Journal of Chromatography B, 998 (2015) 40–44

Contents lists available at ScienceDirect

Journal of Chromatography B
journal homepage: www.elsevier.com/locate/chromb

Validation of a multi-analyte HPLC-DAD method for determination of

uric acid, creatinine, homovanillic acid, niacinamide, hippuric acid,
indole-3-acetic acid and 2-methylhippuric acid in human urine
Daniela Remane a , Soeren Grunwald b,1 , Henrike Hoeke c , Andrea Mueller d ,
Stefan Roeder e , Martin von Bergen b,d,f , Dirk K. Wissenbach b,∗
a
Institute of Forensic Medicine, University Hospital Jena, Jena, Germany
b
Department of Metabolomics, Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany
c
Department of Pharmaceutical and Medicinal Chemistry, Institute of Pharmacy, University of Leipzig, Leipzig, Germany
d
Department of Proteomics, Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany
e
Department of Environmental Immunology, Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany
f
Department of Biotechnology, Chemistry and Environmental Engineering Aalborg University, Aalborg, Denmark

a r t i c l e i n f o a b s t r a c t

Article history: During the last decades exposure sciences and epidemiological studies attracts more attention to unravel
Received 4 September 2014 the mechanisms for the development of chronic diseases. According to this an existing HPLC-DAD method
Received in revised form 15 June 2015 for determination of creatinine in urine samples was expended for seven analytes and validated. Creati-
Accepted 17 June 2015
nine, uric acid, homovanillic acid, niacinamide, hippuric acid, indole-3-acetic acid, and 2-methylhippuric
Available online 24 June 2015
acid were separated by gradient elution (formate buffer/methanol) using an Eclipse Plus C18 Rapid
Resolution column (4.6 mm × 100 mm).
Keywords:
No interfering signals were detected in mobile phase. After injection of blank urine samples signals for
HPLC-DAD
Validation the endogenous compounds but no interferences were detected. All analytes were linear in the selected
Creatinine calibration range and a non weighted calibration model was chosen. Bias, intra-day and inter-day pre-
Uric acid cision for all analytes were below 20% for quality control (QC) low and below 10% for QC medium and
Hippuric acid high. The limits of quantification in mobile phase were in line with reported reference values but had to
Urine be adjusted in urine for homovanillic acid (45 mg/L), niacinamide 58.5(mg/L), and indole-3-acetic acid
(63 mg/L). Comparison of creatinine data obtained by the existing method with those of the developed
method showing differences from −120 mg/L to +110 mg/L with a mean of differences of 29.0 mg/L for 50
authentic urine samples. Analyzing 50 authentic urine samples, uric acid, creatinine, hippuric acid, and
2-methylhippuric acid were detected in (nearly) all samples. However, homovanillic acid was detected
in 40%, niacinamide in 4% and indole-3-acetic acid was never detected within the selected samples.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction
During the last decades exposure research attracts more atten-
tion as several studies found relationships between the “Exposome”
Abbreviations: CREA, creatinine; UA, uric acid; HVA, homovanillic acid; NA,
niacinamide; HA, hippuric acid; IAA, indole-3-acetic acid; 2-MHA, 2-methylhippuric
acid; QC, quality control.
∗ Corresponding author. Fax: +49 341 235 451211.
E-mail address: dirk.wissenbach@ufz.de (D.K. Wissenbach).
1
Current address: IDT Biologika GmbH, Dessau-Rosslau, Germany.
and the development of chronic diseases such as allergy [1,2]. The
Exposome is defined as the sum of all exposures that an individ-
ual is subjected to [3]. Its impact on human health is assessed in
epidemiological cohort studies.
According to this a prospective mother–child cohort study,
“Lifestyle and environmental factors and their influence on New-
born Allergy risk” (LiNA), was initiated 2005 by the Helmholtz
Center for Environmental Research (Leipzig, Germany) and by the
St. Georg Hospital in Leipzig, Germany [4,5]. In addition to, e.g.,
environmental monitoring, bio analysis of urine and blood is con-
ducted for every mother–child pair focusing on I. immunological
parameters; II. external exposures such as pollutants, endocrine

http://dx.doi.org/10.1016/j.jchromb.2015.06.021
1570-0232/© 2015 Elsevier B.V. All rights reserved.
 D. Remane et al. / J. Chromatogr. B 998 (2015) 40–44 41

