High Throughput Identification and Quantification of Anabolic Steroid

Technical Note pubs.acs.org/ac

High Throughput Identification and Quantification of Anabolic Steroid Esters by Atmospheric Solids Analysis Probe Mass Spectrometry for Efficient Screening of Drug Preparations Mickael Doué, Gaud Dervilly-Pinel,* Audrey Gicquiau, Karinne Pouponneau, Fabrice Monteau, and Bruno Le Bizec LUNAM Université, Oniris, Laboratoire d’Etude des Résidus et Contaminants dans les Aliments (LABERCA), Atlanpole-La Chantrerie, CS 50707, Nantes, F-44307, France S Supporting Information *

ABSTRACT: Recent developments in ambient mass spectrometry (AMS), such as atmospheric solids analysis probe (ASAP) mass spectrometry, open a whole new range of possibilities to screen for drug preparations. In this study, the potential of ASAP for the rapid identification and quantification of anabolic steroid esters has been evaluated. These compounds are known to be used both in human and in food producing animals to enhance performances and to improve the rate of growth, respectively. Using a triple quadrupole (QqQ) MS instrument, mechanism of ionization and fragmentation in both positive and negative mode were studied for a range of 21 selected steroid esters (based on testosterone, estradiol, nandrolone, and boldenone) which highlighted common neutral mass loss of 96.1, thus allowing rapid screening in minutes to reveal steroid ester presence with minimal sample preparation. Ester identification is further achieved through an efficient 2 min workflow on a QqQ MS instrument. Moreover, the use of isotope labeled internal standards permitted the quantification of the corresponding steroid esters in selected reaction monitoring (SRM) mode, for the first time in ASAP. This approach was successfully applied for characterization of oily commercial preparations. These results open new perspectives in hormone (and drug) rapid analysis by ASAP-MS in the near future.

A

oily solutions or in tablets as some of the most counterfeited drugs8 whereas antidoping and food safety laboratories examine hair or blood samples for steroid ester contents as an unequivocal proof of misuse.9−14 Forensic laboratories use mainly thin layer chromatography and/or liquid or gas chromatography coupled to mass spectrometry for the analysis of anabolic steroid preparations.15,16 When dealing with biological samples, the main challenge is to detect the xenobiotics of interest present at a trace level in the considered matrix. Therefore, extended and time-consuming purification procedures have been developed, involving liquid−liquid extraction, solid phase extraction, and sometimes a derivatization step to enhance sensitivity.10 The detection step of the steroid esters is mainly performed by liquid chromatography coupled to mass spectrometry (MSn or high resolution mass spectrometry (HRMS))10−14 leading to efficient management of steroid esters issues. Nevertheless, the need for faster multiresidue approaches with both increased sensitivity and specificity is crucial today: the quicker the analysis, the more efficient is the control. In particular, methods allowing rapid characterization of seized preparations are required.

nabolic steroids have been used for decades in sport competitions to enhance athletic performances.1,2 The use of these substances increases lean body mass, strength, and aggressiveness while it decreases recovery time between workouts.3 The use of anabolic steroids in sports has been prohibited by the World Anti-Doping Agency (WADA).4 In addition, the misuse of these substances is no longer exclusive to athletes since teenagers and noncompeting amateurs also use anabolic steroids, changing this issue from being a problem restricted to sport competitions to one of public-health concern.2 In meat producing animals also, such practices may be encountered since the administration of anabolic steroids promotes growth through the increase of lean tissue and body mass.5,6 For food safety concerns, the European Union has prohibited their uses (Directive 96/22/EC).7 Anabolic steroids are usually administered intramuscularly (IM) under their ester forms (synthetically produced) in lipophilic excipient providing longer biological availability and lasting effects. Depending on the length of the side chain, steroid esters show different release times. Thus, mixtures of long- and short-side chain esters are usually administered in order to reach constant doses of active molecules on a wide time scale. Over the last decades, several analytical strategies have been investigated to prevent anabolic steroid (esters) misuse. Forensic laboratories search for anabolic steroids (esters) in © 2014 American Chemical Society

Received: March 25, 2014 Accepted: May 5, 2014 Published: May 5, 2014 5649

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

Technical Note

and negative modes were studied for a large set (n = 21) of steroid esters derived from testosterone, estradiol, boldenone, or nandrolone. These 21 steroid esters have been selected as representative of the most common on Web sites specialized in the anabolic steroids black market. A screening approach based on a simple 3 min workflow was proposed for the rapid specific identification of the anabolic steroid esters. The original use of labeled internal standards enabled accurate quantification, which further strengthens the strategy. The applicability to oily preparations characterization was finally probed.

