Comprehensive phytochemical study of Galianthe thalictroides: isolation of three new compounds, structural revision of β-carbolines, and evaluation of cytotoxic and anti-wnt activities

The medicinal plant G. thalictroides (K. Schum.) E. L. Cabral has been traditionally used in Midwestern Brazil for cancer treatment and prevention. Previous studies demonstrated that the chloroform fraction of its roots exhibits significant cytotoxic activity against cancer cell lines. Phytochemical investigation of this fraction led to the isolation of four cytotoxic compounds, including two β-carboline alkaloids and two cyclopeptides, supporting the traditional use of this species. However, several minor metabolites detected in this active fraction were not obtained in sufficient quantities for complete structural elucidation or biological evaluation. In this context, the present study aimed to expand the chemical characterization of Galianthe thalictroides, with a particular focus on previously unresolved metabolites. Additionally, the study sought to explore other extracts and preparations, including the traditional decoction, to provide a more comprehensive understanding of the phytochemical profile and its potential biological relevance. The roots of G. thalictroides were extracted with ethanol and a traditional water decoction was prepared in parallel. The ethanolic extract was subjected to liquid/liquid partitioning with hexane, chloroform, and ethyl acetate. The extracts, fractions, and traditional decoction were analyzed by ultra-high-performance liquid chromatography coupled with high-resolution mass spectrometry (UHPLC-HRMS/MS) to establish a comprehensive metabolite profile and generate a molecular networking (MN). The chloroform fraction was purified using medium-pressure liquid chromatography (MPLC-UV) followed by semi-preparative HPLC. Structural characterization of the isolated compounds was achieved through UV, Electronic Circular Dichroism, Nuclear Magnetic Resonance, and High-Resolution Mass Spectrometry data. The chemical shifts and coupling constants of β-carboline derivatives were calculated using the CT web application, which performs a complete Quantum Mechanical Spectral Analysis. Finally, the isolated compounds were evaluated for activity in a dose-dependent manner using BT-20 triple-negative breast cancer cells. This was conducted to determine their cytotoxic potencies and specific Wnt-inhibitory activities via a TopFlash reporter assay. Using this approach, 26 compounds were isolated from the chloroform fraction and fully characterized. Notably, this study led to the discovery of two new cyclopeptides and one new β-carboline alkaloid, which represent the most significant and original contributions of this work. In addition, the structures of two previously reported β-carboline alkaloids from this species were revised. Molecular networking–guided MS profiling of the extracts, fractions, and traditional decoction enabled the annotation of 13 additional metabolites, thereby expanding the known chemical diversity of G. thalictroides. Biological evaluation confirmed the cytotoxic activity of known compounds and revealed new active analogues. Bouvardins exhibited strong cytotoxicity (IC 50 0.25–14.00 µM), while anthraquinones and β-carboline alkaloids showed moderate activity (IC 50 23–110 μM and 65–193 μM, respectively). Mechanistic studies revealed that β-carboline derivative 24 selectively inhibited the Wnt signaling pathway (IC 50 ≈ 40 µM), whereas its analogue (25) was inactive. Overall, this study significantly advances the chemical and biological understanding of G. thalictroides by integrating compound isolation, structural characterization, and biological evaluation. The discovery of novel bioactive compounds, together with evidence of pathway-specific effects, highlights the plant’s potential as a source of structurally diverse and pharmacologically relevant metabolites for anticancer research.

The species Galianthe thalictroides (K. Schum.) E. L. Cabral (Rubiaceae) is traditionally used in Mato Grosso do Sul, Brazil, for the treatment and prevention of cancer (Cabral, 2002; Cabral, 2009; Figueiredo et al., 2011; Figueiredo et al., 2016). This ethnopharmacological relevance highlights its potential as a source of bioactive metabolites and provides a strong rationale for further phytochemical and biological investigation.

Consistent with this traditional use, phytochemical studies have begun to uncover bioactive constituents within the species. Previous studies on G. thalictroides have focused mainly on its roots, from which several secondary metabolites have been isolated and characterized. These investigations led to the identification of cytotoxic constituents, particularly β-carboline alkaloids and cyclopeptides (Figueiredo et al., 2011; Figueiredo et al., 2014; Figueiredo et al., 2016). However, despite these important contributions, the chemical profile of the species remains incompletely explored, and several minor constituents have not yet been characterized because of their low abundance.

Among the reported compounds, β-carboline derivatives are of particular interest due to their biological relevance. This scaffold has been shown to modulate Wnt signaling, a pathway that plays a central role in cell proliferation, differentiation, stemness, and tumor progression (Logan and Nusse, 2004; Li et al., 2015; Nusse and Clevers, 2017). Aberrant activation of Wnt signaling is frequently associated with cancer development, therapy resistance, and maintenance of cancer stem cells, making this pathway a major target in anticancer drug discovery. In this context, natural products have emerged as promising sources of anti-Wnt candidates (Blagodatski et al., 2014; Shaw et al., 2019; Blagodatski et al., 2020).

Despite this therapeutic potential, the presence of other bioactive constituents with distinct biological profiles highlights the complexity of this species. In particular, previously reported cyclopeptides from G. thalictroides raise potential safety concerns, as some of these compounds have been associated with toxicity (Figueiredo et al., 2016). This further supports the need for a more comprehensive investigation of the chemical composition and biological properties of this species.

Accordingly, this study extends the work of Figueiredo et al. by conducting an in-depth phytochemical analysis of G. thalictroides roots, including extracts, fractions, and traditional decoctions. Ultra-high-performance liquid chromatography-high-resolution mass spectrometry (UHPLC–HRMS/MS) combined with molecular networking (MN) was employed to characterize the metabolome, while selected minor constituents were isolated, structurally elucidated, and evaluated for their anticancer potential using cytotoxicity (MTT) and anti-Wnt signaling assays.

General experimental procedures

The UV and ECD spectra were recorded on a JASCO J-815 spectrometer (Loveland, CO, United States) in MeCN, using a 1 cm cell. The scan speed was set at 200 nm/min in continuous mode between 500 and 190 nm, with a bandwidth of 1 nm, a data pitch of 0.1 nm, and 2 accumulations. The chiral compounds were dissolved in 1 mL of spectrometric grade solvent, and the optical rotation of plane-polarized light was measured using a Perkin-Elmer 341 electronic polarimeter. The rotation was measured at 589, 578, or 546 nm, and the specific optical rotation values were calculated. 1 H and 13 C NMR data were recorded on an Agilent Varian Inova 500 MHz NMR spectrometer (Palo Alto, CA, United States of America) (499.9 and 125.7 MHz, respectively) or on a Bruker Avance Neo 600 MHz NMR spectrometer equipped with a QCI 5mmCryoprobe and a SampleJet automated sample changer (Bruker BioSpin, Rheinstetten, Germany) (600.2 and 150.9 MHz, respectively). Chemical shifts are reported in parts per million (δ) using the residual solvent signal at δ H 3.31 and δ C 49.0 for CD 3 OD, δ H 2.50 and δ C 39.5 for DMSO- d 6 and δ H 7.26 and δ C 77.2 for CDCl 3 as an internal standard for 1 H and 13 C NMR, respectively, and coupling constants (J) are reported in Hz. Complete assignments were obtained using 2D experiments including COSY, edited-HSQC, HMBC, and NOESY. The chemical shifts and coupling constants of compounds 23-25 were determined using the CT web application (https://ctb.nmrsolutions.fi/), which performs complete Quantum Mechanical Spectral Analysis (QMSA). This tool iteratively optimizes chemical shifts (δ) and coupling constants (J) to minimize differences between experimental and quantum mechanically calculated 1 H NMR spectra, generating high-fidelity HiFSA profiles for precise structural validation (Bhavaraju et al., 2021). All relevant spectral parameters like chemical shifts and coupling constants were thus generated and reported in the description of isolated compounds section. HRESIMS data were obtained on a Micromass-LCT Premier time of flight (TOF) mass spectrometer (Waters, Milford, MA, United States of America) with an electrospray interface and coupled with an Acquity UPLC system (Waters). HPLC-PDA-ELSD data were obtained with an Agilent 1,100 series system (Santa Clara, CA, United States of America) consisting of an autosampler, high-pressure mixing pump, and PDA detector connected to an ELSD (Sedex 85, Sedere Omnilab). MPLC separation was performed using a Buchi 681 pump equipped with a Knauer UV detector and a 460 × 70 mm i.d. column (Büchi, Flawil, Switzerland) loaded with Zeoprep C18 15–25 μm as the stationary phase (Zeochem, Uetikon am See, Switzerland). HPLC-fractionation was performed with an Armen modular spot prep II (Saint-Avé, France) equipped with an UV detector and fraction collector and on a Shimadzu system equipped with an LC-20A module pump, an SPD-20A UV/VIS, a 7725I Rheodyne® valve, and an FRC-10A fraction collector (Shimadzu, Kyoto, Japan). The system was controlled by LabSolutions software, also from Shimadzu.

Plant material

Roots of G. thalictroides were collected in Campo Grande, Mato Grosso do Sul, Brazil, in September 2013. The botanical identification was performed by Prof. Arnildo Pott (Faculty of Pharmaceutical Sciences, Food, and Nutrition - FACFAN, Federal University of Mato Grosso do Sul, Brazil). A voucher specimen (N o. 29003) was deposited at the CGMS Herbarium, Federal University of Mato Grosso do Sul. Access to the genetic heritage was registered under CNPq Process No. 010065/2015-6 prior to the implementation of National System for the Administration of Genetic Heritage and Associated Traditional Knowledge (SisGen) and was subsequently registered in SisGen under N o. A7BF8FF.

Extraction procedures

Air-dried, powdered roots (4.0 kg) of G. thalictroides were extracted with EtOH (5 × 2 L) at room temperature. After concentration in vacuo, the EtOH extract (ER, 81.0 g) was submitted to a liquid/liquid partition between MeOH/H 2 O (9:1; 1,500 mL) and hexane (5 × 300 mL) (fraction ERH); MeOH/H 2 O (1:1; 1,500 mL) and CHCl 3 (8 × 300 mL) (fraction ERC); and MeOH/H 2 O (1:1; 1,500 mL) and EtOAc (5 × 300 mL) (fraction ERA), and leaving the residual hydromethanolic fraction (ERHM). The extracts were concentrated under pressure and later lyophilized to give 6.1 g (Hexane fraction), 13.8 g (chloroform fraction), 38.5 g (EtOAc fraction), and 21.8 g (hydromethanolic fraction). For the metabolomic analyses, the traditional decoction of the plant was also studied. It was prepared in the same way as practiced by the local population from Mato Grosso do Sul, Brazil, for cancer treatment and prevention. It consisted of an infusion of 0.6 g (a half teaspoon) of dried and mild G. thalictroides roots in 500 mL of boiling water. After 15 min, the extract was filtered and then consumed (a water cup, twice a day). For the chemical study, this decoction was filtered and then frozen and lyophilized, resulting in the traditional decoction: TD (56.4 mg).