Table 1
Covered analytes (sorted by retention time), calibration (Cal) level concentrations [mg/L], quality control sample (QC) concentrations [mg/L], observed bias [%], intra-day
precision (intra-day) and inter-day precision (inter-day), and limit of quantification (LOQ) [mg/L] in MP A and urine.

Analytes Calibration level [mg/L] QC low QC medium QC high LOQ

1 2 3 4 5 6 7 c [mg/L] Bias RSD c [mg/L] Bias RSD c [mg/L] Bias RSD MP A urine
[%] [%] [%] [mg/L] [mg/L]

Intra Inter Intra Inter Intra Inter

day [%] day [%] day [%] day [%] day [%] day [%]

UA 100 200 300 450 650 850 1000 150 −5.5 1.8 5.5 500 −2.5 0.3 10.7 900 −0.7 1.3 3.2 25 25
CREA 100 500 750 1000 1250 1500 2000 600 1.0 4.7 5.8 1100 −0.1 5.4 5.4 1700 −0.1 2.8 3.9 100 100
HVA 2.5 5.0 10 25 50 75 100 15 14.9 4.6 17.2 60 1.1 3.2 6.1 90 −2.0 2.4 7.0 2.5 45
NAA 5.0 10 15 25 65 80 100 20 −1.5 5.1 9.0 50 3.6 4.2 4.9 90 1.0 3.0 4.7 5.0 58.5
HA 50 75 100 250 500 750 1000 200 0.4 3.1 4.3 400 0.5 2.5 2.4 900 0.3 1.3 2.3 50 50
IAA 10 25 40 55 70 85 100 20 12.8 3.7 16.9 50 9.3 8.0 10.2 80 6.0 3.0 3.4 10 63
2-MHA 20 30 40 50 70 85 100 25 13.7 8.0 10.4 60 3.7 6.8 8.6 80 4.5 1.7 4.0 20 22.5

UA = uric acid; CREA = creatinine; HVA = homovanillic acid; NAA = niacinamide; HA = hippuric acid; IAA = indole-3-acetic acid; 2-MHA = 2-methylhippuric acid.

