Tentative Structural Assignment of a Glucuronide Metabolite of

Article pubs.acs.org/JAFC

Tentative Structural Assignment of a Glucuronide Metabolite of Methyltestosterone in Tilapia Bile by Liquid Chromatography− Quadrupole-Time-of-Flight Mass Spectrometry Upul Nishshanka, Pak-Sin Chu, Eric Evans, Renate Reimschuessel, Nicholas Hasbrouck, Kande Amarasinghe, and Hiranthi Jayasuriya* Center for Veterinary Medicine, U.S. Food and Drug Administration, Laurel, Maryland 20708, United States ABSTRACT: Methyltestosterone (MT), a strong androgenic steroid, is not approved for use in fish aquaculture in the United States. It is used in the U.S. under an investigational new animal drug exemption (INAD) only during the early life stages of fish. There is a possibility that farmers feed fish with MT to enhance production for economic gains. Therefore, there is a need to develop methods for the detection of MT and its metabolite residues in fish tissue for monitoring purposes. Previously, our laboratory developed a liquid chromatography−quadrupole time-of-flight (LC-QTOF) method for characterization of 17-Oglucuronide metabolite (MT-glu) in bile of tilapia dosed with MT. The system used was an Agilent 6530 Q-TOF equipped with electrospray jet stream technology, operating in positive ion mode. Retrospective analysis of the data generated in that experiment by a feature-finding algorithm, combined with a search against an in-house library of possible MT-metabolites, resulted in the discovery of a major glucuronide metabolite of MT in the bile extracts. Preliminary data indicate it to be a glucuronide of a hydroxylated MT (OHMT-glu) which persists in tilapia bile for at least 2 weeks after dosing. We present the tentative structural assignment of the OHMT-glu in tilapia bile and time course of development. This glucuronide can serve as a marker to monitor illegal use of MT in tilapia culture. KEYWORDS: methyltestosterone, MT-glu, OHMT-glu, QTOF, MS/MS, tilapia, bile

■

INTRODUCTION

The molecular features generated by the algorithm were screened against an in-house library of possible MT metabolites. This exercise led to the discovery of a persistent MT glucuronide in bile samples from dosed tilapia which we designate as OHMTglu (Figure 1). Mass spectrometric characterization of the

The exogenous androgen methyltestosterone (MT) is commonly used in newly hatched tilapia fry for sex reversal.1−3 In the last decades, potential misuse of MT in adult fish has been monitored by analyzing the parent compound, MT. However, MT is extensively metabolized and persists only for a short time after administration. Phase 1 biotransformation products such as reduction of A-ring, 3-oxo, and 4-ene groups4 and oxidation at C6, C-11, C-15, and C-16 have been documented5,6 for MT. Phase 1 biotransformation is followed by sulfate and glucuronide conjugation during phase II biotransformation. A large number of compounds have been detected as glucuronides in fish, confirming that glucuronidation is an important route of metabolism in fish.7,8 On the basis of our previous work with MT-glu, a significant portion of MT residues may persist longer as the glucuronide metabolites in the fish. Therefore, glucuronides could serve as better markers for monitoring purposes than the parent compound MT. In our previously published study, 9 we generated a glucuronide-rich extract of the bile by using a selective anion exchange solid-phase extraction procedure. We performed a retrospective analysis of the data generated in that study to search for additional glucuronide metabolites. We processed our data by the “molecular feature extraction” (MFE) algorithm in Agilent’s Mass Hunter software. MFE is a data mining tool that generates a list of molecular features or chemical compounds with retention time, neutral mass, and abundance. MFE removes areas of noise and groups all of the related ions of compounds (isotopes, charge states, adducts, and multimers) to generate molecular features. This article not subject to U.S. Copyright. Published XXXX by the American Chemical Society

Figure 1. Tentative structure of OHMT-Glu.

OHMT-glu is described in this paper. Finally, the use of OHMTglu as a marker in the screening of illegal use of MT is demonstrated by analyzing bile of dosed tilapia at different time points after administration.

