5.2
Impact Factor
Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
Search in posts
Search in pages
Filter by Categories
Corrigendum
Current Issue
Editorial
Erratum
Full Length Article
Full lenth article
Letter to Editor
Original Article
Research article
Retraction notice
Review
Review Article
SPECIAL ISSUE: ENVIRONMENTAL CHEMISTRY
5.3
Impact Factor
Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
Search in posts
Search in pages
Filter by Categories
Corrigendum
Current Issue
Editorial
Erratum
Full Length Article
Full lenth article
Letter to Editor
Original Article
Research article
Retraction notice
Review
Review Article
SPECIAL ISSUE: ENVIRONMENTAL CHEMISTRY
View/Download PDF

Translate this page into:

Original article
09 2023
:16;
105004
doi:
10.1016/j.arabjc.2023.105004

Metabolic profiling of parthenolide: Characterized by bioactive organosulfur compounds involving in hydrogen sulfide regulation

, , , , , , , ,
a School of Pharmaceutical Sciences, Key Laboratory of Innovative Drug Development and Evaluation, Hebei Medical University, Shijiazhuang 050017, China
b College of Chemical Engineering, Shijiazhuang University, Shijiazhuang 050035, China
c Shijiazhuang Key Laboratory of Targeted Drugs Research and Efficacy Evaluation, Shijiazhuang 050035, China
d Department of Pharmacology, Hebei Medical University, Shijiazhuang 050017, China

⁎Corresponding authors. gaohx686@hebmu.edu.cn (Haixia Gao), rainbowhuo@126.com (Changhong Huo)

Received: , Accepted: ,
© The Author(s) 2023
Disclaimer:
This article was originally published by Elsevier and was migrated to Scientific Scholar after the change of Publisher.
1 Changhong Huo will handle correspondence at all stages of refereeing, publication, and post-publication.

Abstract

Parthenolide (PA) shows promising therapeutic effect on ulcerative colitis and colorectal cancer, and can ameliorate colon inflammation in a gut microbiota-dependent manner. In the present study, an effective UHPLC-QTOF-MS method aided by rat intestinal microbiota biotransformation was applied to characterize 34 metabolites (including a series of characteristic R-Sn-R type organosulfur compounds) of PA for the first time. Noteworthy, organosulfur metabolites PA-S2-PA and PA-S3-PA could react with cysteine to generate H2S via persulfide/polysulfide intermediates and significantly inhibit colonic cancer HT-29 and HCT116 cells viability and pro-inflammatory cytokine IL-6 release. These results indicated that PA and its organosulfur metabolites involved in H2S regulation in intestinal tract, which should be beneficial for a more in-depth understanding of the action mechanism of PA.

Keywords

Parthenolide
Metabolite
UHPLC-QTOF-MS
Hydrogen sulfide
Bioactivity

Abbreviations

PArthenolide

PA

Hydrogen sulfide

H2S

1

1 Introduction

Hydrogen sulfide (H2S) is an important gasotransmitter involving in numerous physiological and pathophysiological processes via the persulfidation of reactive cysteine (Cys) thiols in target proteins. The imbalance of H2S can cause various diseases, such as inflammation, cancer, cardiovascular disease and neurodegenerative disease (Filipovic et al., 2018; Wang, 2012). In the intestinal tract (a major site of H2S production), H2S is produced both endogenously by epithelial cells and exogenously by microbiota. So far, the action mechanisms of H2S on ulcerative colitis (UC) and colorectal cancer (CRC) remain rather confusing and controversial. And the reported deleterious or beneficial role of H2S in inflammation and cancer depends on the model system used, the route of administration, the source of H2S (endogenous vs exogenous), and the bioavailability of H2S (fast- vs slow-releasing H2S donors). However, the concept that H2S is a Janus-faced molecule has been evidenced by mounting researches. On one hand, excessive H2S production from colonic microbial metabolism has clinically relevance in the pathogenesis of UC and CRC. The H2S-detoxifying ability of the colon is decreased in UC and CRC patients. Higher luminal H2S level displays pro-inflammatory effects and affects the mucus integrity. On the other hand, compiling evidences have showed that a “healthy” low level of H2S is likely necessary to combat inflammation, oxidative stress related tissue injury, and cancer (Attene-Ramos et al., 2006; Blachier et al., 2021; Qin et al., 2019; Xiao et al., 2021).

Experimental studies have revealed that orally administered dietary R-Sn-R type organosulfur compounds (n ≥ 1), such as diallyl sulfide (DAS), diallyl disulfide (DADS), and diallyl trisulfide (DATS) derived from garlic, could ameliorate UC and CRC (Fasolino et al., 2015; Mondal et al., 2022). The R-Sn-R type organosulfur compounds have already been verified as thiol-activated H2S donors, and their capabilities to produce H2S are related to substituent group (R-) and the number of linking sulfur atoms (Sn) (Benavides et al., 2007; Liang et al., 2015). Recent studies suggested that the R-Sn-R type organosulfur compounds (n ≥ 2) could react with thiols like free Cys and glutathione to simultaneously generate H2S and persulfide/polysulfide, while the latter were potential signaling molecules which might regulate the activity of enzymes and ion channels, and even the growth of tumors (Filipovic et al., 2018; Ida et al., 2014; Kimura et al., 2013). Moreover, the substituent group (R-) might contribute to the bioactivity of the R-Sn-R type organosulfur compounds by alkylation of Cys thiols in target proteins, and make these compounds show selective cytotoxicity on cancer cells (Czepukojc et al., 2014; Yagdi Efe et al., 2017). It is worth mentioning that previous studies on the R-Sn-R type organosulfur compounds have mainly focused on the compounds owning simple substituent groups (such as allyl and benzyl) (Bolton et al., 2019), the ones owning the substituent group derived from active phytochemical have been neglected.

Parthenolide (PA) is a major bioactive ingredient of various edible and medicinal Asteraceae plants, such as chrysanthemum (Chrysanthemum morifolium) (Sun et al., 2016), feverfew (Tanacetum parthenium) (Freund et al., 2020), Artemisia selengensis, A. anomala, Gynura bicolor, Aster indicus, and Hemisteptia lyrata (Zhou et al., 2022). In recent years, PA has attracted great attention due to its significant anticancer, anti-inflammatory, and redox-modulating activities (Freund et al., 2020). However, although PA shows highly efficacious, its oral bioavailability is as low as 7.78% in rat (Zhao et al., 2016). To improve the bioavailability of PA, several different types of nanostructures have already been developed in the past few years (Guo et al., 2023; Albalawi et al., 2023; Darwish et al., 2019; Gao et al., 2020). Of note, PA displayed promising therapeutic potential on UC and CRC in animal models (Liu et al., 2020; Yang et al., 2013). Interestingly, the anti-inflammatory action of PA in colitis mice disappeared following gut microbiota depletion, while PA could ameliorate colon inflammation of the gut microbiota-depleted mice after fecal microbiota transplantation (Liu et al., 2020). These studies indicated that the metabolites produced by gut microbiota might play important roles in the pharmacological effects of PA, especially on intestinal diseases. However, the metabolism of PA remains unknown. In principle nucleophilic H2S should be captured easily by the α-methylene-γ-lactone (αMγL) functional group of PA. Our previous study isolated two R-Sn-R (n = 1 and 2) type organosulfur compounds from the biotransformation of isoalantolactone (another sesquiterpene lactones possessing αMγL moiety) by rat intestinal microbiota (Yao et al., 2017). Therefore, we speculated that orally administered PA might be possible to generate R-Sn-R type metabolites in vivo and the therapeutic potentials of PA might be related to H2S regulation.

