Abiraterone (CAS: 154229-19-3; Cat Number: SA5840) was purchased from Solarbio Life Sciences Co., Ltd. (Beijing, China). Abiraterone N-Oxide (CAS: 2378463-76-2; Cat Number: A-8216) was purchased from TLC Pharmaceutical Standards Ltd. (Ontario, Canada). The purity of the compounds was higher than 98% (HPLC). HPLC-grade acetonitrile, and methanol were purchased from Fisher Scientific (Fair Lawn, NJ, USA). Sprague-Dawley (SD) rat liver microsomes were purchased from RILD Research Institute for Liver Diseases (Shanghai) Co. Ltd (Cat Number: WWJW). Qualitative identification of ABR metabolites in gut microbiota and liver microsomes and structural analysis using the LC/MSⁿ-IT-TOF system from Shimadzu Corporation (Kyoto, Japan). A small refrigerated high-speed centrifuge (Eppendorf Centrifuge 5424 R) was purchased from Eppendorf (Hamburg, Germany). WH-681 vortex mixer was purchased from Jintan Shenglan Instrument Manufacturing Co., Ltd. (Jintan, China).

Six Sprague Dawley (SD) male rats (200-300 g) were purchased from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All animals were housed in a ventilated room with free access to food and water, a 12-h light and 12-h dark cycle. Temperature was maintained at 20–24°C and humidity at 40–60%. Before the experiment, the rats were fasted for 12 h and had free access to water. This study was conducted under the permission and guidance of the Laboratory Animal Ethics Committee of the Chinese Academy of Medical Sciences and Peking Union Medical College (Approval number: 00005407). All steps were performed in accordance with the Organizational Guidelines and Ethical Guidelines of the Laboratory Animal Ethics Committee.

To identify the metabolites of ABR, an LC/MSⁿ-IT-TOF equipped with an ESI source was used. Analytes were separated using Luna C₁₈-HST column (50 × 2 mm, 2.5 µm, Phenomenex, USA). The mobile phase consisted of formic acid: water (0.1:100, v/v) (as mobile phase A) and acetonitrile (as mobile phase B). The temperature of the column oven was 40°C, and the flow rate was 0.4 mL/min. Mass spectrometry conditions: ionization mode: ESI source, analysis mode: positive and negative ion mode, nebulization gas flow: 1.5 L/min, CDL temperature: 200°C, heating block temperature: 200°C, detector voltage: 1.75 KV, collision energy: 50%, drying gas pressure: 115 KPa, mass spectrometry primary data acquisition range: m/z 100~1000, multi-level data acquisition using automatic mode. The elution gradient conditions (A: B) is shown in Table 1 .

The colon contents of six Sprague-Dawley (SD) rats were collected after sacrifice, and sterilized anaerobic medium (Solarbio Life Sciences Co., Ltd. (Beijing, China) was added with an m/v ratio of 1:20 (g/mL), which was mixed evenly and purged with nitrogen after filtering. The mixture was pre incubated under anaerobic conditions at 37°C for 60 minutes. A methanol solution of ABR (1 mg/ml) was prepared, and 10 μL of this solution was added to a presterilized centrifuge tube (the final concentration of ABR in the system was 10 μg/mL), which was mixed with 990 μL of the preincubated mixture under anaerobic conditions. The drug was incubated with the intestinal microbiota at 37°C for 0, 6, 12 and 24 hours. In addition, negative controls containing heat inactivated intestinal microbiota were incubated with ABR for the same time (24 hours). After the incubation, 3-fold volume (3 mL) of pure methanol solution was added to the incubation system and mixed to stop the reaction and precipitate the protein at 0, 6, 12, 24h. After centrifugation using a small refrigerated high-speed centrifuge at 4°C, 13,400 × g rpm for 10 minutes, 100μL was added to a chromatographic autosampler for LC/MSⁿ-IT-TOF analysis.

The liver microsome incubation system consisted of the following: 5 μL of Sprague–Dawley rat liver microsomes (20 mg/mL), 2 μL of ABR (10 μg/mL, final concentration in the system was 1 µmol/mL), 20 μL of NADPH and 173 μL of Tris/HCl (0.05 mM, pH = 7.4) in a total volume of 200 μL. Culture in a shaking incubator at 37°C and 800 rpm with oxygen. After the incubation, 3-fold volume (600 μL) of pure methanol solution was added to the incubation system and mixed to stop the reaction and precipitate the protein at 0, 15, 30, 60, 90, and 120 min. After centrifugation using a small refrigerated high-speed centrifuge at 4°C, 13,400 × g rpm for 10 minutes, 100 μL was added to a chromatographic autosampler for LC/MSⁿ-IT-TOF analysis.

