliquots of the quality control
samples were frozen at –20°C until use and were analysed in
each series. Both quality control samples were analysed six
times in a row to determine within-day repeatability and on
six different days in each series to determine between-day
repeatability.
Eight individual urine samples with a creatinine content
between 0.4 and 2.5 g/L were used to determine the accu-
racy and imprecision of the analytical method and possible
matrix effects. Each urine sample was analysed as described
above with and without spiking concentrations of 17 and
56 μg/L. A calibration in water was performed in parallel to
a calibration in urine in order to evaluate the influence of
urinary matrix.
GC-MS method
Gas chromatographic analysis was carried out using an
Agilent 7890 A GC and a 5975 C EI/CI mass-selective
detector with a COMBI PAL system autosampler (CTC
Analytics AG, Switzerland). Separation was performed on
a DB 35 capillary column at a constant helium flow of
1.2 mL/min. The chromatographic conditions were as fol-
lows. The initial oven temperature was 100°C for 1 min and
was increased with 20°C/min to 130°C (hold, 1 min). The
temperature was increased with 12.5°C/min to 220°C (hold,
10 min) and finally up to 280°C for bake out (hold, 10 min).
Total run-time was 33.1 min. The transfer line temperature
was 280°C and the injector temperature was 260°C. Injec-
tion was performed in splitless mode (with a purge flow of
50 mL/min after 2 min).
The tert-butyldimethylsilyl derivatives were analysed in
the single ion monitoring mode with electron impact ioni-
zation (EI). EI mass spectra were obtained at an ionisation
energy level of 70 eV. The source temperature was 230°C,
and the quadrupole temperature was 150°C. The dwell time
for each fragment was 100 ms. We used m/z 390 [M-tBu]−
Determination of 2,3-dihydroxypropionamide by GC-MS2433

Page 4

as the quantifier trace and m/z 432 [M-Me]−and m/z 391 as
qualifier traces. The labeled internal standard showed the
same fragmentation pattern, thus m/z 393 was used as
quantifier and m/z 435 as qualifier (see Fig. 2). The deriv-
ative with three tert-butyldimethylsilyl groups (TBDMS)
itself (m/z 447 respectively m/z 450) could not be observed.
For peak verification, we calculated the ratio between quan-
tifier and qualifier ion. Values within the threefold standard
deviation of the mass ratio calculated from the processed
aqueous calibration curve were regarded as acceptable.
Study subjects
In a pilot study, we investigated spot urine samples of 30
persons from the general German population (14 smokers
and 16 non-smokers; smoking status was confirmed by
analysing nicotine, cotinine and 3-hydroxycotinine with an
online LC-LC/MS-MS method after enzymatic hydrolysis
in a slightly modified method according to Tuomi et al.
[30]). These persons had no known occupational exposure
to acrylamide or polyacrylamide. The smokers consumed
between 5 and 20 cigarettes per day. All samples were
stored at –20°C until analysis.
Results
Analytical method
We successfully adapted our previous method developed for
a human metabolism study with labeled material to now be
able to detect and quantify urinary OH-PA levels in urine
samples of the general population [25]. The method to
extract and detect the low molecular weight and rather polar
OH-PA in urine by GC-MS requires a multistep procedure
including “stripping” on a solid phase material, lyophiliza-
tion, derivatization and re-extraction; however, we chose
this approach due to the advantages of the GC-MS in chro-
matographic separation and sensitivity in contrast to possible
HPLC-MS approaches.
Clean-up, derivatization and GC-MS analyses
The clean-up procedure proved effective to extract OH-PA
from the urinary matrix. Given the absolute dryness of the
lyophilized samples, the derivatization procedure was robust
and reliable and relatively easy to perform. The washing and
re-extraction of the tert-butyldimethylsilyl derivatives pro-
duced sufficient clean extracts for GC-MS analyses. The
tert-butyldimethylsilyl derivatives of OH-PA proved to be
ideal for MS in EI-mode with a sufficiently high molecular
weight and a favorable fragmentation (loss of a tert-butyl
moiety) that produced a stable and intensive ion with a
rather high m/z of 390 (m/z 393 for d3-OH-PA). Within
the chromatographic runtime of 33 min, the tert-
butyldimethylsilyl derivates of OH-PA and d3-OH-PA are
well separated from the urinary matrix at all mass traces (see
Fig. 3).
