A sensitive approach for simultaneous quantification of carbonyl and hydroxyl steroids using 96-well SPE plates based on stable isotope coded-derivatization-UPLC-MRM: Method development and application

Article abstract of DOI:10.1039/c8ra01372a

Steroid hormones are crucial substances that mediate a wide range of vital physiological functions. Because of the important biological significance of steroids, this paper presents a new targeted metabolic method based on adding stable isotope tags to hydroxyl containing and carbonyl containing steroid hormones with two pairs of synthesized derivatization reagents: deuterium 4-(dimethylamino)-benzoic acid (D₄-DMBA), and D₅-Girard P (D₅-GP) using of ultra performance liquid chromatography-multiple reaction monitoring (UPLC-MRM). Firstly, an Oasis PRiME hydrophilic-lipophilic balance (HLB) 96-well solid phase extraction plate was used to pretreat a number of biological samples simultaneously. Secondly, hydroxyl and carbonyl steroids were labeled using two pairs of synthetic reagents, namely DMBA and D₄-DMBA, and GP and D₅-GP, respectively. Thirdly, the mixed products were detected using UPLC-MRM and the mass spectroscopy conditions were optimized. Methodology development showed that the sensitivity was enhanced 1 to >500-fold. Finally, the new method was applied to analysis of urine samples of healthy males, females and rats. The results revealed that the method can be sensitive and reliable for simultaneous quantification of steroid hormones containing hydroxyl and carbonyl groups in 12 min in a single run. This method provided a powerful tool for studying the metabolic mechanism of steroids and contributed to the development of targeted metabolomics.

Full text of DOI:10.1039/c8ra01372a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
A sensitive approach for simultaneous
quantification of carbonyl and hydroxyl steroids
using 96-well SPE plates based on stable isotope
coded-derivatization-UPLC-MRM: method
development and application†
Cite this: RSC Adv., 2018, 8, 19713
Chuanxin Liu,‡a Xue Sheng,‡a Yuming Wang,‡a Jia Yin,a Wei Huang,a
a
d
Yunshuang Fan,bc Yubo Li
and Yanjun Zhang
*
*
Steroid hormones are crucial substances that mediate a wide range of vital physiological functions. Because
of the important biological significance of steroids, this paper presents a new targeted metabolic method
based on adding stable isotope tags to hydroxyl containing and carbonyl containing steroid hormones
with two pairs of synthesized derivatization reagents: deuterium 4-(dimethylamino)-benzoic acid (D4-
DMBA), and D5-Girard P (D5-GP) using of ultra performance liquid chromatography-multiple reaction
monitoring (UPLC-MRM). Firstly, an Oasis PRiME hydrophilic-lipophilic balance (HLB) 96-well solid phase
extraction plate was used to pretreat a number of biological samples simultaneously. Secondly, hydroxyl
and carbonyl steroids were labeled using two pairs of synthetic reagents, namely DMBA and D4-DMBA,
and GP and D5-GP, respectively. Thirdly, the mixed products were detected using UPLC-MRM and the
mass spectroscopy conditions were optimized. Methodology development showed that the sensitivity
was enhanced 1 to >500-fold. Finally, the new method was applied to analysis of urine samples of
healthy males, females and rats. The results revealed that the method can be sensitive and reliable for
simultaneous quantification of steroid hormones containing hydroxyl and carbonyl groups in 12 min in
a single run. This method provided a powerful tool for studying the metabolic mechanism of steroids
and contributed to the development of targeted metabolomics.
Received 12th February 2018
Accepted 2nd May 2018
DOI: 10.1039/c8ra01372a
rsc.li/rsc-advances
important role in the maintenance of life, regulation of sexual
function, development of the body and so on.1,2 Many papers in
Introduction
the literature have reported that steroid hormones are associ-
ated with diseases, such as heart disease, high blood pressure,
endocrine disorders, prostate cancer and so on.3–6 Considering
the important biological implications of steroid hormones, it is
necessary to build a sensitive and reliable method to determine
multiple steroid hormones. Quantitatively measuring steroid
hormones will also play a vital role in understanding disease
progression and individual biochemical responses.7
The main methods reported for conventional analysis of
steroid hormones include radio immunoassay (RIA),8 enzyme
immunoassay (EIA),9 and gas chromatography-mass spectrom-
etry (GC-MS).10,11 However, these methods experienced poor
sensitivity, were labor intensive and lacked specicity. Chro-
matographic separation and identication using MS, enabled
rapid qualitative analysis and accurate quantication of
metabolites. Because of the high sensitivity of MS, lower levels
of metabolites could be detected. Therefore, chromatography-
mass spectrometry has an invaluable role in metabolomics
research.12,13 Compared with the MS scan of non-target metab-
olomics, multiple reaction monitoring (MRM) in triple
Steroid hormones are produced by the adrenal cortex, gonads
and placenta, whereas cholesterol, which is in a class of four
ring aliphatic hydrocarbons, is produced by an enzyme cata-
lyzed by cytochrome P450. Steroid hormones have cyclo-
pentane, phenanthrene,
a mother nucleus with unique
physiological and biochemical functions, which play an
aTianjin State Key Laboratory of Modern Chinese Medicine, School of Traditional
Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, 312
Anshan West Road, Tianjin 300193, China. E-mail: yaowufenxi001@sina.com; Fax:
+86-22-59596221; Tel: +86-22-59596221
bState Key Laboratory of Separation Membranes and Membrane Processes, Tianjin
Polytechnic University, Tianjin 300387, China
cSchool of Environmental and Chemical Engineering, Tianjin Polytechnic University,
Tianjin 300387, China
dTianjin State Key Laboratory of Modern Chinese Medicine, Tianjin University of
Traditional Chinese Medicine, 312 Anshan West Road, Tianjin 300193, China.