disruptors, and food ingredients III. the metabolome. Using these
approaches it was shown that the development of chronic diseases
are strongly associated with prenatal and early lifetime exposures
[4–7].
In this study an existing HPLC-DAD-method for the analysis of
urine creatinine (CREA) concentration should be replaced by a fully
validated new HPLC-DAD-method. In addition to CREA, the new
method should also cover uric acid (UA), homovanillic acid (HVA),
niacinamide (NA), hippuric acid (HA), indole-3-acetic acid (IAA) as
well as 2-methylhippuric acid (2-MHA). While urine CREA con-
centration is used for normalization of urine bio analysis results,
the other analytes showed associations either to immune response
(UA, NA), to external exposures such as toluene (HA) or xylene (2-
MHA), to the gut microbiome (IAA, HA), to nutrition (UA,NA,HA) or
the endogenous response to occurring mental stress (HVA) [8–16].
Even if not every single compound is necessarily associated to a sin-
gle exposure (e.g., HA as biomarker for toluene, but also for phenolic
food compounds) they all provide metabolomic marker for certain
lifestyles (including nutrition, mental status, gut microbiome, and
also exposures). So far only a few methods covering some of the
above mentioned analytes in urine were described. For example
Zuo et al. described a HPLC-UV method for CREA and UA, while
Shirao et al. analyzed CREA, HVA and UA using miscellar electroki-
netic chromatography [17,18]. Antunes et al. validated a HPLC-UV
method for CREA and UA in dried urine spots [11]. A liquid chro-
matography mass spectrometry method for CREA, UA, HA, and IAA
was developed by Gu et al. [19]. To reduce sample amount and
work load the new method was supposed to cover all the above
mentioned analytes by using a simple and cost effective sample
preparation and analysis technique.
2. Materials and methods
2.1. Chemicals, reagents, and standards
Formic acid, octanesulfonic acid, phosphoric acid, potassium
hydroxide, and IAA (each analytical grade), methanol and acetoni-
trile (LC–MS grade) were obtained by Merck (Darmstadt, Germany).
Ammonium formate, UA, CREA, HVA, NA, HA, and 2-MHA (each ana-
lytical grade) were obtained from Fluka (Taufkirchen, Germany).
Ultrapure water was produced by a MilliporeQ (MerckMillipore,
Darmstadt, Germany) water purification system. External quality
control (QC) samples for CREA and UA (KU 01/12 A and B) were
obtained from Referenzinstitut für Bioanalytik (Bonn, Germany).
CREA concentrations were 1020.0 and 1080.0 mg/L, UA concentra-
tions were 209.0 and 254.0 mg/L for KUA 1/12 A and B, respectively.
2.2. HPLC-DAD Analysis
A Prominence modular HPLC-system (Shimadzu, Jena,
Germany) consisting of two binary pumps, an autosampler
(tray temperature: 20 ◦ C), a column oven (set to 40 ◦ C), and a diode
array detector (detection wavelength from 200 to 400 nm) was
used in this study. Elution was performed on a ZORBAX Eclipse
Plus C18 Rapid Resolution 4.6 mm × 100 mm (Agilent, Waldbronn,
Germany) at a total flow rate of 1.0 mL/min. The gradient was as
follows 0.00–1.00 min 0% mobile phase (MP) B; 1.00–8.00 min–40%
MP B; 8.00–10.00 min–100% MP B; 10.00–15.00 min holding at
100% MP B; 15.00–16.50 min–0% MP B; 16.50–25.00 min equilibra-
tion at 0% MP B. MP A consisted of 50 mM ammonium formate and
5 mM octanesulfonic acid in 99% water and 1% methanol (adjusted
to pH 6.55 with formic acid). MP B consisted of 50 mM ammonium
formate and 5 mM octanesulfonic acid in 100% methanol (adjusted
to pH 6.55 with formic acid). The injection volume was 10 ␮L;
after each injection the needle was washed using 200 ␮L of an
acetonitrile/water (1:1; v:v) mixture. An inhouse reference library
was created by injecting each compound (c = 1000 mg/L) and stor-
ing the obtained spectra after background subtraction. Qualitative
data were obtained by reference library matching. All quantitative
analyses were achieved by determination of the area under the
curve and calibration models using extracted chromatograms
(233 nm wavelength).
2.3. Preparation of solutions, calibration and quality control
samples
With the exception of HA stock solutions of all analytes
(2000 mg/L for UA, CREA, 1000 mg/L for HVA, NA, IAA, 2-MHA) were
prepared in MP A, and stored at 8 ◦ C until use. Stock solution of
HA (1000 mg/L) was freshly prepared every day in 0.3 M potassium
hydroxide solution according to Kim et al. [20]. Working solutions
were prepared by dilution of the stock solution with the respec-
tive amount of solvent. Calibration samples were prepared at seven
different calibration levels (Cal. 1–7) (Table 1) by dilution of the
appropriate amount of working solution with MP A. Three internal
quality control (QC) samples (low, med, high, concentrations given
in Table 1) were prepared by dilution of the appropriate amount of
a separately prepared working solution with MP A.
2.4. Sample origin and sample preparation
Human blank urine samples (n = 6) were aliquoted and stored
at −20 ◦ C until usage. The calibration, internal QC samples (50 ␮L
solution), urine samples (50 ␮L urine), and external QC samples
42 D. Remane et al. / J. Chromatogr. B 998 (2015) 40–44

(50 ␮L urine) were worked up by dilution with 450 ␮L MP A and

directly analyzed by the HPLC-system.

