Direct quantification of steroid glucuronides in human urine by liquid chromatography–electrospray tandem mass spectrometry
Introduction
Anabolic steroids are the most commonly detected substances prohibited in human sports [1]. Most of these steroids are not naturally present in the human body and therefore, their presence in urine is considered as evidence of drug abuse. Hence, qualitative screening methods are developed for these compounds. The detection of endogenous steroid abuse such as testosterone (T) is more complicated. The determination of concentration ratios for endogenous steroids is normally used for this purpose. The abuse of T is normally controlled by the ratio between urinary concentrations of T and epitestosterone (E) (T/E ratio). Other ratios including concentrations of androsterone and etiocholanolone can also be useful to provide evidence of T, dihydrotestosterne (DHT), dehydroepiandrosterone (DHEA) and alfa reductase inhibitor abuse [2]. In order to establish these ratios accurately, quantitative methods for these analytes are needed.
Endogenous steroids are normally excreted in urine after conjugation in phase II metabolism reactions, of which the most common one is glucuronidation. The determination of these endogenous steroids is normally performed by GC–MS after enzymatic hydrolysis, liquid–liquid extraction (LLE) and derivatization [3], [4], [5]. Besides being time-consuming, this approach can sometimes generate problems associated with the quantitative recoveries in all steps and can potentially make the result obtained for a sample vary between laboratories. One of the most efficient approaches to minimise these problems is the use of labelled internal standards.
Although some methods describe the direct detection of steroid conjugates by GC–MS after derivatization [6], liquid chromatography–tandem mass spectrometry (LC–MS/MS) is reported as a more suitable alternative either in negative ionization mode due to the acidity of the glucuronide moiety [7], [8], [9] or in positive ionization owing to the proton affinity of the steroid keto function [10], [11], [12], [13]. However, in all reported methods a preconcentration step is needed to reach the required sensitivity (low ng/ml). Different preconcentration steps have been described for this purpose such as LLE [10], liquid–solid extraction [11], solid-phase extraction (SPE) [7], [8], [9], [10], [11] or liquid-phase microextraction [10] using between 1 and 5 ml of urine.
The use of a preconcentration step can affect the quantitative characteristics of an LC–MS/MS method because of endogenous interferences [14] which can increase the ion suppression. This fact was observed after different sample pre-treatment methods. Hence, due to the high amount of concentrated interferences LLE was not considered as a suitable approach to validate a quantitative method [10]. SPE is frequently suggested as alternative for LLE. Unfortunately, SPE can also result in signal suppression close to 100% [12]. Additionally, other problems related with the presence of isobaric endogenous interferences in the analysis of steroids have been reported [11], [15] and these can be intensified by the use of preconcentration steps.
The aim of this study is to develop a quantitative method for the direct determination of steroid glucuronides based on LC–MS/MS allowing for a reduction in sample volume, total analysis time and sample manipulation. Two sample pre-treatment approaches are tested in order to meet this aim: filtration and LLE. Finally, the applicability of the developed methods is tested by comparing its results with those obtained by the GC–MS method routinely applied in our laboratory.
Section snippets
Chemicals and reagents
Structures of the selected analytes are shown in Fig. 1. Glucuronides of testosterone (TG), epitestosterone (EPG), androsterone (AG), etiocholanolone (ETG), 19-norandrosterone (NAG), 19-noretiocholanolone (NETG), [2H3]testosterone (TG-d3), [2H3]epitestosterone (EPG-d3), [2H4]androsterone (AG-d4) and [2H4]19-norandrosterone (NAG-d4) were purchased from NMI (Pymble, Australia).
Hydrochloric acid and formic acid were purchased from Merck (Darmstadt, Germany) and Fischer Scientific (Loughborough,
MS/MS optimization
All selected glucuronides exhibited ionization in both positive and negative modes. According to Kuuranne et al. [16], TG and ETG were predominantly detected as [M+H]+ in positive ionization mode due to the high proton affinity of these steroids. On the other hand, AG, ETG, NAG and NETG were mainly detected as [M+NH4]+ and [M+Na]+ as a result of their lower proton affinity. These results are also in agreement with the ionization observed for unconjugated steroids in electrospray ionization [17]
Conclusions
Two quantitative methods based on LC–MS/MS for the determination of some anabolic steroid glucuronides were developed. The product ions resulting from losses of the glucuronide moiety and water are the most abundant ones, the selection of highly specific transitions allowed for an increased sensitivity of the method after optimizing the MS/MS conditions. This was found to be critical for TG and EPG where a large number of endogenous interferences were found. The sensitivity achieved by the
Acknowledgements
Postdoctoral grants by the Spanish Ministerio de Educacion y Ciencia (O.J.P.) and the Flemish Ministry of Culture, Youth, Sports and Brussels (P.V.E., W.V.T., K.D.) are gratefully acknowledged. WADA is thanked for partially funding this work.
References (20)
- et al.
Best Pract. Res. Clin. Endocrinol. Metab.
(2000) - et al.
J. Chromatogr. B
(1996) - et al.
Steroids
(2000) - et al.
Anal. Chim. Acta
(2004) - et al.
J. Chromatogr. B
(2006) - et al.
J. Chromatogr. B.
(2001) - et al.
J. Chromatogr. B
(1997) - et al.
J. Chromatogr. B
(1996) - et al.
J. Am. Soc. Mass Spectrom.
(2000) - et al.
Clin. Chem.
(1997)