E-mail: tianjin_tcm001@sina.com
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c8ra01372a
‡ Co-rst authors.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19713

View Article Online
RSC Advances
Paper
quadrupole-mass spectrometry (QQQ-MS) is increasingly used
for work on targeted metabolism or the concentration analysis
of homologues. Its high throughput and precise monitoring of
complex matrices across a wide range of concentrations are its
inherent advantages.14 Most researchers only focus on steroids
that contain a kind of groups. However, the current study
simultaneously quantied multiple classes of perssad, which
improved the study of diseases related to steroid hormones.
However, steroid hormones increased the difficulty of their
detection because of their low ionization efficiency and low
concentration characteristics.
Experimental
Reagents and materials
Estradiol (E1), estriol (E2) and cortisone were purchased from
the Chinese National Institute for the Control of Pharmaceu-
tical and Biological Products. 17a-Hydroxyprogesterone
(17aOH-PROG), estrone (E3) and pregnanediol were purchased
from TCI (Shanghai) Chemical Industry Development Co., Ltd.
Progesterone (PROG) was purchased from Sigma-Aldrich (USA).
Testosterone (TES) was purchased from the Tianjin Silan
Technology Co., Ltd. Dehydroepiandrosterone (DHEA) was
purchased from Adamas Regent. Pregnenolone (PREG) was
purchased from Ark Pharm. Corticosterone, androsterone and
11-deoxycorticosterone were purchased from Dre (Germany).
5b-Pregnane-3,20-dione, 11b-hydroxyandrost-4-ene-3,17-dione,
Chemical derivatization is
a commonly used sample
pretreatment method in chromatographic analysis. It refers to
the chemical reaction of a special chemical derivatization
reagent with the analyte to change the physical and chemical
properties of the analytes to be tested. Use of derivatization
enhances the ability of analysis and the detection of steroids.15
Traditional quantitative analysis used to use a standard curve
method which is generally used to measure the component in
vivo. However the composition of the actual sample is different
in vivo, and using the traditional methods will tend to bring
some errors to the measurement. Some endogenous substances
cannot be accurately quantied using a standard curve method.
For complex mixtures, such as biological samples and envi-
ronmental samples, the matrix effect can signicantly change
the ionization behavior of target analytes. Even for the same
analyte, the ionization efficiency sometimes varies between two
single runs.16,17 Stable isotopic derivatization is an alternative to
introducing stable isotopes to the analyte, which can be used as
a substitute for the internal standard. Before analysis, an
analogue of the analyte was marked with stable isotopes, and
were added as the internal standard to improve the accuracy by
19-hydroxyandrostenedione,
11b,17a,21-trihydroxy-5b-
pregnane-3,20-dione, 2-methoxyestrone, cortol and tetrahy-
drocortisol were purchased from TRC (Canada). 17a-Hydrox-
ypregnenolone (17aOH-PREG) and tetrahydrocorticosterone
(THB) were purchased from ChromaDex (USA). The purity of all
the standard products was greater than 98%.
HPLC-grade acetonitrile was purchased from Oceanpark
(Gothenburg, Sweden). Distilled water was obtained from
Watsons (Guangzhou, China). Ammonium acetate was
purchased from Sigma-Aldrich (USA). The Ostro 96-well SPE
plate was obtained from Waters Co., Ltd. (USA). Isoprenaline
was obtained from Main Luck Pharmaceutical Inc. (China).
The relationships and structures of all steroid hormone
related metabolic pathways are shown in Fig. 1.
Synthesis of the derivatization reagents
minimizing the matrix effect and ionization difference, so that Ethanol (EtOH) (12 mL), ethyl chloroacetate (6.13 g, 50 mmol)
the absolute/relative quantication will be more accurate when and D5-anhydrous pyridine (3.96 g, 50 mmol) were added to the
compared with the traditional standard curve method.
ask to make a total of 100 mL. The mixture was heated and
Based on the previously described research situation, a new reuxed for 6 h. The resulting mixture was cooled to room
method was proposed which is based on a 96-well solid-phase temperature and then hydrazine hydrate (N2H4$H2O, 2.5 g, 50
extraction (SPE) plate combined with a stable isotope labeling mmol) was added. The product was allowed to precipitate for
technique for simultaneous quantication of steroid hormones 30 min. The solid product was ltered out, washed with cold
using means of UPLC-MRM. Firstly, an Oasis PRiME absolute EtOH, and then dried in a vacuum drying oven to
hydrophilic-lipophilic balance (HLB) 96-well SPE plate was used obtain D5-GP. Finally, nuclear magnetic resonance spectroscopy
to pretreat all the biological samples simultaneously. Then, (NMR), HPLC and MS were used to characterize the conrma-
hydroxyl and carbonyl steroids were labeled using two pairs of tion and purity of the product. Fig. 2A shows the synthetic
synthetic reagents, namely by 4-(dimethylamino)-benzoic acid scheme for D5-GP.
(DMBA) and deuterated-DMBA (D4-DMBA), Girard P (GP) and
Glacial acetic acid (0.6 mL) was added dropwise to 4-carba-
deuterated-GP (D5-GP), respectively. Thirdly, the mixed prod- mate (1.5 g) and dissolved in tetrahydrofuran (18 mL) and
ucts were detected using UPLC-MRM and the mass spectros- 1.8 mL of 20% D2-formaldehyde solution, followed by stirring.
copy conditions were optimized. Methodology development has The solution was kept at room temperature for 1.5 h, then, 0.7 g
shown that the method is sensitive and stable for the quanti- of sodium cyanoborohydride (NaBH3CN) was added to the
cation of steroids. Finally, the new method was applied to solution. Aer 30 min, 3.7 mL of D2-formaldehyde solution was
urine samples from humans and rats. All the results revealed added, followed by 1.5 g of NaBH3CN. The resulting solution
that the method is sensitive and reliable for simultaneous was kept overnight at room temperature. The reaction liquid
quantication of steroid hormones containing hydroxyl and was poured into 50 mL of ice water and extracted twice with
carbonyl groups with a 12 min reaction in complex biological 100 mL of ethyl acetate. The ethyl acetate layer was dried using
samples. The method provides a powerful tool for studying the anhydrous magnesium sulfate and concentrated under pres-
metabolic mechanism of steroids and contributes to the sure. The concentrated product was eluted using 200–300 mesh
development of targeted metabolomics.
column chromatography and silica gel. The elution solvent of
19714 | RSC Adv., 2018, 8, 19713–19723
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 1 All 22 steroid hormone structures and their related metabolic pathways.
methanol/dichloromethane was used with an elution gradient
of 0–10% to obtain the product. Finally, NMR, HPLC and liquid
chromatography-mass spectrometry (LC-MS) were used to
detect and characterize the samples. Fig. 2B shows the synthetic
scheme for D4-4-(dimethylamino)benzoic acid (D4-DMBA).