2.5. Validation experiments

Validation was performed according to international and

national guidelines with respect to selectivity, carry over, limit of
quantification (LOQ), linearity, accuracy, precision, freeze thaw and
processed sample stability [21,22].
Selectivity for exogenous compounds (HA and 2-MHA) was
tested with six different blank urine samples. Selectivity for
endogenous compounds (CREA, UA, HVA, NA, IAA) was tested by
six fold injection of MP A. To test for carry over effects, MP A was
directly analyzed after the injection of Cal. 7 in three independent
runs. The LOQ was set at the lowest calibrator concentration, if
this could reliably be quantified (RSD and bias <20%) and the sig-
nal to noise ratio was above 10:1. A seven point calibration model,
covering characterized concentrations was selected [15]. Linearity Fig. 1. A: Extracted chromatogram (233 nm wavelength) of blank MP A showing no
and calibration model was tested by analyzing six replicates of Cal. interferences. B: Extracted chromatogram (233 nm wavelength) of QC level low (in
1–7. For accuracy and precision, freshly prepared QC samples (low, MP A) showing UA (1), CREA (2), HVA (3), NAA (4), HA (5), IAA (6), and 2-MHA (7).

med, high) were analyzed in duplicate on each of eight days. Accu-

racy was calculated in terms of bias as the percent deviation of the
mean calculated concentration from the corresponding theoretical
concentration [21]. Intra-day precision (intra-day) and inter-day
precision (inter-day) were calculated as relative standard deviation
(RSD), according to Peters et al. [21].. For QC low, the acceptance
criterion was set to 20% CV and RSD, respectively; for QC med and
high to 15% CV and RSD, respectively. To estimate benchtop stabil-
ity of processed sample under the conditions of analysis, a sample
of Cal 4 was prepared in an adequate amount (500 ␮L) and placed
into the autosampler. This sample was analyzed in triplicate on
eight days and resealed after the daily measurement. Freeze–thaw
stability was tested with a spiked urine sample (UA 225 mg/L,
CREA 900 mg/L, HVA 22.5 mg/L, NAA 22.5 mg/L, HA 225 mg/L, IAA
49.5 mg/L, and 2-MHA 45 mg/L) after one and two freeze (23 h at
−20 ◦ C) and thaw (1 h at room temperature) cycles. Stabilities were
tested by comparing absolute peak areas of each analyte. Analytes
were stable, if the average result of the stability samples was within
90–110% of the corresponding average result of the control samples
and the 90% confidence interval of the stability samples was within
80–120% of the corresponding average value of the control samples.
2.6. Proof of applicability
External QC samples (for UA and CREA, KU 01/12 sample
A and B) were used for routine application. Additional three
different internal QC samples (low, medium, high concentra-
tion) were prepared in MP A as described above and also used
for this purpose. In order to proof the applicability, 50 LiNA
cohort study urine samples were analyzed. Results for CREA were
statistically compared to the results obtained with the estab-
lished method for analyzing CREA and to the literature data
[15,23–27]. The participation on the LiNA study was voluntary and
approved by the Ethics Committee of the University of Leipzig
(file refs. 046-2006; 160-2008; 160b/2008; 144-10-31052010;
113-11-18042011; 206-12-02072012). All samples were handled
according to the institutional protocol and regulations.
the endogenous compounds but no other signals were detected.
Hence, the method was shown to be selective. An example chro-
matogram of spiked MP A (QC level low) is given in Fig. 1B. The LOQ
was determined for all analytes at the lowest calibrator concentra-
tion (see Table 1) in MP A. No carry over effects were observed,
which shows that the needle flushing is sufficient in this method.
All analytes were linear in the selected calibration range and a non
weighted calibration model was chosen. The analysis of internal
QC samples showed acceptable results for all analytes at all moni-
tored concentrations (see Table 1). A minimal decrease (5%) of the
peak areas during the eight day lasting bench top stability tests was
only monitored for CREA. However, using this kind of auto sam-
pler, the maximal amount of vials was 80. As the sample to sample
analysis time was 25 min, benchtop stability must only be guaran-
teed for 80 × 25 min = 33.3 h. Variations of the peak areas after two
freeze/thaw cycles were between 5.4% (2-MHA) and 9.1% (HA) with
no significant tendency except for HA, where a decrease of 8.9% was
detected. Nevertheless, as all stability parameters were within the
acceptance criteria, stability was shown for all analytes.
According to literature sufficient LOQ concentrations were
obtained in MP A as given in Table 1 [15]. Unfortunately these could
not be reproduced in authentic urine matrix. A reliable identifica-
tion using library matching, which is in addition to the signal to
noise ratio, a prerequisite of a detection and quantification was only
possible for UA, CREA, HA, and 2-MHA. The obtained LOQs for HA
and CREA were similar as or double than those reported by Antunes
et al. also using an HPLC-DAD analyzer but three fold higher than
those reported for CREA, UA; HA, and 3-IAA using a LC–MS method
[11,19]. The spectra of HVS, IAA, and NSA were different to those
recorded in the library using MP A although the retention time was