Reliability of the method
Calibration graph The calibration graphs (both in water and
urine) were linear for the range investigated (17.2–300 μg/
L) with correlation coefficients higher than 0.99. The slopes
for the calibration in water and urine were the same
(m00.011). In the blank water samples, no OH-PA was
40 6080100120140160180200220240260280300320340360380400420440
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000
220000
240000
m/z-->
Abundance
91
149
390
73
432
m/z 432
NH
OO
Si
Si
O
Si
m/z 390
MW 447 (not detectable)
460
Fig. 2 Mass spectrum of a derivatized OH-PA standard solution (100 μg/L), measured in single-ion mode (SIM). The loss of the methyl
respectively tert-butyl group could take place at each of the three TBDMS (t-butyl-dimethylsilyl) groups
2434 J.M. Latzin et al.

Page 5

detectable. The pooled non-smoker urine used for prepara-
tion of the calibration standards in urine contained a back-
ground level of 23 μg OH-PA/L urine. The detection limit,
defined as signal-to-noise ratio of 3 for the quantifier ion m/
z 390 was 1 μg/L, the limit of quantification, defined as
signal-to-noise ratio of 6 is 3 μg/L.
Precision and accuracy The relative standard deviations for
within-day imprecision were 2.7% at a concentration level
of 35.5 μg/L in Qlowand 3.6% at a concentration level of
176.8 μg/L in Qhigh. The relative standard deviations for
within-day imprecision were 8.2% for Qlowand 1.9% for
Qhigh(see Table 1).
The imprecision was also calculated based on eight indi-
vidual urine samples reflecting a broad range of creatinine
content (creatinine, 0.5–2.5 g/L) and having native OH-PA
concentrations ranging from 18.2 to 109.4 μg/L. Precision
and accuracy data were calculated for the spiked concen-
trations of 17 μg/L (spike 1) and 56 μg/L (spike 2) by
subtracting the native OH-PA concentration from the total
OH-PA concentration of the spiked sample. The relative
standard deviations for this imprecision (reflecting the in-
fluence of different matrix loads) were 2.7% (spike 1) and
3.6% (spike 2).The accuracy of the method calculated from
the mean relative recovery was 97% (83–113%) for spike 1
(17 μg/L) and 101% (90–112%) for spike 2 (56 μg/L) (see
Table 2). The absolute recovery was not determined because
no MTBSTFA derivative of OH-PA was available as a
characterized standard substance.
Results of biological monitoring
We applied this method in a pilot study to 30 urine samples
of the general population. Because smoking is a known
source to AA and therefore a potential confounder of OH-
PA, we selected 14 smokers to be part of the study popula-
tion. We detected and quantified OH-PA in all of the urine
samples. The OH-PA concentrations ranged from 6.8 to
109.4 μg/L (mean, 50.4 μg/L; median, 49.7 μg/L). In all
the samples, the identity of the peak was confirmed by the
qualifier ion. We did not observe any significant differences
17.6017.70 17.8017.90 18.0018.1018.2018.30 18.40 18.50 18.6018.70 18.80 18.90
Time-->
Relative Abundance
m/z 393
m/z 390
m/z 435
m/z 391m/z 432
Fig. 3 Chromatogram of a urine sample (creatinine, 0.63 g/L) spiked with d3-OH-PA (m/z 393 and 435) containing 18.2 μg OH-PA/L (m/z 390,
391 and 432)
Table 1 Precision data
Each, n06Qlow
Qhigh
Mean [μg/L]
Range [μg/L]
Relative standard
deviation [%]
Mean [μg/L]
Range [μg/L]
Relative standard
deviation [%]
35.5
34.5–36.9
2.7
176.8
171.2–189.2
3.6
Within-day
repeatability
39.3
35.5–43.5
8.2
174.6
169.9–178.8
1.9
Between-day
repeatability
Table 2 Accuracy data for OH-PA
n08
(Different individual urines)
Spike I:
16.7 μg/L
Spike II:
55.8 μg/L)
Native OH-PA concentration
[μg/L]
Creatinine range [g/L]
Mean recovery [μg/L]
Recovery (range) [μg/L]
Relative recovery (mean) [%]
Relative recovery (range) [%]
18.2–109.4Recovery
0.54–2.54
16.3
13.9–18.9
97.4
83–113
56.7
50.2–62.5
100.7
90–112
Determination of 2,3-dihydroxypropionamide by GC-MS 2435

Page 6

between smokers (n014) and non-smokers (n016) (see
Table 3). Smokers had a median OH-PA concentration of
49.1 μg/L and non-smokers of 52.5 μg/L. Only after cor-
rection regarding the creatinine content of the urine samples,
smokers had slightly (but not significantly) higher median
concentrations (44.6 μg/g creatinine) compared to non-
smokers (40.9 μg/g creatinine).
