Mass Spectrometry Based Determination of Unconjugated, Sulfated and Glucuronidated Steroid Hormones and Their Metabolites in Biological Fluids

Abstract

Here we present a newly developed and fully validated LC-MS/MS method for the targeted analysis of 20 unconjugated and sulfated steroids in 65 amniotic fluid (AF) samples during mid-gestation. The method requires 600 µL of AF. Sample preparation included protein precipitation, centrifugation, solid phase extraction and derivatization. The method is capable of measuring progesterone, 17α-hydroxyprogesterone, testosterone, estrone, estradiol, estriol, estrone sulfate, 17β-estradiol 3-sulfate, estriol 3-sulfate (E3S), 17β-estradiol 17-sulfate, 16α-hydroxydehydroepiandrosterone sulfate (16OH-DHEAS), dehydroepiandrosterone sulfate (DHEAS), pregnenolone sulfate, 17α-hydroxypregnenolone sulfate, testosterone sulfate, epitestosterone sulfate, dihydrotestosterone sulfate, androsterone sulfate, epiandrosterone sulfate and 3,17β-androstenediol 3-sulfate. The performance of the method was evaluated at three quality control levels for each compound ranging from low limit of quantification to high limit of quantification. The method showed good accuracy and precision. Linearity for each analyte was above 0.99. The limit of quantification ranged between 0.2 and 1.7 ng/mL for the 20 compounds. Except DHEAS, all other sulfated steroids were quantified by LC-MS/MS in AF for the first time. We have produced MS based reference values for 20 steroids in AF of mid-gestation. DHEAS and E3S were both strongly correlated with 16OH-DHEAS, thus confirming the classic concept of the feto-placental unit. Our LC-MS/MS method can be used for future prenatal diagnosis of steroid metabolism disorders. To complement our study of amniotic fluid by LC-MS/MS, gas chromatography-mass spectrometry was employed to characterize the steroidal context of amniotic fluid (n=65; male: female = 35: 30) of mid-gestation (median: 18.8th week, range: 16.0th – 24.6th week) by a comprehensive targeted steroid hormone metabolomics approach. The levels of 52 steroids including pregnenolone and 17-OH-pregnenolone metabolites, dehydroepiandrosterone and its metabolites, progesterone and 17-OH-progesterone metabolites, sex hormones as well as corticosterone and cortisol metabolites were measured. The dominating steroids were the group of pregnenolone and 17-OH-pregnenolone metabolites (mean ± SD: 138.0 ± 59.3 ng/mL), followed by the group of progesterone and 17-OH-progesterone metabolites (107.3 ± 44.3 ng/mL), and thereafter DHEA and its metabolites (97.1 ± 56.5 ng/mL). With respect to sex steroids, only testosterone showed a significantly higher value in male fetuses (p<0.0001) reflecting testicular endocrine activity. Of all estrogen metabolites, estriol showed by far the highest concentrations (33.2 ± 26.1 ng/mL). Interestingly, cortisol metabolites were clearly present (59.6 ± 13.6 ng/mL) though fetal de novo synthesis of cortisol is assumed to start from the 28th gestational week onwards. Our comprehensive characterization of the steroidal milieu in amniotic fluid of mid-gestation shows the presence of all relevant classes of steroid hormones. The steroidal milieu in amniotic fluid mirrors the steroidome of the feto-placental unit. Our set of basic data lays the foundation for further studies characterizing various diseases affecting steroid metabolism. We have developed and validated a LC-MS/MS method for the simultaneous quantification of 15 urinary steroid glucuronides in human urine: androsterone glucuronide, etiocholanolone glucuronide, epiandrosterone glucuronide, dihydrotestosterone glucuronide, dehydroepiandrosterone glucuronide, testosterone glucuronide, epitestosterone glucuronide, estrone glucuronide, 17β-estradiol 17-glucuronide, 17β-estradiol 3-glucuronide, estriol 16-glucuronide, pregnenolone glucuronide, Tetrahydro-11-deoxycorticosterone 3glucuronide, cortisol 21-glucuronide and pregnanediol glucuronide. The method was then successfully applied to 67 urine samples from 5-day-old infants to 17.8-year-old children and adolescents. Free and sulfated steroids were also measured by LC-MS/MS. In addition, GC-MS measurements provided a total concentration of free and conjugated steroids after enzymatic hydrolysis. The total levels of androsterone and etiocholanolone went up to 5820.0 nmol/L and 4017.8 nmol/L through childhood, respectively. They are largely excreted as glucuronides (4374.3 nmol/L and 3588.5 nmol/L, respectively). DHEA was excreted mostly as sulfate in all age groups. Cortisol was present predominantly as sulfate (173.8 nmol/L) in newborns. The levels of sulfated cortisol deceased with age while its glucuronidated form increased. The levels of free cortisol were relatively constant in the urine of children after 1 year of age. Sex hormones were mainly excreted as glucuronide. This study for the first time showed the distribution profile of steroids in free, sulfated and glucuronidated forms.