Preparation of a standard solution
A portion (1 mg) of each of the 22 standard steroid hormone
products was weighed and placed in a 1 mL capacity bottle and
dissolved in methanol to make a 1 mg mLꢀ1 standard solution.
Fig. 2 Reaction schemes for synthesis of the stable isotope-labeling reagent (A) D5-GP and (B) D4-DMBA.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19715

View Article Online
RSC Advances
Paper
Fig. 3 The derivatization scheme involved in this paper: (A) steroid hormones containing hydroxyl groups derivatized by DMBA; (B) steroid
hormones containing carbonyl groups derivatized by GP; (C) steroid hormones containing hydroxyl groups derivatized by D4-DMBA; and (D)
steroid hormones containing carbonyl groups derivatized by D5-GP.
Then the solution was diluted to the desired concentration in
a volumetric ask.
The 12 hour urine of male Wister rats (n ¼ 6) was collected,
ꢂ
and the urine was centrifuged at 4 C and 760 ꢃ g for 15 min.
The supernatant was centrifuged at 1040 ꢃ g for 8 min. Then,
the supernatant was extracted. Aer the supernatant was
combined with sodium azide (10 mL mLꢀ1), the samꢂple tube was
placed upside down. Finally, it was stored at ꢀ80 C in prepa-
ration for further research.
Animal treatment
The experimental animals were purchased from Sibei Fu
Experimental Animal Science and Technology Co., Ltd. (Beijing,
China), and the license number was SCXK (Jing) 2016-0011. The
animal experiments were performed at the Institute of Radia-
tion Medicine, Chinese Academy of Medical Sciences (Tianjin,
China). Male Wistar rats weighing 200 ꢁ 20 g were raised in an
SPF level lab. The rats were housed under the following condi-
tions: ambient temperature of 23 ꢁ 2 ꢂC and humidity of 35 ꢁ
5%. This study was approved by the Animal Ethics Committee
of Tianjin University of Traditional Chinese Medicine under
permit number TCM-2012-078-F01. All experimental proce-
dures were conducted in accordance with Chinese national
legislation and local guidelines.
Use of 96-well SPE plates for pretreatment samples
Urine samples were treated to precipitate the proteins rst.
Acetonitrile (1500 mL) were added to 500 mL of urine to precip-
itate the protein. The resultant mixture was mixed ultrasonically
in cold water for 10 min and vortex mixed for 1 min and then
ꢂ
centrifuged at 14 360 ꢃ g for 15 min at 4 C. The supernatants
were loaded into each well of the Ostro 96-well SPE plate
(Waters, USA) to remove the phospholipids. A vacuum was
applied to the plate for 5–8 min. Then, the pretreatment
samples were collected into centrifuge tubes, and nitrogen (N2)
was used to dry and concentrate the samples before GP and
DMBA derivatization.
Collection of biological samples
From healthy adult males (n ¼ 6) and females (n ¼ 6),
Derivatization
midstream urine was collected in the morning. The urine
samples were centrifuged at 760 ꢃ g for 15 min. The superna- D4-DMBA solution (100 mL, 2 mg mLꢀ1), 4-dimethylaminopyr-
tant obtained was centrifuged at 1040 ꢃ g for 8 min. Then, the idine solution (DMAP; 100 mL, 2 mg mLꢀ1) and 1-ethyl-3-(3-
supernatant was extracted. Samples were handled gently to dimethylaminopropyl) carbodiimide solution (EDC; 100 mL,
avoid turbidity from the precipitate. Aer the sample was 5 mg mLꢀ1) were added successively to the nitrogen dried urine
combined with sodium azide (10 mL mLꢀ1), the sample tube was samples using methylene chloride as a solvent (Fig. 3C). Aer
placed upside down aer adding the preservative so that the rigorous vortex mixing, the solution was transferred to a thick
preservative was dispersed evenly. Finally, it was stored at walled pressure bottle and sealed. The thick walled pressure
ꢂ
ꢀ80 C in preparation for further research.