Discussion
Analytical method
The adaption of the method by Hartmann et al. [25] to
determine environmental OH-PA exposure levels proved to
be robust and reliable. MTBSTFA with pyridine was much
more efficient in derivatization yield than MTBSTFA with
1% TBMCS (tert-butyl-dimethylchlorsilane) as used by
Hartmann et al. [25]. Ten microlitres of pyridine proved
sufficient as a catalyst in this reaction. However, we added
50 μL of pyridine because pyridine also facilitated the
resolvation of the residue. Instead of derivatization over
night as in Hartmann et al. [25], 2 h was sufficient with
our procedure, and no loss of derivates was observed. Com-
plete dryness of the residue/extract before derivatization
with MTBSTFA was essential, which was accomplished
by resuspension of the lyophilizate with acetonitrile and
evaporation to dryness before derivatization. In general, it
is very important not to transfer any precipitate in any step
of the extraction process. Any carry-over of precipitate will
show up in the final toluene solution.
The use of the labeled OH-PA for internal standardization
and quantification adjusted possible analyte losses during
the clean-up procedure and different derivatization yields.
Without using the internal standard, an influence of the
matrix could be observed. In a way, the matrix seemed to
have a stabilizing influence on extraction, lyophilization
and/or derivatization. Peak intensities for spiked standards
and the labeled internal standard were, in general, higher by
a factor 3 to 5 in urine compared to water. Also, peak
intensities for the labeled internal standard varied much
stronger in water than in urine (data not shown).
Biomonitoring
With the method developed above, we are, for the first time,
able to measure the urinary excretion of OH-PA–the major
metabolite of the oxidative pathway of AA metabolism–in
the general population. We detected and quantified OH-PA
in all of the 30 urine samples investigated. Contrary to our
expectations, smokers and non-smokers had similar OH-PA
levels (49.1 μg/L vs. 52.5 μg/L), and the extent of OH-PA in
the urine samples (mean, 50.4 μg/L; range, 6.8 to
109.4 μg/L) was higher than expected. Next to AA contain-
ing foodstuff, tobacco smoke is a major source of AA
exposure. We have previously shown that smokers had 4
respectively 2.4 times higher hemoglobin adduct rates for
the reductive respectively oxidative hemoglobin adducts
than non-smokers [18]. Also, smokers had approximately
four times higher mercapturic acid levels in urine than non-
smokers, both for the reductive AAMA and the oxidative
GAMA [31]. Regarding these parameters, other studies
have similar results [32–37] and show 2.5-fold higher
(AAMA), 1.7-fold higher (GAMA) and a threefold higher
(AA hemoglobin adduct) levels in smokers than in non-
smokers [37] or even tenfold higher (AAMA) and 6.6-fold
higher (GAMA) levels [36].
