Parallel derivatization strategy coupled with liquid chromatography-mass spectrometry for broad coverage of steroid hormones

Journal Pre-proof

Parallel derivatization strategy coupled with liquid chromatography-mass spectrometry for broad coverage of steroid hormones Qian Qin Writing – original draft; Methodology; Investigation , Disheng Feng Writing – review & editing , Chunxiu Hu Writing – review & editing , Bohong Wang , Mengmeng Chang , Xinyu Liu , Peiyuan Yin , Dr. Xianzhe Shi Writing – review & editing , Prof. Guowang Xu Conceptualization; Supervision PII: DOI: Reference:

S0021-9673(19)31141-0 https://doi.org/10.1016/j.chroma.2019.460709 CHROMA 460709

To appear in:

Journal of Chromatography A

Received date: Revised date: Accepted date:

29 May 2019 11 November 2019 12 November 2019

Please cite this article as: Qian Qin Writing – original draft; Methodology; Investigation , Disheng Feng Writing – review & editing , Chunxiu Hu Writing – review & editing , Bohong Wang , Mengmeng Chang , Xinyu Liu , Peiyuan Yin , Dr. Xianzhe Shi Writing – review & editing , Prof. Guowang Xu Conceptualization; Supervision , Parallel derivatization strategy coupled with liquid chromatography-mass spectrometry for broad coverage of steroid hormones, Journal of Chromatography A (2019), doi: https://doi.org/10.1016/j.chroma.2019.460709

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V.

Highlights  A parallel derivatization strategy for steroid hormones was newly proposed.  The approach has a broad coverage of steroid hormones.  The sensitivity of the method was greatly improved by three orders of magnitude.

1

Parallel derivatization strategy coupled with liquid chromatography-mass spectrometry for broad coverage of steroid hormones

Qian Qina,b,#, Disheng Fenga,b,#, Chunxiu Hua, Bohong Wanga,b, Mengmeng Changa,b, Xinyu Liua, Peiyuan Yina, Xianzhe Shia,*, Guowang Xua,*

a.

CAS Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China

b. University of Chinese Academy of Sciences, Beijing 100049, China

# These authors equally contribute to this work.

Correspondence: Dr. Xianzhe Shi, e-mail: [email protected], Tel: 0086-411-84379757, Fax: 0086-411-84379559 Prof. Guowang Xu, e-mail: [email protected], Tel: 0086-411-84379530, Fax: 0086-411-84379559

2

Abstract Steroid hormones are a type of crucial substances that mediate numerous vital physiological functions. The comprehensive detection of steroid hormones can help understand the physiopathologic mechanism of steroid hormone-related diseases. It is very difficult to determine steroid hormones in biological samples due to their low endogenous concentrations and poor ionization efficiency. In this study, an efficient and sensitive approach was developed for profiling steroid hormones by combining liquid-liquid

extraction

and

parallel

derivatization

with

liquid

chromatography-tandem mass spectrometry. Methoxyamine and dansyl chloride were used to derivatize steroid hormones containing carbonyl and phenolic hydroxyl groups, respectively. Our established method achieved simultaneous analysis of carbonyl and phenolic hydroxyl-containing steroid hormones and could cover estrogens, androgens, corticoids and progestogens. Twenty-nine steroid hormones were detected at pg/mL levels with the sensitivity enhanced by three orders of magnitude after derivatization. The linearity (with linear range of 2–4 orders of magnitude), precision (less than 15%) and recovery (71.1-128.7%) were satisfactory for quantitative analysis of steroid hormones. Finally, the established method was successfully employed to the determination of steroid hormones in serum samples of healthy males and females as well as ovarian cancer patients. The results showed that this approach was suitable and reliable for routine test of steroid hormones containing carbonyl and phenolic hydroxyl groups. Key words: simultaneous quantification, parallel derivatization, steroid hormones, liquid chromatography-mass spectrometry

3

1. Introduction Steroid hormones are of exceptional importance to human body, and they act as signaling molecules in various physiological processes including the regulation of transcription, the maintenance of secondary sexual characteristics and the mediation of the immune and endocrine systems. Recently, multiple studies have demonstrated that the metabolism disorder of steroid hormones in the human body is closely related to numerous diseases such as cardiovascular disease, endocrine disorders [1], prostate cancer [2], breast cancer [3] and endometrial cancer [4]. A highly sensitive, stable and reliable analytical method enabling to achieve quantitative profiling of steroid hormones is of great significance for studying the pathogenesis of these diseases. Currently, the detection of single or several steroid hormones has been commonly used in clinical auxiliary diagnosis [5, 6]. However, the concentrations of steroid hormones were individually different and susceptible to many factors such as physiological cycle. Such research strategy based on the detection of few steroid hormones may lead to mis-judgement in the aspect of disease diagnosis and pathological typing. Quantitative metabolic profiling of more steroid hormones enables to not only provide a better understanding of the whole steroid hormone metabolism network, but also produce a more accurate and reliable diagnostic result with reduced false positive or false negative rate. However, most steroid hormones in the human body are at relatively low concentrations, making it great difficulties and challenges for researchers to study the metabolic network of steroid hormones. The detection methods of endogenous steroid hormones mainly included immunoassay, GC-MS and LC-MS [7, 8]. Immunoassays could achieve highly sensitive detection of single hormone, but the specificity, reproducibility and accuracy of this method were insufficient due to severe matrix interference and high batch-to-batch variation of antibodies [9]. LC-MS method can detect multiple steroid hormones with high sensitivity and specificity, which is an ideal technique for high-throughput analysis of the steroid hormones [9-11]. As the physicochemical properties of different steroid hormones are largely different, a variety of ionization techniques (e.g., ESI, APCI, APPI, etc.) and a combination of positive and negative 4

