Figure S1 HPLC assay of TCE. Two saponins were controls in TCE: 5β-furost-Δ25 (27) –en-1β,2β, 3β, 4β, 5β, 6β, 7α, 22ζ, 26-nonaol-26-O-β-D-glucopyranoside (saponin 1) and 5β-furost-Δ25(27)-en-1β, 2β, 3β, 4β, 5β, 7α, 22ζ, 26-octaol-6-one-26-O-β-D-glucopyranoside (saponin 2). HPLC analysis was applied on a Shimadzu 20A series HPLC system consisting of two pumps (LC-20A Solvent Delivery Unit), a column oven (CTO-10ASVP), a SPD detector (SPD-M20AV Photodiode Array Detector) and an LC solution Work Station. TCE was applied to himadzu VP-ODS column (5μm, 150 × 4.6 mm; Shimadzu, Japan) and detected at 203 nm. Column temperature was set up at 25°C and the flow rate was 1 ml/min. The mobile phase was acetonitrile – water two-phase gradient elution.

Figure S2 Effects of TCE on carbontetrachloride (CCl4) -induced hepatitis. Mice were given orally with TCE (75, 150, 300 mg/kg) or Bifendatatum (200 mg/kg) in PBS for five days, respectively. One hour after the final administration, Mice were intraperitoneally injected with CCl4 (0.12%, v/v, dissolved in olive oil, 10 ml/kg bodyweight) to induce acute liver injury, while the mice in the normal group were administered with equal volume of olive oil. The animals were sacrificed 24 h after CCl4 intoxication. Serum was obtained by bleeding for the measurement of alanine transaminase (ALT) and aspartate transaminase (AST) activities respectively. The data indicated the mean ± SEM of experimental animals (n = 10). One-way ANOVA revealed a significant difference at P < 0.01. *P < 0.05, **P < 0.01, vs Con A control; ##P < 0.01, vs normal (Dunnett's test).

Figure S3 Effects of TCE on picryl chloride (PCl)-induced contact hypersensitivity in mice. Mice were sensitized by painting 0.1 ml of 1% PCl in ethanol onto the shaved skin of their abdomen on day 0. Five days after sensitization, they were challenged on right ears with 30 μl 1% PCl in olive oil. At the time of challenge and 5 h later, TCE (75, 150, 300 mg/kg) were given orally twice. (A) Twenty hours after challenge, the ear swelling was evaluated. Each column represents the mean ± s.e.m of 12 mice. (B) Representative microphotographs showing ear histopathologic changes with hematoxylin–eosin staining (original magnification ×200). (C) Microscopic scores of ears (n = 12). One-way ANOVA revealed a significant difference at P < 0.01. *P < 0.05, **P < 0.01, vs Con A control; ##P < 0.01, vs normal (Dunnett's test).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.