Rapid Screening of Anabolic Steroids in Urine by Reactive Desorption

Open Access ... Fast screening for anabolic steroids in whole urine is achieved by combining ... in. thickness (25 μm), was purchased from Small Part...
3 downloads 0 Views 168KB Size
Anal. Chem. 2007, 79, 8327-8332

Rapid Screening of Anabolic Steroids in Urine by Reactive Desorption Electrospray Ionization Guangming Huang,† Hao Chen,‡ Xinrong Zhang,† R. Graham Cooks,*,‡ and Zheng Ouyang*,§

Department of Chemistry, Tsinghua University, Beijing 100084, China, Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, and Weldon School of Biomedical Engineering, Purdue University, West Lafayette, Indiana 47907

Fast screening for anabolic steroids in whole urine is achieved by combining reactive desorption electrospray ionization (reactive DESI) and tandem mass spectrometry. Spray solutions containing hydroxylamine allow heterogeneous reactions of hydroxylamine with the carbonyl group of the steroids during the ionization process. Seven steroids, including a glycosteroid, were examined. The ion/molecule reaction adduct and the oxime formed via its dehydration were observed using reactive DESI; the protonated and sodiated forms of the ionized steroid were also observed both in reactive DESI and in DESI performed without the added hydroxylamine reagent. Paper, glass, and polytetrafluoroethylene were tested as sample substrates, but the glycosteroid was ionized intact without hydrolysis only from polytetrafluoroethylene. Limits of detection for the pure compounds were less than 1 ng, dynamic ranges were typically 2 orders of magnitude, and analysis times were just a few seconds. Concentration levels of ketosteroids in raw urine relevant to screening for sports doping (approximately 20 ng/mL) can be reached using a simple solid-phase microextraction (SPME) preconcentration step. Reactive DESI provided significant improvements in ionization efficiency of these steroids in raw undiluted urine as compared to conventional DESI; suppression effects due to the sample matrix were minimal and the urine matrix had no deleterious effect on steroid detection limits. Tandem mass spectrometry provided confirmation of analyte identification in this rapid screening process. Anabolic steroids have been used increasingly by athletes to alter muscle mass1 and to enhance performance.2 Significant effort has been put into the development of analytical methods for anabolic steroids in urine to control such drug abuse. Methods * Corresponding authors. Professor R. Graham Cooks, Department of Chemistry, Purdue University, West Lafayette, IN, 47907. Phone: (765) 4945262. Fax: (765) 494-9421. E-mail: [email protected]. Professor Zheng Ouyang, Weldon School of Biomedical Engineering, Purdue University, 206 South Intramural Drive, West Lafayette, IN, 47907-2032. Phone: +1 765 494-2214. Fax: +1 765 496-1912. E-mail: [email protected]. † Tsinghua University. ‡ Department of Chemistry, Purdue University. § Weldon School of Biomedical Engineering, Purdue University. (1) Maravelias, C.; Dona, A.; Stefanidou, M.; Spiliopoulou, C. Toxicol. Lett. 2005, 158, 167-175. (2) Deventer, K.; Van Eenoo, P.; Delbeke, F. T. Biomed. Chromatogr. 2006, 20, 429-433. 10.1021/ac0711079 CCC: $37.00 Published on Web 10/05/2007

© 2007 American Chemical Society

ranging from thin layer chromatography,3 through radioimmuno assays4 to enzyme-linked immunosorbent assays,5 have been implemented for steroid screening. However, the accepted analytical method for anabolic steroids and/or their metabolites has long been gas chromatography/mass spectrometry (GC/MS),6-13 due to the highly specific and quantitative information it provides. Liquid chromatography-mass spectrometry (LC-MS) has also been applied successfully.14-18 Because of the complexity of the urine sample, standard sample preparation procedures19 include multiple steps of sample extraction, hydrolysis, and derivatization.20,21 Although the final step of analysis by mass spectrometry takes little time, the sample preparation steps are usually both time-consuming and labor intensive. Because more rapid screening of the steroids in urine is highly desirable, the screening method for steroids in raw urine reported here consists of only two simple steps: (i) a drop of urine is placed on the substrate and (ii) mass spectra are recorded by spraying the (3) Jansen, E.; Vandenbosch, D.; Stephany, R. W.; Vanlook, L. J.; Vanpeteghem, C. J. Chromatogr., Biomed. Appl. 1989, 489, 205-212. (4) Elliott, C. T.; Francis, K. S.; Shortt, H. D.; McCaughey, W. J. Analyst 1995, 120, 1827-1830. (5) Hungerford, N. L.; Sortais, B.; Smart, C. G.; McKinney, A. R.; Ridley, D. D.; Stenhouse, A. M.; Suann, C. J.; Munn, K. J.; Sillence, M. N.; McLeod, M. D. J. Steroid Biochem. Mol. Biol. 2005, 96, 317-334. (6) Sjovall, J.; Axelson, M. J. Steroid Biochem. Mol. Biol. 1979, 11, 129-134. (7) Chung, B. C.; Choo, H. Y. P.; Kim, T. W.; Eom, K. D.; Kwon, O. S.; Suh, J. W.; Yang, J. S.; Park, J. S. J. Anal. Toxicol. 1990, 14, 91-95. (8) Daeseleire, E.; Vandeputte, R.; Van Peteghem, C. Analyst 1998, 123, 25952598. (9) Choi, M. H.; Chung, B. C. Analyst 1999, 124, 1297-1300. (10) Munoz-Guerra, J.; Carreras, D.; Soriano, C.; Rodriguez, C.; Rodriguez, A. F. J. Chromatogr., B 1997, 704, 129-141. (11) Wolthers, B. G.; Kraan, G. P. B. J. Chromatogr., A 1999, 843, 247-274. (12) Buiarelli, F.; Cartoni, G. P.; Amendola, L.; Botre, F. Anal. Chim. Acta 2001, 447, 75-88. (13) Sekera, M. H.; Ahrens, B. D.; Chang, Y. C.; Starcevic, B.; Georgakopoulos, C.; Catlin, D. H. Rapid Commun. Mass Spectrom. 2005, 19, 781-784. (14) Liberato, D. J.; Yergey, A. L.; Esteban, N.; Gomezsanchez, C. E.; Shackleton, C. H. L. J. Steroid Biochem. Mol. Biol. 1987, 27, 61-70. (15) Ferguson, P. L.; Iden, C. R.; McElroy, A. E.; Brownawell, B. J. Anal. Chem. 2001, 73, 3890-3895. (16) Yamamoto, A.; Kakutani, N.; Yamamoto, K.; Kamiura, T.; Miyakoda, H. Environ. Sci. Technol. 2006, 40, 4132-4137. (17) Mazzarino, M.; Botre, F. Rapid Commun. Mass Spectrom. 2006, 20, 34653476. (18) Catlin, D. H.; Sekera, M. H.; Ahrens, B. D.; Starcevic, B.; Chang, Y.-C.; Hatton, C. K. Rapid Commun. Mass Spectrom. 2004, 18, 1245-1249. (19) Leunissen, W. J. J.; Thijssen, J. H. H. J. Chromatogr. 1978, 146, 365-380. (20) Harrison, L. M.; Martin, D.; Gotlin, R. W.; Fennessey, P. V. J. Chromatogr., Biomed. Appl. 1989, 489, 121-126. (21) Kuuranne, T.; Kotiaho, T.; Pedersen-Bjergaard, S.; Rasmussen, K. E.; Leinonen, A.; Westwood, S.; Kostiainen, R. J. Mass Spectrom. 2003, 38, 16-26.

Analytical Chemistry, Vol. 79, No. 21, November 1, 2007 8327

sample with an aqueous solution containing an appropriate reagent. Tandem mass spectrometry is used to confirm the nature of the ions seen in the desorption electrospray ionization (DESI) mass spectrum. The experiment can be done on batches of samples at rates approaching 1 sample/second. As will be seen, the method has high specificity and absolute sensitivity; the concentration sensitivity is such that one simple extraction step is needed to reach action levels for sports doping of anabolic steroids. While mass spectrometry is well-known to allow high specificity and high sensitivity in the analysis of complex mixtures,22-24 and confirmation of identity and quantification of the amounts of particular substances of interest, including parent drugs and their metabolites, can be accomplished by means of tandem mass spectrometry,25 the direct introduction of raw urine samples into mass spectrometers usually results in low sensitivity to the compounds of interest and in rapid contamination of the instrument. The desorption electrospray ionization (DESI) method26 is effective for direct sampling of condensed phase compounds27-29 from a wide variety of substrates in the ambient environment.30-33 Its direct application to urine samples is known to reduce matrix suppression effects and to yield spectra that allow identification of individual constituents, including drugs of abuse and their metabolites.34,35 An additional recent development that greatly improves the selectivity and efficiency with which compounds containing specific functional groups are detected is the use of a DESI spray solution that contains specific reagents intended to allow particular ion/molecule reactions36-38 during the sampling process. In this experiment, reactive DESI, instead of spraying only solvent on the sample surface, reagents are added into the solvent and the reactant ions in the charged microdroplets striking the sample surface react with the targeted compounds to form derivatized products. An analogy can be drawn between reactive (22) Kruger, T. L.; Litton, J. F.; Kondrat, R. W.; Cooks, R. G. Anal. Chem. 1976, 48, 2113-2119. (23) McLafferty, F. W. Acc. Chem. Res. 1980, 13, 33-39. (24) Soni, M.; Bauer, S.; Amy, J. W.; Wong, P.; Cooks, R. G. Anal. Chem. 1995, 67, 1409-1412. (25) Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/Mass Spectrometry: Techniques and Applications of Tandem Mass Spectrometry; VCH Publishers: New York, 1988; pp 311-315. (26) Takats, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471-473. (27) Chen, H. W.; Talaty, N. N.; Takats, Z.; Cooks, R. G. Anal. Chem. 2005, 77, 6915-6927. (28) Bereman, M. S.; Nyadong, L.; Fernandez, F. M.; Muddiman, D. C. Rapid Commun. Mass Spectrom. 2006, 20, 3409-3411. (29) Williams, J. P.; Lock, R.; Patel, V. J.; Scrivens, J. H. Anal. Chem. 2006, 78, 7440-7445. (30) Takats, Z.; Cotte-Rodriguez, I.; Talaty, N.; Chen, H. W.; Cooks, R. G. Chem. Commun. 2005, 1950-1952. (31) Van Berkel, G. J.; Ford, M. J.; Deibel, M. A. Anal. Chem. 2005, 77, 12071215. (32) Kauppila, T. J.; Wiseman, J. M.; Ketola, R. A.; Kotiaho, T.; Cooks, R. G.; Kostiainen, R. Rapid Commun. Mass Spectrom. 2006, 20, 387-392. (33) Kauppila, T. J.; Talaty, N.; Salo, P. K.; Kotiaho, T.; Kostiainen, R.; Cooks, R. G. Rapid Commun. Mass Spectrom. 2006, 20, 2143-2150. (34) Cotte-Rodriguez, I.; Takats, Z.; Talaty, N.; Chen, H. W.; Cooks, R. G. Anal. Chem. 2005, 77, 6755-6764. (35) Talaty, N.; Takats, Z.; Cooks, R. G. Analyst 2005, 130, 1624-1633. (36) Chen, H.; Cotte-Rodriguez, I.; Cooks, R. G. Chem. Commun. 2006, 597599. (37) Cotte-Rodriguez, I.; Chen, H.; Cooks, R. G. Chem. Commun. 2006, 953955. (38) Nyadong, L.; Green, M. D.; De Jesus, V. R.; Newton, P. N.; Fernandez, F. M. Anal. Chem. 2007, 79, 2150-2157.

8328

Analytical Chemistry, Vol. 79, No. 21, November 1, 2007

Scheme 1. Chemical Structures of Seven Anabolic Steroids Used in the Study

DESI and the purely vapor-phase ionization process of chemical ionization. Many standard solution-phase reactions can be implemented under ambient conditions to enhance the selectivity of the DESI sampling process. Besides simple cationization,26,34,38,39 more complex bond forming reactions studied include transacetalization of acylium cations40 and cyclization of phenylboronic acid with cis-diols.36 In this study, hydroxylamine was selected as the reagent to form oximes with anabolic steroid glucuronides as well as neutral steroids.41 EXPERIMENTAL SECTION Steroids, including androstadienedione, stigmastadienone, androsterone hemisuccinate, 5R-androstan-3β,17β-diol-16-one, androsterone glucuronide, epitestosterone, and 6-dehydrocholestenone, were purchased from Steraloids Inc. (Newport, Rhode Island). Their chemical structures are shown in Scheme 1. A porous polytetrafluoroethylene (PTFE) sheet, of 1/16 in. thickness (25 µm), was purchased from Small Parts Inc. (Miami Lakes, FL). Deionized water used in dilutions was obtained using a Milli-Q purification system (Millipore, Bedford, MA). All other reagents were purchased from Sigma-Aldrich (Milwaukee, WI) and used without further purification. Reactive DESI experiments and tandem mass spectrometry were carried out using a Finnigan LTQ mass spectrometer (Thermo Fisher Scientific, Inc., San Jose, CA) fitted with a homebuilt DESI source.42 Except for the urine sample, the steroid (39) Nefliu, M.; Cooks, R. G.; Moore, C. J. Am. Soc. Mass Spectrom. 2006, 17, 1091-1095. (40) Sparrapan, R.; Eberlin, L. S.; Haddad, R.; Cooks, R. G.; Eberlin, M. N.; Augusti, R. J. Mass Spectrom. 2006, 41, 1242-1264. (41) Kushnir, M. M.; Rockwood, A. L.; Roberts, W. L.; Pattison, E. G.; Bunker, A. M.; Fitzgerald, R. L.; Meikle, A. W. Clin. Chem. 2006, 52, 120-128. (42) Takats, Z.; Wiseman, J. M.; Cooks, R. G. J. Mass Spectrom. 2005, 40, 12611275.

Scheme 2. Reaction Between Hydroxylamine and the Carbonyl Group on Steroids

subtraction. All spectra reported are background subtracted, unless specified otherwise.

standards were first dissolved in methanol (5 mg/mL) to form a stock solution and then further diluted with methanol to provide solutions of various concentrations. The stock solutions were kept at 4 °C when not in use. Each sample was prepared by depositing 10 µL of the solution onto an area of 3 mm × 5 mm on the substrate. Four types of substrates were used: polished glass, ground glass, filter paper, and porous PTFE. After the samples had completely dried in air, they were examined using DESI MS. The urine samples were prepared by adding 10 µL of diluted stock solutions into 1 mL of raw urine. The experimental conditions were optimized as follows: for each analysis, the sampling area was about 1 mm × 3 mm. The spray solution of methanol/water (1:1) contained 0.05% acetic acid and 5% hydroxylamine. A flow rate of 3 µL/min, a sprayer-to-surface distance of 1 mm and a spray impact angle of 60° was used. The spray voltage was set at 5 kV, and the nebulizing gas (N2) pressure was 180 psi. A capillary temperature of 200 °C was selected. For each sample, spectra recorded for the blank substrate were used for background

RESULTS AND DISCUSSION Many steroids, including natural and anabolic steroids, contain carbonyl groups, and it has been reported that derivatization by reaction with hydroxylamine in aqueous solution can improve the sensitivity of testosterone analysis by HPLC-MS.41 Reactive DESI provides a convenient way of performing derivatization since reaction can occur in the course of ionization at ambient pressure. Methanol/water containing hydroxylamine was used as the DESI spray solution: protonated hydroxylamine ions are formed during electrospray and carried in microdroplets of solution toward the substrate where they react with neutral steroids (Scheme 2). The scattered progeny droplets containing the derivatized steroids are transferred into the atmospheric interface of the mass spectrometer where they dry to give the ionized, derivatized steroids. The resulting mass spectra display these steroid ions as well as other compounds in urine, such as urea, which also contain carbonyl groups and can compete with the steroids in reacting with hydroxylamine. In this study, the low mass cutoff (LMCO) was set at m/z 200 to eliminate lower mass ions and tandem mass spectrometry was applied to confirm the steroid assignments. Figure 1A shows the MS spectrum of approximately 20 ng of epitestosterone, deposited in an area of 1 mm × 3 mm on a ground glass substrate. In addition to the protonated ([M + H]+, m/z 289) and sodiated ([M + Na]+, m/z 311) ions of epitestosterone,

Figure 1. Reactive DESI analysis of epitestosterone on ground glass, mass spectrum (A) and product ion MS2 spectra of the fragments of the ion/molecule reaction product m/z 322 (B) and the oxime ion m/z 304 (C).

Analytical Chemistry, Vol. 79, No. 21, November 1, 2007

8329

Figure 2. Reactive DESI mass spectra of androsterone glucuronide on (A) ground glass and (B) porous PTFE substrates (without background subtraction). Table 1. Analytical Performance of Reactive DESI for Anabolic Steroidsa limit of detection

steroids androstadienedione stigmastadienone androsterone hemisuccinate 5R-androstan-3β,17βdiol-16-one androsterone glucuronide epitestosterone 6-dehydrocholestenone

monitored ion (m/z)

linear response rangeb (ng)

methanol soln (ng)

spiked raw urine (ng)

318 444 424

1-100 1-100 0.9-40

0.4 0.4 0.3

0.9 1.2 1.0

340

1-40

0.4

1.0

500

2-100

1

322 416

0.6-20 0.8-100

0.2 0.3

10 0.6 0.7

a Porous PTFE was used as the sample substrate. Sampling area was about 1 mm × 3 mm, corresponding to a sample deposition of 2 µL. Three measurements were made for each sample. b Sample prepared in methanol solution. RSDs obtained were between 3.5-7.8%.

the hydroxylamine adduct was observed at m/z 322 and its dehydration product, the protonated oxime, occurred at m/z 304. Tandem mass spectrometry using collision-induced dissociation (CID) was performed to record product ion spectra of these massselected ions (Figure 1B,C). Characteristic fragmentation patterns were obtained. The loss of water from the intact product ion m/z 322 yields the ionized oxime m/z 304 (Figure 1B). Additional fragment ions of m/z 289, 286, and 268 were generated by subsequent loss of a methyl radical, and one and two molecules of water, respectively. The fragment ion of m/z 290 probably arises from the loss of HONH• radical and that at m/z 210 is due to ring cleavage. Other fragment ions observed upon CID include m/z 143, 157, 188, 225, and 251 (Figure 1B). The structural assignment 8330 Analytical Chemistry, Vol. 79, No. 21, November 1, 2007

of the ion m/z 304 as the oxime was also confirmed by MS/MS (Figure 1C). The ion/molecule reaction product ion of m/z 304 yielded fragment ions of 286, 271, and 253 upon CID ascribed to consecutive loss of one water molecule, a methyl radical, and another molecule of water. In addition, characteristic ring cleavages give rise to the fragment ions m/z 124, 138, 152, and 206. Other fragment ions at m/z 176 and 246 are also observed (Figure 1C). The MS/MS spectra of the ions generated by reactive DESI matched the MS/MS spectra of the protonated authentic oxime obtained from normal solution-phase reactions (data shown in Supporting Information). The same set of experiments was conducted using ground glass as the substrate for androstadienedione, stigmastadienone, androsterone hemisuccinate, 5R-androstan-3β,17β-diol-16-one, and 6-dehydrocholestenone. Analogous observations were recorded in the corresponding MS and MS/MS experiments. Comparison of the data recorded for different substrates, ground glass, polished glass, filter paper, and porous PTFE, using androstadienedione, androsterone hemisuccinate, epitestosterone, and 6-dehydrocholestenone, showed no significant differences (data shown in Supporting Information). It is well-known that glycosylated steroids occur in urine and that they usually have relatively low ionization efficiencies. Androsterone glucuronide was therefore selected for study as a typical glucuronide. Typically, hydrolysis is applied before the sample is analyzed by GC/MS or LC-MS.19-21,43-45 In the reactive DESI experiments, androsterone glucuronide was shown to have significantly different characteristics from the other six steroids. (43) Donike, M.; Ueki, M.; Kuroda, Y.; Geyer, H.; Nolteernsting, E.; Rauth, S.; Schanzer, W.; Schindler, U.; Volker, E.; Fujisaki, M. J. Sports Med. Phys. Fitness 1995, 35, 235-250. (44) Gartner, P.; Novak, C.; Einzinger, C.; Felzmann, W.; Knollmuller, M.; Gmeiner, G.; Schanzer, W. Steroids 2003, 68, 85-96. (45) Roig, M.; Segura, J.; Ventura, R. Anal. Chim. Acta 2007, 586, 184-195.

Figure 3. Direct analysis of raw urine samples spiked with steroids using (A) DESI and (B) reactive DESI, and (C) product ion MS2 data on each of the characteristics ions. Porous PTFE substrate with 10 µL of raw urine spiked with 8 ng of androsterone hemisuccinate, 20 ng of 5R-androstan-3β,17β-diol-16-one and 8 ng of epitestosterone.

In the case of this compound only, the response to reactive DESI was strongly dependent on the substrate. When 100 ng of androsterone glucuronide was deposited in an area of 1 mm × 3 mm, no analyte ions were observed for filter paper or for the ground or polished glass substrates (as shown in Figure 2A, without background subtraction); however, the sodiated molecule m/z 489, the product ion m/z 500, and the protonated oxime m/z 482 were each observed with relatively good signal-to-noise ratios when porous nonpolar PTFE was used as the sample surface (Figure 2B). The poor response with paper and glass substrates is presumably caused by strong sample adsorption due to interactions between the hydroxyl groups of the sugar and these polar surfaces. It is also noteworthy that, unlike the other steroids used in this study, the protonated molecule was not observed for androsterone glucuronide, likely due to its relatively low proton affinity. The concentrations of the steroid solutions were systematically varied to allow the preparation of a series of samples with different amounts of each steroid for reactive DESI analysis. Internal standards were used to control the sample-to-sample and run-torun variations. Androstadienedione was used as an internal standard for epitestosterone while epitestosterone was used for androstadienedione, stigmastadienone, androsterone hemisuccinate, 5R-androstan-3β,17β-diol-16-one, androsterone glucuronide, and 6-dehydrocholestenone samples. Every 10 µL of sample deposited on substrate contained 50 ng of internal standard. The limit of detection and the range of concentrations over which responses were linear for each of the steroids are listed in Table 1.

As a demonstration of the effectiveness of reactive DESI in the analysis of steroids in raw undiluted human urine, a urine sample was spiked with three steroids (10 µL of urine with 8 ng of androsterone hemisuccinate, 20 ng of 5R-androstan-3β,17β-diol16-one, and 8 ng of epitestosterone) and analyzed using both DESI and, for comparison, reactive DESI (Figure 3). The peak intensities of the ions generated by reactive DESI are significantly higher, approximately 11-fold for androsterone hemisuccinate, 23-fold for 5R-androstan-3β,17β-diol-16-one, and 8-fold for epitestosterone, than those for the protonated steroid ions generated by conventional (reagentless) DESI (Figure 3A and B). The intensity difference between the derivatized and underivatised steroids shown in Figure 3A,B make it clear that the enhancement achieved by reactive DESI is most effective when suppression effects exist in conventional DESI due to a complex matrix. The identification of the products of reactive DESI was confirmed using MS/MS (Figure 3C). The protonated and sodiated ions were not observed in the reactive DESI spectrum, presumably due to competition for the proton and sodium ions from other compounds in raw urine. The ionized oximes were not observed for the steroids in raw urine analyzed using reactive DESI, a result that is different from the observation made for the analysis of the steroids without matrix. The detection limits for these steroids in urine was less than 1 ng absolute, which is not significantly different from that recorded for the pure samples (Table 1); however, the corresponding solution concentrations are relatively high (>300 ng/ mL, Table 1). To further improve concentration detection limits so as to allow detection of specific steroids from raw urine, a Analytical Chemistry, Vol. 79, No. 21, November 1, 2007

8331

Figure 4. Reactive DESI analysis using carbowax-divinylbezene SPME fiber for prior sampling; (A) raw urine samples without added steroids and (B) with androsterone hemisuccinate, 5R-androstan-3β,17β-diol-16-one and epitestosterone spiked at a concentration of 20 ng/mL each (without background subtraction).

combination of fast preconcentration with reactive DESI was explored. This was done using SPME fibers bearing four types of coating materials including carbowax-divinyl benzene, polydimethylsiloxane, polydimethylsiloxane-divinyl benzene, and polyacrylate (Supelco, Bellefonte, PA). For each test, a SPME fiber was dipped into 1 mL of raw urine spiked with epitestosterone, androsterone hemisuccinate, and 5R-androstan-3β,17β-diol-16-one at chosen concentrations for 2 min. The fiber was dried in air for another 2 min before being exposed to the reactive DESI spray for direct analysis. The fiber coated with carbowax-divinyl benzene was found to be the most effective, and a detection limit of 20 ng/mL was obtained for each of the three steroids from raw urine (Figure 4). The detection of the spiked steroids was confirmed by comparison of the MS2 fragmentations of the ions generated from SPME coupled with reactive DESI with the MS2 spectra of the authentic oxime synthesized in solution and then ionized from a SPME fiber (data shown in Supporting Information). CONCLUSION Rapid screening of steroids in raw urine can be implemented without sample preparation using reactive DESI mass spectrometry. Oxime formation occurs in the course of heterogeneous

8332

Analytical Chemistry, Vol. 79, No. 21, November 1, 2007

phase reaction between the solution-phase hydroxylamine and the solid-phase carbonyl compound, and it is demonstrated to be effective for selective ionization of ketosteroids. Suppression effects due to the complex matrix can be overcome substantially using this method, which gives detection limits under 1 ng. Concentration detection limits are on the order of 20 ng/mL when combined with an SPME preconcentration step. Reactions which are selective for other analytes will be examined in future reactive DESI experiments. ACKNOWLEDGMENT This work was supported by Office of Naval Research (ONR Grant N00014-05-1-0405), the National Science Foundation (NSF Grant CHE-0412782), and National Natural Science Foundation of China (NSFC Grants 20535020 and 20635002). SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review May 24, 2007. Accepted August 27, 2007. AC0711079