3. Results and discussion

As endogenous urinary metabolites are unavoidable substances

in blank urine, selectivity was firstly tested in MP A. As most
of the compounds provided an absorbance maximum of approx.
233 nm, this wavelength was used for all analytes. No interfer-
ing signals were detected at the corresponding analyte retention
times (Fig. 1A). After injection of blank urine samples signals for
Fig. 2. Extracted chromatogram (233 nm wavelength) of an authentic urine sample
showing UA (1), CREA (2), HVA (3), NAA (4) and, HA (5). CREA concentration was
676.7 mg/L. Creatinine normalized concentrations were 586.8 (UA), 140.4 (HVA),
137.4 (NAA), and 663.6 (HA) mg/L/g/L. For IAA (6) the concentration was below
LOQ, while 2-MHA (7) was not detected.
 D. Remane et al. / J. Chromatogr. B 998 (2015) 40–44 43

Fig. 3. Bland–Altman plot/comparison of quantitative CREA results (50 authentic urine samples) obtained by the presented and the established method.

Table 2
Quantification results from n = 50 LiNA urine samples with frequency of quantification [%], concentration range, median, and mean after creatinine normalization [mg/L/g/L]
and literature data of healthy male (m) and female (f) Caucasians for all analytes.

Analytes Frequency of Concentration range Median concentration Mean concentration Literature data
quantification [%] [mg/L/g/L] [mg/L/g/L] [mg/L/g/L]

Min Max Concentration [mg/L/g/L] Reference

UA 100 106.5 2278.6 541.4 597.5 310.9 (m/f) [24]

CREA 100 240 1935 761.7 835.0 1400.0 (m/f) [25]
HVA 40 16.9 327.5 85.9 104 9.0 (m/f) [26]
NAA 4 26.7 47.1 36.9 36.9
HA 100 79.2 2983.5 599.3 696.0 328.0 (f); 363.1 (m/f) [25,27]
IAA 0 0 0 0 0 5.3 (m/f) [25]
2-MHA 98 8.4 69.2 23.5 28.0 22.0 (f) [24]

UA = uric acid; CREA = creatinine; HVA = homovanillic acid; NAA = niacinamide; HA = hippuric acid; IAA = indole-3-acetic acid; 2-MHA = 2-methylhippuric acid.