We did not find such an association for OH-PA between
the smokers and non smokers in our study population. We
can only speculate about the reasons for this, but one reason
may lie in the extent of OH-PA levels we determined. Based
on our knowledge of the AA metabolism and the unambig-
uous identification of OH-PA as an AA metabolite, we
would have expected to find OH-PA in the same concentra-
tion range as the GA mercapturic acid (median non-
smokers, 5 μg/L; median smokers, 19 μg/L according to
Boettcher et al. or medians of 8.7 and 15.0 μg/L according
to Urban et al. or medians of 3 and 20 μg/L according to
Kellert et al.) [19, 25, 27, 36, 37]. The OH-PA levels we
found in the general population were therefore approximately
six to ten times higher than expected from the metabolism of
AAvia GA. As a possible explanation, other sources than AA
respectively. GA might need to be considered for the
formation of OH-PA. Currently, we also cannot rule out
an endogenous formation of OH-PA.
Amoredetailed studymeasuringallavailable biomarkersin
a large population group with smokers and non-smokers could
give a deeper insight in possible OH-PA sources. In such a
study,italsowouldbeinterestingtominimizetheconsumption
of AA containing foodstuff and of course cigarettes.
Summary and conclusions
We developed a reliable and accurate method for the deter-
mination of OH-PA in human urine. The detection limit of
Table 3 Results of the human biomonitoring
S (n014) NS (n016) All (n030)
Median [μg/L]
Range [μg/L]
Median [μg/g creatinine]
Range [μg/g creatinine]
49.1
9.6–109.4
44.6
19.5–95.0
52.5
6.8–79.4
40.9
21.9–180.4
49.7
6.8–109.4
44.4
19.5–180.4
S smokers, NS non-smokers
2436J.M. Latzin et al.

Page 7

1 μg/L was sufficiently low enough to detect OH-PA in
urine of the general population and workers occupationally
exposed to AA/GA. In all 14 smoker and 16 non-smoker
urines, we measured considerably higher amounts of OH-
PA than anticipated from calculations based on quantitative
metabolism data. In contrast to other AA/GA metabolites,
there was no difference in the OH-PA level between smok-
ers and non-smokers. Albeit we unambiguously identified
OH-PA as a major metabolite of the oxidative pathway of
AA metabolism in a previous study, other sources than AA/
GA exposure must significantly contribute to the OH-PA
excretion. Currently, we are not aware of any other possible
source leading to OH-PA excretion, but both other environ-
mental chemicals and an endogenous generation might need
to be considered. Therefore, at least for the background
exposure of the general population we cannot regard OH-
PA as a specific biomarker indicating an exposure to AA/
GA. For an appraisement of this marker and a further risk
assessment of AA/GA, more human data have to be
collected.
Acknowledgement
for their financial support of the project (AN 107/21-1).
We thank the DFG (German Research Foundation)
References
1. IARC (1994) Summary of data reported and evaluation: acrylam-
ide. http://monographs.iarc.fr/ENG/Monographs/vol60/volume60.
pdf. Accessed 19 Dec 2011
2. EPA US (1994) Chemical summary for acrylamide. Office of
Toxic Substances; U.S. Environmental Protection Agency, Wash-
ington DC. http://www.epa.gov/chemfact/s_acryla.txt. Accessed
19 Dec 2011
3. EU (2001) Opinion on the results of the risk assessment of: acrylam-
ide. Scientific Committee on Food, Brussel. http://ec.europa.eu/
health/scientific_committees/environmental_risks/opinions/sctee/
sct_out88_en.htm. Accessed 19 Dec 2011
4. DFG (2002) Maximale Arbeitsplatzkonzentrationen und biologische
Arbeitsstoff Toleranzwerte 2002. Mitteilung 38 der Senatskommission
zur Prüfung gesundheitsschädlicher Arbeitsstoffe. Wiley-VCH,
Weinheim
5. WHO (2002) Health implications of acrylamide in food-report of a
joint FAO/WHO consultation. Geneva, Switzerland. http://www.
who.int/entity/foodsafety/publications/chem/en/acrylamide_full.
pdf. Accessed 19 Dec 2011
6. Stadler RH, Blank I, Varga N, Robert F, Hau J, Guy PA, Robert
MC, Riediker S (2002) Acrylamide from Maillard reaction products.