ion modes is often required to achieve satisfactory coverage and sensitivity. Chemical derivatization is a common sample pretreatment technique that modifies the functional groups on the analytes via the reactions with derivative reagents to augment the UV, fluorescence or mass spectrometric response of the analytes. In LC-MS analysis, the derivative reagent can improve the chromatographic retention of target compounds by altering the active groups of target compounds (e.g., amine groups, carboxyl groups, hydroxyl groups, thiol groups, etc.). Moreover, the ionization efficiency of the modified analytes can be improved, which would effectively increase their detection sensitivity. In addition, the ionization polarities of the modified analytes can be maintained consistent to avoid the switch between various ionization modes of MS [12]. Hydroxylamine [13], Girard's Reagent P (GP) reagent [14, 15] and 2-hydrazinopyridine [16, 17] are used for the derivatization of steroid hormones containing carbonyl groups, while dansyl chloride [18, 19], 1,2-dimethylimidazolium-4-sulfonyl chloride [20], Sanger's reagent and their analogues [21] are used for the derivatization of estrogens containing phenolic hydroxyl groups. However, only a single derivative reagent can’t react with all the steroid hormones. Therefore, parallel derivatization aiming at two different functional groups can expand the scope of the steroid hormones. In this study, we established a highly sensitive method based on liquid-liquid extraction and parallel derivatization by methoxyamine and dansyl chloride combined with LC-MRM-MS to achieve simultaneous analysis of carbonyl and phenolic hydroxyl-containing steroid hormones in a single run. To prove its practicability, the established method was used to the analysis of serum steroid hormones from healthy persons and ovarian cancer patients.

2. Experimental section 2.1. Chemicals and Reagents Pregnenolone corticosterone

(Preg), (B),

progesterone

(Prog),

deoxycorticosterone

17α-hydroxypregnenolone

(DOC),

(17OH-Preg),

17α-hydroxyprogesterone (17-OHP), cortexolone (S) , hydrocortisone (HC), 5

dehydroepiandrosterone (DHEA), estrone (E1), estriol (E3), β-estradiol (E2), testosterone (T), 5α-tetrahydrocorticosterone (5α-DHT), 5β-dihydrotestosterone (5β-DHT), mesterolone (Mest), androsterone (Andro), tetrahydrocortexolone (THS), 5α-tetrahydrocorticosterone (5α-THB), epiandrosterone (EA), etiocholanolone (Etio), 11-oxoetiocholanolone

(O-Etio),

6β-hydroxytestosterone

(6β-HOT),

16α-hydroxytestosterone (16α-HOT), epitestosterone (ET), allotetrahydrocortisol (Allo-THC) and dehydroepiandrosterone sulfate (DHEAS) were obtained from International

Laboratory

USA

(South

San

Francisco,

Calif.,

USA).

4-Androstene-3,17-dione (AN) was obtained from Dr. Ehrenstorfer GmbH (Augsburg, Germany). Aldosterone (Aldo) and 5-dimethylamino-1-naphthalenesulfonyl chloride were acquired from J&K Scientific Ltd. (Beijing, China). Internal standards testosterone-d3 (T-d3) and β-estradiol-d2 (E2-d2), methoxyamine hydrochloride (MOA), 2-hydrazinopyridine (2-HP), butylamine and formic acid were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade organic solvents (acetonitrile, methanol and dichloromethane) were purchased from Merck (Darmstadt, Germany). Sodium carbonate (Na2CO3) and sodium bicarbonate (NaHCO3) were of analytical purity. The ultrapure water was purified by Milli-Q system (Millipore Corporation, Bedford, MA, USA).