In this article, anaerobic incubation with rat intestinal microbiota in vitro, as an effective method to reflect the biotransformation of natural products in the intestinal tract (Yao et al., 2017), was used to study the metabolites of PA. As expected, four novel sulfur-containing metabolites, including three R-Sn-R type organosulfur compounds (PA-Sn-PA, n = 1–3) and an intermediate product (PA-SH) were obtained (Fig. 1). Based on the structures of these metabolites, an ultrahigh-performance liquid chromatography combined with quadrupole time-of-flight mass spectrometry (UHPLC-QTOF-MS) method was developed to identify the metabolites of PA in vivo and in vitro. As the characteristic metabolites of PA in vivo, three organosulfur compounds (PA-Sn-PA, n = 1–3) were further investigated of their inhibitory activities on the proliferation of human colonic cancer cells and the release of LPS-induced pro-inflammatory cytokine IL-6. Their H2S-releasing ability in the presence/absence of Cys and reaction products with Cys were also detected. Our studies indicated that PA and its metabolites involved in the regulation of H2S and could exert beneficial effect on intestinal tract, which provides a novel insight into the underlying action mechanism of PA.

Structures of PA and its organosulfur metabolites obtained from intestinal microbiota biotransformation.
Fig. 1
Structures of PA and its organosulfur metabolites obtained from intestinal microbiota biotransformation.

2

2 Materials and methods

2.1

2.1 Chemicals and reagents

PA was purchased from Baoji Herbest Biotechnology Co. Ltd. (Shanxi, China) with purity ≥ 98%. HPLC-grade acetonitrile was obtained from Fisher Scientific (USA). Formic acid and ammonium acetate were obtained from Sigma-Aldrich (St. Louis, Missouri, USA) and Diamond Technology Incorporation (USA), respectively. De-ionized water was prepared using Milli-Q water purification system (Millipore, Boston, USA). Silica gel (300–400 mesh, Qingdao Marine Chemical Inc., China) was used for column chromatography. The purity of PA-SH and PA-Sn-PA (n = 1–3) were > 93% detected by UPLC-UV. Sodium carboxymethyl cellulose (CMC-Na) was analytical grade (Tianjin Yongda Chemical Corporation, China). Fluorescent probe WSP-5 (CAS: 1593024–78-2) was purchased from Cayman Chemical Company (USA). Other analytical grade reagents were purchased from Tianjin Chemical Corporation (China).

2.2

2.2 Intestinal microbiota biotransformation of PA

Crude extract of intestinal microbiota was obtained from fresh feces of healthy rats via suspended in physiological saline, and added to cultural medium for preparing the culture solution of intestinal microbiota (Yao et al., 2017). Next, PA dispersed in 0.5% (w/v) CMC-Na solution (20 mg/mL) was mixed with the culture solution (1:20, v/v) and incubated under N2 atmosphere at 37 °C for 48 h. The mixture was extracted with equal volume of ethyl acetate for 3 times, and then the evaporated extract was separated by silica gel column chromatography and preparative HPLC. Finally, 4 new biotransformation products named as PA-SH, PA-S-PA, PA-S2-PA, and PA-S3-PA were obtained. On the atmospheric column chromatography, the ratio of sample and silica gel for separation was 1:100 (w/w) and the mobile phase was petroleum ether-ethyl acetate (2:1.1, v/v). The eluents were identified and combined according to their thin layer chromatography behavior. The combined fractions were dried and identified by UHPLC-QTOF-MS. Then, the target fractions were purified by preparative HPLC (acetonitrile–water system). PA-SH was eluted at 22.7 min with the elution condition set as: 0–25 min, 25%-60% acetonitrile (GRACE allsphere ODS column, 250 mm × 22 mm, 5 μm, flow rate = 20 mL/min). PA-S-PA and PA-S2-PA were eluted at 22 min and 32 min with the elution condition of 56% acetonitrile (YMC-ODS-A column, 250 mm × 10 mm, 5 μm, flow rate = 5 mL/min), respectively. PA-S3-PA was eluted at 21 min with the elution condition of 64% acetonitrile (YMC-ODS-A column, 250 mm × 10 mm, 5 μm, flow rate = 5 mL/min). Because PA-SH, PA-S2-PA and PA-S3-PA isolated were unstable (PA-S-PA is stable) during the dryness process (≥45 °C), the dryness condition was adapted as follows: the acetonitrile in the eluents was removed using vacuum rotary evaporator under 15 °C, and the remaining aqueous phase was dried with freeze-drying machine (LGJ-10D, Beijing Sihuan Scientific Instrument Factory Co., China). The structures of 4 products were identified by HR-MS and NMR technology.

2.3

2.3 Animals and drug administration

16 male Sprague-Dawley rats (200 ± 20 g) obtained from Laboratory Animal Center of Hebei Medical University (Shijiazhuang, China) were acclimatized for 7 days before any treatments. All rats were fasted for 12 h with water ad libitum prior to PA administration, and then they were randomly divided into 4 groups: group 1 for blank urine and feces; group 2 for sample urine and feces; group 3 for blank bile; group 4 for sample bile. Groups 2 and 4 were orally given PA dispersed in 0.5% CMC-Na solution (20 mg/mL) at a dose of 100 mg/kg, while groups 1 and 3 were orally administrated with equivalent 0.5% CMC-Na solution without PA. All animal experiments were carried out in accordance with the guidelines of the Chinese Association for Laboratory Animal Sciences and approved by the Animal Ethics Review Committee of Hebei Medical University (Shijiazhuang, China).

2.4

2.4 Sample collection and preparation

Feces and urine samples of groups 1 and 2 were collected for 72 h in metabolic cages. Groups 3 and 4 were anesthetized with 20 % urethane (1–1.5 mL, i.p.) after oral administration, and the bile samples were collected from 0 h to 24 h through bile duct intubation. Microbial biotransformation samples were acquired according to the procedure 2.2. with or without PA. All samples were stored at −20 °C until further processing. The gathered bile, urine, and feces samples and microbial biotransformation samples were separately extracted with equal volume of ethyl acetate for 3 times. All extracts of samples were evaporated to dryness in vacuum drying chamber. After that, the dried samples were dissolved in acetonitrile. Each sample was filtered with 0.22 μm membrane and injected into the UHPLC-MS system at 5 μL per time.

2.5

2.5 UHPLC-QTOF-MS conditions and data processing

UHPLC-QTOF-MS experiments were carried out on a UHPLC system (Shimadzu, Kyoto, Japan) coupled with Triple TOFTM 5600+ system installed ESI source (AB SCIEX, Massachusetts, USA). The chromatographic separations of samples were realized on YMC-UItraHT Pro C18 (100 × 3.0 mm, 2 μm) column (YMC, Japan) with the column temperature at 25 °C. The mobile phase was comprised of water-ammonium acetate (2 mM) (A) and acetonitrile (B). Gradient condition was set as: 0–1 min, 10% B; 1–15 min, 10–60% B; 15–30 min, 60–95% B; 30–40 min, 95% B; 40–45 min, 10% B, and flow rate was 0.4 mL/min. MS analysis was carried out in positive ionization mode. In order to detect as many metabolites as possible, the ionization mode combined monitoring of [M+H]+, [M+NH4]+, and [M+Na]+ was applied. The operating parameters were the following: ion source heater, 450 °C; ion spray voltage, 5.5 kV; declustering potential, 50 eV; collision energy, 30 eV; collision energy spread, 15 eV; nebulizer gas, 55 psi; heater gas, 55 psi; curtain gas, 35 psi. The real-time monitor technique dynamic background subtraction (DBS) and multiple mass defect filter (MMDF) was set to supplement information dependent acquisition method. Of note, the novel organosulfur metabolites obtained in biotransformation were applied to MMDF setting. MetabolitePilot™ 2.0 software (AB SCIEX) was used for the data analyzing. The m/z of product ions in MS/MS was confirmed by mass calculator in PeakView 1.2 (AB SCIEX).

2.6

2.6 Cell culture

Human colonic cancer HT-29 and HCT116 cells (American Type Culture Collection, Manassas, VA, USA) were cultured in McCoy's 5A medium (Biological Industries Israel Beit Haemek Ltd. (BI), Israel) supplemented with 10% fetal bovine serum (FBS, Viva cell, Shanghai XP Biomed Ltd., China) and 1% penicillin–streptomycin antibiotics (BI, Israel). Mouse RAW 264.7 macrophages (Academy of Military Medical Sciences, China) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, GIBICO, Carlsbad, CA, USA) supplemented with 10% FBS and 1% penicillin–streptomycin antibiotics. Both cell lines were incubated at 37 °C in a humidified incubator with 5% CO2 and 95% air.

