In order to confirm whether this “radar” T1R2/T1R3 can “sense” the sweetness of different sweetener, we simulated the model of receptor's structure[1] [2]. It was helpful to observe how the sweeteners binding to T1R2/T1R3 receptor visually. Moreover, we expected to find some unknown sweeteners binding sites based on this model, even some ideal sweeteners are still unknown.


To make the signal input more accurate and reliable, we simulated the T1R2/T1R3 receptor's structure model through SWISS-MODEL. Meanwhile, according to Chemdraw 2D and Chemdraw 3D, we constructed some sweetners' model. And the docking process was performed by using Autodock Vina.


We used homology modeling to obtain the structure of human sweet receptor T1R2/T1R3. According to the crystal protein structure of similar receptor in mice, we only simulated ligand-binding-domain of our receptor.

Fig. 3 The simulated structure of human sweetness receptors' ligand-binding domain (LBD)

Then, taking advantage of software Chemdraw 2D and Chemdraw 3D, we established some natural or artificial sweeteners' structure.

Fig. 4 Three dimensional structure of some sweeteners

Docking process was carried out under Autodock Vina (Fig. 5-7).

Fig. 5. Docking result of different sweeteners Fig. 6. Docking result of aspartame Fig. 7. Docking result of stevioside


Our docking results estimated that ligand-binding domain (LBD) of human T1R2/T1R3 receptors could bind with almost all kinds of known sweeteners. It provided evidence that this receptor has an ability to “taste” the sweetness. Meanwhile, our results also helped us to speculate that the number of both binding-site and binding-molecule also might influence the sense of sweetness.

However, our model still needs optimization. We plan to enhance the docking model and the more professional direct would be supplied by it.