2.2. Stock solutions preparation Twenty-nine hormone standards were separately dissolved in acetonitrile or methanol/water solution. The concentrations of Preg, Prog, DOC, B, 17-OHP, S, HC, DHEA, E1, E2, T, 5α-DHT, 5β-DHT, Mest, Andro, THS, 5α-THB, EA, Etio, O-Etio, 6β-HOT, 16α-HOT, ET, Allo-THC and DHEAS solutions were 1.0 mg/mL. The concentrations of 17OH-Preg and E3 were 0.2 mg/mL. As the internal standard stock solutions, T-d3 was dissolved in acetonitrile with a concentration of 100 μg/mL, and E2-d2 was dissolved in methanol with a concentration of 4.7 mg/mL. The concentrations of MOA and 5-dimethylamino-1-naphthalenesulfonyl chloride were 50 mM and 5 mg/mL, respectively. The internal standard work solutions containing 25 ng/mL T-d3 and 25 ng/mL E2-d2 in dichloromethane were used for the calibration 6

study. The internal standard work solutions containing 1 ng/mL T-d3 and 1 ng/mL E2-d2 in dichloromethane were used for the recovery experiment. All the solutions were stored at -20°C before use.

2.3. Sample collection and preparation The healthy male and female serum samples were collected from Dalian Physical Examination Center, and the serum samples of 5 ovarian cancer patients were collected from Dalian Maternal and Child Health Care Hospital. Serum samples from healthy males and females were mixed to form a mixed blood sample, which was used as the research object for condition optimization and method validation in this study. The serum was thawed at 4°C, and 100 μL of serum was taken in a 1.5 mL centrifuge tube. Each sample was taken in parallel and marked as a and b, respectively. One was added with 1 mL of dichloromethane containing T-d3, and the other was added with 1 mL of dichloromethane containing E2-d2. Both mixtures were vortexed on a shaker for 5 min and then centrifuged (10 min, 14000 rpm, 4°C). Finally, 900 μL of the lower organic layer was transferred to a new 1.5 mL centrifuge tube and lyophilized. In order to maximize the extraction efficiency of steroid hormones from serum, the liquid-liquid extraction effectiveness of dichloromethane and ethyl acetate were compared. Two aliquots of 100 μL serum from the same sample were taken, 1 mL of dichloromethane and 1 mL of ethyl acetate were separately added, both samples were vortexed for 5 min, and then centrifuged at 10°C, 14000 rpm/min for 10 min. Finally, 900 μL of the lower organic layer was transferred to a 1.5 mL centrifuge tube and lyophilized. Each experiment was performed in triplicate.

2.4. Derivatization process For the purpose of optimizing the conditions of derivatization, 50 μL of mixed solution containing 1 μg/mL of Preg, Prog, DOC, B, 17OH-Preg, 17-OHP, S, HC, DHEA, E1, E3, E2, T, 5α-DHT, Andro and Aldo was added with 50 μL of 50 mM 7

MOA or 2-HP methanol solution, the mixture was then vortexed for 30 s. The derivatization reaction solution was incubated in a water bath at 60°C for 30 min, then the tube was placed on ice for 1 min. After centrifugation for 10 min (14000 rpm, 4°C), the supernatant was collected for LC-MS injection analysis. MOA was used as a derivative reagent for the steroid hormones containing carbonyl groups and dansyl chloride for steroid hormones containing phenolic hydroxyl groups. The mechanism of derivatization was shown in Fig. S1. The optimized derivatization process was as follows. Sample a was added with 90 μL of 50 mM MOA methanol solution. After 1 min vortex, the reaction was carried out in a water bath at 60°C for 30 min. Sample b was added with 30 μL of 5 mg/mL of dansyl chloride acetonitrile solution and 30 μL of 0.15 M Na2CO3/NaHCO3 aqueous solution. After 1 min vortex, the tube was placed in a water bath at 60°C for 30 min. Subsequently, 30 μL of 0.5 mol/L aqueous solution of butylamine was added and the tube was put in a water bath at 60°C for another 40 min. After being taken out, the tube was placed on ice for 1 min. Each sample was centrifuged at 14000 rpm/min for 10 min at 4°C. Finally, 60 μL of the supernatant was collected from both a and b samples, and was mixed for LC-MS analysis. A schematic diagram of the hormone analysis process of serum samples was shown in Fig. 1.

2.5. LC-MS analysis To investigate the sensitivity improvement effect of MOA and dansyl chloride, Acquity ultra performance liquid chromatography (UPLC, Waters, USA) coupled with an LTQ Orbitrap mass spectrometer (Thermo Fisher, USA) system equipped with an electrospray ionization (ESI) source was used for LC-MS analysis. Liquid chromatography was carried out on a BEH C18 column (10 cm×2.1 mm, 1.7 μm, Waters, USA). The column temperature was 40°C. The injection volume was 10 μL. The flow rate was 0.25 mL/min. The mobile phases were 0.1% formic acid (v/v) in H2O (phase A) and 0.1% formic acid (v/v) in ACN (phase B). The linear gradient conditions were as follows: 0 min, 30% B; 5 min, 98% B; 8 min, 98 % B; 8.1 min, 30% B; 10 min, 30% B. The total chromatographic run time was 10 min. The mass 8

