New Insights into the Metabolism of Methyltestosterone and Metandienone: Detection of Novel A-Ring Reduced Metabolites

Metandienone and methyltestosterone are orally active anabolic-androgenic steroids with a 17α-methyl structure that are prohibited in sports but are frequently detected in anti-doping analysis. Following the previously reported detection of long-term metabolites with a 17ξ-hydroxymethyl-17ξ-methyl-18-nor-5ξ-androst-13-en-3ξ-ol structure in the chlorinated metandienone analog dehydrochloromethyltestosterone (“oral turinabol”), in this study we investigated the formation of similar metabolites of metandienone and 17α-methyltestosterone with a rearranged D-ring and a fully reduced A-ring. Using a semi-targeted approach including the synthesis of reference compounds, two diastereomeric substances, viz. 17α-hydroxymethyl-17β-methyl-18-nor-5β-androst-13-en-3α-ol and its 5α-analog, were identified following an administration of methyltestosterone. In post-administration urines of metandienone, only the 5β-metabolite was detected. Additionally, 3α,5β-tetrahydro-epi-methyltestosterone was identified in the urines of both administrations besides the classical metabolites included in the screening procedures. Besides their applicability for anti-doping analysis, the results provide new insights into the metabolism of 17α-methyl steroids with respect to the order of reductions in the A-ring, the participation of different enzymes, and alterations to the D-ring.

Due to the 17α-methyl group, the steroids become orally active, because it prevents the first-pass metabolism by hindering the oxidation of the 17β-hydroxy group sterically, while the introduction of a double bond in position 1 was intended to avoid aromatization and reduced the activity of A-ring-reducing enzymes [5][6][7].
Due to the 17α-methyl group, the steroids become orally active, because it prevents the first-pass metabolism by hindering the oxidation of the 17β-hydroxy group sterically, while the introduction of a double bond in position 1 was intended to avoid aromatization and reduced the activity of A-ring-reducing enzymes [5][6][7].
For other steroids with a similar structure, such as dehydrochloromethyltestosterone, there is a metabolite described with a fully reduced A-ring and a rearranged D-ring [30], which was synthesized in 2018 [31,32]. This metabolite led to an extended detection time of the intake for this substance and thereby increased the number of adverse analytical findings.
As the chemical structures of metandienone (12) and methyltestosterone (18) are similar to dehydrochloromethyltestosterone, it is conceivable that intake of these substances results in metabolites with a related structure. The discovery of such new metabolites may help in extending the time of detection after the intake of metandienone (12) or methyltestosterone (18), which would be a considerable contribution to the fight against doping, as cheating may be traced back over a longer period. Additionally, such findings may help to further elucidate the metabolism of synthetic steroids and therefore improve the understanding of human biotransformation. For other steroids with a similar structure, such as dehydrochloromethyltestosterone, there is a metabolite described with a fully reduced A-ring and a rearranged D-ring [30], which was synthesized in 2018 [31,32]. This metabolite led to an extended detection time of the intake for this substance and thereby increased the number of adverse analytical findings.
As the chemical structures of metandienone (12) and methyltestosterone (18) are similar to dehydrochloromethyltestosterone, it is conceivable that intake of these substances results in metabolites with a related structure. The discovery of such new metabolites may help in extending the time of detection after the intake of metandienone (12) or methyltestosterone (18), which would be a considerable contribution to the fight against doping, as cheating may be traced back over a longer period. Additionally, such findings may help to further elucidate the metabolism of synthetic steroids and therefore improve the understanding of human biotransformation.

Post-Administration Urines
Urine samples from the administration trials were analyzed with a GC-QTOF-MS and GC-QQQ-MS after per-TMS derivatization.
The common metabolites of MT (18) and MD (12) were detected by comparison of retention time and quantifier and qualifier transitions, as reported in Table 3. Corresponding chromatograms are available as supplemental material (S3). Table 3. Retention times (GC-QQQ-MS) and ion transitions of currently targeted metabolites in anti-doping analysis.

Post-Administration Urines
Urine samples from the administration trials were analyzed with a GC-QTOF-MS and GC-QQQ-MS after per-TMS derivatization.
The common metabolites of MT (18) and MD (12) were detected by comparison of retention time and quantifier and qualifier transitions, as reported in Table 3. Corresponding chromatograms are available as supplemental material (S3).
The 3α,5β-epi-tetrahydromethyltestosterone was identified as the first peak in positive urine samples of metandienone and methyltestosterone at 9.56 min ( Figure 8).

Urinary Metabolites
As is common in several doping control laboratories, glucuronidated metabolites are enzymatically cleaved and determined as their aglycons together with their analogs that are excreted as unconjugated compounds. Due to the low abundance of some of the target analytes, GC-QQQ-MS in MRM mode is considered as a better-suited technique for metabolite detection after optimization of the ion transitions. As described in the literature [8,24,38], GC-QQQ-MS analysis detected 17α-methyl-5β-androstane-3α,17β-diol (20, MT M1) following the administration of both steroids, MD (12) and MT (18), in all samples. Its 3α,5α-analog (19, MT M2) was detected following the administration of MT (18), while in MD (12) p.a. samples, only very minor corresponding signals were detectable in the 48 h urine and remained unconfirmed due to the low signal-to noise ratio of the qualifier transitions. According to earlier studies, these two metabolites are considered as longest detectable by GC-QQQ-MS after MT (18) administration in GC-MS [38].
Interestingly, the structure of the long-term metabolite of 4-chlorometandienone with modified D-ring structure and a fully reduced A-ring (Sobolevsky's "M3") was assigned to 4α-chloro-17β-hydroxymethyl-17α-methyl-18-nor-5α-androst-13-en-3α-ol by Forsdahl et al. [31]. The metabolites proposed for MT and MD as described above show an inverse stereochemistry at the D-ring in comparison to these assignments.
Additionally, the product of the last synthesis (17β-methyl-5β-androstane-3α,17α-diol, 11) has an inverse D-ring at C17 in comparison to the parent compounds and the fully A-ring reduced metabolites, 17α-methyl-5β-androstane-3α,17β-diol (20, MT M1) and 17αmethyl-5α-androstane-3α,17β-diol (19, MT M2). The latter are formed through reduction of the 1,2-and 4,5-double bond and the 3-oxo group. The 17-epimer was found in the urines after the intake of both mentioned anabolic-androgenic steroids. In the case of MT administration, the metabolite 11 was also described earlier, but found with shorter detection times than the 17α-methyl analogs 19 and 20 [38]. After the intake of MD, this was also found earlier, but with a problem in separation of the four diastereomers [15].
The stereoselectivity of the A-ring reduction is dependent on the parent compound. For metandienone, there is only very limited generation of metabolites with a 5α-structure. This is is likely due to the 1,2-double bond, which inhibits the activity of 5α-reductase [45]. In contrast, methyltestosterone is metabolized to 5α-and 5β-isomers. This substantiates our hypothesis of metabolite generation due to the A-ring structure with a double bond in position 4 and its already saturated positions 1 and 2 in methyltestosterone, while MD (12) has an unsaturated A-ring (i.e., 3-oxo-1,4-diene). Thus, it is reasonable that the 17α-hydroxymethyl-17β-methyl-18-nor-5α-androst-13-en-3α-ol-derivative (8a) is only detectable in p.a. samples of methyltestosterone (18), while the 5β-analog (8) is observed after MT or MD administration. This supports our concept of the order of reductions: if the 1,2-double bond was reduced before the 4,5-double bond, there would have also been 5α-metabolites in p.a. urines of metandienone [8].
Thus, the order of the following two reductions of metandienone (1,2-double bond, 3-oxo group) is not yet confirmed, but it seems to be more likely that the formation of the 3-hydroxy group takes place before the hydrogenation of the 1,2-double bond, because there are known metabolites of metandienone with a 3-hydroxy-1-ene structure but not with a 3-oxo group in a fully reduced A-ring. Both potential ways represent the last step of the proposed formation of the metabolites 8 and 11. They are displayed in Figure 10. The other reactions of the metabolism of both investigated compounds are displayed in Figure 11.

3-oxo group)
is not yet confirmed, but it seems to be more likely that the formation of the 3-hydroxy group takes place before the hydrogenation of the 1,2-double bond, because there are known metabolites of metandienone with a 3-hydroxy-1-ene structure but not with a 3-oxo group in a fully reduced A-ring. Both potential ways represent the last step of the proposed formation of the metabolites 8 and 11. They are displayed in Figure 10. The other reactions of the metabolism of both investigated compounds are displayed in Figure 11. Figure 10. Potential ways of A-ring reduction. Figure 10. Potential ways of A-ring reduction.
Based on preliminary data, the mentioned substances are detected for at least 48 h after the intake of parent compounds. Excretion studies with a higher number of volunteers and prolonged sample collection will be performed in the near future to evaluate the detection windows of the new metabolites.
In addition to that, a potential next step will be the investigation of the substrate specificity of 5α-reductase towards 1,2-ene steroids by means of molecular modelling to elucidate structural requirements for generation of 5α-metabolites of androgenic steroids.
The detection and structure identification of the above-mentioned substances in the urine samples help to gain further insights into human metabolism of metandienone and 17α-methyltestosterone. Due to the similarity of other anabolic androgenic steroids to the investigated compounds, it is probable that other metabolites with related structures may be found in further 17α-methyl steroids. Finally, the results may support the fight against doping by introducing new analytes for screening in anti-doping analysis. ???
Based on preliminary data, the mentioned substances are detected for at least 48 h after the intake of parent compounds. Excretion studies with a higher number of volunteers and prolonged sample collection will be performed in the near future to evaluate the detection windows of the new metabolites.
In addition to that, a potential next step will be the investigation of the substrate specificity of 5α-reductase towards 1,2-ene steroids by means of molecular modelling to elucidate structural requirements for generation of 5α-metabolites of androgenic steroids.
The detection and structure identification of the above-mentioned substances in the urine samples help to gain further insights into human metabolism of metandienone and 17α-methyltestosterone. Due to the similarity of other anabolic androgenic steroids to the investigated compounds, it is probable that other metabolites with related structures may be found in further 17α-methyl steroids. Finally, the results may support the fight against doping by introducing new analytes for screening in anti-doping analysis.

GC-QTOF-MS
High resolution accurate mass analyses were performed on an Agilent GC-QToF 7890B/7250 (Agilent Technologies, Milano, Italy), equipped with an Agilent HP1 column (17 m, 0.20 mm; 0.11 µm) with helium as carrier gas. Injection was performed in split mode with a 1:10 ratio at 280 • C. The oven program had the following heating rates: 188 • C hold for 2.5 min, 3 • C/min to 211 • C and hold for 2 min, 10 • C/min to 238 • C, 40 • C/min to 320 • C, and hold for 3.2 min. The coupled QToF was operated in full scan with an ionization energy of 70 eV. Aberrantly, in LEI an ionization energy of 15 eV was applied. Ions were detected from m/z 50 to 750.

HPLC Purification
The purification of the synthesized reference steroids was performed by semi-preparative HPLC using an Agilent 1260 Infinity Quaternary HPLC system coupled to an Agilent Infinity 1260 diode array detector (Agilent Technologies GmbH, Waldbronn, Germany). Chromatographic separation was achieved on a Hypersil ODS C18 column (pore size: 120 Å, 250 mm length, 10 mm ID, 5 µm particle size, Thermo Scientific, Schwerte, Germany). Isocratic elution was accomplished at a flow rate of 3 mL/min using acetonitrile:water (7:3, v:v) as the mobile phase. The UV signal was monitored at 194 nm.

Nuclear Magnetic Resonance
The nuclear magnetic resonance (NMR) analyses were performed at 500 MHz ( 1 H NMR) and 125 MHz ( 13 C NMR) at 296 K on a Bruker (Rheinstetten, Germany) Avance III instrument equipped with a nitrogen-cooled 5 mm inverse TCI cryoprobe with actively shielded z-gradient coil. Chemical shifts are reported in δ values (ppm) relative to tetramethylsilane. Solutions of about 5 mg of each compound in deuterated dimethylsulfoxide (d 6 -DMSO) were used for conducting 1 H; H,H COSY; 13 C; edited HSQC; HMBC, selective NOE and NOESY experiments. Two-dimensional experiments were recorded in non-uniform sampling (NUS) mode.
Spiro[5 ξ-androstan-17,2 -oxirane]-3ξ-ol (7, 7a) The crude substance was dissolved in dichloromethane, and potassium hydrogen carbonate and meta-chloroperoxybenzoic acid were added. The solution was stirred for 3 h at ambient temperature. Afterwards, the mixture was poured into water and extracted three times with dichloromethane. The organic phases were washed with brine and then dried over sodium sulfate. Further details are disclosed as supplemental information.

Epi-Tetrahydromethyltestosterones
17β-Methyl-5β-androstane-3α,17α-diol (11) A mixture of 450 µL methanol and 50 µL potassium hydroxide solution (5 M) was prepared, and 100 µg epi-methyltestosterone (9) was dissolved. A spatula tip of palladium on charcoal was added, and hydrogen gas flushed through the solution for 5 min. After adding 2 mL of water, the mixture was extracted three times with 3 mL of hexane and evaporated to give the product 10. The residue was dissolved in methanol/water (9:1, v:v) and a spatula tip of sodium borohydride was added. The solution was stirred for one hour at room temperature. After adding ammonium chloride to stop the reaction, potassium hydroxide solution (1 M) was added to yield alkaline solution. Then, the solution was extracted three times with dichloromethane and evaporated to give the product 11.
17β-Methyl-5α-androstane-3α,17α-diol (11a) A spatula tip of epi-mestanolon (10a) was dissolved in 2 mL of absolute THF, 80 µL of K-Selectride was added and the mixture was stirred for 1 h at ambient temperature. Afterwards, 100 µL of aqueous hydrochloric acid (1 M) was added until there was no formation of bubbles anymore. Then, 150 µL of potassium hydroxide solution (1 M) was added and the mixture was extracted three times with 5 mL of hexane. The hexane-phase was evaporated to give the product 11a.

Human Administration Trial
Urine samples out of the stock of the anti-doping laboratory in Rome were available for analysis. Samples collected before and after an oral administration of either MD or MT were used for evaluation of the excretion of the hypothized metabolites. The excretion study with MT was carried out by a healthy male volunteer (Caucasian, 50 years old, 80 kg and normal body mass index). A single oral dose of 10 mg of MT (Metadren ® , Novartis, Basel, Switzerland) was administered. For investigation of MD metabolism, a single oral dose of 5 mg MD (Dianabol ® , Ciba-Geigy, Basel, Switzerland) was administered to a healthy male volunteer (Caucasian, 45 years old, 82 kg and normal body mass index).

Urine Sample Preparation
An aliquot of 6 mL urine was used for the following analysis. As internal standard methyltestosterone (50 µL of a solution of 100 µg/mL) was added. After the addition of 750 µL of phosphate buffer (0.8 M) and 50 µL β-glucuronidase, the mixture was incubated at 55 • C for 60 min. Afterwards, 500 µL of carbonate buffer (20%) was added and the mixture was extracted with 10 mL of TBME. After evaporation, 50 µL of TMIS reagent was added to the sample and the mixture was treated at 75 • C for 20 min before analysis to generate the per-TMS derivatives.
Supplementary Materials: The following are available online: Supplement S1: Table of  Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the School of Pharmaceutical Science and Technology, Tianjin University (date of approval 9 December 2020).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Raw data are stored at the authors.