UHPLC-HRMS/MS analyses

UHPLC-HRMS/MS analyses were performed with a Waters Acquity UPLC system equipped with a PDA detector and hyphenated to a Q-Exactive Focus mass spectrometer (Thermo ScientificTM, Bremen, Germany), employing a heated electrospray ionization source (HESI-II). The system was coupled to a Charged aerosol detector (CAD, Thermo ScientificTM, Bremen, Germany). Control of the instruments was done using Thermo Scientific Xcalibur 3.1 software. The chromatographic separation was done on a Waters BEH C18 column (50 × 2.1 mm, 1.7 µm, Waters, Milford, MA, United States of America), through a linear gradient of 5–100% B over 7 min and an isocratic step at 100% B for 1 min. The mobile phases were: (A) water with 0.1% formic acid and (B) acetonitrile with 0.1% formic acid. The flow rate was set to 600 μL/min, the injection volume was 2 µL (concentration around 5 mg/mL) and the column was kept at 40 °C. ESI parameters were as follows: source voltage, 3.5 kV (pos) and 4.0 kV (neg); sheath gas flow rate (N 2), 55 units; auxiliary gas flow rate, 15 units; spare gas flow rate, 3.0; capillary temperature, 350.00 °C, S-Lens RF Level, 45. The mass analyzer was calibrated using a mixture of caffeine, methionine–arginine–phenylalanine–alanine–acetate (MRFA), sodium dodecyl sulfate, sodium taurocholate, and Ultramark 1,621 in an acetonitrile/methanol/water solution containing 1% formic acid by direct injection. The data-dependent MS/MS (DDA) events were performed as a top3 experiment (the three most intense ions detected in full scan MS). The MS/MS isolation window width was 1 Da, and the stepped normalized collision energy (NCE) was set to 15, 30, and 45 units. In data-dependent MS/MS experiments, full scans were acquired at a resolution of 35,000 FWHM (at m/z 200) and MS/MS scans at 17,500 FWHM both with an automatically determined maximum injection time. After being acquired in an MS/MS scan, parent ions were placed in a dynamic exclusion list for 2.0 s. The PDA wavelength range was from 210 nm to 400 nm with a resolution of 1.2 nm, and the CAD was kept at 40 C. All HRMS runs, in both positive and negative ionization mode, were converted from RAW (Thermo) standard data format to.mzXML format using the MS Convert software (Chambers et al., 2012). The files were treated in MZmine v. 2.53 (Pluskal et al., 2010). The parameter used were as follows: for mass detection, the MS 1 noise level of 2.0 E 4 for positive mode and 1.0 E 4 for negative mode. For MS 2 level 0.0 E 0 was set for both ionization modes. The ADAP chromatogram builder was used and set to a minimum group size of scans of 5, a minimum group intensity threshold of 2.0 E 4 (1.0 E 4 negative), and a minimum highest intensity of 2.0 E 4 (1.0 E 4 negative), and m/z tolerance of 0.005 m /z or 8.0 ppm. Chromatogram deconvolution was performed using the ADAP algorithm (wavelets). The intensity window S/N was used as an S/N estimator with a signal-to-noise ratio set at 10, a minimum feature height at 2.0 E 4 (1.0 E 4 negative), a coefficient area threshold at 110, a peak duration range from 0.01 to 1.0 min, and the RT wavelet range from 0.01 to 0.08 min. The m/z and RT ranges for MS 2 scans paring were set to 0.025 and 0.15, respectively. Isotopes were detected using the isotopes peaks grouper with an m/z tolerance of 5.0 ppm, an RT tolerance of 0.05 min (absolute), the maximum charge set at 1, and the representative isotope used was the most intense. The resulting file was filtered using the peak-list rows filter option to remove features without an MS 2 spectrum associated.

Molecular networking and MS-based annotations

A molecular networking was created from the mgf file exported from MZmine, using the online workflow (https://ccms-ucsd.github.io/GNPSDocumentation/) on the GNPS website (http://gnps.ucsd.edu) (Wang et al., 2016). The precursor ion mass tolerance was set to 0.02 Da, and an MS/MS fragment ion tolerance of 0.02 Da. A network was then created where edges were filtered to have a cosine score above 0.7 and more than 4 matched peaks. The spectra in the network were then searched against GNPS’ spectral libraries. The library spectra were filtered in the same manner as the input data. All matches kept between network spectra and library spectra were required to have a score above 0.7 and at least 4 matched peaks. The works are available in the following links: https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=7c74700c7318428d8d7d25773f39e484. Visualization of results was done in Cytoscape 3.8.0 (Shannon et al., 2003). The output of the GNPS was used to annotate against the in silico ISDB-DNP (Allard et al., 2016), and then the script for taxonomically informed metabolite annotation (Rutz et al., 2019) was used to re-rank and clean the output based on the species.

Isolation of compounds 1-26 using MPLC combined with semi-preparative HPLC-UV

The CHCl 3 fraction (6.0 g) was fractionated using reversed-phase MPLC with a 460 × 70 mm i.d. column (Büchi, Flawil, Switzerland), loaded with Zeoprep C18 15–25 μm as the stationary phase (Zeochem, Uetikon am See, Switzerland). The separation was performed with a mixture of MeOH and H 2 O with 0.1% formic acid as a mobile phase a follow: linear elution of MeOH 15% (0–0.5 h) and a gradient elution from MeOH 15%–30% (0.5–12.2 h), 30%–40% (12.2–16.1 h), linear elution of 40% MeOH (16.1–24.0 h), gradient elution from MeOH 40%–65% (24.0–35.9h) and then a linear elution of 100% MeOH (35.9–37.0 h). The flow rate was 20 mL/min, and the UV absorbance was detected at 254 nm and 280 nm. The separation conditions were optimized at the analytical HPLC level with the same stationary phase chemistry and geometrically transferred by gradient transfer (Challal et al., 2015). The MPLC generated 82 fractions (EC1 to EC 82). Fraction EC-9 consisted of a mixture of 1 and 2 (82.0 mg), and fractions EC-12, EC-15, EC-18, EC-31, EC-41, EC-50, EC-66 and EC-78 consisted of pure 2 (254.6 mg), 3 (213.8 mg), 4 (78.7 mg), 8 (45.9 mg), 11 (34.7 mg), 12 (29.6 mg), 18 (36.1 mg), and 20 (37.2 mg), respectively. The final purification of other MPLC fractions was done using different conditions by semi-preparative HPLC-UV on an Agilent Eclipse XDB Phenyl column (250 × 9.4 mm i.d., 5 μm). Fractionation of EC-(23–29) (223.5 mg) (performed with H 2 O:CH 3 CN 90:10 with 0.1% formic acid as eluent at a flow rate of 4 mL min -1 yielded fractions EC-(23-29)-1 being a mixture of 5 and 7 (2.7 mg), EC-(23-29)-2 as 5 (6.3 mg) and EC-(23-29)-4 yielded 6 (7.4 mg). Fraction EC-(33–37) (193.1 mg) and EC-(57–61) (70.9 mg) were separately fractioned, eluted with isocratic H 2 O:CH 3 CN 75:25 with 0.1% formic at a flow rate of 3 mL min -1 and 4 mL min -1, respectively. Fractions EC-(33-37)-9 yielded 9 (5.1 mg), EC-(33-37)-16 yielded 10 (3.0 mg), EC-(57-61)-2 as a mixture of 13 and 14 (4.7 mg), EC-(57-61)-3 yielded a mixture of 14 and 15 (3.1 mg) and EC-(57-61)-6 yielded 16 (1.9 mg). The purification of EC-(68–69) (38.9 mg) resulted in the isolation of 17 in EC-(68-69)-2 (1.0 mg), after elution with H 2 O:CH 3 CN 73:27 with 0.1% formic acid as eluent at a flow rate of 4 mL min -1. Compound 19 (2.0 mg) was identified in fraction EC-(71-72)-11, after fractionation of EC-(71–72) (130.0 mg) in isocratic H 2 O:CH 3 CN 74:26 with 0.1% formic acid at a flow rate of 4 mL min -1. Fractions EC-79 (11.4 mg) and EC-80 (29.3 mg) were separately fractioned with in H 2 O:CH 3 CN 70:30 at a flow rate of 4 mL min -1, resulting in compound 21 in EC-79-6 (1.4 mg), EC-80-9 (3.3 mg) and EC-80-10 (12.9 mg). Fraction EC-82 (141 mg), after gradient elution of 3 mL min -1 with H 2 O:CH 3 CN 95:5 to CH 3 CN 100% (0–30 min) resulted in the identification of 22 (5.4 mg) 23 (1.2 mg), 24 (9.4 mg), 25 (3.0 mg), and 26 (2.8 mg).

Electronic circular dichroism calculations (ECD)

The absolute configuration of compounds 24 and 25 was based on a comparison between the calculated and experimental ECD (see Figure 3). The calculations were based on the relative configuration determined through NMR 2D ROESY experiments. The structures were used to find the conformers through a random rotor search algorithm (number of conformers, 100) employing the MMFF94s force field in Avogadro v1.2.0 (Hanwell et al., 2012). The conformers were further optimized using PM3 and B3LYP/6-31G (d,p) basis sets in Gaussian 16 software (© 2015-2022, Gaussian Inc., Wallingford, CT, United States of America) with the SCRF model in methanol (Nugroho and Morita, 2014; Mandi and Kurtan, 2019). All optimized conformers were checked for imaginary frequencies. The conformers were subjected to ECD calculations using TD-DFT B3LYP/def2svp as a basis set and an SCRFmodelinmethanolinGaussian16 software. The calculated ECD spectrum was generated in SpecDisv1.71 software (Berlin, Germany). The experimental and calculated spectra are available in the supplementary material. The ECD calculations were performed on the HPC Baobab cluster at the University of Geneva. Graphs were made using GraphPad Prism (version 10.2.3) and chemical structures were drawn with ChemDraw (version 23.0.1), both licensed to OK by the University of Geneva.

Description of isolated compounds

Homovanillyl alcohol 4- O -β-D-glucopyranoside (1): 1 H NMR (CD 3 OD, 500 MHz) δ 2.77 (2H, t, J = 7.0 Hz, H 2 -7), 3.39 (2H, m, H-3′, H-4′), 3.47 (2H, m, H-2′, H-5′), 3.69 (1H, dd, J = 12.6, 5.0 Hz, H-6′b), 3.73 (2H, t, J = 6.9 Hz, H 2 -8), 3.86 (3H, s, CH 3 O-3), 3.88 (1H, m, H-6′a), 6.77 (1H, dd, J = 8.2, 1.9 Hz, H-6), 6.89 (1H, d, J = 1.9 Hz, H-2), 7.09 (1H, d, J = 8.2 Hz, H-5); 13 C NMR (CD 3 OD, 126 MHz) δ 39.8 (C-7), 56.7 (CH 3 O-3), 62.5 (C-6′), 64.3 (C-8), 71.3 (C-4′), 74.9 (C-2′), 77.8 (C-5′), 78.2 (C-3′), 103.1 (C-1′), 114.6 (C-2), 118.3 (C-5), 122.5 (C-6), 135.5 (C-1), 146.4 (C-4), 150.7 (C-3). ESI-HRMS m/z 331.1385 [M + H] + (calculated for C 14 H 21 O 8 =, 331.1390, Δppm = −0.6).

