Physiological processes involving metabolism and behavior, e.g., activity/rest, are generally organized on a cycle of approximately 24 h driven by a circadian rhythm [1
]. The nuclear receptors reversed-viral erythroblastosis α and β (REV-ERB α and β) regulate the expression of core clock proteins and therefore help to modulate the circadian rhythm [1
Modulation of the REV-ERB activity by synthetic agonists, e.g., SR9009 and SR9011 (Figure 1
), alters the expression of genes involved in lipid and glucose metabolism and, therefore, plays an important role in maintaining the energy homeostasis [1
]. Effects of SR9009 and SR9011 observed via in vitro and in vivo animal studies were increased basal oxygen consumption, decreased lipogenesis, cholesterol and bile acid synthesis in the liver, increased mitochondrial content, glucose and fatty acid oxidation in the skeletal muscle and decreased lipid storage in the white adipose tissue [1
The observed increase in energy expenditure and decrease in fat mass make the REV-ERB agonists SR9009 and SR9011 promising drug candidates for the treatment of several metabolic disorders [3
]. At the same time, the increase in exercise capacity observed via in vivo animal studies [6
] makes these compounds also attractive for performance enhancement by athletes. Such use can be classified as doping. The potential interest as doping agents is clearly shown by their popularity in discussion forums on the Internet, where they are mentioned as the ultimate “exercise in a pill” compounds [8
]. Although these REV-ERB agonists are currently still undergoing clinical evaluation and are therefore not approved for therapeutic use, distribution in black market products might be expected as observed before for designer steroids [15
], peptides [18
], several selective androgen receptor modulators (SARMs) [24
] and peroxisome proliferator activated receptor δ (PPARδ) agonists e.g., GW501516 [26
Even though not explicitly mentioned on the Prohibited List published by the World Anti-Doping Agency (WADA), they are indirectly prohibited as non-approved substances (Class S0), but could potentially also be classified as metabolic modulators (Class S4) [27
As illicit use of SR9009 and SR9011 can be anticipated, monitoring of their presence on the market and use by doping control laboratories is recommended. These preventive investigations not only help to close the gap between anti-doping laboratories and the appearance of new doping agents, but also contribute to deter the use of these compounds. It would protect fair play and the health of athletes, as athletes are deterred from use when a substance is detectable.
To allow a fast response to the appearance of new non-approved performance-enhancing substances, in vitro metabolic studies are frequently applied. As the liver is the principal organ for drug metabolism, in vitro models are often based on human liver fractions (e.g., human liver microsomes (HLM)) [28
]. These in vitro studies not only circumvent the ethical objections related to the use of human volunteers for excretion studies, they are more affordable and can be applied rapidly. Moreover, analytically, the clean extracts improve the characterization of metabolites [29
]. However, careful extrapolation of these in vitro studies to real human metabolism should be performed as some metabolic pathways may be over- or under-expressed [28
]. Nevertheless, in vitro metabolic studies allow incorporating metabolites into existing screening methods, which could improve the detection window compared to the parent compound, and they can provide reference material to improve the identification of suspicious doping control samples [29
In the current study, a black market product sold as a performance-enhancing product and labeled to contain SR9009 was purchased over the Internet to verify its content. The analysis resulted in the identification of the mentioned compound by liquid chromatography–(high resolution) mass spectrometry (LC–(HR)MS). Consecutively, HLM were applied to perform metabolic studies of SR9009 and SR9011.
The presence of SR9009 in the black market product indicates that the use of SR9009 is no longer a potential threat, but a real doping threat. Shortly after the purchase of the substance from the gross sale Internet market, the product also appeared for purchase in individual quantities from several Internet suppliers of performance-enhancing substances. This indicates that the substance has now become readily available to athletes and that detection methods need to be developed for the product. Metabolism studies are essential to improve detection (windows) of doping agents. Ideally, human excretion studies can be performed to identify diagnostic metabolites. In case of non-pharmaceutical substances, e.g., SR9009 and SR9011, ethical objections and safety aspects limit the use of human volunteers. Consequently, in vitro metabolic assays (HLM) were applied to perform metabolic studies of SR9009 and SR9011.