Beginning in the 21st century, the introduction of the techniques known as desorption electrospray ionization (DESI) by Cooks et al.17 and as direct analysis in real time (DART) by Cody et al.18 opened the field of ambient mass spectrometry (AMS). Since then, almost 20 different techniques of AMS have been described in the literature,19 showing the high degree of scientific interest for these original ionization/introduction modes. AMS allows direct analysis of samples in their native conditions with little or no sample preparation. A multistep process combining desorption of the sample and ionization of the analytes takes place in the ambient air environment, before analysis by MS, resulting in a much simplified workflow as the chromatographic part is removed. DESI-MS has attracted a lot of attention in forensic science in the last ten years due to its ability to analyze specimens in different forms (e.g., tablets, gels, powders, and so on) and the following advantages: versatility, high throughput, and minimal sample preparation. In this context, the potential of DESI to screen for steroid esters in veterinary drug preparations,20 bovine hair,21 and tissue samples22 has been studied. Authors described the detection of high levels (above 300 μg kg−1) of a few steroid esters (mainly testosterone esters) in all matrices. Although promising, several issues were reported by the authors: the risk of cross-contamination with powder samples20 or the variations of the ion current due to spot heterogeneity and the sputtering character of DESI.21 A recent review23 also pointed out that both repeatability and reproducibility of DESI-MS are poor and therefore cannot currently replace common and routinely used techniques. While other ambient ionization approaches such as DART have been tested for their potential in characterizing chemical residues,24 only desorption atmospheric pressure photoionization (DAPPI) was, besides DESI, reported for the analysis of steroid esters in oil and tablets: the analyses were performed with ion trap MS and DAPPI proven to be rapid (compared to classical gas chromatography (GC)/ MS) and a specific analysis technique, which does not require any sample preparation.25,26 Nevertheless, repeatability and reproducibility of the strategy still need improvement before further implementation as an efficient screening tool. Atmospheric solids analysis probe (ASAP) was first introduced in 2005 by McEwen and co-workers27 and is derived from atmospheric pressure chemical ionization (APCI). Vaporization of samples occurs under the hot nitrogen stream from the electrospray or APCI desolvation gas, and then, the corona discharge needle produces a plasma which ionizes the analytes.27 The overall mechanism results in the production of protonated molecules in the presence of water vapors or radical ions in dry conditions.28 ASAP can be used in conjugation with tandem mass spectrometry (MS/MS) for fragmentation pathways studies or with high resolution mass spectrometry (HRMS) for accurate mass measurements and fingerprinting. One of the main advantages of ASAP is that no vacuum lock is necessary as the sample (solid or liquid) is introduced at atmospheric pressure and can be analyzed in seconds.29 ASAP ionizes both polar and nonpolar small molecules30 and was applied, in the past few years, to the analysis of drugs,30 natural products,31,32 nucleoside,33 polymers,34 coal-related model compounds,35 and steroids.28 Fussell et al. have recently published an assessment of ASAP for the detection of chemicals in food with a special emphasis on pesticides in cereals.29 In this context, the rapid identification and quantification of anabolic steroid esters by ASAP have been assessed. Mechanisms of ionization and fragmentation in both positive