spectrometer operating conditions were set as follows: spray voltage in positive ion mode, 4.5 kV; capillary voltage, 48 V; capillary temperature, 325°C; sheath gas flow rate, 45 units; auxiliary gas flow rate, 10 units; lens voltage, 80 V. The mass spectrometer was scanned within the m/z of 80-800 in positive ion mode, and the resolution was set to 30,000. Methodology and real sample analysis were performed using Acquity ultra performance liquid chromatography (UPLC, Waters, USA) coupled with tandem quadrupole-ion trap mass spectrometer (Q-Trap 5500 MS, AB SCIEX, USA) equipped with an ESI source. Liquid chromatography operating conditions were the same as the descriptions above. Mass spectrometer conditions were set as follows: gas curtain gas, Gas1 and Gas2 (0.241 and 0.276 MPa), spray voltage (5.5 kV), ion source temperature (50°C). The mass spectrometer was run in positive ion multiple reaction monitoring (MRM) mode, and the MRM parameters were optimized. The stable ion pair was used as a quantitative ion pair to analyze the sample. The MRM parameters of each hormone were shown in Table S1.

3. Results and Discussion 3.1. Optimization of LLE solvents and derivative reagents In order to extract as many steroid hormones as possible from serum samples, liquid-liquid extraction (LLE) solvents were optimized. Ethyl acetate and dichloromethane were often used as LLE solvents [22]. In this study, we compared the extraction efficiency of the two kinds of solvents. For most steroid hormones, the extraction efficiency of dichloromethane is slightly higher than that of ethyl acetate. The results were displayed in Fig. S2. Meanwhile, dichloromethane was evaporated more quickly than ethyl acetate, which could save the whole pretreatment time. Thus, dichloromethane was chosen as the extraction solvent. It has been reported that steroid hormones containing hydroxyl group can be derivatized by 4-dimethylamino-benzoic acid (DMBA) [23, 24]. However, the derivatization reaction was carried out with the requirement of catalysts, relatively high temperature and long reaction time. Most steroid hormones containing carbonyl 9

group can be derivatized using hydroxylamine or 2-hydrazinylpyridine (2-HP) under a relatively mild reaction condition. However, this method left out a small number of steroid hormones such as estrogen, which only contain phenolic hydroxyl groups. Dansyl chloride can be used to derivatize these phenolic hydroxyl-containing steroid hormones. In order to expand the coverage of steroid hormones, a parallel derivatization method directed against two different functional groups of steroid hormones was developed. Firstly, thirteen carbonyl steroid hormones and three phenolic hydroxyl steroid hormones were selected for the investigation of different derivative reagents. As exhibited in Fig. 2, the MS response of MOA derivative is higher than that of 2-HP except for three hormones containing phenolic hydroxyl (i.e., estrone, estradiol and estradiol), which only react with dansyl chloride. Theoretically, the alkaline system of the mixed derivatives is not in favor of the ionization of 2-HP derivatives, but it has little influence on MOA derivatives. Therefore, MOA and dansyl chloride were selected to derivatize carbonyl steroid hormones and phenolic hydroxyl steroid hormones, respectively.

3.2. Parallel derivatization method for the coverage of steroid hormones In the following study, 29 steroid hormones were selected to establish and characterize the parallel derivatization method. The 29 steroid hormones covered the most of the important hormones including androgens, estrogens, progestogens and corticosteroids. The metabolic pathway of steroid hormones modified from KEGG pathway is shown in Fig. S3, which represents the core pathway of steroid hormones. Most steroid hormones contain both carbonyl and phenolic hydroxyl groups except for androstenediol. After derivatization, 27 steroid hormones containing carbonyl reacted with MOA, and 3 steroid hormones containing phenolic hydroxyl reacted with dansyl chloride. In addition, estrone (E1) containing both carbonyl and phenolic hydroxyl could react with MOA or dansyl chloride, while the derivatives of dansyl chloride had higher sensitivity. As shown in Fig. S4, the 29 steroid hormones eluted within 2-7 min after derivatization. The results indicated that the parallel derivatization method could cover most of the steroid hormones in the pathway and 10

could be further used to extend the library of steroid standards. Among the 29 steroid standards, there are two carbonyl groups in Prog, DOC, B, HC and AN, etc., and three carbonyl groups in aldosterone. As shown in Fig. S5, the response intensity of the derivatives of single carbonyl group was higher than that of the derivatives of two or three carbonyl groups. It is partly related to the high steric resistance of these hormones containing multiple carbonyl groups, which was not conducive to the derivatization. Therefore, the derivatives of single carbonyl group were selected for quantitative analysis. In order to realize the high throughput analysis, the two derivatives from MOA and dansyl chloride of the same steroid hormones were simultaneously separated after mixed into one sample. However, it is necessary to consider the possibility of cross-reaction of residual derivative reagents. After the derivatization of phenolic hydroxyl steroid hormones by dansyl chloride, butylamine was added to react with the residual excessive amount of dansyl chloride. Thus, the phenolic hydroxyl hormones from another sample would not have a derivatization reaction when mixed into one sample. In addition, as the mixed solution of the two derivatives was an alkaline system, it was further examined whether the carbonyl steroid hormones could react with MOA. As shown in Fig. S6, compared with the neutral condition for derivatization reaction, the carbonyl steroid hormones (e.g., 17OH-Preg, HC, DHEA and AN) didn’t almost react with MOA under alkaline condition even after 55 h at 4C, indicating that the mixed injection of the two derivatives could not cause cross-reaction and would not affect the subsequent quantitative analysis. Besides, the evaluation of derivatization reaction efficiency was performed by comparing the amounts of steroid hormones before and after derivatization reaction. As exhibited in Fig. S7, no underivatized steroid hormones (e.g., Aldo, HC, Mest and E3) were detected. The results showed that the derivatization reactions of steroid hormones with excessive amount of MOA and dansyl chloride were complete. Moreover, the sensitivity of most steroid hormones was enhanced by three orders of magnitude after derivatization, which was displayed in Table S2. 11