Eight metabolites (SR09-1–SR09-8) and fourteen metabolites (SR11-1–SR11-14) were detected in vitro for SR9009 and SR9011 by LC–HRMS. For the characterization of the metabolites, application of high resolution instruments can be advantageous, e.g., the metabolic modifications could be verified by determining the mass deviations of the proposed chemical formulas. To further characterize the proposed structures of the metabolites, LC–HRMS product ion scans were performed. Initially, the parent compounds were studied to determine typical, diagnostic ions. The presence of these diagnostic or modified diagnostic fragment ions can enable the identification of the positions of the metabolic modifications. This metabolite identification is time consuming, but HRMS data can also facilitate the correlation of observed product ions to structure specific fragments of the compounds. Tentative structures of these metabolites, based on LC–HRMS product ion scan data, are presented in Figure 6
. Although modifications were observed in all fragments (A, B, B’ and C), most modifications occurred in the B/B’ fragment of the parent compounds. However, further research will be needed to unequivocally identify the chemical structures of the metabolites.
Similar metabolites for SR9009 were described in the HLM incubation and human excretion urine samples by Sobolevsky et al. [34
]. Metabolites SR09-1 and SR09-6 were not reported in their study. However, two additional metabolites were detected in their study; one metabolite characterized by a loss of D and hydroxylation and another by a loss of C [34
]. This latter metabolite was also described as a major human metabolite [34
]. Considering the structures of SR09-6 (−C+OH) and SR09-7 (−D), these metabolic modifications observed by Sobolevsky et al. seem possible in our in vitro incubation samples [34
Sobolevsky et al. also reported similar metabolites for SR9011, but metabolites SR11-2, SR11-3 and SR11-12 were not reported in their study [34
]. Two additional metabolites were detected in their study, characterized by loss of C or A [34
], with the latter described as a major urinary metabolite [34
]. Since hydroxylated metabolites (SR11-5 and SR11-7), after loss of A or C, were detected, these metabolic modifications observed by Sobolevsky et al. seem also possible in our in vitro incubation samples [34
The highest relative abundances in the HLM incubation samples with SR9009 were observed for metabolites SR09-1, SR09-2, SR09-4(a), SR09-5 and SR09-7. For SR9011, metabolites SR11-1, SR11-3, SR11-5 and SR11-9 have the highest relative abundances in vitro. Therefore, incorporation of metabolites SR09-1, SR09-2, SR09-4(a), SR09-5, SR09-7, SR11-1, SR11-3, SR11-5 and SR11-9 might improve screening for misuse of SR9009 and SR9011. However, it should be noted that this assumption is only based on the relative abundances of the metabolites detected in the HLM incubation samples and that extrapolation from these in vitro studies to the more complex human situation is difficult.
Nevertheless, in the human excretion studies with SR9009 and SR9011 of Sobolevsky et al., metabolites SR09-5 and SR11-5 were indeed described as major metabolites of SR9009 and SR9011, respectively [34
]. In addition, metabolites SR11-6 and SR11-13 were described as major human metabolites of SR9011 [34
The lack of an intact biological system can hamper the extrapolation of in vitro results to the human situation. When applying in vitro techniques, other pharmacokinetic processes, including absorption, distribution and excretion, are not covered. Whereas human administration studies represent a higher complexity, other factors, e.g., phase II metabolism and extrahepatic sites of metabolism, can influence metabolic clearance. Furthermore, HLM consists of enriched drug metabolizing enzymes, which restricts competition for other enzymes and limits the application of HLM for quantitative estimations of specific metabolic pathways.
However, the in vitro approach offers some advantages, such as the production of cleaner and more concentrated extracts, which facilitates the characterization of metabolites, and they allow a fast response to potential new threats. Therefore, HLM can be considered as a valuable ethically acceptable alternative for human metabolic studies of non-pharmaceutical-grade substances.
To circumvent the low extraction recoveries observed when applying a regularly-used sample preparation procedure (enzymatic hydrolysis and LLE at pH 9.5), the combination with a dilute-and-shoot method is recommended.
The presence of SR9009 and SR9011 and their metabolites was verified by retrospective data analysis in 1511 doping control samples. Although misuse of SR9009 and SR9011 could not be demonstrated in our study, it is hoped that other WADA-accredited laboratories will perform a similar retrospective analysis. This would further close the gap between anti-doping laboratories and doped athletes.
