Quantifying human sweetness perception via a TAS1R2/TAS1R3-based model-predicted sweetness score

Artificial sweeteners provide high-intensity sweetness with minimal caloric load, but their evaluation still relies heavily on subjective sensory panels. Psychophysical studies have reported parameters such as sweetness potency, perception threshold, and saturation threshold as key benchmarks. To develop an objective alternative, we developed a TAS1R2/TAS1R3-based calcium imaging system using HEK293T cells to quantify receptor response to five sweeteners. EC₅₀ (potency) and maximal response (efficacy), were obtained from dose-response curves. Inverse regression between EC₅₀ and reported relative sweetness (RS) showed strong correlation (R² = 0.9688, RS ≈ 1.38/EC₅₀), forming the basis for a Model-Predicted Sweetness Scores (MPSS; MPSS = (1.38/EC₅₀) × 10²). MPSS values closely matched known RS and reflected reported detection thresholds. Maximal responses were poorly correlated with RS. These findings suggest that EC₅₀-based modeling offers a receptor-level, perceptually grounded method for sweetness prediction and may serve as a practical alternative to sensory evaluation in sweetener research.