Integrated chemometric and pharmacological evaluation of processing-induced attenuation of hepatotoxicity in...

Polygonum multiflorum (P. multiflorum) is the dried root tuber of P. multiflorum Thunb., a perennial herbaceous plant belonging to the Polygonaceae family. Traditionally, it is believed to possess effects such as nourishing blood, replenishing yin, and moistening the intestines to relieve constipation. Research indicates that the processing procedure significantly alters the chemical composition, pharmacological activity, and hepatotoxicity of substances; however, the underlying mechanisms responsible for attenuating toxicity and enhancing efficacy remain to be clarified. Traditional experience suggests that with an increase in the number of processing cycles (e.g., nine-time processing), their properties shift from purgative to tonic, yet a systematic comparison of the specific differences among samples subjected to varying processing cycles is still lacking.

Based on the central hypothesis that processing alters the chemical composition profile of P. multiflorum and thereby synergistically regulates its toxicity and efficacy, this study established the fingerprint profiles of samples subjected to different processing cycles (0, 3, 6, and 9 times). It systematically compared the hepatotoxicity and efficacy differences between the raw product and its processed products, and focused on identifying the key bioactive components and potential mechanisms responsible for the distinct toxicity and efficacy between the raw product and the nine-steamed-nine-dried product (9x-P. multiflorum) using spectrum-effect relationship analysis.

High-performance liquid chromatography (HPLC) was first employed to establish the chemical fingerprints of the raw P. multiflorum and its processed products after 3, 6, and 9 cycles of steaming with black bean juice (three-steamed-three-dried product, 3x-P. multiflorum; six-steamed-six-dried product, 6x-P. multiflorum; 9x-P. multiflorum), to characterize the dynamic changes in chemical constituents during processing. Zebrafish models of liver injury, intestinal peristalsis, anemia, and immunosuppression were used to systematically compare the hepatotoxicity and efficacy of samples with different processing cycles, with a focus on the toxicity-efficacy shift between raw P. multiflorum (0 cycles) and 9x-P. multiflorum. Hematoxylin and eosin (H&E) staining was performed to observe histopathological changes and evaluate liver injury. Each group contained 10 zebrafish larvae, and data were statistically analyzed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test. Furthermore, a spectrum-effect relationship model was constructed using grey relational analysis (GRA) combined with the entropy weight method (EWM), to correlate common chemical constituents with hepatotoxicity and efficacy indicators. Finally, network pharmacology and molecular docking were integrated to systematically predict and verify the targets and molecular mechanisms of the screened potential active components, especially those that differed significantly between raw P. multiflorum and 9x-P. multiflorum.

Compared with the control group, raw P. multiflorum significantly decreased the liver area-to-lateral body area ratio (LA/BA) by 29% and increased hepatocyte apoptosis by 72.5%, indicating marked hepatotoxicity. H&E staining revealed hepatocyte swelling, vacuolar degeneration, and focal necrosis. In contrast, 9x-P. multiflorum caused no significant increase in apoptosis and increased the LA/BA by 15.5%. Raw P. multiflorum exerted a strong pro-peristaltic effect, reducing the mean gastrointestinal fluorescence intensity (GI FI) by 31.2%, whereas 9x-P. multiflorum decreased it by only 5.7%. For tonic effects, 9x-P. multiflorum increased red blood cell staining intensity by 203% and macrophage count by 82.5%, which was significantly superior to raw P. multiflorum. Spectrum-effect analysis showed that Peak 5 (emodin-8-O-glucoside, EmG), Peak 7 (emodin, Emo), Peak 6 (physcion-8-O-β-D-glucoside, PhG), and Peak 8 (physcion, Phys) were closely correlated with hepatotoxicity and purgative activity. Peak 4 (2,3,5,4'-tetrahydroxystilbene-2-O-β-D-glucoside, TSG) and Peak 7 (Emo) exhibited stronger correlations with tonic effects. Activity validation confirmed that Emo, Phys, and their glucosides showed varying degrees of hepatotoxicity; Emo and Phys displayed significant purgative effects; while TSG and Emo exerted obvious blood-tonifying activity, verifying the reliability of the spectrum-effect analysis. Network pharmacology identified 28 common targets associated with P. multiflorum-induced liver injury, and protein-protein interaction (PPI) network analysis recognized Estrogen Receptor 1 (ESR1) as a hub target. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment revealed significant enrichment in the PI3K-Akt, mitogen-activated protein kinase (MAPK), and estrogen signaling pathways. Molecular docking confirmed that EmG binds strongly to core targets including ESR1 and epidermal growth factor receptor (EGFR) (binding energy < -7.0 kcal/mol).

This study verified the central hypothesis: raw P. multiflorum exhibits prominent hepatotoxicity and purgative effects, whereas processing with black bean juice achieves detoxification and enhances blood-tonifying and immune-enhancing efficacy. Emo, Phys, and their glycosides are the main material basis for hepatotoxicity and purgation, while TSG and Emo mediate the tonic effects. Network pharmacology and molecular docking further revealed that conjugated anthraquinones may synergistically disrupt liver-protective pathways via multiple targets such as ESR1 and EGFR. Processing reduces conjugated anthraquinone content and rewires the regulatory targets from a toxicity network toward a tonic network. These findings provide modern scientific evidence for the traditional Chinese medicine (TCM) theory that "raw P. multiflorum purges while processed P. multiflorum tonifies", and demonstrate the value of integrating spectrum-effect-toxicity correlation with network pharmacology in the mechanistic study of Chinese herbal medicine processing.