Data acquisition and processing were performed with Shimadzu LC-MS Solution (version 5.89, Kyoto, Japan). Two-tailed ANOVA and Student’s t-test were used for statistical analysis with GraphPad Prism Version 9 for macOS (GraphPad Software, San Diego, CA, USA). Data are expressed as the mean ± standard deviation (SD), and p values less than 0.05 were considered statistically significant.

To explore the metabolism of ABR in liver microsomes, we performed in vitro metabolism experiments using a Sprague-Dawley rat liver microsome incubation system (5 μL of Sprague–Dawley rat liver microsomes + 2 μL of ABR + 20 μL of NADPH + 173 μL of Tris/HCl). The relative abundance of ABR in the Sprague-Dawley rat liver microsome incubation system over time is shown in Figure 2A . Figure 2A shows that ABR could be metabolized by Sprague-Dawley rat liver microsomes within 120 min.

Interestingly, we found four metabolites (M1, M2, m3 and M4) (In this study, M1-M5 are the abbreviations for metabolites 1-5, which are synonyms for the unknown metabolites of abiraterone found in this study. No further explanation below.) in the liver microsomal incubation system. Figures 1C , D shows the EIC diagram of metabolites of ABR liver microsome incubated for 0 and 2h. Four hydroxylated metabolites were mainly produced in the metabolism of liver microparticles in vitro (the relative abundance changes in metabolism are shown in Figure 2 BCDE). The possible chemical structures of M1, M2, M3 and M4 are shown in Figure 4 . The retention times of M1, M2, M3 and M4 are: 8.315, 9.820, 11.045, 12.855 min, respectively. M1~M4 are isomers, ion m/z = 366.2597. The MS information of Abiraterone and its metabolites from the liver microparticle incubation system is shown in Table 2 .

To explore whether the gut microbiota is involved in ABR metabolism, the colonic contents of six Sprague-Dawley (SD) rats were incubated with ABR. At the same time, an incubation system of the colon contents inactivated by heating twice was used as a negative control to eliminate the interference of environmental factors such as the culture medium. After the incubation, 3-fold volume (3 mL) of pure methanol solution was added to the incubation system and mixed to stop the reaction at 0, 6, 12, 24h and prepare samples for detection of ABR content in cultures by LC/MSⁿ-IT-TOF. It can be seen from the Figure 3A that the content of ABR in the in vitro incubation system gradually decreased with time. EIC diagram of ABR and its metabolites metabolized by intestinal flora for 0h and 12h in vitro are shown in Figures 3C, D . In contrast, heat-killed gut flora hardly metabolized ABR. This suggests that the decline in ABR is the result of co-metabolism by the live microbiota, which demonstrates the role of the gut microbiota in ABR metabolism.

To explore the metabolites of ABR in the intestinal flora, LC/MSⁿ-IT-TOF was used to search and infer the structures of the metabolites in the reaction system, and a metabolite(M5) was found (the relative abundance changes in metabolism are shown in Figure 3B . EIC diagram of ABR metabolites metabolized by intestinal flora for 12h in vitro is shown in Figure 3D . Possible chemical structure of M5 is shown in Figure 4F ). The retention time of M5 is 14.763 min. However, they were not detected in the inactivated gut microbiota system, suggesting that M5 may be a metabolite produced by the gut microbiota metabolizing ABR.

Firstly, we analyze the cracking law of ABR:

ABR [M+H]⁺ is 350.2601, and the secondary fragment ion is 334.2308, which may be obtained from the loss of methane (CH4) on the 10th or 13th carbon ( Figure 4A ). However, the tertiary fragment ion 170.0997 may be obtained from the C8-C14 position and the single bond of C9-C11 is broken. It can also be judged that the secondary fragment is obtained by the neutral loss of 18Da at the 13th carbon, and the tertiary fragment ion 316.2213, which may be the 3rd carbon. dehydration of hydroxyl. The tertiary fragment ion 196.1190 is obtained by further fragmentation of the C9-C10 single bond and the C7-C8 single bond by the secondary fragment ion 334.2308. Secondary fragment ion 302.2026 results from the loss of one water and two methyl groups from parent ion 350.2601. The secondary fragment ion 156.0879 is obtained by the cleavage of the C12-C13 single bond and the C8-C14 single bond. The mass spectrometry fragmentation rule of ABR is shown in Figure 5 .