bottle was kept at 75 ꢂC for 5 h. Aer cooling the bottle, the
19716 | RSC Adv., 2018, 8, 19713–19723
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Table 1 Mass spectrometry information and conditions of deuterium derivatization substances
Derivatization
reagent
Precursor
ion
Dwell
time (ms)
Collision
energy (eV)
Cone
voltage (V)
RT
Substance
Product ion
2.65
3.25
3.20
3.38
3.58
3.71
3.75
3.75
4.04
4.79
6.10
4.97
5.87
4.69
9.49
4.58
4.66
5.25
5.54
6.53
6.55
8.46
Cortisone
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
D5-GP
499.10
503.20
471.20
409.10
489.20
441.20
427.30
427.20
469.26
455.20
455.20
429.20
453.20
520.20
440.10
518.20
454.10
482.10
498.20
424.20
452.20
472.20
85.10
85.00
85.10
157.00
113.00
357.23
343.30
80.0
0.008
0.008
0.008
0.008
0.008
0.008
0.008
0.008
0.008
0.012
0.012
0.012
0.012
0.012
0.008
0.008
0.008
0.008
0.008
0.008
0.008
0.008
30
30
30
30
48
24
28
26
28
50
32
34
28
24
20
18
22
10
26
16
16
36
70
74
72
62
80
60
58
58
64
68
66
60
58
20
44
24
30
14
52
34
36
48
11b,17a,21-Trihydroxy-5b-pregnane-3,20-dione
17aOH-PREG
E3
THB
11b-Hydroxyandrost-4-ene-3,17-dione
TES
DHEA
11-Deoxycorticosterone
PREG
5b-Pregnane-3,20-dione
Androsterone
PROG
Cortol
E2
Tetrahydrocortisol
19-Hydroxyandrostenedione
17aOH-PROG
Corticosterone
E1
385.37
97.10
85.10
255.30
369.30
152.10
152.10
152.10
152.10
152.00
152.10
152.10
152.10
170.10
D5-GP
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
D4-DMBA
2-Methoxyestrone
Pregnanediol
sample was transferred to a centrifuge tube and centrifuged at an ACQUITY UPLC HSS T3 column (2.1 ꢃ 100 mm, 1.8 mm,
14 360 ꢃ g for 15 min to obtain the supernatant. Then, N2 was Waters). The column temperature was set to 40 ꢂC, and the ow
used to dry solution at 45 C for 2 min. The residue was dis- rate was set to 0.4 mL minꢀ1. The UPLC separation system
ꢂ
solved in 100 mL of 50% acetonitrile and the solution was then included a binary solvent system with mobile phase A (0.1%
centrifuged at 14 360 ꢃ g for 15 min to give the supernatant. ammonium acetate in water) and mobile phase B (acetonitrile).
The supernatant was then ready for use. DMBA solution (100 mL, The gradient proles were as follows: 0 min, (A: 100%); 0–1 min,
2 mg mLꢀ1, DMAP solution (100 mL, 2 mg mLꢀ1) and EDC (A: 100%); 1–2 min, (A: 100–70%); 2–5 min, (A: 70–55%); 5–6
solution (100 mL, 5 mg mLꢀ1) were added successively to the min, (A: 55–30%); 6–9 min, (A: 30–0%); 9–10 min, (A: 0–0%); 10–
standard solution containing the 22 steroids dried by N2 11 min, (A: 0–100%); 11–12 min, (A: 100–100%). The H-CLASS
(Fig. 3A). The rest of the methodology was the same as that used Xevo TQD was equipped with electrospray ionization in posi-
for urine samples.
tive and negative modes. In the positive ion multiple reaction
A dry urine sample was derivatized with 150 mL of D5-GP (10 monitoring (MRM) detectionꢂ mode, the MS parameters were:
mM) reagent in a methanolic solution of 1% acetic acid drying gas temperature: 460 C, capillary voltage: 1.5 kV, cone
(Fig. 3D). The sample was oscillated at room temperature for voltage: 50 V, atomizer: 7 bar, N2 solvent ow rate: 800 L hꢀ1
,
120 min and dried under N2 at 45 ꢂC. The residue was dissolved conical airow: 150 L hꢀ1, impact gas ow: 0.14 mL minꢀ1 and
in 100 mL of 50% acetonitrile, which was centrifuged at 14 360 evaporation gas and auxiliary gas: N2. The scanning mode was
ꢃ g centrifugal force for 15 minutes to give the supernatant. The used in the MRM and two sets of characteristic precursor ion/
supernatant was then ready for use. The dry standard mixture product ion pairs were selected, and their collision energy and
containing the 22 steroids was derivatized with 150 mL GP (10 declustering potential were optimized. A group of precursor
mM) reagent in a methanolic solution of 1% acetic acid ions/product ions with strong abundance was used for quanti-
(Fig. 3B). Other operations were the same as those performed tative analysis. The entire MRM scan was divided into four
with the urine samples.
segments to improve the operation efficiency and shorten the
Each of the two samples was derivatized with deuterated scanning cycle. The optimized parameters for the deuterated
reagents, and the two deuterated derivatization products were derivatization MS of the 22 endogenous steroids are shown in
mixed. The DMBA-standard or GP-standard (4 ng mLꢀ1) was Fig. S1 (ESI†) and Table 1. The optimized parameters and
respectively added to D4-DMBA-urine samples and G5-GP-urine detailed information for the MS of samples of 22 endogenous
samples by 1 : 4 ratio. Aer mixing, the samples were analysed steroids without derivatization are shown in Table S1 (ESI†).
using UPLC-MRM.
Data processing
Apparatus and operation conditions
Data acquisition, peak detection and peak alignment were
UPLC analysis was performed on a UPLC H-CLASS Xevo TQD performed using the Markerlynx function of the MassLynx
system (Waters, USA). Sample aliquot (10 mL) were injected into soware (version 4.1, Waters, USA). MS conditions were taken
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19717

View Article Online
RSC Advances
Paper
Fig. 4 Both D5-Girard P and D4-DMBA chemical synthesis and characterization. (A) showed the successful synthesis of D5-Girard P by 1H-NMR
(400 MHz-D2O); (B) showed the characterization of D5-Girard P by means of LC-MS (ES-API, Pos); (C) showed the successful synthesis of D4-
DMBA by 1H-NMR (400 MHz-DMSO-d6); and (D) showed the characterization of D4-DMBA by means of LC-MS (ES-API, Pos).
from the Intellistart (Waters, USA) using MS soware optimi- GP,30–34
2-hydrazino-1-methylpyridine,35–37
2-hydro-
zation with repeated injections of a single component test zinopyridine,38,39 2-hydrazino-4-(triuoromethyl)-pyrimidine40
solution. To obtain the optimal MS conditions for each analyte, and quaternary aminooxy.41
the parameters were optimized, including the precursor ion, ion However, these derivatization reagents are complicated to
energy, and capillary voltage. Quantitative calculation of use. They usually target only a class of hormones, which
components use soware Targetlynx (Waters, USA).
contains specic structures. DMBA has the advantage of
increasing the ionization efficiency of steroid hormones by
introducing the easily protonated tertiary amines. GP reagent
has the advantage of stably introducing a permanently charged
quaternary amine group to enhance the ionization efficiency. In
this case, the synthesis of D5-GP and D4-DMBA could be used to
provide an easily prepared isotope-labeled steroid hormone.