Feretoside (scandoside methyl ester) (2): [α] 23 D - 41.8 (c 1.22, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ 3.00 (1H, m, H-5), 3.04 (1H, t, J = 7.1 Hz, H-9), 3.21 (1H, t, J = 8.5, 7.9 Hz, H-2′), 3.28 (2H, m, H-4′, H-5′), 3.37 (1H, t, J = 8.5 Hz, H-3′), 3.64 (1H, dd, J = 12.0, 5.1 Hz, H-6′b), 3.76 (3H, s, CH 3 O-11), 3.87 (1H, d, J = 12.0 Hz, H-6′a), 4.19 (1H, d, J = 15.2 Hz, H-10b), 4.34 (1H, dt, J = 15.2, 1.9 Hz, H-10a), 4.55 (1H, dt, J = 4.5, 2.1 Hz, H-6), 4.67 (1H, d, J = 7.9 Hz, H-1′), 5.19 (1H, d, J = 6.3 Hz, H-1), 5.80 (1H, m, H-7), 7.51 (1H, d, J = 1.2 Hz, H-3); 13 C NMR (CD 3 OD, 126 MHz) δ 45.3 (C-5), 46.9 (C-9), 51.8 (CH 3 O-11), 60.8 (C-10), 62.5 (C-6′), 71.2 (C-4′), 74.5 (C-2′), 77.6 (C-3′), 78.1 (C-5′), 82.0 (C-6), 98.0 (C-1), 100.1 (C-1′), 110.6 (C-4), 129.8 (C-7), 147.3 (C-8), 153.6 (C-3), 170.0 (C-11). ESI-HRMS m/z 405.1394 [M + H] + (calculated for C 17 H 25 O 11, 405.1387, Δppm = −0.9).

Asperulosidic acid (3): [α] 23 D +21.6 (c 1.02, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ 2.09 (3H, s, H 3 -13), 2.64 (1H, t, J = 8.2 Hz, H-9), 3.02 (1H, ddd, J = 7.6, 6.1, 1.6 Hz, H-5), 3.26 (3H, m, H-2′, H-4′, H-5′), 3.38 (1H, t, J = 8.7 Hz, H-3′), 3.62 (1H, dd, J = 12.1, 5.7 Hz, H-6′b), 3.86 (1H, dd, J = 12.1, 1.9 Hz, H-6′a), 4.73 (1H, d, J = 7.9 Hz, H-1′), 4.82 (2H, m, H-6, H-10b), 4.94 (1H, dd, J = 14.7, 1.6 Hz, H-10a), 5.07 (1H, d, J = 9.0 Hz, H-1), 6.02 (1H, d, J = 2.1 Hz, H-7), 7.66 (1H, d, J = 1.7 Hz, H-3); 13 C NMR (CD 3 OD, 126 MHz) δ 20.5 (CH 3 -13), 42.3 (C-5), 46.1 (C-9), 62.7 (C-6′), 63.5 (C-10), 71.4 (C-4′), 74.7 (C-2′), 75.1 (C-6), 77.7 (C-3′), 78.3 (C-5′), 100.3 (C-1′), 101.0 (C-1), 108.2 (C-4), 131.7 (C-7), 145.7 (C-8), 155.2 (C-3), 170.6 (C-11), 172.3 (C-12). ESI-HRMS m/z 455.1166 [M + Na] + (calculated for C 18 H 24 O 12 Na, 455.1165, Δppm = 0.2).

Homovanillyl alcohol (4): 1 H NMR (CD 3 OD, 500 MHz) δ 2.73 (2H, t, J = 7.1 Hz, H 2 -7), 3.71 (2H, t, J = 7.1 Hz, H 2 -8), 3.84 (3H, s, CH 3 O-3), 6.65 (1H, dd, J = 8.0, 1.9 Hz, H-6), 6.71 (1H, d, J = 8.0 Hz, H-5), 6.80 (1H, d, J = 1.9 Hz, H-2); 13 C NMR (CD 3 OD, 126 MHz) δ 39.8 (C-7), 56.3 (CH 3 O-3), 64.5 (C-8), 113.7 (C-2), 116.1 (C-5), 122.4 (C-6), 131.8 (C-1), 145.9 (C-4), 148.8 (C-3). ESI-HRMS m/z 191.0694 [M + Na] + (calculated for C 9 H 12 O 3 Na, 191.0684, Δppm = 5.2).

Syringic acid (5): 1 H NMR (CD 3 OD, 500 MHz) δ 3.88 (6H, s, CH 3 O-3, CH 3 O-5), 7.33 (2H, s, H-2, H-6); 13 C NMR (CD 3 OD, 126 MHz) δ 56.3 (CH 3 O-3, CH 3 O-5), 108.0 (C-2, C-6), 122.8 (C-1), 108.0 (C-2), 141.1 (C-4), 148.6 (C-3, C-5), 170.5 (C-7). ESI-HRMS m/z 197.0451 [M–H] - (calculated for C 9 H 9 O 5, 197.0450, Δppm = 0.5).

Protocatechuic acid methyl ester (6): 1 H NMR (CD 3 OD, 500 MHz) δ 3.83 (3H, s, CH 3 O-7), 6.80 (1H, d, J = 8.9 Hz, H-5), 7.41 (2H, m, H-2, H-6); 13 C NMR (CD 3 OD, 126 MHz) δ 52.2 (CH 3 O-7), 115.8 (C-5), 117.4 (C-2), 122.6 (C-1), 123.6 (C-6), 146.2 (C-3), 151.7 (C-4), 168.8 (C-7). ESI-HRMS m/z 169.0498 [M + H] + (calculated for C 8 H 9 O 4, 169.0497, Δppm = 0.1).

Vanillic acid (7): 1 H NMR (CD 3 OD, 500 MHz) δ 7.55 (2H, m, H-2, H-6), 6.83 (1H, d, J = 8.0 Hz, H-5), 3.89 (3H, s, CH 3 O-3); 13 C NMR (CD 3 OD, 126 MHz) δ 56.3 (CH 3 O-3), 113.5 (C-2), 115.5 (C-5), 123.5 (C-1), 124.9 (C-6), 148.3 (C-3), 152.1 (C-4), 170.5 (−7). ESI-HRMS m/z 169.0498 [M + H] + (calculated for C 8 H 9 O 4, 169.0458, Δppm = 0.3).

Scopoletin (8): 1 H NMR (CD 3 OD, 500 MHz) δ 3.91 (3H, s, CH 3 O-6), 6.21 (1H, d, J = 9.4 Hz, H-3), 6.77 (1H, s, H-8), 7.11 (1H, s, H-5), 7.86 (1H, d, J = 9.4 Hz, H-4); 13 C NMR (CD 3 OD, 126 MHz) δ 56.6 (CH 3 O-6), 103.7 (C-8), 109.7 (C-5), 112.3 (C-10), 112.4 (C-3), 145.9 (C-4), 146.9 (C-6), 151.2 (C-9), 152.7 (C-7), 163.9 (C-2). ESI-HRMS m/z 193.0502 [M + H] + (calculated for C 10 H 9 O 4, 193.0501, Δppm = 0.5).

Ciwujiatone (9): [α] 23 D −13.2 (c 0.68, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ 3.30 (1H, overlapped, H-8′), 3.80 (2H, m, H 2 -9′), 3.87 (6H, s, CH 3 O-3′, CH 3 O-5′), 3.93 (6H, s, CH 3 O-3, CH 3 O-5), 4.23 (2H, m, H 2 -9), 4.37 (1H, d, J = 7.6 Hz, H-8), 6.77 (2H, s, H-2′, H-6′), 7.36 (2H, s, H-2, H-6); 13 C NMR (CD 3 OD, 126 MHz) δ 49.8 (C-8), 56.5 (CH 3 O-3′, CH 3 O-5′), 56.8 (CH 3 O-3, CH 3 O-5), 59.8 (C-8′), 69.3 (C-9′), 71.4 (C-9), 85.1 (C-7′), 105.0 (C-2′, C-6′), 107.6 (C-2, C-6), 128.6 (C-1), 132.2 (C-1′), 136.1 (C-4), 142.6 (C-4′), 149.0 (C-3, C-5), 149.1 (C-3′, C-5′), 200.3 (C-7). ESI-HRMS m/z 435.1652 [M + H] + (calculated for C 22 H 27 O 9, 435.1649, Δppm = −0.05).

(+)-Cyclolariciresinol (10): [α] 23 D + 10.5 (c 0.65, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ 1.76 (1H, tt, J = 10.3, 3.6 Hz, H-8′), 2.00 (1H, m, H-8), 2.78 (2H, d, J = 7.7 Hz, H-7), 3.40 (1H, dd, J = 11.3, 4.1 Hz, H-9′b), 3.67 (2H, m, H-9′a, H-9b), 3.71 (1H, dd, J = 10.9, 4.8 Hz, H-9a), 3.78 (3H, s, CH 3 O-3′), 3.81 (1H, d, J = 10.0 Hz, H-7′), 3.81 (3H, s, CH 3 O-3), 6.18 (1H, s, H-5), 6.62 (1H, dd, J = 8.0, 2.0 Hz, H-6′), 6.66 (1H, s, H-2), 6.68 (1H, d, J = 2.0 Hz, H-2′), 6.74 (1H, d, J = 8.0 Hz, H-5′); 13 C NMR (CD 3 OD, 126 MHz) δ 33.3 (C-7), 39.7 (C-8), 47.7 (C-8′), 47.8 (C-7′), 56.1 (CH 3 O-3, CH 3 O-3′), 62.0 (C-9′), 65.6 (C-9), 112.1 (C-2), 113.5 (C-2′), 115.8 (C-5′), 117.1 (C-5), 122.9 (C-6′), 128.7 (C-1), 133.9 (C-6), 138.3 (C-1′), 145.0 (C-4), 145.3 (C-4′), 146.9 (C-3), 148.8 (C-3′). ESI-HRMS m/z 361.1635 [M + H] + (calculated for C 22 H 25 O 6, 361.1648, Δppm = −2.75).

Ethyl 3,4-dihydroxybenzoate (11): 1 H NMR (DMSO- d 6, 500 MHz) δ 1.27 (3H, t, J = 7.1 Hz, CH 3 -9), 4.22 (2H, q, J = 7.1 Hz, H-8), 6.81 (1H, d, J = 8.2 Hz, H-5), 7.31 (1H, dd, J = 8.0, 2.1 Hz, H-6), 7.37 (1H, d, J = 2.1 Hz, H-2); 13 C NMR (DMSO- d 6, 126 MHz) δ 14.0 (CH 3 -9), 59.7 (C-8), 114.9 (C-5), 115.8 (C-2), 120.4 (C-1), 121.5 (C-6), 144.8 (C-3), 149.2 (C-4), 165.4 (C-7). ESI-HRMS m/z 183.0648 [M + H] + (calculated for C 9 H 11 O 4, 183.0655, Δppm = −1.99).