2.7

2.7 Measurement of cell viability with CCK-8 assay

Cell viability was determined by CCK-8 kit (Report, China) following manufactory’s manual. Cells were plated in 96-well plates with a confluence of 60% before treating. After indicated treatment with PA or PA-Sn-PA (n = 1–3) for 24 h, 10 μL CCK-8 reagents were added to the medium and incubated for 1 h at 37 °C. Then, the optical density at 450 nm was measured by Spectra Max microplate reader (Molecular Devices, USA). Each experiment was triplicate independently.

2.8

2.8 Determination of IL-6 production

The concentrations of IL-6 in the cell culture media were determined by commercial IL-6 ELISA kit (MultiSciences (Lianke) Biotech Co., Ltd., Zhejiang, China). RAW 264.7 macrophages were seeded in 24-well plates with a density of 3.6 × 104 cells/well and pre-incubated for 24 h. Cells were pretreated with PA or PA-Sn-PA (n = 1–3) diluted with 1% FBS-DMEM medium for 1 h, and stimulated with lipopolysaccharide (LPS, 1 μg/mL) for another 24 h. Then, the culture media were collected and centrifuged at 1780 rpm for 10 min. The IL-6 concentrations of supernatant were determined following the instruction of the manufacturer. The optical density at 450 nm was measured by Spectra Max microplate reader. Each experiment was triplicate independently.

2.9

2.9 Detection and identification of reaction products from PA-Sn-PA (n = 1–3) and Cys

To match the UPLC system (ACQUITY UPLCTM H-Class TUV, waters, USA) and the solubility of compounds, reaction solution was adjusted to 50% PBS-acetonitrile. PA-Sn-PA (n = 1–3) were dissolved in acetonitrile with the concentration of 1 mM. Cys was dissolved in PBS (phosphate buffer saline, 10 mM, pH 7.2–7.4). To detect the reaction products, PA-Sn-PA (n = 1–3) with a final concentration of 500 μM were mixed with Cys at the molar ratio of 1:1 and 1:4, respectively. The reaction solution was detected by UPLC at 210 nm in 30 min. The mobile phase consisted of water (A) and acetonitrile (B). The elution requirement was set as: 0–2 min, 20% B; 2–15 min, 20–80% B; 15–25 min, 80–95% B; 25–35 min, 95% B; 35–36 min, 95–10% B; 36–40 min, 10% B, and flow rate was 0.4 mL/min. Additionally, the reaction products were identified by UHPLC-QTOF-MS. The mobile phase was composed of water-0.1% (v/v) formic acid (A) and acetonitrile (B), and the elution requirement was the same as UPLC system. The MS condition was the same as procedure 2.5 without MMDF settings.

2.10

2.10 Capture and identification of reaction intermediates from PA-Sn-PA (n = 2–3) and Cys

Sulfide intermediates was captured by monobromobimane (MBB) and identified by UHPLC-QTOF-MS (Liang et al., 2015). Firstly, 20 μM PA-S3-PA or PA-S2-PA was mixed with 20 μM Cys in 50% PBS-acetonitrile, and reacted for 30 min or 1 h, respectively. Then 200 μM MBB was added with the final concentration of 20 μM, and reacted for another 40 min. Next, 5 μL reaction solution was injected to UHPLC-QTOF-MS for analysis. The UHPLC mobile phase was water-0.1% (v/v) formic acid (A) and acetonitrile (B), and the elution requirement was set as: 0–5 min, 5% B; 5–20 min, 5–90% B; 20–25 min, 90% B; 25–26 min, 90–5% B, 26–30 min, 5% B. The MS condition was the same as procedure 2.5 without MMDF settings.

2.11

2.11 H2S releasing ability assay

The fluorescence intensity generated from the reaction of H2S and fluorescent probe WSP-5 was measured by F-7100 fluorescence spectrophotometer (Hitachi High-Tech Group, Japan). WSP-5 and PA-Sn-PA (n = 1–3) were dissolved in DMSO as 1 mM stock solutions, respectively. Cys was dissolved in PBS with the concentration of 1 mM, which should be used in 24 h. Surfactant cetyl trimethyl ammonium bromide (CTAB) was dissolved in EtOH (20 mM) and diluted with PBS to a final concentration of 1 mM. As the final reaction solution, 10 μL WSP-5 stock solution was mixed with 960 μL PBS containing CTAB, then 10 μL DMSO (blank control) or PA-Sn-PA (n = 1–3) stock solution and 40 μL PBS or Cys stock solution were added, respectively. The basic fluorescence intensity of reaction solution without PA-Sn-PA (n = 1–3) was monitored within 6 h with or without 40 μM Cys. The fluorescence intensity changing of 10 μM PA-Sn-PA (n = 1–3) in the reaction solution with or without 40 μM Cys was monitored at emission wavelength (λem) 525 nm with excitation wavelength (λex) at 502 nm for 6 h, respectively.

2.12

2.12 Statistical analysis

The cell viabilities and IL-6 concentrations were exhibited as mean ± standard deviation (M ± SD). Statistical tests were performed with GraphPad Prism Software version 8.0.2. The statistical significance was analyzed with Student’s t-test or one-way analysis of variance (ANOVA) and set at * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

3

3 Results and discussion

3.1

3.1 Structural elucidation of biotransformation products (PA-SH, PA-S-PA, PA-S2-PA, and PA-S3-PA)

PA-SH was isolated as a white amorphous powder. Its HR-MS ion at m/z 305.1188 [M+Na]+ (calcd 305.1187) suggested a molecular formula of C15H22O3S, 34 Da more than that of PA, implying a hypothetical addition of one H2S molecule at Δ11,13 or Δ1,10 of PA. In the NMR data (Table 1) of PA-SH, the absence of one characteristic sp2 methylene signals [δH 6.33 (1H, d, J = 3.7 Hz) and 5.62 (1H, d, J = 3.2 Hz); δC 121.3] for CH2-13 of parent drug PA (Jacobsson et al., 1995) and the presence of an additional sp3 methylene signals [δH 3.18 (1H, m) and 2.70 (1H, m); δC 22.0] indicated that the addition reaction of H2S should occur at Δ11,13, i.e., one -SH group was connected with C-13. The HMBC and 1H–1H COSY correlations also supported this deduction. Moreover, the NOESY correlations of H-5 (δH 2.81)/H-7 (δH 2.53) and H-6 (δH 3.90)/H-15 (δH 1.33) suggested that the stereo configurations of C-4, C-5, C-6, and C-7 of PA-SH should be identical with those of PA. The strong NOESY correlation of H-6/H-11 (δH 2.67) indicated H-11 was in β-orientation. Therefore, PA-SH was established as 13-sulfydryl-11β,13-dihydroparthenolide.