Based on the previously described investigation, in this study
D5-GP and D4-DMBA reagents were synthesized. D5-GP reagent
(9 g) was obtained with a yield of 89%. For the D4-DMBA reagent
(350 mg), the yield was 23.3% as determined using NMR and
Results and discussion
Characterization and selection of derivatization reagents
Using suitable derivatization reagents is the key for detecting
and quantifying steroids. Papers in the literatures report that
reagents for the derivatization of hydroxyl groups were mainly
picolinic acid,18 fusaric acid,19 2-uoro-1-methylpyridinium
p-toluenesulfonate,20,21 isonicotinoyl azide,22 and DMBA.23
Reagents for the derivatization of carbonyl groups are mainly
hydroxylamine,24–27 methoxylamine,28 Girard reagent T,29
19718 | RSC Adv., 2018, 8, 19713–19723
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Table 2 Lower limit of detection (LLOD), LLOQ, linear range, regression coefficient and stability for methodology developmenta
Derivatization LLOD
reagent
LLOQ
Linear
Regression
Stability within
Steroid hormones
(pg mLꢀ1
)
(pg mLꢀ1
)
range (pg mLꢀ1
)
coefficient (R2) 24 h RSD (%)
17aOH-PROG
E1
DMBA
DMBA
DMBA
DMBA
DMBA
DMBA
DMBA
DMBA
DMBA
GP
GP
GP
GP
GP
GP
GP
GP
GP
2
5
2
5
20
10
20
50
100
<2
2
2
2
2
5
2
5
5
5
5–20 000
0.7944
0.9822
0.9881
0.9942
0.9820
0.9941
0.9911
0.9915
0.9709
0.9999
0.9979
0.9991
0.9984
0.9981
0.9997
0.9994
0.9993
0.9999
0.9909
0.9967
0.9986
1.3
3.2
6.2
3.1
12
10
20
50
50
20
50
200
200
5
5
5
5
5
10
10
10
20
50
1000
5000
10–20 000
20–20 000
50–20 000
50–20 000
20–20 000
50–20 000
100–40 000
200–40 000
5–20 000
5–20 000
5–20 000
5–20 000
5–20 000
10–20 000
10–20 000
10–20 000
20–20 000
50–40 000
1000–40 000
5000–40 000
2-Methoxyestrone
Tetrahydrocortisol
Corticosterone
E2
Pregnandiol
19-Hydroxyandrostenedione
Cortol
Androsterone
11-Deoxycorticosterone
THB
5b-Pregnane-3,20-dione
PREG
E3
TES
17aOH-PREG
DHEA
Cortisone
15
3.4
ND
1.0
2.5
5.3
8.8
1.4
ND
7.2
6.9
1.5
12
GP
20
500
2500
2.0
ND
ND
11b,17a,21-Trihydroxy-5b-pregnane-3,20-dione GP
PROG
GP
a
ND: not detected.
HPLC-MRM. The purity of synthetic D5-GP and D4-DMBA which simultaneously contained carbonyl and hydroxyl groups,
reached above 95% (Fig. 4)
were difficult to quantify. In the early stage, two reagents were
used for the derivatization. The results are compared and
shown in Table S4 (ESI†). Aer optimization, the reagent with
the best response value was used as the nal derivatization
reagent.
The UPLC-MRM conditions were optimized for the efficient
separation of isomers and the enhancement of the mass
response. For the mobile phase, the study compared formic acid
water to ammonium acetate by the effect of the response value
of steroid hormone detection. The results indicated that
ammonium acetate can increase the ionization ratio of steroid
hormones. The efficient separation of isomers was obtained
using acetonitrile, water, and ammonium acetate as the mobile
phase. Good MS results were dependent on the ions from the
ammonium acetate. The optimized chromatographic peaks are
presented in Fig. S1 (ESI†).
Optimization
In the derivatization of hydroxyl steroid hormones, methylene
chloride was used as a solvent. These steroids were derivatized
with DMBA (D4-DMBA) with DMAP and EDC as the catalysts.
DMAP and DMBA (D4-DMBA) could be dissolved in dichloro-
methane, whereas EDC dissolved better in water. Therefore, in
the early stage, EDC was dissolved in water to compare the MS
response of the products aer derivatization. The results are
shown in Table S2 (ESI†). Of the 22 steroids, only eight
substances could be detected, and their response value was
lower than that of the products identied using dichloro-
methane as a solvent. These results were probably because of
the density of water and dichloromethane. This led to a deriva-
tization reaction with a response that was not complete at the
interface.
Methodology development
In order to remove the endogenous interfering substances
from the urine samples and to optimize the sample pretreat- Steroid hormone mixture solutions of 2, 5, 10, 20, 50, 100, 200,
ment method, a 96-well semi-automatic SPE method was used 500, 1000, 2000, 5000, 10 000, 20 000, 40 000 pg mLꢀ1 were
to handle a large number of samples.42,43 This method can subjected to extraction, derivatization and UPLC-MRM analysis
remove phospholipid and proteins, which were the main to evaluate the lowest limit of quantication (LLOQ) and linear
components to interfere with the detection of endogenous range for the 22 derivatized steroid hormones. LLOQ was
substances.