Nature 419:449–450
7. Mottram DS, Wedzicha BL, Dodson AT (2002) Acrylamide is
formed in the Maillard reaction. Nature 419:448–449
8. LGL Bayern (2009) (Bavarian national office for health and food
security) Untersuchungsergebnisse: Acrylamid in Lebensmitteln.
http://www.lgl.bayern.de/lebensmittel/chemie/toxische_
reaktionsprodukte/acrylamid/ue_2009_acrylamid.htm. Accessed
19 Dec 2011
9. Tareke E, Rydberg P, Karlsson P, Eriksson S, Törnqvist M (2002)
Analysis of acrylamide, a carcinogen formed in heated foodstuffs.
J Agric Food Chem 50:4998–5006
10. JECFA (2005) Summary and conclusions of the sixty-fourth meet-
ing of the Joint FAO/WHO Expert Committee on Food Additives
(JECFA). ftp://ftp.fao.org/es/esn/jecfa/jecfa64_summary.pdf.
Accessed 19 Dec 2011
11. Smith CJ, Perfetti TA, Rumple MA, Rodgman A, Doolittle DJ
(2000) “IARC group 2A Carcinogens” reported in cigarette main-
stream smoke. Food Chem Toxicol 38:371–383
12. Schettgen T, Broding HC, Angerer J, Drexler H (2002) Hemoglobin
adducts of ethylene oxide, propylene oxide, acrylonitrile and
acrylamide-biomarkers in occupational and environmental medicine.
Toxicol Lett 134:65–70
13. Sumner SC, Williams CC, Snyder RW, Krol WL, Asgharian B,
Fennell TR (2003) Acrylamide: a comparison of metabolism and
hemoglobin adducts in rodents following dermal, intraperitoneal,
oral, or inhalation exposure. Toxicol Sci 75:260–270
14. Miller MJ, Carter DE, Sipes IG (1982) Pharmacokinetics of acryl-
amide in Fisher-344 rats. Toxicol Appl Pharmacol 63:36–44
15. Calleman CJ, Bergmark E, Costa LG (1990) Acrylamide is me-
tabolized to glycidamide in the rat: evidence from hemoglobin
adduct formation. Chem Res Toxicol 3:406–412
16. Sumner SC, Fennell TR, Moore TA, Chanas B, Gonzalez F,
Ghanayem BI (1999) Role of cytochrome P450 2E1 in the metab-
olism of acrylamide and acrylonitrile in mice. Chem Res Toxicol
12:1110–1116
17. Bergmark E, Calleman CJ, Costa LG (1991) Formation of hemoglo-
bin adducts of acrylamide and its epoxide metabolite glycidamide in
the rat. Toxicol Appl Pharmacol 111:352–363
18. Schettgen T, Rossbach B, Kutting B, Letzel S, Drexler H, Angerer
J (2004) Determination of haemoglobin adducts of acrylamide and
glycidamide in smoking and non-smoking persons of the general
population. Int J Hyg Environ Health 207:531–539
19. Boettcher MI, Schettgen T, Kutting B, Pischetsrieder M, Angerer J
(2005) Mercapturic acids of acrylamide and glycidamide as bio-
markers of the internal exposure to acrylamide in the general
population. Mutat Res 580:167–176
20. Segerbäck D, Calleman CJ, Schroeder JL, Costa LG, Faustman
EM (1995) Formation of N-7-(2-carbamoyl-2-hydroxyethyl)guanine
in DNA of the mouse and the rat following intraperitoneal adminis-
tration of [14C]acrylamide. Carcinogenesis 16:1161–1165
21. da Gamboa CG, Churchwell MI, Hamilton LP, Von Tungeln LS,
Beland FA, Marques MM, Doerge DR (2003) DNA adduct forma-
tion from acrylamide via conversion to glycidamide in adult and
neonatal mice. Chem Res Toxicol 16:1328–1337
22. Sumner SC, MacNeela JP, Fennell TR (1992) Characterization and
quantitation of urinary metabolites of [1,2,3-13C]acrylamide in rats
and mice using
Chem Res Toxicol 5:81–89
23. Fennell TR, Friedman MA (2005) Comparison of acrylamide
metabolism in humans and rodents. Adv Exp Med Biol 561:109–
116
24. Kopp EK, Dekant W (2009) Toxicokinetics of acrylamide in rats
and humans following single oral administration of low doses.
Toxicol Appl Pharmacol 235:135–142
25. Hartmann EC, Latzin JM, Schindler BK, Koch HM, Angerer J
(2011) Excretion of 2,3-dihydroxy-propionamide (OH-PA), the
hydrolysis product of glycidamide, in human urine after single oral
dose of deuterium-labeled acrylamide. Arch Toxicol 85:601–606
26. Sweeney LM, Kirman CR, Gargas ML, Carson ML, Tardiff RG
(2010) Development of a physiologically-based toxicokinetic
model of acrylamide and glycidamide in rats and humans. Food
Chem Toxicol 48:668–685
27. Hartmann EC, Boettcher MI, Bolt HM, Drexler H, Angerer J
(2009) N-Acetyl-S-(1-carbamoyl-2-hydroxy-ethyl)-L-cysteine
(iso-GAMA) a further product of human metabolism of acrylamide:
comparison with the simultaneously excreted other mercapturic
acids. Arch Toxicol 83:731–734
13C nuclear magnetic resonance spectroscopy.
Determination of 2,3-dihydroxypropionamide by GC-MS 2437

Page 8

28. Fennell TR, Sumner SC, Snyder RW, Burgess J, Spicer R, Bridson
WE, Friedman MA (2005) Metabolism and hemoglobin adduct
formation of acrylamide in humans. Toxicol Sci 85:447–459
29. Jaffe M (1886) Ueber den Niederschlag welchen Pikrinsaeure im
normalen Harn erzeugt und über eine neue Reaktion des Kreatinins.
Z Physiol Chem 10:391–400 [in German]
30. Tuomi T, Johnsson T, Reijula K (1999) Analysis of nicotine, 3-
hydroxycotinine, cotinine, and caffeine in urine of passive smokers
by HPLC-tandem mass spectrometry. Clin Chem 45:2164–2172
31. Boettcher MI, Angerer J (2005) Determination of the major mer-
capturic acids of acrylamide and glycidamide in human urine by
LC-ESI-MS/MS. J Chromatogr B Analyt Technol Biomed Life Sci
824:283–294
32. Fennell TR, Sumner SC, Snyder RW, Burgess J, Friedman MA
(2006) Kinetics of elimination of urinary metabolites of acrylamide
in humans. Toxicol Sci 93:256–267
33. Bjellaas T, Stolen LH, Haugen M, Paulsen JE, Alexander J, Lundanes
E, Becher G (2007) Urinary acrylamide metabolites as biomarkers for
short-term dietary exposure to acrylamide. Food Chem Toxicol
45:1020–1026
34. Scherer G, Engl J, Urban M, Gilch G, Janket D, Riedel K (2007)
Relationshipbetweenmachine-derivedsmokeyieldsandbiomarkersin
cigarette smokers in Germany. Regul Toxicol Pharmacol 47:171–183
35. Vesper HW, Bernert JT, Ospina M, Meyers T, Ingham L, Smith A,
Myers GL (2007) Assessment of the relation between biomarkers
for smoking and biomarkers for acrylamide exposure in humans.
Cancer Epidemiol Biomarkers Prev 16:2471–2478
36. Kellert M, Scholz K, Wagner S, Dekant W, Volkel W (2006)
Quantitation of mercapturic acids from acrylamide and glycida-
mide in human urine using a column switching tool with two trap
columns and electrospray tandem mass spectrometry. J Chromatogr
A 1131:58–66
37. Urban M, Kavvadias D, Riedel K, Scherer G, Tricker AR (2006)
Urinary mercapturic acids and a hemoglobin adduct for the dosim-
etry of acrylamide exposure in smokers and nonsmokers. Inhal
Toxicol 18:831–839
2438 J.M. Latzin et al.