Table 1 1H and 13C NMR data of PA-SH, PA-S-PA, PA-S2-PA and PA-S3-PA (CDCl3, in ppm).
position PA-SH a PA-S-PA a PA-S2-PA b PA-S3-PA b
δC δH (J in Hz) δC δH (J in Hz) δC δH (J in Hz) δC δH (J in Hz)
1/1′ 125.4 5.27, ddd (11.8, 3.7, 1.6) 125.3 5.29, ddd (12.1, 3.8, 1.7) 125.2 5.22, dd (12.0, 3.8) 125.5 5.25, dd (12.9, 3.7)
2/2′ 24.1 2.43, m
2.22, m		 24.1 2.43, m
2.21, m		 24.1 2.39, m
2.19, m		 24.1 2.38, m
2.19, m
3/3′ 36.6 2.18, m
1.27, m		 36.6 2.16, m
1.28, m		 36.6 2.14, m
1.23, m		 36.5 2.14, m
1.27, m
4/4′ 61.5 61.5 61.5 61.6
5/5′ 66.3 2.81, d (9.0) 66.3 2.81, d (9.0) 66.4 2.75, d (9.0) 66.1 2.78, d (9.0)
6/6′ 82.3 3.90, t (9.0) 82.4 3.88, t (9.0) 82.5 3.88, t (9.0) 82.6 3.89, t (9.0)
7/7′ 46.1 2.53, m 47.2 2.39, m 47.23 2.36, m 47.8 2.40, m
8/8′ 29.9 1.90, m
1.72, m		 29.9 1.97, m
1.69, m		 29.9 2.00, m
1.69, m		 29.9 2.01, m
1.70, m
9/9′ 41.0 2.32, dd (13.3, 6.5)
2.12, m		 40.9 2.32, dd (13.3, 6.4)
2.17, m		 40.9 2.30, dd (13.4, 6.4)
2.15, m		 40.7 2.30, dd (13.4, 6.4)
2.16, m
10/10′ 134.3 134.4 134.5 134.2
11/11′ 49.8 2.67, m 48.4 2.69, dt (12.2, 4.2) 47.3 2.77, m 47.7 2.81, m
12/12′ 174.8 175.2 175.2 174.8
13/13′ 22.0 3.18, m
2.70, m		 31.5 3.18, dd (13.9, 3.9)
3.04, dd (13.9, 4.6)		 36.4 3.28, dd (14.1, 4.1)
3.14, dd (14.1, 5.2)		 37.8 3.44, dd (14.4, 4.3)
3.29, dd (14.4, 5.4)
14/14′ 16.9 1.74, s 16.9 1.73, s 16.9 1.70, s 16.9 1.71, s
15/15′ 17.2 1.33, s 17.2 1.33, s 17.2 1.30, s 17.2 1.30, s
a Measured at 500 (1H) MHz and 125 (13C) MHz; b Measured at 600 (1H) MHz and 150 (13C) MHz.

PA-S-PA, a white amorphous powder, was identified as C30H42O6S established by the HR-MS [M+Na]+ ion at m/z 553.2593 (calcd 553.2600). However, it is interesting that only 15 carbon signals were presented in the 13C NMR spectrum (Table 1), which implied that PA-S-PA should have a symmetrical skeleton. The 1H and 13C NMR spectra were similar to those of PA-SH, except the signals of C-7 (Δ δ = + 1.1), C-11 (Δ δ = − 1.4), C-12 (Δ δ = + 0.4), and C-13 (Δ δ = + 9.5), which proposed that PA-S-PA was composed by two 11,13-dihydroparthenolide molecules linking through a sulfur atom. The fragment ion [M+H-C15H22O3S]+ at m/z 249 in the MS spectrum further supported the above conclusion. The NOESY correlations of H-5 (δH 2.81)/H-7 (δH 2.39) and H-6 (δH 3.88)/H-15 (δH 1.33) and H-11 (δH 2.69) confirmed that the stereo skeletons of two 11,13-dihydroparthenolide units of PA-S-PA was in agreement with that of PA-SH. Then, PA-S-PA was identified as 13,13′-thiobis(11β,13-dihydroparthenolide).

PA-S2-PA was obtained as a white amorphous powder. Its molecular formula was assigned as C30H42O6S2 by the HR-MS ion at m/z 585.2315 [M+Na]+ (calcd 585.2321), 32 Da more than that of PA-S-PA. Comparison of its 1H NMR and 13C NMR data (Table 1) with those of PA-S-PA suggested that PA-S2-PA should possess two same 11β,13-dihydroparthenolide units. A significant difference was that the resonance of C-13 exhibited downfield shift (Δ δ = + 4.9), implying that PA-S2-PA was the corresponding disulfide dimer of 11β,13-dihydroparthenolide. The fragment ion [M+H-C15H22O3S]+ at m/z 281 in the MS spectrum and NOESY experiment could confirm that the structure of PA-S2-PA was 13,13′-dithiobis(11β,13-dihydroparthenolide).

PA-S3-PA is a white amorphous powder, whose HR-MS showed [M+Na]+ ion at m/z 617.2058 (calcd 617.2041), with the formula composition of C30H42O6S3. According to the 1H NMR and 13C NMR data (Table 1), PA-S3-PA should also have two 11β,13-dihydroparthenolide units. The downfield shift (Δ δ = + 1.4) of C-13 and the similar NOESY correlations with those of PA-S2-PA confirmed that PA-S3-PA was 13,13′-trithiobis(11β,13-dihydroparthenolide).

The 1D and 2D NMR and MS spectra of PA-SH, PA-S-PA, PA-S2-PA, and PA-S3-PA are showed in Figs. S1-S32.

3.2

3.2 Identification of PA and its metabolites by UHPLC-QTOF-MS

UHPLC-QTOF-MS is a powerful tool for studying drug metabolism, and MMDF technique has been successfully applied in MS data on-line acquisition of potential metabolic pathway. In this study, based on the organosulfur compounds (PA-Sn-PA, n = 1–3) separated from intestinal microbiota biotransformation and identified by NMR, 10 MDF forms were set: (1) parent drug PA: 141.2447 mDa; (2) PA + glucuronide: 173.3329 mDa; (3) PA + sulfate: 100.8001 mDa; (4) PA + glutathione: 225.0521 mDa; (5) PA + acetylcysteine: 171.5598 mDa; (6) PA-S-PA: 270.2113 mDa; (7) PA-S2-PA: 242.2831 mDa; (8) PA-S3-PA: 214.3549 mDa; (9) PA-S4-PA: 186.4267 mDa; (10) PA-S5-PA: 158.4985 mDa. In data processing, common metabolite forms of PA tend to be captured as [M+H]+ form, while organosulfur compounds tend to be captured as [M+NH4]+ form. Thus, 2 adduct ion forms ([M+H]+, [M+NH4]+) were used in metabolites searching, and [M+Na]+ form was applied for confirmation (Yuan et al., 2022).

34 metabolites of PA were identified or tentatively identified in vivo and in vitro, including 15 common I phase metabolites (M1-M15) and 19 sulfur-containing metabolites (M16-M34), whose detailed information are listed in Table 2. The deduced structures and MS/MS spectra of these metabolites are showed in Fig. S34. The detailed analytical process of common I phase metabolites is described in Supporting information. The separated extracted ion chromatograms of PA and its metabolites are presented in Fig. S35, respectively.