■

MATERIALS AND METHODS

Chemicals and Reagents. LC-grade water used in reagent preparations was from a Milli-Q water purification system (Millipore, Bedford, MA). Methanol (MeOH), acetonitrile (ACN), and ethyl Received: January 23, 2015 Revised: April 24, 2015 Accepted: May 17, 2015

A

DOI: 10.1021/jf506151g J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Figure 2. MS/MS of OHMT-glu (A) and MS/MS of MT-glu (B), CE 30. acetate (EtOAc) were of high-performance liquid chromatography (HPLC) grade (Burdick & Jackson, Muskegon, MI). Ammonium Hydroxide (NH4OH), potassium dihydrogen phosphate (KH2PO4), dipotassium hydrogen phosphate (K2HPO4), and formic acid were of reagent grade supplied by Fisher Scientific (Pittsburgh, PA). Oasis MAX strong anion exchange (3 mL, 60 mg) solid-phase extraction (SPE) cartridges were obtained from Waters Corp. (Milford, MA). The lyophilized enzyme β-glucuronidase type VII A from Escherichia coli (20 000U) was purchased from Sigma Chemical Co. (St. Louis, MO). Fish Dosing. Fish were dosed with MT at the Center for Veterinary Medicine aquaculture facility. Tilapia (Oreochromis species, not sex reversed) was obtained from Aquasafra Inc. (Bradenton, FL) as ∼1 g fry. Fish were raised in freshwater recirculating systems consisting of two 500 gal tanks at a temperature range of 20−30 °C until they reached a suitable size for dosing (400−700 g). Tilapia were given a single oral dose of 60 mg/kg body weight MT and sacrificed at 1, 3, 7, and 14 days after the dose. One fish was used for each time point. Bile samples were removed and stored at −80 °C until analysis. Control bile was also removed from a fish that was not fed with MT. The experimental protocol was approved by the Animal Care and Use Committee at the Office of Research, Center for Veterinary Medicine, U.S. Food and Drug Administration, and all procedures were conducted in accordance with the principles stated in the Guide for the Care and Use of Laboratory Animals (2011) and the Animal Welfare Act of 1966 (P.L. 89-544), as amended. Sample Preparation. Bile from dosed and control fish were weighed (∼100 mg each) into 15 mL polypropylene centrifuge tubes. A 1 mL sample of 5% aqueous ammonium hydroxide solution was added to each sample and vortexed for 30 s. The extracts were subjected to SPE cleanup by an automated SPE system (RapidTrace workstation by Caliper Life Science, Hopkinton, MA). Waters Oasis MAX, SPE cartridges were preconditioned with 2.5 mL of MeOH (42 mL/min) followed by 2.5 mL of water (42 mL/min). After the samples were loaded (1 mL/min), cartridges were rinsed successively with 2 mL of 0.5 M NH4OH in 5% MeOH (42 mL/min) followed by 2 mL of 20% MeOH (42 mL/min). Cartridges were dried by passing 6 mL of air (42

mL/min). Glucuronides were eluted with 2 mL of freshly prepared 2% formic acid in MeOH (1 mL/min). The eluents were evaporated to dryness by blowing N2 at 40 °C. Samples were reconstituted to 200 μL with MeOH/water (60/40 v/v) and transferred to autosampler vials for mass spectral analysis after filtering through a 0.2 μm nylon syringe filter. Enzymatic Hydrolysis of Bile from Dosed Fish with βGlucuronidase from E. coli. Glucuronide enriched fraction from SPE of control and dosed fish bile (3 days) was evaporated to dryness and dissolved in 200 μL of 0.25 M potassium phosphate buffer (pH ∼ 6.8). A 40 μL (1000U) of β-glucuronidase enzyme from E. coli in 0.25 M phosphate buffer was added. The phosphate buffer was prepared by mixing 55 mL of 0.5 M dipotassium hydrogen phosphate and 45 mL of 0.5 M potassium dihydrogen phosphate. The sample was incubated overnight in a shaking water bath at 37 °C and was extracted twice with 300 μL of ethyl acetate. The combined ethyl acetate extract was evaporated to dryness under nitrogen in a TurboVap evaporator at 40 °C. The residue was reconstituted in 200 μL of 60:40 methanol:water. The sample was transferred to autosampler vial after filtering through a 0.2 μm nylon syringe filter for mass spectral analysis. Liquid Chromatography−Quadrupole Time-of-Flight Mass Spectrometry (LC-QTOFMS) Analysis. Chromatographic separation was carried out using a HPLC system (Agilent Series 1100 or 1290 Palo Alto, CA) equipped with a reversed phase C18 analytical column of 150 × 2.0 mm, 3 μm particle size (Phenomenex Luna C18) with a guard column of the same packing. The column oven was held at 40 °C, and the autosampler tray was maintained at 20 °C. The injection volume was 5 or 10 μL. Mobile phase A and B were 0.1% aqueous formic acid and ACN with 0.1% formic acid, respectively. The optimized chromatographic method on the Agilent 1290 pump was a linear gradient of 40− 90% ACN in 10 min, hold at 90% ACN for 2 min, linear gradient of 9040% ACN in 1 min, followed by equilibration at 40% ACN for 12 min before the next injection. The optimized method for Agilent 1100 pump was a linear gradient of 30−90% ACN in 10 min, hold at 90% ACN for 5 min, linear gradient of 90−30% ACN in 1 min, followed by equilibration at 30% ACN for 9 min before the next injection. Flow rate was 0.2 mL/ min for both gradients. The HPLC system was connected to a B