3.3. Method validation To evaluate the reliability of the developed method, the linearity, limit of detection (LOD), limit of quantitation (LOQ), intraday/interday precision, reproducibility, stability and recovery were further investigated. As exhibited in Table 1, good linearities over wide concentration ranges (2–4 orders of magnitude) and high corresponding regression coefficients (from 0.981 to 0.999) were acquired. Except for aldosterone, the LOD values were 0.5-5 pg/mL, the LOQ values were 1.5-15 pg/mL for most of steroid hormones indicating the high detection sensitivity of the method. Spiked serum samples at three concentration levels (0.1 ng/mL except of Aldo with 1 ng/mL, 5 ng/mL and 25 ng/mL) were extracted and derivatized to investigate the recovery of the method. The relative recovery (Rec) was calculated by the following formula: Rec= (A1 –A2) / A3

(1)

where A1, A2 and A3 referred to the relative peak area of steroid hormones in spiked serum samples, normal serum samples and standard samples, respectively. As a result, the recoveries of the derivatives of 29 steroid hormones at low, medium and high concentration levels was 72.1-125.7%, 71.1-126.7% and 79.6-128.7%, respectively. For the investigation of repeatability, reproducibility and stability, serum samples spiked with a mixed standard solution was extracted and derivatized. For the evaluation of repeatability, six replicate spiked serum samples were carried out at the same time. Six spiked serum samples were prepared and analyzed on three different days for the evaluation of reproducibility. As shown in Table S2, the repeatability and reproducibility of the steroid hormones were within 15% and 20%. A spiked serum sample placed in the autosampler at 4C was analyzed after 0, 6, 12, 18, 24, 30 and 48 h, respectively, to test the stability of the derivatives of steroid hormones. It displayed that the RSDs of 29 derivatives of steroid hormones were less than 17%, showing that the derivatized serum samples were stable at 4C within two days.

3.4. Determination of steroid hormones in serum samples from healthy subjects 12

and ovarian cancer patients The established parallel derivatization and LC-MRM-MS method was applied to analyze the steroid hormones in serum samples of healthy male, female and ovarian cancer patients. It enabled to detect androgens, estrogen, progesterone and corticosteroids, showing a good feasibility of the method. The MRM chromatograms of steroid hormones detected from healthy female serum samples were shown in Fig. 3. Univariate analysis showed that Prog, HC, E1, T, 5α-THB, 16α-HOT and allo-THC had significant difference (p < 0.05) between healthy male and female serum samples, indicating that hormone sterol metabolism is closely related to the gender. Prog, S, HC, AN, E2, E3, T, AN, 5α-THB, Etio, ET, allo-THC and DHEAS in serum samples had a significant difference (p < 0.05) between healthy female and ovarian cancer patients, indicating that ovarian cancer involved significant changes of steroid hormones in peripheral circulation system. Five steroid hormones included Aldo, 5α-DHT, 5β-DHT, Mest and 6β-HOT weren’t be quantitated due to their very low concentrations in serum samples. The information of serum steroid hormone levels in 5 healthy male, 5 female and 5 ovarian cancer patients was shown in Table 2. It can be known that most steroid hormones were at pg level while few steroid hormones (e.g., 17-hydroxyprogesterone and DHEA) were at ng level, which was consistent with the reports in the literatures [22, 25, 26].

4. Conclusions In this study, a novel method based on parallel derivatization was developed to simultaneously analyze steroid hormones containing carbonyl or phenolic hydroxyl group covering four categories of androgens, estrogen, progesterone and corticosteroids. MOA and dansyl chloride were used to derivatize steroid hormones containing carbonyl group and phenolic hydroxyl group, respectively. The sample pretreatment process of the method was simple and rapid, and the derivatization reaction conditions were mild. In addition, the analytical characteristics of the method were satisfactory for simultaneous analysis of hormone steroids with good linearity, precision, repeatability, stability and recovery. The method showed good practicability 13

in studying steroid hormones in relation to ovarian cancer. Significant changes were found in a variety of serum steroid hormones in ovarian cancer patients. Forthcoming studies will continue to expand the detection coverage of steroid hormone metabolism network and apply it to the research of other hormone-related diseases.