determined at the signal to noise ratio of 10. The results show
Use of DMBA introduced tertiary amines, which were easily that the corresponding non-deuterated steroid hormone LLOQ
protonated, to label the hydroxyl groups. GP had the ability to ranged from 5 pg mLꢀ1 to 40 000 pg mLꢀ1. Derivatization
introduce permanently charged quaternary amine groups to reagents were synthesized for labeling the steroid hormones to
label the carbonyl groups. To identify the structures of the 22 improve the sensitivity 1 to >500-fold with LLOQs ranging from
steroids, hydroxyl steroids were derivatized using DMBA and 5 pg mLꢀ1 to 5000 pg mLꢀ1. Comparison of sensitivity of the
carbonyl steroids were derivatized using GP. The steroids, UPLC-MRM detection of steroid hormones with and without
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19719

View Article Online
RSC Advances
Paper
19720 | RSC Adv., 2018, 8, 19713–19723
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 5 Peaks of quantification based on D/H stable isotope DMBA and GP labeling: extracted ion chromatogram of heavy and light labeled
substances.
derivatization by DMBA and GP are shown in Table S3 (ESI†). analysis to evaluate the inter-batch and intra-batch precision (n
The results for the derivatized steroid hormones showed a good ¼ 6). The RSD values were also calculated. The results showed
linear range. The regression coefficients ranged from 0.7944– that the RSD range of the inter-batch accuracy was less than
0.9999 (Table 2).
15%. The RSD of intra-batch range accuracy was also less than
Recoveries were investigated at low, medium and high levels. 15%. These results showed that the precision complied with the
The different volumes of mixed steroid solutions were added to requirements of the methodology (Table 3).
500 mL of quality control (QC) samples. Then extraction, deriv-
atization and UPLC-MRM analysis were performed to evaluate
the recoveries. The results showed that the recovery rate
complies with the requirements of methodological (Table 3).
Quantitative analysis of targeted metabolism
An analyte quantication approach based on the MRM with
stable isotope labeled standard analytes provided absolute
steroid hormone quantication values, which were based on the
QC samples (500 mL) were analyzed at 0, 4, 8, 12, 16, 20,
24 hours to test their stability within 24 hours. The relative
addition of known quantities of stable isotope labeled internal
standard deviation (RSD) value was calculated. The results
standards which were added to the urine samples. Quantication
was achieved by comparing ion signals from the isotope labeled
standards and the analytes. The biological samples labeled by
showed that for the steroid hormones, the RSD were close to
15%, indicating that the method was stable (Table 2).
A mixed steroid solution was added to a 500 mL QC sample
deuterium derivatization reagents were mixed with a certain
and then subjected to extraction, derivatization and UPLC-MRM
proportion of a steroid hormone standard labeled by non-
Table 4 Steroid levels in urine of healthy adult males, females and ratsa
Average content of urine
Average content of urine
samples in male (pg mLꢀ1
)
samples in female (pg mLꢀ1
)
Derivatization
reagent
Average content of urine
45
Substance
This study
Reference
This study
Reference 45
samples in male rats (pg mLꢀ1
)
Cortol
E2
DMBA
DMBA
DMBA
DMBA
DMBA
DMBA
GP
GP
GP
GP
GP
1326
—
13 340
46.14
3166
396.4
12.15
37.88
23 479
ND
51.04
244.9
15 988
5089
34.13
ND
—
34
2588
1013
20
ND
216.1
884.1
122.8
52.89
74.95
38 766
3343
188.1
228.3
4234
366
902
466
41
—
—
508
877
—
16 626
5329
ND
ND
—
25.22
2812
ND
14.54
84.00
ND
ND
ND
ND
16.68
1602
ND
ND
ND
Tetrahydrocortisol
Corticosterone
E1
2-Methoxyestrone
Cortisone
17aOH-PREG
E3
THB
TES
DHEA
11-Deoxycorticosterone
PREG
5b-Pregnane-3,20-dione
Androsterone
—
—
ND
242
—
17 884
6851
35
ND
—
5999
GP
GP
GP
GP
1989
43.41
75.74
262.8
1145
ND
2773
GP
3529
312.9
a
ND: not detected. —: relevant data not found.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19721

View Article Online
RSC Advances
Paper
deuterium derivatization reagents. Chromatograms showed two
chromatographic peaks, which had different numbers of
deuterated elements with m/z at the same retention times (RT).
For example, the m/z value of the derivative products of D5-GP will
be ve more than that of the GP derivatives. The response value
of the peak was proportional to the content of the product. So,
accurate quantication of analytes was obtained using this
technology, and the formula of quantication was:
Conclusion
There is increasing evidence that steroid hormones play an
important role in the development and progression of various
diseases and quantitative metabolomics. However, few methods
were established for detecting and quantifying steroid hormones
comprehensively and high-sensitively. In this study, a stable
isotope labeled derivatization method, using HPLC-MRM, was
developed for targeted analysis of steroid hormones, which
contain both carbonyl and hydroxyl groups. DMBA, D4-DMBA,
GP, and D5-GP were synthesized as novel derivatization reagents
for labeling steroid hormones and use of these improved the
sensitivity from 1 to >500-fold with LLOQ ranging from 5 pg mLꢀ1
to 5000 pg mLꢀ1. The method had good inter-batch accuracy,
intra-batch accuracy and stability over 24 h, and almost all the
RSD were lower than 15%. The results showed that accurate
relative and absolute quantication can be achieved using stable
isotope labeling strategies. In addition, the method was
successfully applied to the determination of healthy urine steroid
hormones in males and females and in male rats. This method of
derivation of steroid hormones based on stable isotope labeling
techniques using the UPLC-MRM analysis platform can be used
in steroid hormone metabolomics studies. Finally, the method
established in this study not only accurately detected and quan-
tied multiple steroid hormones in 12 min, but also provides
guaranteed detection and quantication of steroid hormones in
complex biological samples.
Area (internal standard)/area (analyte) ¼ concentration (internal
standard)/concentration (analyte)
In Fig. 5, several substances are listed which include the
peaks of the urine samples and inter standard solutions. The
ion pairs of the D4-DMBA and the DMBA derivatized steroid
hormone were found in
a
UPLC-MRM run. DMBA-
corticosterone was determined to have an m/z of 494.2230,
whereas D4-DMBA-corticosterone showed an m/z of 498.2139.