4. Materials and Methods
4.1. Chemicals and Reagents
A black market gross sales product, claiming to contain SR9009, was purchased over the Internet (to prevent athletes from purchasing this product, further details remain confidential). The internal standard (IS) 17α-methyltestosterone was obtained from Organon (OSS, Noord-Brabant, The Netherlands). 4-Hydroxytamoxifen-d5 was purchased from Toronto Research Chemicals (TRC, Toronto, ON, Canada). The reference standard of methandienone was obtained from the National Measurement Institute (NMI, Sydney, North Ryde, Australia). Reference material of SR9009 and SR9011 was purchased from Calbiochem (Merck Chemicals, Nottingham, UK) and Xcess Bio (San Diego, CA, USA), respectively. Pooled HLM from 20–30 donors, the nicotinamide adenine dinucleotide phosphate (NADPH) regenerating system Solutions A and B and phosphate buffer pH 7.4 all from Gentest were purchased from Corning (Amsterdam, The Netherlands). β-Glucuronidase/arylsulfatase from Helix pomatia was from Roche Diagnostics (Mannheim, Germany). Ethanol and ammonium acetate (NH4OAc) were purchased from Biosolve (Valkenswaard, The Netherlands). Formic acid (HCOOH), ammonium formate (NH4OOCH) and methanol (MeOH) were obtained from Fisher Scientific (Loughborough, UK). tert-Butyl methyl ether (TBME) was purchased from Macron-Avantor (Deventer, The Netherlands). Sodium acetate (NaOAc), sodium hydrogen carbonate (NaHCO3), potassium carbonate (K2CO3), ammonium iodide (NH4I) and acetic acid (HOAc) were from Merck (Darmstadt, Germany). LC-grade water, LC-grade MeOH and LC-grade acetonitrile (ACN) were purchased from J.T. Baker (Deventer, The Netherlands). Nitrogen (N2) and oxygen-free nitrogen (OFN) was delivered by Air Liquide (Bornem, Belgium).
Liquid Chromatography–(High Resolution) Mass Spectrometry (LC–(HR)MS)
The same conditions were applied for all liquid chromatography (LC) experiments using a Thermo Finnigan Surveyor Autosampler Plus and an MS Pump Plus (Thermo Scientific, Bremen, Germany). A SunFire™ C18 column (50 mm × 2.1 mm i.d., 3.5 μm) from Waters (AH Etten-Leur, The Netherlands) was applied for the LC separation at a flow rate of 250 μL/min. A volume of 25 μL was injected using the no waste injection mode. LC-grade water (Solvent A) and LC-grade MeOH (Solvent B) both with 1 mM NH4OAc and 0.1% HOAc were used as the mobile phase. In the gradient program, the percentage of Solvent B was linearly changed as follows: 0 min, 15%; 1 min, 15%; 6.5 min, 70%; 14 min, 75%; 16.0 min, 100%; 16.9 min, 100%; 17 min, 15% and 20 min, 15%. The total run time was 20 min for the LC methods.
The low resolution experiments were performed using a TSQ Quantum Discovery MAX triple quadrupole mass spectrometer (Thermo Scientific). For these experiments, a full-scan method was applied with a scan range of m/z 100–500 in both positive and negative mode. The other MS conditions included interface: electrospray ionization (ESI); capillary voltage: 3.5 kV; source temperature: 350 °C; sheath gas (N2) pressure: 50 (arbitrary units); auxiliary gas (N2) pressure: 20 (arbitrary units); tube lens offset: 100 V; scan time: 0.5 s.
An Exactive mass spectrometer (Thermo Scientific) was applied for the high resolution-tandem mass spectrometry (HR–MS(/MS)) experiments. The instrument operated in a full-scan (both positive and negative) mode with a scan range of m
100–2000 at a resolving power of 50,000 and a data acquisition rate of 2 Hz. LC–HRMS product ion scans were performed by a Q-Exactive mass spectrometer (Thermo Scientific) for the structural investigation of metabolites. Therefore, the protonated and deprotonated molecules were selected as precursor ions with an isolation window of 1.0 m
at a resolving power of 70,000, and collision energies of 15, 25, 35 and 45 eV were applied. For both LC–HRMS instruments, the other MS parameters were identical to the low resolution instruments, except for spray voltage: 4 kV, source temperature of 250 °C and heated ESI (HESI) (probe heater at 300 °C). For the assay validation, an LC–HRMS screening method, as indicated in [33
], was applied.
