Biotransformation by beta glucosidase enhances anti inflammatory metabolites in licorice using untargeted metabolomics

Abstract

Licorice (Glycyrrhiza uralensis) has traditionally been used as a food-derived herbal remedy for inflammation; however, the anti-inflammatory potential of its fermented extract in skin health is still unclear. This study investigated fermented licorice extract (FLE) for its effects against glyoxal-derived advanced glycation end products (GO-AGEs) and ultraviolet B (UVB)-induced skin inflammation in HaCaT keratinocytes. At 10 µg/mL, FLE reduced IL-6 levels by 46% and TNF-α levels by 52%, and significantly lowered PGE₂ levels. Mechanistic evaluation showed that FLE suppressed inflammatory signaling pathways, particularly nuclear factor-κB (NF-κB) and mitogen-activated protein kinases (MAPKs). Untargeted metabolomics identified fermentation-enhanced bioactive metabolites, including glycyrrhetic acid-3-O-glucuronide, 18β-glycyrrhetic acid, 24-hydroxyglycyrrhetic acid, and isoliquiritigenin, which correlated with anti-inflammatory activity. Notably, 18β-glycyrrhetic acid and isoliquiritigenin exhibited potent antiglycation effects and cytokine suppression. These results suggest that fermentation enhances the bioactive profile of licorice, supporting its potential as a functional ingredient for managing skin inflammation from GO-AGEs and UVB exposure.

Fig. 1. Antiglycation and anti-inflammatory effects of ULE and FLE.

A GO-affinity assay. B GO-AGEs-breaker assay. C GO-AGEs formation assay. D Cell viability in HaCaT keratinocytes. E Cell viability in GO-AGEs-treated and UVB-irradiated HaCaT keratinocytes. F–I Inflammatory cytokine production in GO-AGE- and UVB-treated HaCaT keratinocytes, including (F) IL-1β, (G) IL-6, (H) TNF-α, and (I) PGE₂. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. ULE-treated group; ^$p < 0.05, ^$$p 0.01, and ^$$$p < 0.001 vs. GO, GO-AGE-^, or GO-AGEs/UVB-treated group; ^###p < 0.001 vs. Control group (untreated cells). Results are expressed as the mean ± SEM (n = 3).

Fig. 2. Expression levels of RAGE and related inflammation signals in GO-AGE- and UVB-treated HaCaT human keratinocytes either treated with ULE or FLE.

A RAGE; (B) COX-2; (C) IL-1β; (D) MAPKs pathway proteins (p38, ERK, and JNK); (E) NF-κB and IκB-α protein levels. ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 vs. Control (untreated) group; ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001 vs. GO-AGEs/UVB-treated group. Results are expressed as the mean ± SEM (n = 3).

Fig. 3. Multivariate statistical analysis of differential metabolites between ULE and FLE.

A PCA score plot showing overall metabolic variation. B S-plot from OPLS-DA illustrating key differential metabolites contributing to the separation between ULE and FLE. C OPLS-DA score plot demonstrating group discrimination. D Cross-validation plot validating the OPLS-DA model robustness. E Heatmap of hierarchical clustering analysis of differential metabolites in ULE and FLE. Metabolite names are labeled in red to indicate increased levels and in black to indicate decreased levels after fermentation of licorice by L. mesenteroides.

Fig. 4. Pearson correlation matrix between differential metabolites and antiglycation and anti-inflammatory activities.

The color scale ranges from −1 (blue, strong negative correlation) to +1 (red, strong positive correlation). Circle sizes represent the absolute correlation strengths based on each correlation coefficient. Differential metabolite names are written in red (if increased) and black (if decreased) based on their relative levels following fermentation of licorice by L. mesenteroides.

Fig. 5. Proposed bioconversion pathways of glycosidic triterpenoids and flavonoids in licorice following fermentation by L. mesenteroides.

A Triterpenoid derivatives: Bioconversion of glycyrrhizin and licorice saponin G2 into aglycone and hydrolyzed forms via β-glucosidase activity. B Flavonoid derivatives: Conversion of isoliquiritin apioside into isoliquiritin and isoliquiritigenin through sequential deglycosylation. Selected compounds among the differential metabolites were compared based on their abundance in ULE and FLE. The y-axis of each graph represents the peak areas (log¹⁰) of each metabolite in ULE and FLE. Differential metabolite names are displayed in red (if increased) and black (if decreased) according to their relative levels during the fermentation of licorice by L. mesenteroides. *p < 0.05, **p < 0.01, and ***p < 0.001 vs. ULE. Results are expressed as the mean ± SEM (n = 3).

Fig. 6. Antiglycation and anti-inflammatory effects of 18β-glycyrrhetic acid.

A GO-affinity assay; (B) GO-AGEs breaker assay; (C) GO-AGEs formation assay; (D) Cell viability in GO-AGEs/UVB-treated HaCaT human keratinocytes; (E–H) Levels of proinflammatory markers including (E) IL-1β, (F) IL-6, (G) TNF-α, and (H) PGE₂ in GO-AGEs/UVB-treated HaCaT cells. ^###p < 0.001 vs. Control (untreated) group; ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001 vs. GO, GO-AGE-, or GO-AGEs/UVB-treated group. Results are expressed as the mean ± SEM (n = 3).

Fig. 7. Antiglycation and anti-inflammatory effects of isoliquiritigenin.

A GO-affinity assay; (B) GO-AGEs breaker assay; (C) GO-AGEs formation assay; (D) Cell viability in GO-AGEs/UVB-treated HaCaT human keratinocytes; (E–H) Levels of proinflammatory markers including (E) IL-1β, (F) IL-6, (G) TNF-α, and (H) PGE₂ in GO-AGEs/UVB-treated HaCaT cells. ^###p < 0.001 vs. Control (untreated) group; ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001 vs. GO, GO-AGE-, or GO-AGEs/UVB-treated group. Results are expressed as the mean ± SEM (n = 3).