Chemical Profiling and UPLC-qTOF-MS/MS-Based Metabolomics of Three Different Parts of Edgeworthia chrysantha and Identification of Glucose Uptake-Enhancing Compounds

Abstract

Background/objectives: Edgeworthia chrysantha is rich in coumarin and flavonoid dimers, which may exhibit diverse pharmacological activities. However, to date, no metabolomics studies have been conducted and its bioactive constituents related to glucose metabolism remain uncharacterized. This study aimed to conduct a comprehensive chemical analysis combined with bioactivity assays to evaluate its efficacy in promoting glucose uptake.

Methods: Chemical profiling of three parts (leaf, stem, and root) of E. chrysantha was performed using UPLC-Q-TOF-MS/MS spectrometry, followed by metabolomics analysis. Based on the chemical profiles and glucose uptake activity, compounds were isolated from the root. Their structures were elucidated using spectroscopic techniques, including UV, NMR, and mass spectrometry. The glucose uptake activity of the isolated compounds was assessed using a 2-NBDG assay.

Results: Metabolic analysis revealed distinct chemical compositions among the plant parts. Dimeric coumarins and biflavonoids were abundant in the root, whereas flavonoid monomers were predominant in the leaf. Bioactivity-guided isolation yielded nine compounds (1-9), among which compound 1, a newly identified coumarin glycoside, exhibited significant glucose uptake-enhancing activity. Molecular docking analysis further suggested that compound 1 activates AMPK through an allosteric site, thereby promoting glucose uptake.

Conclusions: These findings provide a comprehensive chemical and metabolomic characterization of E. chrysantha and highlight its potential as a functional food ingredient for glucose-lowering effects.

Keywords: Edgeworthia chrysantha; Thymelaeaceae; anti-diabetic; diabetes; glucose uptake; metabolomics.

Figure 1

Nine secondary metabolites (1–9) were isolated and identified from the roots of Edgeworthia chrysantha, including various coumarin glycosides, aglycones, and biflavonoid derivatives.

Figure 2

(A) Key HMBC correlations of the new compound 1, highlighting the long-range proton–carbon couplings that were critical for establishing the connectivity between the coumarin skeleton, the glucose moiety, and the HMG side chain; (B) ECD spectrum of compound 4, displaying characteristic cotton effects that facilitated the determination of its axial chirality and absolute configuration. (C) ¹H NMR and (D) ¹³C NMR (D) spectra of new compound 1.

Figure 3

(A) PCA loading plot illustrating the metabolic distribution across three distinct plant parts (leaf, stem, and root) of Edgeworthia chrysantha; (B) PLS-DA scatter plot showing clear group separation based on metabolic profiles; (C) Heatmap visualization of normalized and clustered metabolite intensities across leaf, stem, and root samples, providing an overview of the relative abundance and grouping of structurally related compounds; (D) Volcano plot representing differential metabolite expression between leaf and stem tissues; (E) Volcano plot displaying differential metabolites between root and stem samples.

Figure 4

(A) Mass fragmentation patterns and their corresponding chemical structures; (B) MS1 and MS2 spectra of edgeworic acid, illustrating the precursor ion and key product ions resulting from characteristic bond cleavages.; (C) Cluster visualization from feature-based molecular networking (FBMN) analysis. Each differently colored box represents a distinct class of derivatives. Each node represents a detected ion, with edges indicating spectral similarity based on MS/MS fragmentation patterns. Structurally related metabolites, such as dimeric coumarins and flavonoid glycosides, formed distinct clusters, facilitating compound annotation and classification.

Figure 5

Evaluation of glucose uptake in 3T3-L1 adipocytes using the fluorescent glucose analog 2-NBDG, with insulin (100 nM) as positive control and isolated compounds 1–9 (each 20 µM) as test samples. Fluorescence intensity was measured at excitation/emission wavelengths of 450/535 nm. Data are presented as mean ± SD. * p < 0.05, ** p < 0.01 versus control.

Figure 6

(A) Three-dimensional visualization of compound 1 bound to the allosteric site of AMPK, located at the interface between the α-subunit kinase domain (green) and the β-subunit CBM (blue). (B) Overlay of AMPK structure with compound 1 (green), highlighting its orientation in the allosteric pocket. (C) Molecular conformation of compound 1, showing its coumarin dimer core, β-D-glucose, and HMG group positioned for optimal interactions. (D) Three-dimensional interaction diagram depicting key hydrogen bonds and hydrophobic contacts between compound 1 and surrounding AMPK residues.