Author Contributions Section Name Qian Qin Disheng Feng Chunxiu Hu Bohong Wang Mengmeng Chang Xinyu Liu Peiyuan Yin Xianzhe Shi Guowang Xu

Contributions Writing - original draft; Methodology; Investigation Writing - review & editing Writing - review & editing Data curation Data curation Resources Funding acquisition Writing - review & editing Conceptualization; Supervision

Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments We gratefully acknowledge the financial support of the projects from National Key Research

and

Development

Program

of

China

(No.2017YFC0906900,

2016YFC1303100) and National Natural Science Foundation of China (No. 21575142, 21874130, 21435006).

References [1] J. Nielsen, S. Oliver, The next wave in metabolome analysis, Trends Biotechnol., 23 (2005) 544-546. 14

[2] P.H. Gann, C.H. Hennekens, J. Ma, C. Longcope, M.J. Stampfer, Prospective study of sex hormone levels and risk of prostate cancer, J. Natl. Cancer Inst., 88 (1996) 1118-1126. [3] I.A. Blair, Analysis of estrogens in serum and plasma from postmenopausal women: Past present, and future, Steroids, 75 (2010) 297-306. [4] A. Lukanova, E. Lundin, A. Micheli, A. Arslan, P. Ferrari, S. Rinaldi, V. Krogh, P. Lenner, R.E. Shore, C. Biessy, P. Muti, E. Riboli, K.L. Koenig, M. Levitz, P. Stattin, F. Berrino, G. Hallmans, R. Kaaks, P. Toniolo, A. Zeleniuch-Jacquotte, Circulating levels of sex steroid hormones and risk of endometrial cancer in postmenopausal women, Int. J. Cancer, 108 (2004) 425-432. [5] K. Yamashita, Y. Miyashiro, H. Maekubo, M. Okuyama, S. Honma, M. Takahashi, M. Numazawa, Development of highly sensitive quantification method for testosterone and dihydrotestosterone in human serum and prostate tissue by liquid chromatography-electrospray ionization tandem mass spectrometry, Steroids, 74 (2009) 920-926. [6] N. Janzen, S. Sander, M. Terhardt, U. Steuerwald, M. Peter, A.M. Das, J. Sander, Rapid steroid hormone quantification for congenital adrenal hyperplasia (CAH) in dried blood spots using UPLC liquid chromatography-tandem mass spectrometry, Steroids, 76 (2011) 1437-1442. [7] M. Szecsi, I. Toth, J. Gardi, M. Vecsernyes, J. Nemeth, J. Julesz, HPLC-RIA analysis of the ectopic cortisol production in a cancerous pancreas tumor, J. Biochem. Bioph. Methods, 69 (2006) 51-55. [8] K. Mitamura, T. Nakagawa, K. Shimada, M. Namiki, E. Koh, A. Mizokami, S. Honma, Identification of dehydroepiandrosterone metabolites formed from human prostate homogenate using liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry, J. Chromatogr. A, 961 (2002) 97-105. [9] S.J. Soldin, O.P. Soldin, Steroid Hormone Analysis by Tandem Mass Spectrometry, Clin. Chem., 55 (2009) 1061-1066. [10] M.M. Kushnir, A.L. Rockwood, W.L. Roberts, B. Yue, J. Bergquist, A.W. Meikle, Liquid chromatography tandem mass spectrometry for analysis of steroids in clinical 15

laboratories, Clin. Biochem., 44 (2011) 77-88. [11] T. Higashi, Y. Shibayama, I. Kawasaki, K. Shimada, Determination of salivary 17-ketosteroid sulfates using liquid chromatography-electrospray ionization-mass spectrometry, J. Pharm. Biomed. Anal., 43 (2007) 1782-1788. [12] F. Xu, L. Zou, Y. Liu, Z. Zhang, C.N. Ong, Enhancement of the capabilities of liquid chromatography –mass spectrometry with derivatization: general principles and applications, Mass Spectrom. Rev., 30 (2011) 1143-1172. [13] P. Keski-Rahkonen, K. Huhtinen, M. Poutanen, S. Auriola, Fast and sensitive liquid chromatography-mass spectrometry assay for seven androgenic and progestagenic steroids in human serum, J. Steroid Biochem. Mol. Biol., 127 (2011) 396-404. [14] A.E. DeBarber, W.E. Connor, A.S. Pappu, L.S. Merkens, R.D. Steiner, ESI-MS/MS quantification of 7 alpha-hydroxy-4-cholesten-3-one facilitates rapid, convenient diagnostic testing for cerebrotendinous xanthomatosis, Clin. Chim. Acta, 411 (2010) 43-48. [15] A.E. DeBarber, Y. Sandlers, A.S. Pappu, L.S. Merkens, P.B. Duell, S.R. Lear, S.K. Erickson, R.D. Steiner, Profiling sterols in cerebrotendinous xanthomatosis: utility of Girard derivatization and high resolution exact mass LC-ESI-MS(n) analysis, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 879 (2011) 1384-1392. [16] T. Higashi, T. Nishio, N. Hayashi, K. Shimada, Alternative procedure for charged derivatization to enhance detection responses of steroids in electrospray ionization-MS, Chem. Pharm. Bull., 55 (2007) 662-665. [17] N. Nadarajah, Ø. Skadberg, J. Adaway, C. Brede, Multiplexed analysis of steroid hormones in saliva by LC-MS/MS with 2-hydrazinopyridine derivatization, Clinical Mass Spectrometry, 4-5 (2017) 1-10. [18] X. Xu, J.M. Roman, H.J. Issaq, L.K. Keefer, T.D. Veenstra, R.G. Zieger, Quantitative measurement of endogenous estrogens and estrogen metabolites in human serum by liquid chromatography-tandem mass spectrometry, Anal. Chem., 79 (2007) 7813-7821. [19] X. Xu, T.D. Veenstra, S.D. Fox, J.M. Roman, H.J. Issaq, R. Falk, J.E. Saavedra, 16