■

EXPERIMENTAL SECTION Chemicals and Materials. The reference steroid esters, including testosterone acetate (T Ac), testosterone benzoate (T Bz), testosterone cypionate (T Cy), testosterone decanoate (T Dc), testosterone heptanoate (T Hp), testosterone propionate (T Pr), boldenone acetate (B Ac), boldenone benzoate (B Bz), boldenone propionate (B Pr), boldenone undecylenate (B Un), nandrolone benzoate (N Bz), nandrolone decanoate (N Dc), nandrolone laurate (N Lr), nandrolone propionate (N Pr), nandrolone phenylpropionate (N PhPr), estradiol 17-cypionate (E2 17Cy), estradiol 17-enanthate (E2 17En), estradiol 17propionate (E2 17Pr), estradiol 17-valerate (E2 17Vl), estradiol 17-decanoate (E2 17Dc), and estradiol 3-benzoate (E2 3Bz), were obtained from either Steraloids Inc. Ltd. (London, England) or Sigma-Aldrich (St Louis, USA). Testosterone acetate-d3 (T Ac-d3), testosterone benzoate-d3 (T Bz-d3), testosterone decanoate-d3 (T Dc-d3), and testosterone heptanoate-d3 (T Hp-d3) were used as labeled internal standards (IS) and were kindly provided by RIKILT (Wageningen, The Netherlands). All reference standards have been characterized by GC/MS and GC/MS/MS before use in order to check their purity. Ethanol and methanol were of analytical grade quality and purchased from Carlo Erba Reagents (Rodano, Italy). Sample Preparation. Each steroid ester stock solution was prepared at 1 mg mL−1 by dilution in an appropriate volume of ethanol and was stored at −20 °C. The working standard solutions were prepared by diluting stock solutions in methanol. Four oily solutions, commercially available on the Internet, were analyzed: BOLD-UN 200 (B Un at 200 mg mL−1), TESTO-PR 100 (T Pr at 100 mg mL−1), Laurabolin (N Lr at 25 mg mL−1), and Sustanon (T Pr at 30 mg mL−1, T PhPr at 60 mg mL−1, T Isocaproate (T Iso) at 60 mg mL−1, and T Dc at 100 mg mL−1). Before analysis, oily solutions were diluted 1000-fold in methanol. For quantification purposes, a mixture of the 4 internal standards (T Ac-d3, T Bz-d3, T Dc-d3, and T Hp-d3) at 1 mg L−1 each (final concentration) was added before dilution. All samples were analyzed using a glass melting point capillary dipped in the corresponding solution; the sample was directly introduced into the source. Prior to use, the glass capillary was rinsed with a mixture of water/ethanol (50:50, v/v) and baked in the ion source at 600 °C for at least 30 s to clean the surface of the capillary from any residual contamination. ASAP-MS Instrumentation. The steroid esters were analyzed using a Waters ASAP Probe either on a XEVO TQS (triple quadrupole, QqQ) instrument or on a SYNAPT G2-S (Q-TOF) instrument (Waters, Milford, MA, USA). The instruments were operated with the same optimized source conditions in both positive and negative APCI mode. Common setting parameters shared by both modes were as follows: 5650

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

Technical Note

Table 1. Specific Ions and Common Ions Observed for the 21 Selected Steroid Esters in Scan and Product Ion Scan of the Corresponding [M + H]+ or [M − H]− in Both Positive and Negative Mode (APCI+/APCI−) scan and product ion scan of [M − H]−

scan and product ion scan of [M + H]+ monoisotopic mass (g mol−1)

specific ions APCI+ (m/z)

steroid base

compound

testosterone

T Ac

330.2

331.2 [M + H]+

T Pr

344.2

345.2 [M + H]+

T Bz

392.2

T Hp

400.3

T Cy

412.3

T Dc

442.3

E2 17Pr

328.2

393.3 [M + 105.0 401.3 [M + 85.0 413.3 [M + and 79.0 443.4 [M + 105.0 329.2 [M +

E2 17Vl

356.2

357.2 [M + H]+

E2 3Bz

376.2

377.2 [M + H]+; 359.3; 105.0 and 77.0

E2 17Hp

384.3

E2 17Cy

396.3

E2 17Dc

426.3

385.3 [M + H]+ and 85.0 397.3 [M + H]+ and 79.0 427.3 [M + H]+; 409.3 and 273.2 [E2 + H]+

B Ac

328.2

329.2 [M + H]+

B Pr

342.2

343.2 [M + H]+

B Bz

390.2

B Un

452.3

391.2 [M + H]+ and 105.0 453.3 [M + H]+

N Pr

330.2

N Bz

378.2

N PhPr

406.3

N Dc

428.3

N Lr

456.4

estradiol

boldenone

nandrolone

common ions APCI+ (m/z)

specific ions APCI− (m/z)

289.2 [T + H]+; 271.2 [T + H − H2O]+; 253.2 [T + H − 2H2O]+; 175.1; 109.0 and 97.0