identical. This might be explained by co-elution of broad matrix
peaks leading to a high background, which cannot be countervailed
by using automatic background compensation algorithms. A library
search against reference spectra obtained in matrix was not suitable
to compensate this problem. Nevertheless, unacceptable quantifi-
cation results for six different urine specimen spiked at the LOQ
were obtained for HVA, NAA, and 3-IAA and therefore, LOQs were
adopted (Table 1).
3.1. Proof of applicability
External QC samples (for UA and CREA, KU 01/12 sample A and
B) were used during the validation process and for testing routine
application. Bias and precision results for both QC samples were
within the acceptable range of ±15%. Therefore, calibration in MP
A was shown approvable for the analysis of UA and CREA.
50 authentic LiNA urine samples were analyzed with the
described method. A representative chromatogram for an authen-
tic urine samples is given in Fig. 2. Quantitative CREA results were
compared with those obtained using the method described by
Rolle-Kampczyk et al. [23]. As shown in Fig. 3 low differences (rang-
ing from −120 mg/L to +110 mg/L) were observed. These differ-
ences were not influenced by the absolute concentration of CREA.
However, the new method (mean concentration = 942.4 mg/L) pro-
vided 29.0 mg/L lower mean CREA concentrations – given in the
Bland–Altman plot (mean of differences, X-axis)– in comparison
to the established method (mean concentration = 971.4 mg/L). This
negligible difference of approximately 3% may be explained by dif-
ferent calibrations, calibration models, weighting or dilution biases.
Only a few times NAA (4%) and no IAA (0%) were quantified
in the 50 LiNA samples as indicated in Table 2. Even if urine LOQ’s
seemed unsuitable to cover normal urine concentrations those ana-
lytes should be included in the described method. If high urine
concentrations were observable for those compounds this infor-
mation can be used for stratification in statistical analyses. HVA
was quantified in 40% of the 50 LiNA urine samples. Interestingly
the median and mean concentrations were approx. 10 times higher
in comparison to the reference values described by Guneral and
Bachmann [26]. In comparison to literature data higher values were
observed for UA, HVA, and HA. For CREA lower concentrations were
determined as depicted in Table 2. As reference and obtained data
were normalized to CREA, this could be one explanation for higher
values. Obtained 2-MHA concentrations were in the same range as
described by Inoue et al. [24]. For NAA no reference data for healthy
Caucasian male or females (>18 years old) was available.
Summary
The analytical portfolio for analysis of cohort study urine sam-
ples was successfully extended by additional six analytes. All
compounds showed sufficient validation parameters according to
recent guidelines. Routine application was tested for 50 authentic
urine samples. Comparison of the new method with the established
method showed a high level of agreement for quantitative CREA
results.
44 D. Remane et al. / J. Chromatogr. B 998 (2015) 40–44
Acknowledgments
The authors thank their colleagues Sven Baumann, Ralph Fel-
tens, Susanne Pfeiffer, and Brigitte Winkler for scientific discussion
and technical assistance.
References
[1] P.J. Lioy, S.M. Rappaport, Environ. Health Perspect. 119 (2011) A466.
[2] E.S. Epel, E.H. Blackburn, J. Lin, F.S. Dhabhar, N.E. Adler, J.D. Morrow, R.M.