(±)-Epiphyllocoumarin (12): [α] 23 D −0.1 (c 1.10, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ 2.87 (1H, dd, J = 17.2, 2.7 Hz, H-4b), 2.96 (1H, dd, J = 17.2, 4.5 Hz, H-4a), 4.30 (1H, t, J = 4.5, 2.7 Hz, H-3), 5.02 (1H, s, H-2), 6.09 (1H, d, J = 9.5 Hz, H-12), 6.36 (1H, s, H-6), 6.80 (1H, d, J = 8.2 Hz, H-5′), 6.85 (1H, dd, J = 8.2, 2.0 Hz, H-6′), 7.04 (1H, d, J = 2.0 Hz, H-2′), 8.14 (1H, d, J = 9.5 Hz, H-11); 13 C NMR (CD 3 OD, 126 MHz) δ 29.4 (C-4), 66.5 (C-3), 80.7 (C-2), 95.4 (C-6), 103.7 (C-8), 105.4 (C-10), 110.0 (C-12), 115.2 (C-2′), 116.0 (C-5′), 119.3 (C-6′), 131.4 (C-1′), 141.1 (C-11), 146.1 (C-4′), 146.2 (C-3′), 153.4 (C-9), 156.0 (C-7), 162.1 (C-5), 164.3 (C-13). ESI-HRMS m/z 343.0815 [M + H] + (calculated for C 18 H 15 O 7, 343.0819, Δppm = 1.9).

Ethyl caffeate (13): 1 H NMR (CD 3 OD, 500 MHz) δ 1.31 (3H, t, J = 7.2 Hz, H-11), 4.21 (2H, q, J = 7.2 Hz, H-10), 6.25 (1H, d, J = 15.9 Hz, H-8), 6.77 (1H, d, J = 8.2 Hz, H-5), 6.94 (1H, dd, J = 8.2, 2.1 Hz, H-6), 7.03 (1H, d, J = 2.1 Hz, H-2), 7.53 (1H, d, J = 15.9 Hz, H-7); 13 C NMR (CD 3 OD, 126 MHz) δ 14.4 (C-11), 61.2 (C-10), 114.9 (C-2), 115.1 (C-8), 116.3 (C-5), 122.7 (C-6), 127.5 (C-1), 146.5 (C-7), 146.6 (C-3), 149.3 (C-4), 169.0 (C-9). ESI-HRMS m/z 209.0811 [M + H] + (calculated for C 11 H 13 O 4, 209.0814, Δppm = 0.5).

Ethyl-4-hydroxybenzoate (14): 1 H NMR (CD 3 OD, 500 MHz) δ 1.36 (3H, t, J = 7.2 Hz, H-9), 4.31 (2H, q, J = 7.2 Hz, H-8), 6.82 (2H, d, J = 9.0 Hz, H-3, H-5), 7.87 (1H, d, J = 9.0 Hz, H-2, H-6); 13 C NMR (CD 3 OD, 126 MHz) δ 14.4 (C-9), 61.4 (C-8), 115.9 (C-3, C-5), 122.2 (C-1), 132.4 (C-2, C-6), 163.2 (C-4), 167.9 (C-7). ESI-HRMS m/z 167.0705 [M + H] + (calculated for C 9 H 11 O 3, 167.0695, Δppm = −3.2).

Ethyl syringate (15): 1 H NMR (CD 3 OD, 500 MHz) δ 1.38 (3H, t, J = 7.7 Hz, H-9), 3.89 (6H, s, CH 3 O-3, CH 3 O-5), 4.34 (2H, q, J = 7.2 Hz, H-8), 7.32 (2H, s, H-2, H-6); 13 C NMR (CD 3 OD, 126 MHz) δ 14.4 (C-9), 61.6 (C-8), 107.8 (C-2, C-6), 121.5 (C-1), 141.9 (C-4), 148.9 (C-3, C-5), 168.0 (C-7). ESI-HRMS m/z 227.0917 [M + H] + (calculated for C 11 H 15 O 5, 227.0914, Δppm = 0.3).

Cleomiscosin A (16): 1 H NMR (CD 3 OD, 500 MHz) δ 3.54 (1H, dd, J = 12.7, 3.8 Hz, H-9′b), 3.85 (1H, m, H-9′a), 3.87 (3H, s, CH 3 O-6), 3.88 (3H, s, CH 3 O-3′), 4.20 (1H, dt, J = 8.0, 3.8, 2.4 Hz, H-8′), 5.04 (1H, d, J = 8.0 Hz, H-7′), 6.31 (1H, d, J = 9.5 Hz, H-3), 6.81 (1H, s, H-5), 6.86 (1H, d, J = 8.1 Hz, H-5′), 6.94 (1H, dd, J = 8.1, 2.0 Hz, H-6′), 7.05 (1H, d, J = 2.0 Hz, H-2′), 7.89 (1H, d, J = 9.5 Hz, H-4); 13 C NMR (CD 3 OD, 126 MHz) δ 56.5 (CH 3 O-3′), 61.5 (CH 3 O-6, C-9′), 77.9 (C-7′), 79.7 (C-8′), 101.7 (C-5), 112.0 (C-2′), 112.9 (C-10), 113.8 (C-3), 116.1 (C-5′), 121.8 (C-6′), 128.3 (C-1′), 133.0 (C-8), 139.4 (C-7), 139.8 (C-9), 146.1 (C-4), 147.4 (C-6), 148.5 (C-4′), 149.2 (C-3′), 163.1 (C-2). ESI-HRMS m/z 387.1069 [M + H] + (calculated for C 20 H 19 O 8, 387.1080, Δppm = 2.8).

1-Methoxy-2,7-dihydroxyanthraquinone (17): 1 H NMR (CD 3 OD, 500 MHz) δ 3.90 (3H, s, CH 3 O-1), 7.15 (1H, dd, J = 8.6, 2.6 Hz, H-6), 7.23 (1H, d, J = 8.5 Hz, H-3), 7.50 (1H, d, J = 2.6 Hz, H-8), 7.99 (1H, d, J = 8.5 Hz, H-4), 8.10 (1H, d, J = 8.6 Hz, H-5); 13 C NMR (CD 3 OD, 126 MHz) δ 61.4 (CH 3 O-1), 112.6 (C-8), 121.7 (C-3), 121.9 (C-6), 126.2 (C-4), 130.6 (C-5), 134.4 (C-12), 147.1 (C-1), 157.7 (C-2), 162.3 (C-7), 181.5 (C-10). ESI-HRMS m/z 271.0598 [M + H] + (calculated for C 15 H 11 O 5, 271.00604, Δppm = −0.81).

5β-hydroxy-RA-XII (18): [α] 23 D – 81.3 (c 1.26, MeOH). 1 H NMR (CD 3 OD, 500 MHz) δ Ala-1: 1.23 (3H, d, J = 7.0 Hz, H 3 -β), 4.44 (1H, q, J = 7.0 Hz, H-α), Ala-2: 1.30 (3H, d, J = 7.0 Hz, H 3 -β), 4.69 (1H, q, J = 7.0 Hz, H-α), Tyr-3: 2.92 (3H, s, NCH 3), 3.22 (1H, dd, J = 13.7, 4.9 Hz, H-β1), 3.30 (1H, m, H-β2), 3.78 (3H, s, CH 3 O), 3.89 (1H, m, H-α), 6.88 (2H, d, J = 8.5 Hz, H-ξ1, H-ξ2), 7.09 (2H, d, J = 8.5 Hz, H-δ1, H-δ2), Ala-4: 1.10 (3H, d, J = 6.7 Hz, H 3 -3β), 4.87 (1H, m, H-α), Tyr-5: 3.27 (3H, s, NCH 3), 5.05 (1H, d, J = 2.4 Hz, H-β), 5.34 (1H, d, J = 2.4 Hz, H-α), 6.95 (1H, dd, J = 8.7, 2.4 Hz, H- ξ1), 7.32 (1H, dd, J = 8.4, 2.4 Hz, H- ξ2), 7.44 (1H, dd, J = 8.4, 2.3 Hz, H-δ1), 7.52 (1H, dd, J = 8.7, 2.3 Hz, H-δ2), Tyr-6: 2.66 (3H, s, NCH 3), 3.03 (1H, m, H-β1), 3.07 (1H, m, H-β2), 4.55 (1H, dd, J = 11.0, 4.5 Hz, H-α), 4.72 (1H, d, J = 2.2 Hz, H-δ2), 6.67 (1H, dd, J = 8.5, 2.2 Hz, H-δ1), 7.10 (1H, d, J = 8.5 Hz, H- ξ1), Glc: 3.44 (2H, m, H-4′, H-5′), 3.50 (1H, t, J = 9.1 Hz, H-3′), 3.58 (1H, t, J = 8.4 Hz, H-2′), 3.72 (1H, dd, J = 12.1, 3.8 Hz, H-6′b), 3.89 (1H, m, H-6′a), 5.00 (1H, d, J = 7.7 Hz, H-1′); 13 C NMR (CD 3 OD, 126 MHz) δ Ala-1: 20.8 (CH 3 -β), 48.7 (C-α), 173.1 (C=O), Ala-2: 16.1 (CH 3 -β), 45.6 (C-α), 174.4 (C=O), Tyr-3: 33.5 (C-β), 40.0 (NCH 3), 55.5 (CH 3 O), 68.5 (C-α), 114.9 (C-ξ), 131.3 (C-δ), 131.4 (C-γ), 159.9 (C-ζ), 171.0 (C = 0), Ala-4: 18.6 (CH 3 -β), 47.7 (C-α), 173.1 (C=O), Tyr-5: 33.7 (NCH 3), 55.5 (C-α), 78.8 (C-β), 125.1 (C-ξ1), 126.6 (C-ξ2), 127.5 (C-δ2), 129.9 (C-δ1), 140.5 (C-γ), 160.4 (C-ζ), 171.5 (C=O), Tyr-6: 29.6 (NCH 3), 36.6 (C-β), 58.6 (C-α), 116.0 (C-δ2), 118.6 (C-ξ1), 122.5 (C-δ1), 131.8 (C-γ), 145.4 (C-ζ), 154.4 (C-ξ2), 171.7 (C=O), Glc: 62.4 (C-6′), 71.2 (C-4′), 74.7 (C-2′), 77.7 (C-3′), 78.0 (C-5′), 102.7 (C-1′). ESI-HRMS m/z 935.4036 [M + H] + (calculated for C 46 H 59 N 6 O 15, 935.4025, Δppm = 0.6).