L.K. Keefer, R.G. Ziegler, Measuring fifteen endogenous estrogens simultaneously in human urine by high-performance liquid chromatography-mass spectrometry, Anal. Chem., 77 (2005) 6646-6654. [20] L. Xu, D.C. Spink, 1,2-Dimethylimidazole-4-sulfonyl chloride, a novel derivatization reagent for the analysis of phenolic compounds by liquid chromatography

electrospray

tandem

mass

spectrometry:

Application

to

1-hydroxypyrene in human urine, Journal of Chromatography B-Analytical Technologies in the Biomedical and Life Sciences, 855 (2007) 159-165. [21] T. Nishio, T. Higashi, A. Funaishi, J. Tanaka, K. Shimada, Development and application of electrospray-active derivatization reagents for hydroxysteroids, J. Pharm. Biomed. Anal., 44 (2007) 786-795. [22] D.R. Taylor, L. Ghataore, L. Couchman, R.P. Vincent, B. Whitelaw, D. Lewis, S. Diaz-Cano, G. Galata, K.M. Schulte, S. Aylwin, N.F. Taylor, A 13-Steroid Serum Panel Based on LC-MS/MS: Use in Detection of Adrenocortical Carcinoma, Clin. Chem., 63 (2017) 1836-1846. [23] W. Dai, Q. Huang, P. Yin, J. Li, J. Zhou, H. Kong, C. Zhao, X. Lu, G. Xu, Comprehensive and Highly Sensitive Urinary Steroid Hormone Profiling Method Based on Stable Isotope-Labeling Liquid Chromatography Mass Spectrometry, Anal. Chem., 84 (2012) 10245-10251. [24] C. Liu, X. Sheng, Y. Wang, J. Yin, W. Huang, Y. Fan, Y. Li, Y. Zhang, 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, Rsc Advances, 8 (2018) 19713-19723. [25] C. Wang, C. Wu,

L.

Zhang, J. Zhang,

Ultraperformance

Liquid

Chromatography-Tandem Mass Spectrometry Method for Profiling Ketolic and Phenolic Sex Steroids Using an Automated Injection Program Combined with Diverter Valve Switch and Step Analysis, Anal. Chem., 88 (2016) 7878-7884. [26] T.F. Yuan, J. Le, Y. Cui, R. Peng, S.T. Wang, Y. Li, An LC-MS/MS analysis for seven sex hormones in serum, J. Pharm. Biomed. Anal., 162 (2018) 34-40. 17

Table 1 Linear range, R2, LOD, LOQ and recovery of 29 steroid hormones derivatives Linear range Steroid hormones

(ng/mL)