4.3. In Vitro Incubation Studies
Prior to the in vitro metabolic studies, the black market product containing SR9009 and available reference material of SR9011 and SR9009 were analyzed by LC–(HR)MS for purity verification.
The reaction mixtures for the in vitro metabolic assays (phase I) consisted of 0.1 M phosphate buffer (pH 7.4), a 1.3 mM NADPH regenerating system and the test compound, at a final concentration of 40 μg/mL, dissolved in ethanol (maximum 1% ethanol). The reaction mixtures were first pre-incubated for 5 min at 37 °C using the Eppendorf thermomixer comfort (Rotselaar, Belgium). The enzymatic reactions were then initiated by adding pooled HLM, to obtain a final protein concentration of 0.5 mg/mL. The final samples (250 μL) were incubated at 37 °C for 2, 4 or 18 h. At the appropriate time, the enzymatic reactions were terminated by adding 250 μL of ice-cold MeOH and transferring the tubes into an ice bath for 15 min.
Control samples were used to verify the enzymatic reactions and the stability of the test compounds. These control samples included system blank samples, which did not contain the test compound, and substrate stability samples (blank), which consist of all of the reaction components, except the microsomal proteins. In a positive control sample, the reference standard of methandienone was incubated with HLM.
4.4. Assay Validation
The qualitative determination of SR9009 and SR9011 was validated in human urine regarding specificity, extraction recovery and the limit of detection (LOD) according to the Eurachem guidelines [35
] and in compliance with the WADA International Standards for Laboratories (ISL) [36
Specificity was tested during the validation procedure by checking for possible interfering peaks in the extracted ion chromatograms (EIC) at the expected retention times for SR9009 and SR9011.
The LOD was defined as the lowest concentration that can be detected in ten human urine samples with a signal to noise ratio (S/N) higher than three. Ten different blank human urine samples (seven male and three female; pH-range from 4.72–6.89; specific gravity between 1.006 and 1.035) were spiked at 2, 5, 10 and 20 ng/mL for SR9009 and at 2, 5, 10, 20 and 50 ng/mL for SR9011. For the assay validation, 4-hydroxytamoxifen-d5 was used as the internal standard. Blank urines and distilled water samples spiked only with this IS were also included. Sample preparation was performed as described in the following section (4.5. Sample Preparation). The samples were analyzed according to an existing LC–HRMS screening method [33
] providing the data necessary to determine the LOD.
The extraction recovery of SR9009 and SR9011 during sample preparation was determined at 20 ng/mL and at 50 ng/mL, respectively. Recovery was calculated by comparison of the mean peak area of the analytes for urine samples spiked before and after sample preparation (hydrolysis/liquid–liquid extraction (LLE), without dilution fraction). Therefore, the results of the ten blank urines applied for the determination of the LOD spiked with SR9009 and SR9011 before sample preparation were used. Another batch of the same blank urines was spiked with SR9009 and SR9011 after sample preparation. Additionally, matrix effects were studied at 20 ng/mL for SR9009 and at 50 ng/mL for SR9011. Matrix effects were measured by comparing the peak area (A) of the analytes spiked in 10 extracted urines and in a neat standard solution following the formula:
4.5. Sample Preparation
The samples of the in vitro metabolic assays were analyzed by direct injection on LC–HRMS, after removal of the enzymatic proteins. Briefly, the HLM incubation samples were first centrifuged at 4 °C (12,000× g, 5 min) followed by transferring 400 μL into new tubes. Fifty microliters of the internal standard (IS) 17α-methyltestosterone (2 μg/mL) were added to all samples.
The sample preparation procedure for the assay validation and the, 1511 previously analyzed, routine doping control samples was described before [33
]. Briefly, a 10-fold dilution fraction of the urine sample (with 1 mM NH4
OOCH/0.01% HCOOH in H2
O/ACN (95/5)) was used to reconstitute the dried extract of this urine sample. This extract was obtained after enzymatic hydrolysis (by β-glucuronidase/arylsulfatase from Helix pomatia
at pH 5.2 and 56 ± 5 °C) and LLE (by TBME at pH 9.5) of the urine sample.