5β-Hydroxy-RA-I (19): [α] 23 D −26.0 (c 0.70, MeOH). 1 H NMR (DMSO- d 6, 500 MHz) δ Ala-1: 1.06 (3H, d, J = 6.8 Hz, H 3 -β), 4.36 (1H, m, H-α), 7.94 (1H, d, J = 8.7 Hz, NH), Ser-2: 3.49 (1H, dd, J = 10.5, 6.4 Hz, H-β1), 3.67 (1H, m, H-β2), 4.61 (1H, m, H-α), 8.48 (1H, d, J = 7.6 Hz, NH), Tyr-3: 2.91 (3H, s, NCH 3), 3.12 (2H, d, J = 7.9 Hz, H-β), 3.72 (3H, s, CH 3 O), 3.79 (1H, m, H-α), 6.81 (2H, d, J = 8.5 Hz, H-ξ1, H-ξ2), 7.15 (2H, d, J = 8.5 Hz, H-δ1, H-δ2), Ala-4: 0.95 (3H, d, J = 6.8 Hz, H 3 -β), 4.69 (1H, p, J = 6.8 Hz, H-α), 6.61 (1H, d, J = 8.3 Hz, NH), Tyr-5: 3.13 (3H, s, NCH 3), 4.94 (1H, brs, H-β), 5.19 (1H, d, J = 2.5 Hz, H-α), 6.18 (1H, d, J = 9.4 Hz, OH-β), 6.88 (1H, dd, J = 8.7, 2.3 Hz, H-ξ2), 7.29 (1H, dd, J = 8.4, 2.3 Hz, H-ξ1), 7.37 (1H, dd, J = 8.4, 2.3 Hz, H-δ1), 7.42 (1H, dd, J = 8.7, 2.3 Hz, H-δ2), Tyr-6: 2.51 (3H, s, NCH 3), 2.78 (1H, dd, J = 16.9, 12.3 Hz, H-β1), 3.01 (1H, d, J = 16.9 Hz, H-β2), 4.32 (1H, dd, J = 11.7, 3.8 Hz, H-α), 4.61 (1H, s, H-δ2), 6.49 (1H, d, J = 8.3 Hz, H-δ1), 6.69 (1H, d, J = 8.3 Hz, H-ξ1); 13 C NMR (DMSO- d 6, 126 MHz) δ Ala-1: 20.8 (CH 3 -β), 46.7 (C-α), 171.5 (C=O), Ser-2: 50.4 (C-α), 60.6 (C-β), 171.0 (C=O), Tyr-3: 32.2 (C-β), 39.2 (NCH 3), 54.8 (CH 3 O), 67.2 (C-α), 113.5 (C-ξ1, C-ξ2), 130.3 (C-δ1, C-δ2), 130.9 (C-γ), 157.7 (C-ζ), 168.0 (C=O), Ala-4: 18.0 (CH 3 -β), 45.7 (C-α), 170.8 (C=O), Tyr-5: 32.8 (NCH 3), 53.4 (C-α), 77.2 (C-β), 123.8 (C-ξ2), 125.3 (C-ξ1), 126.2 (C-δ1), 128.7 (C-δ2), 139.2 (C-γ), 159.1 (C-ζ), 169.9 (C=O), Tyr-6: 28.9 (NCH 3), 34.8 (C-β), 57.2 (C-α), 114.9 (C-δ2), 116.5 (C-ξ1), 121.5 (C-δ1), 127.4 (C-γ), 143.6 (C-ζ), 151.1 (C-ξ2), 169.2 (C=O). ESI-HRMS m/z 789.3457 [M + H] + (calculated for C 40 H 49 N 6 O 11, 789.3463, Δppm = −0.4).

Bouvardin (20) [α] 23 D −112.1 (c 1.12, MeOH). 1 H NMR (CDCl 3, 500 MHz) δ Ala-1: 1.30 (3H, d, J = 6.8 Hz, H 3 -β), 4.35 (1H, m, H-α), 6.48 (1H, d, J = 6.8 Hz, NH), Ala-2: 1.36 (3H, d, J = 7.0 Hz, H 3 -β), 4.77 (1H, p, J = 7.9, 7.0 Hz, H-α), 6.41 (1H, d, J = 7.9 Hz, NH), Tyr-3: 2.87 (3H, s, NCH 3), 3.37 (2H, m, H-β), 3.61 (1H, dd, J = 10.5, 4.8 Hz, H-α), 3.79 (3H, s, CH 3 O), 6.84 (2H, d, J = 8.1 Hz, H-ξ1, H-ξ2), 7.05 (2H, d, J = 8.1 Hz, H-δ1, H-δ2), Ala-4: 1.10 (3H, d, J = 6.6 Hz, H 3 -β), 4.88 (1H, p, J = 7.9, 6.6 Hz, H-α), 6.67 (1H, d, J = 7.9 Hz, NH), Tyr-5: 3.33 (3H, s, NCH 3), 5.06 (1H, brs, H-β), 5.35 (1H, d, J = 2.2 Hz, H-α), 6.97 (1H, dd, J = 8.6, 2.4 Hz, H-ξ2), 7.25 (1H, dd, J = 8.3, 2.4 Hz, H-ξ1), 7.38 (1H, dd, J = 8.3, 2.4 Hz, H-δ1), 7.50 (1H, dd, J = 8.6, 2.4 Hz, H-δ2), Tyr-6: 2.73 (3H, s, NCH 3), 2.90 (1H, m, H-β1), 3.10 (1H, m, H-β2), 4.35 (1H, m, H-α), 4.36 (1H, d, J = 2.1 Hz, H-δ2), 6.52 (1H, dd, J = 8.3, 2.1 Hz, H-δ1), 6.83 (1H, d, J = 8.3 Hz, H-ξ1); 13 C NMR (CDCl 3, 126 MHz) δ Ala-1: 20.4 (CH 3 -β), 47.8 (C-α), 172.3 (C=O), Ala-2: 16.4 (CH 3 -β), 45.0 (C-α), 172.5 (C=O), Tyr-3: 32.6 (C-β), 39.6 (NCH 3), 55.3 (CH 3 O), 68.2 (C-α), 114.1 (C-ξ1, C-ξ2), 130.3 (C-δ1, C-δ2), 130.7 (C-γ), 158.4 (C-ζ), 168.2 (C=O), Ala-4: 18.5 (CH 3 -β), 46.6 (C-α), 172.3 (C=O), Tyr-5: 33.0 (NCH 3), 53.9 (C-α), 78.5 (C-β), 124.2 (C-ξ2), 125.6 (C-ξ1), 126.8 (C-δ1), 128.4 (C-δ2), 139.9 (C-γ), 158.6 (C-ζ), 170.8 (C=O), Tyr-6: 29.3 (NCH 3), 35.9 (C-β), 57.6 (C-α), 112.7 (C-δ2), 116.0 (C-ξ1), 121.6 (C-δ1), 127.2 (C-γ), 143.2 (C-ζ), 150.9 (C-ξ2), 169.9 (C=O). ESI-HRMS m/z 773.3507 [M + H] + (calculated for C 40 H 49 N 6 O 10, 773.3502, Δppm = −0.3).

5β-Hydroxy-RA-III (21): 1 H NMR (CDCl 3, 500 MHz) δ Ala-1: 1.33 (3H, d, J = 6.8 Hz, H 3 -β), 4.45 (1H, p, J = 6.8 Hz, H-α), 6.45 (1H, d, J = 6.8 Hz, NH), Ser-2: 3.81 (1H, m, H-β1), 3.93 (1H, m, H-β2), 4.62 (1H, m, H-α), 6.90 (1H, brs, NH), Tyr-3: 2.97 (3H, s, NCH 3), 3.36 (2H, m, H-β), 3.66 (1H, dd, J = 10.3, 5.3 Hz, H-α), 3.80 (3H, s, CH 3 O), 6.85 (2H, d, J = 8.3 Hz, H-ξ1, H-ξ2), 7.08 (2H, d, J = 8.3 Hz, H-δ1, H-δ2), Ala-4: 1.10 (3H, d, J = 6.6 Hz, H 3 -β), 4.88 (1H, p, J = 7.8, 6.6 Hz, H-α), 6.71 (1H, d, J = 7.8 Hz, NH), Tyr-5: 3.34 (3H, s, NCH 3), 5.06 (1H, dd, J = 6.5, 2.1 Hz, H-β), 5.38 (1H, d, J = 2.1 Hz, H-α), 6.48 (1H, d, J = 6.5 Hz, OH-β), 7.01 (1H, dd, J = 8.6, 2.3 Hz, H-ξ2), 7.25 (1H, dd, J = 8.4, 2.3 Hz, H-ξ1), 7.36 (1H, dd, J = 8.4, 2.3 Hz, H-δ1), 7.50 (1H, dd, J = 8.5, 2.3 Hz, H-δ2), Tyr-6: 2.75 (3H, s, NCH 3), 2.92 (1H, dd, J = 18.5, 3.0 Hz, H-β1), 3.15 (1H, dd, J = 18.5, 11.9 Hz, H-β2), 3.95 (3H, s, CH 3 O), 4.34 (1H, d, J = 2.0 Hz, H-δ2), 4.39 (1H, dd, J = 11.9, 3.0 Hz, H-α), 6.58 (1H, dd, J = 8.3, 2.0 Hz, H-δ1), 6.82 (1H, d, J = 8.3 Hz, H-ξ1); 13 C NMR (CDCl 3, 126 MHz) δ Ala-1: 20.6 (CH 3 -β), 47.8 (C-α), 172.5 (C=O), Ser-2: 48.4 (C-α), 61.6 (C-β), 171.9 (C=O), Tyr-3: 32.7 (C-β), 40.3 (NCH 3), 55.3 (CH 3 O), 68.4 (C-α), 114.2 (C-ξ1, C-ξ2), 130.2 (C-δ1, C-δ2), 158.4 (C-ζ), 167.8 (C=O), 130.3 (C-γ), Ala-4: 18.6 (CH 3 -β), 46.4 (C-α), 172.0 (C=O), Tyr-5: 33.0 (NCH 3), 53.9 (C-α), 78.5 (C-β), 124.3 (C-ξ2), 125.7 (C-ξ1), 126.7 (C-δ1), 128.3 (C-δ2), 139.3 (C-γ), 158.9 (C-ζ),170.7 (C=O), Tyr-6: 29.2 (NCH 3), 35.7 (C-β), 56.2 (CH 3 O), 57.7 (C-α), 112.5 (C-ξ1), 113.2 (C-δ2), 121.0 (C-δ1), 127.6 (C-γ), 146.7 (C-ζ), 152.9 (C-ξ2), 170.0 (C=O). ESI-HRMS m/z 803.3613 [M + H] + (calculated for C 41 H 51 N 6 O 11, 803.3607, Δppm = 0.3).