R2

LOD

LOQ

Recovery

pg/mL

pg/mL

low

medium

high

Preg

0.02-200

0.982

1.0

3.0

105.5

106.1

106.9

Prog

0.02-500

0.999

1.0

3.0

104.9

71.1

79.6

DOC

0.02-500

0.994

1.0

3.0

87.3

78.9

81.4

B

0.02-500

0.999

2.0

6.0

103.8

100.8

102.7

Aldo

1-500

0.999

200.0

600.0

72.9

79.7

83.7

17OH-Preg

0.02-1000

0.998

5.0

15.0

92.9

117.4

100.9

17OHP

0.02-500

0.997

1.0

3.0

77.6

80.4

84.7

S

0.02-500

0.993

5.0

15.0

91.9

102.3

99.5

HC

0.01-500

0.999

5.0

15.0

84.7

92.0

87.2

DHEA

0.01-500

0.995

1.0

3.0

125.7

103.6

112.2

AN

0.01-250

0.999

1.0

3.0

73.9

91.2

98.6

E1

0.02-200

0.981

0.5

1.5

106.4

108.1

111.6

E3

0.02-200

0.984

0.5

1.5

72.1

41.6

60.1

E2

0.02-200

0.987

0.5

1.5

103.0

89.0

99.4

T

0.02-200

0.987

1.0

3.0

87.1

101.4

100.6

5α-DHT

0.02-1000

0.997

5.0

15.0

82.7

102.5

101.9

5β-DHT

0.05-500

0.996

5.0

15.0

99.6

111.8

101.1

Mest

0.05-500

0.992

2.0

6.0

82.2

98.1

104.2

Andro

0.05-200

0.998

2.0

6.0

84.2

110.0

105.5

THS

0.05-200

0.989

2.0

6.0

76.7

1.4

13.7

5α-THB

0.05-500

0.990

2.0

6.0

107.3

126.7

103.8

EA

0.02-500

0.988

2.0

6.0

86.9

115.0

108.9

Etio

0.02-200

0.988

2.0

6.0

89.9

110.3

105.2

O-Etio

0.02-200

0.986

2.0

6.0

97.9

122.0

128.7

6β-HOT

0.02-200

0.996

2.0

6.0

102.5

123.3

110.2

16α-HOT

0.02-200

0.983

1.0

3.0

122.3

101.6

97.2

ET

0.02-200

0.987

1.0

3.0

123.4

100.8

100.8

Allo-THC

0.02-200

0.983

2.0

6.0

106.4

89.5

100.5

DHEAS

0.02-500

0.989

5.0

15.0

99.2

78.1

33.1

18

Table 2 Comparison of steroid hormones in serum from healthy male (M), healthy female (F) and ovarian cancer (OC) patients Steroid hormones Preg Prog DOC B 17OH-Preg 17OHP S HC DHEA AN E1 E3 E2 T Andro THS 5α-THB EA Etio O-Etio 16α-HOT ET Allo-THC DHEAS

This study (mean±SD, ng/mL) Male 0.116±0.049 N.D 0.047±0.028 N.Q 2.764±0.621 0.065±0.036 N.Q 28.116±8.644 2.291±0.714 0.242±0.016 0.069±0.004 0.042±0.011 0.042±0.006 1.310±0.495 0.168±0.063 1.531±0.784 0.203±0.063 0.098±0.047 0.055±0.016 0.116±0.060 0.026±0.008 N.Q 9.636±1.526 23.837±25.675

Female 0.097±0.036 0.086±0.087 0.024±0.022 N.Q 2.041±1.219 0.037±0.031 N.Q 14.602±2.571 2.512±1.036 0.291±0.119 0.054±0.012 0.034±0.008 0.071±0.039 0.082±0.027 0.114±0.04 1.080±0.288 0.104±0.024 0.059±0.013 0.049±0.022 0.100±0.009 0.015±0.007 0.005±0.002 4.596±1.286 9.294±2.636

Ovarian Cancer 0.073±0.062 N.D 0.014±0.006 0.062±0.016 3.869±5.077 0.018±0.009 0.059±0.035 27.433±8.005 1.429±1.355 0.077±0.034 0.050±0.008 N.D N.D 0.036±0.008 0.038±0.001 0.905±0.310 0.205±0.083 0.027±0.015 0.023±0.011 0.088±0.026 0.018±0.008 0.002±0.001 9.158±3.299 2.881±1.935

Ref.[25,26] (ng/mL) Male a

0.09-0.1 0.03-0.04a 1.0-2.0a 0.3-0.4a 2.0-3.0a 0.15-0.2a 0.01-0.03b 0.08-0.10a 0.01-0.03b 2.0-3.0a -

P-value

Female

M/F

OC/F

a

0.489 S 0.178 N.A 0.283 0.224 N.A 0.022 0.705 0.412 0.042 0.203 0.162 0.005 0.150 0.281 0.021 0.139 0.623 0.588 0.036 N.A 0.001 0.275

0.488 S 0.409 S 0.473 0.258 S 0.020 0.196 0.013 0.542 S S 0.016 0.011 0.381 0.050 0.006 0.058 0.362 0.517 0.046 0.033 0.003

0.1-0.2 0.4-0.5a 1.0-1.5a 0.3-0.4a 2.0-3.0a 0.2-0.3a 0.01-0.03b 0.1-0.15a 0.01-0.03b 0.1-0.2a -

Note：The concentrations were presented as “mean value±standard deviation”. N.D refers to Not Detected. N.Q refers to Not Quantified. N.A refers to Not Acquired. S refers to significant. a represents Ref. 25, b represents Ref. 26.

19

Figure Captions Fig. 1 Schematic diagram of steroid hormones analysis process of serum samples (DC: dansyl chloride)

Fig. 2 MS intensity of steroid hormones derivatized by different derivative reagents

Fig. 3 MRM chromatograms of 27 steroid hormones derivatized by methoxamine and dansyl chloride in serum from healthy female

20