Article

Figure 3. MS/MS of OHMT (in-source generated 319).

Figure 4. MS/MS spectra of C-6α (A), C-6β (B), and C-11α (C) hydroxy-MT standards. quadrupole time-of-flight mass spectrometer model 6530 from Agilent (Agilent Technologies, Palo Alto, CA) equipped with electrospray jet stream technology operating in the positive ion mode, using the following operation parameters: capillary voltage 3500 V; nebulizer pressure, 50 psi; drying gas, 12 L/min; gas temperature, 325 °C; nozzle voltage, 1000 V; fragmentor voltage, 165 V; skimmer voltage, 65 V; and octapole RF, 750 V. LC/MS accurate mass spectra were recorded across the range of m/z 50−1000 at 4 GHz high resolution. Data were collected

in both centroid and profile formats and processed with Mass Hunter software. Accurate mass measurements of each peak from the extracted ion chromatograms were obtained by means of a calibrant solution delivered by an external isocratic pump. This solution contains the internal reference masses of purine (C5H4N4) at m/z 121.0509 and HP921 [hexakis-(1H, 1H, 3H-tetrafluoro-pentoxy) phosphazene] (C18H18O6N3P3F24) at m/z 922.009. The OHMT-glu ion at calculated exact mass of m/z 495.2589 was extracted from the total ion C



Article

Figure 5. Extracted ion (m/z 319.2268) chromatograms of hydroxylated MT standards (6α, 6β, and 11α) and hydrolyzed bile (OHMT). chromatogram and selected as the precursor ion for collision-induced dissociation (CID) experiment. For targeted tandem mass spectrometry (MS/MS) analysis, product ion scan range was 50−500 Da. A retention time window was set at ±0.5 min, and the isolation width was set at medium (∼4 m/z). Collision energy of 30 eV was used for MS/MS experiments to obtain maximum intensity of fragments at m/z 319.2265, 301.2156, and 283.2056.

■

for them. We took advantage of published literature of steroid fragmentation pathways and compared them with our experimental observations. This is a difficult strategy because small differences in chemical structure such as additional double bonds and methyl groups can result in a significantly different fragmentation pathway of steroids.10 The MS/MS analysis of OHMT-glu at a collision energy (CE) of 30 eV (Figure 2A) showed an abundant peak at m/z 319.2265 (C20H31O3) [M + H − 176 Da]+ due to the loss of a dehydrated glucuronide molecule (176 Da). Successive losses of water from m/z 319 were observed at m/z 301.2156 (C20H29O2), 283.2056 (C20H27O), and 265.1946 (C20H25) for [M + H − glu − H2O]+, [M + H − glu −2H2O]+, and [M + H − glu − 3H2O]+, respectively, indicating three oxygens in OHMT-glu structure. The MS/MS spectrum of our previously synthesized MT-glu showed corresponding losses of dehydrated glucuronide molecule followed by only two consecutive water losses (Figure 2B). This further confirms that OHMT-glu bears an additional hydroxyl group in the aglycone in comparison to MT-glu. We designate the aglycone of OHMT-glu as OHMT. Our challenge was to place the additional hydroxyl group in OHMT. Hydroxylation is a common metabolic pathway of anabolic steroids. Therefore, an understanding of fragmentation of hydroxylated steroids is very important in their structure elucidation. Also, an in-depth comparison of fragmentation of known related hydroxylated MT analogs with that of OHMT is a useful strategy for structure elucidation. The hydroxylation can be at any of the ring carbons or one of the three methyl groups. Hydroxylation has been reported previously at C6, 7, 11, 15, and