6- O -methylbouvardin (22): [α] 23 D −150.0 (c 0.36, CHCl 3). 1 H NMR (CDCl 3, 500 MHz) δ Ala-1: 1.30 (3H, d, J = 6.9 Hz, H 3 -β), 4.37 (1H, m, H-α), 6.47 (1H, d, J = 6.6 Hz, NH-Ala-1), Ala-2: 1.36 (3H, d, J = 7.0 Hz, H 3 -β), 4.78 (1H, p, J = 7.0 Hz, H-α), 6.37 (1H, d, J = 8.0 Hz, NH), Tyr-3: 2.87 (3H, s, NCH 3), 3.35 (2H, m, H-β), 3.61 (1H, dd, J = 10.6, 5.2 Hz, H-α), 3.80 (3H, s, CH 3 O), 6.84 (2H, d, J = 8.6 Hz, H-ξ1, H-ξ2), 7.05 (2H, d, J = 8.6 Hz, H-δ1, H-δ2), Ala-4: 1.10 (3H, d, J = 6.7 Hz, H 3 -β), 4.87 (1H, p, J = 6.7 Hz, H-α), 6.68 (1H, d, J = 8.0 Hz, NH), Tyr-5: 3.33 (3H, s, NCH 3), 5.06 (1H, d, J = 10.3 Hz, H-β), 5.37 (1H, d, J = 2.2 Hz, H-α), 6.42 (1H, d, J = 10.3 Hz, OH-β), 7.01 (1H, dd, J = 8.6, 2.4 Hz, H-ξ2), 7.24 (1H, dd, J = 8.4, 2.4 Hz, H-ξ1), 7.37 (1H, dd, J = 8.4, 2.4 Hz, H-δ1), 7.49 (1H, dd, J = 8.6, 2.4 Hz, H-δ2), Tyr-6: 2.74 (3H, s, NCH 3), 2.91 (1H, d, J = 18.6 Hz, H-β1), 3.14 (1H, dd, J = 18.6, 11.8 Hz, H-β2), 3.95 (3H, s, CH 3 O), 4.34 (1H, d, J = 2.2 Hz, H-δ2), 4.37 (1H, m, H-α), 6.58 (1H, dd, J = 8.4, 2.2 Hz, H-δ1), 6.81 (1H, d, J = 8.4 Hz, H-ξ1); 13 C NMR (CDCl 3, 126 MHz) δ Ala-1: 20.6 (CH 3 -β), 47.9 (C-α), 172.4 (C=O), Ala-2: 16.5 (CH 3 -β), 44.9 (C-α), 172.5 (C=O), Tyr-3: 32.7 (C-β), 39.9 (NCH 3), 55.4 (CH 3 O), 68.4 (C-α), 114.2 (C-ξ1, C-ξ2), 130.4 (C-δ1, C-δ2), 130.6 (C-γ), 158.5 (C-ζ), 168.3 (C=O), Ala-4: 18.7 (CH 3 -β), 46.5 (C-α), 172.3 (C=O), Tyr-5: 33.2 (NCH 3), 54.0 (C-α), 78.6 (C-β), 124.4 (C-ξ2), 125.8 (C-ξ1), 127.0 (C-δ1), 128.4 (C-δ2), 139.5 (C-γ), 158.9 (C-ζ), 170.9 (C=O), Tyr-6: 29.3 (NCH 3), 35.8 (C-β), 56.3 (CH 3 O), 57.8 (C-α), 112.6 (C-ξ1), 113.3 (C-δ2), 121.1 (C-δ1), 127.8 (C-γ), 146.9 (C-ζ), 153.2 (C-ξ2), 169.9 (C=O). ESI-HRMS m/z 787.3664 [M + H] + (calculated for C 41 H 51 N 6 O 10, 787.3665, Δppm = 0.1).

(14 S,15 S,18 R,19 R)-12-methoxy-14,14-dimethyl-3b,4,5,6-tetrahydro-7H-6a,4-(epoxymethano)cyclopenta [c]indolo[3,2,1-ij][1,5]naphthyridin-7-ol (23): [α] 23 D +12.0 (c 0.50, MeOH). 1 H NMR (DMSO- d 6, 600 MHz) δ 0.92 (1H, ddd, J = −13.2, 12.0, 2.7 Hz, H-17b), 1.35 (3H, s, H 3 -22), 1.35 (1H, dddd, J = −12.9, 12.0, 6.3, 2.9 Hz, H-16b), 1.37 (3H, s, H 3 -21), 1.49 (1H, dddd, J = −13.2, 8.8, 6.3, 1.6 Hz, H-17a), 1.82 (1H, dddd, J = −12.9, 8.8, 2.7, 1.9 Hz, H-16a), 2.80 (1H, ddd, J = 2.9, 1.5, 0.9 Hz, H-15), 3.95 (1H, ddd, J = 1.9, 1.6, 1.5 Hz, H-14), 4.03 (3H, s, CH 3 O-9), 6.00 (1H, d, J = 7.7 Hz, H-19), 6.50 (d, J = 7.7 Hz, OH-19), 6.85 (1H, d, J = 8.0 Hz, H-10), 7.32 (1H, d, J = 8.2 Hz, H-12), 7.55 (1H, dd, J = 8.2, 8.0 Hz, H-11), 7.81 (1H, dd, J = 5.3, 0.9 Hz, H-6), 8.27 (1H, d, J = 5.3 Hz, H-5); 13 C NMR (DMSO- d 6, 126 MHz) δ 22.0 (C-16), 25.5 (CH 3 -22), 28.7 (CH 3 -21), 30.8 (C-17), 46.4 (C-15), 46.9 (C-14), 55.5 (CH 3 O-9), 74.6 (C-19), 83.4 (C-20), 88.1 (C-18), 101.2 (C-10), 103.3 (C-12), 111.1 (C-8), 114.5 (C-6), 123.7 (C-7), 129.4 (C-11), 131.7 (C-2), 139.2 (C-5), 140.8 (C-13), 141.8 (C-3), 156.5 (C-9). ESI-HRMS m/z 351.1692 [M + H] + (calculated for C 21 H 23 N 2 O 3 +, 351.1703, Δppm = −3.13).

(14 R,15 R,18 S)-1-(hydroxymethyl)-3-(2-hydroxypropan-2-yl)-2-(5methoxy-9H-b-carbolin-1-yl)-cyclopentanol (24): 1 H NMR (DMSO- d 6, 500 MHz) δ 1.28 (3H, s, H 3 -22), 1.44 (1H, dddd, J = −12.2, 8.8, 6.1, 0.8 Hz, H-17″), 1.45 (3H, s, H 3 -21), 1.47 (1H, dddd, J = −12.7, 12.0, 6.1, 3.2 Hz, H-16″), 1.80 (1H, ddddd, J = −12.7, 8.8, 2.8, 1.2, 1.1 Hz, H-16′), 2.17 (1H, dddd, J = −12.2, 12.0, 2.8, 0.8 Hz, H-17′), 2.70 (1H, dddd, 3.2, 1.2, 0.8, 0.8 Hz, H-15), 3.70 (1H, d, J = −11.7 Hz, H-19″), 3.77 (1H, d, J = −11.7 Hz, H-19′), 3.93 (1H, ddd, J = 1.1, 0.8, 0.8 Hz, H-14), 4.03 (3H, s, CH 3 O-9), 6.77 (1H, d, J = 7.9 Hz, H-10), 7.25 (1H, d, J = 8.2 Hz, H-12), 7.49 (1H, dd, J = 8.2, 7.9 Hz, H-11), 7.94 (1H, d, J = 5.2 Hz, H-6), 8.21 (1H, d, J = 5.2 Hz, H-5), 11.28 (1H, s, NH-1); 13 C NMR (DMSO- d 6, 126 MHz) δ 22.3 (C-16), 25.9 (CH 3 -22), 29.3 (C-17), 29.6 (CH 3 -21), 49.7 (C-15), 51.6 (C-14), 55.5 (CH 3 O-9), 61.5 (C-19), 78.5 (C-20), 90.5 (C-18), 100.3 (C-10), 104.8 (C-12), 110.4 (C-8), 114.9 (C-6), 126.8 (C-7), 129.2 (C-11), 133.5 (C-2), 136.9 (C-5), 141.8 (C-13), 142.3 (C-3), 156.3 (C-9). ESI-HRMS m/z 353.1856 [M + H] + (calculated for C 21 H 25 N 2 O 3 +, 353.1860, Δppm = −1.13), m/z 351.1697 [M - H] - (calculated for C 21 H 23 N 2 O 3 −, 351.1714, Δppm = −4.84).

1-methyl-3-(2-hydroxypropan-2-yl)-2 (5-methoxy-9H-b-carbolin-1-yl)-cyclopentanol (25): [α] 23 D + 36.8 (c 0.38, MeOH). 1 H NMR (DMSO- d 6, 500 MHz) δ 1.26 (3H, s, H 3 -22), 1.40 (3H, s, H 3 -19), 1.45 (3H, s, H 3 -21), 1.47 (1H, ddd, J = −12.0, 9.0, 5.7 Hz, H-17″), 1.60 (1H, ddd, J = −12.9, 11.5, 5.7 Hz, H-16″), 1.77 (1H, ddd, J = −12.9, 9.0, 2.8 Hz, H-16′), 2.28 (1H, ddd, J = −12.0, 11.5, 2.8 Hz, H-17′), 2.52 (1H, s, H-15), 3.75 (1H, s, H-14), 4.03 (3H, s, CH 3 O-9), 6.76 (1H, d, J = 7.9 Hz, H-10), 7.25 (1H, d, J = 8.1 Hz, H-12), 7.47 (1H, dd, J = 8.1, 7.9 Hz, H-11), 7.91 (1H, d, J = 5.2 Hz, H-6), 8.24 (1H, d, J = 5.2 Hz, H-5), 11.22 (1H, s, NH-1); 13 C NMR (DMSO- d 6, 126 MHz) δ 18.6 (CH 3 -19), 22.8 (C-16), 25.6 (CH 3 -22), 33.3 (C-17, CH 3 -21), 50.1 (C-15), 53.3 (C-14), 55.2 (CH 3 O-9), 77.9 (C-20), 86.5 (C-18), 99.9 (C-10), 104.4 (C-12), 110.0 (C-8), 114.2 (C-6), 125.9 (C-7), 128.6 (C-11), 133.4 (C-2), 137.0 (C-5), 141.4 (C-13), 142.1 (C-3), 155.9 (C-9). ESI-HRMS m/z 337.1906 [M + H] + (calculated for C 21 H 25 N 2 O 2 +, 337.1911, Δppm = −1.48), m/z 335.1763 [M-H] - (calculated for C 21 H 23 N 2 O 2 −, 335.1765, Δppm = −0.60).

1-Methoxy-2-hydroxyanthraquinone (26): 1 H NMR (DMSO- d 6, 500 MHz) δ 3.83 (3H, s, CH 3 O-1), 7.25 (1H, d, J = 8.6 Hz, H-3), 7.85 (2H, m, H-6, H-7), 7.91 (1H, d, J = 8.6 Hz, H-4), 8.12 (2H, m, H-5, H-8); 13 C NMR (DMSO- d 6, 126 MHz) δ 60.2 (CH 3 O-1), 121.6 (C-3), 124.6 (C-11), 125.1 (C-4), 126.0 (C-5), 126.5 (C-8), 126.6 (C-10), 133.7 (C-6), 133.8 (C-7), 147.9 (C-1), 159.4 (C-2), 181.0 (C-12), 182.3 (C-9). ESI-HRMS m/z 255.0664 [M + H] + (calculated for C 15 H 11 O 4, 255.0657, Δppm = 2.7).

Cell lines and culture conditions

The triple-negative breast cancer (TNBC) cell line BT-20 was obtained from ATCC (cat. HTB-19) and maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Fisher Scientific) supplemented with 10% fetal calf serum (FCS) and 1% penicillin-streptomycin. The cells were cultured at 37 °C in a humidified atmosphere containing 5% CO 2, unless otherwise specified.

