A Novel Quantitative Prediction Approach for Astringency Level of Herbs Based on an Electronic Tongue

BACKGROUND: The current astringency evaluation for herbs has become dissatisfied with the requirement of pharmaceutical process. It needed a new method to accurately assess astringency. METHODS: First, quinine, sucrose, citric acid, sodium chloride, monosodium glutamate, and tannic acid (TA) were analyzed by electronic tongue (e-tongue) to determine the approximate region of astringency in partial least square (PLS) map. Second, different concentrations of TA were detected to define the standard curve of astringency. Meanwhile, coordinate-concentration relationship could be obtained by fitting the PLS abscissa of standard curve and corresponding concentration. Third, Chebulae Fructus (CF), Yuganzi throat tablets (YGZTT), and Sanlejiang oral liquid (SLJOL) were tested to define the region in PLS map. Finally, the astringent intensities of samples were calculated combining with the standard coordinate-concentration relationship and expressed by concentrations of TA. Then, Euclidean distance (E(d)) analysis and human sensory test were processed to verify the results. RESULTS: The fitting equation between concentration and abscissa of TA was Y = 0.00498 × e((−X/0.51035)) + 0.10905 (r = 0.999). The astringency of 1, 0.1 mg/mL CF was predicted at 0.28, 0.12 mg/mL TA; 2, 0.2 mg/mL YGZTTs was predicted at 0.18, 0.11 mg/mL TA; 0.002, 0.0002 mg/mL SLJOL was predicted at 0.15, 0.10 mg/mL TA. The validation results showed that the predicted astringency of e-tongue was basically consistent to human sensory and was more accuracy than E(d) analysis. CONCLUSION: The study indicated the established method was objective and feasible. It provided a new quantitative method for astringency of herbs. SUMMARY: The astringency of Chebulae Fructus, Yuganzi throat tablets, and Sanlejiang oral liquid was predicted by electronic tongue. Euclidean distance analysis and human sensory test verified the results. A new strategy which was objective, simple, and sensitive to compare astringent intensity of herbs and preparations was provided. Abbreviations used: CF: Chebulae Fructus; E-tongue: Electronic tongue; Ed: Euclidean distance; PLS: Partial least square; PCA: Principal component analysis; SLJOL: Sanlejiang oral liquid; TA: Tannic acid; VAS: Visual analog scale; YGZTT: Yuganzi throat tablets.