329.2 [M − H]− and 59.0 [Ac]− 343.2 [M − H]− and 73.0 [Pr]− 391.2 [M − H]−; 121.0 [Bz]− and 77.0 399.2 [M − H]− and 129.1 [Hp]− 411.2 [M − H]− and 141.1 [Cy]− 441.2 [M − H]− and 171.1 [Dc]− 327.2 [M − H]− and 73.0 [Pr]− 355.2 [M − H]− and 101.0 [Vl]− 375.2 [M − H]−; 271.2 [E2 − H]− and 121.0 [Bz]− 383.2 [M − H]− and 129.0 [Hp]− 395.2 [M − H]− and 141.0 [Cy]− 425.3 [M − H]−; 271.2 [E2 − H]− and 171.1 [Dc]− 327.2 [M − H]−; 311.2 and 59.0 [Ac]− 341.2 [M − H]−; 325.2 and 73.0 [Pr]− 389.2 [M − H]− and 121.0 [Bz]− 451.3 [M − H]−; 435.3 and 183.1 [Un]− 329.2 [M − H]− and 73.0 [Pr]− 377.2 [M − H]−; 121.0 [Bz]− and 77.0 405.3 [M − H]− and 149.0 [PhPr]− 427.3 [M − H]− and 171.1 [Dc]− 455.4 [M − H]− and 199.1 [Lr]−

H]+ and H]+ and H]+; 107.0 H]+ and H]+

331.2 [M + H]+ and 95.0 379.2 [M + H]+ and 105.0 407.3 [M + H]+ and 105.0 429.3 [M + H]+; 155.1; 95.0 and 81.0 457.4 [M + H]+ and 95.0

255.2 [E2 + H − H2O]+; 159.1; 135.0 and 107.0

287.2 [B + H]+; 269.2 [B + H − H2O]+; 173.1; 147.1; 135.0 and 121.1

275.2 [N + H]+; 257.2 [N + H − H2O]+; 239.2 [N + H − 2H2O]+; 145.1 and 109.0

source housing temperature was set at 150 °C, desolvation gas flow and cone flow at 500 and 150 L/h, respectively, and nebulizer at 4 bar. For fragmentation pathways study and identification of unknown samples, the APCI probe temperature was set at 450 °C whereas for quantification purposes a linear gradient of temperature was applied from 100 to 600 °C in 15 s and held at 600 °C thereafter. MS current and voltage were optimized in both modes in order to improve detection sensitivity. In positive mode, corona current was set to 3 μA, cone voltage to 45 V, and source offset to 70 V, while in negative mode corona current was set to 5 μA, cone voltage to 35 V, and source offset to 80 V. For scan, precursor ion scan, product ion scan, or neutral loss scan with QqQ MS, data were acquired in “Multi Channel Analysis” which consists in accumulating profile scans over a given period (i.e., 50 s). As signal data accumulates more rapidly than random noise, this mode allows the signal data to be emphasized and the signal-tonoise ratio to be improved. In selected reaction monitoring (SRM) experiments, collision energy (CE) was optimized for

common ions APCI− (m/z) 269.2 [T − H − H2O]−

253.2 [E2 − H − H2O]−

267.2 [B − H − H2O]−

no common ion

each compound and argon was used as collision gas at 0.15 mL min−1. With the Q-TOF instrument, data were acquired in scan mode (in the range of 50 to 500 m/z) in high resolution mode.

■

RESULTS AND DISCUSSION Determination of Desorption Temperatures. The first optimized parameter was the APCI probe temperature since a complete vaporization of compounds without thermal degradation is the main objective to reach higher signal intensity. For the 21 steroid esters and the 4 labeled internal standards, temperature gradient chromatograms were recorded by increasing the temperature manually by 50 °C every 20 s. The corresponding determined desorption temperatures are presented in Table S-1 (Supporting Information) and ranged from 200 to 400 °C. As expected, esters with a short side chain (i.e., steroid Ac or Pr) showed the lowest desorption temperatures (from 200 to 250 °C) whereas those exhibiting a long side chain (i.e., steroid Dc) showed the highest desorption temperatures (from 350 to

5651

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

Technical Note

Figure 1. Schematic workflow proposed to identify steroid esters in less than 3 min.

400 °C). Influence of the steroid base seemed to be weak as steroids with the same side chain showed identical desorption temperature (e.g., T Bz, E2 3Bz, B Bz, and N Bz have the same desorption temperature observed at 300 °C). For fragmentation pathway study and identification of unknown samples, the APCI probe was thereafter set at 450 °C to ensure complete desorption. Particular attention was paid to the stability of the esters, and no thermal degradation was observed for any of the compounds of interest at this temperature. Mechanisms of Fragmentation in Both Positive and Negative Modes. All steroid esters were analyzed independently in both positive and negative scan modes (range from 50 to 500 m/z) with a QqQ MS instrument. Examples of corresponding obtained mass spectra are presented in Figure S1a,c for T Ac and Figure S-2a,c for E2 17Dc (direct introduction at 10 μg mL−1) (Supporting Information). Neither in positive nor in negative mode could any of the 21 analyzed steroid esters be observed under radical ion forms. The presence of water vapor in the source favors the formation of the protonated or deprotonated molecules as already observed by Ray et al. for steroids.28 Thus, all steroid esters formed either [M + H]+ or [M − H]− species depending on the polarity applied. In positive mode, the base peak was observed related to the steroid moiety (i.e., testosterone (T), estradiol (E2), nandrolone (N), or boldenone (B)) and corresponded to the steroid moiety with a loss of water, except for a few testosterone esters which showed the [M + H]+ ions as the most intense (i.e., T Ac (Figure S-1a, Supporting Information), T Pr, and T Cy). In negative mode, the base peak always corresponded to the [M − H]− except for E2 17Dc which was observed as 271.2 m/z corresponding to [E2 − H]− (Figure S-2b, Supporting Information). In order to study the fragmentation of the selected steroid esters, product ion scan experiments on the [M + H]+ and the [M − H]− ions, at 20 and 30 eV (collision energies), respectively, were performed. Examples of obtained mass spectra are presented in Figure S-

1b,d for T Ac and Figure S-2b,d for E2 17Dc (Supporting Information). The results are summarized in Table 1, emphasizing the specific and common ions within the same steroid family. In general, fewer fragmentations could be observed in the negative mode as compared to the positive mode. In APCI+, the most common fragmentation pattern observed for all the esters was the loss of the esters side chain and the subsequent observation of the free steroids as already reported in ESI+.36 Within the same steroid family, several common ions could be observed resulting from the fragmentation of the steroid itself. In APCI-, the only shared ions were the free steroid with a loss of water for testosterone, estradiol, and boldenone; no common ions could be observed within the nandrolone esters family. In negative mode, the direct observation of the ester side chains whatever the steroid moiety considered (e.g., 59.0 m/z for the acetate, 171.1 m/z for the decanoate, etc.) was found to be interesting since it immediately informs on the presence of such ester based compounds. Combining negative and positive modes therefore gives access to valuable complementary pieces of information: steroid moiety in positive mode and ester side chains in negative mode. Development of a Simple Workflow to Identify Steroid Esters. In addition to these results on steroid esters fragmentation, it was further observed for the 21 compounds of interest that a common neutral loss of 96.1 in APCI+ resulted from the following transitions: 271.2 > 175.1 (T base), 269.2 > 173.1 (B base), 257.2 > 161.1 (N base), and 255.2 > 159.1 (E2 base). It should be noted that specific product ion of m/z 175 only for testosterone esters (and not for free testosterone) in ESI+ was already reported by You et al; unfortunately, the fragmentation pathway generating this ion is still unknown.36 Specific ions 173 m/z, 161 m/z, and 159 m/z, belonging to boldenone, nandrolone, and estradiol esters, respectively, were also observed with collision induced dissociation experiments of the corresponding free steroids. Nevertheless and to the best of 5652

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

Technical Note

Figure 2. Workflow applied on B Bz at 10 μg mL−1 in MeOH. (a) Neutral mass loss spectrum of 96.1. (b) Precursor ion scan mass spectrum of 269.2 m/z. (c) Product ion scan spectrum of 389.3 m/z.

To further assess the relevance of the proposed strategy, the next step consisted in blind analysis of unknown mixtures of steroid esters (up to three) prepared within the laboratory. Once again, all steroid esters were identified using the developed workflow. Workflow Application on Oily Preparations. The developed workflow was applied on diluted oily solutions of BOLD-UN 200, TESTO-PR 100, and Laurabolin, confirming the presence of B Un, T Pr, and N Lr, respectively. The workflow was then applied on Sustanon solution (mixture of T Pr, T PhPr, T Iso, and T Dc); corresponding mass spectra of neutral loss scan and precursor ion scan are presented in Figure 3. As expected, all testosterone esters could be identified according to the developed protocol (data not shown). Neutral loss scan of 96.1 further revealed the presence of other molecule(s) through the detection of an ion at m/z 269.2 potentially corresponding to a boldenone moiety. Precursor ion scan mass spectrum of 269.2 m/z revealed the presence of four pseudo molecular ions [M + H]+ at 361.2, 403.3, 437.3, and 459.3 m/z which were then further investigated through product ion scans of the corresponding [M − H]− to show ions at 73.1, 115.1, 149.1, and 171.1 m/z corresponding to the following ester side chains: Pr, PhPr, Iso, and Dc, respectively. The mass difference between the pseudo molecular ions and the side chain esters was 304 Da which does not support the idea of the presence of any boldenone esters (as boldenone has a mass of 286 Da). Such a mass difference could be attributed to hydroxy-testosterone esters. Such a hypothesis was further confirmed by accurate mass measurement with the Q-TOF MS instrument. To validate this identification, the fragmentation pathway should be further studied together with the analysis of corresponding standard solutions, either in GC/MS/MS or LC-

our knowledge, no information regarding their fragmentation origins are reported in the literature. A first screening strategy, based on the specific neutral loss of 96.1, has therefore been set up. All steroid esters were analyzed with neutral loss scan mode and showed the expected ions (i.e., 271.2 m/z for T esters, 269.2 m/z for B esters, 257.2 m/z for N esters, and 255.2 m/z for E2 esters). Thus, this first approach allowed one to detect the presence of steroid esters (or free steroid in the case of N, B, and E2) in just a few minutes. Nevertheless, no information concerning the identification of the steroid esters was provided. Thus, on the basis of precedent results about steroid ester fragmentation (see section Mechanisms of Fragmentation in Both Positive and Negative Modes), two additional steps were proposed: (i) precursor ion scan mode in APCI+ of the detected specie(s) in order to obtain the mass of the [M + H]+ and (ii) product ion scan mode in APCI− of the corresponding [M − H]− to identify the ester side chain. The resulting workflow is summarized in Figure 1 and has been applied for all the steroid ester standards in order to check its relevance and efficiency. As expected, each steroid ester was identified in 3 min. Corresponding mass spectra for B Bz are presented in Figure 2 as an example. As can be observed, the neutral mass loss spectrum in APCI+ showed only one intense peak at 269.2 m/z corresponding to the boldenone moiety with a loss of water. The presence of boldenone in the solution is thus detected. Then, precursor ion scan of 269.2 m/z in positive mode allowed one to detect the [M + H]+ species at m/z 391.3 and thus determine the mass of the compound (390.3 g mol−1). Finally, product ion scan in negative mode of the corresponding found molecular mass further allowed ester moiety identification (121.1 m/z corresponding to benzoate ester). 5653

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

Technical Note

Figure 3. Workflow applied on Sustanon solution diluted by a factor of 1000 in MeOH. (a) Neutral mass loss spectrum of 96.1. (b) Precursor ion scan mass spectrum of 269.2 m/z. (c) Precursor ion scan mass spectrum of 271.2 m/z.

determined for the 21 steroid esters, and results are presented in Table S-3 (Supporting Information). Regarding repeatability (n = 6) of the obtained areas for the different steroid esters, variations never exceeded 20% and more than half did not exceed 10%. Limits of quantification (LOQ) were determined by the analysis of the steroid esters at the corresponding level with a signal-to-noise ratio of 10:1. Almost all the steroid esters exhibited LOQ at 1 ng L−1 with the exception of several esters exhibiting a long side chain with a LOQ determined at 10 ng L−1 (i.e., T Hp, T Dc, E2 17Cy, and E2 17Dc). For the linearity, all the steroid esters considered in this study exhibited a linear correlation coefficient (R2) superior to 0.950. The best results were obtained for testosterone esters with R2 above 0.990, easily explainable with the use of the corresponding labeled isotope standards. This method was then applied for the quantification of B Un, T Pr, and N Lr in BOLD-UN 200, TESTO-PR 100, and Laurabolin solutions. According to the concentrations, indicated on the bottles, the obtained results showed a maximal bias of 7%. Thus, a robust method allowing quantification of steroid esters in oily solution was developed with ASAP-MS, which showed performances as never reported before for any AMS strategy.

MS/MS. In summary, the developed protocol allowed one to (i) confirm the presence of the four testosterone esters and (ii) investigate potential degradation of the commercial solution through the detection of byproducts. The developed workflow can therefore also be applied to check and ensure the purity and the stability of oily preparations. Quantification of Steroid Esters by ASAP-MS. For quantification purposes, all the steroid esters were thus analyzed separately by a QqQ MS instrument in APCI+ mode, selected as the most sensitive mode. Cone voltage and capillary potential were optimized for each molecular ion in order to reach the highest signal for the [M + H]+ ion. Then, collision induced dissociation (CID) energy was optimized for each molecular ion to settle the two most intense transitions, one used for quantification and the second for identification purposes as indicated in Table S-2 (Supporting Information). The relative ion ratios of the identification transitions were calculated on the basis of standard solutions. Tolerance limits of ±20%, ±25%, and ±30% were applied for ions with relative abundances above 50%, between 20% and 50%, and below 20%, respectively. In order to avoid possible bias resulting from the sampling volume, isotope labeled internal standards were used for the quantification. Since all corresponding isotope labeled internal standards are not commercially available, we selected four isotope labeled testosterone esters. The length of the ester side chain was the criteria of choice for the selection of the most relevant internal standard and, therefore, covered the range of esters encountered. Method optimization was performed on standard solutions since the matrix effect resulting from the oily solution is negligible due to 1000-fold dilution in MeOH. The sensitivity (limit of quantification), the repeatability at 1 mg L−1, and the linearity (from 100 ng L−1 to 10 mg L−1) were

■

CONCLUSION In this study, we assessed the benefits of ASAP-MS for high throughput identification and quantification of anabolic steroid esters for efficient screening of drug preparation. The workflow allowed fast screening in minutes, with minimal sample preparation, of oily preparations and can be easily extended to tablets. Such an approach can be very useful to support the control and enforcement authorities. Moreover and for the first time, ASAP-MS was used for quantification thanks to isotope labeled internal standards. At the moment, the only limitation of this approach is its lack of automation. 5654

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655

Analytical Chemistry

■

Technical Note

(24) Hajslova, J.; Cajka, T.; Vaclavik, L. Trends Anal. Chem. 2011, 30, 204−218. (25) Luosujärvi, L.; Arvola, V.; Haapala, M.; Pól, J.; Saarela, V.; Franssila, S.; Kostiainen, R.; Kotiaho, T.; Kauppila, T. J. Eur. J. Pharm. Sci. 2008, 34 (2008), S29. (26) Kauppila, T. J.; Flink, A.; Haapala, M.; Laakkonen, U. M.; Aalberg, L.; Ketola, R. A.; Kostiainen, R. Forensic Sci. Int. 2011, 210, 206−212. (27) McEwen, C. N.; McKay, R. G.; Larsen, B. S. Anal. Chem. 2005, 77, 7826−7831. (28) Ray, A.; Hammond, J.; Major, H. Eur. J. Mass Spectrom. 2010, 16, 169−174. (29) Fussell, R. J.; Chan, D.; Sharman, M. Trends Anal. Chem. 2010, 29, 1326−1335. (30) Petucci, C.; Diffendal, J. J. Mass Spectrom. 2008, 43, 1565−1568. (31) McEwen, C. N.; Gutteridge, S. J. Am. Soc. Mass Spectrom. 2007, 17, 1274−1778. (32) Lindberg, J.; DerMarderosian, A. Planta Med. 2012, 78, PJ65. (33) Rozenski, J. Int. J. Mass Spectrom. 2011, 304, 204−208. (34) Smith, M. J. P.; Cameron, N. R.; Mosely, J. A. Analyst 2012, 137, 4524−4530. (35) Wang, S. Z.; Fan, X.; Zheng, A. L.; Wang, Y. G.; Dou, Y. Q.; Wei, X. Y.; Zhao, Y. P.; Wang, R. Y.; Zong, Z. M.; Zhao, W. Fuel 2014, 117, 556−563. (36) You, Y.; Uboh, C. E.; Soma, L. R.; Guan, F.; Li, X.; Liu, Y.; Rudy, J. A.; Chen, J.; Tsang, D. J. Chromatogr., A 2011, 1218, 3982− 3993.

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +33 2 40 68 78 80. Fax: +33 2 40 68 78 78. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS We thank the “Region Pays de la Loire, France” for its financial support. REFERENCES

(1) Van Eenoo, P.; Delbeke, F. T. J. Steroid Biochem. Mol. Biol. 2006, 101, 161−178. (2) Sjöqvist, F.; Garle, M.; Rane, A. Lancet 2008, 371, 1872−1882. (3) Kintz, P. Leg. Med. 2003, 5, S29−S33. (4) The World Anti Doping Code. The 2014 Prohibited List International Standard; World Anti-Doping Agency: Montreal, 2013. (5) Mooney, M. H.; Elliott, C. T.; Le Bizec, B. Trends Anal. Chem. 2009, 28, 665−675. (6) Pinel, G.; Weigel, S.; Antignac, J. P.; Mooney, M. H.; Elliott, C.; Nielen, M. W. F.; Le Bizec, B. Trends Anal. Chem. 2010, 29, 1269− 1280. (7) Council Directive 96/22/EC of 29 April 1996. Official Journal of the European Communities; European Union, 1996; No. L 125/3. (8) Da Justa Neves, D. B.; Marcheti, R. G. A.; Caldas, E. D. Forensic Sci. Int. 2013, 228, e81−e83. (9) Scarth, J. P.; Kay, J.; Teale, P.; Akre, C.; Le Bizec, B.; De Brabander, H. F.; Vanhaecke, L.; Van Ginkel, L.; Points, J. Drug Test. Anal. 2012, 4, 40−49. (10) Bichon, E.; Béasse, A.; Prevost, S.; Christien, S.; Courant, F.; Monteau, F.; Le Bizec, B. Rapid Commun. Mass Spectrom. 2012, 26, 819−827. (11) Kaabia, Z.; Dervilly-Pinel, G.; Hanganu, F.; Cesbron, N.; Bichon, E.; Popot, M. A.; Bonnaire, Y.; Le Bizec, B. J. Chromatogr., A 2013, 1284, 126−140. (12) Gray, B. P.; Viljanto, M.; Bright, J.; Pearce, C.; Maynard, S. Anal. Chim. Acta 2013, 787, 163−172. (13) Strano-Rossi, S.; Castrignanò, E.; Anzillotti, L.; Odoardi, S.; DeGiorgio, F.; Bermejo, A.; Pascali, V. L. Anal. Chim. Acta 2013, 793, 61−71. (14) Duffy, E.; Mooney, M. H.; Elliott, C. T.; O’Keeffe, M. J. Chromatogr., A 2009, 1216, 8090−8095. (15) De Wasch, K.; De Brabander, H.; Courtheyn, D.; Van Peteghem, C. Analyst 1998, 123, 2415−2422. (16) Daeseleire, E.; Vanoosthuyze, K.; Van Peteghem, C. J. Chromatogr., A 1994, 674, 247−253. (17) Takáts, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471−473. (18) Cody, R. B.; Laramée, J. A.; Durst, H. D. Anal. Chem. 2005, 77, 2297−2302. (19) Venter, A.; Nefliu, M.; Graham Cooks, R. Trends Anal. Chem. 2008, 27, 284−290. (20) Nielen, M. W. F.; Hooijerink, H.; Claassen, F. C.; Van Engelen, M. C.; Van Beek, T. A. Anal. Chim. Acta 2009, 637, 92−100. (21) Nielen, M. W. F.; Nijrolder, A. W. J. M.; Hooijerink, H.; Stolker, A. A. M. Anal. Chim. Acta 2011, 700, 63−69. (22) de Rijke, E.; Hooijerink, D.; Sterk, S. S.; Nielen, M. W. F. Food Addit. Contam. 2013, 30, 1012−1019. (23) Morelato, M.; Beavis, A.; Kirkbride, P.; Roux, C. Forensic Sci. Int. 2013, 226, 10−21. 5655

dx.doi.org/10.1021/ac501072g | Anal. Chem. 2014, 86, 5649−5655