Elsevier

Steroids

Volume 70, Issue 8, July 2005, Pages 499-506
Steroids

Formation of 19-norsteroids by in situ demethylation of endogenous steroids in stored urine samples

https://doi.org/10.1016/j.steroids.2005.01.005Get rights and content

Abstract

The formation of 19-norsteroids by demethylation of endogenous steroids in stored urine samples was observed. Suspicious urine samples (i.e. containing trace amounts of 19-norandrosterone and 19-noretiocholanolone) were selected and spiked with deuterated analogues of androsterone and etiocholanolone at concentrations corresponding to high endogenous levels (4 μg/mL). After incubation, respective 19-norsteroids (19-norandrosterone-d4 and 19-noretiocholanolone-d5) were identified in these samples by high-resolution mass spectrometry. The transformation of the 5β-isomer (etiocholanolone) yields about three-fold higher concentrations, compared to the 5α-isomer. A significant temperature dependence was observed by comparison of reaction kinetics at room temperature (23 ± 2 °C) and 37 °C. Concentrations of 19-norandrosterone-d4 and 19-noretiocholanolone-d5, respectively, were 2.7 and 3.6 times higher at elevated temperature. The conversion of androsterone-d4 to 19-norandrosterone-d4 did not exceed a relative amount of 0.1%. Incubation of the urine samples with androsterone-d4-glucuronide led to the production of 19-norandrosterone-d4-glucuronoide. A partial stabilization was observed after addition of metabolic inhibitors (e.g. EDTA). The application of the incubation experiments described may contribute to the clarification of adverse analytical findings regarding low levels of 19-norsteroid metabolites.

Introduction

Oxidation and subsequent demethylation of the angular 19-methyl group is an essential step in the biosynthesis of estrogens. Several endogenous steroids like testosterone, androstenedione or 16α-hydroxydehydroepiandrosterone are known to undergo this biochemical reaction. Therefore, the observation of intermediate 19-norsteroids in specimens of various species (e.g. urine of stallion [1], boar, pregnant cow [2], goat, sheep and mare [3]) is consistent. Moreover, there is clear evidence for the occurrence of 19-norandrosterone (NA) in humans, e.g. during pregnancy or the ovulation period of the menstrual cycle [4]. Findings of low concentrations in men were reported [5], but often also attributed to the contamination of food [6], [7] or nutritional supplements [8]. The investigation of a correlation between low concentrations of NA and physical exercise led to contradictory results [9], [10]. Finally, norethisterone based contraceptives interfere with the analyses due to the formation of NA as minor metabolite [11], which can be easily recognised by identification of 5β-tetrahydronorethisterone at an appropriate level and does not constitute a problem. Any administration of 19-norsteroids in humans leads predominantly to formation of glucuronides of the 5α-metabolite NA and the 5β-metabolite noretiocholanolone (NE), regardless the steroid administered (mostly nortestosterone (NT, Fig. 1), norandrostenedione or norandrostene-3β,17β-diol) or the administration route (oral, transdermal, buccal or intramuscular injection of esters). A threshold value of 2 ng/mL NA (glucuronide, calculated as the unconjugated form [12]) in urine was implemented in athletes drug testing to avoid adverse analytical findings due to low endogenous levels and are subject of controversial discussions ever since.

Additional criteria to ascertain analytical NA findings, e.g. a reference to endogenous steroid levels, are common practice, but not officially adopted. These criteria include the calculation of NA concentrations relative to androsterone (A), already proposed in the early 1990s by Donike; cut-off ratios (androsterone + etiocholanolone (E))/NA of 1000 (male) or 500 (female) were proposed later [13]. The investigation of biotransformation kinetic yielded other qualitative indications in excretion studies. The ratio of 5α- and 5β-metabolites (NA/NE) showed a significant correlation to the structure of the administered 19-norsteroid (esp. position of the double bound) [14]. Similarly, the relative amounts of sulphates versus glucuronides were reported to be influenced by the origin of the steroid [15]. However, doping controls represent isolated metabolic snapshots, influenced by the unknown variables dosage, time and route of application, accurate structure of the administered steroid derivative(s) as well as other parameters (e.g. ethnical influences on metabolism). It is not likely, that this complex matter can be sufficiently elucidated by investigation of an individual sample.