The difference in m/z values showed that four hydrogen atoms
had been replaced by four deuterium atoms. The D4-DMBA-
analyte mixed with the DMBA-standard as internal standards,
can be used to provide accurate and convenient quantications
of steroid hormones. The traditional standard curve method
was more suitable for testing of samples in vitro. For substances
in vivo, the standard curve method had a large error because of
different compositions.44,45 Compared with the standard curve
method, the isotope labeling method had more advantages in
reducing the matrix effect and retaining stability.
Conflicts of interest
Application of the method for quantication
The authors declared no competing nancial interest and no
conicts.
In this study, urine samples from six healthy females and males
were subjected to steroid hormone quantication. The levels of
steroid hormones in the urine samples of six rats were also
quantied. GP derived steroid hormones and DMBA derived
steroid hormones were detected in non-hydrolytic urine
samples using UPLC-MRM analysis. The levels of these steroids
were denitively quantied. The quantitative results are shown
in Table 4. As expected, urine samples from males had higher
levels of androgens. For example, TES in male urine samples
was 3.8 times higher than in female samples and androsterone
in male urine samples was 2.4 times higher than in female
samples. More estrogen was found in female sample than in
male samples. For example, the content of E1 in female urine
samples was 4.4 times higher than in male samples, and E3 in
female samples was 3.7 times higher than in male samples. The
results obtained were consistent with those found in the liter-
ature.45 The urine samples from healthy males had higher
hormone levels than the urine samples from male rats. For
example, DHEA from healthy males was 3.2 times higher than
in the samples from male rats. Androsterone from healthy
males was 8.9 times higher than in the samples from male rats.
Some substances like cortisone did not be detected because
activity of 17-hydroxylase is so low in rodents that cortisol and
cortisol can not be synthesized. Also, some progesterone
steroids will be detected only in pregnant females. All these
results are consistent with actual situations.
Acknowledgements
This project was supported by the National Natural Science
Foundation of China (No. 81273998), and the Research
Programs of Application of Basic and Frontier Technology in
Tianjin (15JCYBJC29400). We also thank KareBay Biochem, Inc
for help with the organic synthesis.
Notes and references
1 L. Shi, C. Maser-Gluth and T. Remer, Steroids, 2008, 73,
1446–1451.
2 M. M. Kushnir, A. L. Rockwood, W. L. Roberts, B. Yue,
J. Bergquist and A. W. Meikle, Clin. Biochem., 2011, 44, 77–88.
3 P. Jiang, W. Dai, S. Yan, Z. Chen, R. Xu, J. Ding, L. Xiang,
S. Wang, R. Liu and W. Zhang, J. Ethnopharmacol., 2011,
138, 530–536.
4 J. Nielsen and S. Oliver, Trends Biotechnol., 2005, 23, 544–546.
5 H. N. Bui, E. A. Struys, F. Martens, W. de Ronde,
L. M. Thienpont, P. Kenemans, M. O. Verhoeven,
C. Jakobs, H. M. Dijstelbloem and M. A. Blankenstein, Ann.
Clin. Biochem., 2010, 47, 248–252.
19722 | RSC Adv., 2018, 8, 19713–19723
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
6 T. F. Kalhorn, S. T. Page, W. N. Howald, E. A. Mostaghel and 27 D. Borrey, E. Moerman, A. Cockx, V. Engelrelst and
P. S. Nelson, Rapid Commun. Mass Spectrom., 2007, 21, 3200–
3206.
M. R. Langlois, Clin. Chim. Acta, 2007, 382, 134–137.
28 H. N. Bui, E. A. Struys, F. Martens, R. W. De, L. M. Thienpont,
7 K. Yamashita, Y. Miyashiro, H. Maekubo, M. Okuyama,
S. Honma, M. Takahashi and M. Numazawa, Steroids,
2009, 74, 920–926.
P.
Kenemans,
M.
O.
Verhoeven,
C.
Jakobs,
H. M. Dijstelbloem and M. A. Blankenstein, Ann. Clin.
Biochem., 2010, 47, 248–252.
´
´
´
´
8 M. Szecsi, I. Toth, J. Gardi, M. Vecsernyes, J. Nemeth and 29 D. Tamae, M. Byrns, B. Marck, E. A. Mostaghel, P. S. Nelson,
J. Julesz, J. Biochem. Biophys. Methods, 2006, 69, 51–55.
9 S. Benhaiem, M. Dehnhard, R. Bonanni, H. Hofer,
W. Goymann, K. Eulenberger and M. L. East, Gen. Comp.
Endocrinol., 2012, 178, 265–271.
P. Lange, D. Lin, M. E. Taplin, S. Balk, W. Ellis, L. True,
R. Vessella, B. Montgomery, I. A. Blair and T. M. Penning,
J. Steroid Biochem. Mol. Biol., 2013, 138, 281–289.
30 K. Karu, M. Hornshaw, G. Woffendin, K. Bodin,
¨
10 K. Mitamura, T. Nakagawa, K. Shimada, M. Namiki, E. Koh,
A. Mizokami and S. Honma, J. Chromatogr. A, 2002, 961, 97–
105.
M. Hamberg, G. Alvelius, J. Sjovall, J. Turton, Y. Wang and
W. J. Griffiths, J. Lipid Res., 2007, 48, 976–987.
31 W. J. Griffiths, M. Hornshaw, G. Woffendin, S. F. Baker,
¨
11 J. Y. Moon, K. M. Mcnamara, J. J. Lee, B. C. Chung, H. Sasano
and M. H. Choi, J. Steroid Biochem. Mol. Biol., 2017, 178, 185–
192.