Table 2 Summary of PA metabolites detected in vivo and in vitro.
No. Identification Molecular formula Rt
(min)		 [M+Na]+
(Error, ppm)		 [M+NH4]+/[M+H]+a
(Error, ppm)		 MS/MS fragment ions
(From MS [M+NH4]+/[M+H]+)		 Bb Ub Fb Mb
M0 Parent C15H20O3 14.92 271.1303
(−0.7)		 266.1745
(−2.2)		 266, 249, 231, 213 203, 185, 159, 145, 131, 119, 105, 91 + + + +
M1 Introduction of H2O C15H22O4 7.83 289.1403
(−2.7)		 284.1848
(−3.0)		 284, 267, 249, 231, 203, 185, 159, 145, 131, 105, 91 − + + −
M2 Introduction of H2O C15H22O4 8.23 289.1409
(−0.4)		 284.1851
(−1.7)		 284, 267, 249, 231, 203, 185, 157, 145, 131, 119, 105, 91 − + + −
M3 Introduction of H2O C15H22O4 10.35 289.1409
(−0.5)		 267.1585*
(−2.1)		 267, 249, 231, 203, 185,159,147, 131, 105, 91 − + + −
M4 Introduction of H2O C15H22O4 10.82 289.1419
(3.1)		 267.1598*
(2.6)		 267, 249, 231, 203, 185,159,145, 131, 105, 91 − + + +
M5 Oxidation C15H20O4 6.28 287.1249
(−1.8)		 265.1430*
(−1.6)		 265, 247, 229, 201, 183, 159, 145, 131, 119, 105, 91 − + − −
M6 Oxidation C15H20O4 7.79 287.1248
(−2.0)		 265.1430*
(−1.8)		 265, 247, 229, 201, 183, 159, 145, 131, 119, 105, 91 + + − −
M7 Oxidation C15H20O4 8.31 287.1260
(2.1)		 282.1702
(0.8)		 282, 265, 247, 229, 201, 183, 159, 145, 131, 119, 105, 91 + − − −
M8 Oxidation C15H20O4 9.61 287.1260
(2.1)		 265.1435*
(0.4)		 265, 247, 229, 201, 183, 157, 145, 131, 119, 105, 91 + + − −
M9 Introduction of H2O + Oxidation C15H22O5 5.86 305.1361
(0.6)		 300.1798
(−2.5)		 300, 283, 265, 247, 219, 201, 173, 159, 145, 131, 119, 105, 91 − + + −
M10 Introduction of H2O + Oxidation C15H22O5 6.66 305.1355
(−1.3)		 283.1531*
(−3.1)		 283, 247, 229, 219, 201, 187, 183, 173, 159, 131, 105, 91 + + − −
M11 Hydrogenation C15H22O3 15.10 273.1455
(−2.2)		 251.1637*
(−2.0)		 251, 233, 215, 205, 187, 159, 145, 131, 119, 105, 91 − + + +
M12 Introduction of two H2O C15H24O5 5.50 307.1513
(−0.9)		 302.1956
(−1.9)		 302, 285, 267, 249, 231, 203, 185, 157, 145, 131, 119, 105 − + + −
M13 Hydrogenation + Demethylation C14H20O3 8.14 – 237.1479*
(−2.8)		 237, 219, 201, 191, 177, 159, 145, 131, 105 − + − −
M14 Methylation C16H22O3 16.25 285.1475
(4.8)		 263.1640*
(−0.5)		 263, 245, 227, 217, 199, 187, 185, 159, 145, 131, 119, 105, 91 − − − +
M15 Ketone Formation C15H18O4 7.14 – 263.1271*
(−2.6)		 263, 217, 199, 189, 175, 173, 171, 157, 143,131, 119, 105, 91 − + − −
M16 PA-SH C15H22O3S 15.79 305.1175
(−2.3)		 283.1367*
(1.7)		 283, 265, 247, 231, 219, 185, 159, 145, 131, 119, 105, 91 − − + +
M17 PA-S-PA C30H42O6S 19.55 553.2581
(−2.4)		 548.3021
(−3.5)		 548, 531, 513, 495, 467, 449, 265, 249, 231, 213, 185, 159, 145, 131, 105 − − + +
M18 PA-S-PA C30H42O6S 20.11 553.2583
(−2.1)		 548.3027
(−2.4)		 548, 531, 513, 495, 467, 449, 265, 231, 213, 185, 159, 145, 131, 119, 105 + − + +
M19 PA-S-PA C30H42O6S 20.66 553.2575
(−3.4)		 548.3023
(−3.1)		 548, 531, 513, 495, 467, 449, 265, 249, 231, 213, 185, 159, 145, 131, 105 − − + +
M20 PA-S2-PA C30H42O6S2 21.25 585.2297
(−3.1)		 580.2739
(−3.9)		 580, 545, 527, 499, 281, 263, 231, 203, 185, 159, 145, 131, 119, 105 − − + +
M21 PA-S2-PA C30H42O6S2 21.62 585.2296
(−3.2)		 580.2749
(−2.1)		 580, 545, 527, 499, 281, 263, 231, 203, 185, 159, 145, 131, 119, 105 + − + +
M22 PA-S2-PA C30H42O6S2 21.83 585.2298
(−3.0)		 580.2737
(−4.2)		 580, 545, 527, 499, 315, 281, 263, 231, 203, 185, 159, 145, 131 − − + +
M23 PA-S3-PA C30H42O6S3 22.40 – 612.2461
(−3.4)		 612, 577, 559, 313, 295, 277, 263, 229, 185, 159, 145 − − + +
M24 PA-S3-PA C30H42O6S3 22.90 – 612.2456
(−4.2)		 612, 577, 559, 313, 295, 281, 277, 263, 229, 185, 159, 145 + − + +
M25 PA-S3-PA C30H42O6S3 23.64 – 612.2509
(4.4)		 612, 577, 559, 531, 313, 295, 281, 277, 263, 229, 185, 159, 145, 131 − − + +
M26 PA-S4-PA C30H42O6S4 24.41 649.1728
(−4.4)		 644.2169
(−5.2)		 644, 609, 591, 563, 345, 327, 281, 263, 231, 185, 159, 145 − − + +
M27 PA-S4-PA C30H42O6S4 24.64 649.1733
(−3.7)		 644.2176
(−4.2)		 644, 609, 591, 345, 327, 281, 263, 231, 185, 159, 145 + − + +
M28 PA-S4-PA C30H42O6S4 25.37 649.1722
(−5.3)		 644.2172
(−4.7)		 644, 609, 345, 327, 281, 263, 231, 213, 185, 159, 145 − − + +
M29 PA-S5-PA C30H42O6S5 26.41 681.1442
(−5.2)		 676.1885
(−5.7)		 676, 641, 623, 359, 327, 281, 263, 229, 185, 159 − − + +
M30 PA-S5-PA C30H42O6S5 25.96 681.1443
(−5.0)		 676.1889
(−5.1)		 676, 641, 623, 359, 327, 281, 263, 229, 185, 159 − − + +
M31 PA-S-PA + oxidation C30H42O7S 16.42 569.2547
(0.7)		 564.2984
(−1.0)		 564, 547, 529, 511, 483, 465, 281, 249, 231, 213, 185, 159, 145, 131, 105 + − + −
M32 PA-S-PA + oxidation C30H42O7S 14.81 569.2535
(−1.4)		 564.2977
(−2.3)		 564, 547, 529, 511, 465, 281, 249, 231, 213, 185, 159, 145, 131, 105 − + + −
M33 PA-S2-PA + oxidation C30H42O7S2 15.69 601.2249
(−2.5)		 596.2691
(−3.2)		 596, 561, 543, 525, 497, 297, 281,263, 231, 185, 159, 145 − − + −
M34 N-acetylcysteine conjugation C20H29NO6S 8.31 434.1613
(−1.2)		 – 305, 271, 186 + + + −
a molecular ions marked with * are from [M+H]+.
b B: bile; U: urine; F: feces; M: microbial biotransformation; +: detected; −: undetected.

3.2.1

3.2.1 Fragmentation study of PA

Under the present experimental condition, PA was eluted at 14.92 min and presented 3 adduct ions of [M+H]+ (m/z 249.1478), [M+NH4]+ (m/z 266.1746), and [M+Na]+ (m/z 271.1303). The MS/MS spectrum and fragment pattern of [M+NH4]+ are shown in Fig. S33. Product ion at m/z 249 was formed by the loss of NH3, and 4 representative fragment ions at m/z 231 [M+NH4-NH3-H2O]+, 213 [M+NH4-NH3-2H2O]+, 203 [M+NH4-NH3-H2O-CO]+, and 185 [M+NH4-NH3-2H2O-CO]+ were detected. Other typical product ions observed at m/z 159, 145, 131, 119, 105, and 91 should be due to the cleavage of germacrane scaffold of PA. It is worth noting that the appearance of 3 fragment ions at m/z 159, 145, and 131 should have important diagnostic value for the elucidation of PA derivatives, because they were formed by the successive losses of five-membered lactone ring, Me-14, and Me-15 from germacrane skeleton and retained unbroken ten-membered ring structure.

3.2.2

3.2.2 Sulfur-containing metabolites identification

M16 was extracted from the total ion chromatogram at m/z 283.1367 ([M+H]+, C15H22O3S), 34 Da higher than that of PA. The characteristic fragment ions at m/z 265, 247, 219, and 185 were derived from the successive losses of H2O, H2O, CO, and H2S (Fig. S34I). Because the retention time, molecular composition, parent ion and fragment ions of M16 were identical with those of PA-SH, M16 was unequivocally identified as 13-sulfydryl-11β,13-dihydroparthenolide.

M17-M19 were eluted at 19.55, 20.11, and 20.66 min, and showed [M+Na]+ ions at m/z 553.26 and [M+NH4]+ ions at m/z 548.30, respectively. Their theoretical molecular compositions were all inferred as C30H42O6S, identical with that of PA-S-PA (Fig. S34J). The appearance of product ions at m/z 265 [M+NH4-NH3-H2O-C15H20O3]+, 249 [M+NH4-NH3-C15H22O3S]+, and 231 [M+NH4-NH3-C15H22O4-H2S]+ indicated that these metabolites were composed of two 11,13-dihydroparthenolide molecules linked at C-13 by a sulfur atom, as same as PA-S-PA. Based on the same retention time with PA-S-PA obtained from biotransformation, M18 was finally identified as 13,13′-thiobis(11β,13-dihydroparthenolide), while M17 and M19 should be the stereoisomers of M18.

