A computer-assisted method for the generation of pseudoreceptor models is presented together with two practical applications. From a three-dimensional alignment of known histamine H(4) receptor ligands, a pseudoreceptor model of the putative ligand binding site was constructed and used for virtual screening of a large collection of commercially available compounds. Two bioactive chemotypes were retrieved, demonstrating the general applicability of the approach. The pseudoreceptor model was also used to find the putative ligand binding pocket within the transmembrane receptor domain. For each frame of a molecular dynamics simulation of a homology-based H(4) receptor model, we automatically extracted potential ligand binding pockets and used their compatibility with the pseudoreceptor as a selection criterion. The best-matching pocket fits perfectly with existing mutation data and previously published hypotheses suggesting Glu182(5.46) as the preferred binding partner of a positively charged moiety of H(4) receptor ligands. This new pseudoreceptor approach has demonstrated its suitability for both structure-based prioritization of protein receptor models, and ligand-based virtual screening with the aim to perform scaffold hopping.

No abstract

Kinks have been observed to provide important functional and structural features for membrane proteins. Despite their ubiquity in membrane proteins, and their perceived importance, no protein modeling methods explicitly considers kinks. In spite of the limited data for transmembrane proteins, we were able to develop a knowledge-based modeling method for introducing kinks, which we demonstrate can be exploited in modeling approaches to improve the quality of models. The work entailed a thorough analysis of the available high resolution membrane protein structures, concomitantly demonstrating the complexity of the structural considerations for kink prediction. Furthermore, our results indicate that there are systematic and significant differences in the sequence as well as the structural environment between kinked and nonkinked transmembrane helices. To the best of our knowledge, we are reporting a method for modeling kinks for the first time.

Hypothalamic functions, including feeding behavior, show a high degree of plasticity throughout life. Doublecortin (DCX) is a marker of plasticity and neuronal migration expressed in the hypothalamus. Therefore, we wanted to map the fate of DCX(+) cells in the arcuate nucleus (ARC) of the hypothalamus. For this purpose, we generated a BAC transgenic mouse line that expresses the inducible recombinase CreER(T2) under control of the DCX locus. Crossing this line with the Rosa26 or Ai14 reporter mouse lines, we found reporter(+) cells in the ARC upon tamoxifen treatment. They were born prenatally and expressed both DCX and the plasticity marker TUC-4. Immediately after labeling, reporter(+) cells had an enlarged soma that normalized over time, suggesting morphological remodeling. Reporter(+) cells expressed β-endorphin and BSX, neuronal markers of the feeding circuit. Furthermore, leptin treatment led to phosphorylation of STAT3 in reporter(+) cells in accordance with the concept that they are part of the feeding circuits. Indeed, we found a negative correlation between the number of reporter(+) cells and body weight and epididymal fat pads. Our data suggest that DCX(+) cells in the ARC represent a cellular correlate of plasticity that is involved in controlling energy balance in adult mice.