3D QSAR of serotonin transporter ligands: CoMFA and CoMSIA studies

The main purpose of this study was the investigation of quantitative structure-activity relationships of serotonin transporter ligands with regard to the future development of potential new and selective PET radiotracers for the serotonin transporter. A heterogeneous data set of 19 selective and non-selective serotonin reuptake inhibitors was used. Affinity data for both the serotonin transporter and the norepinephrine transporter was available. As a necessary prerequisite for our 3D QSAR studies a reasonable alignment of the compounds was developed using GASP. It was based on an existing pharmacophore model. In addition to the widely used CoMFA method, the somewhat newer CoMSIA method was applied. Statistically reliable CoMFA models for both the serotonin transporter (q(2) = 0.538) and the norepinephrine transporter (q(2)=0.445) were developed, further improving the internal predictability by applying region focusing for the serotonin transporter (q(2)=0.674). These models were compared with the CoMSIA models for the serotonin and the norepinephrine transporter that yielded comparable cross-validated correlation coefficients (q(2)=0.531 and q(2)=0.502, respectively). Certain structural features that are distinctive of each transporter and important for high binding affinity were identified. Highly comparable results were obtained for CoMFA and CoMSIA. Both methods were applied to elucidate structural requirements for serotonin transporter selectivity. The resulting CoMSIA map provides important information for lead optimization with respect to selectivity enhancement.