A. Lockhart, S. Heidelberger, M. Gustafsson, J. Sjovall and
Y. Wang, J. Proteome Res., 2008, 7, 3602–3612.
32 M. Ogundare, S. Theolopoulos, A. Lockhart, L. J. Hall,
¨
12 K. Fang, X. Pan, B. Huang, J. Liu, Y. Wang and J. Gao,
Chromatographia, 2010, 72, 949–956.
E. Arenas, J. Sjovall, A. G. Brenton, Y. Wang and
W. J. Griffiths, J. Biol. Chem., 2010, 285, 4666–4679.
13 J. Nielsen and S. Oliver, Trends Biotechnol., 2005, 23, 544– 33 W. J. Griffiths, Y. Wang, K. Karu, E. Samuel, S. Mcdonnell,
546.
M. Hornshaw and C. Shackleton, Clin. Chem., 2008, 54,
1317–1324.
34 A. E. Debarber, Y. Sandlers, A. S. Pappu, L. S. Merkens,
P. B. Duell, S. R. Lear, S. K. Erickson and R. D. Steiner, J.
Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2011, 879,
1384–1392.
´
´
´
´
14 M. Szecsi, I. Toth, J. Gardi, M. Vecsernyes, J. Nemeth and
J. Julesz, J. Biochem. Biophys. Methods, 2006, 69, 51–55.
15 L. H. Qian, N. G. Li, Y. P. Tang, L. Zhang, H. Tang, Z. J. Wang,
L. Liu, S. L. Song, J. M. Guo and A. W. Ding, Int. J. Mol. Sci.,
2011, 12, 8208–8216.
16 N. Guo, C. Y. Peng, Q. F. Zhu, B. F. Yuan and Y. Q. Feng, Anal. 35 T. Higashi, A. Yamauchi, K. Shimada, E. Koh, A. Mizokami
Chim. Acta, 2017, 967, 42–51. and M. Namiki, Anal. Bioanal. Chem., 2005, 382, 1035–1043.
17 D. L. Di, H. Benabdelkamel, D. Taverna, S. Indelicato, 36 Y. Shibayama, T. Higashi, K. Shimada, A. Odani,
D. Aiello, A. Napoli, G. Sindona and F. Mazzotti, Anal.
Bioanal. Chem., 2015, 407, 5835–5842.
A. Mizokami, H. Konaka, E. Koh and M. Namiki, J.
Chromatogr. B, 2009, 877, 2615–2623.
18 K. Yamashita, T. Kawahata, M. Takahashi and 37 T. Higashi, A. Nagahama, N. Otomi and K. Shimada, J.
M. Numazawa, J. Mass Spectrom. Soc. Jpn., 2009, 57, 75–80.
19 K. Yamashita, K. Yamazaki, S. Komatsu and M. Numazawa,
J. Am. Soc. Mass Spectrom., 2010, 21, 249–253.
20 Z. O'Brien, N. Post, M. Brown, A. Madan, T. Coon, R. Luo and
T. A. Kohout, J. Chromatogr. B, 2009, 877, 3515–3521.
21 H. H. Adomat, O. S. Bains, J. M. Lubieniecka, M. E. Gleave,
E. S. Guns, T. A. Grigliatti, R. E. Reid and K. W. Riggs, J.
Chromatogr. B, 2012, 902, 84–95.
Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2007, 848,
188–199.
38 T. Higashi, T. Nishio, S. Uchida, K. Shimada, M. Fukushi and
M. Maeda, J. Pharm. Biomed. Anal., 2008, 48, 177–182.
39 Y. Shibayama, T. Higashi, K. Shimada, K. Kashimada,
T. Onishi, M. Ono, K. Miyai and S. Mizutani, J. Chromatogr.
B, 2008, 867, 49.
40 Y. Weng, F. Xie, L. Xu, D. Zagorevski, D. C. Spink and
X. Ding, Anal. Biochem., 2010, 402, 121–128.
41 M. Starweinstock, B. L. Williamson, S. Dey, S. Pillai and
S. Purkayastha, Anal. Chem., 2012, 84, 9310.
22 T. Higashi, H. Yokoi, Y. Nagura, T. Nishio and K. Shimada,
Biomed. Chromatogr., 2008, 22, 1434–1441.
23 S. Boenzi, F. Deodato, R. Taurisano, D. Martinelli,
¨
¨
D. Verrigni, R. Carrozzo, E. Bertini, A. Pastore, 42 S. Salihovic, A. Karrman, G. Lindstrom, P. M. Lind, L. Lind
C. Dionisivici and D. W. Johnson, Clin. Chim. Acta, 2014,
437, 93–100.
24 P. Keski-Rahkonen, K. Huhtinen, M. Poutanen and
and B. B. Van, J. Chromatogr. A, 2013, 1305, 164–170.
43 S. Tulipani, R. Llorach, M. Urpi-Sarda and C. Andres-
Lacueva, Anal. Chem., 2013, 85, 341–348.
S. Auriola, J. Steroid Biochem. Mol. Biol., 2011, 127, 396–404. 44 C. Wang, C. Wu, Z. Lin and J. Zhang, Anal. Chem., 2016, 88,
25 M. M. Kushnir, A. L. Rockwood, W. L. Roberts, E. G. Pattison, 7878.
A. M. Bunker, R. L. Fitzgerald and A. W. Meikle, Clin. Chem., 45 W. Dai, Q. Huang, P. Yin, J. Li, J. Zhou, H. Kong, C. Zhao,
2006, 52, 120–128. X. Lu and G. Xu, Anal. Chem., 2012, 84, 10245–10251.
26 T. F. Kalhorn, S. T. Page, W. N. Howald, E. A. Mostaghel and 46 C. Wang, H. Kong, Y. Guan, J. Yang, J. Gu, S. Yang, G. Xu and
P. S. Nelson, Rapid Commun. Mass Spectrom., 2007, 21, 3200–
A. Chem, Anal. Chem., 2005, 77, 4108–4116.
3206.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 19713–19723 | 19723