M20-M22 were eluted at 21.25, 21.62, and 22.40 min, and possessed [M+NH4]+ ions at m/z 580.27, implying that they had the same molecular composition of C30H42O6S2. The characteristic ions at m/z 315 [M+NH4-NH3-C15H20O3]+, 297 [M+NH4-NH3-C15H20O3-H2O]+, 281 [M+NH4-NH3-C15H22O3S]+, 263 [M+NH4-NH3-C15H22O3S-H2O]+, and 231 [M+NH4-NH3-C15H20O3-H2S-H2O-S]+ (Fig. S34K) showed that M20-M22 possessed the same planar structure as PA-S2-PA. Among M20-M22, M21 and PA-S2-PA standard isolated had the same retention time.

M23-M25 were eluted from 21.83 to 23.64 min. Their [M+NH4]+ ions provided a common molecular formula of C30H42O6S3, 32 Da higher than those of PA-S2-PA. The typical fragment ions at m/z 313, 281, 263, and 229 suggested that M23-M25 were the trisulfide dimer derivatives of PA (Fig. S34L). M24 was determined as PA-S3-PA on the basis of their same retention times.

M26-M28 were eluted from 24.41 to 25.37 min and the corresponding [M+NH4]+ ions were detected from m/z 644.2169 to 644.2176, whose molecular composition was C30H42O6S4. The crucial fragment ions at m/z 609, 591, 345, 281, and 231 implied that the cleavage pathway of M26-M28 were similar to that of PA-S3-PA (Fig. S34M). Hence, M26-M28 were defined as the tetrasulfide analogues of PA-S3-PA (named as PA-S4-PA).

M29 and M30 were eluted at 26.41 and 25.96 min respectively. Their [M+NH4]+ ions appeared at m/z 676.1885 and 676.1889 with the molecular composition of C30H42O6S5. The characteristic product ions at m/z 641 and 623 were all 32 Da higher than the related ions of M26-M28. Other typical product ions at 359, 327, 281, 263, and 229 could also be observed (Fig. S34N). Then, M29 and M30 were identified as the pentasulfide dimer derivatives of PA (named as PA-S5-PA).

M31 and M32 were detected at 16.42 and 14.81 min respectively, and showed [M+NH4]+ ions at m/z 564.2984 and 564.2977, 16 Da more than that of PA-S-PA. The characteristic product ions at m/z 281 and 249 suggested that M31 and M32 were the mono-oxidation derivatives of PA-S-PA (Fig. S34O). In a similar way, M33 (m/z 596.2691 [M+NH4]+, C30H42O7S2) eluted at 15.69 min was identified as the mono-oxidation derivative of PA-S2-PA (Fig. S34P).

M34 was eluted at 8.3 min with the [M+Na]+ ion at m/z 434.1598, 163 Da more than that of PA. Its theoretical molecular composition was C20H29NO6S. Based on the distinctive fragment ions at m/z 290 [M+Na-CO2]+, 305 [M+Na-C5H7NO3]+, 271 [M+Na-C5H9NO3S]+, and 186 [M+Na-C15H20O3]+, M34 was deduced to be the N-acetylcysteine addition product of PA (Fig. S34Q). Considering that the addition reaction easily takes place at Δ11,13 (Freund et al., 2020), M34 was identified as 13-N-acetylcysteinyl-11,13-dihydroparthenolide.

3.2.3

3.2.3 Metabolic profile of PA

In brief, there were 18 metabolites identified in the intestinal microbiota biotransformation sample and 34 metabolites identified in vivo, including 10 from the bile, 15 from the urine, and 26 from the feces. The proposed metabolic pathways of PA are shown in Fig. 2. The metabolic reactions involved introduction of H2O, oxidation, hydrogenation, methylation, demethylation, N-acetylcysteine conjugation, and a distinctive pattern: addition of H2S and further reaction to generate PA-Sn-PA type organosulfur compounds (n = 1–5). The results indicated that PA would undergo an uncommon metabolic pathway after oral administration, in which the αMγL group of PA first took place the Michael addition with microbial-derived H2S to produce PA-SH. Then, PA-SH as an intermediate further reacted to generate PA-Sn-PA (n = 1–5) and their oxidation products. Obviously, the organosulfur compounds were the characteristic metabolites in feces and bile (Fig. S36), whose sulfur source (microbial-derived H2S) and R-Sn-R structure feature hinted that they might play a dual role in modulating H2S of intestinal tract.

Proposed metabolic pathways of PA.
Fig. 2
Proposed metabolic pathways of PA.

3.3

3.3 Inhibition of PA-Sn-PA (n = 1–3) and PA on the proliferation of human colonic cancer cells

HT-29 and HCT116 cell lines were applied to evaluate the potential anticancer activity of PA-Sn-PA (n = 1–3) and parent drug PA. As shown in Fig. 3A, PA and 3 metabolites all displayed cell proliferation inhibition. PA-S3-PA and PA-S2-PA showed lower half maximal inhibitory concentrations (IC50) than PA, while PA-S-PA exhibited higher IC50. The results indicated that PA-Sn-PA type metabolites should contribute to the final anticancer bioactivities of PA. Relative dose–response relationship and specific IC50 data was exhibited in Fig. S37.

Potential inhibition activities against colonic cancer cells viability and IL-6 release of PA and PA-Sn-PA (n = 1–3). A. IC50 Values of PA and PA-Sn-PA (n = 1–3) against HT-29 and HCT116 cells. B. Effects of PA and PA-Sn-PA (n = 1–3) on LPS-induced IL-6 release in RAW 264.7 macrophages. Error bars indicate SD of the mean. The statistical significance was signed with * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Fig. 3
Potential inhibition activities against colonic cancer cells viability and IL-6 release of PA and PA-Sn-PA (n = 1–3). A. IC50 Values of PA and PA-Sn-PA (n = 1–3) against HT-29 and HCT116 cells. B. Effects of PA and PA-Sn-PA (n = 1–3) on LPS-induced IL-6 release in RAW 264.7 macrophages. Error bars indicate SD of the mean. The statistical significance was signed with * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

3.4

3.4 Inhibition of PA-Sn-PA (n = 1–3) and PA on the LPS-induced pro-inflammatory cytokine in RAW 264.7 macrophages

To screen the anti-inflammatory activity of PA-Sn-PA (n = 1–3) and PA, RAW 264.7 macrophages were pretreated with non-cytotoxic concentrations (Fig. S38) of PA-Sn-PA (n = 1–3) and PA for 1 h and stimulated with LPS to induce inflammation. Then the concentrations of IL-6 in cell culture media were detected. As shown in Fig. 3B, PA-S2-PA, PA-S3-PA and PA displayed inhibition activity against LPS-induced IL-6 release in a dose-dependent manner, while PA-S-PA didn’t. Particularly, the inhibition activity of PA-S2-PA and PA-S3-PA was comparable to parent PA at 0.25 μM. The results meant that PA-Sn-PA type metabolites also play a role in the anti-inflammatory activity of PA.