MTT assay

To assess cell viability, BT-20 TNBC cells were seeded at a density of 2,000 cells per well into a transparent, cell culture-grade 384-well plate. After 24 h, the maintenance medium was replaced with 50 μL of fresh medium containing six serial dilutions of the test compounds. Following a 3-day incubation, cell viability was determined by replacing the medium with 50 μL of a 0.5 mg/mL Thiazolyl Blue Tetrazolium Bromide (MTT) solution prepared in phosphate-buffered saline (PBS). The cells were then incubated at 37 °C for 3 h, after which the solution was removed, and 25 μL of dimethyl sulfoxide (DMSO) was added to each well. Absorbance was measured at 570 nm using a Tecan Infinite M200 PRO plate reader. The experiments were performed in N = 3 independent repeats, with n = 2 technical repeats per concentration.

Luciferase-based assay for wnt-dependent transcriptional activity

Purified Wnt3a was obtained from stably transfected mouse L-cells (Willert et al., 2003), with modifications from our laboratory (Xu et al., 2020). BT-20 TNBC cells, stably transfected with the M50 Super 8×TopFlash plasmid, were seeded in 20 μL of maintenance medium per well into white, tissue-culture-treated 384-well plates at a density of 6,000 cells per well. For additional transfection, the cells were co-transfected with the pCMV-RL plasmid, which constitutively expresses Renilla luciferase, using the XtremeGENE 9 transfection reagent following the manufacturer’s instructions. After 24 h, the medium was removed, and the cells were treated with 2-fold serial dilutions of the test compounds in 10 μL of maintenance medium per well. Following 1 h of preincubation, Wnt3a was added at a final concentration of 500 ng/mL in an additional 10 μL per well.

After 24 h of incubation, the medium was removed, and luminescence was measured using a Tecan Infinite M200 PRO plate reader equipped with a two-channel dispensing unit. Firefly and Renilla luciferase activity were sequentially detected by injecting 15 μL of their respective buffer solutions (Boudou et al., 2022). Dose-response data from both the MTT assay and the luciferase assay were analyzed using GraphPad Prism 9, and IC 50 values were determined. The experiments were performed in N = 3 independent repeats, with n = 2 technical repeats per concentration.

Isolation and structure elucidation

In previous investigations, the chloroform fraction obtained from the roots of G. thalictroides yielded structurally interesting bioactive compounds, notably belonging to the β-carboline alkaloid and cyclopeptide classes (Figueiredo et al., 2011; Figueiredo et al., 2014; Figueiredo et al., 2016). Building on these findings, this fraction was reinvestigated using modern analytical approaches.

To enable a comprehensive and reproducible analysis of its chemical composition, the chloroform fraction was regenerated under controlled conditions. It was re-prepared from freshly extracted plant material following previously established extraction and fractionation procedures (Figueiredo et al., 2011; Figueiredo et al., 2016). This approach ensured methodological consistency and provided sufficient material for subsequent analyses.

The chloroform fraction was fractionated by medium-pressure liquid chromatography (MPLC-UV) according to a method developed in our laboratory (Challal et al., 2015). Some of the fractions obtained were purified by semi-preparative HPLC-UV, resulting in the isolation of 26 compounds (1–26). All the isolated compounds were submitted to nuclear magnetic resonance (NMR) and high-resolution mass spectrometry (HRMS) analysis for structural identification. Based on this analysis the compounds were identified as homovanillyl alcohol 4- O -β-D-glucopyranoside (1) (Liu et al., 2006), feretoside (2) (Otsuka et al., 1991; Guvenalp et al., 2006), asperulosidic acid (3) (Kim et al., 2005), homovanillyl alcohol (4) (Kringstad and Morck, 1983), syringic acid (5) (Liao et al., 2014), protocatechuic acid methyl ester (6) (Miyazawa et al., 2003), vanillic acid (7) (Frezza et al., 2025), scopoletin (8) (Jerezano et al., 2011), ciwujiatone (9) (Chen et al., 2012), cyclolariciresinol (10) (Abe and Yamauchi, 1989), ethyl 3,4-dihydroxybenzoate (11) (Sawai et al., 2005), epiphyllocoumarin (12) (Tang et al., 2013), ethyl caffeate (13) (Shi et al., 2008), ethyl-4-hydroxybenzoate (14) (Chae, 2008), ethyl syringate (15) (Bao et al., 2009), cleomiscosin A (16) (Ray et al., 1985), 1-methoxy-2,7-dihydroxyanthraquinone (17) is, to the best of our knowledge, described here for the first time as an anthraquinone isolated from a natural source, bouvardin (20) (Boger et al., 1994), 5β-hydroxy-RA-III (21) (Figueiredo et al., 2016), 6- O -methylbouvardin (22) (Boger et al., 1994), and 1-methoxy-2-hydroxyanthraquinone (26) (Wu et al., 2009).

Among the isolated compounds, three are novel natural products (18, 19, and 23), and their structural elucidation is detailed below. Additionally, the structures of two alkaloids (24 and 25), previously reported (Figueiredo et al., 2011; Figueiredo et al., 2014), were revised (Figure 1).

As reported in the previous study, G. thalictroides was shown to produce two Rubiaceae-type cyclopeptides, namely, 5β-hydroxy-RA-III and 6- O -methylbouvardin (Figueiredo et al., 2016). In the present work, these compounds were re-isolated as compounds 21 and 22, respectively, along with bouvardin (20) and two newly identified members of this family (18 and 19). All these cyclopeptides share a hydroxy group located at the β-carbon of the tyrosine-5 residue.

When compared to bouvardin (20), the NMR spectra of compound 18 showed additional signals corresponding to a glucose moiety at δ H 5.00 (1H, d, J = 7.7 Hz, H-1′), 3.58 (1H, t, J = 8.4 Hz, H-2′), 3.50 (1H, t, J = 9.1 Hz, H-3′), 3.44 (2H, m, H-4′, H-5′), 3.89 (1H, m, H-6′a), 3.72 (1H, dd, J = 12.1, 3.8 Hz, H-6′b). It was positioned on C-ζ of Tyr-6 thanks to the HMBC correlation from the anomeric proton H-1′ to C-ζ at δ C 145.4, from the aromatic protons H-δ2 and H-δ1 of Tyr-6 at δ H 4.72 (1H, d, J = 2.2 Hz) and 6.67 (1H, dd, J = 8.5, 2.2 Hz), respectively to C- ζ and from the NOESY correlation between H-ξ1 of Tyr-6 at δ H 7.10 (1H, d, J = 8.5 Hz) and H-1′ of the glucose. The HRMS data (m/z 935.4036 [M + H] +) confirmed the presence of a glucose moiety, and thus compound 18 was identified as the new cyclopeptide 5β-hydroxy-RA-XII. The nomenclature of this compound follows the established convention for RA-series cyclopeptides originally reported from Rubia species, where structural variants are defined relative to known scaffolds such as RA-XII, with substituent modifications (e.g., hydroxylation at C-β of fifth amino acid) indicated by appropriate prefixes (Morita et al., 1992).

The HRMS of compound 19 showed an HRMS ion at m/z 789.3457 [M + H] + corresponding to the molecular formula of C 40 H 48 N 6 O 11 and indicating that 19 has an additional oxygen compared to bouvardin (20). This was confirmed by the lack of methyl doublet signal of Ala-2 in the NMR spectra of 19 and the presence of an additional oxymethylene at δ H 3.49 (1H, dd, J = 10.5, 6.4 Hz, H-β1), 3.67 (1H, m, H-β2), and δ C 60.6. Compound 19 was thus a new cyclopeptide named 5β-hydroxy-RA-I.

Compounds 24 and 25 were obtained as pale-yellow amorphous powders. Their 1 H and 13 C NMR spectra were consistent with those of the β-carboline alkaloids previously described for another specimen of G. thalictroides and again re-isolated in the present study (Figueiredo et al., 2011; Figueiredo et al., 2014). However, a careful re-examination of the spectroscopic data, combined with advanced analytical approaches, led to a structural revision of these two alkaloids, representing one of the most significant findings of this work.

In the earlier reports, these compounds were assigned as 9-methoxy-β-carboline derivatives bearing, at C-3, a cyclopentane moiety substituted on one side by a 2-(2-hydroxypropyl) chain and, on the other side, by a tertiary alcohol linked to either a hydroxymethyl group (24) or a methyl group (25) (Figueiredo et al., 2011; Figueiredo et al., 2014). The present study provides a revised structural interpretation that refines the original structural proposals and expands our understanding of this class of β-carboline alkaloids in G. thalictroides (Figure 1).

The ions observed in the HRMS spectra for these reported structures were previously assigned to the [M + H-H 2 O] + adducts. It has now been proposed that, to instead match the [M + H] + adducts, an additional cyclisation between the two tertiary alcohols at C-18 and C-20 occurs, forming an ether and an extra five-membered ring, as illustrated in Figure 1. The chemical structures are in agreement with all 1D and 2D NMR data and the HRMS protonated ions at m/z 353.1866 [M + H] + and 337.1915 [M + H] + corresponding to the molecular formulas C 21 H 25 N 2 O 3 and C 21 H 25 N 2 O 2, for compounds 24 and 25, respectively (Supplementary Figures S2, S3, Supporting Information). A comparative analysis of the 13 C NMR chemical shifts for carbons C-18 and C-20 with structurally related compounds reported in the literature supports the presence of a C18–O–C20 ether linkage rather than two independent hydroxyl substituents at C18 and C20. Specifically, in hypermonones C and F (Zeng et al., 2021), carbon C-15 (substituted with two methyl groups and one hydroxyl group) resonates at δ C 69.6. In contrast, carbons C-21 and C-26, which are linked via an ether bond, appear at δ C 78.9 and 85.0, respectively. In the current study, carbons C-18 and C-20 resonate at δ C 90.5 and 78.5 in compound 24, and at δ C 86.5 and 77.5 in compound 25, respectively, consistent with the presence of an ether bridge between these two carbons. In addition, the consistency between these structures and their respective NMR data were evaluated using the CT web application (https://ctb.nmrsolutions.fi/), which performs a complete Quantum Mechanical Spectral Analysis (QMSA). All relevant spectral parameters like chemical shifts and coupling constants were thus generated and reported in the “description of isolated compounds” section.

The relative stereochemistry of the aliphatic bicyclic moieties of 24 and 25 was found to be as shown in Figure 1, based on relevant NOE interactions observed in both compounds; i.e., from H-14 to H-15, H 3 -21 and H 3 -19 (or H 2 -19), from H 3 -22 to H-16a and H-17b (Figure 2). The absolute stereochemistry configuration of 24 was established using Electronic Circular Dichroism (ECD). The experimental ECD trace was compared to the traces of the two possible absolute stereoisomers (according to the relative stereochemistry established before), calculated using time-dependent density functional theory (TD-DFT) (Grauso et al., 2019; Mandi and Kurtan, 2019). The experimental trace matched that of the isomer 14 R,15 R,18 S (Figure 3). Compound 24, after corrected assignment of the aliphatic moieties, was identified as (14 R,15 R,18 S)-1-(hydroxymethyl)-3-(2-hydroxypropan-2-yl)-2-(5-methoxy-9H-b-carbolin-1-yl)-cyclopentanol (24).