RESULTS AND DISCUSSION

We utilized the MFE algorithm in the Q-TOF Mass Hunter qualitative analysis software to detect the molecular features in the glucuronide-rich extract of dosed tilapia bile. For the extraction of molecular features, we used a peak height filter of >15 000 in the compounds filter tab. The charge state was limited to 1. Ion species were limited to the positive ions H+ and Na+. The 523 molecular features extracted under these criteria were then searched against our in-house database of MT metabolites. We used the criteria of mass and a mass accuracy tolerance of 5 ppm for the library search to identify OHMT-glu of the correct formula C26H38O9 at m/z 495.2579 with a mass accuracy of 2 ppm. The in-house MT metabolite database was constructed to include the exact masses and the molecular formulas of mono-, di-, and tri-hydroxylated MT and their glucuronides. According to the formula C26H38O9, OHMT-glu has an additional hydroxyl function. The QTOF high-resolution mass spectrometry (HRMS) approach utilized provided us with exact masses of good mass accuracy to determine the chemical formula of OHMT-glu as well as its fragment ions. The correct formula generated for the fragment ions enabled us to propose structures D



Article

Scheme 1. Proposed Mechanism for the Formation of m/z 243 and 225

16 of MT and related compounds.7,11−13 To simplify the discussion, the nominal masses are sometimes used in this paper. To find further structural information on OHMT, we optimized the fragmentation voltage at 200 V in the MS1 experiment to produce in-source generated fragment ion at m/z 319 (protonated OHMT) from precursor ion at m/z 495 (the protonated OHMT-Glu). We then recorded the MS/MS spectrum of the in-source generated OHMT (m/z 319) (Figure 3) and compared it with those generated from the three commercially available standards C-6α, -6β, and -11α hydroxyMT (Figure 4). Mass spectra of hydroxylated MT standards and OHMT showed significant ions at m/z 301, 283, and 265 indicating three sequential losses of water (Figures 3 and 4). In addition, two other common ions were observed at m/z 243 and 225. These can be attributed to the loss of 58 Da (acetone) from fragment ions at m/z 301 and 283, respectively.14 However, the lower m/z region of product ion spectra of the three standards looked quite different from that of OHMT. We observed very prominent ions at m/z 97.0630 (C6H9O) and 109.0642 (C7H9O) in the MS/MS spectra of OHMT (Figure 3). In fact, these two ions were the most abundant ions in its MS/MS spectrum. However, these ions were either absent or of very low abundance in the MS/MS spectra of the 3 hydroxylated MT standards (Figure 4).

Williams et al.15 extensively studied the abundance of m/z 97 and 109 ions in monohydroxylated testosterones. According to their study, the hydroxylation of testosterone at C-2, C-6, C-7, C11, and C-19 positions greatly reduces the relative abundances of the m/z 97 and 109 ions. Therefore, the presence of intense fragment ions at m/z 97 and 109 in the MS/MS spectra of OHMT provides strong support to rule out hydroxylation at C-2, C-6, C-7, C-11, and C-19 positions. In an attempt to characterize the aglycone OHMT further, the glucuronide extract from the anion exchange SPE of bile was successfully hydrolyzed by β-glucuronidase enzyme from E. coli in 0.25 M phosphate buffer (pH ∼6.8) to produce hydroxylated MT analogs. One major and a few other minor isomeric hydroxylated MT analogs were detected in the hydrolysis reaction (Figure 5). They were less polar, and their retention times did not match with that of commercially available hydroxy MT standards C-6α, C-6β, and C-11α (Figure 5). Therefore, the additional hydroxyl group cannot be on either C-6 or C-11α of the OHMT. Moreover, hydroxylation of methyl groups at C-17, C-18, and C-19 can also be ruled out because of the absence of a characteristic 30 Da (formaldehyde) loss in the spectrum of OHMT.14,16 As reported previously, 17-alkyl steroids (MT, ethisterone, oxymestrone, and mibolerone) show a neutral loss of 58 Da corresponding to a loss of acetone molecule due to E