The increase in 19-norsteroid related doping cases in the late 1990s paralleled the appearance of prohormones on the market (e.g. DHEA, androstenedione, 19-norsteroids) and the contamination of nutritional supplements in particular. However, due to the prevalence of latent sources of NA, it is hardly possible to trace an individual analytical finding back to the relevant origin.

Contrary to suggestions that technical improvement of the steroid screening might have contributed to the increase in 19-norsteroid doping tests, the technique (high-resolution mass spectrometry, HRMS) routinely used in our laboratories, remained mainly unchanged during the past 10 years. Long-term observations of the few cases with low concentrations of NA and NE in a total number of roughly 30,000 doping control samples revealed a possible correlation with the amounts of the endogenous steroids A and E.

The aim of the work was to study an eventual conversion of A and E to NA and NE, respectively, by incubating urine samples with the deuterated steroids A-d4 and E-d5 and with A-d4-glucuronide. The identification of deuterated NA and NE could prove the relevance of this transformation in selected, stored urine samples.

NA-glucuronide is described to be stable under various storage conditions [16], and a screening result can usually be confirmed in new aliquots of A- or B-sample regardless their different storage conditions. However, a few experiences in our laboratories, where re-analysis of A- and B-samples revealed differences in NA and NE concentrations greater than expected, support the need for such an investigation.

In a former case of a female athlete, the result of an elevated NA concentration (6.4 ng/mL, R.S.D. 5.6%) could not be verified in the confirmation analysis. The B-sample (frozen at −20 °C when arriving the laboratory) contained 4.2 ng/mL (R.S.D. 3.8%) of NA while the concentration of the original sample, which was routinely re-analysed during the confirmation, remained similar (7.1 ng/mL). Although the overall uncertainty was only partially represented by the relative standard deviation, the latter was suitable for the direct comparison of both concentrations, which were examined in the same batch where other contributions (day-to-day or inter-laboratory uncertainty) are negligible. Moreover the amounts of NE varied significantly (3.3 ng/mL versus 7.1 ng/mL).

Section snippets

Urine samples

The selection of suitable samples was based on the routine doping analysis carried out in the doping control laboratories Kreischa and Oslo. Urine samples with concentrations of NA or NE higher than 0.5 ng/mL (concentration corrected to a specific gravity (s.g.) of 1.020 by the formula conc.1.020=(1.0201)/(s.g.1)×conc.measured) but lower than the doping threshold1 were collected during a time

Formation of NA-d4 and NE-d5 by decomposition of A-d4 and E-d5

The formation of corresponding 19-norsteroids after incubation (12 h at 37 °C) of urine samples with deuterated A and E was observed in four of the five suspicious samples (Table 3), while sample S5, the only one with a NA/NE ratio greater than one, was negative. Additionally, NA-d4 and NE-d5 were identified in one of the samples with low, but presumably exogenous amounts of NA and NE (X1).

Sample S3 (Fig. 2, Fig. 3) was used for further quantification and kinetic studies (Sections 3.3 Kinetic of

Conclusion

The formation of NA-d4 and NE-d5 could be confirmed by unequivocal identification of the metabolites after incubation of suspicious urine samples with A-d4 and E-d5.

This transformation reaction occurs in free steroids as well as in corresponding glucuronides, the latter yielding predominantly glucuronides of the 19-norsteroids.

The reaction is significantly dependent on temperature and substrate concentration, the conversion rate of the 5β-isomer (E-d5) is about three times higher compared to

Acknowledgements

The support of this project by the World Anti Doping Agency (WADA) and Federal Institute of Sports Science (BISp, Germany) is gratefully acknowledged. We also acknowledge the performance of the microbiology analysis by Prof. Signe Holta Ringertz, Aker University Hospital, Oslo, Norway.

References (16)

There are more references available in the full text version of this article.

Cited by (0)

View full text