Compound 23, obtained as a pale-yellow amorphous powder, showed a [M + H] + at m/z 351.1692 corresponding to C 21 H 23 N 2 O 3 (Δppm = −3.13) (Supplementary Figure S4, Supporting Information). Its NMR spectra closely resembled those of β-carboline alkaloids 24 and 25, indicating that 23 also contains a disubstituted β-carboline moiety bearing a methoxyl group at C-9. Compared to 25, an additional signal at δ H 6.00 (1H, d, J = 7.7 Hz, H-19) and δ C 74.6 (CH-19) was observed, whereas methyl CH 3 -19 was missing. The HMBC correlations from this methine H-19 to the tertiary CH-14 at δ C 46.9, to the quaternary carbon C-18 at δ C 88.1 and to the aromatic sp 2 carbon C-2 at δ C 131.7 and the NOE correlations from H-19 to the aromatic proton H-12 at δ H 7.32 (1H, d, J = 8.2 Hz) and to methylene protons H 2 -17 at δ H 0.92 (1H, ddd, J = −13.2, 12.0, 2.7 Hz, H-17b) and 1.49 (1H, dddd, J = −13.2, 8.8, 6.3, 1.6 Hz, H-17a), allowed to link H-19 to N-1 of the β-carboline, and to C-18 of the pentane moiety. To fit the molecular formula and the 13 C NMR chemical shift (δ C 74.6), a hydroxyl group was positioned at C-19. The NOE correlations observed from the methine H-14 at δ H 3.95 (1H, ddd, J = 1.9, 1.6, 1.5 Hz) to the methyl H 3 -21 at δ H 1.37 (3H, s) and the methine H-15 at δ H 2.80 (1H, ddd, J = 2.9, 1.5, 0.9 Hz), from the methyl H 3 -22 at δ H 1.35 (3H, s) to H-16a at δ H 1.82 (1H, dddd, J = −12.9, 8.8, 2.7, 1.9 Hz) and H-17a at δ H 1.49 (1H, dddd, J = −13.2, 8.8, 6.3, 1.6 Hz), and from H-19 to H 2 -17 (Figure 4) enabled the relative configuration of compound 23 to be determined. 1 H iterative full-spin analysis (HiFSA) was performed using the CT web application (https://ctb.nmrsolutions.fi/) to calculate the chemical shifts and coupling constants of this compound.

To establish the absolute configuration of compound 23, its experimental ECD spectrum was compared with the TD-DFT-calculated spectra of the two possible absolute stereoisomers, consistent with the previously determined relative stereochemistry (Grauso et al., 2019; Mandi and Kurtan, 2019). The experimental trace matched that of isomer 14 S,15 S,18 R,19 R (Figure 3). The structure of the β-carboline alkaloid 23, which is being reported for the first time, was established as 14 S,15 S,18 R,19 R -12-methoxy-14,14-dimethyl-3b,4,5,6-tetrahydro-7H-6a,4-(epoxymethano)cyclopenta [c]indolo[3,2,1-ij][1,5]naphthyridin-7-ol (Figure 1).

Biological activity

This work enabled the identification of both previously reported compounds and novel derivatives that had not yet been experimentally characterized. The biological effects of these compounds were subsequently evaluated on the proliferation of the triple-negative breast cancer cell line BT-20 (Figure 5; Table 1). The results convincingly demonstrate that the anticancer activity of this plant is likely mediated by a complex mixture of constituents, particularly members of the bouvardin family, which exhibit cytotoxic effects with varying degrees of potency.

Among these, bouvardin (20) and 5β-hydroxy-RA-III (21) displayed the most pronounced activity, with IC 50 values in the range of 0.25–2.70 μM against cancer cells, consistent with previously reported data (Jolad et al., 2002). In addition, several compounds exhibited moderate tumor-suppressive effects, notably anthraquinones (IC 50 23–110 µM) (17 and 26) and β-carboline derivatives (IC 50 65–193 µM) (24 and 25).

Mechanistic insights

The mechanisms of action of bouvardins as inhibitors of protein biosynthesis (Zalacaín et al., 1982), as well as those of anthraquinones, which act as sources of reactive oxygen species (ROS) and serve as important scaffolds in several FDA-approved anticancer drugs, have been extensively studied (Malik et al., 2021). Building on this established knowledge, the activity profile observed in the present study provides further insight into structure–activity relationships within the bouvardin series.

Specifically, the reduced potency of compounds 19 and 21 (14 and 2.7 µM, respectively), compared to 20 and 22 (0.25 and 0.27 µM, respectively), suggests that the presence of a hydroxyl group in the Ala-2 position induces a negative effect on biological activity. However, a more detailed mechanistic interpretation remains limited by the absence of structural data on the bouvardin–target complex.

In contrast, the low activity of glucoside 18 (10 µM), which lacks the hydroxyl group in the Ala-2 position, is more plausibly attributed to reduced cellular uptake and limited target engagement, likely due to the presence of a highly polar sugar moiety. Furthermore, the relatively small difference in potency observed between compounds 19 and 21 or between 20 and 22 suggests that the macrocyclic system tolerates modifications at the ζ position of Tyr-6, indicating that this site may be less critical for maintaining biological activity.

Beyond the bouvardin series, several additional compounds exhibited weaker tumor-suppressive activity, most notably anthraquinones (17 and 26) and β-carboline derivatives (24 and 25). In contrast to bouvardins, the mechanism underlying the activity of the β-carboline derivatives remained unclear. Nevertheless, compounds belonging to this scaffold have previously been reported as potent inhibitors of Wnt signaling (Li et al., 2015), prompting an evaluation of whether the activity observed for these derivatives could be associated with modulation of this key oncogenic pathway (Shaw et al., 2019). Accordingly, Wnt pathway inhibition was assessed using the TopFlash assay (Pellissier et al., 2021), which measures Wnt3a-induced pathway activation in BT-20 triple-negative breast cancer cells. The results indicate that compound 24 acts as a selective inhibitor of Wnt signaling, with an IC 50 of approximately 40 µM (∼14 μg/mL) (Figure 5).

In contrast, the closely related β-carboline analog 25 did not exhibit specific inhibitory activity toward the Wnt pathway. Compared to compound 24, this analog lacks a hydroxyl group at C-19 position, which may be critical for activity, potentially affecting its ability to interact with key molecular targets involved in pathway modulation. Moreover, both compounds displayed substantially lower potency, approximately 10–20-fold reduced compared to previously reported Wnt-targeting β-carbolines (Li et al., 2015). These findings suggest that the derivatives identified in this study do not represent optimal scaffolds for effective Wnt pathway inhibition. A similar rationale may explain the lack of activity observed for compound 23. In this case, the indole NH group is likely involved in intramolecular cyclization, rendering it unavailable for interaction with biological targets, whereas this functionality remains accessible in active analogs such as compound 24.

Taken together, these findings suggest that the chloroform fraction may function as a multi-component anticancer system rather than as a source of a single active compound. The coexistence of broadly cytotoxic constituents and pathway-selective Wnt inhibitors points to a coordinated pharmacological profile arising from the combined presence of multiple bioactive molecules. Such a composition is likely to affect several aspects of cancer biology simultaneously, including key Wnt-driven processes such as tumor progression, therapy resistance, disease recurrence, cancer stem cell (CSC) maintenance, epithelial–mesenchymal transition (EMT), and apoptosis. The concurrent modulation of these mechanisms may enhance the overall biological effect, as inhibition of the Wnt pathway could increase the susceptibility of cancer cells to cytotoxic stress.

Although this hypothesis requires further experimental validation, the data indicate that interactions among the different constituents may contribute to enhanced biological activity beyond that expected from individual components alone. Multi-component phytotherapeutic systems are known to exert their effects through complex interactions among their constituents, leading to multi-target modulation (Efferth and Koch, 2011). Overall, the chemical composition of G. thalictroides root extract provides a compelling basis for further investigation into naturally derived, multi-target approaches aimed at addressing tumor heterogeneity and therapeutic resistance.

Metabolite profiling

To obtain an overview of the chemical composition of G. thalictroides, roots were extracted with ethanol and also prepared as a traditional water decoction. The ethanolic extract was subsequently fractionated using solvents of increasing polarity to yield hexane, chloroform, and ethyl acetate fractions. The samples included in the UHPLC-HRMS/MS-based metabolomic analysis were the crude ethanolic extract, its corresponding fractions (hexane, chloroform, and ethyl acetate), and the traditional decoction. All of these samples were analyzed to detect compounds based on their mass and fragmentation patterns, and the resulting data were used to construct a molecular networking (MN) (Supplementary Figure S1, Supporting Information), where nodes represent compounds and edges indicate structural similarity. The MN comprised 2066 nodes (1,298 connected components and 768 singletons) linked by 3,075 edges. Through this process, 284 compounds were tentatively annotated (Allard et al., 2016) using ISDB-DNP databases and a taxonomical filter prioritizing metabolites reported in related plant families (Rutz et al., 2019), leading to 13 compounds previously identified in Rubiaceae (Supplementary Table S1, Supporting Information). Several clusters were identified: C1 and C2 correspond to indole alkaloids typical of Rubiaceae (Rumbero and Vazquez, 1991; Martins and Nunez, 2015; Boucherle et al., 2016), while C3 contains cyclopeptides previously reported in this species (Xu et al., 2013; Figueiredo et al., 2016).

Within the framework of molecular networking-based annotation, the 26 isolated compounds validated the network assignments and enabled the precise attribution of nodes to specific structures. In particular, Rubiaceae-type cyclopeptides (18–22) were mapped to cluster C3, while three β-carboline alkaloids (23–25) were assigned to cluster C1 (Supplementary Figure S1, Supporting Information). Interestingly, although β-carboline alkaloids and Rubiaceae-type cyclopeptides have previously been reported as cytotoxic (Figueiredo et al., 2011; Figueiredo et al., 2014; Figueiredo et al., 2016), they appear to be absent or present only in trace amounts in the traditional decoction (Supplementary Figure S1, Supporting Information).

This study provides a comprehensive phytochemical and biological characterization of G. thalictroides. A total of 26 compounds were isolated and structurally elucidated, including the discovery of two new cyclopeptides (18, 19) and one new β-carboline alkaloid (23), representing significant additions to the chemical space of this species. In addition, the structural revision of two β-carboline alkaloids (24, 25) constitutes a major advancement, refining the current understanding of this class of natural products.

The integration of UHPLC–HRMS/MS-based metabolite profiling and molecular networking enabled the annotation of 13 additional metabolites, demonstrating the effectiveness of metabolomic approaches in expanding chemical coverage and supporting targeted isolation strategies.

The identification of potent cytotoxic compounds, particularly bouvardins, together with the detection of pathway-specific activity such as selective Wnt signaling inhibition by β-carboline derivative 24 highlights the presence of multiple and complementary mechanisms of action.

Collectively, these findings suggest that G. thalictroides constitutes a rich and underexplored source of bioactive metabolites with anticancer properties. These results provide a solid foundation for future pharmacological investigations and support the progression toward in vivo studies and drug development efforts.