Article

Scheme 2. Proposed Alternative Mechanism for the Formation of m/z 225

moiety, which generates ambiguity about positioning of the glucuronide in OHMT-glu. To confirm the structure of OHMTglu we need to compare the mass spectrometric and LC data (retention time) with a standard. Unfortunately, there are no commercial standards available for OHMT-glu or OH-MT for comparison purposes. The multistep procedure to synthesize OHMT-glu was beyond the scope of our study. We were also limited by the minute amount of OHMT-glu present in the bile to purify sufficient quantity to get NMR data to support the structure elucidation. However, by a careful study of the steroid fragmentation pathways, we were able to propose a tentative structure for OHMT-glu to best fit all our experimental data. This is a useful strategy in cases in which a chemical standard is not available. Typical extracted ion chromatograms of the parent ion at m/z 495.2589 for [M + H]+ in bile of dosed fish over a 14 day period are shown in Figure 6. All bile samples from dosed fish contained OHMT-glu. The control bile was confirmed negative. A rough estimate of the OHMT-glu concentrations in bile as compared to a MT-glu standard over 14 days are reported in Table 1. The previously reported MT-Glu is not detected 7 days after administration.9 However, the OHMT-glu stays in the bile for at

fragmentation of the D ring. Also, it has been stated that this neutral loss can be seen after losing a water molecule from the molecular ion.14 According to the proposed mechanism (Schemes 1 and 2), the molecular ion at m/z 319 loses a water molecule to generate an ion at m/z 301 followed by a loss of a neutral acetone molecule to generate the ion at m/z 243. On the basis of this mechanism, C-16 cannot be substituted with a hydroxyl group because we see a significant D ring fragment at m/z 243 in OHMT. A careful in-depth analysis and comparison of our mass spectrometric data against published literature has enabled us to eliminate hydroxylation of most of the carbons in OHMT except for the 15 and 12 positions of the nucleus. As postulated in the proposed pathways (Schemes 1 and 2), formation of the two ions we observe at m/z 243 and 225 can be better explained by locating the new hydroxyl group at C-12 than at C-15 in OHMT. Similar mechanisms have been proposed by Pozo et al. in a study to elucidate urinary metabolites of fluoxymesterone.14 The m/z 225 ion can originate from m/z 243 (Scheme 1) or the ion at m/z 283 (Scheme 2). Therefore, we are proposing C-12 as the most probable position for the hydroxyl group in OHMT (Figure 1). The glucuronidation can be at C-17 or on the new hydroxyl F



Article

Figure 6. Extracted ion (m/z 495.2579) chromatograms of OHMT-glu in bile of control and dosed fish. Note that the y-scales for each extracted ion chromatogram are different.

Notes

Table 1. Levels of OHMT-glu in Dosed Tilapia days after dosing

concentration of OHMT-glu (μg/mL)

1 3 7 14

2.2 2.7 0.2 0.07

The authors declare no competing financial interest.

■

(1) Gale, W. L.; Fitzpatrick, M. S.; Lucero, M.; Contreras-Sanchez, W. M.; Schreck, C. B. Masculinization of Nile tilapia (Oreochromis niloticus) by immersion in androgens. Aquaculture 1999, 178 (3−4), 349−357. (2) Kuwaye, T. T.; Okimoto, D. K.; Shimoda, S. K.; Howerton, R. D.; Lin, H. R.; Pang, P. T. K.; Grau, E. G. Effect of 17α-methyltestosterone on the growth of the euryhaline tilapia, Oreochromis mossambicus, in fresh water and in sea water. Aquaculture 1993, 113 (1−2), 137−152. (3) Pandian, T. J.; Sheela, S. G. Hormonal induction of sex reversal in fish. Aquaculture 1995, 138 (1−4), 1−22. (4) Cravedi, J. P.; Delous, G.; Rao, D. Disposition and Elimination Routes of 17-α-Methyltestosterone in Rainbow Trout (Salmo gairdneri). Can. J. Fish. Aquat. Sci. 1989, 46, 159−165. (5) Stanley, S. M. R.; Smith, L.; Rodgers, J. P. Biotransformation of 17alkylsteroids in the equine: gas chromatographic-mass spectral identification of ten intermediate metabolites of methyltestosterone. J. Chromatogr., Biomed. Appl. 1997, 690, 55−64. (6) Dumasis, M. C. In vivo biotransformation of 17-α-Methyltestosterone in the horse revisited: identification of 17- hydroxymethyl metabolites in equine urine by capillary gas chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2003, 17, 320−329. (7) Clarke, D. J.; George, S. G.; Burchell, B. Glucuronidation in fish. Aquat. Toxicol. 1991, 20 (1−2), 35−56.