Cawthon, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 17312.
[3] C.P. Wild, Int. J. Epidemiol. 41 (2012) 24.
[4] G. Herberth, M. Bauer, M. Gasch, D. Hinz, S. Röder, S. Olek, T. Kohajda, U.
Rolle-Kampczyk, M. von Bergen, U. Sack, M. Borte, I. Lehmann, Maternal and
cord blood miR-223 expression associates with prenatal tobacco smoke
exposure and low regulatory T-cell numbers, J. Allergy Clin. Immunol. 133 (2)
(2014) 543–550, e4.
[5] D. Hinz, M. Bauer, S. Roder, S. Olek, J. Huehn, U. Sack, M. Borte, J.C. Simon, I.
Lehmann, G. Herberth, LS Group, Allergy 67 (2012) 380.
[6] K. Weisse, I. Lehmann, D. Heroux, T. Kohajda, G. Herberth, S. Roder, M. von
Bergen, M. Borte, J. Denburg, Clin. Exp. Allergy 42 (2012) 1337.
[7] K.M. Junge, F. Hornig, G. Herberth, S. Roder, T. Kohajda, U. Rolle-Kampczyk, M.
von Bergen, M. Borte, J.C. Simon, D. Heroux, J.A. Denburg, I. Lehmann, Clin.
Immunol. 152 (2014) 68.
[8] M.S. Gersch, R.J. Johnson, Nephrol. Dial. Transplant. 21 (2006) 3046.
[9] G. Ng, E.M. Chau, Y. Shi, Arch. Immunol. Ther. Exp. (Warsz) 58 (2010)
273.
[10] C.M. John, R. Ramasamy, G. Al Naqeeb, A.H. Al-Nuaimi, A. Adam, Curr. Med.
Chem. 19 (2012) 5181.
[11] M.V. Antunes, C.G. Niederauer, R. Linden, Clin. Biochem. 46 (2013) 1276.
[12] T. Sumiyoshi, T. Yotsutsuji, M. Kurachi, H. Itoh, K. Kurokawa, O. Saitoh,
Neuropsychopharmacology 19 (1998) 70.
[13] A.L. Kau, P.P. Ahern, N.W. Griffin, A.L. Goodman, J.I. Gordon, Nature 474 (2011)
327.
[14] M. Li, M. Wang, S.M. Donovan, Semin. Reprod. Med. 32 (2014) 74.
[15] D.S. Wishart, C. Knox, A.C. Guo, R. Eisner, N. Young, B. Gautam, D.D. Hau, N.
Psychogios, E. Dong, S. Bouatra, R. Mandal, I. Sinelnikov, J. Xia, L. Jia, J.A. Cruz,
E. Lim, C.A. Sobsey, S. Shrivastava, P. Huang, P. Liu, L. Fang, J. Peng, R. Fradette,
D. Cheng, D. Tzur, M. Clements, A. Lewis, A. De Souza, A. Zuniga, M. Dawe, Y.
Xiong, D. Clive, R. Greiner, A. Nazyrova, R. Shaykhutdinov, L. Li, H.J. Vogel, I.
Forsythe, Nucleic. Acids Res. 37 (2009) D603.
[16] R.W. Pero, Curr. Clin. Pharmacol. 5 (2010) 67.
[17] Y. Zuo, C. Wang, J. Zhou, A. Sachdeva, V.C. Ruelos, Anal. Sci. 24 (2008) 1589.
[18] M.K. Shirao, S. Suzuki, J. Kobayashi, H. Nakazawa, E. Mochizuki, J. Chromatogr.
B Biomed. Sci. Appl. 693 (1997) 463.
[19] L. Gu, X. Wang, Y. Zhang, Y. Jiang, H. Lu, K. Bi, X. Chen, J. Sep. Sci. 37 (2014)
1058.
[20] K.M. Kim, G.N. Henderson, X. Ouyang, R.F. Frye, Y.Y. Sautin, D.I. Feig, R.J.
Johnson, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 877 (2009) 2032.
[21] F.T. Peters, M. Hartung, M. Herbold, G. Schmitt, T. Daldrup, F. Musshoff,
Toxichem Krimtech 76 (2009) 185.
[22] I.H.T. Guideline, Website: http://www.ich.org/fileadmin/Public Web Site/ICH
Products/Guidelines/Quality/Q2 R1/Step4/Q2 R1 Guideline.pdf (2010).
[23] U.E. Rolle-Kampczyk, M. Rehwagen, U. Diez, M. Richter, O. Herbarth, M. Borte,
Arch. Environ. Health 57 (2002) 326.
[24] O. Inoue, K. Seiji, T. Suzuki, T. Watanabe, H. Nakatsuka, H. Satoh, M. Ikeda,
Bull. Environ. Contam. Toxicol. 47 (1991) 204.
[25] S. Bouatra, F. Aziat, R. Mandal, A.C. Guo, M.R. Wilson, C. Knox, T.C. Bjorndahl,
R. Krishnamurthy, F. Saleem, P. Liu, Z.T. Dame, J. Poelzer, J. Huynh, F.S. Yallou,
N. Psychogios, E. Dong, R. Bogumil, C. Roehring, D.S. Wishart, PLoS One 8
(2013) e73076.
[26] F. Guneral, C. Bachmann, Clin. Chem. 40 (1994) 862.
[27] N.G. Psihogios, I.F. Gazi, M.S. Elisaf, K.I. Seferiadis, E.T. Bairaktari, NMR
Biomed. 21 (2008) 195.