3.5

3.5 Reactions of PA-Sn-PA (n = 1–3) and Cys

The reaction products of PA-Sn-PA (n = 1–3) and Cys were detected by UPLC and UHPLC-QTOF-MS, preliminarily. Corresponding UPLC chromatograms were showed in Fig. 4. PA-S3-PA dissolved in 50% PBS-acetonitrile could generate trace amounts of PA-S2-PA and PA-S4-PA, which indicated that the S—S—S bond of PA-S3-PA was labile and tended to rearranged into complex mixtures of PA-Sn-PA. When equivalent Cys was added, PA-S3-PA decreased and accompanied by the increase of PA-S2-PA, PA-S4-PA and PA-S5-PA, and the appearance of 3 new peaks at retention times 6.9 min, 7.5 min, and 12.4 min. When 4 equiv of Cys was added, PA-Sn-PA (n = 2–5) reduced and 3 new peaks increased further. By UHPLC-QTOF-MS analysis, the new peaks were identified as PA-S-Cys, PA-S2-Cys, and PA-SH (Fig. 4A). In addition, PA-S3-Cys was discovered by UHPLC-QTOF-MS (retention time 7.86 min). The relative HR-MS spectra are listed in Figs. S39-S42. PA-S2-PA (Fig. 4B) didn’t produce obviously other sulfur-containing dimers without Cys, which matches the regularity that S—S bond is more stable than S—S—S bond (Arbach et al., 2019; Liang et al., 2015). When PA-S2-PA was mixed with Cys, PA-S-Cys and PA-SH were monitored. As for PA-S-PA, no more compounds generated with or without Cys (Fig. 4C).

Liquid chromatograms of the reaction products of PA-Sn-PA (n = 1–3) and Cys with the molar ratio of 1:1 and 1:4 at 210 nm. (A) PA-S3-PA, (B) PA-S2-PA, and (C) PA-S-PA.
Fig. 4
Liquid chromatograms of the reaction products of PA-Sn-PA (n = 1–3) and Cys with the molar ratio of 1:1 and 1:4 at 210 nm. (A) PA-S3-PA, (B) PA-S2-PA, and (C) PA-S-PA.

To ascertain a more integrated reaction course, the highly reactive sulfide intermediates of the reaction between PA-Sn-PA (n = 2–3) and Cys were captured by MBB (Liang et al., 2015). In the reaction of PA-S3-PA and Cys (Fig. 5A), Cys-S-bimane (peak 2, Fig. S43) and Cys-SS-bimane (peak 3, Fig. S44) were identified, demonstrating the appearance of Cys-SH and Cys-SSH. Peak 5 eluted at 11.85 min was identified as H2S derivative (Fig. S45), indicating the generation of H2S molecule. Peaks 7–9 were identified as PA-S-bimane, PA-SS-bimane, and PA-SSS-bimane (Fig. S46-S48), which were from PA-SnH (n = 1–3), respectively. In addition, MBB (peak 6, Fig. S49) and corresponding Cys derivative (Cys-bimane, peak 1, Fig. S50) were also detected. Surprisingly, trace PA-Cys (peak 4, Fig. S51) was detected under this condition. In the reaction of PA-S2-PA and Cys (Fig. 5B), corresponding MBB derivatives of Cys-SH (peak 2), H2S (peak 5), PA-SH (peak 7) and PA-SSH (peak 8) were detected, while the MBB derivatives of Cys-SSH and PA-SSSH were not. Similarly, product PA-Cys (peak 4) was discovered.

Total XIC of the reaction sulfide intermediates of PA-S3-PA (A) or PA-S2-PA (B) with equivalent Cys captured by MBB derivatization. (Peak 1: Cys-bimane; peak 2: Cys-S-bimane; peak 3: Cys-SS-bimane; peak 4: PA-Cys; peak 5: sulfide bimane; peak 6: MBB; peak 7: PA-S-bimane; peak 8: PA-SS-bimane; peak 9: PA-SSS-bimane.
Fig. 5
Total XIC of the reaction sulfide intermediates of PA-S3-PA (A) or PA-S2-PA (B) with equivalent Cys captured by MBB derivatization. (Peak 1: Cys-bimane; peak 2: Cys-S-bimane; peak 3: Cys-SS-bimane; peak 4: PA-Cys; peak 5: sulfide bimane; peak 6: MBB; peak 7: PA-S-bimane; peak 8: PA-SS-bimane; peak 9: PA-SSS-bimane.

On the basis of these results, we speculated that PA-Sn-PA (n ≥ 2) like garlic-derived diallyl sulfides could undergo thiol-disulfide exchange reaction and α-carbon nucleophilic substitution to generate the corresponding persulfide/polysulfide and H2S (Fig. 6) (Arbach et al., 2019; Benavides et al., 2007; Khodade et al., 2022; Liang et al., 2015).

Proposed reaction of PA-S2-PA (a) and PA-S3-PA (b) with Cys to produce persulfide/polysulfide and H2S.
Fig. 6
Proposed reaction of PA-S2-PA (a) and PA-S3-PA (b) with Cys to produce persulfide/polysulfide and H2S.

3.6

3.6 H2S releasing ability of PA-Sn-PA (n = 1–3) in the presence/absence of Cys

Fluorescent probe WSP-5 is a H2S-sensitive chemical reagent (Peng et al., 2014). Here, we used WSP-5 to compare the H2S release profile of PA-Sn-PA (n = 1–3) (Fig. 7). As expected, in the presence of Cys (40 μM), PA-S3-PA (10 μM) exhibited fast H2S release in 30 min, and then became slower in 4 h. PA-S2-PA (10 μM) sustained H2S release at a relatively steady rate within 6 h, but the release rate was slower than that of PA-S3-PA. The H2S release difference between PA-S2-PA and PA-S3-PA might be attributed to the fact that S—S bond is less stable than C—S bond (Steudel, 2002). So, PA-S3-PA tends to undergo thiol-disulfide exchange reaction to generate H2S rapidly at the initial reaction stage, while PA-S2-PA tends to generate H2S sluggishly via α-carbon nucleophilic substitution (Liang et al., 2015). Of course, PA-S2-PA, as a product of PA-S3-PA and Cys, would also affect the H2S release profile of PA-S3-PA with the progress of reaction. It should be noted that, without Cys in the test system, PA-S3-PA could also spontaneously release H2S at a slower rate while PA-S2-PA only showed detectable H2S release at a sluggish rate after 2 h. PA-S-PA (10 μM) showed no obvious H2S release in 6 h with or without Cys. Overall, these results confirmed that PA-Sn-PA (n ≥ 2) possessed H2S slow-releasing ability, and the ability could be accelerated by Cys.

H2S release profile of PA-Sn-PA (n = 1–3) in the presence or absence of Cys (40 μM) monitored by fluorescent probe WSP-5 at λem = 525 nm with λex = 502 nm. The concentrations of PA-Sn-PA (n = 1–3) were all 10 μM.
Fig. 7
H2S release profile of PA-Sn-PA (n = 1–3) in the presence or absence of Cys (40 μM) monitored by fluorescent probe WSP-5 at λem = 525 nm with λex = 502 nm. The concentrations of PA-Sn-PA (n = 1–3) were all 10 μM.

4

4 Conclusion

In summary, an effective UHPLC-QTOF-MS method combined with rat intestinal microbiota biotransformation was successfully used for the characterization of 34 metabolites of PA in vivo and in vitro. Besides the common metabolic products (originated from introduction of H2O, oxidation, hydrogenation, methylation, demethylation, and N-acetylcysteine conjugation), a series of novel organosulfur compounds (PA-Sn-PA, n = 1–5) and their related oxidation products together with an addition product of H2S were identified. The appearance of the addition product of H2S and PA-Sn-PA type metabolites means that PA has the capacity to down-regulate the higher luminal H2S level under intestinal pathological status. However, due to the fact that the organosulfur metabolites (n ≥ 2) could react with Cys to slowly regenerate H2S via persulfide/polysulfide intermediates and exhibit significant bioactivities, it is reasonable to speculate that PA and its organosulfur metabolites can relieve colonic diseases through dual regulating H2S level. Recently, H2S-based therapeutic approaches have already been proposed, and several small molecules involving in H2S modulation have moved into clinical trials. Therefore, our study may have an extraordinary significance in explaining the therapeutic material basis of PA on UC and CRC.

CRediT authorship contribution statement

Qiqi Zhou: Investigation, Data curation, Writing – original draft. Lu Gao: Validation, Investigation. Yannan Ji: Methodology. Xiaoding Zhang: Methodology, Investigation. Ningning Shi: Methodology. Jia Liu: Resources. Pengbo Tang: Resources. Haixia Gao: Funding acquisition. Changhong Huo: Project administration, Writing – review & editing, Funding acquisition.