least up to 14 days after administration. Because it persists in bile longer, OHMT-Glu is a better metabolite to monitor in bile than the previously reported MT-glu. MT is not approved for use in aquaculture in the United States. Therefore, this method can be used for monitoring its illegal use and for surveillance purposes. In this paper we have attempted to show the strength of HRMS in providing the exact mass of OHMT-glu with high mass accuracy to arrive at the correct molecular formula for structure elucidation purposes. In addition, utilization of retrospective analysis of data generated in the past to discover other unknown metabolites is also highlighted.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: 240 402 6688. Fax (240) 264-8401. E-mail: hiranthi. [email protected]. G



Article

(8) Hintikka, L.; Kuuranne, T.; Leinonen, A.; Thevis, M.; Schanzer, W.; Halket, J.; Cowan, D.; Grosse, J.; Hemmersbach, P.; Nielen, M. W. F.; Kostiainen, R. Liquid chromatographic-mass spectrometric analysis of glucuronide-conjugated anabolic steroid metabolites: method validation and interlaboratory comparison. J. Mass Spectrom. 2008, 43 (7), 965−973. (9) Amarasinghe, K.; Chu, P. S.; Evans, E.; Reimschuessel, R.; Hasbrouck, N.; Jayasuriya, H. Development of a Fast Screening and Confirmatory Method by Liquid Chromatography−Quadrupole-Timeof-Flight Mass Spectrometry for Glucuronide-Conjugated Methyltestosterone Metabolite in Tilapia. J. Agric. Food Chem. 2012, 60 (20), 5084−5088. (10) Guan, F.; Soma, L. R.; Luo, Y.; Uboh, C. E.; Peterman, S. Collision-Induced Dissociation pathways of Anabolic Steroids by Electrospray Ionization Tandem Mass Spectrometry. J. Am. Soc. Mass Spectrom. 2005, 17, 477−489. (11) Biddle, S. T. B.; O’Donnell, A.; Houghton, E.; Creaser, C. Metabolism of methyltestosterone in the greyhound. Rapid Commun. Mass Spectrom. 2009, 23, 713−721. (12) Gauthier, J.; Goudreault, D.; Poirier, D.; Ayotte, C. Identification of drostanolone and 17-methyldrostanolone metabolites produced by cryopreserved human hepatocytes. Steroids 2009, 74, 306−314. (13) Musharraf, S. G.; Ali, A.; Khan, N. T.; Yousuf, M.; Choudhary, M. I.; Rahman, A. Tandem mass spectrometry approach for the investigation of the steroidal metabolism: Structure-fragmentaion relationship (SFR) in anabolic steroids and their metabolites by ESIMS/MS analysis. Steroids 2013, 78, 171−181. (14) Pozo, O. J.; Van, W.; Deventer, K.; Van, E.; Delbeke, F. T. Elucidation of urinary metabolites of fluoxymesterone by liquid chromatography−tandem mass spectrometry and gas chromatography−mass spectrometry. J. Mass Spectrom. 2008, 43 (3), 394−408. (15) Williams, T. M.; Kind, A. J.; Houghton, E.; Hill, D. W. Electrospray collision-induced dissociation of testosterone and testosterone hydroxy analogs. J. Mass Spectrom. 1999, 34, 206−216. (16) Pozo, O. J.; Van, E.; Deventer, K.; Hernandez, F.; Delbeke, F. T. Detection and characterization of anabolic steroids in doping analysis by LC-MS. TrAC, Trends Anal. Chem. 2008, 27 (8), 657−671.

H


Tentative Structural Assignment of a Glucuronide Metabolite of

Recommend Documents