The [M+H]⁺ of metabolites M1, M2, M3 and M4 is 366.2573, which are isomers. The molecular weight is 16Da more than that of ABR, suggesting that there is one more oxygen atom in the molecule. These four metabolites were obtained by oxidation reaction.

The [M+H]⁺ of M1 is 366.2573, and the secondary fragment ion 334.2327 is consistent with ABR. Fragments 207.1398 and 156.1029 prove that the oxidation position is not on pyridine ring, five membered ring and C18; The secondary fragment 262.1769 is consistent with the secondary fragment information of M4, indicating that the oxidation position is not on the methyl group, so the possible position of the oxidation position of M1 is C12, C11, C9, C8, C6 or C7. Because the parent ion 366.2573 goes to the secondary fragment ion 334.2317 and loses 32 Da, and the CH₃OH structure is excluded for the above reasons, then only 1 H₂O and 1 CH₄ can be lost at the same time and form a double bond, the parent ion to the secondary 318.1700 loses 48Da, it is speculated that 1 H₂O and 2 CH₄ are lost. Therefore, M1 may be oxidized to hydroxyl at C12 or C9, but it is more likely to oxidize at C12 from 262.1769 fragment information. The multistage mass spectrometry information of M1 is shown in Table 1 . The MS fragmentation rule of M1 is shown in Figure 6A . The possible chemical structure of Abiraterone metabolite M1 is shown in Figure 4B .

The secondary fragments of M2 and M3 are 350.2615, which may be obtained by the loss of CH₄ (16Da) by the parent ion. Because M2 and M3 contain fragments 156.0858, it is speculated that the oxidation position is not on the pyridine ring, five membered ring and C18. However, M2 contains the same fragment 318.2209 as M1, which may be that the parent ion 366.2567 loses a CH₄ and a CH₃OH, and M2 may be oxidized to methoxy at C19.

There is no 318.2209 in the secondary fragment of M3. The secondary fragment 248.2133 of M3 further indicates that the oxidation position is not on the five membered ring and pyridine ring. The Multistage mass spectrometry information of M2 and M3 are shown in Table 1 . The MS fragmentation rule of M2 and M3 are shown in Figures 6B, C . The possible chemical structure of M2 and M3 are shown in Figures 4C, D .

The [M+H] ⁺ of M4 is 366.2597, and the secondary fragment ion is 334.2191, 316.2325, 262.2264. This mass spectrometry information is consistent with Abiraterone N-oxide, a metabolite of ABR in the literature. By comparing the retention time and mass spectrometry information with Abiraterone N-oxide standard samples (The chemical structure of M4 (Abiraterone N-oxide) is shown in Figure 1B . The EIC diagram of Standard Abiraterone N-Oxide is shown in Figure 1E ), we can determine that M4 is Abiraterone N-oxide. The multistage mass spectrometry information of M4 is shown in Table 1 . The MS fragmentation rule of M4 is shown in Figure 6D . The chemical structure of M4 (Abiraterone N-oxide) is shown in Figures 1B , 4E .).

The fragment of intestinal metabolite M5 has great changes with ABR, suggesting that the reaction may be more complex. The [M+H] ⁺ of the metabolite M5 is 389.3033. The molecular weight of M5 is 39.10 Da more than that of ABR. The parent ion 389.3033 to the secondary ion 371.3003 lose an H₂O, indicating that there is a hydroxyl group in the structure and that the hydroxyl group at the C3 position of ABR remains unchanged. The third-order fragment 325.2909 is obtained from the loss of 46da of fragment ion 371.3003, indicating the possible loss of CO and H₂O at the same time or the loss of neutral molecular formic acid. Because there is no fragment related to ABR in the fragment, there may be no pyridine ring in the metabolite. The remaining structural molecular weight of ABR without pyridine ring is 274.23, while the [M+H]⁺ of metabolite is 389.3033, with a difference of 115.07. The structural type of ABR is like cholesterol. It is speculated that ABR may lose pyridine ring under the action of intestinal bacteria, connecting a long-chain carboxylic acid will produce a neutral loss of 46 Da, that is, the loss of carboxylic acid. Therefore, the group connected on the five membered ring is C₆H₁₂O₂. Therefore, it is speculated that the chemical structure of M5 is shown in the Figure 4F , and the possible cracking is shown in the Figure 6E .

Summarizing the cleavage laws of the above metabolites, we summarize the metabolic pathway of abiraterone in the gut microbiota system and liver microsomes (Abiraterone is metabolized in the gut microbiota to produce M5 and produce M1-4 in liver microsomes), as shown in Figure 7 .