Acknowledgments

This work is supported by the National Natural Science Foundation of China (81872770, 81871027) and the Natural Science Foundation of Hebei Province (H2020206596).

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. , , , , , . Parthenolide phytosomes attenuated gentamicin-induced nephrotoxicity in rats via activation of Sirt-1, Nrf2, OH-1, and NQO1 axis. Molecules. 2023;28(6):2741.
    [CrossRef] [Google Scholar]
  2. , , , , , , , . Antimicrobial garlic-derived diallyl polysulfanes: Interactions with biological thiols in Bacillus subtilis. Biochim. Biophys. Acta. Gen. Subj.. 2019;1863(6):1050-1058.
    [CrossRef] [Google Scholar]
  3. , , , , . Evidence that hydrogen sulfide is a genotoxic agent. Mol. Cancer Res.. 2006;4(1):9-14.
    [CrossRef] [Google Scholar]
  4. , , , , , , , , , . Hydrogen sulfide mediates the vasoactivity of garlic. Proc. Natl. Acad. Sci. U S A.. 2007;104(46):17977-17982.
    [CrossRef] [Google Scholar]
  5. , , , , , , . Production of hydrogen sulfide by the intestinal microbiota and epithelial cells and consequences for the colonic and rectal mucosa. Am. J. Physiol. Gastrointest. Liver Physiol.. 2021;320(2):G125-G135.
    [CrossRef] [Google Scholar]
  6. , , , , . Effects of sulfane sulfur content in benzyl polysulfides on thiol-triggered H2S release and cell proliferation. Free Radic. Biol. Med.. 2019;131:393-398.
    [CrossRef] [Google Scholar]
  7. , , , , , , , , , , , , . Synthetic polysulfane derivatives induce cell cycle arrest and apoptotic cell death in human hematopoietic cancer cells. Food Chem. Toxicol.. 2014;64:249-257.
    [CrossRef] [Google Scholar]
  8. , , , , , , , , , . Novel targeted nano-parthenolide molecule against NF-kB in acute myeloid leukemia. Molecules. 2019;24(11):2103.
    [CrossRef] [Google Scholar]
  9. , , , , , , . Orally administered allyl sulfides from garlic ameliorate murine colitis. Mol. Nutr. Food Res.. 2015;59(3):434-442.
    [CrossRef] [Google Scholar]
  10. , , , , . Chemical biology of H2S signaling through persulfidation. Chem. Rev.. 2018;118(3):1253-1337.
    [CrossRef] [Google Scholar]
  11. , , , , . Advances in chemistry and bioactivity of parthenolide. Nat. Prod. Rep.. 2020;37(4):541-565.
    [CrossRef] [Google Scholar]
  12. , , , , , , , . Nano magnetic liposomes-encapsulated parthenolide and glucose oxidase for ultra-efficient synergistic antitumor therapy. Nanotechnology. 2020;31(35):355104
    [CrossRef] [Google Scholar]
  13. , , , , , , , , , . Nano parthenolide improves intestinal barrier function of sepsis by inhibiting apoptosis and ROS via 5-HTR2A. Int. J. Nanomedicine.. 2023;18:693-709.
    [CrossRef] [Google Scholar]
  14. , , , , , , , , , , , , , , , , . Reactive cysteine persulfides and S-polythiolation regulate oxidative stress and redox signaling. Proc. Natl. Acad. Sci. U S A.. 2014;111(21):7606-7611.
    [CrossRef] [Google Scholar]
  15. , , , . Sesquiterpene lactones from Michelia champaca. Phytochemistry. 1995;39(4):839-843.
    [CrossRef] [Google Scholar]
  16. , , , , , , . Development of hydropersulfide donors to study their chemical biology. Antioxid. Redox Signal.. 2022;36(4–6):309-326.
    [CrossRef] [Google Scholar]
  17. , , , , , , . Polysulfides are possible H2S-derived signaling molecules in rat brain. Faseb. J.. 2013;27(6):2451-2457.
    [CrossRef] [Google Scholar]
  18. , , , , . Diallyl trisulfide is a fast H2S donor, but diallyl disulfide is a slow one: the reaction pathways and intermediates of glutathione with polysulfides. Org. Lett.. 2015;17(17):4196-4199.
    [CrossRef] [Google Scholar]
  19. , , , , , , , , , , , , , . Parthenolide ameliorates colon inflammation through regulating Treg/Th17 balance in a gut microbiota-dependent manner. Theranostics.. 2020;10(12):5225-5241.
    [CrossRef] [Google Scholar]
  20. , , , , , , , . Garlic constituents for cancer prevention and therapy: From phytochemistry to novel formulations. Pharmacol. Res.. 2022;175:105837
    [CrossRef] [Google Scholar]
  21. , , , , , , , , . Fluorescent probes based on nucleophilic substitution-cyclization for hydrogen sulfide detection and bioimaging. Chemistry. 2014;20(4):1010-1016.
    [CrossRef] [Google Scholar]
  22. , , , , , , , , , . Hydrogen sulfide protects against DSS-induced colitis by inhibiting NLRP3 inflammasome. Free Radic. Biol. Med.. 2019;137:99-109.
    [CrossRef] [Google Scholar]
  23. , . The chemistry of organic polysulfanes R-Sn-R (n > 2) Chem. Rev.. 2002;102(11):3905-3945.
    [CrossRef] [Google Scholar]
  24. , , , , , , , , . Continuous separation and preparation of parthenolide from Chrysanthemum morifolium with high-speed counter-current chromatography. Shandong Sci.. 2016;29(03):18-22.
    [CrossRef] [Google Scholar]
  25. , . Physiological implications of hydrogen sulfide: a whiff exploration that blossomed. Physiol. Rev.. 2012;92(2):791-896.
    [CrossRef] [Google Scholar]
  26. , , , . The role of H2S in the gastrointestinal tract and microbiota. Adv. Exp. Med. Biol.. 2021;1315:67-98.
    [CrossRef] [Google Scholar]
  27. , , , , , , , , , , , , , . Tubulin-binding anticancer polysulfides induce cell death via mitotic arrest and autophagic interference in colorectal cancer. Cancer Lett.. 2017;410:139-157.
    [CrossRef] [Google Scholar]
  28. , , , , , , , , . Parthenolide from Parthenium integrifolium reduces tumor burden and alleviate cachexia symptoms in the murine CT-26 model of colorectal carcinoma. Phytomedicine. 2013;20(11):992-998.
    [CrossRef] [Google Scholar]
  29. , , , , , , , , , . Study on the metabolites of isoalantolactone in vivo and in vitro by ultra performance liquid chromatography combined with Triple TOF mass spectrometry. Food Chem.. 2017;214:328-338.
    [CrossRef] [Google Scholar]
  30. , , , , , . Rapid screening and identification of metabolites of murpanicin in rats by UHPLC/Q-TOF-MS/MS combined with diagnostic fragment ions (DFIs) and multiple mass defect filter. J. Pharm. Biomed. Anal.. 2022;213:114679
    [CrossRef] [Google Scholar]
  31. , , , , , . Determination of parthenolide in rat plasma by UPLC-MS/MS and its application to a pharmacokinetic study. J. Pharm. Biomed. Anal.. 2016;119:99-103.
    [CrossRef] [Google Scholar]
  32. , , , , , , , , . Comparison on contents of parthenolide in leaves of eight Magnoliaceae species and twenty-three Asteraceae species. J. Plant Resour. Environ.. 2022;31(02):85-87.
    [CrossRef] [Google Scholar]

Appendix A

Supplementary material

The supporting information showed the corresponding NMR spectra and MS spectra of organosulfur metabolites and reaction products. The analytical process of common Ⅰ phase metabolites were described in supporting information. The detailed cytotoxic data of PA and organosulfur compounds were also displayed. Supplementary data to this article can be found online at https://doi.org/10.1016/j.arabjc.2023.105004.

Appendix A

Supplementary material

The following are the Supplementary data to this article:

Supplementary data 1

Supplementary data 1


Fulltext Views
415

PDF downloads
298
View/Download PDF
Download Citations
BibTeX
RIS
Show Sections

Suggested read for related articles: